LARGE - SCALE MAPPING AND GEOMORPHOMETRY OF UPLAND PERIGLACIAL LANDSCAPES IN EASTERN BERINGIA By Clayton W. Queen A THESIS Submitted to Michigan State University i n partial fulfillment of the requirements for the degree of Geography Mast er of Science 2018 ABSTRACT LARGE - SCALE MAPPING AND GEOMORPHOMETRY OF UPLAND PERIGLACIAL LANDSCAPES IN EASTERN BERINGIA By Clayton W. Queen In subarctic and alpine regions, an altitudinal and latitudinal zone exists between the treeline and glaciers i n which a distinct set of periglacial landforms and processes dominated . Several prominent periglacial geomorphologists have recently questioned whether periglacial processes can give rise to a charact eristic periglacial landscape . This thesis tests the hypothesis that such landscapes do indeed exist in upland , cold, nonglacial environments. T he landscape was examined through a multiscale approach using large - scale mapping and geomorphometry. Study sites were chosen from the largely unglaciated eastern Be ringia , together form ing a transect across this region. An additional site was mapped that represented a high alpine site, in which periglacial geomorphic processes are currently active. At each site, large - scale geomorphological maps of cryoplanation terr aces (CTs) were generated using traditional field techniques and computer - generated mapping based on digital elevation models (DEMs). Geomorphometric analysis was conducted at both the landform and landscape scale s for sites in eastern Beringia . This an alysis focused on the identification of CTs and the the Yukon - Tanana Upland physiographic province . Results from these analyses indicate there is a distinct periglacial landscape , composed of an assemblage of interconnected forms . C ryoplanation landforms function as the foundation of this periglacial landscape . Copyright by CLAYTON W. QUEEN 2018 iv ACKNOWLEDGEMENTS I would like to express my sincere gratitude to all who helped make this thesis possible. It i s appropriate to start by thanking all those involved with the Circumpolar Active Layer Monitoring (CALM) program for first introducing me to periglacial research and providing me with many subsequent opportunities. I would like to thank t he United States Permafrost Association for funding this research. Funding from Michigan State University Department of Geography and College of Social Science was essential in allowing me to travel to and conduct this research. Finally, I would like to extend my gratitu de to Northern Michigan University for the use of equipment . Dr. Ashton Shortridge of Michigan State University inspired me to pursue geomorphometry in this thesis. His inspiration and advice were invaluable. Dr. Shortridge s d oor was always open and he was very willing to take the time to help when difficulties arose . I am indebted to Dr. Alan Arbogast of Michigan State University , who , not only served as an invaluable mentor while conducting this research but generously provided the opportunity to con tinue to expand my education beyond periglacial problems . Dr. Arbogast could always be counted on for good advice and an honest answer. Dr. Frederick Nelson of Michigan State University first introduced me to the periglacial realm and fostered my interes t in periglacial geomorphology. As an advisor and mentor, he provided the impetus for this research and continued support throughout the project. Dr. Nelson deserves special thanks for the many trips to East Lansing, Skype calls, and edits of v papers. He we nt above and beyond to provide the opportunity to conduct this research and I could not have finished this work without his steadfast guidance and enthusiasm . I would like to express my heartfelt gratitude to Kelsey Nyland for h er mentorship and assistanc e in the field and through many aspects of this project . This work c ould not have been done without her support through many long hours of field and office work. I would like to express my thanks to my parents for their support of my education and encoura gement to obtain higher education and enriching experiences. Finally, I would like to thank my wife, Katie, for her support of my graduate education and the pursuit of this project . I could not have gotten through this without the motivation and support she provided through the long hours of studying, writing, and through the many days spent apart. Thank you! vi TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ....................... ix LIST OF FIGURES ................................ ................................ ................................ ..................... x KEY TO ABBREVIATIONS ................................ ................................ ................................ ...... x ii i Chapter 1 Introduction ................................ ................................ ................................ ......... 1 1.1 The Concept of Landscape ................................ ................................ ........................... 1 1.1.1 History of the landscape concept : ................................ ................................ ................. 2 1.1.2 Landscape ecology : ................................ ................................ ................................ ....... 3 1.1.3 Definition of landscape for this work : ................................ ................................ ........ 4 1.1.4 Scale of the landscape : ................................ ................................ ................................ .. 4 1. 2 Periglacial Geomorphology ................................ ................................ ......................... 5 1. 3 Cryoplanation Terraces ................................ ................................ ................................ . 7 1. 3 .1 Morphology : ................................ ................................ ................................ .................. 7 1. 3 .2 Hist ory : ................................ ................................ ................................ .......................... 9 1. 3 .3 Nivation : ................................ ................................ ................................ ........................ 10 1. 4 Periglacial Assemblages ................................ ................................ ............................... 11 1. 5 Study Area Beringian Uplands ................................ ................................ ................. 1 6 1. 5 .1 Glacial and environmental history : ................................ ................................ ............... 1 7 1. 5 .2 Periglacial studies in B eringia : ................................ ................................ ...................... 1 8 1. 6 Statement of Problem ................................ ................................ ................................ ... 1 9 1. 7 Hypothesis ................................ ................................ ................................ ........................ 2 1 1. 8 Methods ................................ ................................ ................................ ............................ 21 1. 8 .1 Geomorphologic analysis and periglacial assemblages : ................................ ............... 21 1. 8 .2 Periglacial geomorphometry : ................................ ................................ ........................ 2 5 1. 8 .3 Specific geomorphometry : ................................ ................................ ............................ 2 6 1. 8 .4 General geomorphometry : ................................ ................................ ............................ 2 7 Chapter 2 Mapping and Multiscale Analysis of Periglacial Assemblages .............. 2 9 2.1 Background ................................ ................................ ................................ ...................... 2 9 2.1.1 Perigla ci al form communities : ................................ ................................ ...................... 31 2.1.2 Periglacial assemblage : ................................ ................................ ................................ . 32 2.2 Geomorphological Mapping in Alaska and British Columbia ............................. 33 2.2.1 Study sites : ................................ ................................ ................................ ..................... 3 3 2.2.2 Mapping m ethods : ................................ ................................ ................................ ........ 3 6 vii 2.2.3 Mapping r esults : ................................ ................................ ................................ ............ 3 9 2.2.4 Th e facies concept and the periglacial landscape : ................................ ........................ 61 2.3 Conclusion ................................ ................................ ................................ ........................ 63 Chapter 3 Specific Periglacial Geomorphometry ................................ .......................... 6 5 3.1 Review of Geomorphometry ................................ ................................ ........................ 6 5 3.1.1 Evolution of geomorphometry : ................................ ................................ ..................... 6 6 3.1.2 Terrain analysis : ................................ ................................ ................................ ............ 67 3.2 Introduction to Periglacial Geomorphometry ................................ ......................... 6 8 3.2.1 Periglacial geomorphometry applied to the Beringian landscape : ............................... 6 9 3.3 Specific Geomorphometry of Cryoplanation Terraces ................................ .......... 6 9 3.3.1 Slope and aspect calculations : ................................ ................................ ...................... 71 3.3.2 Tangential and profile curvature calculations : ................................ ............................. 72 3.3.3 Low - pass filters : ................................ ................................ ................................ ............. 73 3.3.4 Elevation - relief ratio : ................................ ................................ ................................ ..... 74 3.3.5 Semi - automated identification of cryoplanation terraces (CTAR) : ............................... 75 3.4 Results from Semi - automated Identification of Cryo p lanation Terraces ........ 7 7 3.5 Conclusion ................................ ................................ ................................ ........................ 8 9 Chapter 4 Genera l Periglacial Geomorphometry ................................ .......................... 91 4.1 Introduction ................................ ................................ ................................ ..................... 91 4.1.1 Hypsometry : ................................ ................................ ................................ .................. 92 4.1.2 Application of hypsometry in general periglacial geomorphometry : ........................... 92 4.2 General Periglacial Geomorphometry : Methods ................................ ................... 94 4.2.1 Hypsometric integral for the Yukon - Tanan a Upland : ................................ ................... 95 4.3 Results of General Periglacial Geomorphometry ................................ ................... 9 7 4.3.1 Hypsometric analysis of the Yukon - Tanana Upland : ................................ .................... 9 7 4.3.2 Hypsometric analysis of the Basin and Range : ................................ ............................. 100 4.4 Conclusion ................................ ................................ ................................ ........................ 102 Chapter 5 Conclusion s ................................ ................................ ................................ .......... 10 7 5.1 Summary of Research ................................ ................................ ................................ ... 10 7 5.1.1 Large - scale geomorphological mapping : ................................ ................................ ..... 10 7 5.1.2 Specific periglacial geomorphometry : ................................ ................................ ........... 10 9 5.1.3 General periglacial geomorphometry : ................................ ................................ .......... 1 10 5.2 I s There a Distinctly Periglacial Landscape? ................................ ............................ 1 12 5.3 Recommendations for Future Work ................................ ................................ .......... 1 14 5 .3.1 Geomorphological mapping : ................................ ................................ ......................... 1 15 5.3.2 Specific periglacial geomorphometry : ................................ ................................ ........... 1 15 5.3.3 General periglacial geomorphometry : ................................ ................................ .......... 1 16 5.3.4 Cryoplanation terraces : ................................ ................................ ................................ . 11 7 viii A PPENDIX ................................ ................................ ................................ ................................ 1 1 8 REFERENCES ................................ ................................ ................................ ............................ 1 3 6 ix LIST OF TABLES Table 2 . 1 . Relative sizes, elevatio n, and orientations of terraces at study sites ....................... 35 Table 2 . 2 . Periglacial features found at the study sites : ................................ ............................ 42 Table 2 . 3 . Comparison of sorted stripes at Frost Ridge, Mount Fairplay, and Iron Creek : ........ 5 7 Table 3 . 1 . Summary of accuracy assessment for CT AR : ................................ ............................ 87 Ta ble 4 . 1 . Descriptive statistics of watershed size : ................................ ................................ .... 9 6 Table 4.2 . Sites for general geomorphometric analysis : ................................ ............................ 9 7 Table 4.3 . Comparison of HI for YTU sites : ................................ ................................ ................. 9 8 x LIST OF FIGURES Figure 1 . 1 . Photos of cryoplanation terraces at Eagle Summit, AK : ................................ .......... 8 Fi gure 1 . 2 . : ................................ ................................ ................................ 1 3 Figure 1.3 . A conceptual diagram of the periglacial landscape : ................................ ............... 14 Figure 1 . 4 . Proposed topoclimatic model for periglacial facies : ................................ ................ 1 5 Figure 1 . 5 . Study sites in eastern Beringia : ................................ ................................ ................ 1 6 Figure 2.1 . Illustration of the periglacial a ssemblage : ................................ ............................... 34 Figure 2.2 . Idealized sketch of cryoplanation terraces : ................................ ............................. 35 Figure 2.3 . Panoramic photos of study sites : ................................ ................................ ............. 3 7 Figure 2.4 . Study methods at Frost Ridge, BC : ................................ ................................ ........... 40 Figure 2.5a . T opography of Eagle Summit, AK : ................................ ................................ ......... 43 Figure 2.5b . Bedrock geology of Eagle Summi t, AK : ................................ ................................ .. 4 4 Figure 2.5c . Periglacial Features at Eagle Summit, AK : ................................ ............................. 4 5 Figure 2.5d . Subset of terraces at Eagle Summit, AK : ................................ ................................ 4 6 Figure 2.6a . T opography of Mount Fairplay, AK : ................................ ................................ ....... 4 7 Figure 2.6b . Bedrock geology of Mount Fairplay, AK : ................................ ............................... 4 8 Figure 2.6c . Periglacial feature s at Mount Fairplay, AK : ................................ ........................... 4 9 Figure 2.6d . Subset of terraces at Mount Fairplay, AK : ................................ ............................. 50 Figure 2.7a . T opography of Iron Creek, AK : ................................ ................................ ............... 51 Figure 2.7b . Surficial geology of Iron Creek, AK : ................................ ................................ ........ 52 Figure 2.7c . P eriglacial features at Iron Creek, AK : ................................ ................................ .... 53 Figure 2.7d . Subset of terrace s at Iron Creek, AK : ................................ ................................ ..... 5 4 Figure 2.8 . Geomorphological map of Frost Ridge, BC : ................................ ............................. 5 5 Figure 2.9 . Bedrock Geology of Frost Ridge, BC : ................................ ................................ ........ 5 6 Figure 2.10 . P rofiles of snowbanks at Frost Ridge, BC : ................................ .............................. 59 xi Figure 2.1 1 . Patterned ground from slope at Eagle Summit, AK : ................................ .............. 60 Figure 2.1 2 . Comparison of geomorphically active and inactive rock clasts : ............................ 61 Figure 3 . 1 . Flow diagram of the methods for the Semi - Automated Identification of CTs : ........ 7 6 Figure 3 . 2 . Map of sites w h ere CTAR was tested : ................................ ................................ ...... 7 7 Figure 3 . 3 . Map of slopes at Eagle Sum mit, AK : ................................ ................................ ........ 7 8 Figure 3 . 4 . Profile curvature map of Eagle Summit, AK : ................................ ............................ 7 9 Figure 3 . 5 . Tangential curvature map of Eagle Summit, AK : ................................ ..................... 80 Figure 3 . 6 . Map of ERR at Eagle Summit, AK : ................................ ................................ ............ 82 Figure 3 . 7 . CTAR accuracy assessment at Eagle Summit, AK : ................................ ................... 84 Figure 3 . 8 . CT AR accur acy assessment at Mount Fairplay, AK : ................................ ................. 8 5 Figure 3 . 9 . CT AR accuracy assessment at Kwethluk River, AK : ................................ ................. 8 6 Figure 3.10 . CT AR accuracy assessment at Sittookooyook River, AK : ................................ ....... 8 7 Figure 3.11 . CT AR accuracy assessment at Indian Mountain, AK : ................................ ............. 8 8 Figure 4 . 1 . The Yukon - Tanana Upland : ................................ ................................ ...................... 93 Figure 4 . 2 . Map of study locations for general geomorphometric analysis : ............................. 9 6 Figure 4.3 . Hypsometric curves for Y ukon T anana U pland study sites : ................................ ..... 9 9 Figure 4.4 . Map of Basin and Range study area , NV : ................................ ................................ 101 Figure 4.5 . Hypsometric curves for Basin and Range study sites : ................................ .............. 103 Figure A 1 . Hypsometric curves and integrals for 100 watershed size at Eagle Summit, AK : .... 11 9 Figure A 2 . Hypsometric curves and integrals for 1,000 watershed size at Eagle Summit, AK : . 1 20 Figure A 3 . Hypsometric curves and integrals f or 10,000 and 100,000 watershed size at Eagle Summit, AK : ................................ ................................ ................................ ................................ 1 21 Figure A 4 . Hypsometric curve and integral for entire Yukon - Tanana Upland : .......................... 1 22 Figure A 5 . Hypsometric curves and integrals for Sixtymile Butte : ................................ ............. 1 23 Figure A 6 . Hypsometric curves and integrals for wa tersheds at Cleary Summit, AK : ............... 12 4 Figure A 7 . Hypsometric curves and integrals for watersheds at Gilles Creek, AK : .................... 12 5 Figure A 8 . Hypsometric curves and integrals for watersheds at Boundary, AK and Yukon Territory : ................................ ................................ ................................ ................................ .... 12 6 xii Figure A 9 . Hypsometric curves and integr al for the glaciated site at Yukon River : .................. 12 7 Figure A 10 . Hypsometric curves and integrals for glaciated site at Fortymile River : ................ 12 8 Figure A 1 1 . Hypsometric curves and integrals for glaciated site at Healy River : ...................... 12 9 Figure A 1 2 . Hypsometric curves and integr als for glaciated site at Salcha River : ..................... 1 30 Figure A 1 3 . Hypsometric curves and integrals for glaciated site at Bear Creek : ....................... 1 31 Figure A 1 4 . Hypsometric curves for 1,000 watershed size in the Basin and Range : ................. 1 32 Figure A 1 5 . Hypsometric curves for 10,000 waters hed size in the Basin and Range : ............... 13 3 Figure A 1 6 . Hypsometric curves for 100,000 watershed size in the Basin and Range : ............. 1 3 4 Figure A 17 . Hypsometric curves for 1,000,000 watershed size in the Basin and Range : .......... 1 3 5 xiii KEY TO ABBREVIATIONS CT Cryoplanation Terrace CT AR Cryoplanation Terrace semi - Automated Recognition YTU Yukon - Tanana Upland DEM Digital Elevation Model DSM Digital Surface Model ERR Elevation Relief Ratio HI Hypsometric Integral SETSM Surface Extraction from TIN - based Searchspace Minimalization TIN T riangular Irregular Network USGS United States Geological Survey 1 Chapter 1 Introduction has been questioned in recent literature, spurring speculation about the importance of periglacial processes i n landscape - scale geomorphology. This growing body of literature includes French (2016), who postulated that cold, nonglaciated upland landscapes in Yukon Territory may be composed of inherited features, and that periglacial processes may only be responsib le for surficial forms. Andr é effectiveness of the mechanical weathering traditionally assumed to be associated with periglacial geomorphic activity. These crit ics based their claims on studies conducted on discrete periglacial landforms. However, few studies have focused on assemblages of features at varying scales and included landscape geometry in their analysis. This thesis seeks to resolve the question of th through application of mapping, geomorphometric analysis, and spatial - analytic techniques to the unglaciated upland landscapes of Interior and western Alaska . An under investigated class of landforms known as cryoplanation terraces (CTs) abound s in this region . This thesis also examines the proposition that CTs form the foundational building blocks of periglacial assemblages. 1.1 The Concept of Landscape defining principle for the discipline. Outside geography, the term landscape has often been used by artists, travelers, and members of the general public. Because of the many commonplace uses of th is term, a definition of the te rm for use in scientific research is essential. With the central 2 focus of this thesis being the periglacial landscape, an understanding of the term landscape is necessary. 1.1.1 History of the landscape concept: The concept of landscapes can be traced ba ck to the middle ages, where the term land was introduced as a term for a physical location . The term land was originally used as a political construct that relates to the social structure of medieval Europe (Olwig 2002, p. x iii). I n Germany , land was even tually expanded upon to became landschaft , a somewhat ambiguous term, which Hartshorne (1939) explained varied landschaft in German, was not clarified in its translation to the English term landscape . This ambiguity led to a loss of meaning and a lack of definition in the field of geography (Hartshorne 1939) . Early definitions of landscape and landschaft were based largely on the visual ( Penck 192 7 ; Waibel 1933; Gran ö 19 29 cf. Hartshorne 1939). Hartshorne (1939) expressed the need for a clearer term to describe the landscape. He reasoned that increased clarity was essential because the term was indistinguishable from the terms region and area . Based upon a review of the literature, Hartshorne (1939) was able to define landscape Hartshorne further clarified that the terms natural landscape and cultural landscape are actual landscape elements and not distinct types of landscape. Since landscape was translated to English and dispersed through the field of geography in the early 20 th century, there have been a series of refinements to the definition. As the term evolved over time, the focus of the geomorphic study of landscapes did as well. In the early 20 th century, the Davisian landscape evolution theory of normal erosion (Davis 1909) dominated the 3 study of geomorphic landscapes. Those who studied landscapes through Davisia n theory, focused on visual observations of the structure, process, and stage of the landscape. Such visual observations relied on regional geography (e.g., Fenneman 1914) and the delineation of landscapes through regionalization of geomorphic processes an d forms. Towards the middle of the 20 th century, the focus of landscapes shifted to one of the basic units of a landscape, landforms. The climatic origin of landforms, and subsequently landscapes, became a central focus of this era (e.g., Peltier 1950). Th is period was also the beginning of quantitative geomorphology (e.g., Strahler 1950). In the 1950s and the development of the quantitative revolution in geography, landscapes were described using quantitative measures that relate form to processes (e.g., S trahler 1952; Mark 1983). Today, studies of landscapes have expanded topically, with research investigating the paleoclimatic significance of the landscape, morphological studies of the landscape, and the sedimentary budgets of landscape elements (Haschenb urger and Souch 2004). As the focus of landscape evolution changed, so did the definition. What started out as highly ambiguous and being defined by the field of view of an individual, became defined by the specific landforms mak ing up a distinct area. 1 .1.2 Landscape ecology : While the quantitative revolution was transforming the field of geomorphology, the increase in spatial analytical techniques filtered into ecology, sparking the subfield of landscape ecology. The primary goal of landscape ecology is to understand the spatial variation of ecological processes (Cain, Bowman, and Hacker 2011). This subfield provides a method for understanding landscape elements and their relation to ecology and the physical geography of a landscape. The field often focu ses on patterns and spatial heterogeneity in the landscape (Turner 2005). Landscape ecology is necessary to this 4 discussion , as the subdiscipline constitutes one of the major perspectives on the landscape (Higgins et al. 2012). 1.1.3 Definition of land scape for this work: For the purposes of this thesis, landscape is conceptualized geomorphologically: it consists of a complex of landforms that vary in size, composition, and morphology across space (Cain, Bowman, and Hacker 2011). More specifically, the landscape is an assemblage of landforms arranged on the land surface. This landscape assemblage may contain forms that are inherited or azonal in character (King 1950; Ambrose 1964; Twidale 1972; cf. Haschenburger and Souch 2004). In this thesis, a nalysis was conducted to understand the landscape at multiple scales. Mapping and geomorphometry were used to understand the basic units of the landscape, the landforms. At a broader scale, analysis was conducted across a periglacial region in an attempt to ident ify a distinct signature of periglacial environments. 1.1.4 Scale of the landscape: Scale is a central component in the study of geography, ecology, and many other disciplines. Without understanding the scale of a study incorrect conclusions may be draw n. In this work, the question of what the appropriate scale i s to study the periglacial landscape is posed . A review of literatures related to defining the scale of landscapes shows that there is no clearly stated quantitative definition of landscape scale , although there is general agreement in the discussion of landscapes that they are scale dependent (e.g. Forman and Godron 1981; Withers and Meentemeyer 1999; Wu and Hobbs 2002; Higgins et al. 2012) . One of the few studies to actually provide dimensions to landscape was Delcourt and Delcour (1988), who defined the scope of landscape ecology as covering those areas that are meso scale in nature , from 1 km to 100 km in width. Other more 5 descriptive definitions have been proposed, such as Parker and Bendix ( 1996 ) who stated that landscape i s intermediate between landforms and region. Forman and Godron (1981) provide both descriptive and quantitative bounds. When defining a landscape, they state that the landscape is geomorphically homogenous areas that vary i n size but have distinct boundaries. Their discussion further defines the size as kilometers wide. While there appears to be some agreement that the size of landscapes fall into the meso scale range or that of tens of kilometers, this lack of a formal, agr eed upon landscape scale must be addressed. Higgins et al. (2012) advocates that this sort of integrated landscape concept would be able to transcend disciplines. However, they also point out that there appears to be a degree of disagreement over how to de fine and view scale within and between disciplines , an issue that must be resolved . 1.2 Periglacial Geomorphology The term periglacial climatically controlled features adjacent to the Plei This first definition limited its utility in describing processes in regions such as Beringia, which were largely unglaciated during the Pleistocene. In the mid - 1900s the International Geographical Union (IGU) rec ognized the subdiscipline of periglacial geomorphology. During the same period, a published outlet for periglacial research, the journal Biuletyn Peryglac j alny , was introduced ( 1954 ) . These events formalized and expan ded the definition to include all cold - climate, non - glacial processes and landforms (French 2017). The broadening of the definition, which explicitly excludes proximity to glaciers as a criterion (Washburn 1980), allows for an increased range of study. 6 Although its existence is broadly accepte d, a definition for the term periglacial landscape has been the subject of scholarly debates (e.g., André 2003 ; French and Thorn 2006; French 2016; French 2017 ) . These discussions are exacerbated by the nebulous character of the the lack of precision in the definition of the periglacial zone . The term zone latitudinal and altitudinal zonation of landscapes. Tricart and Cailleux (1972) provide d a detailed accounting of the zonation of landscapes and associated processes. A zonal process is one associated with a distinct climatic regime within which that process or landform can be found. Other processes are azonal, in that they are found across all climatic zones and lack a well - defined limit. Some scientists (e.g., Tricart 1967; Péwé 1969) provide evidence that the periglacial zone is linked with underlying permafrost and can thus be defined using the thermal properties of frozen ground, i.e., areas in which subsurface la yers remain at or below 0 o C. However, this concept has been disregarded by most because not all landforms considered periglacial are associated with permafrost (French 2017). Other attempts to create quantitative climatic boundaries for periglacial region s have been proposed . A few of th ose boundaries are: a mean annual air temperature of between - 1 5 o C and - 1 o C (Peltier 1950), - 12 o C and 2 o C (Wilson 1968), and below 3 o C (French 2017). Although these boundaries may have merit, it is more common to e ncounter open - ended definitions that propose general criteria for the periglacial landscape, such as intense frost action and lack of perennial snow (i.e., Washburn 1980). Most definitions of periglacial zones describe them as being poleward of (or in alpi ne areas above) the tree line and excluding glacial processes. Given the prevalence of such unrestricted definitions, it is not surprising that some 7 periglacial landscapes could be construed as i nherited landscapes covered with surface decorations (Bir ot 1968). A holistic examination of the landscape that includes features involving entire slope sequences is necessary to better understand whether distinctively periglacial landscapes exist and what can or should be included in that conception. 1.3 Cryop lanation Terraces 1.3.1 Morphology: Cryoplanation terraces (CTs) are large periglacial landforms characterized by alternating treads and risers, giving the appearance of giant staircases (Figure 1.1). The risers (scarps) are typically covered with clastic rubble, while the nearly planar treads are a mosaic of vegetation, rock debris, and surficial periglacial landforms. One of most comprehensive examinations of cryoplanation terraces was conducted by Reger (1975), who mapped and classified 686 cryoplanation varied in size and morphology, with scarps ranging from 3 to 76 m in height and slopes of 9 - 32 o . Additionally, Reger found that treads range in length from as little as 5 m to hundreds of meters, and have considerably gentler slopes (Reger 1975) . The overall size of these landforms can be anywhere from 3,000 to 845,000 m 2 (Reger 1975) and is such that they constitute entire slope sequences. A l though structural control has not been shown to be a dominant factor in cryoplanation terrace formation (Washburn 1980 p. 240 ) , l ithology has been suggested as important (French 2017), with numerous observations made regarding the geologic attributes of terraces. Eakin (1916) , in his initial survey of terraces in the Yukon - Koyukuk Region, found that they are best developed in areas of Mesozoic rock composed of granite, greenstone, and quartzite schist. Reger (1975) 8 rock as a common bedrock type in whi ch cryoplanation terraces are formed. Reger also found that well - developed terraces were not common on fine - to medium - grained sedimentary rocks, an observation that appears consistent with other work (Demek 1969). Figure 1.1 . Photos of cryoplanation terraces at Eagle Summit, AK : A) series of cryoplanation terraces. The tread and riser (scarp) are highlighted to indicate the various components of the cryoplanation system. T he tread is 200 m long and the scarp is 2 5 m in height. B) scarp of cryoplanation terrace, view to east. C) tread of cryoplanation terrace in part A , view to west. Note the rubble at the base of scarp and patterned ground beyond. 9 1.3.2 History: Geologic observations of a general nature were common among early 20 th century U.S. Geolog ical Survey ( USGS ) scientists operating in Alaska (e.g., Moffit 1905; Prindle 1905; Mertie 1937) and provided a baseline for geomorphic research across Alaska. These studies were necessarily of a general nature and l acked in - depth investigations of landforms such as cryoplanation terraces. Additionally, in the intervening years the paradigm of landscape creation has shifted away from the Davisian model that influenced these pioneers. Today, those hundred - year old stud ies provide concise descriptions and powerful insights and form a useful basis for contemporary investigations into periglacial landscapes. Currently, the predominant hypotheses regarding the origin of CTs holds that they form through a climatically cont rolled suite of processes ( nivation ) involving intensified weathering and sediment transport in the vicinity of late - lying snowbanks (Cairnes 1912). Proponents of this hypothesis cite research showing that cryoplanation terraces exhibit altitudinal similar ities with glacial cirques (Nelson 1989; Nelson and Nyland 2017) and display poleward orientation similar to that of glacial cirques, also indicating a climatic origin (Nelson 1998) . Others state that cryoplanation terraces are related to permafrost and other periglacial processes such as solifluction (Reger and Péwé 1976) . Critics of the climati c - origin hypothesis of CT development raise several issues. French (2016) questioned the moisture regime available in periglacial regions, pointing out that many landforms found in periglacial regions show similarities to those in arid regions. Washburn (1980 p. 240 ) cited the pre sence of patterned ground on terrace treads as evidence that material transport on the terraces is too weak to remove material from terrace treads. As Thorn and Hall (2002) point out, however, little field - based research focusing on CTs has been undertaken in recent years. 10 1.3.3 Nivation: Weathering and transportation processes associated with late - lying snow patches , termed nivation by Matthes (1900), ha s been cited as being responsible for the formation of CTs. Matthes postulated that mechanical weatheri ng and mass wasting are intensified in the vicinity of snowbanks. Since Matthes first proposed the idea, process - oriented nivation studies (e.g., Thorn 1976; Berrisford 1991) have increased understanding about how snow patches enlarge initial topographic i rregularities on slopes, and the specific mechanisms involved. Thorn (1976) calculated long - term erosion rates in the Colorado Front Range and stressed the importance of both chemical and mechanical weathering for the creation of nivation hollows. He concluded that the nivation process suite is not effective enough to support long - held notions about the existence of a form continuum extending from nivation hollows through to glacial cirques. Nivation research has been expanded by others, including Ballantyne (1978) , who investigated the hydrologic significance of the nivation system. Hall (1993) studied the breakdown of bedrock under the influence of late - lying snow, and Caine (1992) studied the sediment loads of snowpatches. et al . (2018) calcula ted a long - term erosion rate of 0.77± 0.12 mm/yr in a nivation hollow on James Ross Island near the northeastern tip of the Antarctic Peninsula . The aforementioned studies are just a few that have continued to refine the understanding of localized er osion associated with late - lying snow. Cairnes (1912), although not widely credited, was the first to suggest that the nivation suite of geomorphic processes is responsible for development of cryoplanation landforms. While this interpretation has been acc epted by many (e.g., Demek 1969; Reger 1975), recent literature (i.e., Thorn and Hall 2002; French 2017) points out the lack of field - based process 11 investigation, highlighting the need for research that would lead to detailed understanding of the dynamics responsible for the formation of cryoplanation landforms. 1.4 Periglacial Assemblages periglacial realm . This term encompasses all areas that are very cold , but not glacial. Many of those areas have only recently become periglac i al by virtue of deglaciation . Because much of the periglacial realm has not been operated on by periglacial processes for long enough, only a relatively small subset of the periglacial realm could feasibly have for med a periglacial landscape. This thesis proposes that the periglacial landscape is , in turn, made up of an assemblage of various landforms ( Figure 1.3 ). In the German periglacial literature, this periglacial landscape is described as a series of the research in periglacial geomorphology has been focused on small features such as sorted patterned ground and solifluction lobes (e.g., Troll 1944; Furrer and Dorigo 1972 ; Graf 1973; Rudberg 197 7 ) that constitute the components of form communities. Birot (1968, 125) evolution. The concept of a periglacial assemblage advanced in this thesis extends over a wide range of landform size , including small (e.g., sorted patterned ground) , meso - forms (e.g., solifluction lobes and terraces) and macro - forms ( l andscape - scale features such as cryoplanation terraces and cryopedime nts ). Although this array of landforms may s eem to be a chaotic mixture , the various forms ar e closely related through a common origin : f reezing and thawing of earth materials in the presence of water. 12 The periglacial assemblage is composed of variously s ized features that, considered as a unit, create a distinctive pattern extending across a broad spectrum of geographical scale. This concept was introduced indirectly in an unpublished report by Brunnschweiler (1965) and is illustrated in Figure 1.2 as the altiplanorium. conceptualizing the periglacial assemblage in the context of cryoplanation (altiplanation) terraces, large periglacial landforms that remain inadequately studied from process - based and geomorp hometric perspectives. A potentially useful approach to the periglacial landscape problem is through the concept of facies , units of rock or sediment formed in distinct environments and recognizably distinct from adjacent units. The facies concept can be u tilized in understanding and creating a general framework for the environments in which periglacial features form. The concept was developed by Madole (1972) , who investigated the spatial distribution of Neoglacial facies in the Colorado Front Range. His hypothesis was that if the spatial arrangement of facies could be linked to their genesis then they could be significant i ndicators of past depositional environments at the local scale. Morris and Olyphant (1990) creating a physical model for facies deposition. This model incorporated topoclimate and geology to predict the type of lithofacies that would form under various environmental conditions. A similar model, as proposed in Figure 1.4 , would be useful in understanding and expanding on the concept of periglacial facies . 13 Figure 1.2 . : The altiplanorium is a n integrating concept based on assemblages of minor periglacial features superimposed on much larger erosional periglacial landforms (altiplanation [cryoplanation] terraces). Figure from Brunnschweiler (1965). 14 Figure 1.3. A conceptual diagram of the periglacial landscape : This diagram show s how the periglacial landscape is made up of an assemblage of individual landforms. Landforms are the most basic unit of this study . A repeating series of these landforms is an assemblage, w hich in turn makes up the periglacial landscape. The periglacial realm consists of all these elements and all other periglacial features. This graphic i llustrates how the landscape scales from the individual periglacial landforms to an assemblage, which in turn creates a landscape. The periglacial landscape, assemblage, and landforms all fall into the periglacial realm. Scale indicated is non - linear and numbers are an approximation. 15 In establishing a model for periglacial facies, certain morphologic and climatic factors were accounted for. Because of their relatively large size and dominance on the periglacial landscape, CTs are the logical basis for this model. A review of the literature indicates that potential factors influencing the formation of cryo planation terraces include climate (Reger and Péwé 1976; Nels on 1998; Nelson and Nyland 2017) , bedrock type and structure ( Reger 1975; French 2017) , and topography (Nelson 1979b, 1989). The concept of periglacial facies, which is closely related to the assemblages discussed above, will facilitate understanding about the interplay between environmental parameters and how they are incorporated into periglacial landscape development. Figure 1 .4 . Proposed topoclimatic model for periglacial facies : A proposed model for the development of a periglacial landscape and associated landforms (green bubbles) based on a series of filters (blue bubble) and processes (brown bubbles). The model assumes that climate is the driving force behind periglacial environments. 16 1.5 Study Area Beringian Uplands Beringia (Figure 1.5) is the landmass north of 55 o N extending roughly from the Lena River in eastern Siberia to the Mackenzie River in western Canada. Beri ngia is composed of three subregions: (a) eastern Beringia, the area between the former Bering Land Bridge and the Mackenzie River; (b) central Beringia, the former Bering Land Bridge (now the Bering Strait and islands within); and (c) western Beringia, ex tending west from the land bridge to the Lena River in Central Siberia. Figure 1.5 . Study sites in e astern Beringia : The Bering Land Bridge (dark grey) and Pleistocene glaciers (light grey) are delineated to provide context . Study sites across Alaska and in British Columbia are indicated by black dots. Map adapted from Brubaker et al. (2005). 17 1.5.1 Glacial and environmental history: During the last glacial maximum (LGM) Beringia remained largely unglaciated, with only scattered mountain glaciers existing at higher elevations of the generally ice - free corridor between the Alaska and Brooks Ranges. The southern and eastern boundaries of Beringia were defined by the large ice sheets. The Cordilleran Ice Sheet coalesced with mountain glaciers in the Alaska R ange at the southeastern edge of Beringia. On the eastern edge of Beringia, the Laurentide Ice Sheet met with the Cordilleran, cutting off this ice - free region from the rest of North America (Brigham - Grette 2001). During the LGM, the climate in eastern Ber ingia was highly continental, owing to the physical barriers to the Arctic and Pacific Oceans created by the Brooks and Alaska Ranges. Because of its proximity to the ocean, central Beringia experienced a more marine climate with milder temperatures and a mesic environment. At higher elevations in eastern Beringia mean annual temperatures and precipitation decreased, leading to a cold, semi - arid environment (Hopkins et al. 1982; Elias 2001). Vegetation in central Beringia was mesic tundra dominated by mois ture - loving plants (Elias et al. 1997). Eastern Beringia was dominated by steppe tundra, a biome that is non - existent today, although the region was dominated by tundra plants that could survive with little moisture (Guthrie 1982). Paleoecologists disagree about whether Beringia was an arboreal refugium, but nonetheless it is unlikely that trees were present in upland areas of eastern Beringia (Hopkins, Smith, and Matthews 1981; Brubaker et al. 2005). Around 18,000 years ago, two concurrent events changed the nature of the Beringian landscape: (1) the ice sheets began to retreat; and (2) the Bering Land Bridge became inundated with water as sea level rose (Brubaker et al. 2005). Subsequent changes in the 18 environment resulted in a decrease in continentality in eastern Beringia . Nelson and Nyland that contemporary eastern Beringia is still experiencing a pronounced continentality gradient that increases with distance f rom the Bering Sea. The opening of a corridor between the Laurentide and Cordilleran Ice Sheets allowed trees to migrate north and vegetation to transition to the tundra vegetation observed today (Brubaker et al. 2005). 1.5.2 Periglacial studies in Bering ia: Although periglacial studies have been numerous in eastern Beringia, much of the research is fragmented and focused on discreet (specific) landforms, including patterned ground (e.g., Church, Péwé, and Andersen 1965 ), rock glaciers (e.g., Wahrhaftig an d Cox 1959), frost mounds (e.g., Mackay 198 6 ), and solifluction (e.g., Hanson 1950; Millar 2005). Curiously, only a few published studies have been devoted specifically to cryoplanation landforms in Beringia (Cairnes 1912; Eakin 1916; Obruchov 1937; Chaiko 1988; Péwé 1970; Reger and Péwé 1976). The dissertation by Reger (1975), although qualitative, stands out conceptually as the most comprehensive work ever produced about cryoplanation terraces in eastern Beringia. Moreover, the thesis contains the most co mprehensive CT data set ever assembled. Unfortunately, with the exception of one short 1975) data set to demonstrate similarities in the geographic distribution (Nelson 1989), elevation trends (Nelson and Nyland 2017), and orientation (Nelson 1998) of CTs and glacial cirques. However, no major work has been conducted that focuses on periglac ial assemblages in the Beringian landscape, 19 1.6 Statement of Problem Beringia should provide an ideal example of an upland periglacial landscape. Owing to ultracontintental climatic con ditions during Pleistocene cold intervals, most of Beringia remained unglaciated during the entire Quaternary Period ( Figure 1.5 ). Periglacial processes have, the oretically , operated over extended periods, possibly in nearly continuous fashion, throughout much of this vast area over the past 1.8 Ma. According to the tenets of climatic geomorphology (e.g., Birot 1968; Tricart and Cailleux 1972; Büdel 1977), this largely unbroken period of cold nonglacial conditions should, in principle, have resulted in a va st region in which the imprint of periglacial conditions prevails in the form of characteristic landform assemblages (cf. André 2003; French 2016). Periglacial landforms have been documented in a variety of sizes throughout Beringia, from small (needle ic e forms and patterned ground) and medium - sized constructional features (pingos and solifluction sheets) to large erosional forms encompassing entire slope sequences (cryoplanation terraces and cryopediments). Together, these landforms can be viewed as cons spectrum of geographical scale. To date, however, this concept has not been applied in unglaciated Beringia, which ironically is one of the very few extensive regions on earth in which periglacial processes have operated continuously over an extended period of geological time. Spatially oriented analytic and statistical methods can be applied to periglacial assemblages in much the same way that they are applied to ecologi cal associations in landscape ecology (see review by Turner 2005). Progress has been achieved along these lines in the periglacial realm (see work by Hjort et al. 2007, 2010, and references therein), but most such 20 studies have been concerned with features in the smaller part of the periglacial landform scale continuum, and usually in polygenetic landscapes. This thesis applies the form community concept in the periglacial realm to answer the question: is there a characteristic periglacial landscape? This q uestion has been gaining prominence in periglacial geomorphology (Andr é 2003; French and Thorn 2006; French 2015; French 2017) . Beginning with Andr é ncreased skepticism about the existence of a dominantly periglacial character exist and suggested that cryoplanation features may be inherited from past warm inter vals. However, these studies focused on a single scale, and a holistic analysis utilizing a multi - scale approach has not been undertaken. This thesis explores ma pping and spatial - analytic techniques to assemblages of periglacial landforms; and (b) use of geomorphometric procedures to identify cryoplanation landforms and to ascertain whether the topography of eastern Beringia exhibits distinctive and characteristic geometric and hypsometric signatures. This topical approach is highly unusual because it extends across a wide spectrum of geographical scale, is focused on a cold region that remained largely unglaciated throughout the Quaternary and considers both const ructional and erosional periglacial landforms. 1.7 Hypothesis This thesis tests the hypothesis that , a characteristic periglacial landscape does exist. Evidence for this assertion will be found in the existence of a repeating mosaic of periglacial 21 assembla ges over a continuum of spatial scale and in the distinctive geomorphometric properties ( signature ) of upland Beringian topography. Th e thesis will evaluate this hypothesis through two related methodologies: (1) documentation, primarily through mapping, of the existence of repeating sedimentological and periglacial landform assemblages, extending across multiple spatial scales; and (2) analysis of periglacial landforms and Beringian terrain, using both specific and general geomorphometric tools. These met hodologies are described below. 1.8 Methods Study locations were determined in conjunction with a larger project investigatin g cryoplanation terraces on a transect across eastern Beringia. Alaskan study sites include cryoplanation landforms near Iron Cree k (Seward Peninsula), Eagle Summit (interior Alaska), and Mount Fairplay (near the Alaska - Canada border). Together, these sites extend through terrain rising from 300 m above sea level in the west to over 1400 m in the east. A site at Frost Ridge in the Ca thedral Massif near Atlin, British Columbia was chosen because the suite of periglacial processes collectively known as nivation were demonstrably active during the mid - 1970s (Nelson, personal communication 2017). 1.8.1 Geomorphologic analysis and perigla cial assemblages: Geomorphological mapping is the cartographic expression of geomorphic processes and landforms. It is often an integrated representation of climate, soils, landforms, topography, geology, and other factors (Smith, Paron, and Griffiths 2011) (Bishop et al. 2012 pp. 5). 22 Distinctly geomorphological mapping methods can be traced back over 100 years to Passarge (1914), who is often credited with creating the first geomorphological map (Smit h, Paron, and Griffiths 2011). The basis for modern geomorphological mapping in the United States goes back to the late 1800s, however, when regional geomorphology was standard practice for geomorphologists. Mapping at that time was at small spatial scales (i.e., involving large areas) and often focused on specific themes (Thornbury 1965). Powell (1895), while operating under the regionalist viewpoint, mapped the physiographic regions of the United States. This map definitively separated the United States i nto regions and spurred further work in the coming era. A shift in thought with regard to geomorphological mapping began after creating the geomorphologic maps of Fenneman (1917) created a standardized process for descriptive physiographic mapping. By the 1950s the geomorphologic paradigm had shifted away from the Davisian idea of landscape genesis, changing the nature of geomorphology and its influence on geomorphologic mapping (Bishop et al. 2012) to a more process - oriented focus that necessarily changed the spatial scale at which investigation could take place. While geomorphological mapping in the United Stat es (e.g., Thornbury 1965) and Canada (Bostock 1970) was focused on regional aspects, much of the scholarly progress in the subject at more local scales came from Europe. In Europe, the precursors to geomorphologic mapping were mostly descriptive, with car tographic representation of terrain being dominant (Smith, Paron, and Griffiths 2011). Passarge (1914) is considered to be the first to publish a geomorphologic map, with only sporadic geomorphic maps following until after World War II. With the war neces sitating a 23 better understanding of terrain, cartography and mapping became a priority in Europe. After World War II, geomorphological mapping became more commonplace, with countries such as France, Czechoslovakia, and Hungary focused on mapping based on li thologic units, while Germany, Poland, and the Soviet Union considered landform - creating processes (Bishop et al. 2012). Small - scale mapping was also practiced widely in the Soviet Union (Smith, Paron, and Griffiths 2011). Countrywide geomorphological map ping became common practice in Europe, after Klimaszewski (1956, 1963) and Galon (1962) mapped Poland and Annaheim (1956) mapped Switzerland. In the latter half of the 20 th century, European geomorphological mapping focused on applications such as engineer ing, environmental science, natural resource management, and related fields. Because of the high level of interest in applied geomorphologic mapping, Europeans have contributed significantly to the standardization of mapping symbols and legends (Demek 1972 ; Demek and Embleton 1976). With the Cold War and subsequent advances in aerial imagery and satellite technology, digital image analysis became standard practice in geomorphological mapping. Advances in digital technology, specifically remote sensing, greatly increased the scale and breadth of geomorphological mapping (Smith, Paron, and Griffiths 2011). Today, the use of remotely sensed imagery is combined with digital elevation models (DEMs) to create large - scale representations of geomorphic processe s. In hard - to - reach locations such as the Arctic, digital technology is fast becoming the standard (Evans 2012) . Given the large expense of conducting research in the Arctic, using DEMs and other remotely sensed information can save time and money. Advances in technology have led to increased spatial resolution, ma king more geomorphic information available (Evans 2012). 24 While geomorphological mapping is often focused on terrain under temperate and alpine climates, some studies have applied the concept to periglacial features. Caine (1972), Rudberg (1972), Graf (19 73), Åkerman (1980), Hjort et al. (200 7 , 2010), and several of the authors in the edited volume by Poser (1976), for example, have focused on periglacial geomorphological mapping and other forms of spatial analysis. Karte (1979, 1982, 1983) provided an ove mapping of periglacial phenomena. For the purposes of this project, large - scale (e.g., 1: 10,000) geomorphologic maps will depict aspects of geology, general and periglacial geomo rphology, and vegetation at the study sites. The surficial geology of the study areas has been documented by previous researchers (i.e., Foster 1967) and will be compiled via literature review. Collection of ma pping data was accomplished through a combination of general field survey techniques and acquisition of digital elevation models (DEMs). Sampling of terrace geomorphology involved collecting an extensive group of photos of periglacial features, identificat ion and classification of surface geomorphic features, identification of vegetation, and the spatial zonation of these features. Identification in the field was followed with laboratory review using photographs. Surveys of terrace profiles were conducted d uring the summers of 2016 and 2017. At each site a general survey was undertaken that incorporated the entire terrace and was used to inform decisions in the field. Base maps were obtained using the 2 m resolution Arctic DEM (Polar Geospatial Center 2017), as well as maps created by Reger (1975) . The mapping phase of this work was concluded in the lab by combining data into a GIS 25 using ArcGIS and other geospatial software. Mapping symbols were modified from Demek (1972) and Gustavsson et al. (2006). 1.8.2 Periglacial geomorphometry: Geomorphometry is the quantitative study of the land surface (Pike, Evans, and Hengl 2009) and lies at the intersection of mathemati cs, geosciences, and computer science. Evans (1972, 1987) divided the study of geomorphometry into two distinct categories: specific and general geomorphometry. Specific geomorphometry is the study of individual (discreet) landforms of specific origin or m orphological characteristics, while general geomorphometry investigates the continuous land surface (Pike, Evans, and Hengl 2009). In the context of this thesis, specific geomorphometry will focus on the cryoplanation landforms themselves and general geomo rphometry will focus on extensive tracts of unglaciated uplands in eastern Beringia. Geomorphometry has been studied since the late 1700s, but its origin dates to the beginning of human attempts to understand the shape of the earth through mathematics and visualization. Much of the evolution in the discipline of geomorphometry took place in Germany and Austria, where the subject has evolved with earth science, especially geography (Pike, Evans, and Hengl 2009). Today, the study of topography through contou r maps and drawings (e.g., Clarke 1966) has been largely replaced with digital technology and computer algorithms. With the new generations of space satellites and a variety of new types of airborne sensors, the resolution of digital information has increa sed rapidly. Now, with LiDAR (Light Detection And Ranging) and radar applications, the resolution of digital mapping has even become an issue, with too much information available, creating problems related to storage and analysis (Evans 2012). 26 Applicatio ns of geomorphometry in the periglacial realm have been few (e.g., Nelson 1998). Most geomorphometric studies have focused on fluvial and glacial landscapes (e.g., s imilar to that provided by King (1982) for glacial geomorphometry, which has since developed into a mature subject making extensive use of contemporary technology and tectonic theory (e.g., Mitchell and Montgomery 2006; Sternai et al. 2011). Chapters 3 and 4 constitute an initial attempt at developing a new field of study, tentatively named periglacial geomorphometry. This is accomplished in the context of cryoplanation landforms as a method for recognizing and analyzing periglacial landscapes. Application of the tools of specific and general geomorphometry to digital representations of topography will help to assess the validity of concerns raised by French (2016) and others skeptical about the validity of the cryoplanation concept, as well as the notion of a characteristic periglacial landscape. Two general goals are involved: (1) identification of discreet cryoplanation landforms through manipulation of digital reg ional - scale analysis of the geometric and hypsometric properties of the topography. 1.8.3 Specific geomorphometry: To achieve the goals laid out above, specific criteria for the identification of CTs have been developed. It was anticipated that the charact eristic sharp topographic break at CT scarp - tread junctions, together with adjacent flats (treads) will constitute the primary pattern - recognition tool for identifying the locations of individual CTs. Using the Alaska 2 m DEM ( Polar G eospatial Center, 2017 ), these inflections should be recognizable through a kernel operator (moving window operation) (Clarke 1995 pp. 55) searching for minima and maxima in the derivatives of altitude. Remotely sensed imagery, 27 - scale map of t he Indian Mountain area (Péwé and Reger 1.8.4 General geomorphometry: General geomorphometric tools involve hypsometric analysis , which treats the distribution of the ground surface area of a lan dmass with respect to altitude. The procedure used in this thesis is hypsometric integral , a measure of actual landmass volume with respect to a reference solid with the same basal (map - view) configuration. This section of the th esis will employ the elevation - relief ratio , a measure of the ratio of upland to lowland (Pike and Wilson, 1971) proposed by Wood and Snell (1960). The elevation - relief ratio was mathematically proven by Pike and Wilson (1971) to be equivalent to the hypso metric integral proposed by Strahler (1952), but easily computed from DEMs. Nelson (1979b) conducted preliminary analysis indicating that the hypsometric integral may be a valuable tool for identifying a periglacial signature in the landscape. Owing to the unusual preponderance of flat - which is the opposite of what studies in areas of fluvial erosion conc lude. The corresponding hypsometric curve may be unusually convex, showing a large proportion of the land area at relatively high elevation, and thus providing a signature of terrain developed under periglacial conditions. This analysis expand s upon Nelso - relief ratio to conduct a multi - scale analysis of watersheds across the Yukon - Tanana Upland (Wahrhaftig 1965). 28 The methods and analyses proposed above will help to quantitatively address concerns in recent literature (e.g., André 2003; French 2016) regarding the existence of a characteristic periglacial landscape. By identifying periglacial landscapes in the context of cryoplanation terraces links will be established, across a range of spatial scales, between landform assemblages, terrain geometry, and periglacial conditions. 29 Chapter 2 Mapping and Multiscale Analysis of Periglacial Assemblages -- V. Van Gogh characteristics of the Arctic is the abundance and arrangement of a number of small, relatively uniform, features such as oriented lakes, drumlins, meander scars, ice - wedge polygons, string bogs, solifluction stripes, and block fields. Indeed, the repetiti ve nature over large areas of such forms has prompted many visitors to use the term monotonous H. J. Walker 1983 i b out of many, one ), created the perfect literary image of the assemblage that forms Arctic and sub - Arctic periglacial landscapes. The concept of a periglacial assemblage is simple; there exists a series of periglacial features that merge and superimpose, forming an assemblage. 2.1 Background G eocryology consists of two closely interrelated areas of scientific inquiry: the study of permafrost (including engineering) and of relatively small and discreet surficial landforms. Permafrost , earth material continuously at or below 0 o C for two or more years (Washburn 1980), is unquestionably important in the periglacial realm. Permafrost is a central focus of geocryology and is necessary for numerous periglacial landforms such as pingos, ice - wedge polygons, and gelifluction terraces (French 2017). Howev er, by most definitions, permafrost is not necessary for landforms to be classified as periglacial (Washburn 1980). Landforms frequently referred to as periglacial may be formed through polyzonal or azonal processes that are not reliant on permafrost. For example, frost shattering of rock clasts and bedrock due to 30 freeze - thaw cycles is distinctly periglacial. Some features, such as cryoplanation terraces, have not been studied enough to fully understand their relation to permafrost (French 2017). Although s ome writers have asserted a strict relationship between CTs and permafrost, (Reger and Péwé 1976) most (e.g., Demek 1969; French 2017; Ballantyne 2018) state that it is not necessary for CT formation. Small (cm to 1 m) and medium (meter to decameter) siz ed periglacial landforms have been the subject of much study in periglacial geomorphology. Beginning with early exploratory expeditions to the Arctic and Antarctic in the 1800s (French 2017), qualitative observations of these landforms were common. In the early 1900s, quantitative study began to supplement the qualitative observations by prospectors and explorers in the Arctic and Antarctic regions. With a few exceptions (Lozinski 1909, 1912; Andersson 1906) it was not until the mid - 20 th century that proces s - based scientific understanding of periglacial landforms really developed. Lozinski (1909, 1912) is widely credited with coining the term periglacial and Andersson (1906), with his studies on solifluction, was one of the early pioneers of the subject. Stu dies of solifluction (e.g., Benedict 1970; Price 1974; Matsuoka 2001) have shown it to be an effective mass movement process involving the transport of material downslope in periglacial areas. Gelifluction, a closely related process, relies on material flo wing slowly over permafrost. In this thesis, the term solifluction will be retained because it is more general and because the presence of permafrost at all locations has not been confirmed. Patterned ground features, which Ballantyne (2018, pp. 148) desc (1944) was among the first investigators to undertake comprehensive, worldwide study of the 31 morphology and genesis of patterned ground. Many s tudies followed that served to better understand how these features formed (e.g., Washburn 1956; Goldthwaite 1976; Ballantyne 1996; Kling 1997; Boelhouwers et al. 2003). Similarly, frost boils are of considerable interest in periglacial studies (e.g., Mack ay 1980; Walker et al. 2004). The study of clastic fragments and how they react to periglacial processes (e.g., Corte 1963), is closely related to patterned ground. Frost jacking (Dyke 1984; Anderson 1988) is commonly observed in periglacial areas, as are blockfields (Rapp 1967; French 1987). Blockfields have also been used to identify past periglacial landscapes (Park Nelson, Nelson, and Walegur 2007). Although the list of studies conducted on small landforms is long and comprehensive, there have been few English - language studies that provide a link between these features and assemblages of periglacial Curiously, with the exception of a few in - depth publications (e.g., Demek 1969; Reger and Péwé 1976; Nelson 1989; Thorn and Hall 2002), little attention has been given to the foundational building blocks of the periglacial assemblage: cryoplanation terraces. 2.1.1 Periglacial form communities: s (1944) survey of the periglacial realm, German - language literature is in sharp contrast to the focus on singular types of periglacial feature that has dominated Anglophone periglacial science. Karte (1979) reviewed the concept of periglacial form commun ities, widely investigated in German - language literature, in his discussion of periglacial zonation. This concept is similar to the ecological concept of community that occur together community concept applies the ecological definition to periglacial forms. The periglacial form 32 community concept can thus be expressed as a group of periglacial forms that interact with each other in the same geographic space. This concept establishes a link between the individual forms, the processes that form them, and the surrounding forms (Kelletat 1970; Garleff 1970; Karte 1979). By allowing features to interact over multiple scales, knowledge of each individual and its formation helps to explain the geographic location of other, similar features, thereby forming an assemblage. This concept of an assemblage of periglacial features standing as a distinct landscape is in direct oppositi on to the views of André (2003) and French (2016). 2.1.2 Periglacial assemblage: Based on the periglacial form communities discussed in the German - language literature (e.g., Karte 1979) and the qualitative observations of various authors, the periglacial assemblage could be viewed as the core of the periglacial landscape, with cryoplanation terraces and cryopediments functioning as the foundational building blocks. The periglacial assemblage is defined formally in this thesis but has been conceptualized pr eviously in figures and diagrams by multiple authors. Brunnschweiler (1965) created a schematic of a periglacial assemblage in what he termed the altiplanorium. term was based on his observations and detailed sketches (Figure 1.2) of altip lanation terraces (now known as cryoplanation terraces). The term was derived as an analog for the long - existing geological features an ti clinorium and synclinorium , large structural features on which minor folds are superimposed. In Europe, Karte (1979, 14 6 - 147) expressed the periglacial assemblage as a series of landforms from small, patterned ground features and frost - jacked clasts to graphical expression of the peri glacial form community concept. Åkerman (1980) provided a similar graphic based on empirical data from West Spitsbergen. Reger (1975) described the 33 periglacial assemblage in an idealized diagram of a system of cryoplanation terraces (Figure 2.2). In his fi gure, Reger illustrates cryoplanation terraces as having a series of smaller, surficial periglacial features inscribed on them, an observation consistent with those presented in subsequent paragraphs of this thesis. These detailed descriptions of the inter relationship between small features, CTs, and the periglacial landscape reinforce the concept of the landscape integrating numerous, small landforms, they also indicat e that this is indeed a visually distinctive landscape, created by the effects of periodic freezing and thawing, and is composed of distinctly periglacial features. 2.2 Geomorphological Mapping in Alaska and British Columbia Geomorphological mapping is t he cartographic representation of climate, soils, landforms, topography, geology, and other environmental factors (Smith et al. 2011). It is a representation of the terrain surface and environmental factors that define the geomorphologic processes shaping it. Conceptually, geomorphological mapping partitions terrain into distinct spatial units (Bishop et al. 2012). In this study, geomorphological mapping focus es on presenting field - collected data to show the individual forms that make up the periglacial ass emblage. 2.2.1 Study sites: Mapping and analysis for this chapter was conducted at the four sites shown in Figure 1.5 . Three sites were chosen to represent landforms along an east - west Transect 1 ) corresponds to the gradient of mean annual air temperature and continentality along the same path, (Nelson and Nyland 2017, Figure 3). The terrain at each of the sites is relatively similar, with flat - topped 34 ridges and steep - sided valleys. Ridges are fes tooned with small periglacial features. Hillslopes are in almost all cases slopes of transportation, and valley bottoms connect with fluvial networks. The fourth study site (Frost Ridge) is located in the Cathedral Massif near Atlin, British Columbia (Figu re 2.3C). In the 1970s Frost Ridge was described as a site of active nivation and other periglacial processes in a thesis by Nelson (1979a). Further study at this location allows for a comparison between this site, where active nivation processes are ongoi ng, and the dormant cryoplanation terrace sites in Alaska (Figures 2.2 A and B). Figure 2.1. Illustration of the periglacial assemblage : This figure represents the idealized the periglacial landscape and is redrawn from a diagram by Karte (1979) in his discussion of periglacial form communities. 35 Figure 2.2 . I dealized sketch of cryoplanation terraces : The terrace figure (redrawn from Reger 1975 fig. 17 p. 47) is a less detailed expression of the periglacial assemblage shown by Brunnschweiler (1965) and Karte (1979), but serves to show that this idea, while not specifically expressed, has been pervasive in the literature. Table 2.1 . Relative sizes, elevation , and orientation s of terraces at study sites Site Terrace Size (ha) Elevation (m.a.s.l.) Orientat ion Iron Creek 0.23 - 0.96 600 East Eagle Summit 0.77 - 8.86 1,113 West Mount Fairplay 3.01 - 3.98 1,689 North Frost Ridge 5.12 1,676 Predominantly North At each of the three sites in Alaska, cryoplanation terraces and smaller features are present, althoug h the r elative sizes of features and terraces vary between sites (Table 2.1). A wealth of periglacial features exists at the Iron Creek site and provides an excellent example of the ubiquity of small features in the periglacial landscape. At Eagle Summit, periglacial features are abundant, although an extensive mat of tundra vegetation obscures much of the terrace 36 treads. In the higher elevations at Mount Fairplay, multiple periglacial features not present at the other sites are found, such as tors and talu s slopes. In contrast, at high elevation, on the lee side of the coastal mountain range east of Juneau, Alaska, Frost Ridge is a is home to active periglacial processes and landforms . Numerous periglacial features, active nivation hollows, and late - lying s nowbanks can be found at this site. The similarities between Iron Creek and Frost Ridge are worth further exploration due to the climatic gradient and elevation difference. Terraces surveyed in this study ranged in orientation across the north, west, and east directions. Nelson (1998) performed directional statistical analysis of terraces identified by Reger (1975) and found they were predominately located in the northwest and northeast quadrants in interior Alaska. Orientation was more diffuse in western Alaska, possibly in response to a dominance of cloudy conditions near the Bering Land Bridge coast (Reger 1975; Guthrie 1982). 2.2.2 Mapping methods: Geomorphological mapping was conducted through both field - based procedures and digital terrain analysis. Field - based mapping was accomplished using traditional mapping and surveying methods. At all study sites, initial topographic surveys were conducted with a hand level and stadia rod. The topographic profiles were used to identify the locations of terraces, based on morphologic characteristics. Once identified, the terrace area was surveyed with handheld GPS. Scarp and tread lengths and angles were measured using a laser range finder and clinometer. Surface periglacial forms were identified, located with GPS , and dimensions were measured. A series of form communities was delineated on at least one terrace tread at each study site. Also noted were areas where debris was likely to flow off the 37 edge of treads, or topographic depressions in terrace sides. Several solifluction features on the side slopes were measured and geolocated. Figure 2.3 . Panoramic photos of study site s : A) Eagle Summit Terrace 10, looking west. Terrace tread is roughly 4 hectares. B) Mount Fairplay Terrace 25 and 24 looking north. The tot al area occupied by the terraces at Mount Fairplay is roughly 14 ha. C) Frost Ridge looking northeast. The tread at Frost Ridge is roughly 5 ha. Treads at Eagle Summit and Mount Fairplay are vegetated, but contain scattered periglacial features, namely sor ted stripes and nets. Frost Ridge is largely devoid of vegetation, except in the greenish - toned area in the upper left of the photo. Numerous sorted stripes follow the fall line across the tread. Snow surveys were conducted at Frost Ridge, where snowbanks remained on the landscape at the time of the study (late July, 2017). Snowbank perimeter and surface were surveyed with GPS and hand surveying methods (Figure 2.4D). Ground penetrating radar (GPR) was run along a north - south transect across the snowbank s urface (Figure 2.4A) to obtain a 38 subsurface topographic profile. At three locations on each snowbank snow pits were dug to check the accuracy of the GPR and to study the stratigraphy of the snowbank (Figure 2.4B). Data were compiled from multiple literatur e sources. Base maps showing surficial and bedrock geology were digitized from Wiltse et al. (1995) for Eagle Summit, Foster (1967) for Mount Fairplay, and Till et al., (2010) for Iron Creek. Geologic maps were superimposed with terraces and periglacial fe atures. The ArcticDEM used in this analysis was obtained from the Polar Geospatial Center at the University of Minnesota (Polar Geospatial Center 2017 ). Digital mapping was conducted using basic digital terrain analysis in ArcGIS and GRASS geospatial sof tware. GIS mapping at the Alaskan study sites was done using the Arctic DEM, a freely available two - meter resolution DEM derived from optical satellite imagery (Polar Geospatial Center 2017). Due to a crosshatch artifact in the ArcticDEM and the high varia bility of this 2 - meter resolution product, a 9x9 neighborhood low - pass filter was applied to the DEM before calculating the terrain derivatives of slope and aspect. Surface flow accumulation was also calculated on the DEM to identify topographic depression s and areas where material is likely to flow off terrace treads. No high - resolution DEM exists for the Frost Ridge study site, so a map by Cialek (1977) was used to provide elevation and topographic information. The final maps were compiled in ArcGIS usi ng both field - collected and DEM - derived data. Map symbols were modified from Demek (1972). This modification of symbols is reflective for large - scale mapping. Add reflective of current science. One attempt has been made at updating symbols for geomorphological mapping (Gustavsson et al. 2006), but the periglacial symbols were 39 secondary to that project and are not visually appealing. Additionally, the traditional method of geomorphological mapping created maps that were difficult to interpret, so maps in this thesis were created as a series that reflects the various elements (topography, geology, and landfor ms) of a geomorphological map. 2.2.3 Mapping r esults : Results from the geomorphological mappings indicate that there is both diversity and similarity in features between sites. Across all of the sites in Alaska the general topography remained relatively co nstant, with the landscape dominated by flat - topped ridges. Generally, the terraces surveyed in this thesis were ridgetop terraces (Reger 1975) that descend from a flat summit down into a valley. At Eagle Summit (Figure 2.5a), a series of five terraces wer e surveyed, from the summit to the lowest terrace, 100 meters below. At Mount Fairplay (Figure 2.6a), the terraces were separated by a large slope that divided the terrace series. No terrace existed on the summit of the mountain, but numerous terraces were identified on all ridge spurs from the summit. On the Seward Peninsula, Iron Creek, (Figure 2.6a), the terraces are ill formed, with a summit type terrace, followed by two terraces on a ridge. The valleys on either side of the ridge at Iron Creek are stee p sided, all with running water at the base. The side - slope topography at Frost Ridge (Figure 2.8) appears to have been affected by a valley glacier during the Wisconsin glacial period (Slupetzky and Krisai 2009 and references therein), although the top of Frost Ridge does have a large terrace, indicating a periglacial environment that exists in close proximity to glacial landscapes. On the north side of the Frost Ridge study site, terraces are smaller, with snowbanks obscuring much of the terrace. Topograp hic profiles done with GPR (Figure 2. 10 ) indicate that nivation hollows are forming under the snowbanks, incising themselves into the hillslope and creating terraced forms that 40 extend longitudinally across the slope subparallel to the contour. Nelson (per sonal communication, 2017) found cold (subzero) multi - year ice containing sediment bands in these snowbank positions in the mid - 1970s, implying geomorphic quiescence at that time, and indicating that periglacial processes have become more active under rece nt climate warming. Figure 2. 4. Study methods at Frost Ridge , BC : A) GPR survey of snowbanks. B) Measurements of snow temperature and density. C) Collection of suspended sediment at the downslope edge of a snowbank. D) Survey of snowbanks, surrounding top ographic depressions, and periglacial features. Geology varied between the sites. Observations of terraces from Eakin (1916), Mertie (1937), and Reger (1975) show that cryoplanation terraces are found on many bedrock types. Data collected verify those o bservations. In the east, at Mount Fairplay, the bedrock was primarily volcanic (Figure 2.6b), being dominated by mafic and felsic volcanic rock. At the Eagle 41 Summit study area in the central Yukon - Tanana Upland, bedrock is primarily quartzite (Figure 2.5b ). At the westernmost site (Iron Creek), bedrock is Ordovician schist (Figure 2.7b). Across the study sites, mapping of the periglacial assemblage showed the presence of many similar features, summarized in Table 2.2. Sorted stripes are important features appearing at all sites, in varying sizes. At Eagle Summit (Figure 2.5c and 2.5d), Mount Fairplay (Figure 2.6c and 2.6d), and Iron Creek (Figure 2.6c and 2.6d), sorted stripes can be found running from the scarp toward the edge and sides of the tread. At Fr ost Ridge, those stripes are observed at the base of snowbanks, with water moving through them. These stripes channelize the water away from the scarp, moving material with it and progressively eroding the base of the scarp. The stone stripes channelize su spended sediment over the sides through spillways. These spillways have solifluction lobes immediately downslope in almost all cases. The few cases where this is not true are where the spillway enters another stone stripe. Using a surface flow accumulati on algorithm derived from the ArcticDEM, the movement of water and suspended sediment across terrace treads was modeled. While there are limitations to the specific terrain modeling performed, it does help confirm field observations that indicate areas whe re material is transported off terrace treads. 42 Table 2.2 . Periglacial features found at study sites : Sites are arranged from west to east. An X represents the feature present at the site. This list is not necessarily comprehensive and is based on qu alitative observation of features. Feature Iron Creek Eagle Summit Mount Fairplay Frost Ridge Sorted stripes X X X X Frost Jacked Rocks X X Sorted Patterned Ground X X X X Frost Boils X X Rubble covered Scarps X X X X Solifluction Lobes X X X X Needle Ice Creep X Nivation Hollow X X X X Exposed Bedrock X X X ? Solifluction Terrace X 43 Figure 2.5a . Topography of Eagle Summit , AK : Inset map in lower left shows the larger area around the Eagle Summit study site, with areas where Reger ( 1975) delineated terraces outlined in black and white. The larger map in the upper right give topographic c ontours derived from the Arctic DEM (Polar Geospatial Center 2017). The b ackground color is an indication of the degree of slope . From this image, cr yoplanation terraces can easily be seen based on the slope, with low angle slopes representing treads and subsequent high angle slopes representing the scarps . 44 Figure 2.5b . Bedrock geology of Eagle Summit, AK : A dapted from Wiltse et al. 1995 , t his map sho ws uniform quartzite throughout the study site , however, consistent with observations by Reger (1975) the underlying geology varies between the sites described in this chapter. 45 Figure 2.5c . P eriglacial features at Eagle Summit , A K : The map in the lower left corner indicates locations of terraces mapped by Reger (1975) , showing that in the general area there are numerous CTs . In the upper right map, c ryoplanation terraces and other periglacial features have been identified and delineated. . On the lowest terrace, the tread was subdivided into a series of zones where similar periglacial features were found in high density. The upper treads show the location s of various periglacial features. Downslope from the CTs, the hillside is covered in solifluction fea tures. Grey areas are continuations of the flat - topped ridges and are covered with small periglacial features similar to those mapped on terraces. 46 Figure 2.5d . Subset of terraces at Eagle Summit, AK : Large - scale map of terraces 7 and 6 at the Eagle Summ it study site. This view of the terraces allows for closer inspection of the mapped periglacial features. A pattern begins to emerge on this terrace. Sorted stripes are apparent at the scarp - tread junction. With distance from the scarp, however, other patt erned ground features become more prominent, especially sorted circles and nets. 47 Figure 2.6a . Topography of Mount Fairplay, AK : Inset map in lower left shows the larger area around the Mount Fairplay study site, with areas where Reger (1975) delineated terraces outlined in black and white. The larger map in the upper right give topographic contours derived from the Arctic DEM (Polar Geospatial Center 2017). The background color is an indication of the degree of slope. From this image, cryoplanation terra ces can easily be seen based on the slope, with low angle slopes representing treads and subsequent high angle slopes representing the scarps 48 Figure 2.6b . Bedrock geology of Mount Fairplay, A K : A dapted from Foster ( 1963 ) , this map shows that b edrock at th is site is mostly volcanic, with Quaternary sediment in the valleys, indicating the location of eroded debris. The terraces identified and discussed in this section are found on mafic and felsic volcanic rock. 49 Figure 2.6c . P eriglacial features at Mount Fairplay, AK : The map in the lower left corner indicates locations of terraces mapped by Reger (1975), showing that in the general area there are numerous CTs. In the upper right map, cryoplanation terraces and other periglacial features have been identifie d and delineated. On one of the lower terrace s , the tread was subdivided into a series of zones where similar periglacial features were found in high density. The upper treads show the location of various periglacial features. Downslope from the CTs, the h illside is covered in solifluction features. Grey areas are continuations of the flat - topped ridges and are covered with small periglacial features similar to those mapped on terraces. 50 Figure 2.6d . Subset of terraces at Mount Fairplay, AK : Large - scale ma p of terraces 24 and 25 at the Mount Fairplay study site. This view of the terraces allows for closer inspection of the mapped periglacial features. A similar pattern to the one observed at Eagle Summit can be seen here, with sorted stripes at the scarp tr ead junction and other patterned ground features (sorted circles and nets) farther out. 51 Figure 2.7a . Topography of Iron Creek, AK : Inset map in lower left shows the larger area around the Iron Creek with delineated terraces outlined in black and white. The larger map in the upper right displays topographic contours derived from the Arctic DEM (Polar Geospatial Center 2017). The background color is an indication of the degree of slope. From this image, cryoplanation terraces can easily be seen based on th e slope, with low angle slopes representing treads and subsequent high angle slopes representing the scarps. 52 Figure 2.7b . Surficial geology of Iron Creek area, A K : The map is adapted from Till et al. (2011). Bedrock at this site is uniformly schist. 53 Fi gure 2.7c . P eriglacial features at Iron Creek, A K : Cryoplanation terraces are outlined with individual features noted on terrace surface. Side slopes are covered in solifluction features. Grey areas are continuations of the flat - topped ridges and are covere d with small periglacial features similar to those mapped on terraces. 54 Figure 2.7d . Subset of terraces at Iron Creek, AK : This view of the terraces allows for closer inspection of the mapped periglacial features. The same pattern observed at Mount Fairpla y and Eagle Summit was observed at Iron Creek, with sorted stripes at the scarp - tread junction and other patterned ground features (sorted circles and nets) away from the scarp. 55 Figure 2.8 . Geomorphological map of Frost Ridge, BC : One terrace has been d elineated on top of the ridge, with incipient terraces and late lying snowbanks have been noted. The ridge is a periglacial site between two glacial valleys. The numbers next to snow banks correspond with GPR profiles in Figure 2.8. 56 Figure 2.9. Bedrock g eology of Frost Ridge, BC : The study areas is uniformly sedimentary rock. Geologic data adapted from the Atlas of Canada (1958) . Measurements of sorted stripes were undertaken at Frost Ridge, Eagle Summit, and Mount Fairplay. Table 2.3 shows the results of these measurements, allowing for comparison of the stripe widths. As the slope leveled out, the stripes give way to other forms of sorted 57 patterned ground, primarily circles and nets. A calculation of slopes from the ArcticDEM on terrace treads at Eagle S ummit and Mount Fairplay show a series of circular shapes that are interpreted to be sorted patterned ground ( Figure 2.11 ). Soil pits dug across the terrace treads provide further evidence for the interpretation that patterned ground features exist across the terrace. Clasts exhibiting varying degrees of weathering were found in the majority of soil pits. Together, these findings make the case that the terraces have been forming for long periods of time and help to link terraces in Alaska with those studied at the more active Frost Ridge site. Table 2.3 . Comparison of sorted stripes at Frost Ridge, Mount Fairplay, and Iron Cree k : All measurements are in meters. Stone stripes were not measured at Eagle Summit but appear to be of similar magnitude to those at Mount Fairplay. Frost Ridge Mount Fairplay Iron Creek Mean (m) 1.40 3.43 1.81 Min (m) 0.50 2.30 0.90 Max (m) 4.40 4.80 3.50 Between treads, some similar features were noted. All terrace scarps are covered with angular clastic rubble, apparently piece s of frost - shattered bedrock derived from bedrock outcrops. At the base of these scarps, where the slope is reduced, the rubble is in many places channelized into sorted stripes. In some places, the stripes grade into sorted nets and garlands. Farther fr om the scarps, where the slope is reduced even more, the sorted stripes become sorted polygons. Nelson (1979a) reported that on Frost Ridge water continued to flow through the coarse sections of sorted circles, probably continuing to carry sediment. In the middle of the terraces at Eagle Summit and Mount Fairplay, patterned ground gives way to wet tundra vegetation, at least at the surface. At the far side of the terrace, the wet tundra becomes smaller patterned ground features and then gravel pavement, whi ch exists on the 58 outer edge of the terrace treads. Surrounding the terraces on all sides, except those that lead to another terrace, are slopes of transportation with solifluction features extending to streams and then rivers in valley bottoms (Figure 2.5a , 2.6a, 2.7a). The middle sections of terraces exhibited differences, depending on the morphology and vegetation cover. Terraces at Mount Fairplay have frost boils and wet tundra areas in the mid - terrace region. At the Eagle Summit site, the center is a mo saic of tundra vegetation and sorted polygons. Both sites have areas of saturated peat with a high frost table (~30cm) in early July. The wetness of these peat areas is attributable to their occupation of low - lying areas on the terrace tread. The periglac ial features identified on the terraces are similar at the different sites (Table 2.2), although their morphology differs. Features at the Eagle Summit and Mount Fairplay sites appear to be dormant . Evidence for contemporary geomorphic dormancy includes th e high percentage of clasts with nearly complete lichen cover ( Figure 2.12 A) and nearly complete vegetation cover on many terrace treads. These observations appear to be consistent with the literature (e.g., Reger 1975; Reger and Péwé 1976) , although Reger (1975) and Reger and Péwé (1976) describe relict rather than dormant cryoplanation terraces in Alaska. However, contrary to the conclusions of Reger and Péwé (1976), these terraces are not relict, but in a semi - dormant state where the nivation processes a re not as strong and the moisture supply is not as high as during the LGM. Cursory evidence does suggest that freeze - thaw related processes and landforms are still operating on the terraces. One metric indicating that there are still limited periglacial pr ocesses acting upon these sites can be indicated by counts of recent rock fracture. Counts of rock clasts at three sites indicate that roughly 2% of rocks were fractured. Frost boils 59 are present and active at the Eagle Summit, Mount Fairplay, and Iron Cree k sites. At Iron Creek, the features appear only slightly more active. Figure 2. 10 . Profile of snowbanks at Frost Ridge , BC : Positions of these snowbanks are indicated in Figure 2.8. The dark grey line is the surface of the snowbank, the dashed white line is the surface of the ground below the snowbank. The white line flattens out beneath the snowbank indicating these are actively eroding and have the potential to become terraces over time. Figure courtesy of K. Nyland 2017 . Solifluction appears to be acti ve at both the Eagle Summit and Mount Fairplay sites. Although this study did not include instrumental observations of solifluction movement, evidence of downslope movement of material was observed at the Mount Fairplay site and repeat observations at the Eagle Summit site show burst solifluction lobes that must have involved rapid movement of large volumes of sediment. At Mount Fairplay, evidence of moving 60 water was seen at the scarp - tread junction and digital terrain analysis modeling of surface hydrologi c flow on the tread agrees with field observations of water moving towards the sides of terraces and into solifluction lobes on the terrace sides. Clasts within nivation hollows near the Eagle Summit site, where snow remained into early July, had less lich en cover and the snow supplied moisture to the area. Nelson ( p ersonal c ommunication, 2017) observed that snow remained well into August at these locations in the early 1980s. Similarly, at Frost Ridge, clasts and material in the sorted stripes were largely free from lichen ( Figure 2.12 b) and appeared to be subject to significant activity by periglacial processes, although the features have undergone some settlement since the mid - 1970s (Nelson personal communication 2017). At Frost Ridge, the sorted stripes were observed through a crude sediment trap to be channeling water, a phenomenon also noted by Nelson (1979a) at the same location. Figure 2.11 . Pattered ground from slope at Eagle Summit, AK : The results of DEM (Arctic DEM) derived slope calculations at Eagle Summit study site, terraces 6 and 7 are outlined in red. The white is interpreted to be patterned ground features that have since been buried by vegetation. 61 Figure 2.12 . Comparison of geomorphically active and inactive rock clasts : a) Lichen covered rock clasts in a stone stripe at the base of the scarp between Eagle Summit terrace 7 and terrace 6. b) lichen free sorted stripe at Frost Ridge. This sorted stripe appears to be active, with water running through its base. 2.2.4 The facies concept an d the periglacial landscape: The concept of periglacial facies is not new to the periglacial literature, having been initially proposed by Lozinski (1909) and recently mentioned by French (2017, pp. 3). A more extensive exploration can be found in Richmond (1962) and Morris (1981). However, it does appear that the linkage between the periglacial facies concept and the periglacial landscape has been extended only as far as block fields and rock glaciers. This thesis proposes expanding upon the facies concept to include periglacial forms that, pursuant to Madole (1972), can give indication of environmental conditions. Conversely, by pursuing an understanding of these facies and their depositional environments, the conditions necessary for the formation of peri glacial landforms can be deduced. Given what is known about cryoplanation terraces and the environment of Beringia during the Last Glacial Maximum, as well as the formation of small periglacial landforms, a conceptual representation of the periglacial lan dscape can be created. Morris and Olyphant (1990) conceptualized a physical model that relates topoclimatic factors to the deposition of 62 facies. A similar model is appropriate here ( Figure 1.4 ). This model would link the periglacial landscape to surface fe atures by providing an understanding of the environmental factors necessary for the formation of various periglacial forms, including cryoplanation terraces. Climate is the driving force behind the periglacial environment (Washburn 1980, 2 ), but local fact ors play a major role in the manifestation of various landforms (Morris and Olyphant 1990). These local factors, or filters, include, topography, vegetation, and lithology. By providing minute adjustments to the climate, the filters will produce just enoug h variation to alter the physical manifestation of the landscape. In the model shown in Figure 1.4 , the filters yield three possible forms, glaciers, cryoplanation terraces, and surface periglacial landforms. It has been shown that CTs occupy the same clim ate space as glacial cirques (Nelson 1989; Nelson and Nyland 2017) and thus some simple factors might affect the mass balance of snow and thereby cause the formation of vastly different landforms. If, for example, snowbank orientation is such that the snow banks melt before the end of the year most years, nivation will occur and CTs will form. If snowbanks are not able to remain on the landscape for longer than in the immediate surroundings due to less favorable topographic position or a combination of aspec t and lithology, smaller surface features will likely form instead . Evidence for this depositional model can be seen at Eagle Summit, within close proximity to the study site, CTs, a glacial cirque, and small features are all present. A map by Péwé, Burban k, and Mayo (1967) indicates that a glacial cirque is present on the east side of the study site ridge; if this interpretation is correct, its form has been modified significantly in postglacial time. On the west and north sides, cryoplanation terraces are found, and among both of these features surface periglacial features 63 have been documented. The features likely formed after the others, owing to changes in environmental conditions. 2.3 Conclusion This chapter examines the spatial location and arrangemen t of periglacial features on the Beringian landscape through detailed mapping . By formalizing the idea of form communities in the periglacial landscape, the patterns of observed features become a larger assemblage, rather than a series of small, essentiall y disconnected, features. Large - scale mapping of periglacial sites across Alaska indicate that this landscape is indeed a system of interconnected parts that, together as a periglacial assemblage , constitute s a distinct landscape. At all sites, a series of form communities creates a repetitive pattern. The similarities between sites that range in elevation from 600 to 1600 m.a.s.l. and are separated by distances of over 1000 km show that this landscape is not unique to one site, but can be found in many Ber ingian uplands, which are in large part a periglacial domain. This pattern can be seen as a series of small, micro - forms that embroider the larger meso - scale features, CTs and cryopediments. The distinct nature of some of the micro - forms and the processes , which are connected to the formation of the larger forms, comprise a distinct and characteristic landscape. The mapping efforts described above solidify the notion that cryoplanation terraces are the foundation of the periglacial landscape. It is thro ugh this large - scale mapping of terraces that the identification of the relationship between small - and large - scale periglacial features is shown. The identified relationship between the unquestionably periglacial features and larger slope - scale features p rovides a description of the periglacial landscape. This shift in thinking will 64 allow for a critique of French (2016) and André (2003) that relies on the periglacial attributes of the small features at a landscape scale. Owing to their relatively small si ze, the periglacial microforms (e.g., sorted patterned ground, frost boils, frost - jacked clasts) mapped in this study are difficult to discern on most air photographs. Field mapping is laborious and time - consuming, and only relatively small areas can be ma pped without investment in a major field campaign. In some cases (e.g., Evans 2017), high - resolution air photos facilitate mapping of more extensive areas than those covered here. Progress in this area of research is most likely to come, however, from use of unmanned aerial vehicles or UA S s (Whitehead and Hugenholtz 2014), and from small - satellite technology (Kulu 2018; Mascaro 2018), which can provide resolution of as little as 0.3 m. 65 Chapter 3 Specific Periglacial Geomorphometry This chapter focuse s pr imarily upon understanding the signature of periglacial processes on the landscape and predicting the locations of specific forms (CTs) using semi - automated methods. In the previous chapter, it was concluded that there is indeed a distinct landscape compos ed of a repeating series of interconnected forms and sediments, an assemblage of forms, that makes the periglacial landscape distinctive. This chapter uses specific geomorphometric methods to both recognize and delineate cryoplanation terraces and to demon strate their ubiquity and distinctive imprint on the landscape. Through a multi - method and multi - scale approach, a semi - automated identification sequence will provide a method for distinguishing the periglacial landscape. 3.1 Review of Geomorphometry Geom orphometry is the quantitative study of the landscape through the intersection of geomorphology, mathematics, and computer science (Pike et al. 2009). By defining landscape attributes in a mathematical construct, the landscape is demarcated through a series of geometric shapes. Landforms have a specific geometry and the conglomeration of these landforms together creates a landscape that varies in geometry over space. The study of geomorphometry has roots in basic cartographic and spatial analysis. However, t he subject has increased in popularity with the availability of high - capacity et al. 2009). Computer - aided spatial analytics has also greatly increased the speed and accuracy of applying geomor phometric methods, making this technique more commonplace. As the methods used 66 in geomorphometry have evolved, the predominant use has been to analyze and understand fluvial landscapes. A smaller subsection of the literature on geomorphometry has been appl ied to glacial landscapes. Even less has been done to apply geomorphometric techniques to periglacial landscapes. 3.1.1 Evolution of geomorphometry: The study of geomorphometry is interwoven with the subjects of terrain analysis and terrain modeling (Pike 2002). Early studies in geomorphometry were conducted using terrain data collected by pioneers of geography, such as Alexander Von Humboldt (Pike 2002). Methods of analysis have evolved over time with the increased accuracy of topographic maps through to t he use of computers to perform landscape analysis (Pike et al. 2009). As of 2002, over 4 , 000 publications related to geomorphometry had been written, beginning in the nineteenth century. These publications span many disciplines, demonstrating that the subj ect is of considerable importance to many scientists (Pike 2002). With the quantitative revolution in geomorphology and its shift from primarily geography to geology (Church 2005), the nature of landforms became an important part of geomorphologic study. A review of land form geography (Zakrzewska 1967) built upon an earlier review of morphometric studies (Hammond 1964) to broaden the scope of inquiry from those of basic morphologic parameters such as slope, local relief, and texture. Through these basic parameters, various landscape indices were developed, forming the foundation of geomorphometry. Wood and Snell (1960) pioneered this landform analysis using their terrain index, which was furthered by Pike (1988), who identified a series of landscape para meters and used them to predict a geomorphic signature. Further work has sought to identify ways in which simple measures such as slope and relief can be put together in an automated procedure 67 to classify landforms (Dikau et al. 1991, 1995). Together, the identification of landforms constitutes a primary subfield of geomorphometry. A major shift in the field of geomorphometry came when Evans (1972, 1987) subdivided the subject further into specific and general geomorphometry. Specific geomorphometry is con cerned with discreet or individual landforms, while general geomorphometry focuses on the geomorphic landscape, or continuous land surface. This subdivision of the two types of study allows for a multiscale approach to understanding the landscape. Modern g eomorphometry relies on digital elevation models (DEMs) for elevation information. DEMs are gridded matrices with each node (pixel) representing a specific elevation. Another version of an elevation model is a digital surface model (DSM). The primary diffe rence between a DSM and a DEM is that a DSM represents the surface of the earth and will often include vegetation, buildings, or other surface covering, while DEMs are bare - earth model s of the land surface (Bishop and Shroder 2004). 3.1.2 Terrain analysi s: With the increased use of DEMs in geomorphometry, it is important to understand the limitations of DEMs. A DEM is a discrete dataset that segments the elevations of the study area into a regularly spaced series of nodes, the locations of which are arbit rary with respect to the topographic surface they represent (Mark 1975). The discrete division between neighboring pixels poses a problem in modeling the continuous land surface. To solve the problem of continuous data in a discrete space, several techniqu es have been proposed that allow for the derivation of landscape attributes from these discrete models. For example, slope, the first vertical derivative of terrain, can be linked with the concept that the landscape represents a function. The derivative of a function is its slope at a point (Sharpnack 68 and Akin 1969). On a DEM, each grid node represents an elevation value, or point, and should have a corresponding slope value. The second vertical derivative of terrain, profile curvature, can again be traced to the mathematical relationship between a function and its second derivative, the rate of change of slope. Through consideration of terrain by these derivatives it can be computationally manipulated, allowing for highly sophisticated quantitative analysis and comparison. 3.2 Introduction to Periglacial Geomorphometry Periglacial geomorphometry is proposed here as a subdiscipline of geomorphometry that has not been extensively explored in the literature. An extensive collection of work addressing geomorph ometry in cold regions has focused on identifying glacial cirques (e.g., Eisank et al. 2010; Eisank, Smith, and Hiller 2014; Wagner 2018). Concepts related to p eriglacial geomorphometry w ere touched upon briefly in an unpublished study by Nelson (1979b), w hich provides a description of the need for such a subject, including an application to cryoplanated terrain. Further research proposed by Nelson (1979b) relates to the periglacial landscape, particularly cryoplanated uplands, and states that these lands capes should have a distinct hypsometric signature that allows it to be distinguished from other landscapes, a topic featured prominently in Chapter 4 of this thesis. geomorphometric analysis of cryoplanation terraces (CTs) using directional statistics (e.g., Mardia 1972 ; Fisher 1993 ) . This work was analogous to work in glacial geomorphometry pioneered by Evans (1977). Additional geomorphometric analysis by Nelson and Nyland (2017) demonstrated that cryoplanation terraces exhibit elevational trends similar to those of glacial 69 cirques, indicating that cryoplanation terraces are climatically controlled and thus useful in understanding the origins of periglacial landscape s. 3.2.1 Periglacial geomorphometry applied to the Beringian landscape: This research applies the concepts and lessons from the broader subject of geomorphometry to the periglacial environment of eastern Beringia. The subdiscipline of periglacial geomorph ometry will thus have the goal of understanding the periglacial landscape through analysis of the character and form of cold - climate, nonglacial landforms. Periglacial geomorphometry will be discussed in the context of cryoplanation terraces and how they f orm the fundamental underpinning of the upland periglacial landscape. The analysis undertaken in this thesis uses a multiscale approach that incorporates both specific (this chapter) and general (Chapter 4) geomorphometry. 3.3 Specific Periglacial Geomorp hometry of Cryoplanation Terraces Chapter 2 concluded that cryoplanation terraces can be regarded as the foundational building blocks of upland periglacial landscapes. It follows that no investigation into the geomorphometric signature of the upland perigl acial landscape would be complete without specific geomorphometric analysis of cryoplanation terraces. This chapter discusses the application of multiple geomorphometric methods to individual and series of cryoplanation terraces at the previously discussed sites of Mount Fairplay and Eagle Summit ( Figure 1.5 ) along with several sites mapped by Reger (1975) and previously glaciated sites mapped by Péwé, Burbank, and Mayo (1967) . It is well understood that defining a landform based on morphometric parameter s cannot be done without some degree of subjectivity (Pike 2001). With respect to cold - climate 70 landforms, Evans and Cox (1974) discussed the difficulties involved in achieving an operational definition for glacial cirques, a relatively simple geomorphic fe ature. Wide variation in cirque morphology occurs in glaciated uplands, and it is virtually impossible to avoid at least some subjectivity in conducting inventories of these features. Operational definitions of cryoplanation terraces face similar challenge s. Despite a wide range of CT morphological attributes and size, the definition of a cryoplanation terrace has been addressed primarily through qualitative verbal description, sometimes augmented by semi - quantitative measures, such as ranges of scarp and t read dimensions, slope angles, and topographic position (e.g., Demek 1969; Reger 1975). Increasing availability of high - resolution representations of the land surface facilitates the creation of quantitative definitions of landforms. These, in turn, can be used through geomorphometric analysis to further the goals of geomorphology (Evans 2012). This chapter makes a first attempt at developing an objective, semi - automated method for recognizing and delimiting cryoplanation terraces using high - resolution digi tal elevation models. A series of terrain attributes were calculated in an effort to identify which would be most effective for the detection of terraces. First, slope and aspect algorithms were applied to the study areas. Second, tangential and profile cu rvature were calculated for the study sites. Third, three low - pass filters were applied to the study sites and the previous two methods were applied to these smoothed DEMs. Fourth, the elevation - relief ratio (Pike and Wilson 1971) was calculated based on a grid over the study sites. Finally, all of these parameters were used collectively to create a semi - automated method for detecting the locations of CTs in the landscape. 71 3. 3.1 Slope and aspect calculations: The calculation of slope and aspect from a DE M is common and many methods exist to determine these attributes. In a basic sense, slope is the rate of change in elevation from point A to point B and can be calculated by dividing rise by the run. Given a continuous function, slope becomes the first ver tical derivative of that function. In a raster environment, such as a DEM, where data comprise a discrete grid of values, defining a continuous function is not possible. This means that the slope is dependent on the resolution of the DEM. To work around th e lack of continuity, various algorithms have been developed, one of the more common being the Horn (1981) algorithm. Horn (1981) takes the eight cells surrounding the target raster cell and averages the east - west and north - south gradients (Equation 3.1). These averages are used to designate a value for each raster pixel that is indicative of the rate of change in elevation between the surrounding pixels. Aspect (orientation) is, in the simplest terms, the direction faced by the slope. For a continuous fun ction, aspect is the first horizontal derivative of elevation. The calculation of aspect commonly uses the same eight neighboring cells used in calculating slope. In this case, the algorithm identifies the change in both the horizontal and vertical directi ons. A straightforward method for calculating aspect is given in Equation 3.2 (Olaya 2009, 146). Using the algorithms discussed above and the Arctic DEM (Polar Geospatial Center 2017), slope (Figure 2.4a; 2.12a) and aspect maps were calculated for each of the study sites. A visual inspection of these maps shows significant areas of nearly flat (<5 °) slopes. These flat areas are primarily cryoplanation terrace treads or river valley bottoms. Owing to irregularities in the DEM and the fine resolution of the d ata, slope maps at 2 m resolution are too variable for 72 this analysis, and a different method for performing these calculations was developed , as described below . Equation 3.1 . Equation 3.2 . 3.3.2 T angential and profile curvature calculations: Curvature is the second derivative of elevation, or the rate of change of slope. Slopes can be flat, concave, or convex. To calculate curvature, it is necessary to specify the direction for this calculation. Pr ofile curvature is the rate of change from the bottom to the top of the slope (Equation 3.3). Tangential curvature is equivalent to plan curvature, except in that it adds a slope as a factor. It has been shown that A B C D E F G H I 73 tangential curvature is a better measure in specific geomorphometry (Olaya 2009). Tangential (plan) curvature is simply the rate of change from side to side (Equation 3.4). Following Olaya (2009) , the equations used in a GIS environment to calculate profile and tangential curvature are shown in E quations 3.3 and 3.4 respectively. Equation 3.3 . Equation 3.4 . As with slope, the use of profile and tangential curvature was limited when applied to the 2 - meter Arctic DEM, as curvature is highly variable. Inspection of the curvature maps reveals that they indicate locations of cryoplan ation terraces. However, analysis of these locations proved to be unnecessarily complicated. 3.3.3 Low - pass filters: The Arctic DEM was created using automated photogrammetry software, the Surface Extraction from TIN - based Searchspace Minimalization (SE TSM) (Polar Geospatial Center 2017). The SETSM software extracts information from optical satellite imagery. Although the software provides a visually appealing product, a series of crosshatch artifacts appear in the data. Noh and Howat (2015) demonstrated that the SETSM algorithm 74 was highly accurate in both horizontal and vertical directions, although they failed to recognize the artifacts produced by this method. The crosshatch artifact visible in the ArcticDEM (Polar Geospatial Center 2017) inhibits cert ain geomorphometric calculations, including the accurate identification of cryoplanation terraces. To reduce the artifact, a low - pass filter was run on the DEM. A low - pass filter is a neighborhood (kernel) operation that takes the average of neighboring c ells to fill in the target cell. For example, using a 3 X 3 neighborhood, the filter would take the three neighboring cells in either direction and average them to replace the target cell value. For this project, 3 X 3, 9 X 9, 21 X 21, and 75 X 75 neighbor hoods were used. Some filters were applied to excessively smooth out the data for a specific purpose (i.e., the 75 X 75 filter), others were compared for optimal smoothing to remove artifacts and variability. Once low - pass filters were applied to the DEM, slope and aspect algorithms were run on the smoothed DEM. From these new slope maps various geomorphometric analyses were conducted. 3.3.4 Elevation - relief ratio: Strahler (1952) developed the hypsometric integral (HI) as a method to classify and compare method of calculation involves planimetry from maps, and is both laborious and time - identical measure, th e elevation - relief ratio (ERR) (Equation 3.5), which is simple to calculate on DEMs (Pike and Wilson 1971). With advances in computer - aided analysis, the ERR has been used frequently in geomorphology and allows for comparison of landscapes, often at a wate rshed scale. This method extends beyond the comparison of watersheds and can be used to predict features on the landscape. Pérez - Peña et al. (2009) developed a method that used 75 the ERR to predict the location of fault scarps. By resampling the DEM and appl ying the ERR to each grid node, the location of steep scarps becomes apparent. A similar method was applied to study sites in this chapter as a means to identify the locations of terrace scarps. Equation 3.5 . 3.3.5 Semi - automat ed identification of cryoplanation terraces (CT AR) : Descriptive studies of cryoplanation terraces have identified basic characteristics of terra ces that can provide a baseline for their identification. Demek (1969) and Reger (1975), the two major reviews of the cryoplanation landform literature, are in agreement that terrace treads range in slope from 0 o to 5 o . Adjacent to terrace treads, scarps r ise at angles of 9 to 32 o , providing a distinct contrast between low - and high - angle slopes. This morphological description creates the primary parameters for CT recognition (identification) . Using rasters derived from the ArcticDEM and described in the pr eceding sections, this semi - automated method ( CTAR ) is shown as a flow diagram in Figure 3.1. The procedure uses both a 75 x 75 neighbor average kernel and a 21 x 21 neighbor average kernel to create low - pass filters in GRASS GIS (GRASS Development Team 20 18). The 75 x 75 filtered DEM is then used to calculate flow accumulation using the GRASS r.watershed command (Gruber and Peckham 2009). The absolute value of the output flow accumulation is taken and accumulations of more than 100 pixels are filtered out. By filtering out the higher flow accumulations, only the ridges are being used in subsequent calculations. 76 To address the high variability of the Arctic DEM (Polar Geospatial Center 2017), slope was calculated on the 21 x 21 neighborhood raster, which gr eatly reduce d variation in the DEM. The slope raster wa s then combined in an if - then statement with the flow accumulation raster to recognize CTs. In this study, a threshold value of 5 o slopes, based on the observations of Demek (1969) and Reger (1975), wa Fairplay, Indian Mountain, Kwethluk River, and Sittookooyook River study sites, each of which contains numerous CTs (Figure 3.2). Figure 3.1 . Flow diagram of the methods for the Semi - Automated I dentification of CTs : Darker boxes are output rasters, lighter boxes are the operations performed , and descriptions of these operations can be seen on the sides . The use of 75 x 75 and 21 x 21 low pass filters w as chose n because they provide the optimal res olution s to perform the various methods outlined here. The procedure was created and run in GRASS GIS software. 77 3.4 Results from S emi - automated I dentification of C ryoplanation T erraces In the initial exploratory analysis of cryoplanation terraces various terrain parameters were calculated and filtered out to find the most appropriate measure. The calculation of slope allowed for visual identification of cryoplanation terraces (Figure 3.3). Gentler slopes are associated with valley bottoms, summit flats, a nd cryoplanation terraces. The steeper slopes coincided with valley walls and terrace scarps. CTs were easily recognized on the low - angle slopes and were used in subsequent calculations. The calculation of aspect, however, did not provide any additional in formation on the location of cryoplanation terraces and was not used for identification. Figure 3.2 . Map of sites where the CTAR was teste d : Each site was previously mapped by Reger (1975) and used to test the accuracy of the identification sequence. Als o noted on the map is the extent of Pleistocene Glaciers (light grey) and the Bering Land Bridge (dark grey) adapted from Brubaker et al. (2005). 78 When the terrain derivative of curvature was calculated, visual identification of terraces was possible in bo th the profile and tangential curvature maps. Profile curvature resulted in the identification of scarps (Figure 3.4), while tangential curvature showed the nearly flat treads of the terraces (Figure 3.5). CTs were easily recognized, but also misidentified other landforms (i.e. , solifluction lobes and nivation hollows) when curvature was used in the semi - automated method for CT identification . Figure 3.3 . Map of slopes at Eagle Summit , AK : Computation of slope provid e s an example the visual impression g iven by the slope maps created for both Eagle Summit and Mount Fairplay. Flat - topped ridges and terraces are highlighted in lighter tones; scarps and valley walls are darker. Coordinates for this map are UTM Zone 6N. 79 Figure 3.4. Profile curvature map o f Eagle Summit , AK : Red values are concave, blue values are convex. Terraces and summit flats (indicated by grey arrows) are immediately apparent in this image as lighter blues with scarps visible as dark blues and reds. Coordinates on this map are given as UTM Zone 6N. 80 Figure 3 . 5. Tangential curvature map of Eagle Summit , AK : Red values are concave, blue values are convex. Terraces and summit flats (indicated by grey arrows) are immediately apparent in this image because of the clustering of flat to sligh tly convex slopes (light blue tints) . Coordinates on this map are given as UTM Zone 6N. 81 Smoothing the DEM before performing calculations allowed for better generalization of terrain derivatives. This generalization helped to visualize the terrain and to p erform terrain analysis. An optimal neighborhood size was identified from five different resolution filters by performing the identification sequence with each one and comparing the result. The optimal resolution varied depending on the calculation perform ed. Slope was calculated using a 21 x 21 low - pass filter and the watershed analysis was done using a 75 x 75 low - pass filter. Computation of the ERR using a gridded approach helped to identify terrace scarps (Figure 3.6) , although mathematically it was n ot ideal for identification of terraces. The large variation caused by the scale of the grid and the scale of the DEM inhibited the utility of this method. While performing additional analysis on the ERR raster would likely have yielded a favorable result, this method was more complicated than was deemed necessary. The goal of this analysis is to provide a method that can be applied to a DEM and will identify the location s of CTs. This identification / recognition procedure ( CTAR ) was applied to sites identi fied by Reger (1975): Eagle Summit, Mount Fairplay, Indian Mountain, Kwethluk River, and Sittookooyook River (Figure 3.1). The first two locations were visited in the field and large - scale mapping was conducted on terraces in that location (Chapter 2). Al l of the sites named in the previous paragraph were originally mapped by Reger (1975), who used 1:63,360 scale USGS topographic quads to discern areas in which CTs may be common. Reger, who is a highly accomplished photo interpreter (e.g., Krieg and Reger 1976, 1982) then used the best available imagery to delineate individual terraces on the quadrangles. The fact that CTs are recognizable on such low - resolution maps demonstrates their prominence in the Beringian landscape. 82 Figure 3.6 . Map of ERR at Eagle Summit, AK : This m ap shows the elevation relief ratio calculated from a 5 m grid. Lighter areas have a higher ERR, dark areas have a lower ERR. Cryoplanation terraces can be seen as areas of non - uniform ERR values. Coordinates around ERR map are from UTM Z one 6N. 83 The scale used by Reger was much coarser than that used in the present study. Indian Mountain is the only exception to this scale restriction, as Reger extensively mapped terraces in the field at that site, using a topographic base map prepared e xclusively for his project. As expected, more terraces were identified using CTAR than were mapped by Reger (1975) (Table 3.1). It is important to note that the semi - automated method, while performing well in most cases, did misidentify terraces in some lo cations, particularly at the Sittookooyook River site. The high degree of error seen at the Si tt ookooyook site could be an artifact of holes in the DEM at that location. Because of the data gap, the low - pass filter values may be artificially lower and flow accumulation will not have the upstream contributing area necessary for proper analysis. In addition, the nature of the terrain at this site, which contains relatively large low - lying areas, creates false positives. Comparison of results from the CTAR me means to check the effectiveness of the new method. Table 3.1 summarizes the accuracy of this were found. The method for i dentifying CTs worked well at Eagle Summit and Mount Fairplay. At Sittookooyook River, Indian Mountain, and Kwethluk River, the sequence had to be adjusted slightly. Due to large low - lying areas adjacent to the study sites, an additional filter was necessa ry. At these sites, all areas that were less than the mean elevation were filtered out to exclude the flat, low - lying regions. 84 Figure 3 . 7 . CTAR accuracy assessment at Eagle Summit, AK : Comparison of the semi - automated identification of cryoplanation terraces developed in this study to the terraces identified by Reger (1975) at Eagle Summit. Coordinates in this map are measured in meters from UTM Zone 6N. Note that terraces identified by CTAR consistently fall within the bounds of CTs delineated by Reg er (1975). 85 Figure 3 . 8. CTAR accuracy assessment at Mount Fairplay, AK : Comparison of semi - automated method developed in this study to the cryoplanation terraces identified by Reger (1975) at Mount Fairplay. Coordinates are UTM Zone 7N. 86 Figure 3 . 9. CTAR accuracy assessment at Kwethluk River, AK : Comparison of CTs at Kwethluk River identified using the semi - automated approach developed in this study compared with those mapped by Reger (1975). Coordinates are in UTM Zone 4N. 87 Figure 3.10. CTAR accuracy asse ssment at Sittookooyook River, AK : Comparison of CTs at Sittookooyook River identified using the semi - automated approach developed in this study compared to those mapped by Reger (1975). The semi - automated approach to finding CTs was less accurate than expe cted at this site, with relatively large, low, flat areas identified and large holes in the data. Coordinates are UTM Zone 3N. Table 3.1 . Summary of accuracy assessment for CTAR : T he number of terraces identified by the algorithm compared to the number of t erraces identified by Reger (1975). The CTAR identified more terraces than Reger, again underscoring the ubiquity of these features. Site Name Total Number of Terraces Identified Number of Terraces Identified by Reger Number of Identifie d Percentage Identified Eagle Summit 47 29 29* 100 Mount Fairplay 50 25 25* 100 Kwethluk River 166 146 131 90 Sittookooyook River 86 71 58 82 Indian Mountain 74 88 66 75 * Size of these sites is smaller than the area mapped by Reger (1975) due to lac k of full coverage by the Arctic DEM. 88 Figure 3.11. CTAR accuracy assessment at Indian Mountain, AK : Comparison of CTs at Indian Mountain identified using the semi - automated approach developed in this study compared to those mapped by Reger (1975). The h igh resolution base map used by Reger (1975) at this site makes it an important test site for the CTAR method. Coordinates are UTM Zone 5N. 89 3.5 Conclusion The Semi - Automated Identification of CTs ( CTAR ) uses basic terrain parameters to identify the locati on of cryoplanation terraces on the landscape. In a multistep process whereby, certain areas are identified based on morphological characteristics, the CTAR filters out areas that are in valleys and only searches ridgetops or mountain summits for flat area s, i.e., areas with slopes less than 5 o . Because CTs are found in upland periglacial environments, some locations require that low - lying areas be filtered out. To assess accuracy, CTAR was tested at five sites identified by Reger (1975), distributed across eastern Beringia. Overall, the CTAR (1975) CTs, the CTAR was able to identify potential terraces that had not been mapped previously. Overall, CTAR constitut es a substantial methodological advance for the delineation of CTs. The ability to identify CTs based on a DEM provides a useful tool for recognition of upland periglacial environments. The application of geomorphometry to periglacial geomorphology provid es a fresh analytical technique for scientists who work on periglacial problems. With high - quality DEMs and the increased processing power of contemporary computers, periglacial geomorphometry has potential to provide a less expensive alternative to many f orms of field - based work in periglacial regions. This chapter provides one example of the application of periglacial geomorphometry to answer the research question is there a distinct periglacial landscape? . Through specific geomorphometric analysis, thi s chapter draws the conclusion that there is indeed a distinct periglacial landscape that can be identify by objective methods. Cryoplanation 90 terraces are larger landforms than other periglacial features and occupy a substantial portion of the upland Berin gian landscape. The simplicity of identifying CTs on a DEM provides a methodology for recogni zing and interpret ing the periglacial landscape quickly and relatively easily. Although the methodology is not completely automated and requires advanced knowledg e about environmental factors affecting the terrain, geomorphic features in the study area besides CTs, and knowledge about the morphology of CTs, the methodology can be adapted for use across the periglacial realm. The ability to recognize where CTs are l ocated furthers the capability of geomorphometric analysis and facilitates geomorphological mapping. 91 Chapter 4 General Periglacial Geomorphometry 4.1 Introduction Following Evans (1972, 1987), the subject of geomorphometry can be subdivided naturally int o two subdisciplines, specific (treating discrete landscape units) and general geomorphometry (treating landscapes as continuous surfaces). This chapter focuses on general geomorphometry in the Yukon - Tanana Upland (YTU), a largely unglaciated region in eas tern Beringia that has been affected by periglacial processes throughout most of the Quaternary (Mertie 1937; Wahrhaftig 1965; Péwé , Burbank, and Mayo 1967). The YTU is bounded by the Yukon River in the north and the Tanana River in the south (Figure 4.1). General geomorphometric analysis in this study uses the methods developed by Strahler (1952) and Pike and Wilson (1971) , discussed briefly in Chapter 3. The work presented in this chapter is of a purely exploratory nature and seeks evidence about whether a hypsometric - topped ridges and hills in the YTU, the hypsometric inte gral should be relatively large, i.e., of a magnitude similar to that of a young landscape, as proposed by Strahler (1952). To provide terraces may be inherited feat ures formed under warm desert conditions, a similar analysis has also been conducted on an area in the Basin and Range physiographic province of southwestern USA. 92 4.1.1 Hypsometry: A potentially useful intersection between geomorphometry and periglacial g eomorphology is the hypsometric analysis developed by Strahler (1952) as a method for classifying the landscape based on a function that relates area to altitude. This analytical technique, called hypsometry , has two chief components: the hypsometric curve and the hypsometric integral (HI). The hypsometric curve is a graph relating area to altitude, while the hypsometric integral is the area under the curve. Strahler (1952) used the hypsometric curve to classify landscapes based on their developmental stage . Hypsometry was developed The hypsometric curve, as originally conceived by Strahler (1952), relates form to the Davisian concepts of stage (youth, maturity, and o ld - age). Although the Davisian model is now outdated, the concept of hypsometry remains a useful tool for understanding the character of the landscape. In a modern context, the hypsometric integral can be useful for comparing multiple landscape types based on their geometric characteristics. 4.1.2 Application of hypsometry in general periglacial geomorphometry: In an upland periglacial landscape, Nelson (1979b) suggested that hypsometric curves should be convex - upward , with an associated integral greater t han 0.5. Strahler (1952) interpreted convex - upward hypsometric curves as representative of younger landscapes. While this may apply (under Davisian assumptions) to the humid, mid - latitude fluvial environments Strahler was studying, the interpretation chang es significantly when discussing the periglacial landscape. It is likely that landscapes with numerous cryoplanation terraces would exhibit hypsometric curves with convex - upward shapes. High values of the hypsometric integral and convexity in the hypsometr ic curve may be expected in the YTU because, under the nivation model of 93 cryoplanation, the landscape erodes by parallel retreat of slopes over a large elevational range, leaving more of the reference solid intact than in a typical mature fluvial environme nt. Under this interpretation, the convex shape of the hypsometric curve would not necessarily be indicative of age, but rather of the relatively large proportion of land area remaining at higher elevation, even after long intervals of erosion. Figure 4 .1. The Yukon - Tanana Upland : S hown in the center of the map, the YTU is indicated in relation to the rest of Alaska. Application of geomorphometric techniques to periglacial landscapes allows for quantitative characterization of those landscapes and provid es a means for comparison with landscapes formed under different environmental conditions. French (2016) asserted that cryoplanation terraces and cryopediments in northern Yukon Territory may be inherited features, formed under warm and dry conditions, and therefore would be similar to features found in contemporary arid regions. This interpretation follows André (2003) in holding that distinctively periglacial landscapes do not exist. The penultimate section of this chapter (Section 4.3.2) applies the meth ods used for the Yukon - Tanana Upland to an arid environment in the 94 identifying differences between the periglacial landscape and those in other parts of the world, t his chapter uses quantitative evidence and semi - quantitative analysis to assess whether the concept of a distinctly periglacial upland landscape is warranted. 4.2 General Periglacial Geomorphometry: Methods The concept of a periglacial landscape involves a diverse assemblage of features that creates a distinct landscape (Chapter 2). Based on ideas from Nelson (1979b), the methods outlined here will investigate whether periglacial areas yield specific ranges of hypsometric integral values and characteristi initial work on the hypsometric integral at Indian Mountain near Hughes, Alaska provided a starting point upon which this study was built . Because the hypsometric integral was developed to investig ate watersheds, this analysis will focus on watersheds as the basic unit to stratify the Yukon - Tanana Upland region. Geographic scale is an important consideration in exploratory analysis. Because there can be no a priori way of knowing that environmental influences and geomorphic processes operate independent of scale (Phillips 2004), it is critical in exploratory work to operate over a range of spatial scale (e.g., Nelson et al. 1999) in order to determine the most appropriate scale(s) upon which to focu s. This point is reflected in the recommendations of a study by Hurtrez, Sol, and Lucazeau (1999), which found results from hypsometric analysis are highly dependent on drainage area. In this study, hypsometric analysis was applied to a series of nested wa tersheds of progressively larger extent and hydrological characteristics. 95 4. 2.1 Hypsometric integral for the Yukon - Tanana Upland: A DEM for the Yukon - Tanana Upland (YTU) was mosaiced together in ArcGIS from a series of ASTER Global DEM (GDEM) tiles (ASTER GDEM 2018). Using the r.watershed tool in GRASS GIS (GRASS Development Team 2018), a series of nested watersheds were delimited. The GRASS command r.watershed delineates watersheds based on a single flow accumulation (D8) algorithm. At the head of the bas in, pixels are assigned a value of zero, and at the mouth of the basin the flow accumulation will equal the threshold value set by the user (Hofierka, Mit á á, and Neteler 2009) . The watershed thresholds used here increase by factors of 10, from 100 to 1, 000,000. Watershed sizes are summarized by descriptive statistics in Table 4.1. The smallest basins are on the size of individual ridges, while the largest basins involve second - or third - order streams. The sites in unglaciated parts of the Yukon - Tanana Up land are a subset of those identified by Reger (1975) (Table 4.2), o ne of which were mapped and analyzed in Chapters 2 and 3. Originally, the site at Mount Fairplay was included in this analysis . A choice was made to replace the Mount Fairplay site with Si xtymile Butte based on work by Wi l lgoose and Hancock (1997) that showed that catchment geometry strongly influenced the hypsometric curves. Because watersheds at the Mount Fairplay site are more elongated than the other sites, the resultant curves were mor e concave and had lower integrals. The rest of the sites were chosen to provide coverage across the YTU. Glaciated sites (Table 4.2) were chosen from the map of glaciations in the YTU ( Péwé, Burbank, and Mayo 1967). Sites were identified to correspond with areas mapped as glaciated during Late Pleistocene ( Wisconsin ) and older ( Illinoian ) glaciations. Glaciated watersheds were chosen by virtue of being in close proximity to points identified from the map by Péwé, Burbank, and Mayo (1967). 96 Table 4.1 Des criptive statistics of watershed size : Watershed sizes are expressed in square kilometers Basin Size Area (km 2 ) Min Mean Median Max 100 0.00051 0.09262 0.07524 18.79299 1,000 0.00033 0.70437 0.50005 22.59464 10,000 0.00052 4.79647 1.57025 57.51812 100 ,000 0.00052 26.21953 0.30715 431.2464 Figure 4.2 . Map of study locations for general geomorphometry : Map of the study sites in Table 4.1. Glaciated sites are represented by black circles, unglaciated sites are white hexagons. Computation of the hypso metric curve was done by extracting elevations from the DEM as a table with the number of pixels at each elevation. The area of each pixel was calculated from the cell resolution (22.78 m) and multiplied by the number of cells with the same 97 elevation to ge t an area for each elevation value. To produce the hypsometric curve, elevations were normalized based on the ratio of the elevation above each threshold to the total elevation. Area was normalized by taking the area at each elevation and dividing it by th e total area. Hypsometric curves were plotted as elevation vs. area with axes ranging from zero to one. This method was applied to each watershed computed for each site. Table 4.2 . Sites for general geomorphometric analysis : Sites are all located within t he Yukon - Tanana Upland. Unglaciated sites were identified by Reger (1975), glaciated sites were chosen based on the map by Péwé, Burbank, and Mayo (1967). Site Name Glaciated or Unglaciated Latitude Longitude Eagle Summit Unglaciated 65.47 145.34 Sixtym ile Butte Unglaciated 63. 56 14 3 . 11 Gilles Creek Unglaciated 64.46 145.81 Cleary Summit Unglaciated 65.03 147.43 Boundary Unglaciated 64.07 140.98 Yukon River Glaciated 65.54 143.77 Fortymile River Glaciated 64.22 143.39 Healy River Glaciated 64.15 14 4.10 Salcha River Glaciated 64.94 144.56 Bear Creek Glaciated 65.16 145.35 4.3 Results from G eneral P eriglacial G eomorphometry 4.3.1 Hypsometric analysis of the Yukon - Tanana Upland: Hypsometric analysis of the Yukon - Tanana Upland facilitated classific ation of the landscape based on a quantifiable parameter allowing for comparison with other sites. Because the Yukon - Tanana Upland is known to have remained largely ice free during the last glacial interval ( Péwé, Burbank, and Mayo 1967) and to contain num erous cryoplanation terraces, this analysis provides a means by 98 Results from this analysis show trends both in size of watershed and between glaciated and unglaciated terrain. In general, as watershed area increases, the ERR decreases ( Appendix ; Table 4.3). Not surprisingly, the highest values of the ERR, above 0.5, for unglaciated sites were found at the smallest watershed size (Figure 4.3 , Appendix ). The lowest ERR value (0.10) was calculated from the entire Yukon - Tanana Upland. This low value is indicative of the relatively high proportion of the terrain occupied by river valleys around the YTU. The analysis of the entire Yukon - Tanana Upland might also be skewed by the arbitrary boundaries placed on the DEM and the inclusion of a portion of the Yukon Flats. Perhaps a better measure of the low values would be to include the largest watershed size calculated (1,000,000), which yielded an average ERR of 0.23 (Appendix). This value is again lower due to the high proportions of river valleys included in this watershed. Table 4.3 Comparison of HI for YTU sites : Eagle Summit, Mount Fairplay, Gilles Creek, Cleary Summit, and Boundary (Reger, 1975) were never glaciated during Pleistocene cold intervals. Sites A - E were chosen from areas that were glaciated during the Pleistocene ( Péwé, Burbank, and Mayo 1967) . Average Hypsometric Integral by Watershed Size Site Name 100 1,000 10,000 100,000 1,000,000 Eagle Summit 0.61 0.56 0 .52 0.40 Sixtymile Butte 0. 61 0. 50 0. 49 0. 50 Gilles Creek 0.44 0.46 0.41 0.37 0.27 Cleary Summit 0.51 0.57 0.54 0.35 0.50 Boundary 0.62 0.57 0.54 0.43 0.48 Yukon River 0.50 0.50 0.36 0.23 0.29 Fortymile River 0.49 0.23 0.36 0.37 0.37 Healy River 0 .43 0.43 0.52 0.50 0.37 Salcha River 0.35 0.31 0.36 0.40 0.40 Bear Creek 0.40 0.37 0.39 0.39 0.36 99 Figure 4.3 . Hypsometric curves for Yukon - Tanana Upland study sites : The top panel shows the unglaciated sites, while the bottom panels shows hypsometric curves for the glaciated sites within the YTU. Individual basin morphology and hypsometric curves are shown in Appendix. Mean elevation relief ratio is indicated in the lower left of each graph. 100 The form of h ypsometric curves is highly scale dependent and change as a result of the size of the watershed under consideration. Upward convexity is reflective of hillslope mass wasting processes, while upward concavity indicates dominance by fluvial processes (Willgoose and Hancock 1998 p. 621). At the local sca le (smallest watersheds), the curves are slightly convex, with flat segments ( benching ) in the middle (Figure 4.3 and individual graphs in Appendix). These flat areas are likely to be indications of the presence of flat benches, possibly cryoplanation te rraces, although these phenomena also occur in other environments (e.g., El Hamdouni et al. 2008) and are worthy of further analysis to determine the nature of this artifact. As basin size increases, curves for some of the watersheds in unglaciated terrai n become increasingly concave - upward and smooth out, while most retain convexity. At the 1 , 000 watershed size (Figure 4.3B), the curves are tightly clustered and show pronounced upward convexity. This is in distinct contrast to the curves representing the watersheds developed in glaciated terrain, which again are tightly clustered but show pronounced upward concavity. HI values reflect this disparity between glaciated and unglaciated terrain at this scale (Table 4.3). These characteristics indicate that th is scale may be most appropriate for further investigation , a lthough these contrasts are retained at the 10,000 watershed size . At the scale of the entire Yukon - Tanana Upland, the curve is highly concave upward , resembling what Strahler (1952) would have c alled an old age landscape (Appendix, Figure B4). 4.3.2 Hypsometric analysis of the Basin and Range: To address the concerns of French (2016), the methods used in the YTU were applied to terrain in a warm desert environment of the Basin and Range region of Nevada, USA (Figure 4.4). The Basin and Range study area is 101 bounded by the 117 th meridian in the east, as identified by Dohrenwend (1984). The rest of the study area contains multiple sites where Dohrenwend (1984) identified numerous small nivation holl ows. Size of the study area was chosen to incorporate the range of elevations from valley bottom to mountain top. Decreasing watershed areas were calculated in GRASS using the r. watershed tool, with threshold values of 10 6 , 10 5 , 10 4 , and 10 3 . Five randoml y selected sites were used as study sites, with the ERR and hypsometric curve calculated for each watershed size. For the smallest watershed (threshold of 1,000) two additional study sites were selected. Figure 4.4 . Map of Basin and Range study area , NV : Coordinates are given in NAD 1983 UTM Zone 11N. 102 The Basin and Range region (Thornbury 1965, Chapter 24) is climatically very different from the study sites in Alaska. Although the higher peaks in the region were once glaciated, in general the region has r emained unglaciated and is largely a warm desert environment (Dohrenwend 1984). This environment provides an ideal contrast to the Yukon - Tanana Upland. Across watershed scales, the hypsometric integral again decreased with increasing basin size (Figure 4. 5, Appendix Figures A 1 4 - A17 ). In the Basin and Range, however, the decrease was not as drastic as it was in the Yukon - Tanana Upland. The smallest watershed has an average ERR of 0.46, less than that of the Yukon - Tanana Upland. The ERR for the largest basin was 0.29, similar to that of the periglacial region , and again reflecting the dominance of fluvial processes . Local hypsometric curves for the Basin and Range area showed a different character from those of the Yukon - Tanana Upland (Figure 4.5). With one e xception, the hypsometric curves for the Basin and Range were concave upward . 4.4 Conclusion The use of hypsometric analysis for quantitative analysis of the periglacial landscape is a powerful tool. Results from this analysis allow for the comparison of watersheds between unglaciated and glaciated areas. This analysis also allows for comparison with different landscape types to identify similarities or differences. 103 Figure 4.5 . Hypsometric c u rves for Basin and Range study sites : Comparison of hypsometr ic curves for the Basin and Range Region at varying basin sizes. Mean elevation relief ratio is indicated in the lower left of each graph. 104 Comparison of the hypsometric curves between the glaciated and unglaciated watersheds of the Yukon - Tanana Upland sho wed differences, especially at intermediate scales. At the scale of the smallest watersheds there appears to be enough variation in the data at both the glaciated sites and the unglaciated sites that no significant trends can be identified. At the broadest scale, owing to the large river valleys throughout the YTU, the geomorphometric signature is that of a mature fluvial landscape. However, in the moderately sized watersheds, the character of the hypsometric curves is both qualitatively and quantitatively different for glaciated and unglaciated terrain. This result indicates an appropriate scale for further analysis. Results from the analysis of the Yukon - Tanana Upland show that subareas of the YTU are experiencing what early quantitative geomorphologists w ould have called a youthful stage of development. These authors might have ascribed this youthful stage to recent uplift events. The interpretation offered here, however, is that observed ERR values in the medium - sized YTU watersheds are instead artifact s of the prolonged operation of periglacial processes. While landscapes take long periods of time to form and few regions have undergone periglacial conditions long enough to achieve a p. 225 - - age stage of the periglacial cycle . The Yukon - Tanana Upland is, however, one of the oldest and best - developed in the periglacial realm. Despite its nearly unique character in this regard, it is clear that the YTU has not reached anything resembling the end result of t he periglacial cycle envisioned by Peltier (1950). Indeed, because YTU has been subject to periglacial conditions primarily during Quaternary time (Birot 1968), its distinctiveness lies largely its ubiquitous unconsumed upland remnants. These upland remn ants are occupied by 105 geomorphometric signature. The contrast found between the glaciated and unglaciated areas of the YTU shows that there is a difference in character be tween periglacial regions and those affected by glaciation. This difference provides another indication that the periglacial landscape is distinct. It also represents an initial step in our understanding of how periglacial landscapes can be characterized q uantitatively and the imprint of periglacial processes identified. of the Basin and Range stands in distinct contrast with the Yukon - Tanana Upland. Curves generated for th e Basin and Range are almost all concave, with no significant benching artifacts and lower values of the elevation - relief ratio. Dohrenwend (1984) identified a number of small nivation hollows in the upper reaches of the Basin and Range landscape. These ho llows are much smaller than the cryoplanation terraces found in Alaska, explaining the lack of noticeable benching in the hypsometric curves for the Basin and Range study. The ERR for the entire Basin and Range study area is slightly lower than that of the composite Yukon - Tanana Upland. Taken as a whole, this chapter provides another indication that there is a distinct upland periglacial landscape, with cryoplanation terraces as its foundation. However, this chapter treats only one area of general geomorph ometry. The results of this analysis indicate rich potential for application of other types of general geomorphometric analysis (e.g. Evans and Cox 1974; Florinsky 2017) in periglacial terrain, especially Beringia. As an initial, exploratory step, this stu dy can be regarded as having met with success, although much more work 106 remains. Extensions to the work will necessarily involve improved sampling frames and consideration of other Beringian provinces. 107 Chapter 5 Conclusions 5.1 Summary of Research The pe riglacial landscape is formed through cold climate, non - glacial processes. First coined by Lozinski (1909), the term periglacial was originally intended to describe those areas adjacent to Pleistocene glaciers. Since that first definition, the term has bee n modified to refer to all processes and areas that are part of the cold regions of the world outside of the glacial realm (French 2017). A natural consequence of this terminological expansion and the increased investigations of the cold regions of the ear th have led many to deduce that there is indeed a periglacial landscape. However, in recent years, several studies (e.g., André 2003; French and Thorn 2006; French 2016) have questioned whether there are distinct landscape s formed by perig lacial processes. This skepticism has potential to undermine the discipline of periglacial geomorphology and to distract attention away from process - based studies. This thesis provides an objective assessment of this critique. 5.1.1 Large - scale geomorpho logical mapping: André existence of a distinctively periglacial landscape is based primarily on the idea that periglacial processes are not sufficiently powerful to effect landscape - scale modification. Many of the studies conduc ted in periglacial geomorphology are concerned primarily with the numerous small, surface periglacial landforms. André (2003) is correct that these features alone are operating on a much more local scale and are not sufficient to form a landscape, especial ly an erosional landscape. However, this criticism fails to take into account two important considerations: (1) that cryoplanation landforms involve entire slope sequences; and (2) that 108 cold, nonglacial landscapes are composed of assemblages small forms th at, collectively, impart a striking and decidedly periglacial character to the terrain . Chapter 2 of this thesis used large - scale geomorphological mapping to identify the components of this periglacial assemblage and to examine how cryoplanation terraces fit into that assemblage. Mapping for this project was conducted at three sites (Mount Fairplay, Eagle Summit, and Iron Creek) that together form a transect across Alaska, which comprises most of eastern Beringia. Beringia was largely unglaciated during th e Pleistocene and provides one of the best opportunities to examine the concept of a periglacial landscape, owing to the relatively long period of time in which periglacial processes have operated (Birot 1968). At each site, a series of traditional surveyi ng and terrain analysis methods were performed in order to create maps of the periglacial features present. In addition to the sites in Alaska, a high - elevation site in British Columbia was investigated in order to document features in an environment in wh ich periglacial processes are currently active. Geomorphological mapping revealed a repeating pattern of landforms at and between sites. Similar features (sorted stripes, sorted polygons, solifluction lobes, tundra vegetation, cryoplanation terraces, etc .) were found at all of the sites. Not only were similar features found at all sites, but the spatial arrangement of these features followed a pattern in relation to cryoplanation terraces. This arrangement is well described as an assemblage of periglacial features that interact to form a geomorphically integrated landscape. It is through this interaction that large features such as cryoplanation terraces are able to form, showing that the form communities span a range of geographic scale. This repeating pa ttern indicates that these features are interacting and that together they can affect landscape - scale geomorphic change. 109 5.1.2 Specific periglacial geomorphometry: Geomorphometric analysis has been conducted in one way or another for hundreds of years. A s a formal subdiscipline of geomorphology, however, it is relatively young and has experienced accelerated growth since the digital revolution (Pike, Evans, and Hengl 2009). The use of geomorphometry in landscape and landform analysis has been integral to the understanding of a variety of different geomorphic subfields. Many studies have used geomorphometry in fluvial, glacial, and mountain landscapes, but very few studies have applied geomorphometric methods to periglacial problems. The need for a subfield of periglacial geomorphometry is necessary to strengthen theoretical and methodological ties between periglacial geomorphology and the parent discipline. Specific geomorphometry is concerned with discrete landforms and is often used to identify landform s (Evans 1987). Chapter 3 uses specific geomorphometric methods to provide an objective, semi - automated means by which to identify cryoplanation terraces. Parametrization of cryoplanation terraces and the use of these parameters to identify the location of CTs complements the methods used in Chapter 2 to examine the concept of a distinctive periglacial landscape. Cryoplanation terraces form the foundation or base of the periglacial assemblage. Specific geomorphometric analysis for this thesis used the two - meter resolution Arctic DEM to identify terrain parameters that can be used to identify the locations of cryoplanation terraces. The final semi - automated algorithm for identifying CTs on the landscape used contrasting slope facets as the primary identific ation parameter. The accuracy of the algorithm was assessed by comparing its output with an extensive dataset produced by 110 more traditional map and air photo interpretation ( Reger 1975). The accuracy assessment showed a high degree of agreement between the two methods of CT identification. Findings from this analysis confirm that cryoplanation terraces are ubiquitous in largely unglaciated upland periglacial regions. The findings show that at every study site, they ascend ridgetops and that nearly all of t he hill - and ridgetop s are summit - type CTs. In concert with the data from large - scale mapping, this sequence demonstrates that the foundation of the periglacial assemblage extends across eastern Beringia. 5.1.3 General periglacial geomorphometry: Genera l geomorphometry is concerned with the continuous land surface and is useful for identifying geomorphic signatures. Given the question of whether there is a distinctively periglacial landscape, a general geomorphometric investigation is a necessary compo nent of the analysis. In Chapter 4, hypsometric analysis was used to compare periglacial, glacial, and warm desert landscapes and to discern morphological differences between them. Hypsometric analysis was pioneered by Strahler (1952) and has since seen n umerous applications in geomorphology. This analysis proceeded from a deduction that, owing to the extensive flat - crest ed ridges and hills in Beringian, characteristic hypsometric curves and values of the hypsometric integral will result. Because a relativ ely large proportion of the terrain exists at higher elevations, the hypsometric curve should be convex upward, and the associated integral should be relatively high (above 0.5). In contrast, glaciated landscapes should have more area at lower elevations, with concave hypsometric curves and relatively low values of the hypsometric integral. 111 Hypsometric analysis was conducted primarily in the Yukon - Tanana Upland, an extensive area in eastern Beringia, in which cryoplanation terraces abound. This analysis u sed sites, Eagle Summit and Mount Fairplay, were also used for large - scale mapping (Chapter 2). Contrasting glaciated areas within the Yukon - Tanana Upland were i dentified from Péwé , Burbank, and Mayo (1967). Watersheds of varying sizes were calculated from a DEM. Threshold values, corresponding to upstream contributing area, were chosen that increase in size by powers of 10 from 100 to 1,000,000. For each water shed centered on the study sites and at each watershed scale the elevation and area were extracted. From these data, the elevation - relief ratio (ERR) of Pike and Wilson (1971) was calculated and the hypsometric curves plotted. This same process was used fo r both unglaciated and glaciated sites allowing for the comparison of hypsometric desert landforms, the hypsometric analysis was conducted for an area in the Basin and Range Province of southwestern USA. Results from the general geomorphometric analysis provid e the necessary quantitative characterization of the periglacial landscape. As the scale of the watersheds became more localized, the hypsometric curves beca me increasingly convex and the associated integral s increased. Information in the hypsometric integral increased with localization. Significant benching could be seen in the curves, which is interpreted to be the result of terraces on the landscape. When t he comparison with the glaciated landscapes was done, the associated integrals showed significant differences. This difference is best shown at the 1,000 basin sizes 112 (~1 km 2 ). At this scale, the glacial landscapes had lower integrals than those of the ungl aciated landscapes, and their hypsometric curves were more concave - upward . The comparison with watersheds in the Basin and Range region showed that there are substantial differences between the two landscapes. The trend of decreasing integrals and curves w as seen in this region as well , although the decrease was not as sharp as those observed in the Alaskan periglacial landscape s . This comparison is, however, only an initial exploratory step, and further analysis is necessary to unambiguously identify diffe rences between the two landscape types. 5.2 Is There a Distinctly Periglacial Landscape? André (2003) was among the first to express skepticism about the existence of a distinctively periglacial landscape by questioning the geomorphic effectiveness of sm all surficial periglacial forms. André (2003) provided a well - thought - out argument that disputed the existence of periglacial landscapes. However, the work fell into a pitfall typical of contemporary periglacial geomorphology: it focused narrowly on specif ic small landforms. Many of the studies conducted in periglacial geomorphology in North America are focused on specific landforms . I nterconnection s between units of the landscape are rarely discussed. The periglacial form community concept was popular in the German language literature into the 1980s but, ironically, diminished just as the rise of geographic information science (e.g., Goodchild 1992) could have led to great progress in the field. This work has largely been ignored in English - language litera ture. A broader focus that identifies interconnections between periglacial features would allow for a more thoroughly integrated geomorphology. The idea of an integrated periglacial geomorphology would address (2003) concerns by identifying how the smaller forms often observed can coalesce to form a distinct landscape. 113 The concerns of French (2016) are primarily related to his objections to standard narratives depicting cryoplanation terraces and cryopediments as having been formed through perigla cial processes. This thesis identifies cryoplanation terraces as the foundation for the periglacial landscape assemblage, a finding that contradicts Since Cairnes (1912) first postulated a link between CT - like features and nivation, a great deal of evidence (e.g., Demek 1969; Reger 1975; Nelson 1989, 1998) has indicated a link between CTs and climate. After several decades of often speculative critique (see Thorn and Hall 2002) that has increased skepticism about the notion that nivati on plays a significant role in CT genesis, fresh evidence supporting the validity of the nivation hypothesis of cryoplanation terrace formation is accumulating (e.g., Nelson and Nyland 2017). Such evidence could lead to cryoplanation terraces soon being vi ewed as unambiguously periglacial by most geomorphologists. This thesis provides additional evidence in support of the nivation hypothesis, in the form of large - scale maps that link the active nivation process and the cryoplanation terraces found throughou t unglaciated eastern Beringia. The processes documented at Frost Ridge are forming incipient terraces that resemble the larger and presumably more mature terraces in central Alaska. Together, this growing collection of evidence indicates a climatic origin for CTs. The question of what constitutes a landscape is a complicated one. French (2016) and André (2003) conclude that there is not a characteristic periglacial landscape, but their conception of landscape is unknown. Neither study provided an expli cit definition of the term. It is a given that the formation of a landscape requires long intervals of geological time, which are unusual in the periglacial realm. Many contemporary periglacial regions have only 114 recently been affected by periglacial proces ses, and insufficient time has elapsed for periglacial processes to have had substantial effects on landscape development. However, if there is an extensive and distinct periglacial landscape, it is almost certainly in Beringia (Birot 1968). Even so, becau se periglacial processes have been active in Beringia only since the beginning of the Quaternary (~ 1.8 million years), this landscape may be relatively young. Stated in Davisian terms, any postulated cycle of cryoplanation (Peltier 1950; Demek 1969) has n ot reached completion in many locations. However, there is a distinct character to upland Beringia, and at the mesoscale (1 - 10 km 2 ), this region constitutes a periglacial landscape. Evidence for this character has been presented in this thesis in the form of a repeating pattern of interacting features that, together, comprise a morphologically distinctive assemblage. The combination of the assemblage and the signatures found through geomorphometric analysis convincingly demonstrates that there is a distinct character to the landscape. The key to understanding the periglacial landscape lies in the recognition that cryoplanation terraces form the foundation of a multiscale assemblage. The answer to the question of whether this region constitutes a periglacial landscape is a resounding yes at the scale of cryoplanation terraces and individual mountains, but at regional scales exploratory hypsometric analysis indicates a less definitive answer is appropriate. Scale dependence is not uncommon in geomorphic work (Phillips 2004; Rasemann et al. 2004; Hurtrez, Sol, and Lucaseau 1999), and at regional and subcontinental scales Beringia is best characterized as polygenetic. 5.3 Recommendations for Future Work Continuation of this work would be useful for furthering o ur collective understanding about the place of cryoplanation terraces in mainstream periglacial geomorphology and 115 contradicts those who challenge the concept of a characteristic periglacial landscape. Arising from this analysis are several avenues for fu ture work that follow the general themes of this thesis: (1) geomorphological mapping; (2) periglacial geomorphometry; and (3) refined defini tion of the periglacial landscape. 5.3.1 Geomorphological mapping: Continuation of geomorphological mapping will provide further insight into the nature of the periglacial assemblage. The first step in this mapping effort would be to increase the sample size of periglacial features identified on terrace treads. Metrics on the size, shape, and morphology of the featu res are highly useful and an increased number would provide the ability to quantify the parameters of the assemblage. With the increased availability of remotely sensed imagery, particularly that obtained from unmanned aerial vehicles and small satellite technology , much of this mapping could be done inexpensively, at high resolution, and over extensive areas. Ground truthing of remote sensing data using rigorous sampling frameworks would allow for thorough evaluation of this work and would facilitate the creation of a wide range of large - scale maps of periglacial assemblages. 5.3.2 Specific periglacial geomorphometry: Periglacial geomorphometry has great potential to address the research questions of periglacial geomorphologists. While the algorithm explo red in this thesis achieved success, further refinement would help to reduce errors and potentially create a fully automated algorithm. This refinement might use additional terrain analysis to further filter out areas where CTs are absent or to identify th e precise locations of scarp - tread junctions. 116 The algorithm proposed in this thesis was necessarily tested at a relatively small number of sites. Given the imminent prospect of full coverage of Beringia by the Arctic DEM, these procedures can soon be appl ied to much larger areas. The algorithm could be applied to a large region, such as the Yukon - Tanana Upland, as a way to identify the periglacial nature of that area and to identify those locations where CTs are ubiquitous in the landscape. By applying the algorithm over a larger area, the boundaries of the periglacial landscape could be defined. 5.3.3 General periglacial geomorphometry: The hypsometric analysis performed here demonstrates that a morphometric signature exists in terrain subject to erosio n under periglacial conditions and provides good indication of differences between glacial and periglacial landscapes at certain scales. Additional work to provide a large sample size for statistical analysis would be beneficial for this comparison. Once c oncrete, quantifiable differences are identified, these can be used, in conjunction with spatial - analytic methods, to predict the locations of areas in which periglacial erosion has been substantial. One particularly interesting avenue for further analy sis was postulated by Mertie (193 7 ). In his travels through the Yukon - Tanana Upland, Mertie (193 7 ) suggested that if all the flat - topped ridges were connected they would form a gently sloping elevated plain extending from east to west. Mertie also explicit ly linked this surface with nivation and altiplanation (cryoplanation). Nelson and Nyland (2017) noted that the elevation trends of CTs in eastern Beringia follow a similar trend, rising from the Bering Sea to the Alaska - Yukon border. The consistency of th ese observations indicate s of analysis would provide an interesting and possibly definitive theoretical study. 117 5.3.4 Cryo p lanation terraces: As the foundation of the periglacial landscape, it is importa nt to understand the genesis of cryoplanation terraces. Further investigations of cryoplanation terraces to understand how they form are necessary. Process - based investigations of nivation processes and their relation to the periglacial assemblage are esse ntial for understanding how CTs relate to the broader periglacial landscape. Further work at sites such as Frost Ridge, where periglacial processes are highly active, would provide a useful juxtaposition with the older dormant terraces in Alaska. Understa nding the age of terraces and the length of time they take to form would provide an indication of the amount of time the periglacial landscape has been forming. If an understanding of the time CTs take to form can be achieved, these features could bolster the notion of the characteristic upland periglacial landscape. 118 APPENDI X 119 Chapter 4 Hypsometric Curves and Basin Information Figure A 1 . Hypsometric curves and integrals for 100 watershed size at Eagle Summit, AK : The Hypsometric I ntegrals are shaded in the map and the resulting hypsometric curves for the watersheds outline the map. Coordinates on the map are from UTM Zone 6N. 120 Figure A 2 . Hypsometric curves and integrals for 1,000 watershed size at Eagle Summit, AK : The Hypsometric I ntegrals are shaded in the map and the resulting hypsometric curves for the watersheds outline the map. Coordinates on the map are from UTM Zone 6N. 121 Figure A 3 . Hypsometric curves an d integrals for 10,000 and 100,000 watershed sizes at Eagle Summit, AK : Hypsometric integral for the watershed thresholds of 10,000 (blue) and 100,000 (green) with the resulting hypsometric curves on the outside. This region i s centered on Eagle Summit. 122 Figure A 4 . Hypsometric curve and integral for entire Yukon - Tanana Upland : Coordinates on the map are in meters. 123 Figure A 5 . Hypsometric curves a nd integral for Sixtymile Butte, AK : The numbers on the left indicate watershed size. 124 Figure A 6 . Hypsometric curves and integral for Cleary Summit, AK : The numbers on the left indicate watershed size. 125 Figure A 7 . Hypsometric curves and integral for Gilles Creek, AK : The numbers on the left indicate watershed size. 126 Figure A8 . Hypsometric curves and integral fo r Boundary, AK and Yukon Territory : The numbers on the left indicate watershed size. 127 Figure A 9 . Hypsometric curves and integral for the glaciated site at Yukon River : The numbers on the left indicate watershed size. 128 Figure A 1 0 . Hypsometric curves and integral for glaciated site at Fortymile River : The numbers on the left indicate watershed size. 129 Figure A 1 1 . Hypsometric curves and integral for glaciated site at Healy Rive r : The numbers on the left indicate watershed size. 130 Figure A 1 2 . Hypsometric cu rves and integral for glaciated site at Salcha River : The numbers on the left indicate watershed size. 131 Figure A 1 3 . Hypsometric curves and integral for glaciated site at Bear Creek : The numbers on the left indicate watershed size. 132 Figure A 1 4 . Hypsometr ic curves for the 1,000 watershed size in the Basin and Range : Watershed threshold is 1,000 and the watersheds were chosen at random. 133 Figure A 1 5 . Hypsometric curves for the 1 0 ,000 watershed size in the Basin and Range : Watershed threshold is 1 0 ,000 and t he watersheds were chosen at random. 134 Figure A 1 6 . Hypsometric curves for the 1 00 ,000 watershed size in the Basin and Range : Watershed threshold is 1 00 ,000 and the watersheds were chosen at random. 135 Figure A 17 . Hypsometric curves for the 1 ,000 ,000 water shed size in the Basin and Range : Watershed threshold is 1 ,000 ,000 and the watersheds were chosen at random. 136 REFERENCES 137 REFERENCES Å kerman, J. (1980). Studies on periglacial geomorphology in West Spitsbergen. Meddelanden Fran Lunds Universitets Geografisk a Instituion Avhandlingar , 84, 297 pp. Ambrose, J.W. (1964). Exhumed palaeoplains of the Precambrian shield of North America. American Journal of Science , 262, 817 - 857. Andersson, J.G. (1906). Solifluction: a component of subaerial denudation. Journal of Geology, 14, 91 - 211. André, M. (2003). Do periglacial landscapes evolve under periglacial conditions? Geomorphology , 52, 149 - 164. Annaheim, H. (1956). Zur frage der geomorphogischen Kartierung. Petermanns Geographische Mitteillungen , 103, 315 - 319. ASTER G DEM NASA and METI. V2. NASA EOSDIS Land Processes DAAC, USGS Earth Resources Observation and Science (EROS) Center, Sioux Falls, South Dakota https://lpdaac.usgs.gov . Accessed January 15, 2018. Atlas of Canada. (195 8). Canadian Department of Mines and Technical Surveys. 4 pp. Ballantyne, C.K. (1978). The hydrologic significance of nivation features in permafrost areas. Geografiska Annaler: Series A, Physical Geography , 60(1/2), 51 54. Ballantyne, C.K. (1996). Formati on of miniature sorted patterns by shallow ground freezing: a field experiment. Permafrost and Periglacial Processes , 7 (4), 409 - 424. Ballantyne, C.K. (2018). Periglacial Geomorphology . Hoboken, NJ: John Wiley & Sons. 472 pp. Benedict, J.B. (1970). Downslop e soil movement in a Colorado alpine region: rates, processes and climatic significance. Arctic and Alpine Research, 2, 165 - 226 . Berrisford, M.S. (1991). Evidence for enhanced mechanical weathering associated with atches, Jotunheimen, Norway. Permafrost and Periglacial Processes , 2(4), 331 - 340. Birot, P. (1968). The Cycle of Erosion in Different Climates . Berkeley , CA : University of California Press. 144 pp. Bishop, M.P., James, A.L., Shroder, J.F., and Walsh, S.J. (2012). Geospatial technologies and digital geomorphological mapping: Concepts, issues, and research. Geomorphology , 137, 5 - 26. 138 Bishop, M., and Shroder, J.F. (2004). Geographic Information Science and Mountain Geomorphology . Chichester, UK: Praxis. 486 pp . Boelhouwers, J., Holness, S., and Sumner, P. (2003). The maritime Subantarctic: a distinct periglacial environment. Geomorphology , 52 (1 - 2), 39 - 55. Bostock, H.S. (1970). Physiographic Regions of Canada . Ottawa: Geological Survey of Canada, Map 1254A, scal e 1:5,000,000. Brigham - Grette, J. (2001). New perspectives on Beringian Quaternary paleogeography, stratigraphy, and glacial history. Quaternary Science Reviews , 20, 15 - 24. Brubaker, L . B., Anderson , P.M. , Edwards , M.E. , and Lozhkin , A.V . (2005). Beringia a s a g lacial r efugium for b oreal t rees and s hrubs: New p erspectives from m apped p ollen d ata. Journal of Biogeography , 32(5), 833 48. doi:10.1111/j.1365 - 2699.2004.01203.x. Brunnschweiler, D.H. (1965). The morphology of altiplanation in interior Alaska. Unpu blished report to the U.S. Army Cold Regions Research and Engineering Laboratory, Hanover, NH. 48 pp. Büdel, J. (1977). Klima - Geomorphologie translation by Fischer, L. and Busche, D. (1982). Climatic Geomorphology . Princeton, NJ : Princeton University Press. 443 pp. Caine, N. (1972). The distribution of sorted patterned ground in the English Lake District. Révue de Géomorphologie Dynamique , 21, 49 - 56. Caine, N. (1992). Sediment transfer on the floor of the Martinelli Snowpatch, Colorado Front Range, U.S.A. Geografiska Annaler: Series A, Physical Geography , 74(2/3), 133 144. Cain, M.L., Bowman, W.D. , and Hacker, S.D. (2011). Ecology . Sunderland, MA : Sinauer. 648 pp. Cairnes, D.D. (1912). Differential erosion and equiplanation in portions of Yukon and Alaska. Bulletin of the Geological Society of America , 23, 333 348. Chaiko, A.V. (1988). Morphology a nd o rigin of a ltiplanation t erraces under c ryozone c onditions. Novosibirsk. (in Russian.) Church, M. (2005). Continental drift. Earth Surface Processes and Landforms , 30, 129 - 130. Church, R.E., Péwé, T.L., and Andresen, M.J. (1965). Origin and environment al significance of large - scale patterned ground, Donnelly Dome are a , Alaska. U.S. Army Cold Regions Research and Engineering Laboratory Research Report , 159, 77 pp. Cialek, C.J. (1977). The Cathedral Massif, Atlin Provincial Park, British Columbia . Scale 1:20,000. Pacific Science Center. Seattle, WA . 139 Clarke, J.I. (1966). Morphometry from maps. In: Dury, G.H. (ed.). Essays in Geomorphology . London: Heinemann , 235 - 274. Clarke, K.C. (1995). Analytical and Computer Cartography . Englewood Cliffs, NJ: Prentice H all. 334 pp. Corte, A.E. (1963). Vertical migration of particles in front of a moving freezing plane. US Army Cold Regions Research and Engineering Laboratory Research Report , 105, 8 pp. Davis, W.M. (1909). Geographical E ssays . New York, NY: Ginn and Co. 777 pp. Delcourt, H.R., and Delcourt, P.A. (1988). Quaternary landscape ecology: relevant scales in space and time. Landscape Ecology , 2 (1), 23 - 44. Demek J. ( 1969 ) . Cryoplanation terraces, their geographical distribution, genesis and development. Rozpravy Ceskoslovenski Akademie ved, Rada, Matematickych a Prirodnich Ved , 79(4), 80 pp. Demek, J. ( Ed. ). (1972). Manual of Detailed Geomorphological Mapping . IGU Commission for Geomoropholgical Mapping. Academia, Prague , 368 pp. Demek, J. and Embleton, C. ( Eds .) . (1976). Guide to medium - scale geomorphological mapping . IGU Commission for Geomorphological Mapping and Surveying. Academia, Prague , 348 pp. Dikau, R., Brabb, E.E., Mark, R.K. (1991). Landform classification of New Mexico by computer. U.S. Geological Su rvey Open File Report 91 - 634, 15 pp. Dikau, R., Brabb, E.E., Mark, R.K., Pike, R.J. (1995). Morphometric landform analysis of New Mexico. Zeitschrift fur Geomorphologie Supplementband , 101, 109 - 126 . Dohrenwend, J.C. (1984). Nivation landforms in the west ern Great Basin and their paleoclimatic significance. Quaternary Research , 22 (3), 275 - 288. Dyke, L.S. (1984). Frost heaving of bedrock in permafrost regions. Association of Engineering Geologists Bulletin , XXI(4), 389 - 405. Eakin, H.M. (1916). The Yukon - Ko yukuk region, Alaska. U.S. Geological Survey Bulletin 631 , 88 pp. and Blaschke, T. (2010). Developing a semantic model of glacial landforms for object - based terrain classification the example of glacial cirques. GEOBIA , 1682 - 1777. Eisank, C., Smith, M., and Hillier, J. (2014). Assessment of multiresolution segmentation for delimiting drumlins in digital elevation models. Geomorphology , 214 , 452 - 464. 140 El Hamdouni, R., Irigaray, C., Fernández, T., Chacón, J., and Keller, E.A. (2008). Assessment of relative active tectonics, southwest border of the Sierra Nevada (southern Spain). Geomorphology , 96 (1), 150 - 173. Elias, S.A., Short, S.K., and Birks, H.H. (1997). Late Wisconsin environments of the Bering land bridge. Palaeogeography, palaeoclimatology, palaeoecology , 136 (1 - 4), 293 - 308. Elias, S.A . (2001). Beringian paleoecology: Results from the 1997 workshop. Quaternary Science Reviews , 20, 7 - 13. Evans, I.S. (1972). General geomorphometry, derivatives of altitude, and descriptive statistics. In: Chorley, R. J. ( E d.). Spatial Analysis in Geomorph ology . 17 - 90 pp. New York, NY: Harper & Row . Evans, I.S. (1977). World - wide variations in the direction and concentration of cirque and glacier aspects. Geografiska Annaler: Series A, Physical Geography , 59 (3 - 4), 151 - 175. Evans, I.S. (1987). The morphom etry of specific landforms. In: Gardiner, V. ( E d.), International Geomorphology 1986 . New York: Wiley, 105 - 124. Evans, I.S. (2009). Allometric development of glacial cirques: an application of specific geomorphometry. Proceedings of Geomorphometry , 248 - 253 . Evans, I.S. (2012). Geomorphometry and landform mapping: What is a landform? Geomorphology , 137(1), 94 106. http://doi.org/10.1016/J.GEOMORPH.2010.09.029 Evans, I.S., and Cox, N. (1974). Geom orphometry and the operational definition of cirques. Area, 6, 150 - 153. Evans, D.J., Kalyan, R., and Orton, C. (2017). Periglacial geomorphology of summit tors on Bodmin Moor, Cornwall, SW England. Journal of Maps , 13 (2), 342 - 349. Fenneman, N.M. (1914). Ph ysiographic boundaries within the United States. Annals of the Association of American Geographers , 4 (1), 84 - 136. Fenneman, N.M. (1917). Physiographic division of the United State s . Annals of the Association of American Geographers , 6, 19 - 28 . Fisher, N.I. (199 3). Statistical Analysis of Circular Data . Cambridge: Cambridge University Press, 279 pp. Florinsky, I.V. (2017). An illustrated introduction to general geomorphometry. Progress in Physical Geography , 41 (6), 723 - 752. Forman, R.T., and Godron, M. (1981) . Patches and structural components for a landscape ecology. BioScience , 31 (10), 733 - 740. 141 Foster, H.L. (1967). Geology of the Mount Fairplay Area Alaska. United States Geological Survey Bulletin , 1241, 18 pp. French, H.M. (1987). Periglacial processes and landforms in the Western Canadian Arctic. In: Boardman, J. (Ed.), Periglacial Processes and Landforms in Britain and Ireland. Cambridge, UK: Cambridge University Press , 27 - 43 . French, H.M. (2016). Do periglacial landscapes exist? A discussion of the upland landscapes of Northern Interior Yukon, Canada. Permafrost and Periglacial Processes , 27(2), 219 228. French, H.M. (2017). The Periglacial Environment (4th ed.). Hoboken, NJ: John Wiley & Sons. French, H.M., Thorn, C.E. (2006). The changing nature of peri glacial geomorphology. Geomorphologie: Relief, Processes , Environment , 3, 165 - 174 Furrer G. and Dorigo G. (1972). Abgrenzung und Gliederung der Hochebirgsstufe der Alpen mit Hilfe von Solifluktionsformen. Erdkunde , 26, 98 - 107. Galon, R. (1962). Instruction to the detailed geomorphological map of the Polish lowland. Geography and Geomorphology Department , Polish Academy of Sciences, Torun. Garleff, K. (1970). Verbreitung und Vergesellschaftung rezenter Periglazialerscheinungen in Skandinavien. Gottinger Geog raphicsche Abhandlungen , 51, 7 - 66. Goldthwait, R.P. (1976). Frost sorted patterned ground: A review. Quaternary Research, 6, 27 - 35 . Goodchild, M.F. (1992). Geographical information science. International Journal of Geographical Information Systems, 6, 31 - 4 5. Graf K. (1973). Vergleichende b etrachtungen zur s olifluktion in verschiedenen b reitenlagen. Z eitschrift für G eomorphologie S upplementband, 16, 104 - 154. Gran ö , J.G. (1929). Rein geographie: Eine methodologische studie beleuchtet mit beispielen aus Finnla nd und Estland. Acta Geographica , 2(2). GRASS Development Team. (2018). GRASS GIS 7.0.7svn Reference Manual. < https://grass.osgeo.org/grass70/manuals/index.html > . Accessed April 21, 2018. Gruber, S., and Peckham, S. (2009) Land - surface parameters and ob jects in hydrology. In: Hengl, T. and Reuter, H.I. (Eds.) 1 71 - 194 . Geomorphometry: Concepts, Software, Applications. Developments in Soil Science, 33. Oxford, UK. Elsevier. Gustavsson, M., Kolstrup, E., and Seijmonsbergen, A.C. (2006). A new symbol - and - GIS based detailed geomorphological mapping system: renewal of a scientific discipline for understanding landscape development. Geomorphology , 77, 90 - 111. 142 Guthrie, R.D. (1982) . Mammals of the mammoth steppe as paleoenvironmental indicators. In : Paleoecology o f Beringia . Hopkins, D.M., Matthews, J.V ., Schweger, C.E., and Young , S.B. (Eds.) New York: Academic Press , 307 26. Hall, K. (1993). Enhanced bedrock weathering in association with late - lying snowpatches: evidence from Livingston Island, Antarctica. Earth Surface Processes and Landforms , 18, 121 129. doi.org/10.1002/esp.3290180204 Hammond, E.H. (1964). Analysi s of properties in land form geo scale land form mapping. Annals of the Association of American Geographers , 54 (1), 11 - 19. Hanson, H. C. (1950). Vegetation and soil profiles in some solifluction and mound areas in Alaska. Ecology , 31 (4), 606 - 630. Hartshorne , R. (1939). The nature of geography: A critical survey of current thought in the light of the past. Annals of the Association of American Geographers, 29(3), 173 - 412. Haschenburger, J.K., Souch, C. (2004). Contributions to the understanding of geomorphic landscapes published in the Annals. Annals of the Association of American Geographers , 94(4), 771 - 793. Hengl, T., Reuter, H.I. ( Eds. ). (2009). Geomorphometry: Concepts, Software, Applications. Developments in Soil Science , 33, 765 pp . Higgins, S., Maho n, M., and McDonagh, J. (2012). Interdisciplinary interpretations and applications of the concept of scale in landscape research. Journal of E nvironmental M anagement , 113 , 137 - 145. Hjort, J., Etzelmüller, B., and Tolgensbakk, J. (2010). Effects of scale an d data source in periglacial distribution modelling in a high Arctic environment, western Svalbard. Permafrost and Periglacial Processes , 21, 345 - 354. Hjort, J., Luoto, M., and Seppälä, M. (2007). Landscape scale determinants of periglacial features in sub arctic Finland: a grid - based modelling approach. Permafrost and Periglacial Processes , 18, 115 - 127. Hofierka, J., Mit á á, H., Neteler, M. (2009) . Geomorphometry in GRASS GIS. In: Hengl, T., Reuter, H.I. (Eds.) Geomorphometry: Concepts, Software, Applicat ions. Developments in Soil Science. 33, 1 - 28. Hopkins, D.M., Matthews, J.V., Schweger, C.E., Young, S.B. (Eds.) (1982). Paleoecology of Beringia. Academic Press, New York, NY. 489 pp. 143 Hopkins, D.M., Smith, P.A., and Matthews, J.V. (1981). Dated Wood f rom Alaska and the Yukon: Implications for f orest r efugia in Beringia. Quaternary Research , 15 (3), 217 49. doi:10.1016/0033 - 5894(81)90028 - 4. Horn, B.K. (1981). Hill shading and the reflectance map. Proceedings of the IEEE , 69 (1), 14 - 47. Hurtrez, J. - E., So l, C., and Lucazeau, F. (1999). Effect of drainage area on hyp s ometry from an analysis of small - scale drainage basins in the Siwalik Hills (Central Nepal). Earth Surface Processes and Landforms, 24, 799 - 808. Karte, J. (1979). Räumliche Abgrenzung und regi onale Differenzierung des Periglaziärs. Bochumer Geographische Arbieten , 35, 211 pp. Karte J. (1982). Development and present state of German periglacial research in Arctic and alpine environments. Biuletyn Peryglacjalny , 29, 183 - 201. Karte J. (1983). Per iglacial phenomena and their significance as climatic and edaphic indicators. GeoJournal , 7, 329 - 340. Kelletat, D. (1970). Rezente periglazialerscheinungen im Schottischen Hochland. Untersuchungen zu ihrer verbreitung und vergesellschaftung. G öttinger Geog raphische Abhandlungen, 5, 67 - 140. King, C.A.M. (1982). Morphometry in glacial geomorphology. In: Coates, D.R. (Ed.). Glacial Geomorphology . Amsterdam: Springer. 147 - 162. Geological Society of Americ a Bulletin , 106, 101 - 131. Klimaszewski, M. (1963). The principles of geomorphological map of Poland. Geographic Studies 46, 69 - 70. Klimaszewski, M. (1965). The principles of the geomorphological map of Poland. Przeglad Geografizny , 28, 32 - 40. Kling, J. (1 997). Observations on sorted circle development, Abisko, northern Sweden. Permafrost and Periglacial Processes , 8 (4), 447 - 453. Holocene nivation rates on deglaciated surfaces of James Ross Island, Antarctica. In: Deline, P., Bodin, X., and Ravenel, L. (eds.) 5 th European Conference on Permafrost Book of Abstracts . Chamonix, France: International Permafrost Association, 519 - 520. Kreig, R.A. and Reger, R.D. (1976). Preconstruction terrain ev aluation for the Trans - Alaska Pipeline Project. In: Coates, D. (ed.) Geomorphology and Engineering . Stroudsburg, PA: Dowden, Hutchinson and Ross, 55 - 76. 144 Kreig, R.A. and Reger, R.D. (1982). Airphoto Analysis and Summary of Landform Soil Properties Along the Route of the Trans - Alaska Pipeline System . Geological Report 66 , Alaska Division of Geological and Geophysical Surveys, 149 pp. Kulu, E. (2018). Nanosatellite database by Erik. http://www.nanosats.eu/ . Accessed Ma y 12, 2018. Lozinski, W. (1909). Uber die mechanische verwitterung der sansteine im germassingten klima. Mathematiques, et Naturelles, 1, 1 - 25. Lozinski, W. (1912). Die pe riglaziale fazies der mechanichen verwitterung. 11 th International G eological C ongress, Stockholm, 1910, Compte rendu, 1039 - 1053. Mackay, J.R. (1980). Deformation of ice - wedge polygons, Garry Island, Northwest Territories. Current R esearch, part A. Geolog ical Survey of Canada , Paper 80 , 287 - 291. Mackay, J.R. (1986). Frost Mounds. In: French, H.M. (Ed.). 363 - 364. Focus: Permafrost Geomorphology, The Canadian Geographer , 30. Madole, R.F. (1972). Neoglac i al facies in the Colorado Front Range. Arctic and Alp ine Research , 4(2), 119 130. Mardia, K.V. (1972). Statistics of Directional Data . London: Academic Press, 357 pp. Mark, D.M. (1975). Geomorphometric parameters -- a review and evaluation. Geografiska Annaler Series A , 57, 165 - 177. Mark, D.M. (1983). Relation s between field - surveyed channel networks and map - based geomorphometric measures, Inez, Kentucky. Annals of the Association of American Geographers, 73, 358 - 372. Mascaro, J. (2018) Satellite Imagery. https://www.pgc.umn.edu/events/video - online - pgc - seminar - planets - joseph - mascaro - 2 - 22/ . Accessed April 20, 2018. Matsuoka, N. (2001). Solifluction rates, processes and landforms: A global review. Earth Science Re views, 5(2) , 107 - 134. Matthes, F.E. (1900). Glacial sculpture of the Bighorn Mountains, Wyoming. United States Geological Survey 21st Annual Report 1899 - 1900 , 167 - 190. Mertie, J.B. ( 1937 ). The Yukon - Tanana region, Alaska. U.S. Geological Survey Bulletin 87 2, 276 pp. Millar, S.W. (2005). Fabric variability associated with periglacial mass - wasting at Eagle Summit, Alaska. Geomorphology , 72 (1 - 4), 222 - 237. 145 Mitchell, S.G. and Montgomery, D.R. (2006). Influence of a glacial buzzsaw on the height and morphology of the Cascade Range in central Washington State, USA. Quaternary Research , 65, 96 - 107. Moffit, F.H. (1905). The Fairhaven gold placers, Seward Peninsula, Alaska. U.S. Geological Survey Bulletin 247, 85. Morris, S.E. (1981). Topoclimatic factors and the dev elopment of rock glacier facies, Sangre de Cristo Mountains, southern Colorado. Arctic and Alpine Research , 329 - 338. Morris, S.E. and Olyphant, G.A. (1990). Alpine lithofacies variation: Working toward a physically - based model. Geomorphology , 3, 73 90. Nel son, F.E., (1979a). Patterned G round in the Juneau Icefield R egion, Alaska British Columbia . M .S. thesis, Department of Geography, Michigan State University, 136 pp. Nelson, F.E. (1979 b analysis. Unpublished paper, Department of Geography, University of Michigan, 68 pp. + appendices. Nelson, F.E. (1989). Cryoplanation terraces: Periglacial cirque analogs. Geografiska Annaler , 71(1/2), 31 41. Nelson, F.E. (1998). Cryoplanation terrace orie ntation in Alaska. Geografiska Annaler , 80(2), 135 - 151. DOI: 10.1111/j.0435 - 3676.1998.00032.x Nelson, F.E. (2017). Personal Communication: Activity of Periglacial Landforms on the Cathedral Massif, northwestern British Columbia . Nelson, F.E. and Nyland, K. E. (2017). Periglacial cirque analogs: Elevation trends of cryoplanation terraces in eastern Beringia. Geomorphology , 293, 305 317. Nelson, F.E., Shiklomanov, N.I., and Mueller, G.R. (1999). Variability of active - layer thickness at multiple spatial scales , north - central Alaska, U.S.A. Arctic, Antarctic, and Alpine Research , 31(2), 158 - 165. Noh, M.J., and Howat, I.M. (2015). Automated stereo - photogrammetric DEM generation at high latitudes: Surface Extraction with TIN - based Search - space Minimization (SETSM) validation and demonstration over glaciated regions. GIScience and Remote Sensing , 52 (2), 198 - 217. Obruchev, S.A. (1937). Soliflukcionnye (nagornyye) terrasy i yikh genezis no osnovanii rabot v Chukotsom kraye. Problemy Arktiki , 3: 27 - 48, 4: 57 - 83. Lening rad. Olaya, V. (2009). Basic l and - s urface p arameters. In: Hengl, T. and Reuter, H.I. (Eds.) 141 - 169. Geomorphometry: Concepts, Software, Applications. Developments in Soil Science, 33. Oxford, UK. Elsevier. 146 Olwig, K.R. (2002). Landscape, N ature, and the B ody P R enaissance to N ew W orld . Madison, WI. The University of Wisconsin Press , 298 pp. Pannekoek, A.J. (1967). Generalized contour maps, summit level maps, and streamline maps as geomorphol o gical tools. Zeitschrift für Geom orphologie , 11, 169 - 182. Park Nelson, K.J., Nelson, F.E., and Walegur, M.T. (2007). Periglacial Appalachia: palaeoclimatic significance of blockfield elevation gradients, eastern USA. Permafrost and Periglacial Processes , 18 (1), 61 - 73 . Parker, K.C., and Be ndix, J. (1996). Landscape - scale geomorphic influences on vegetation patterns in four environments. Physical Geography , 17 (2), 113 - 141. Passarge, S. (1914). Morphologischer atlas , Erl ä uterungen zu l ief. 1 , Morphologie des Messtischblates Stadtremba (1:25,0 00). Mitteilungen der Geographischen Gesellschaft Hamburg, 28 Friederichsen. Peltier, L.C. (1950). The geographic cycle in periglacial regions as it is related to climatic geomorphology. Annals of the association of American Geographers , 40 (3), 214 - 236. P Proceedings of the American Philosophical Society, 66, 621 - 644. and Delgado, J. (2009). Differentiating geology and tectonics using a spatial a utocorrelation technique for the hypsometric integral. Journal of Geophysical Research: Earth Surface , 114 (F2). 15 pp. Péwé T.L. ( ed. 1969). The P er iglacial E nvironment. Montreal: McGill 487 pp. Péwé T.L. (1970). Altiplanation terraces of early Quaternary age near Fairbanks, Alaska. Acta Geographica Lodziensia , 24, 357 - 363. Péwé, T.L, Burbank L., and Mayo L.R. (1967). Multiple glaciation of the Yukon - Tanana Upland, Alaska. U.S. Geological Survey Miscellaneous Investigations Map I - 507 , scale 1:500,000. Péwé, T.L. and Reger, R.D. (1969). Geologic - p hysiographic m ap of Indian Mountain and v icinity, Hughes A02 q uadrangle , Alaska. Progress report to U.S. Army Cold Regions Research and Engineering Laboratory, Grant No. DA - AMC - 27 - 021 - 68 - 624. Phillips, J.D. (2004). Independence, contingency, and scale linkages in physical geography. In: Shepard, E. and McMaster, B. (Eds.). 86 - 100. Scale & Geographic Inquiry . Malden, MA: Blackwell Publishing. Pike, R.J. (1988). The geometric signature: qua ntifying landslide - terrain types from Digital Elevation Models. Mathematical Geology , 20, 491 - 511. 147 Pike, R.J. (2001). Scenes into numbers: Facing the subjective in landform quantification. In: Hoffman, R.R., Markman, A.B. (Eds.) 83 - 114. Interpreting R emot e S ensing I magery -- H uman F actors. Boca Raton, FL: Lewis Publishers (CRC). Pike, R.J. (2002). A bibliography of terrain modeling (geomorphometry), the quantitative representation of topography. U . S . G eological S urvey Open File Report 02 - 465 . 158 pp. Pike, R .J., Evans, I.S., and Hengl, T. (2009). Geomorphometry: A brief guide. In: Hengl, T. and Reuter, H.I. (Eds.). Geomorphometry: Concepts, Software, Applications . Developments in Soil Science, 33, 3 30. Pike, R.J. and Wilson, S.E. (1971). Elevation - relief r atio, hypsometric integral, and geomorphic area - altitude analysis. Geological Society of America Bulletin , 82, 1079 - 1084. Price, L.W. (1974). The developmental cycle of solifluction lobes. Annals of the Association of American Geographers, 64(3), 430 - 438. Prindle, L.M. (1905). The gold placers of the Fortymile, Birch Creek, and Fairbanks regions, Alaska. U.S. Geological Survey Bulletin 251, 89 pp. Polar Geospatial Center. Arctic DEM. https://www.pgc.umn .edu/data/arcticdem/ . Accessed March 4, 2017. Poser H. ( E d.) . ( 1976 ) . Formen, Formengesellschaften, und Untergrenzen in den heutigen Periglazialen Höhenstufen der Hochgebirge Europas und Africas zwischen Arktis und Äquator: Bericht über eine Symposium. Ab handlungen der Akademie der - physikalische Klasse , Dritte Folge, 31: 354 pp. Powell, J.W. (1895). Physiographic R egions of the United States . American Book Company, New York. Rapp, A. (1967). Pleistocene activity and Holocene stability of hillslopes, with examples from Scandinavia and Pennsylvania. 40, 229 - 244. Rasemann, S., Schmidt, J., Schrott, L., and Dikau, R. (2004). Geomorphometry in mountain terrain. In: Bis hop, M., and Shroder, J.F. (Eds.) Geographic Information Science and Mountain Geomorphology . Chichester, UK: Praxis , 101 - 145. Reger, R.D. (1975). Cryoplanation Terraces of Interior and Western Alaska . Ph.D. dissertation, Department of Geology Arizona Stat e University, 226 pp. + maps, appendices. Reger, R.D., and Péwé, T.L. (1976). Cryoplanation terraces: Indicators of a permafrost environment. Quaternary Research , 6, 99 109. 148 Richmond, G.M. (1962). Quaternary stratigraphy of the La Sal Mountains, Utah. U.S. Geological Survey Professional Paper 324 , 135 pp. Rudberg, S. (1972). Periglacial zonation -- a discussion. Göttinger Geographische Abhandlungen , 60, 221 - 233. Rudberg, S. (1977). Periglacial zonation in Scandinavia. In Poser, H. (ed.) pp. 92 - 104. Formen, F ormengesellschaften, und Untergrenzen in den heutigen Periglazialen Höhenstufen der Hochgebirge Europas und Africas zwischen Arktis und Äquator: Bericht über eine Symposium. Mathematische - physikali sche Klasse , Dritte Folge , 354 pp . Sauer, C.O. (1925). The morphology of landscape. University of California Publications in Geography , 2(2), 19 - 54. Sharpnack, D.A. and Akin, G. (1969). An algorithm for computing slope and aspect from elevations. Photog rammetric Engineering , 35(3), 247 - 248. Slupetzky, H. and Krisai, R. (2009). Indications of Late Glacial to Holocene fluctuations of Cathedra l Massif Glacier, Coastal Range (northern British Columbia, Canada). Zeitschrift f ür Gletscherkunde und Glazialgeolo gie , 43/44, 187 - 212 . Smith, M., Paron, P., and Griffiths, J. (2011). Geomorphological Mapping: Methods and Applications (1st ed.). Oxford, UK: Elsevier , 612 pp. Sternai, P., Herman, F., Fox, M.R., and Castelltort, S. (2011). Hypsometric analysis to identify spatially variable glacial erosion. Journal of Geophysical Research -- Earth Surface , 116(F3), DOI: 10.1029/2010JF001823. of slope development viewed in the light of recent quantitative investigations. Annals of the Association of American Geographers, 40, 209 - 213. Strahler, A.M. (1952). Hypsometric (area - altitude curve) analysis of erosional topography. Geological Society o f America Bulletin , 63, 1117 - 1142. Thorn, C.E. (1976). Qualitative evaluation of nivation in the Colorado Front Range. Geological Society of America Bulletin 87, 1169 1168. Thorn, C.E., and Hall, K. (2002). Nivation and cryoplanation: the case for scrutin y and integration. Progress in Physical Geography , 26, 533 550. Thornbury, W.D. (1965). Regional Geomorphology of the United States . New York: Wiley, 609 pp. 149 Till, A.B., Dumoulin, J.A., Werdon, M.B., and Bleick, H.A. (2011). Bedrock geologic map of the Sew ard Peninsula, Alaska, and accompanying conodont data. U.S. Geological Survey Scientific Investigations Map 3131. Tricart, J. (1967). Le modelé des régions périglaciaires. In Tricart, J. and Cailleux, A., Trait é de g é omorphologie , 2. Paris Sedes, 512pp. Tricart, J. and Cailleux, A. (1972). Introduction to Climatic Geomorphology . New York , NY : St. Martin's Press . 295 pp. Troll C. ( 1944 ) . Strukturböden, Solifluction und Frostklimate der Erde. Geologische Rundschau 34: 545 - 694. English translation: Structure Soils, Solifluction, and Frost Climates of the Earth. U.S. Army Snow Ice and Permafrost Research Establishment, Wilmette, IL., 1958, 121 pp. Turner, M.G. (2005). Landscape ecology in North America: Past, present, and future. Ecology , 86, 1967 - 1974. Twidal e, C.R. (1972). Landform development in the Lake Eyre region, Australia. Geographical Review, 62, 40 - 70. Wagner, K. (2017). Geographic Information Systems and Glacial Environments. In : Menzies, J. and van der Meer, J.J.M. (Eds.) 503 - 536 pp. Past Glacial E nvironments (Second Edition) , Elsevier, Cambridge, MA. Wahrhaftig , C. (1965). Physiographic Divisions of Alaska. U.S. Geological Survey Professional Paper 482 , 52 pp. + maps, appendices. Wahrhaftig, C. and Cox, A. (1959). Rock glaciers in the Alaska Range. Geological Society of America Bulletin, 70, 383 - 436. Waibel, L., (1933). Was verstehen wir unter landschaftskunde. Geographical Anz eiger ., 34, 197 - 207. Walker, H.J. (1983). E. pluribus unum: Small landforms and the Arctic landscape. In : Gardner, R. and Scoging, H. (Eds.) 39 - 55 pp. Mega - Geomorphology. Oxford, UK: The British Geomorphological Research Group, Clarendon Press . Walker, D.A., Epstein, H.E., Gould, W.A., Kelley, A.M., Kade, A.N., Knudson, J.A. , Krantz, W.B., Michaelson, G., Peterson, R.A., Pin g, C.L., and Raynolds, M.K. ecosystems: complex interactions between landforms, soils, vegetation and climate. Permafrost and Periglacial Processes , 15 (2), 171 - 188. Washburn, A.L., (1956). Classification of patterned ground and review of suggested origins. Geologic al Society of America Bulletin, 67, 823 - 865. 150 Washburn, A.L. (1980). Geocryology: A S urvey of P eriglacial P rocesses and E nvironments . New York , NY : John Wiley and Sons. 406 pp. Whitehead, K., and Hugenholtz, C.H. (2014). Remote sensing of the environment w ith small unmanned aircraft systems (UASs), part 1: A review of progress and challenges. Journal of Unmanned Vehicle Systems , 2 (3), 69 - 85. Willgoose, G., and Hancock, G. (1998). Revisiting the hypsometric curve as an indicator of form and process in transp Earth Surface Processes and Landforms: The Journal of the British Geomorphological Group , 23 (7), 611 - 623. Wilson, L. (1968). Morphogenetic classification. In Fairbridge, R.W. ( e d.) The encyclopedia of geomorphology . New York: Reinho ld Book Corporation , 717 - 729 . Wiltse, M.A., Reger, R.D., Newberry, R.J., Pessel, G.H., Pinney, D.S., Robinson, M.S., and Solie, D.N. (1995). Bedrock geologic map of the circle mining district, Alaska. 1:63,360. U.S. Geological Survey. Withers, M.A., and Meentemeyer, V. (1999). Concepts of scale in landscape ecology. In Landscape Ecological Analysis (pp. 205 - 252). Springer, New York, NY. Wood, W.F., and Snell, J.B. (1960). A quantitative system for classifying landforms . Quartermaster Research and Engineering Command , Report No. TR - EP - 124. Natick, MA . Wu, J., and Hobbs, R. (2002). Key issues and research priorities in landscape ecology: an idiosyncratic synthesis. Landscape E cology , 17 (4), 355 - 365. Zakrzewska, B. (1967). Trends and methods in land form geography. Annals of the Association of American Geographers , 57 (1), 128 - 165.