AHETHCD FER 1 .—UTILIZI~' Tne technoh :: pEanners and 01 tez‘mica1 expertis 2:;ears obscure. We, in a non-ts 5’311-sca1e co1or an area in Michige ird11-sca1e (1:112 Vincipefl data so». 13-51jet aircrafl 53339 Administrat‘ my Remte Sensir ABSTRACT A METHOD FOR ANALYZING THE ENVIRONMENT OF AN URBANIZING AREA --UTILIZING HIGH ALTITUDE COLOR INFRARED PHOTOGRAPHY by M. Susan Beckman Thomason The technology of remote sensing has recently become available to planners and others interested in the land resource. Without technical expertise, the utility of remotely sensed imagery Often appears obscure. The objective of this research effort is to ex- plore, in a non-technical manner, the usability of high-altitude, small-scale color infrared imagery in analyzing the environment of an area in Michigan experiencing increased development pressure. Small-scale (1:113, 500) color infrared photography was used as the principal data source for this study. The imagery was taken by an RB-57 jet aircraft on May 7, 1974 for the National Aeronautic and Space Administration and was Obtained by the Michigan State Univer- sity Remote Sensing Project. The study begins by providing a basic introduction to pertin- ent remote sensing concepts and briefly discusses film, photo inter- pretation, and availability of imagery in Michigan. It then reviews the state-of—the-art in environmental analysis methodology focusing on the purposes for which methods have been developed and the types of environmental factors which are used in making environmental evaluations. The study then describes the test site or study area in Michigan ‘Sr thCh an i 13erd5ve1ope entfreU 1'” m !53’3230- 5’” I’f stuff area :Esfificaticn The claS :dorinfrared fiatare 163 a :‘re ifegery ET“ :Lne: relief :muege, amen Efiifl concerr immature and 1 Touawith lim‘ An eva1ue We and the in W are discus Inventoried M. Susan Beckman Thomason for which an inventory-classification system and analysis method are later developed. The study area is 144 square miles in size and lies entirely in Allegan County between the cities of Grand Rapids and Kalamazoo. Growth characteristics and the environmental setting of the study area are discussed, prior to development of an inventory classification system. The classification system developed for use with the small—scale color infrared imagery in the study area inventories parcels of land that are 160 acres in size. Such a parcel is .25 by .25 inches on the imagery employed. The major categories of data inventoried in- clude: relief, land cover, land use, transportation routes, drainage, amenities/disvalues, residential settlement patterns and special concerns. The existing study area is primarily agricultural in nature and the inventory-classification system is designed for those with limited experience in photo interpretation. An evaluation of the physical environment of the study area is made and the implications for accommodating future development are noted. The general interactions between development and the environ- ment are discussed and followed by a five-step method for analyzing the inventoried data. The method focuses on existing community de- velopment as it appears on the photographs, and identifies environ- mentally sensitive features and potential development constraints. Use of the method is illustrated in the assessments made for all existing communities in the area as well as other non-community- related locations. Observations and recommendations are advanced for all development areas. Findings from the research effort conclude that small-scale :iorinfrared sr'drcnr'enta1 a Ise‘ul inforfai certification caee1opnent. E :3':erning urba .-.-_~T;‘n should re cezisions reg-31 Search effort WET! are at if M. Susan Beckman Thomason color infrared imagery provides substantial detail for a general environmental analysis. The inventory and analysis method generates useful information about existing development patterns and enables identification Of potential problems regarding accommodation of future development. Such information may be used to direct new policies concerning urban growth and expansion or to highlight locations which should receive additional environmental study prior to making decisions regarding development or non-development. The entire re- search effort is most applicable to regional studies covering areas which are at least several townships in size. A .VETHCID F1 --UTIL A METHOD FOR ANALYZING THE ENVIRONMENT OF AN URBANIZING AREA --UTILIZING HIGH ALTITUDE COLOR INFRARED PHOTOGRAPHY By M. Susan Beckman Thomason A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER IN URBAN PLANNING School Of Urban Planning and‘ Landscape Architecture 1977 The writer 1! flaming at Michig sac-em effort. Sar‘ard Farness a! Planning and Land graduate studies. I wish to e Timers of the Mi For1973 to 1975 tant in the Renot hronautic and S; 5111 (NASA NGL 22 To my husba socouragement thI ACKNOWLEDGMENTS The writer wishes to thank Myles Boylan, Professor of Urban Planning at Michigan State University who served as advisor in this academic effort. Sincere appreciation is also extended to Professor Sanford Farness and all other faculty members of the School of Urban Planning and Landscape Architecture for their guidance during my graduate studies. I wish to express my appreciation to Bill Enslin and other members of the Michigan State University Remote Sensing Project. From 1973 to 1975, the writer served as a Graduate Research Assis- tant in the Remote Sensing Project which is funded by a National Aeronautic and Space Administration grant to Michigan State Univer- sity (NASA NGL 23-004-083). To my husband Arlen, I wish to extend special thanks for his encouragement throughout preparation of this thesis. ii ‘1' i: ~n~- r bod I i IHULL .STOF FIGUFP G ‘5‘ .H.:‘AED‘I ’ PIT I. ‘ “JV ‘ U1‘ 0 "'~".7~ J‘rrliR I IMTRQSU FIT? Phot AUDI Surr 11 STATE.( METH335 Com; SPEc SOL, TABLE OF CONTENTS Page LIST OF TABLES ........................ v LIST OF FIGURES ....................... vi INTRODUCTION ......................... 1 CHAPTER I INTRODUCTION TO REMOTE SENSING ............ 6 Film ........................ 11 Photo Interpretation ................ 12 Availability of Imagery .............. 15 Summary ...................... 16 II STATE-OF-THE-ART REVIEW OF ENVIRONMENTAL ANALYSIS METHODS ........................ 19 Comprehensive Methods ............... 21 Special Purpose Methods .............. 30 Sources with Conceptual Input to the Study of Environmental Analysis .............. 32 Summary ...................... 33 III SELECTION AND DESCRIPTION OF THE STUDY AREA ...... 35 Growth Characteristics of the Study Area ...... 37 Environmental Characteristics of the Study Area . . 42 Summary ...................... 50 IV INVENTORY AND ANALYSIS METHOD ............. 53 Development of the Inventory ............ 53 Interpretation Methods .............. 59 Classification System ............... 68 Field Checks .................... 78 Summary ...................... 79 V EVALUATION OF THE PHYSICAL ENVIRONMENT OF THE STUDY AREA AND IMPLICATIONS FOR DEVELOPMENT ......... 81 Interactions Between Development and the Environment ................... 83 Analysis Method and Application to Study Area . . . 93 Analysis of Community-Related Development ..... 98 Analysis of Non-Community-Related Development . . . 110 Summary ...................... 112 H CCNCLUSIO Usatii Invent Oeve Study 3.5391311 A Ni 53 19‘.- ~.":.‘JIX B Codi fi' , _ . - :.3LICSF‘APHY . . CHAPTER Page ‘ VI CONCLUSIONS AND RECOMMENDATIONS ............. 115 Usability of Color Infrared Imagery ......... 115 Inventory/Analysis Method and Implications for Development .................... 117 Study Area Recommendations .............. 120 APPENDIX A Mission 273 Flight Specifications ......... 123 APPENDIX B Coding Sheets and Recorded Inventory Data ..... 126 BIBLIOGRAPHY .......................... 147 iv Study Ar Employre EmpIDyTe Michigan 11 . . . Coding 5‘ Examples Specific; Approxir; LIST OF TABLES Table . Page 1 Study Area Population Growth .............. 4O 2 Employment by Place of Work Data Summarized ...... 41 3 Employment by Place of Work Data ............ 43 4 Michigan Land Use Classification System - Levels I and II ........................... 55 5 Coding System ..................... 57 6 Examples of Field Checked Sites ............ 78 A-l Specifications for Mission 273, Roll 10 ........ 124 A-2 Approximate Conversion Factors for 1:113,SOO ...... 125 E1ectronagne ERTS-l Cover R8-57 COIOr Study Area 1 Kalamazoo Ri Land Divisic R3‘57 Color tion ActuaI RB-57 CO1or tion Actual RB~57 Color tiOn Actual RB‘57 COIOr “0" ACtua1 GFIdded StUC Eva] “at 1' On I‘ Figure 0'1wa 03 10 11 12 A-1 LIST OF FIGURES Electromagnetic Spectrum ................ ERTS-l Coverage of Lansing - Grand Rapids Area ..... RB-57 Color Infrared Imagery .............. Study Area in Allegan County .............. Kalamazoo River Drainage Basin ............. Land Division Map of Study Area ............ RB—57 Color Infrared Image: Black and White Reproduc- tion Actual Size Mission 273 - Frame Number 10-0046 RB-57 Color Infrared Image: Black and White Reproduc- tion Actual Size Mission 273 - Frame Number 10-0045 RB-57 Color Infrared Image: Black and White Reproduc- tion Actual Size Mission 273 - Frame Number 10-0044 RB-57 Color Infrared Image: Black and White Reproduc- tion Actual Size Mission 273 - Frame Number 10-0043 Gridded Study Area ................... Evaluation Map of the Study Area ............ Mission 273 Flight Line Coverage ............ vi Page 7 9 13 38 45 48 6O 61 62 63 65 95 123 INTRODUCTION In recent years there has been an increasing interest in the natural environment. According to the First Annual Report Of the Council on Environmental Quality: Historians may one day call 1970 the year of the environment. They may not be able to say that 1970 actually marked a sig- nificant change for the better in the quality of life; in the polluting and fouling of the land, the water, and the air; or in health, working conditions, and recreational opportunity. Indeed, they are almost certain to see evidence of worsened environmental conditions in many parts Of the country. Yet 1970 marks the beginning of a new emphasis on the environ- ment--a turning point, a year when the quality of life has become more than a phrase; environment and pollution have be- come everyday words; and ecology has become almost a religion to some of the young.1 This new awareness and interest is largely a response to prob- lems that have been compounded over many years. These problems in- clude: the unrestricted growth of cities, scatteration of development, consumption of prime agricultural land and open space and, in general, a wasting Of our resources. The increased emphasis on the environment is having profound effects on the planning profession. The influence is especially strong in the area of land use allocation because of the close rela- tionship that exists between "natural qualities of the land and the activities making use of the land."2 According to an Environmental Protection Agency report issued in February 1974, land use planning is now evolving on three fronts with: 1) an increasing concern for the natural environment and a concomitant questioning of the economic assumption that growth is always beneficial; 2) a search for information on how to allocate urban activ- ities while preserving the integrity of environmental systems; and 3) a shift in approach from the long range master plan to implementation and action.3 These new considerations in land use planning reflect the need to improve the decision-making processes which affect the allocation and use of land and natural resources. Historically, land use de- cisions have been made on the basis of short-term economic consider- ations and political exigency. Numerous public officials and concerned citizens fear that many decisions of public concern are being based on factors "unrelated or contrary to the real concerns of sound land use policy."4 Policy conflicts that arise are generally settled by negotiation and political bargaining rather than through the use of empirical information on environmental implications of proposed decisions. In order to improve the decision-making process, reliable data on the physical environment is essential. This data should include information on components such as: soils, land use, vegetation, topography, water characteristics, settlement patterns, unique physical features, and location of transportation routes. The data must then be synthesized so that an evaluation of the area can be made. Various methods have been developed for analyzing the physical- environment. A state-of-the-art review of these methods reveals 2.23:: no one aPPI‘OE :‘s‘:uation. A: 52‘1ity Of inform :‘feIiness, cost, ‘:.srethods ind: :costraints to "r :erhaps a functic istic of the new Remote sens serve as an extra 1‘9 Physicai env Rerote sensors 1' theme SYSIEm WWII/e 0f thi sca1ec010r Infr imajgr data $01. $36 by MP”Iners 391151119 Whose m that no one approach has been found to be the best for every type of situation. A common limitation of all the methods is the avail- ability Of information on the physical environment including its timeliness, cost, scale and quality. Many applications of the var- ious methods indicate that the users were forced by time or other constraints to "make do" with the best data available. "This is perhaps a function of the urgency of information demands character- istic of the new environmental emphasis in planning."5 Remote sensing is a recently developed technology which can serve as an extremely valuable means of obtaining information on the physical environment since large areas can be rapidly covered. Remote sensors include multi-spectral scanners, radar, passive microwave systems, high altitude aerial cameras and others. The Objective Of this study is to investigate the utility of small scale color infrared imagery taken from high altitude aircraft as a major data source for environmental analysis. It is intended for use by planners and others with little technical expertise in remote sensing whose need for the ability to make environmentally informed decisions regarding development at a large scale would benefit from the use of remote sensing technology. The material of this thesis is presented in the following components: (1) a brief introduction to remote sensing providing a basic understanding of pertinent remote sensing concepts; (2) a state-of-the-art review of environmental analysis methods to determine the types of assessments which have been made, why they have been developed, and what envir- onmental factors they have considered; (3) selection and description of a site in Michigan covered by small-scale high—altitude color-infrared imagery to be used as a study area in which to test a new inventory and analysis method; (4) development and application of the inventory and analysis method in the study area; (5) evaluation Of data generated from the inventory/analysis method and its implications for accommodating future devel- opment or non-development within the study area; and (6) expansion of the findings from the above components to generate workable indicators Of environmental conditions and implications for development inside and outside the study area. Throughout Michigan, remotely sensed imagery is increasingly being used by various planning organizations. Present users of remote sensing include state natural resource and highway Officials, both state and regional land use planning personnel, as well as private citizens with a diversity of interests. The increasing use of such imagery by so many indicates that planners not yet knowledgeable of the technology will likely be exposed to it in the foreseeable future. Despite historically overenthusiastic claims for magic results from remote sensing technology, it must be emphasized that it is only a tool, but a tool that can reduce the time and cost involved in assembling and analyzing contemporary and useful data on the physical environment. Remote sensing is envisaged as a tool that can be used by professional planners, planning commissions, city councils, county boards, state legislatures, park and recreation officials, land use commissions, students, as well as business and industrial organizations. 1Council Iiashington, D 2Marion Pmss,1965), 3Edward R.Stan1and, P and Controls T 131+ , PP. 133 4. _ uohn F. :ccnonics EXP-‘5 S, . kaiser FOOTNOTES 1Council on Environmental Quality, Environmental Quality. (Washington, D.C.: Council on Environmental Quality, 1970), p. 5. 2Marion Clawson, Land Use Information (Baltimore: John Hopkins Press, 1965), p. 17. 3Edward Kaiser, K. Elfers, S. Cohn, P. Reichert, M. Hufschmidt, R. Stanland, Promotinngnvironmental Quality_Through Urban Planning and Controls (Washington, D.C.: Environmental Protection Agency, 1974), pp. 108-109. 4John F. Timmons, Guidelines for Developing_$tate and National Public Land Use Policy, (Ames, Iowa: Iowa Agriculture and Home Economics Experiment Station), p. 4. 5Kaiser gt_gl,, p. 133. In very tne nature Of a definition, sensing devic with the tem “the detectic distant 59,15. nal scanners Mgh‘aultudi Light, Energy are CI tIPes are di‘ continuum in the SIEctrOmi rent1ya T‘Em fr0m the mm 82111415 11.1 eql d1 ffeI‘Ent ban CHAPTER I INTRODUCTION TO REMOTE SENSING In very general terms, remote sensing is defined as "detecting 6 Under such the nature Of an Object without actually touching it.f a definition, the camera can be considered as the original remote sensing device. With the elaborate technology currently associated with the term, remote sensing can more functionally be defined as "the detection, recognition or evaluation of objects by means of distant sensing or recording devices."7 Such devices include ther- mal scanners in orbiting satellites and aerial cameras mounted in high—altitude aircraft. Light, x-rays, radiant heat, radio waves and other forms of energy are composed of electromagnetic waves. The various energy types are differentiated by their wavelengths and frequencies. The continuum in which each energy type is assigned a range is known as the electromagnetic or EM spectrum. In remote sensing, energy types are conventionally distinguished by wavelength (see Figure 1). Cur- rently, remote sensing concentrates on wavelengths in the range of .3 microns to about 3 centimeters, although the EM spectrum ranges from the minute gamma rays to the very long radio waves. One micron equals lu equals .OOOOOOl meter. Remote sensors record both reflected and emitted energy in different bands of the EM spectrum. Ground objects reflect and_ 6 Inc. In. I: {1111 ‘00. \O— £..¢I.O)O§ EEOQ‘ 'EC— I Ezguumam owpmcmmsogpumpm .p mgsm_m «coco... .. aw. am at» m. a... v. 1.1 4. u u u I. u n n J.“ a u . e _ . o . N “L“ E . E u L v - E — M .L- E n w o m u .. m o m..." u m .. m v A H n ”L. n t. . ..l 1 3332mm»... /, mam: / pro... mums"; s >3 \\ I II I s I I II II Q \ I II I s \ I I II I: s \ i I II [I s I I III It: s. \ I I I III! III s \\ I I II II \ \ I II I s \ ’ I I Q I I II s \ I I II I \ I, II II ~ \ R b r P F 1? H H F 1? W't" ’H’P’ ~F\ F L r 1’1 ’ m m w at: £403. «a... u " huJo; >¢¢ A. . . 0.048 u><30¢o§ ouzficuz.“ (Chi.- ><¢ I x (8340 I" (m A. 5.409 5.2 I... 3.00.. 6.9 (1. two. er. H... :13. amp. m. 5:100. tho. Eu. «.1 «m «(6 50:22.0} an: ener-QO’ 1” d. gymsition. Be< sensor to record ‘erentiated. Th1 tcnal and/Dr coll The greater the inshed. The ch navelength band Iced by remote EIEQh is One typ In the EM 3313-1. sensors III-areas conventi g'aphl'c emulsior 0r emitted energ rem"WIT! tre The human Spear“: 1.e. i that range beCaL 1'2 mic"Ons. dtIECted by ”On- One tYDe ( StaImEr (I188 ) 8 emit energy in different amounts as a function of time and material composition. Because of these differences, it is possible for a sensor to record a scene so that objects can be identified and dif- ferentiated. The appearance of Objects is generally rendered in tonal and/or color contrast between the Object and its background. The greater the contrast, the more easily the object can be distin- guished. The characteristic appearance of an object in a particular wavelength band is known as its tonal signature. The displays pro- duced by remote sensors are known collectively as imagery. A photo- graph is one type of imagery which records visible light on film. In the EM range in which remote sensing concentrates (.3u to 3 cm.), sensors include cameras, scanners, radiometers and others. Whereas conventional cameras directly record an image onto a photo- graphic emulsion or film, scanners and radiometers record reflected or emitted energy signals as electronic impulses which can then be mechanically translated into a photograph-like image. The human eye is sensitive to only a small region of the EM spectrum, i.e. from .4 to .7 microns. Photographic film expands that range because it is made sensitive to wavelengths from .3 to 1.2 microns. Wavelengths of greater or lesser magnitude must be detected by non-photographic sensors. One type of non-photographic sensor is the multi-spectral scanner (M55). The M55 is a line scanning device which uses an oscillating mirror to simultaneously scan the terrain passing beneath the spacecraft in several wavelength bands. The scanner produces four synchronous images, each at a different wavelength band of the EM spectrum. Figure 2 is an example of the imagery produced by the lANSlEI‘ IAN ERTS'I COVERAGE OF lANSING ' GRAND RAPIDS AREA Figure 2 I55 aboard the firs LAISSAI - 1). Sam .56 to .63» (yello to .73.. Band 6 i .81. and Band 1 t; tar-d accentuates Band A, the short “Penetration and ' ination of depth 335d 5 senses a “I atmsnheric 0f the land sm 10’19 Infrared reflected by 1 rapid (“SC-Tin. ADOthey known as a t1 251 mums. EEIttEd by We active supp]ies can dfite. “E “\St “SEQ s 10 M55 aboard the first Earth Resources Technology Satellite (ERTS/ LANDSAT - 1). Band 4 or Channel 4 is sensitive to wavelengths of .50 to .60u (yellow-green light). Band 5 records red light at .60 to -70u- Band 6 is sensitive to near infrared radiation of .70 - .80u and Band 7 to intermediate infrared of .80 to 1.10u. Each band accentuates different characteristics Of the earth's surface. Band 4, the shortest wavelength, permits the greatest underwater penetration and hence is sometimes useful in qualitative discrim- ination of depth and/or turbidity in standing bodies of water. Band 5 senses a longer wavelength and as a result reduces the amount of atmospheric attenuation or blue light permitting a sharper view of the land surface than does Band 4. Bands 6 and 7 record the long infrared wavelengths. The large amounts Of infrared energy reflected by land features and absorbed by water bodies enables rapid discrimination of the land/water interface with Bands 6 and 7. Another type of scanner senses radiant heat rather than light. Known as a thermal scanner, this device senses in the range of 3 to 20 microns. Temperature differentials in fractions of a degree Centigrade can be detected with this instrument. Passive microwave instruments record the energy naturally emitted by surface features in wavelengths of .75 mm to 1 meter. The active counterpart is radar; active because the sensor itself supplies the initial energy signal and records its return. Radar can detect longer wavelengths of 1 mm. to 3 meters. Despite the existence of and increasing research on highly sophisticated sensors, the camera is still the most extensively used form of remote sensing. The modern aerial camera is generally equipped with ve 11.1tip1e lens ca of different fil nost widely acq: ible and reflect soectrum as rec< Black and “0*" as Panchn Black and white near infrared we CORVEHLIOI colors 51m11ar 1 aerial photogral fively easy to . The fnm 1 sing is c010,. 1., This type of fl" Camuflage becaL cantly differem records infrarec alth0ugh 1!: Canr over com,entmna altitudes. N AS A 50.000 feet. 60 ate LEIlUdtes COn v 11 equipped with very high quality lenses and rapid shutter speeds. Multiple lens cameras are available and permit the simultaneous use of different film/filter combinations. With the aerial camera, the most widely acquired and usable type of imagery is that of the vis- ible and reflectance infrared (not thermal) portions of the EM spectrum as recorded on film. Film Black and white film sensitive to all the visible colors and known as panchromatic is the most familiar type of aerial coverage. Black and white infrared film is also available and is sensitive to near infrared wavelengths. Conventional color film, film which renders an object in colors similar to those seen by the human eye, is also used in aerial photography. Because the colors are familiar, it is rela- tively easy to identify ground objects. The film type that is of increasing importance in remote sen- sing is color infrared film (CIR) also known as false-color film. This type of film was originally used by the military for detecting camouflage because natural vegetation and camouflage have signifi- cantly different infrared reflective properties. Because CIR records infrared wavelengths, it is able to penetrate haze and smog, although it cannot penetrate clouds. This is a substantial advantage over conventional color, especially when coverage is taken at high altitudes. NASA-sponsored RB-57 jet aircraft typically fly at 60,000 feet. 60,000 feet of smog and haze over an urban area severely attenuates conventional color film which is sensitive to the relatively snort Haw-length The Chemic print reflected tight as blue. on the average» wavelengths. 81 the imagery ”I“? are 3 shows the filters. The to infrared film an dent in this imau darker yellow fi' tontrasts are mu: demonstrates the Land/Hater inter-1 tI'nsuished by the urban and non-urb color infrared fi Sensing tools in 1971), 12 short wavelength Of blue light. The chemical composition of CIR film has been designed to print reflected infrared energy as red, red light as green and green light as blue. Vegetation appears in varying shades of red because, on the average, vegetation reflects twice as much infrared as green wavelengths. Blue and violet light is reduced or eliminated from the imagery through the use of yellow ("minus blue") filters. Fig- ure 3 shows the tonal difference resulting from use of different filters. The top image at a scale of 1:120,000 was taken with infrared film and a light yellow filter. A blue-green cast is evi- dent in this image. The lower image was shot simultaneously with a darker yellow filter. Most of the blue has been removed and color contrasts are much more pronounced. This example of tonal contrast demonstrates the important differentiations possible with CIR film. Land/water interfaces are dramatic and vegetation types can be dis- tinguished by the varying shades of red. Differentiating between urban and non-urban areas is also made easier. For this reason, color'infrared film is becoming one of the most widely used remote sensing tools in urban and land use analysis (Dueker and Horton, 1971). Photo Interpretation Many of the more sophisticated remote sensors record electro- magnetic energy as electronic impulses which must be mechanically converted to a tonal image. For example, the ERTS I (LANDSAT 1) images in Figure 2 were produced by an electron beam recorder which translated the electronic impulses that had measured intensities of Muskegon, Michigan Top Photo - 1:120,000 Bottom Photo - 1:60,000 Figure 3. RB-57 Color Infrared Imagery reflected l earth. The is that dor. and delinea cludes fiel The amount irage is 1a image, the available, analysis of 5129. shape 15 Primaril from the 99. lng; We ”10$“ Our knowled. Studia With a Part. POrt by Shel l4 reflected light from satellite, to a photograph-like image here on earth. The most prevalent type of image interpretation, however, is that done manually. Interpretation involves the identification and delineation of features from the photographs or images and in- cludes field checks and use of available collateral information. The amount of information that can be extracted from a particular image is largely dependent on the scale and quality of the aerial image, the ability of the interpreter, and the collateral data available,e.g. topographic maps. Interpretation is based on an analysis of the apparent properties of surface features including size, shape, tone, texture, pattern, shadow and association. It is primarily a deductive process in which the interpreter proceeds from the general to the specific. But the imagery "gives us noth- ing; we must extract information from the imagery as a function of our knowledge and core of experience."8 Studies have repeatedly shown that the person most familiar with a particular area is often the best interpreter, eventhough previously untrained in image interpretation. According to a re- port by Shelton gt_a1,, Ideally, the user should do the interpretation; he would, during that process gain an understanding of the study area far deeper than is usually possible by other methods. He also would have an advantage over most interpreters by virtue of his existing knowledge in the study area. An interpreter is doing more than just recognizing certain land uses or phys- ical features; in the case of land use particularly, he is interpreting events and conditions as they affect land use and as they have observable characteristics which serve to identify them.9 There a currently ava imagery from black and whi (one inch equ frai'e depicti type of image of LAADSAT fr n0t substanti as color cor; Quentially e; 511 a scale 01 all LANDSAT 5 A seCOr PrOsran whicl ”3'" ”Ar 22, 15 Availabilitygof Imagery There are four principal types of remotely sensed imagery currently available in Michigan. At the smallest scale is MSS imagery from ERTS, now known as LANDSAT (refer to Figure 2). This black and white imagery at an approximate scale of 1:1,000,000 (one inch equals 16 miles) generally comes in 10" x 10" frames--each frame depicting an area of approximately 115 miles square. This type of imagery is suitable for macro-regional analysis. Enlargement of LANDSAT frames is possible, but "detail on interpretability does not substantially increase."10 Selected MSS images can be reproduced as color composites. Three black and white spectral bands are se- quentially exposed through different color filters onto color film at a scale of l:l,000,000, but color composites do not exist for all LANDSAT scenes. A second type of imagery was produced by the NASA Skylab Program which consisted of one unmanned and three manned missions from May 22, 1973 through February 8, 1974. The spacecraft travelled in an equatorial orbit 270 miles above the earth and crossed Mich- igan in a northeast-southwest direction. Most of the southern half of the lower peninsula was covered at scales of 1:1,000,000 and/or 1:3,000,000. Skylab imagery consists of black and white, color and color infrared photography. The third type of imagery, and the type with which this thesis is primarily concerned, is high altitude, color infrared photography. This coverage is obtained principally by RB-57 jet aircraft at an altitude of approximately 60,000 feet (refer to Figure 3). Two camera systems typically provide imagery at scales of 1:60.000 and 1:120,030. A1 The individual Fear frames at to the size 01 The four altitude photc and Conservati Each frare is 1:15.840, 1:23 Post widely us Addition Tent a“deficies the US. Fores Department of DEPAItment of Imagery as We] The tech: [071319 . X. 8 U1: a expert to Util- Ing pemllt the Able fl" 16 1:120,000. At 1:60.000, one inch is approximately equal to one mile. The individual frames of this CIR imagery are generally 9" x 9". Four frames at the scale of 1:120,000 cover an area almost equal to the size of Eaton County,Michigan.]]' The fourth type of generally available imagery is medium- altitude photography provided by the Agricultural Stabilization and Conservation Service (ASCS) and other government agencies. Each frame is typically black and white panchromatic at scales of 1:15,840, l:20,000 or 1:40,000. This type of imagery has been the most widely used in various types of planning activities. Additional aerial imagery is available from various govern- ment agencies at the state and federal levels. Examples include the U.S. Forest Service, U.S. Geological Survey, the Michigan Department of State Highways and Transportation, and the Michigan Department of Natural Resources. The types and scales of their imagery as well as the areas covered, are variable. Summary The technology of remote sensing is highly sophisticated and complex. But as the "canned program" permits the non-computer expert to utilize the computer, so too can some introductory train- ing permit the layman to effectively use much of the output avail- able from remote sensors. The primary importance of remote sensing lies in applications rather than hardware; integrating remotely- sensed data with other conventional sources of resource information to plan, to make policy, and to act. Remote sensing, in its current state of development, cannot supply all the Oh; sis of the envirov and many organiza' use could substan sensed inagery is inventories; to r in time. The ad.- ies include its t is the intent of however, where Inc of remote SEIlSlna scape; processes The HldeSp which deVETOped covery and new d andHSis- The n these m9thods to l7 supply all the physical information necessary for a complete analy- sis Of the environment. Its capabilities are extensive, however, and many organizations concerned with natural resources and land use could substantially benefit from its use. Increasingly, remotely sensed imagery is being used for land use/cover and natural resources inventories; to record what is on the ground at a particular point in time. The advantages to using remote sensing for such inventor- ies include its timeliness, large area coverage and economy. It is the intent of this thesis to go beyond the inventory state, however, where most work has been done to investigate the utility Of remote sensing in processes which analyze and evaluate the land- scape; processes which are referred to as "environmental analysis." The widespread attention on the natural and cultural environment which developed in the late 1960's and early 1970's spurred redis- covery and new development Of numerous methods for environmental analysis. The next chapter presents a state-of-the-art review of these methods to provide an understanding of the environmental anal- ysis process and serve as a basis for development of the environmental analysis method in Chapters IV and V. 6James P. Sch Sensing for EDVIrnn Wsconsin, 1916, 71. Eugene Av o1is,Hinnesota: E 8 . (LP. Weacher uEnvironnental Qu Protection Agency, 9R.L. Shelton Natural Resources I lcrk: Center for A mliark c. Sull Use Classification tniversity, 1973 , H1bid., p, 5, 18 FOOTNOTES 6James P. Scherz and Alan R. Stevens, An Introduction to Remote Sensing for Environmental Monitoring (Madison, Wisconsin: University of Wisconsin, 1970), p. 2. 7T. Eugene Avery, Interpretation of Aerial Photographs (Minneap- olis, Minnesota: Burgess Publishing Company, 1968), p. 135. 8C.P. Weatherspoon et al., Proceedings of the Second Conference on Environmental Quality Sensors (Washington, D.C.: Environmental Protection Agency, 1973), p. v-60. 9R.L. Shelton et al., Land Use, Environmental Features and Natural Resources Inventory_of the Hudson River Valley (Ithaca, New York: Center for Aerial Photographic Studies, 1969), p. CL-2. 10Mark C. Sullivan, Image Interpretation for a Multi-Level Land Use Classification SystemT(East Lansing, Michigan: Michigan State University, 1973), p. 5. 11 Ibid., p. 5. STATE OF THE Land use and environmental aware United States. Em established in ma”) land. Among the Inc achusetts, river ba shorelines in Wiscc and tallies, 1971). tending respect for the integration Of the environment. CHAPTER 11 STATE OF THE ART REVIEW OF ENVIRONMENTAL ANALYSIS METHODS Land use and natural resources planning based on increased environmental awareness has been rapidly developing across the United States. Environmentally protective legislation has been established in many locations for various categories or types of land. Among the more prominent are protection of wetlands in Mass- achusetts, river basins in New England, bay areas in San Francisco, shorelines in Wisconsin and agricultural land in Hawaii (Bosselman and Callies, 1971). Various organizations and individuals are de- manding respect for the ecosystem, protection of open space, and the integration of development plans with provisions protective of‘ the environment. The increased emphasis on protection of the natural and human environment has stimulated the development and rediscovery of methods for analyzing and evaluating man's biotic and abiotic surroundings. This chapter describes eighteen environmental analysis methods which can be considered representative of the state-of-the-art. The vast najority of them were developed in the late 1960's and early 1970's. The process of environmental analysis consists of essentially three steps. The first involves the identification of a problem or formulation of a purpose to which an analysis method is designed to respond. The second step is an inventory of pertinent environmental l9 20 data. The third step consists of an analysis and evaluation of the inventory data as it relates to the identified purpose. The eighteen methods described in this chapter were developed for oftentimes substantially different purposes. But despite such differences and although specific definitions and level of detail vary, at least some aspects of the following factors were generally inventoried: topography, soils, land cover/use, geology, water re— sources and climate. The greatest weakness in the methods reviewed in this chapter involves the third step Of the environmental analysis process; i.e., the evaluation of inventoried data. Most methods fail to clearly define the specific steps used in evaluating inventory data; a fact which prevents results from being readily duplicated. It appears that this lack of step-by-step evaluative instruction may be partially intended, underscoring the fact that the evaluation process often requires subjective judgments when specific scientific parameters are unavailable. In addition, there may be a general re- luctance to transpose detailed evaluations of natural environments and man's interactions with them from one geographical region to another because of known and unknown variations which may be present. There is rui singularly best environmental analysis method for all situations. The impressions given above may explain why the eventual development of any universally applicable method is considered un- likely. Such an observation does not diminish the importance of environmental analysis methods. The continuing need for a better understanding of the environment requires refinements and new develop- ments in the methodology of environmental analysis. The envi broadly categ: corparison of priate perspec is identified. tors used to 6 M93, those nany component for rather Spe area fer recre a much narrow because the me to sPQleic [TE approacn t0 til sents writings 1.871113 includln analyses, the addreSSed and tens. lihen ad show he cons 21 The environmental analysis methods which follow have been broadly categorized and are summarized in a common format to permit comparison of methods. In order to place each method in an appro— priate perspective, the purpose for which each method was developed is identified. Then each method is briefly described and the fac- tors used to evaluate the environment are outlined. Comprehensive methods, those which evaluate extensive or sizable areas and/or many components of the ecosystem are discussed first. Next, special purpose methods are considered. They represent techniques developed for rather specific reasons such as appraising the potential of an area for recreational activities. Special purpose methods involve a much narrower analysis than do comprehensive methods. Finally, because the methodology of environmental analysis is not limited to specific methods, a list of various sources with a conceptual approach to the subject has also been provided. This list repre- sents writings which discuss environmental analysis in general terms including topics such as the rationale for conducting such analyses, the basic man-environment interactions which must be addressed and general suggestions on development of evaluative sys- tems. When additional detail is required, the original documents should be consulted. Comprehensive Methods 1. Dee, Norbert, Janet Baker, Neil Drobny and Ken Duke. "An Environmental Evaluation System for Water Resource Planning." Water Resources Research, Vol. 9, NO. 3, (June 1973) PURPOSE: To evaluate the environmental impacts of proposed water resource projects. 22 METHOD: This is a highly detailed method which assigns numerical values to various environmental components in an attempt to quan- titatively measure the impact Of a particular action. The four major categories evaluated include: ecology, environmental pol- lution, esthetics, and human interest. The critical stage in this method is assignment of numerical values and weights. Gentili, Joseph and Mitchell J. Lavine, eds. Owego Environmental Study. Ithaca, New York: Cornell University, 1974. PURPOSE: To provide a basis for decisions relating to proposed alterations of the Owego, New York landscape. METHOD: Initially, several maps of various natural resource features were prepared. The integration of this data was made in a "compatibility matrix" which represents a series of decisions concerning relationships between the inventoried natural features and prospective land uses. The impact of natural systems on land uses is considered. The reciprocal relationship, the im- pact of the land uses on natural systems, is not considered. Ratings resulting from the matrix analysis are: Not Compatible, Low Compatibility, Moderate Compatibility, and Full Compatibility. Hills, G. Angus. The Ecological Basis for Land Use Planning. Ontario, Canada: Ontario Department of Lands and Forests, 1961. PURPOSE: To establish a scientifically derived frame-work for various resource management problems. METHOD: The Angus Hills Method (as outlined in Landscape Archi- tecture (January 1968), p. 147 and here cited verbatim) I. CLASSIFICATION 1. Given a total study area. 2. The total site area is divided into regions which are then further subdivided several times into land areas of decreas— ing size. II. 4. 5. Q'UQVOI—mm A poss physic Each p for ge ments EVALUAT Site t tO pro Landsc evalua (a) I 23 (a) Site Regions, by defining areas of broad climatic similarities; (b) Landtypes, by separating areas of differing geologic composition; (c) Physiographic site classes, by separating areas of differing microeclimate; (d) Physiographic site types, by separating areas of differing micro-landform variation. 3. A possible range of general land-uses is determined, and physical requirements for each use is identified. 4. Each physiographic site type is subdivided into site_phases for general kinds of uses according to the physical require- ments for the use. II. EVALUATION 4. Site types (see 2) are grouped into common landscape units to provide a basis for evaluation. 5. Landscape units are subdivided into landscape subunits to evaluate a particular use with respect for: (a) Its use capability: i) At the local level ratings are established on the basis of the sub-unit's inherent features; ii) At the broad level on the basis of the subunit's geographic context; (b) Its use suitability, on the basis of the sub-unit's present condition. (c) Its use feasibility, on the basis of present forecast socio-economic climate. 6. A land-use is recommended for each landscape sub-unit. 7. Multiple 1and-uses are recommended for each landscape unit. Lewis, Philip Jr. "Quality Corridors for Wisconsin." Landscape Architecture, Vol. 54 (January 1964): 100-108. PURPOSE: To identify and evaluate resource patterns to help guide future growth. METHOD: The Philip Lewis method (as outlined in Landscape Architecture (January 1968), p. 147, and here cited verbatim) 1. Given a total study area. 2. Kinds of uses (or activities) to be planned for are identified and the use requirements (or use criteria) are established. 3. A case study area is selected within which: (a) Resources which meet the above use criteria are identified. (b) Major resources are inventoried and located within case study area (as patterns) or separate transparent overlays. (c) Patterns of major resources are combined into one pattern. (d) Additional resources are inventoried and located (as patterns) on separate transparent overlays. (e) purpc (f) patte (9) and a 4. Inven area. 5. Inven StUGy 6. Point 7. Point 8. Dewar and a 9. Limit are i each Linear pa cases, tn floodplai - McClellan PURPOSE; Urbanizin METHOD; Used. It the Veriol a Potent]:a the f0110h IEPIUreS, 91 ageruI 24 (e) A variety of patterns may be identified for special purposes from a few additional resources. (f) Patterns of additional resources are combined into one pattern. (9) Patterns of major and additional resources are compared and a correlation between the two is established. Inventory of major resources takes place over total study area. Inventory of additional resources takes place over total study area. Points are assigned to major and additional resources. Points are totaled to identify relative priority areas. Demand for planned uses is established and final priorities and areas are defined. Limitations of each priority area to specific kinds of uses are identified and specific kinds of uses are assigned to each area. ‘ LO comm 01 ink Linear patterns or corridors are generally identified. In most cases, these corridors consist of areas containing wetlands, water, floodplains, sandy soils near water, and steep topography. McClellan, G. ed. "Environmental Planning in Waterloo County," The Waterloo CountyaArea Selected Geographical Essays. Waterloo, Canada: University of Waterloo, 1971. PURPOSE: To develop an optimal land use plan for 1990 in an urbanizing area. METHOD: This was primarily an inventory in which overlays were used. It was possible, however, to vary the weights assigned to the various factors analyzed. Areas were identified whiCh had a potentially high value for the development of one or more of the following: a) recreation, b) water quality, c) unique natural features, d) wildlife, e) groundwater recharge, f) forested areas, 9) agriculture, and h) unique cultural features. McClellan, G. ed. "Ecological Concepts of Subdivision Design," The Waterloo County_Area Selected Geographical Essays. Waterloo, Canada: University of Waterloo, 1971. PURPOSE: "To identify and articulate those environmental condi- tions which may influence, either positively or negatively, the various ft METHOD: 1 specialist areas not itative en The in fauna, cli. phology, SI of the Pepi undevelopet 0f the envi standing de t0 identify science inp a” handlin enVlI‘Onment reached the and Sut‘face ' ”who. Ian HallaCe- McH. PURPOSE. T! EXlsting C01 t0 the envll the “intrins "ETHOD: The “(1%! Jan 25 various forms of urban land use." (p. 256) METHOD: This method used an environmental team approach including specialists and a team coordinator. The procedure was to subject areas not yet developed, to an intensive quantitative and qual- itative environmental analysis. The initial inventory included the following factors: flora, fauna, climatic elements, water quantity and quality, geomor- phology, soils, slope patterns and land history. In the words of the report, the "survey defined the natural diversity of the undeveloped areas and in doing so, gave detailed understanding Of the environmental interactions of the site. With the under- standing derived from this comprehensive survey, we were able to identify several situations where phyto-sociological and earth science inputs would have been most useful in improving the over- all handling of development and in creating a more viable natural environment in an urban context. Based on this research, we reached the conclusion that soils, geomorphology, phytosociology and surface water are the constant prime elements in the ecolog- ical survey." (p. 257) McHarg, Ian. Plan for the Valleys. Philadelphia, Pennsylvania: Wallace-McHarg Associates, 1963. PURPOSE: To conduct a resource inventory and to analyze the existing conditions and pressures to demonstrate what can happen to the environment if development proceeds without considering the "intrinsic suitability" of a given area. METHOD: The Ian McHarg method (as outlined in Landscape Archi- tecture, January 1968, p. 147, and here cited verbatim) Diagrammati 1. Given a 2. Feature structu 3. Feature total 5 4. InventO related 5. Limitin and rel. 6. Relativ. 7. Having a formulai 8. Potentia the lim‘ 9. Possible against 10. Land are 11. Conflict by compa 12- Demand f area 1‘80 as suita ' ”filthy. Rich. LEQQ- Thesit PURPOSE: To Ian-made feat residential a METHoo; An 0 following fac ])a9rlCUltUr‘ both ground ar woodland, 6) t dd d N 26 Diagrammatic Outline of the Analysis Procedure Given a total study area. Features of natural or cultural processes are defined and structured into broad resource categories. Features of natural or cultural processes are inventoried for total study area. Inventory of features is supplemented with descriptions of related processes. Limiting factors of each process described are identified and related to their features. Relative value is attributed to processes. Having attributed value to the processes, principles are formulated relative to limiting factors. Potential land-uses are considered and possible effect on the limiting factors is identified. Possible effects of each land-use requirement is weighed against each limiting factor. Land areas suitable to each land-use are compared. Conflicts between overlapping suitable uses are resolved by comparing economic return. Demand for these possible uses is established for their area requirements, and compared with other areas identified as suitable. do to 0) NOW 01 b no N—0 Maltby, Richard A. A Method for Determining_the Use Potential of Land. Thesis for M.U.P., Michigan State University, 1965. PURPOSE: To consider the suitability of natural resources and man-made features for supporting agricultural, recreational, residential and industrial development. METHOD: An outline of the land user's requirements for the following factors was to be a guide for assigning such uses: 1) agricultural soil capability, 2) soil permeability, 3) water-- both ground and surface, 4) land drainage, timber resources or woodland, 6) type of terrain and degree of slope, 7) wetlands, 8) highway, roads and accessibility, 9) railroad facilities, and 10) central water supply and sewage disposal. The following maps were manually compiled for the analysis: Soil Permeability, Groundwater Supply Potential, Lake Size, River and Stream Size, Woodlands, Wetlands, Slope, Roads and Railroads. MEIR-OI judgme and th. inProve The prc Problen the pr “(my H PlOying The of Compu ‘=20,ooo could be was Wade thSDOrt and acces: aCtiVlties tErms 0f v 27 Murray, Timothy, Peter Rogers, David Sinton, Carl Steinitz, Richard Toth, Douglas Way. Honey Hill: A Systems Analysis for Planning the Multiple Use of Controlled Water Areas. Cambridge, Massachusetts: Harvard University, 1971. PURPOSE: As stated on page 393 of the report itself, the method developed was in response to the following question, How can one propose a solution to a planning problem, in this case the multiple use for a reservoir site, when the activity demands on the site are unclear and when the costs and benefits to be derived from any proposal must be eval- uated in terms of multiple objectives including dollar bene- fits and costs, environmental resource quality and the social benefits of consumer satisfaction? METHOD: Two approaches were developed, “that of professional judgment proposals evaluated and improved by a simulation model, and that of a linear program generated proposal evaluated and improved through the use of a simulation model." (page 397) The procedures developed included the traditional components of problem identification, data inventory and analysis. It was in the preparation, evaluation and improvement of plans that the Honey Hill Method diverged from the traditional, chiefly by em- ploying simulation models. The data inventory was computer-compatible and made much use of computer graphic displays. Aerial photos at approximately l:20,000 were one data source. A selection of activities which could be part of a development proposal for their study area was made and consisted essentially of recreation, residences and transportation infra-structure. Space Standards, site, service and accessibility requirements were established for the various activities. Resource systems at various sites were assessed in terms of vulnerability and demands were assessed for the possible 10. 11. 12. 28 activities. Costs to provide the facilities and incomes to be derived from participation in the types of activities were also considered. Ohio Department of Natural Resources. Land Capability Analysis: The Wolf Creek Project. Columbus, Ohio: Ohio Department Of Natural Resources, February, 1974. PURPOSE: To evaluate the ability of land to accommodate various types of development and of incorporating this evaluation into the planning process. METHOD: The land capability analysis relies almost exclusively on characteristics of soils in determining capabilities. This analysis is based on an inventory of the physical variables in an area and an assessment of how their properties can affect or be affected by the following types of land uses: residential, commercial, industrial, agricultural, transportation, recreation, conservation and underground utilities. This is a computer- based data system. Sharpe, Carl and Donald L. Williams. "The Making of an Environ- mental Fit." Landscape Architecture, Vol. 62 (April 1972), pp. 210-215. PURPOSE: TO select an area with potential for accommodating a new town. METHOD: This analysis combined McHarg's approach in the assembly of natural resource data and Lewis's concept of natural amenities and existing land use. The overall integration of data was done on the computer with a cluster analysis program. Toth, Richard E. Criteria for Evaluating the Valuable Natural Resources of the TIRAC Region. Stroudsburg, Pennsylvania: Tocks Island Regional Advisory Council, 1968. 13. 29 PURPOSE: To develop guidelines for controlling growth and expan- sion. METHOD: Eight factors of the natural resource base are inven- toried and mapped. They include geology, climate, water, soils, vegetation, wildlife, slope and visual aspect. The possible effect of these factors on various "man-activity" components such as settlement, transportation or recreation, are identified in a matrix. Natural features are thus viewed as constraints on various types of development. Vermont State Planning Office. The Vermont Interim Land Cap- ability Plan. Montpelier, Vermont: Vermont State Planning Office, June 1971. PURPOSE: "To identify land and water resource opportunities and limitations relevant to the continuing evolution of Vermont settlement." METHOD: A series of maps were created for each Vermont county depicting: generalized land use, surface waters and drainage divides, limitations for development, capability for agriculture and forestry (high potential agricultural soils), and unique fragile areas. Any development was to be planned so as to relate logically to established settlements, to avoid areas where environmental, historical or educational damage would occur, to conform with known environmental limitations, and to avoid the displacement of important non-urban uses relying upon basic characteristics of the land. 30 Special Purpose Methods* 1. John A. Dearinger (Kentucky), 1968 PURPOSE: To evaluate and rank the potential Of small watersheds near urban areas for various recreational activities. FACTORS EVALUATED: topography, geology, soils, hydrology, vege- tation, climate. Cultural information includes land use, land use capability, transportation, water pollution, and historical sites. METHOD: This inventory results in eleven major elements: climate, scenery, natural environments, historical value, soils, water quality, water quantity, fish populations, human population char- acteristics, local and tourist access, and "disvalues" such as junkyards. The process of evaluating the elements is quite subjective and the importance of each element is determined by multipliers supposedly based on Observations made in Kentucky. 2. R. Burton Litton, Jr. (California), 1968 PURPOSE: A description and analysis of the visual forest land- scape. FACTORS EVALUATED: Form, spatial definition and light, distance, observer position and‘sequence. METHOD: The identification of the variables above is followed by an identification of landscape composition characteristics including: panoramic, feature, enclosed, and focal landscapes. \ i The source for the special purpose methods is: A:Comparative Study of Resource Analysis Methods by Carl Steinitz, Timothy Murray, ‘Vid Sinton, and Douglas Way. Cambridge, Massachusetts: Harvard University, 1969. 31 This method is an attempt to integrate design with natural re- sources data and has been used on highway corridors, etc. Soil Conservation Service, 1966 PURPOSE: To provide a method for county and state agencies to appraise the potentials for various outdoor recreational develop- ments. FACTORS EVALUATED: Climate, scenery and scenic areas, wildlife, people, proximity and access, rural ownership and land use patterns. METHOD: The procedure is to inventory variables that are con- sidered important in making;an outdoor recreation plan (emphasis is on soils and water). Local groups are to determine priorities and therefore the ordering of variables is not explicitly directed. Edward A. Williams (California), 1969 PURPOSE: The planning and acquisition of major open land areas near metropolitan areas; holdings of 90 or more acres. FACTORS EVALUATED: Open land is divided into groupings of: 1) land unsuited for urbanization because Of natural hazards or constraints; 2) class 1 and 2 soils, specialty agriculture, special geological characteristics, major wildlife habitats and ecologies, high scenic quality; 3) proximity to urban centers. METHOD: This method is concerned exclusively with various types of open space. Potential areas were evaluated by the factors listed above and those areas were then overlain on an existing open space map. Means for protecting open space, primarily legal, were examined and implementation strategies were developed. The study which formed the basis of this technique is "Open Spaces-- 32 the Choices Before California." Ervin Zube (Massachusetts), 1968 PURPOSE: To map and define visually homogeneous areas within the study area. FACTORS EVALUATED: Contrast, spatial variety, enclosure, and major water features. METHOD: This method involves a broad classification scheme for large areas. A linear program model was also adapted to Zube's work. ~Suitabi1ity for a particular use is to be "a function of the county's land resource and landscape unit characteristics." Constraints result in recommendations to either maintain exist- ing visual conditions, add open space; add forest; add water; add water and open space, etc. Sources with Conceptual Input to the Study Of Environmental Analysis Burchell, Robert W. and David Listokin. The Environmental Impact Handbook. New Brunswick, New Jersey: Rutgers University, 1975. Davis, Charles M. A Study_of the Land Type. Durham, North Car- olina: Army Research Office, March T969. Dickinson, Robert E. Regional Ecology--The Study of Man's Environ- ment. New York: John Wiley and Sons, Inc., 1970. Fabos, Julius Gy. "An Analysis of Environmental Quality Ranking Systems," Recreation Symposium Proceedings, U.S. Department of Agriculture, 1971. Graham, Edward H. Natural Principles of Land Use. New York: Oxford University, T944. Hackett, Brian. Landscape Planning. Newcastle-Upon-Tyne, England: Oriel Press Limited, 1971. Isachenko, A.G. Principles of Landscape Science and Physical- Geographic Regjpnalization. Translated by N.J. Rosengren. Carlton, Victoria, Australia: Melbourne University Press, 1973. TO. 11. 12. 13. 14. 15. 33 Land Use Analysis Laboratory. A Land Classification Method for Land Use Planning. Ames, Iowa: Iowa State University, 1973. Linville, Jack, Jr. and Ron Davis. The Political Environment - An Ecosystems Approach to Urban Management. Washington, D.C.: American Institute of Planners, 1976. Lyle, John and Mark von Wodtke. "An Information System for Environmental Planning," Journal of the American Institute Of Planners, Vol. 40. No. 6, (November 1974): 394-413. Odum, Eugene. "The Strategy of Ecosystem Development," Human Identity in the Urban Environment. eds. Gwen Bell and Jacqueline Tyrwhitt. Baltimore, Maryland: Penguin Books, Inc. 1972. Sorenson, Jens C. "Some Procedures and Programs for Environmental Impact Assessment," Environmental Impact Analysis: Philospphy and Methods. eds. Robert Ditton and Thomas Goodale. Madison, Wisconsin: University of Wisconsin, 1972. Steinitz, Carl. "Landscape Resource Analysis--the State of the Art," Landscape Architecture, Vol. 60, No. 2 (January 1970): 101-105. Warner, Maurice L. and Edward H. Preston. A Review of Environ- mental Impact Assessment Methodologies. Washington, D.C.: Office of Research and Development of the Environmental Protec- tion Agency, April 1974. Weatherspoon, C.P., J.N. Rinker, R.E. Frost and T.E. Eastler. "Remote Sensor Imagery Analysis for Environmental Impact Assess- ment." Proceedings of the 2nd Conference on Environmental Quality Sensors. Washington, D.C.: Environmental Protection Agency, 1972. Summary The increased emphasis on protection of the natural and human environment has stimulated the development and rediscovery of methods for analyzing and evaluating man's biotic and abiotic surroundings. Numerous methods exist but no one method can claim to be appropriate for all needs. Variations in available environmental data and in the interpretation of such data due to geographic differences, require that any method must be adapted to fit the needs of a particular study. 34 This chapter contains summaries of eighteen environmental anal- ysis methods identifying the purpose for which each method was devel- oped and briefly describing the factors used to evaluate the environ- ment. NO attempt was made to include all methods, but the review can be considered comprehensive and representative of the state-Of- the-art. As was evident in reviewing the eighteen methods in this chap- ter, much diversity is found in the purposes for which individual methods haVe been developed and in the factors which were employed to make environmental evaluations. Certain limitations regarding environmental data are shared by all methods, however, because any environmental evaluation can only be as accurate as the data on which it is based. Continuously expanding programs and legislation pro- tective of the environment imply a need for improved methods of evaluation and better data about the environment to serve as input to those methods. Some of the problems of data base availability include cost of acquisition, timeliness, accuracy, scale and comprehensibility. This thesis explores the role of a new data source, high altitude, color infrared photography, in the field of environmental analysis; a source which can improve the overall quality of the data base for environmental analysis. In the established sequence of this thesis, the step which fol- lows the review of environmental analysis methods involves selecting a site in Michigan in which to test the utility of high altitude color infrared photography. Chapter III identifies the selected site and describes its environmental and growth characteristics. CHAPTER III SELECTION AND DESCRIPTION OF THE STUDY AREA The area between the cities of Grand Rapids and Kalamazoo, Michigan contains extensive agricultural and recreational acreage. Grand Rapids and Kalamazoo, major urban areas of southwestern MichJ igan with 1970 populations Of 197,649 and 85,555 respectively are expanding and little analysis of the impending growth problems of the area lying between these cities has been done. The major Objective of this research effort is to investigate the utility of small scale color infrared imagery as a data source for environmental analysis in areas under pressure for development and urbanization. Development; as defined by the American Law In- stitute in A Model Land Development Code, means "the carrying out of building, engineering, mining or other operations in, on or under land, or the making of any material change in the use of any build- ings or other land."12 The term development is.widely used in legal as well as plan- ning disciplines and conceptually the term "can be divided into tan- gible and intangible activities. The tangible activities include I the construction and alteration of buildings and the other activities typically associated with changes in the use of the land. The intan- gible activities include a limited number of types of transactions in land which are very directly related to the type of development 35 36 that may occur on the land. These transactions include the dividing of land into parcels and the creation or extinguishment of rights of access or riparian rights."13 The area lying between Grand Rapids and Kalamazoo is exper— iencing increased pressure for development. In other words, the de- mand for urban and built-up land is increasing. On March 1, 1974, the NASA-sponsored Project for the Use of Remote Sensing in Land Use Policy Formulation at Michigan State University requested high- altitude RB-57 color infrared coverage of the urbanizing corridor between the cities of Kalamazoo and Grand Rapids, Michigan. This request was part Of the Remote Sensing Project's continuing research effort to experiment with applications of remote sensing to resource analysis and land use planning in Michigan. Color infrared photog- raphy was requested because it is highly useful in land use analysis, as was explained in Chapter I. The mission, Mission 273, was flown on May 7, 1974 at an approximate altitude of 56,750 feet. Color infrared imagery at scales of l:56,570 and 1:113,500 was later pro- vided to the MSU Project. Figure A-l in Appendix A indicates the flight lines and resultant coverage at both scales. Specifications for Mission 273 and approximate conversion factors are also given in Appendix A. For use as a test site in the environmental analysis which follows in Chapters IV and V, a study area was arbitrarily delineated within the total area covered by the imagery. The study area is a north-south corridor extending from the northern to the southern boundary of Allegan County. US-131, a major arterial from the Michigan/Indiana border north to Petoskey, and a limited access 37 freeway between Grand Rapids and Kalamazoo roughly approximates the center line of the corridor which extends laterally about three miles east and three miles west of the arterial. (See Figure 4.) Portions of eight townships, all within Allegan County, are included in the study area, namely: Dorr, Leighton, Hopkins, Wayland, Watson, Martin, Otsego and Gun Plain. The study area or test site is 24 miles long from north to south and 6 miles wide from west to east. This l44 square mile area comprises approximately l7 percent of the total land area of Allegan County. Four overlapping frames of color infrared imagery at a scale of l:ll3,500 were used for data extrac— tion and interpretations. Black and white prints of the original color infrared imagery are provided as Figures 7 through 10 in Chapter IV. . The first portion of this chapter describes the growth char- acteristics of the study area supporting the contention that this area is being exposed to increasing development pressure. The second 'portion of the chapter describes the general physical and biotic characteristics of the study area; a prerequisite to any environmental evaluation of the land. Growth Characteristics of the Study Area The entire State of Michigan lies within the drainage basin of the Great Lakes. Studies compiled by the Great Lakes Basin Commis-. sion in 197514 indicate that urban demands for land in the Basin will increase substantially and that much of this expansion (86 percent) twill occur within the Standard Metropolitan Statistical Areas (SMSA's) by 1980. Grand Rapids and Kalamazoo are both classified as SMSA's. 38 11114.240 8100' Ali! Figure 4. Study Area in Allegan County 39 Within the Grand Rapids SMSA approximately 25,000 acres are now considered urban and built-up while in the Kalamazoo SMSA the figure is about one-half that amount. Between l980 and 2000, urban and built—up land in both SMSA's is expected to grow by approximately l0,000 acres. In the year 2020, nearly 50,000 acres in Grand Rapids and 25,000 acres in Kalamazoo is projected to be urban and built-up. This expansion will affect surrounding areas outside the SMSA's. The rate and direction of urban expansion is influenced by many factors. Population, population growth, economic activity, and transportation networks are some of the more important fac- tors. These affect growth around existing urban areas and also between centers of economic activity.15 The area lying between Grand Rapids (Kent County) and Kalam- azoo (Kalamazoo County) is not part of either SMSA. However, urban and built-up land in this region which includes the study area, is Projected to increase by 5,000 acres between the present time and 1930. and to increase by another l0,000 acres between l980 and 2000. To demonstrate the current growth relationship between the StUdy area and the cities of Grand Rapids and Kalamazoo, population QVOWth was analyzed. Table l lists the population growth of the 130W05hips and major cities included in the study area between 1960 and 1970. Those townships closest to the cities of Grand Rapids (DOW and Leighton Townships) and Kalamazoo (Otsego and Gun Plain Townships) exhibited some of the highest rates of growth in the study area. Infbrmation gathered from the l970 Census of Population con- Geming the place of work of a sample group of Allegan County resi- dents and shown in Table 2, serves to illustrate the interde- Pendence of Allegan County with the urban areas of Grand Rapids and Ka‘ am 200 . . 40 Table l. Study Area Population Growth Population 1970 1960 % Change Dorr Township 3,055 2,313 32.1 Gun Plain Township 3,231 2,796 15.6 Hopkins Township 2,084 1,766 18.0 Hopkins Village 566 556 1.8 Leighton Township 2,354 1,951 20.7 Martin Township 2,125 1,963 8.3 Martin Village 502 483 3.9 Otsego City 3,957 4,142 -4.5 Otsego Township 3,721 2,564 45.1 Plainwell City 3,195 3,125 2.2 Watson Township 1,331 1,065 25.0 Wayland City 2,054 2,019 1.7 Wayland Township 1,661 3,450 -51.9 Total 1960 Population --- 28,193 Total 1970 Population --- 29,836 1970 Test Site Density*---103.6 persons per square mile (*288 square miles for the eight townships noted above) Average density for all of Allegan County---80.6 persons per square mile 41 Table 2. Employment by Place of Work Data Sumnarized Dorr Townshfl) (directly south of Grand Rapids) 0f'680 persons employed: 50% worked in Grand Rapids (city) 15% in other areas of Kent County 31% in Allegan County Leighton Township (east of Dorr Township) 0f'746 persons reporting: 31% worked in Grand Rapids (city) 13% in other areas of Kent County 34% in Al1egan County Gun Plain Township (north of Kalamazoo) Of 1192 persons emp1oyed: 24% worked in Kalamazoo (city) 13% in other areas of Kalamazoo County 62% in Allegan County In Plainwe1l, the major city in Gun Plain Township, of 1039 persons reporting: 24% worked in Kalamazoo (city) 10% in other parts of Kalamazoo County 63% in Allegan County (£5890 Township (west of Gun Plain Township) Of 1264 persons employed: 20% worked in Kalamazoo (city) 10% in other areas of Kalamazoo County 67% in Allegan County I".QI£fl§gg, the major city in Otsego Township, of 1239 persons emp] o.Yed: 17% worked in Kalamazoo (city) 3% in other areas of Kalamazoo County 78% in Allegan County For all eight townships, of 9042 persons reporting, 1495 or 16:5 Percent worked in Grand Rapids and areas of Kent County; 1738 0" 19.2 percent in Kalamazoo City and County; and 5436 or 60.1 percent 42 in Allegan County. Table 3 lists complete employment by place of work findings from the U.S. 1970 Census which pertain to the study area. ‘ In sunmary, growth in the study area, as it is related to pop- ulation and employment is currently influenced by nearness to the cities of Grand Rapids and Kalamazoo. In addition, the increased demand for urban and built-up land in the Grand Rapids and Kalamazoo SMSAs is projected to spillover to the area which lies between them; an area which includes the test site. Environmental Characteristics of the Study Area As discussed in Chapter II, any evaluation of a land area re- quires consideration of major features of the natural environment such as: climate, soils, water resources, topography, geology and land cover/use. This portion of Chapter 111 describes in general terms, the environmental setting of the study area and thus lays the foundation for development of the environmental inventory and anatysis method in Chapters IV and V. The references used for the descriptions which follow oftentimes addressed areas of land much 1“‘98!" in size than the study area. At the time this study was initiated, there had been no detailed environmental inventories pl‘bHShed for either the townships in the study area or for Allegan County, The Allegan County Soil Survey, as an illustration, was made i" 1901 by the Soil Conservation Service and is out-of—print. As a "esult of the lack of specific information, it was necessary to gene"alize facts from sources of a broader scope, and apply them to the Study area. Table 3. Employment by Place of Work Data :g.ug:kPe;sons Percentage of Total Number of , Place of Work 1 Total Surveyed Townshisa or City of Residence 1 Selected Areas Employed Surveyed Area Number Surveyed During Aneiglected Census Week Dorr Township 680 2 341 50 3 104 15 10 214 31.5 Hopkins Township 820 2 98 12 3 45 5 8 25 3 10 489 60 Leighton Township 746 2 230 31 3 206 28 8 19 -- 9 9 -- 10 256 34 Marthi'Township 544 2 29 5 3 15 3 8 99 18 9 20 4 10 366 67 Gun Plain Township 1192 8 291 24 9 152 13 10 742 62 15 4 -- Otsego City 1239 3 6 -- 8 209 17 9 42 3 10 969 78 Otsego Township 1264 a 251 20 9 121 10 10 851 67 15 12 l Plainwell City 1039 3 6 -- 8 251 24 9 108 10 10 658 63 Watson Township 279 2 6 2 8 56 20 9 35 12 10 167 60 “ayland City 756 2 177 23 3 84 ll 9 15 2 10 465 62 ”Maud Township 483 2 112 23 3 36 7 8 28 6 9 7 l 10 259 54 \ fl 1A sauple population of employed persons over 14 years of age who worked during the Census week 1 ma Number Place of Work Grand Rapids (City) Remainder of Kent County Kalamazoo (City) Remainder of Kalamazoo County Al legan County Battle Creek (City) wooooww it I I II | I o I I I I I o I n I I I 0 I I I —l—J 44 Climate The climate of the study area can be characterized as humid- continental. Humidity is moderate and temperatures range from an average of 24°F. in February to 73° F in June. Precipitation aver- ages 33 inches per year, snowfall averages 50 inches per year, and severe storms are uncommon. Prevailing winds are from the southwest, a factor to be con- sidered when planning the location of facilities which generate air pollutants or odors. This information can also be utilized when locating residences in order to take advantage of the cooling effect of the winds. Although the area's climate is somewhat moderated by Lake Michigan, as is the case for the entire lower peninsula of the state, the study area lies east of the famed fruit production belt of Mich- igan which parallels the Lake Michigan shoreline. Geology/Hydrology The study area lies within the Kalamazoo River basin which en- compasses 2060 square miles. The Kalamazoo River flows west by north- west and terminates at its confluence with Lake Michigan. The main branch of the Kalamazoo crosses the study area as do two of its major tributaries, the Rabbit River and Gun River. (See Figure 5.) The bedrock underlying the study area consists of formations dating back approximately 325 million years, i.e. Mississippian. "0511 of the Kalamazoo River Basin of which the study area is part, 15 l"iderlain by Coldwater shale. In addition, a narrow band of the Marshall sandstone formation crosses the area in a northwest-south- west alignment between Muskegon and Battle Creek. But as is the :mem mmmc_mga gm>_m oonEmpmx .m mesa?“ } 2: :22 .3545; \ .\ \ \ ,5 33:3; 4 .Honac 9 Q m“ sexygV \. m. x... .\ n. 25:: :2, 3.. . ne_auu ==..u 46 case fbr most lower peninsula communities, bedrock geology has little effect on the economy or development of the area since no bedrock outcroppings occur. The entire surface is masked by the unconsol- idated materials of glacial drift. Surface formations in the area were formed by the final recession of the Wisconsin glacier which began about 20,000 years ago. The low, rolling relief consists of outwash plains, moraines and lake bed deposits. Allegan County is not covered by a comprehensive groundwater report of any detail. In general, however, water is likely to be more available from glacial deposits than from bedrock. This availability in glacial drift and alluvium is quite variable. The more produc- tive aquifers (over 500 gallons per minute) are likely to occur in thick sand and gravel deposits near streams. The poorest aquifers would probably be found in thin glacial drift deposits or in the clay or silt till and lake deposits. High yield aquifers have been found in areas of the Kalamazoo River Basin including the Otsego/Plainwell area. Natural streamflow in the Kalamazoo River Basin fluctuates with the seasons; high in the spring and low in the winter. But the streams within the project area are not characterized by rapid erosion and their sediment content is generally low. Because the soils in the basin are generally porous, infiltration is increased thus re- ducing run-off peaks and equalizing the groundwater supply which reaches the streams. Most floods in the basin have occurred after heavy spring rains or snowfalls when the ground is already partially saturated. Minor flooding damages have occurred within the study area in Otsego and 47 Plainwell, chiefly because of encroachment on the floodplain. Wetlands There are few large lakes within the study area and no exten- sive wetland complexes remain. The area is crossed by a network of artificial drainage channels or straightened natural channels which were developed primarily for improving drainage of agricultural fields. The area does, however, have a large number of small ponds and pitted areas that are at least seasonally filled with water. Soils and Topography There is no up-to-date soil survey for Allegan County but accord- ing to a 1968 publication of the Michigan State University Cooperative Extension Service, Soils of Michigan, two major land divisions are found in the study area. Land Division "T" is composed primarily of the Miami and Conover soil associations, while division "V" is com- prised of Fox and Oshtemo soil associations. (See Figure 6.) The soils in Land Division "T" were derived chiefly from limy loam glacial till. Drainage varies from good to poor in depressions and natural drainageways. Topography is nearly level to rolling and is favorable for tillage operations. Topography in Division "V" also ranges from nearly level to rolling and many of the level outwash areas are strongly pitted. Soils can be tilled easily and are moder- ately productive. A lack of moisture holding capacity and low natural fertility, however, are limiting factors in crop production. Two other land divisions are found in the study area; Land Division "0" is composed primarily of the Rubicon, Grayling and Roselawn soil series and was formed from materials low in lime. -..———___.__—.._. —. . . g _. 71:32: 2.: :— I : " -—'-"-'-';:;_zz:_=:: . :' __———_-,:..—_——':: ..- .,.-.-. :‘ ————72 Tu . E =:'i'.=::.= , -‘-.,_...__.‘_- .:;_;j——- T ‘.. ..,... :_______...:_;2 v 5:: 2&522: - .5 _:='—_—*:-.-:.;_,* , I :g=_ =__ —.___-: :3“ "E van. _ .._r____________-- lillllllulllllll Till H) I” I H ”Egg-:2: ll: '1. ._.————.—_.,. ..__ __.1 E:E.=~=~:EE ‘1 uuo omsms i—EEETEEZES Willi Efgéi—Eg l" [in “.1 ...__._..__..__. 11111) W 1' .1 ll. Ill _ __ _————-——— —_ _.—.___ ___— _———-— —- ._ —_ — ___——-— -—_.—— ___ __————-—-—. ,— .— 111 := " FM = :_—_'__.___—::1:_:':_l _ _.-.. _____*:—__EE"='_-' -.____. __ __—____._ __ -—___;-- _ _—_.—_._.:‘_ = __——:_____z __._.= :-_____-—.__— ::—.__:::::--; , '-.-.-.-..- .-.-.--_.->. ‘‘‘‘‘ —___—_—-=-——___:._—_—:::= '7 ::;.:-;=: :3: ._ _ —.=.-_—;.; =_—':::-:_- ‘¢ t'_'—_'. ——_=EE: _.-.~' 1 r; h '3 —, ___:___ . ’ . —~—.—_—.—.-_.__2.. A-__ ~ hf; "“‘ ' ' ' 7'... '1 3’s: 1&5". -:‘ .; '. in meqafi‘ m. ~ .1 up, M '2'- WWWM 111 Figure 6. Land Division Map of Study Area 49 They are mainly well drained sands with topography ranging from level plains to hilly uplands. Natural fertility is low, as is moisture holding capacity. Wind erosion may be a problem when the soil is tilled. Land Division "Z" contains organic soils (muck and peat). Fertility, control of water table and wind erosion are major problems for successful use of these soils. . It should be noted that the glacial materials responsible for soils in most of the lower peninsula of Michigan differed signif- icantly in texture and as a result, natural drainage conditions can vary from well-drained to poorly drained in very short distances. Land Cover and Agriculture The study area was entirely forested when first settled with mixed southern hardwoods and oak-hickory associations. The area now dramatically illustrates the farming pursuits which have occurred. The majority of the land is now or has previously been in cultivation. Wooded areas are scattered and generally small in size. Professor J.0. Veatch in SpecialfBulletin 231 (October 1941), of the Michigan State University Agricultural Experiment Station entitled "Agricultural Land Classification and Land Types of Mich- igan," presented a dot map of Michigan at a scale of 1:750,000 show- ing proportions of agricultural land in Michigan with high (first class), medium (second class), low (third class), and very low (fourth class) values for general farming. According to a 1972 report by Whiteside and Schaner, Veatch's generalized map is still valid. This map indicates that most of the land in the study area consists of first and second class agricultural land. 50 A glance at an aerial photograph or a drive through the area confirms that much of the land can be and is being farmed. However, a significant amount of land during a field survey in August 1975, appeared idle. This seemed particularly true in areas closest to urban centers. Statistics from the U.S. Department of Commerce Census of Agriculture between 1964 and 1969 showed a decrease of 196 farms in Allegan County in those five years. The number of farms of less than 100 acres decreased from 1,565 to 1,481 (-5.4%) and the number of farms between 100 and 219 acres dropped from 859 to 740 (-13.9%). Checking the pattern of land ownership on the Allegan County Plat maps, it is clear that very few large parcels are held in single ownerships. It thus seems likely that the small farms are most prevalent but declining. 1% Color infrared imagery covering an area under increasing pres- sure for development which roughly approximates a corridor from Grand Rapids to Kalamazoo, was obtained by the MSU Remote Sensing Project. A portion of the area covered was designated as a test site to be used in evaluating the envirbnmental analysis method in Chapters IV and V of this thesis. The 144 square mile test site or study area lies entirely in Allegan County. Growth in the study area as it is related to population and employment is currently influenced by nearness to the cities of Grand Rapids and Kalamazoo. In addition, the increased demand for urban and built up land in the Grand Rapids and Kalamazoo SMSAs is projected to spillover to the area which lies between them; an area 51 including the test site. The natural features of the study area described in this chap- ter lend themselves to agriculture and most of the land is now or has previously been in cultivation. The potential impacts of urban expansion in what is now a predominantly rural, agricultural area include: 1) suburban land value appreciation far above farm land prices and the resultant loss of agricultural land; 2) increased costs of public services to be born by local governments which often lack a large commercial and/or industrial tax base; costs which are particularly high when development occurs in the rural fringe; and 3) increased pressure for development of open space and other nat- ural features. Development pressures in the study area indicate that early planning for growth and identification of areas most suited for development should be undertaken now. The following chapter pre- sents the inventory and analysis method to be used in evaluating the study area environment and the implications for development therein. 52 FOOTNOTES 12American Law Institute, A Model Land Development Code (Philadelphia: American Law Institute, April 15, 1974), p. 21. 13 14Great Lakes Basin Commission, Appendix 13 - Land Use Manage- ment (Ann Arbor: Great Lakes Basin Commission, 1975), pp. 73-79. 15 Ibid., p. 23. Ibid., p. 74. CHAPTER IV INVENTORY AND ANALYSIS METHOD The purpose of environmental inventory is to provide data about selected components of the natural environment; information which can then be used as a basis for evaluating the environment. Depending on the type of evaluation to be made, different components of the environment will assume varying degrees of importance. Illustrations of different types of evaluation that have been made are provided in Chapter II which summarizes the purposes for which various environ- mental analysis methods were developed. In order to analyze and evaluate the environment in the study area in Allegan County as it relates to accommodating increased amounts of development, it is necessary to first inventory those natural and man—made features which will be important for land use decisions and resource protection. In addition, since the objective of this re- search effort is to employ small scale color infrared imagery in a non-technical manner as the primary data source, it is necessary to integrate the data requirements for environmental evaluation with a realistic level of detail to be interpreted fronithe imagery. Development of the Inventory, There is an abundance of information concerning the development of inventory systems. In Michigan, for example, the Office of Land 53 54 Use of the Department of Natural Resources, developed a land cover/ use classification system for the entire state.16 The system was developed with four levels of detail for characterizing mapped areas with scales ranging from l:250,000 to l:1,000,000 (Level I), 1:125,000 to l:250,000 (Level II), 1:50,000 to l:125,000 (Level III), and 1:24.000 to 1:50,000 (Level IV). The system was designed for compatibility with aerial photography and other remotely sensed imagery as well as with computers and illustrates a general trend in that "plans for inventories almost always now imply or specify an associated computer-based information system, particularly as the tasks are defined in terms of larger areas of land than can be "17 Recently developed environmental assessed by direct observation. analysis methods such as numbers 9-11 in Chapter II which are computer compatible, further substantiate this trend. Because of the increasing emphasis on computer utilization, and for the convenience found by numerically coding data for large areas, the inventory method developed for this.research effort was designed tolxacomputer compatible. Although no computer storage or manipula- tion was employed, all data were recorded on computer data sheets according to a numerical coding system. In the interest of standardization and the advantages that a statewide system would offer, the potential for extensive use of the Michigan Land Use Classification System was investigated. Accord- ing to that System, a combination of Levels II and III could have been best employed (see Table 4) since the scale of the color infra- red photography was l:113,500. The Michigan System heavily emphasizes land cover and land use inventory data. Because of the investigative 55 Table 4. Michigan Land Use Classification System - Levels I and 11* Level I Level II 1--Urban and Built-up 2——Agriculture 3--Rangeland 4--Forest Land 5--Water 6--Wetlands 7--Barren ll--Residentia1 12--Commercial, Services and Institutional l3--Industrial l4--Transportation, Communication and Utilities 16--Mixed l7--Extractive l9--Open and Other 21--Cropland, Rotation and Permanent Pasture 22--0rchards,Bushfruits, Vineyards and Ornamental Horticulture areas 23--Confined feeding operations 29--Other Agricultural land 3l--Herbaceous rangeland 32--Shrub rangeland 4l--Broadleaved 42--Coniferous 43--Mixed 51—-Streams and Waterways 52—-Lakes . 53--Reservoirs 54--Great Lakes 61--Forested 62--Non-Forested 72--Beaches and Riverbanks 73--Sand other than beaches 74-—Bare, exposed rock 75--Transitional areas 79--Other * Only those categories applicable to the imagery and study area were here reproduced. nature of this study and its attempt to obtain more analytical infor- mation, it was determined that development of a new classification 56 system was required. Consequently, use of the Michigan Land Use Classification System was minimal. The inventory classification system developed in this research effort was designed for use with small scale, high altitude color infrared imagery as the principal data source, which unavoidably required some photo interpretation. The system was also designed fbr use by individuals with little technical expertise in remote sensing such as planners who would benefit by the ability to utilize available imagery in assessing and directing development decisions in a more environmentally informed manner. Many operations which utilize remotely sensed imagery oftentimes request that specialists prepare the initial inventory. This is often the case when a plan- ning agency, for example, wishes to obtain such an inventbry but feels it lacks either the time, hardware or manpower with sufficient expertise to perform the initial interpretation. The objective of the classification system and methods presented herein, was to devise a technique for use with color infrared-imagery that would be under- standable, easily reproducible by other planners and relatively in- expensive to employ assuming the imagery was available. The degrees of complexity and sophistication of the tools used in this system were designed to reflect those resources typically available to the intended users including the resources of manpower, time and hardware. The classification system initially developed in this research effort employed the major categories A through H as listed in Table 5, as well as many of the specific numbered items. The system was developed as a result of the following actions: (1) a review of those factors inventoried by various other environmental analysis Table 5. Coding System 57 Relief (Colums 9-12) - level - undulating hilly Land Cover (Columns l4-28) open land urban and built up forested other wetland . major lake (plus 100 acres) minor lake (less 100 acres) najor stream minor stream Land Use (Columns 30-44) cultivated cropland tree farm orchard/vineyard urban non-residential park/playground residential undetermined/other extractive no apparent use/idle . Drainage (Columns 46-50) - very poor (wetland) - poor - moderate - well wN—‘W O-P- owN—‘O O Accessory Information (Predom- inant condition) frontage developed frontage moderately developed (at least cleared) frontage undeveloped (thick vegetation) under construction none Accessory Information airport rail station mobile home park none Transportation Routes (Columns 52—56) US 131 interchange US 131 paved road bituminous road divided highway railroad road bridge no roads gravel,soil surfaced or similar Amenities/Disvalues (Columns 58-62) golf course athletic field track junkyard/dump cemetery Plum-112350 - 23456789 H11 ‘ h’a in CO Uh Uh Nil Her fir cat 58 Table 5. (Continued) F (Continued) 6 - other 7 - drive-in theatre 0 - none G. Residential Settlement (Columns 64-65) G. Accessory Information 1 - low density 1 - no pattern 2 - medium density 2 - scattered 3 — high density 3 — linear 5 - undetermined 4 - clustered or 0 - no evidence of residential settlement neighborhood 5 - subdivision Special Concerns (Columns 67-71) artificial drainage (straight channel) residential construction other construction or indeterminate construction water impoundment structure (dam) other ponds eroded areas evident oil wells none QmNO‘U'l-DWN-HI I I I I I l I I I methods in Chapter II; (2) a review of the Michigan Department of Natural Resources classification system which can employ remote sensing imagery; and (3) a review of those natural and man-made factors which could be reasonably extracted from the imagery by a non-expert and which could influence the quality of development and the quality of the surrounding environment. As work progressed and familiarity with the imagery was achieved, alterations, additions and deletions were made to an original format. The system shown in Table 5 is the final classification utilized. Complete definitions of the various categories follow the discussion of interpretation methods. FOUl l:113,500 Black and trating t Figures 7 the only for illum M- 1 network fc For this s Jacket for then Outli RaDl'dograp I01- Inst: dimensions jacket to n. ground Sect the one ”1116 Earth; i.e. Each Photogn sides and 1 a mile the dlst graphic Cente SEQ' The 9nd to Wm IhIS quartEr 59 Interpretation Methods Four color infrared (CIR) transparencies at a scale of l:113,500 depicting the 144 square mile study area, were used. Black and white prints of the color infrared transparencies illus- trating the actual size and scale of the originals are provided in Figures 7 through 10. A 10X (power) Seerite Tripod Magnifier was the only magnification equipment used. A fluorescent light table for illuminating the transparencies was also required. Step_l. Michigan's original land survey system and resultant road network form, in essence, a grid pattern or network on the ground. For this study, each transparency was placed in a plastic (acetate) jacket for protection. All square mile sections on the ground were then outlined directly on the jacket in black drawing ink using a Rapidograph 3060 No. 00 pen point. (Illustrated in Figures 7 through 10). Instead of making an acetate overlay with standard section dimensions drawn in, the sections were outlined directly on the jacket to more closely follow the actual configuration of the ground sections. These ground sections are slightly distorted from the one mile square dimension because of the sperical shape of the earth; i.e. all lines on the ground are not truly perpendicular. Each photograph was gridded at least 1 and one-eighthS'hwflifrom the sides and l and seven-eighths inchfrom the top and bottom to mini- mize the distortion which increases with distance from the photo- graphic center or principal point. Stgp_g, The second step was to draft a detachable quarter-section grid to overlay on the full sections already drawn. (See Figure 7.) This quarter section grid, because of its standard dimension of Figure 7_ RB 6O Black and White Reproduction Actual Size Mission 273 - Frame Number 10-0046 RB-57 Color Infrared Image: Figure 7. Figum 8. l 61 ‘11 “‘3 aiIP’; q‘ J .' ' f l‘ | l I {J Figure 8. RB-57 Color Infrared Image: Black and White Reproduction Actual Size Mission 273 - Frame Number 10-0045 62 Figure 9. RB—57 Color Infrared Image: Black and White Reproduction Actual Size Mission 273 - Frame Number 10-0044 Figure lo. 63 Figure 10. RB-57 Color Infrared Image: Black and White Reproduction Actual Size Mission 273 - Frame Number 10-0043 approximat drawn full characteri the lines conformit: (160 acre: the smalll was chose' to retrie Highways divided 1 and from a quarter 94d was area as 1 The "Ortl t0 the nc The SOUtI to the 5‘ southeasi w1th the This ”Um! infT‘ared It to ident- 64 approximately .25 inch by .25 inch did not precisely align with each drawn full section. It was determined, however, that the removable characteristic of the quarter section grid permitted viewing beneath the lines and this was a more important capability than complete conformity with the full section grid lines. Each quarter section (160 acres) is henceforth referredtto as a grid cell, and represents the smallest data gathering unit for this study. This 160 acre unit was chosen because it could be easily outlined and rapidly identified to retrieve inventory data. Step_§. Using a base map from the Michigan Department of State Highways and Transportation, the study area in Allegan County was divided into a grid. (See Figure 11.) Running from west to east and from north to south, each grid cell on the map corresponded to a quarter mile section on the ground. The northwest corner of the grid was labelled (01,01) as X and Y coordinates, and is the same area as the northwest one-quarter of Section 3 in Dorr Township. The northeast corner of the grid was labelled (12,01) and corresponds ' to the northeast one-quarter of Section 4 in Leighton Township. The southwest corner of the grid was labelled (01,48) and corresponds to the southwest one-quarter of Section 34 in Otsego Township. The southeast corner of the grid was labelled (12,48) and corresponds with the southeast one-quarter of Section 33 in Gun Plain Township. This numbering system was also applied to the grid made on the color infrared photographs and is shown in Figures 7 through 10. I It was recognized early in this effort, that the capability to identify the grid cell, the ground location and the extracted in- terpretation data was of critical importance. The township, range Coordinates V 65 c—— I Coordinates __. 01 02 03 04 05 06 07 08 09 10 ll 12 Figure 11. 6} Gridded Study Area 66 and section layout of Michigan made it relatively easy to delineate standard size, squared areas on the grid overlays and have them correspond with ground locations. The numbering system as devised with X and Y coordinates, made it possible to rapidly correlate coded information with actual ground locations and cells on the imagery. §t§p_4, Data Coding Forms of the Michigan State University Computer Laboratory were used as formats to record the data. Each form pro- vided 30 rows and 80 columns for recording information. As utilized, each row represented a different grid cell. (See "Key Coding Sheet" in Appendix B.) A total of 20 forms were used to record data for the 576 grid cells; only six rows of the twentieth sheet were re- quired. Two types of data were recorded in this classification system: area data and occurence data. Area data which was recorded in deciles (percentage groupings in increments of ten), is found in the following categories: A. Relief, 8. Land Cover,,C. Land Use, and D. Drainage. These categories of information are applicable to entire cells but in differing proportions. For example, between 10-20 percent of a cell could be hilly and the remaining 80-90 percent might be level. The gptjrg_cell has some kind of topographic characteristic, however. Decile groupings were recorded on the coding sheets in the fol- lowing manner: Range of Percentage Symbol Appearing in Coding Sheet 0-10 percent 10-20.percent 20-30 percent 30-40 percent 40-50 percent «th-‘C For any data cat~ the exam would ha Sum to '” illustra' Furthermi inch by ‘ much fasi Deded if 0f Clarii in area ( Enabled 1 digit an‘ for less The type Incl F' Amenii as a Dave couid DOt "Ca eithEr er COOSj del‘e 67 Range of Percentage Symbol Appearingpin Coding Sheet 50-60 percent 60-70 percent 70-80 percent 80-90 percent 90-100 percent tomeU'l For any single cell, the aggregation of percentages for any area data category was equal to 9, and represented the entire cell. In the example of topographic relief just mentioned, hilly topography would have been coded "l" and the level topography "8." The numbers sum to "9." An advantage of using decile ranges is that it clearly illustrates that figures derived in the interpretation are estimates. Furthermore, since each grid cell on the imagery measured only .25 inch by .25 inch, the decile system made ocular estimation of size much faster. Speed of interpretation would have been severely im- peded if an exact percentage were required. As an additional note of clarification, it should be remembered that the "0" decile coding in area data categories does not represent "none." Use of the zero enabled the coding of percentages to be maintained with only one digit and provided a means to indicate area data which accounted for less than ten percent of a grid cell area. The second type of data recorded was occurrence data. This type includes the following categories: E. Transportation Routes, F. Amenities/Disvalues, and H. Special Concerns. Information such as a paved road or a golf course is either present or absent and could not always be recorded in percentages. "Category D. Settlement Patterns" does not fall neatly into either area or occurrence data categories. It could technically be considered an "area" concept, but because of the small scale of the image cells broken Use, 0 terns, here de R maps C01 covered The rema ”t9 Quad (1973 ph (1972 phi Bec (15 minUt minute qu jettive n Scale, the LEVE UDdL Hill 68 imagery, this assessment was generalized or averaged for entire cells and thus percentages were not given. Classification System The information to be obtained with the coding system was broken down into eight categories: Relief, Land Cover, Land Use, Drainage, Transportation Routes, Residential Settlement Pat- terns, Amenities/Disvalues, and Special Concerns. Each category is here defined and its associated problems discussed. Category A. Relief (Coding Sheet Columns 9-12) Relief was approximated for each cell from USGS topographic maps covering the test site. The bulk of the site (72.9%) was covered on the 15-minute quadrangle titled Wayland, Michigan (1959). The remainder of the test site was covered on four USGS 7 1/2 min- ute quadrangle maps: Otsego (1973 photo revised), Kalamazoo NE ‘ (1973 photo revised), Cutlerville (1972 photo revised) and Caledonia (1972 photo revised). .. I Because the scales of the available topographic maps differed (15 minute quadrangles give 20-foot contour intervals and 7 1/2 minute quadrangles give lO-foot contours) and because of the sub- jective nature of assessing topography for a large area at a small scale, the following classes of slope were approximated: Level: 0-6% Undulating: 6-25% Hilly: over 25% Lar from 1anc (l) forested (2) degree of natural 5 (3) broken do into upla 0P9” char to areas tracts. "011 consi as "fores fifty per ”OUS tree from bare (4) e90rie3 M with hUge Substat'i 0] 69 Category 8. Land Cover (Columns 14-28) Land cover comprises the earth's surface cover and is distinct from land use (Category C). For this analysis, land cover included: (1) Open land; land which is undeveloped or vegetated but not forested or wetland. (2) Urban and built-up; an item used in instances where the degree of urbanization was extensive enough to obscure the original natural surface. (3), Forested. In many classification schemes, forests are broken down into hardwoods and conifers and then further subdivided into upland and lowland species. Because of the agricultural and open character of the study area, forest cover was generally confined to areas along water-courses and in scattered woodlots; not in large tracts. Detailed forest-type analysis and inventory was therefore not considered necessary in this study. In those areas categorized ‘ as "forested," tree crown coverage was generally estimated at least fifty percent. Because the imagery was taken in May and the decid- uous trees had no leaves, this percent coverage was approximated from bare, deciduous branches. (4) Other. This category was used only twice when other cat- egories were inapplicable. Those instances occurred in Cells (10,12) with huge water storage ponds and (12,46) which housed a large power substation. (5) Wetlands. Because the imagery was flown in May, 1974, following a particularly wet spring season, the wet areas appeared quite prominently on the imagery. No attempt was made to separate forested and non-forested wetlands, chiefly because the deciduous trees h could n wetland high-we only wh zone, 03 (t Classifi (7 Classifi (8. (9] dnce. Land Study area 70 trees had no leaves. Seasonally wet areas were also identified but could not be distinguished from all-season wetlands. Ponds and small wetlands were sometimes difficult to differentiate because of the high-water levels. Consequently, ponds were classified as wetlands only when emergent vegetation was apparent; generally in the littoral zone, or edge. (6) Major Lake; over 100 acres in size (taken from the DNR Classification System). (7) Minor Lake; less than 100 acres (taken from the DNR Classification System). (8) Major Stream; the main branch of the Kalamazoo River. (9) Minor Stream; other water courses with a natural appear- ance. Category 8. Land Cover - Accessory Information (Columns 16, 19, 22, 25, 28) This category applied to lake or river frontage properties only and was intended to express their predominant condition relative to development. 1 - developed 2 - moderately developed; at least cleared of forest cover as in farm areas 3 - undeveloped; natural forest cover along banks remained intact 4 - frontage under construction 0 - information not applicable Category C. Land Use (Columns 30-44) Land use represents the principal activities occurring in the study area at the time the imagery was flown. The following land uses were inventoried: (1) (2) (3) (4) (5) (6) (7) (8) 71 Cultivated cropland; farm land in production with soil tillage or crops evident. Tree Farm; evident plantings of conifers. Orchard/vineyard. The pattern for this category is easy to identify when large areas were so used, but small ones were less likely to be correctly identified. It was also not always possible to determine if the orchards were active or inactive. Urban non-residential: This was used when it was evident that commercial/institutional/industrial or similar non- residential development was present. Park/playground; determined by presence of athletic fields, baseball diamonds, etc. Residential; areas in which housing is the principal use. Density of housing was not considered in this category, but residential was indicated only when sizable enough to warrant a decile range in a cell. Other or undetermined. Often by checking contiguous areas it was possible to get some idea of the activity in ques- tion. For example, Cell (04,42) has 60-70 percent classi- fied as other or undetermined. It is a massive construc- tion site and was noted as such in the margin of the coding sheet. Extractive; presence of sand or gravel surface excavations. This category was checked on USGS topographic sheets, which were used to supplement.information obtained on extractive operations. Sixteen sites were recorded from the topographic maps that had not been identified on the imagery. These sixteen sites were subsequently re-checked on CIR photos flown June 4, 1974 at a scale of 1:31.680. The results of the check were as follows: - one of the sites in which a gravel pit was supposedly located is now within a massive construction site and is no longer intact; i.e. Cell (04,42) - one other site in Cell (01,08) is now a lake surrounded by a subdivision; - eleven of the supposed sand and gravel pits are no longer visible. In most cases vegetation covers all all or nearly all of the sites; - three sites remained questionable as gravel pits in an active state. These three sites were, however, recorded as extractive. (9) No Apparent Use/Idle. "Idle" described open areas where vegetation oftentimes was starting to obscure the formerly even edges of farm fields and where various non-cultivated land 1 detail USE C3 T red filr the basi (l 72 plants and trees appeared randomly through the fields. This category also accounted for those areas covered by wetlands, forests and other natural features which appeared to accommodate none of man's activities. Category C. Land Use - Accessory Information (Columns 32, 35, 38, 41, 44) This category provided for the identification of features of land use which could be seen on the imagery, but which were more detailed than the classification components in the principal land use category. airport rail station mobile home park none OwN—I II Category D. Drainage (Columns 46-49) The absence of much vegetative cover and the ability of infra- red film to detect moisture differences in the bare soil provided the basis for this comparative category. (1) Very poor drainage. This designation applies to pond/ lake/wetland/watercourse andiwherever surface water appeared to be present. Soil if not water-covered, appeared very black indicating a high moisture content. It cannot be assumed that such soils were alluvial, however; only that they were quite wet. (2) Poor drainage. Tones appeared dark grey but it was evi- dent that drainage was better than in the very poor category. (3) Moderate drainage. Tone was lighter than in the poor category indicating somewhat drier conditions. Some mottling of soils was also usually evident. (4) Well drained. Tone of ground was very light indicating dry conditions. Relative to the other assessments of drainage, this is the lightest in tone. Depart an ess the f0 1 l l ( ( ( 01 Quality by the r (MDSHT) lnous r0. road and mine Gene Bit gTa inol COUI ness tYDe OQSS 73 Category E. Transportation Routes (Columns 52-56) These were derived from a county map prepared by the Michigan Department of State Highways and Transportation. Accessibility is an essential element in the development potential of any area and the following types of transportation were identified: (1) US 131 Freeway interchange (2) US 131 Freeway (3) paved road (4) bituminous road (5) gravel, soil surfaced or similar (6) divided highway (not a freeway) (7) railroad (9) road bridge (0) no roads Only the highest quality road in a cell was recorded. The quality ranking of roads as used in the General Highway County Maps by the Michigan Department of State Highways and Transportation (MDSHT) is as follows: freeway, divided highway, paved road, bitum- inous road and gravel or soil-surfaced. For clarification, the paved road and bituminous road definitions are here given as used to deter- mine General Highway County Map classitications. Bituminous Road - An earth road, a soil-surfaced road, or a gravel road to which has been added, by any process, a bitum- inous surface course with or without a seal coat. The base course of which is a nonrigid type and the combined thick- ness of surface and base is less than 7 inches. Paved Road - A bituminous road as described or higher surface type, the base course of which is a rigid type of any thick- ness or a nonrigid type of such thickness that the total depth of surface and base is 7 inches or more in compacted thickness. The divided highway component, (6) in the inventory was added after much inventory work was completed. As a result, the road quality sequence of the MDSHT was not preserved for this inventory. In conducting this Transportation Route inventory, if a road was locat cell. T? shared a correctly gating t1 access fl viding 1c Thi those ty; POtential Parcel. (1) This of deVElop me “We 74 was located on the border of a grid cell it was recorded for that cell. This implies that duplication occurred when two or more cells shared a road at their boundary. As a result, it is not possible to correctly determine the number of roads in the study area by aggre- gating the roads by grid cell. In addition, if the US-131 limited access freeway was located in a cell, the highest quality road pro- viding local service was also recorded. Category F. Amenities/Disvalues (Columns 58-62) This category of occurrence data was incorporated to identify those types of uses which could add or detract from the development potential of a land parcel, depending on the ultimate use of that parcel. (1) golf course (2) athletic field (3) track (race track, go kart, etc.) (4) junkyard or dump (5) cemetery; the topographic maps supplied most of this in- formation since grave stones and driveway patterns in small cemeteries could not be accurately identified at this scale. (6) other; this category was clarified in the margin of coding sheets. (7) drive-in theatre (0) none identified Category G. Residential Settlement Patterns (Columns 64-65) This category was designed as an aid in recognizing the types of development existing in the area. All density determinations were subjectively derived relative to residential settlement in the study area is generali ment may b1 tunes in e each cell. ately be c acre was I (ll (2) aid in and the identif 75 study area only. Each designation of low, medium, or high density is generalized or averaged for the entire cell, eventhough the settle- ment may be present only in part of the cell. The number of struc- tures in each call were visually approximated and then averaged for each cell. Because an actual count of residences could not accur- ately be derived, a traditional density figure in number of homes per acre was not given. (1) Low Density; less than two dwellings per acre. This was the major type of settlement in the test site accounting for 526 of the 576 cells. (2) Medium Density; between two and five dwellings per acre. A total of 31 cells were included in this category. They were generally located in or near towns like Dorr, Moline, Wayland, Shelbyville, Martin and near Plainwell and 0t- sego; on good roads near the towns, and peripheral to high density areas. (3) High Density; six or more structures per acre. Only 8 cells were so identified and recorded. They were either in Plainwell, Otsego or Wayland--the major communities in the test site. (0) No evidence of residential settlement. This grouping was used when the natural cover of an area appeared intact or was completely farmed and where no structures were dis- cerned. Only 11 of the 576 Cells fell into this category. Category G. Residential Settlement - Accessory Information - Pattern of Settlement This category of accessory information was developed as an aid in assessing the probable efficiency of residential settlement and the resulting service requirements. The following patterns were identified. (1) No pattern; used when no settlement was apparent (2) Scattered (3) Linear or Strip (4) Th provide . problene 76 (4) Clustered or Neighborhood (5) Subdivision. Newer than the neighborhoods, many of the homes which fell into this group were apparently constructed as a unit development. Category H. Special Concerns This group represents features identified on the imagery that provide additional insight to the characteristics and potential problems of a particular parcel. (1) Artificial drainage or channel straightening present (2) Residential construction in progress (3) Indeterminate construction in progress (4) Water impoundment (dam) (5) Other (6) Ponds (7) Eroded areas evident; exposed ground without protective vegetation (8) Oil well (determined from topographic maps) Miscellaneous Points Regarding Coding Copies of the actual coding sheets are shown in Appendix B. The following miscellaneous points should be noted. 1) The information in Columns 58-62 concerning transportation facilities was gathered from the Allegan County Highway map produced by the Michigan Department of State Highways and Transportation in September 1969. 2) On the original coding sheets the information gathered from the USGS topographic maps was recorded in blue to distinguish it from information derived from the remotely sensed imagery. 77 Reproduction of the color differences was not possible for this text. Color coding may be of advantage to other users of this system, how- ever, and should be considered. 3) Space for general comments was available on the left side of the coding sheets preceding the columns with coding data. This area was often used to document information recorded as "other." For example in Cell (02,43) a sewage treatment plant was identified and recorded under the Amenities/Disvalues category as "6-Other," but was documented in the area to the left of Column 1. 4) Gravel and sand pits were often suspected from the imagery and were re-checked on the topographic maps. Oil wells and some sand and gravel operations were recorded from the topographic map that had not been seen on the imagery. Subsequent checks of the data recorded from the topographic maps concerning such features generally indicated that information on these maps was largely out- dated in relation to sand and gravel excavations. 5) In Columns 52-56, if more than one railroad was identified in a cell, the code for railroad (7) was recorded more than once. 6) In certain instances it was difficult to determine whether or not seemingly idle or unused fields were cultivated and lying fallow, or used as pastureland, or permanently idle. In most situa- tions of this nature, the assumption was made in favor of an agri- cultural use and assignment of deciles reflected this assumption. It was considered preferable to say farmland was producing than to assume it permanently idle. 78 Field Checks A total of three field trips were made to the test site during this study. The first was made in June of 1975 before any photo in- terpretation was done. The purpose of this first trip was to become acquainted with the study area on a very general basis and to obtain an idea of the character of the environment. This field trip con- sisted of a "windshield survey" only. The second and third field trips were made in August and October of 1975, respectively. Each trip was made with specific sites identified for checking. The sites checked on the October 12 field trip are given in Table 6. Table 6. Examples of Field Checked Sites. Cell Number EagmrIggggis Correct Use ifiiglgiiy 07-02 junk cars Auto junkyard yes 06-03 junkyard Auto jhnkyard yes 07-09 cemetery Go-Kart track no 08-15 junk cars Packing Co. with Truck bodies scattered no 07-21 wetland NOT ACCESSIBLE -- 12-21 landfill dump with exposed sand no 08-31 lake or wetland lake yes 06-34 junkyard dump yes 10-36 landfill abandoned excavation no 08-38 multi-family farm buildings no 04-42 Construction site Construction site yes 07-42 dotted area Skeet and trap shooting range no 12-18 marl deposits INACCESSIBLE -- k the inver land resid data I study photo grid CI for eac ception was obt. from th' in Chapi 79 Summary An inventory or classification system was developed for use in the study area with small scale color infrared imagery. Eight major inventory categories were established, namely: relief, land cover, land use, drainage, transportation routes, amenities/disvalues, residential settlement patterns, and special concerns. Inventory data was gathered in a four step process which involved dividing the study area into one-quarter mile-square grid cells, correlating the photo grid with actual ground locations, numbering the individual grid cells with X and Y coordinates, and recording all interpreted data for each grid cell in a computer coding sheet format. With the ex- ception of topographic and transportation route information all data was obtained from small-scale color infrared imagery. Data generated from this inventory provides the basis for the environmental analysis in Chapter V which follows. 80 FOOTNOTES 16Michigan Department of Natural Resources, Michigan Land Use Classification System - Part 3 (Lansing, Michigan: Michigan Depart- ment of Natural Resources, February, 1974). 17R. Shelton and E. Hardy, Design Concepts for Land Use and Natural Resources Inventories and Information Systems (Ann Arbor: University of Michigan, 1974), p. 3. CHAPTER V EVALUATION OF THE PHYSICAL ENVIRONMENT OF THE STUDY AREA AND IMPLICATIONS FOR DEVELOPMENT This chapter presents an evaluation of the physical environ- ment in the study area based on data gathered from the inventory and analysis method in Chapter IV and describes the method used to make the evaluation. Although the state-of-the-art review of envir- onmental analysis methods in Chapter 11 revealed that an abundance of methods have been developed, procedures for actually evaluating data obtained from various inventories are obscure. In the case of evaluating data derived from remotely sensed imagery there is very little written on procedures which have previously been tested that can be applied to planning in large areas undergoing urban expansion. This chapter describes such a procedure. The most important land use trend in the study area for the next several years will be increasing urbanization. The location of urban development will have a significant impact on agriculture, recreation, and the entire natural resource base. It is therefore important to understand the existing natural and urban conditions, and their probable interactions which might affect the location and soundness of development. It is readily acknowledged that interpre- tations and evaluations made herein will serve only in an advisory and educational role in the eventual location of development. Future 81 82 development in the study area will result from a complex interaction of political, economic, cultural, institutional and environmental factors. In consideration of this fact, no attempt has been made to identify a particular quantity of land that will be or should be developed. As Anthdny Downs stated, Urban development and growth involves dozens of important variables each of which could reasonably take on several dif- ferent future values. Some of these variables are: 1. Location of new growth in relation to existing metropol- itan areas. 2. Contiguity of new growth to smaller existing communities located beyond the continuously build-up portions of met- ropolitan areas (including outside such areas). Type of planning control. Level of quality standards required in new constructions. Degree of public control over new urban development. Degree of public subsidy for new urban development. Nmmbw Distribution of housing subsidies among various income groups. 8. Degree of social class integration. 9. Degree and nature of racial integration. 10. Mixture of transportation modes. Just considering these ten variables, and several arbitrarily chosen values for each one, yields at least 93,322 logically possible combinations--each representing a potential form of future urban growth.‘8 Environmental constraints may also be considered a variable in development decisions and it is with such environmental constraints that this chapter is largely concerned. The analysis which follows is presented in two parts. Part A provides background information on the interactions between development and the natural environment. It iiiv of’ fiVI erIV‘ USE OF vari maps co Scale ( necessai C911. wt 83 It is presented on a category by category basis as followed in the inventory of Chapter IV, and applies conventional interpretations of development constraints where appropriate. Part B provides a five-step method for evaluating the inventory data and assesses the environment of developing locations in the study area to illustrate use of the method. Interactions Between Development and the Environment Category A. Relief In analyzing topographic variation in the study area, broad ranges of slope were recorded: 0-6 percent (level), 6-25 percent (undulating), and over 25 percent (hilly). Slope classes commonly used in soils mapping in Michigan are much more detailed as f01lows: A 0-2 percent, 8 = 2-6 percent, C = 6-12 percent, 0 = 12-18 percent, E 18-25 percent, F = 25-35 percent, and G = over 35 percent (Whiteside, Schneider and Cook, 1968). The very broad ranges used in this study were selected because of variations in scale of United States Geological Survey topographic maps covering the area. Most of the area was covered on the small scale (1:62,500), fifteen minute quadrangle. In addition, it was necessary to generalize relief over the basic data unit, the grid cell, which was one quarter mile square. Such generalization could be rapidly made with the broad ranges of slope. In general, the topography in the study area has relatively little variation. There is less than a 200 fOOt difference in ele- vation between the lowest and highest points in the 144 square mile study area. The lowest point is an elevation of 695 feet in Section 84 23 of Otsego Township on the banks of the Kalamazoo River. The highest elevation is 885 feet and is located in Section 29 of Way- land Township. Infbrmation gathered on relief in this study is not suffic- iently detailed to permit site planning but it does enable one to visualize the topographic character of the site. This capability becomes important when determining developable areas since different uses require differing terrains. Scenic sites with rolling to hilly surroundings are a particularly valuable resource in the study area since most places are level. Slope is a significant factor in urban development since building costs generally increase with steep slopes. In addition, accelerated erosion becomes an important concern in areas of steep slope cleared of stabilizing vegetation. Farming likewise is poorly suited to such areas. No arbitrary determinations of which slope ' can or cannot be developed are here advanced. As Carl Steinitz noted, Topographic slope is a prime example of an objectively meas- urable variable used in a variety of interpretations, notably for its impact on urbanization. The cut-off point between "good" and "bad" slope for urbanization ranges anywhere from 9 to 45 degrees. Thus it appears not to be the data inventor which distinguishes a method, but interpretation of the data. 9 Category 8. Land Cover The surface of the study area as seen from the air and recorded in the inventory of Chapter IV, consists chiefly of open land, wood- lots, wetlands, urban/built-up areas, lakes and streams. This rela- tively simple categorization of the study area's major land cover features can be very useful in a variety of ways. 85 The inventory data in this category can assist in understand- ing the hydrologic regime of the area. For example, areas with vegetative cover are able to hold precipitation and facilitate its percolation into the soil. Conversely, areas where vegetation has been removed or where the surface has been sealed, cause increased rates of runoff and consequently decreased groundwater recharge. The various types of land cover represent habitats for differ- ent species of plants and animals. The amounts of such land areas are important in determining wildlife support capabilities. Tran- sitional or interface zones between cover types such as open fields and woodlots are generally supportive of a highly diverse plant and animal community. This is known as the "edge effect" and indications of the areas likely to support wildlife can be obtained from the inventory. Forested areas or woodlands play an important aesthetic role in the countryside and near developments. They moderate the effects of wind, and temperature, absorb certain air pollutants and may serve to buffer noise. Woodlands stabilize soil and slow runoff from precipitation. Wetlands are perhaps the most critical natural feature in the environment. In those few cases where wetlands are found adjacent to a lake, particular care should be taken for their protection. Dr. Christa Schwintzer of the University of Wisconsin described the rela- tionship of wetlands to lake water quality in the following manner. Wetlands are important in determining the quality of lake waters and the seasonal flow rates into the lakes. In addi- tion, they affect the quality and quantity of stream flow and ground water. They improve water quality by filtering out wastes and breaking them down into simple inorganic compounds Fu ch fou lea irr Y‘eCl 86 which can be taken up by the vegetation. Wetlands also reg- ulate the seasonal flow of water by storing water temporarily during periods of high input due to snow-melt and rainfall, discharging it during periods of low input. The effectiveness of individual wetlands in performing these functions depends on many characteristics including size, location, soil, vege- tation, and relationship to the regional water table. Aside from the value of wetlands for water quality and their obvious use for waterfowl habitat, wetlands are important to urban areas. If located upstream of an urban area, wetlands store surface runoff and help alleviate the threat of floods. The soils or sub- stratum in a wetland are generally organic and require extensive and costly modification in order to support foundations. Also, the lo- cation of the water table in wetlands can serve as a relatively rapid indicator of groundwater levels and the feasibility of successful septic tank operations. Streams and creeks are highly valuable as well. Aesthetically, a high quality stream environment draws recreational enthusiasts. Functionally, though, streams are a seasonal source of groudwater re- charge. They represent corridors for the migrations of birds, water- fowl, and various land animals. Development of a watershed often leads to increased runoff and sedimentation which can result in irregular water flows in the streams as well as irregular groundwater recharge. Lakes are even more susceptible to the negative impacts of such development and are prime attractors for this development, which often takes the form of second homes on the lakefront. The study area is bounded by urban centers of employment, and pressure for lakefront development in the currently rural settings is sure to increase. The evidence of lakefront and other shoreline development 87 was recorded in the inventory. Category C. Land Use Existing use of land in the study area has been inventoried as principal activities that can be rapidly identified on small scale color infrared imagery. Because activities conducted on any single parcel of land can influence activities conducted on nearby parcels, an important consideration in locating development is compatibility with surrounding land. The land use category provides for a rapid identification of various types of activities and can be used to assess probable compatibility of proposed developments with existing uses. The principal land use in the study area is agriculture. Much has been written about the increasing encroachment of urban develop- ment into some of the best agricultural land in the nation. The trend is similar in Michigan. As Whiteside and Schaner stated, Cities were initially developed on or near the best agricul- tural 1and because during this formative period of city growth, Michigan had an agrarian economy. Food (particularly bulky and/or perishable produce) could be marketed near where it was produced without spoilage in transit.20 As the labor demands of agriculture decreased and rural to urban area migration accelerated, the swelled populations of cities forced expansion and "as a consequence, the urban area growth has covered much of our better farmland."2] Substantial acreagestmam>_z: 9:5 z “'3 __..__-_._‘.._.__.. .. m--.. l'T... ~_- o a -~ g. E : 5Q®b~§®o~bfifc~c~c~0~o~0§o~ mamu~o~c~o~c~o~o~o~u~m a 3 ‘3 -_ -_ 317. m u 3‘ \ _W W... ___ WW - - , ,L a ‘ :. o\\NN\oN\ Go °°°_-- \ on _0535350-4“ ‘3}: g g ~- -— 1»: r! H; III I .. '* 'é‘,‘ 0- : —— — ~ — :+—_- ‘ > ‘ ‘."":‘_ ' _" ' __: ___;‘T' " -3? O- “ \Q-_\.§- -____...,.._-.__ ___-- ‘3 0 w— m .- 3 __ Q\°°Q\Qs -_.-.__._. I‘!,_ z = a - Q M a f 2 n 3;\ “31°0- ....__-_ 5”__>__‘ - g ‘ 3 - ‘3 Wfi:\§§u§\\- ___ __ _3 - d g - I: I g : \h\'n.n$~\<~5<~{«\75\ \-§-°h\"“.\°‘\°‘\° .- Q U a. g; 0~§~0~Q~$O~0~O~g~°~§~g~ WWR‘NG-wa-mmmcxmmo. “94137 2 2 E oLfi-NQWo-SbfiQ KQQ \o-SbVN75}‘n'ln \\°o5°66-00~Q~0~Q. 3 3 \\\\\\\\\\\\\\\\‘Q\§§\§~§\§\‘\\ \§_:;' 5: U 2‘ .. m-..._...__JJ-".- :2 _ ___"- _ - -_-_-_.-. :g_::* a —- _————-~ ——---—-—-- - 2 — ————-- —--—- —-~::» g p """ «cum '"_'“‘“‘ ’“ “"7 “‘ ‘““""’ ” _j‘ 2" g \ K“‘\ '___---__’ ' "-1 fl 2 . "f”i-I‘WWh‘“ ._ H W ' _ ‘ _" M ' .IE‘ ’3 a m7 [123232\ PCQ .9O°oo~. Q b Qm °~o' "'- "L: 5" L72: xx \ \\:_\\m'mm'~\’\\ ‘“" gm} \ - 1"; 3 g L1 | . I_I . . . a mm o 7“ .._.._ H . o a ' , ' ' 2 \q~\0§;_§_\_q QQQ$\\QQ~‘OQ~ {16‘ Qc~\%Q~Q '3‘ :: «_Hn mjmm‘fi—fiim‘lfiQ_6~:: W\~_N N\55)\m_\\_\ ‘\ 1 g I ‘-\ S .__.”¢ '.--.’..\. !I_tI c ' I ‘ 1‘ ‘ ‘ ‘ ' I ‘ | ' ‘ ‘E 3 noqx\QQNEGHQQQQQOQV-oosoQUOQ‘QQc-O z 2 ”mmhlffilh'kflb fing§mthlsmfioonhbmhmhh~m -; 2 .--_.-~- -_ -_.- "Cd-H1 ‘3’ “““ 5'“ 'Z’“ ' =;—. - — v—_--—- ——- -——-—-— —- -- — -- -— - ——- --—-—~- J \6 9 °~°’°‘ “V- 517% ..°:9~!°.-=-_:9m o~v~-__ _, o 4) 6670‘ _N o~_ _ _ _5 _Ac~0~0~ $95 o~o~o~_o~s: . __W .. .5. 7 hot>®o~o~ NQ3~_§Q¢I>_¢€¢;6—Nm;v- Lbdkoo~§"cfi€im9h 2 « SE“N“mmflflflfifih@9¢11yvVYVY°°OQQQH a .. # - - ~ 3 . mm mm ifi’é‘fi‘m'fi'fi’m"m”m€dwh§j:m”nimmp:§jmcfiyfy"§v$yf g .. cococoocouwooooob‘ongooééobéqca. 4 _W -L......_.._ .... W. O u ._ - ._ -- . - .._._..u5.. .- .r- ‘—-7- -:::"'—-'“~'* :— ‘3 E a” i9. & Ct-2 132 «tau ..,.l.!m......:u......l.1.n....mnw.l..rvmmmmmmmmm«plumm...LLnEw2:. mun. rnu.”"u.urnnn_._ Eggfim .« . x . m . Hm . m 0 ¥‘ \ «x o M Q. \ mm s. 5., mmo map jg bu mx o 1 5 \ \¢ mm n\ on.m. mmfl.» +V b w\ a m m \ . Mm \ Q Q \r vw «Q fix o m h \ \m Mm m\ 60 Q v ma» 2‘ Q \ s 5 u mm _. h..\m-. mxsueo v» q n\ c we \ \s _ .29 om- ..lem2n »‘ a m\ o n T u a. \.,xg.§: an h.,; -QaQqs w» .d‘n\ a “m_ w T “ .ym mzwvfi ,un;;:m.ow\wr v» ; at a fii_& o -. fl." “x.s\ _mm.mbg_hhgfig $w q d\ o_ gwwg mm H. w . . % +2 ..._ m:_sm.mL »+._ m_«T 6 ,f.m 6 f. . m <5 %\.. w ..NQ+<&.Qm. w‘_ \.u< Q fi__w o. 1“ _ Hm m2.“ .. .m\<.mmqm MQ w: \ U m;_¢ _j "m . cm,m_ mv 1 R\ _ o _ Emm;.. g _. Mfluai. f .. _Qmmi. a. U. m. . a .u m:_¢;m._ :_ m 3. $3. . _ w. an. wv .\ h\ _ o _. may a“ on m _ _ >9 u; m a Q \m cm . «v _Q d\ _ G _. _t= ¢__6. . m _ ”m s: . . bQNOm An N 3. . .3 . .0 . a... .M 31..., M .. . _: a: _ . t. o... . uw Q ".fi\ _ _Q _ M. .m w “on . . M ~ . _ u _ max . n . . 9‘ Om 9* ..Q .“S mQ . . m . .M. . .. M m _ _ . _ _ _ VH5 H . m w W; w 6.3.. o .9. _fi; ...EQ f..m”_... Q.“ .0 fi\ _. F . n m_ __ _$w M a . w M. ‘ — mm“..$<.. ...m . w ‘. 3 ..i. ”o.m_._m._aw§ 17:7 S, ; 2:... g» L» «\ -.crmwf: g -_Wrg.+h.;m .-,m..¢2 .. ~\->m ammo. u 523:32.“ o..o...o330 7...... 3 .2: ..Lrazn a. 3.3.... v3: 2.... :2? 2...-.. .7. .. .. . . _ .:- urrwfl ""PF I «A a La. onwul... .. +flmy .Mr #3 WFVHJL. ”HF”. ER...» ...1 . .._ w H. ..u m. :3 oz 5.3: 2.2 0268 4:3 I 50:20:. .2223 _1- . a .0 Q .0: § 51.3.30: >h-m8m>_z= mb‘hm 2(0_IU.2 “V! ,Hfiu..- ‘ $00 0 m0x&M»M&~ ummmmnnm. RVVvvvvvwfivw v §’ \Vx QVBOQQQOQQp fl. xiv 09° *Vurw QSJPQ .... -_....._.. . 133 luau Q“. 8 k 34. E .g 53.0: ai¢c8us 2.9. 0283 4:3 I €052.09: $5.23 5.5.522: :45 2(9322 . so oosonvnua «Q- o. .. c.n.v_9n...o_o-no nann— ‘l’ I: . . ..I. :I. 2032.0ua0 <~=m¢u>=s= wh:m 20:23.8 3:. 135 _ _ . “UL-.. ...L.. a: 3105.0 040C OOuOnvnn_ m \ o¥c~c-’6~o-'c-‘c~'c~ o~ c~o~ Q~Q~o~ Qboom L.. ___‘QV‘ 5805 02.900 (...(0 .I >80»(80¢(a awhamiOU VS»? .3380: >.—_muw>.23 whip“ Z¢O_IU_E ns...¢7-fi:<. .’W¥CU.ECO.A.>:.w 20:233.. «:3 15.1... 3&4 0.2.25 52...: 136 Ofliouzu 0.5;. .83 71.5.9 .0 _ . . _. . . __ _ _ a. ...M . . .. _ a 2 . . 2.3:. 2...»? 3.... _.~ ...TT...T..= . . _L; , . . . “L. m. _ ,. . m u . m L. . . _ .._ . .. . . . . . 3a.: . E . M .3 E _ s _ CW ..wNGQm—F _ H. . T L “M _ m. QM” my) .6 \_m_ Kc H\ _ w \ «r in «£6 110$ 11:. 2:. onuoaon~_ :35 02.000 ‘h‘n I >80h(82‘.- Inhattou O.“ 8Q\ 33 .V-e .3358: >h-mdw>-z: mh‘hm 2‘Q_IU—z 1#:~E§L.7cw 20.2.3.3 «:3 137 .08 138: 1.9. 0233 «:3 I 59232: 35.23 5.2322: un80h<¢0¢¢a smug-5200 >._._m¢w>_23 wbcm 205.330 (:8 vll1,'ln.. ...1.II n 11‘911 139 #1.: L L s goo o. «:40 (_n c n .. ‘ m p L q_—n. admut to. m. m .- A C .. . .. , v - a J o ‘61 <3 G QT Q>f59~- -wm hwooooo 0‘ C) ounOnvnn_ .23.. 0233 (in I 59:09: 35.83 on a 8 .2. £3322: ~25 23.5.: .:...:: 3.1...." 5:5... .2. 140 «lav ON 8*\ 32 .wffiNv-MN “3%!“ '5 ' Ten—v7 {BI—W \\\\\\WN§\\\\\\\\NN\ .Né .01 1:00.; Cu:(-8at g; oononvnn— L _w _m . & osc- q~ »‘§‘@\9»\ o~ _ mo. Ly :9 . E , L9 :2 . L” L. . .. .v H .m . E k _ L _Q. VF... I’. 9: V9“ V. ‘k§-\V‘ Tm \\\\_\ 1.“— .93" .‘_ .‘. KTQ as” Q q 9 . . _. _L .m . . . n~: r .. ”...._......,....n..fl L . _ _ Thinuqui 2.9. 0283 <29 I 29:23 55.53 5.3322: 32‘: 243.222. $5.321 ’ovcdficohzm 20:233.. «:3 1.1!..171. 1 I! 141 a ..o uh\ .05. 17“. \\\\\s\‘ IE. li‘gt 2.0; 02300 (p¢Oh<§(._ 32.2100 55322: .35 23.222 $.30 p oosOnvn«_ infrl‘ 3391:1356 89:33.0 :1. 142 IF" «Iau . 83¢ OINOnvnn— \ \\\\\ ~D 0‘83... . __0',‘ . _- 0‘ @0‘0‘0‘0‘6‘ \\\\\\\\\ \\\\‘\‘\\\\\\\\\\ \\\\ K-$-.;T ;. Us ? __ .x 6 E . .c._ m C" n. a. I .2. .:...:. 32.0. 02.000 <23 I {oi-0:. 35.200 an a Q .2. 03% 5.2.8.. 5.22:2: :45 2.6.20.2 i--. i a... d.¢IO:<. 10.5!!! toxicm 20:23.3 143 093-20 0...; :3 0. L... 3.... c e a. -3 _r;. « L m! mm . . . . . . . . . . _.. _ . _ _— . .. . p _ I. _ .3. t..__o— h. N. mu‘o 0M._ WW. ”DO mm Qrfl ..MN. _0M. _0b_ .‘N xx 0%. . mu «0.:o> «0 _ofi mm. _\« x~._ 6N. 11...... \ . x x 1.2.... . x x . x «it... xx xx xx :0. . . OOxOnvnn— =3 . 280; 02.000 80h(§0¢(u nub-.0300 Q“... Ax 8... .V‘I‘ .:..::8... >.—_m¢w>.2= mpcm 10...».33 «:3 146 «Ian 0N8 0“ u.‘. E .2 138.; 1.2 0253 «:3 I 502-0.4.. 35.33 3.2322: and: z<0_:u.£ . SECSm '_—.~_ .\m_~:__— ‘ \ ~ \ w m canonVna. zo¥$vfico&$cm 20.3338 :3 3......‘5 cal BIBLIOGRAPHY BIBLIOGRAPHY American Law Institute. A Model Land Development Code. Philadel- phia, Pennsylvania: AmeFTCan Law Institute,_April l5, 1974. American Society of Photogrammetry. Manual of Photographic Interpre- tation. Washington, D.C.: American Society of Photogrammetry, 1960. Anderson, James. A Land Use Classification System for Use with Remote Sensor Data. Washington, D.C.: U.S. Government Printing Office, l972. Avery, T. Eugene. Interpretation of Aerial Photographs. Minneapolis, Minnesota: Burgess Publishing Company, l968. Bosselman, Fred and David Callies. The Quiet Revolution in;Land Use Control. Washington, D.C.: Council on Environmental Quality, l97l. Burchell, Robert W. and David Listokin. The Environmental Impact Handbook. New Brunswick, New Jersey: Rutgers University, l975. Clawson, Marion with Charles L. Stewart. Land Use Information - A Critical Survey_of U.S. Statistics Including‘Possibilities for Greater Uniformity. Baltimore: John Hopkins Press, l965. Council on Environmental Quality. Environmental Quality, Washington, D.C.: Council on Environmental Quality, 1970. Davis, Charles. A Study of the Land Type. Durham, North Carolina: Army Research Office, March 1969. Dee, Norbert, Janet Baker, Neil Drobny and Ken Duke. "An Environmental Evaluation System for Water Resource Planning." Water Resources Research. Vol, 9, No. 3, (June 1973). Dickinson, Robert E. Regional Ecology - The Study of Man's Environ- ment. New York: John Wiley and Sons, Inc., l970. Downs, Anthony. "Alternative Forms of Future Urban Growth in the United States." Journal of the American Institute of Planners, XXXI (January l97DII Dueker, Kenneth J. and Joseph S. Drake. The Land Use Policy_and Planning_Assistance Act of l972: Information Systems ImpTications. I47 148 Technical Report No. 14, Institute of Urban and Regional Re- search, Iowa City, Iowa: University of Iowa, October 1972. Dueker, Kenneth J. and Frank Horton. Utility_of Remote Sepsjng_Data for Urban Land Use Planning, Iowa City, Iowa: UnIversify off— Iowa, n.d. Estes, John E. and Leslie W. Senger, eds. Remote Sensin --Techniques for Environmental Analysis. Santa Barbara, California: Hamil- ton Publishing Company, 1974. Fabos, Julius Gy. "An Analysis of Environmental Quality Ranking Systems," Recreation Symposium Proceedings. Washington, D.C.: U.S. Department of Agriculture, 1971. Gentili, Joseph and Mitchell J. Lavine, eds. Owengnvironmental Study. Ithaca, New York: Cornell UniversIty, 19742 Governors Special Commission on Land Use. Land Use Report - Appen- dix B. Lansing, Michigan: Office of the Governor, January 5, I972. Graham, Edward H. Natural Principles of Land Use. New York: Oxford University, 1944. Great Lakes Basin Commission. Appendix 3 - Geologyand Groundwater. Ann Arbor, Michigan: Great Lakes BasIn Commissioni—1975. Appendix 13 Land Use and Management. 1975. Appendix 14 Floodplains. 1975. Appendix 19 Economic and Demographic Studies. 1975. Hackett, Brian. LandscapePlanning. Newcastle-Upon-Tyne, England: Oriel Press Limited, 1971. Hardy, Ernest E. Interpretation of Aerial Photographs for Land Use Classification. Paper presented at the Conference on Land Use Information and Classification Co-Sponsored by the Depart- ment of Interior and National Aeronatic and Space Administra- tion, Washington, D:C.: June 28-30, 1971. Higgs, Gary and Mark Sullivan. A Comparative Analysis of Remote Sensing Scale/System Attributes for a MGlti-LeVEl Land Use Classification System. ‘East Lansing, MIchigan: Project for the Use of Remote Sensing in Land Use Policy Formulation, Michigan State University, 1973. Hills, G. Angus. The Ecological Basis for Land Use Planning, Ontario, Canada: Ontario Department of'Lands and'Forests, 1961. 149 Hitchcock, H.C., T.L. Cox, F.P. Baxter, C.W. Smart. "Soil and Land Cover Overlay Analyses," Photo rammetric En ineering and Remote Sensing. Vol. 41, No. 12, (December i975). pp. lSl9-1524} Isachenko, A.G. Principles of Landscape Science and Physical- Geographic Regionalization. Translated by N.J. Rosengren. Carlton, Victoria, Australia: Melbourne University Press, 1973. ' Kaiser, Edward, K. Elfers, S. Cohn, P. Reichert, M. Hufschmidt, R. Stanland. Promoting_Environmentalygualitnyhrough Urban Plan- ning and Controls. Washington, D.C.: Environmental Protection Agency, 1974. Keys, Dale L. Land Develppment and the Natural Environment: Estima- ting Impacts. Washington, D.C.: The Urban Institute, 1976. Land Use Analysis Laboratory. A Land Classification Method for Land Use Plannipg. Ames, Iowa: Iowa State University, 1973. Lewis, Philip Jr. "Quality Corridors for Wisconsin." Landscape - Architecture. Vol. 54 (January 1964):. lOO-lO8. Linville, Jack Jr. and Ron Davis. The Political Environment - An Ecosystems Approach to Urban Management. Washington, D.C.: American Institute of Planners, 1976. Lyle, John and Mark von Wodtke. "An Information System for Environ- mental Planning," Journal of the American Institute of Planners, Vol. 40, No. 6, (November 1974): 394-413. McClellan, G. ed. "Ecological Concepts of Subdivision Design," The Waterloo County Area Selected Geographical Essgys. Water- loo, Canada: University of Waterloo, 1971. McHarg, Ian. Plan for the Valleys. Philadelphia, Pennsylvania: Wallace-McHarg Associates, 1963. Maltby, Richard A. A Method for Determining the Use Potential of Land. Thesis for M.U.P., East Lansing, Michigan: Michigan State University, 1965. Michigan Cooperative Extension Service. County and Regional Facts for Planning and Develppment Region 8. Lansing, Michigan: Michigan Cooperative Extension Service, n.d. Michigan Department of Natural Resources. Mapping_and Photo ra h Directory. Lansing, Michigan: Department of Naturaigesources, July 1975. Michigan Department of Natural Resources. Michigan Land Use Class- ification System,iPart 3 - Classification and Definitions. Lan- sing, Michigan: Department of Natural Resources, February 1974. 150 Michigan State University Remote Sensing Project. Users Guide to High Altitude Imagery of Michigan. East Lansing, Michigan. Michigan State University, Aprii_1973. Michigan Water Resources Commission. The Water Resources of the Lower Lake Michigan Drainage Basin. Lansing, Michigan: Michigan Department of Conservation, August 1968. National Academy of Sciences. Urban Growth and Land Develepment - The Land Conversion Process. Washington, D.C.: National Academy of Sciences, 1972. Nieman, Bernard J. Jr. Handbook: Applications of Remote Sensing and ComputepyTechniques for Recreation Planning, vol. 1. Madison, WiSconsin: UniVersity of Wisconsin, March 1974. Odum, Eugene. "The Strategy of Ecosystem Development,“ Human Iden- tity in the Urban Environment. eds. Gwen Bell and Jacqueline Terhitt. Baltimore, Maryland: Penguin Books, Inc., 1972. Ohio Department of Natural Resources. Land C_pability Analysis: The Wolf Creek Project. Columbus, Ohio: _OhioDepartment of Natural Resources, February 1974. Real Estate Research Corporation. The Costs of Sprawl. Washington, D.C.: U.S. Government Printing Office, l974. Scherz, James P. and Alan R. Stevens. An Introduction to Remote Sensing_for Environmental Monitoring, Madison, Wisconsin: University of Wisconsin, 1970. Sharpe, Carl and Donald L. Williams. "The Making of an Environmental Fit." Landscape Architecture, Vol. 62 (April 1972): 210-215. Shelton, Ron L., H.W. Charnley, T.A. Dimock, D.J. Belcher. Land_Use, Environmental Features and Natural Resources Inventory ef the Hudson River Valley. Ithaca, New York: Center for Aerial Photographic Studies, 1969. Shelton, Ronald L. and Ernest E. Hardy. Design Concepts for Land Use and Natural Resource Inventories and Information§ystems. Paper prepared for Ninth International Symposium on Remote Senzing of the Environment, Ann Arbor, Michigan, April 16, 97 . Sorenson, Jens C. "Some Procedures and Programs for Environmental Impact Assessment" Environmental Impact Analysis: Philosophy and Methods. eds. Robert Ditton and Thomas Goodale. Madison, Wisconsin: University of Wisconsin, 1972. Steinitz, Carl. "Landscape Resource Analysis - The State of the Art." Landscape Architecture, Vol. 60, No. 2 (January 1970): 101-105. 151 Steinitz, C., T. Murray, D. Sinton, and D. Way. A Comparative Study of Resource Analysis Methods. Cambridge, Massachusetts: Harvard University, 1969. Sullivan, Mark C. and Gary Higgs. .Image Interpretation for a Multi- Level Land Use Classification System. East Lansing, Michigan: Michigan State University, 1973. Thurow, Charles, William Toner and Duncan Erley. Performance Con- trols for Sensitive Lands. Chicago: American‘SociEty of' Planning Officials, June 1975. Timmons, John F. Guidelines for Developing_State and National Public Land Use Policy, Ames, Iowa: Iowa Agriculture and Home Economics Experiment Station, 1973. Toth, Richard E. Criteria for Evaluating_the Valuable Natural Re- sources of the TIRAC Region. Stroudsburg, Pennsylvania: Tacks Island Regional Advisory Council, 1968. U.S. Department of Commerce. Bureau of the Census. 1974 Census of Agriculture - Preliminary_Report,eAllegan County, Michigan. Washington, D.C.: U.S. Government Printing Office, 1976. Veatch, J.O. Agricultural Land Classification and Land_Types of Michi an. SpeciaIiBulletin 231. East Lansing, MiEhigan: Michigan State University Agricultural Experiment Station, October 1941. Vermont State Planning Office. The Vermont Interim Land Capability_ Plan. Montpelier, Vermont: Vermont State Plannihg Office, June 1971. Warner, Maurice L. and Edward H. Preston. A Review of Environmental Impact Assessment Methodologies. Washington, D.C.: Oiiice of Research and Development of the U.S. Environmental Protection Agency, April 1974. Way, Douglas. Terrain Analysis. Stroudsburg, Pennsylvania: Dowden, Hutchinson andRoss, Inc., 1973. Weatherspoon, C.P., J.N. Rinker, R.E. Frost and T.E. Eastler. "Remote Sensor Imagery Analysis for Environmental Impact Assess- ment," Proceedings of the Seeend Conference on Environmental Qualipy Sensors. Washington, D.C.: Environmental Protection Agency, 1973. West Michigan Regional Planning Commission. Fact Book Planning and Development Region 8. Grand Rapids, Michigan: West Michigan Regional Planning Commission, January 1974. 152 Whiteside, Eugene and Don Schaner. Michigan Agriculture: Agriculturel Land Use Trends and Future Needs for Agricfilture Lands. East“ Lansing, Michigan: Michigan State University, April l972. Whiteside, E.P., I.F. Schneider, and R.L. Cook. Soils of Michigan. East Lansing Michgian: Michigan State University, 19681 William, Spangle and Associates. Application of Earth Science Info:: mation in Urban Land Use Planning. Springfield,_Virginia:"" National Technical Information Service, 1974. Williams and Works. Weyland - Natural Resources; Foundations for Planning, Part 4. Grand Rapids, Michigan: _Wiliiams and' Works, 1963. Wolf, Paul R. Elements of Photogrammetry. New York: McGraw Hill, Inc., 1974. ”'ll'lllll ll (11] 1111111111111 17'“