THE EFFECT OF FREEZENG AND THAWSNG 0N SOEL MOISWRE‘. BULK DENSITY, AND SHEAR S‘YRENGTH UNDER OPEN AND FOREST CONDITIONS Thesis for the Dogma of pk. D. MICHEGAN STATE UNEVERSETY Arthur William Krumbach, Jr. 1950 0-169 This is to certify that the thesis entitled The Effect of Freezing and Thawing on Soil Hoisture, Bulk Density, and Shear Strength Under Open and Forest Conditions presented by Arthur William Krumbach, Jr. has been accepted towards fulfillment of the requirements for Ph.D. Forestry degree in Date 9 December Infomatic changes during needed on the ' timing freezir. cipitation, e; ABSTRACT THE EFFECT OF FREEZING AND THAWING ON SOIL MOISTURE, BULK DENSITY, AND SHEAR STRENGTH UNDER OPEN AND FOREST CONDITIONS by Arthur William.Krumbach, Jr. Information is available on soil moisture contents and bulk density changes during freezing or thawing periods. However, information is needed on the soil.moisture, bulk density, and shear strength regime during freezing and thawing periods, and their relation to soil, pre- cipitation, and vegetative cover. From.November 1959 to May 1960 soil moisture, bulk density, and shear strength were studied in the upper 15 in. of two medium-textured soils in Kent and Clinton Counties, in the lower peninsula of Michigan. In each county three plots were established, one with hardwood cover, one with herbaceous cover, and one bare. Texture, organic matter content, specific gravity, and.Atterberg limits were determined for each soil. Depth, density, and water equiva- lent of snow and frost; frost type; soil moisture; bulk density; and shear strength were sampled.periodically in 3-in. layers. Daily air temperatures were similar in both counties but Kent County received.more snow. During the period of continuous snow cover the Kent County Bare, Herbaceous, and Hardwood.plots averaged 3.55, 1.21, and 1.23 in. more snow than the Clinton County plots. Prior to being covered with snow, bare plots in both counties contained.more frost, and to a greater depth, then plots with vegetal cover. Amount and depth of frost were about equal in plots under similar cover. Depth of freezing in Clinton County was correlated with air Arthur William Krumbach, Jr. temperature during the eight-day period before a sample date. Depth of frost on the hardwood plot was about 3.5 times less than the herbaceous plot, and frost in the herbaceous plot about 2.2 times less than in the bare plot. In Kent County depth of freezing was consistently less in the hardwood~covered.plot than the other two plots. Bulk density of each 3-in. depth was inversely related to frost depth in the 15-in. soil layer and to moisture content in the same 3-in. depth. The relation exists because there is a repetitive proc- ess of’moisture moving into areas, freezing, expanding, and causing more free pore space. Moisture continued to move from the snowpack through this pore space--even though concrete frost was present throughout the study period. That moisture was constantly moving through the frozen soil helps explain why moisture could not be statistically correlated with frost depth. Also, moisture may exist in frozen soil in the vapor as well as liquid phase, and may or may not be present in frozen soil. Shear strengths were below the prefreeze level while the soils were thawing. However, after soils thawed and moisture contents had dropped below prefreeze level, shear strengths remained unusually low for a period of time. This is ascribed to the coating of soil aggre- gates with a thick moisture firm early in the thaw period. THE EFFECT OF FREEZING AND THAWING ON SOIL MOISTURE, BULK DENSITY, AND SHEAR STRENGTH UNDER OPEN AND FOREST CONDITIONS By Arthur William Krumbach, Jr. A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 1960 c? A2372 7/2i/5/ "Neither snow, nor rain, nor heat, nor gloom of night stays these couriers from the swift com- pletion of their appointed rounds" Inscribed over Central Post Office New York City, New York ACKNOWLEDGEMENTS This study was accomplished as part of the research program on trafficability of soils being conducted by the Vicksburg Research Center of the Southern Forest Experiment Station in c00peration with the U. S. Army Engineer Waterways Experiment Station. The author is grateful to the Southern Forest Experiment Station and the Army Mobility Research Center for making this arrangement pos- sible, and wishes to extend his thanks to P. A. Briglieb, Director, and R. S. Campbell, Chief, Range and Watershed Management Division of the Southern Forest Experiment Station, and to Mr. S. J. Knight, Chief, Army Mobility Research Center, and his staff for their advice and support. Acknowledgement is also due to the staff of the Lansing Research Center, Lake States Forest Experiment Station, for their assistance dur— ing the field phase of the work. Sincere appreciation is extended to all those, named and unnamed, who have assisted in the course of this study -- from initial planning through long hours in the field and at the calculator, to the typing and final cepy. I am indebted further to all the members of my doctoral committee, Drs. D. P. White, T. D. Stevens, V. J. Rudolph,)%.]3. Erickson, T. D. Eatery and Prof. Ivan F. Schneider, for valuable advice and counsel, and to Dr. Forest Stearns for his constant editorial serv- ices and suggestions from the beginning of the study. Identification of soils on the study plot by Mr. Edward Redmond and Prof. Schneider was a necessary step for which assistance the author is grateful. Especial thanks are due Dr. White for his guidance through all phases of the study. ii CONTENTS ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . . . LITERATURE . . . . . . . . . . . . . . . . . . . . . . Temperature Conditions Required for Soil Freezing . Soil Moisture During Freezing Periods . . . . . . . Influence of Precipitation and Vegetation on Soil Freezing and Thawing . . . . . . . . Influence of Soil Color, Evaporation, and Microrelief on Soil Moisture During Freezing and Thawing . . . Conclusions from Literature . . . . . . . . . . . . LOCATION AND DESCRIPTION OF STUDY AREAS . . . . . . . . . General . . . . . . . . . . . . . . . . Location . . . . . . . . . . . . . . Land Use and Vegetation . . . . . . . . . . . . . . Soil Descriptions . . . . . . . . . . . . . . . Climate . . . . . . . . . . . . . . . . . . . . EXPERIMENTAL PROCEDURE . . . . . . . . . . Design of Sampling Plots . . . . . . . . . . . Installation of Plots . . . . . . . . . . . . Sampling Procedures . . . . . . . . . . . . . . . . Collection of Data . . . . . . . . . . . . . . . . . RESULTS 0 O O O O O O O O O O O 0 O 0 Static Soil Properties . . . . . . . . . . . . . Soil Freezing and Thawing . . . . . . . . . . . Frozen Cores and Frost Type . . . . . . . . . . . . Soil Moisture and Bulk Density . . . . . . . . . . . Snowfall and Peak Moisture Contents . . . . SWEY AND CONCLUS IONS O O O O O O O O O O O O O O O O 0 Influence of Vegetation and Snow on Frozen Soil . . Soil Moisture in the l5-in. Depth . . . . . . . . . Soil Moisture and Bulk Density . . . . . . . . . . . Shear Strength . . . . . . . . . . . . . . . . . . . L ITERATLJRE C ITm O O O O O O I O O O O O O O 0 APPENDIX A: TABLES . . . . . . . . . . . . . . . . . . APPENDIX B: DESCRIPTION AND USE OF THE CONE PENETROMETER . VITA O O O O O O O O O O 0 iii 15 2b. 26 3O 3O 3O 35 A2 53 55 55 55 59 67 67 72 112 1uo 161 161 162 163 16 3 165 173 181 185 LIST OF TABLES (Continued) Tables Page 21 Estimation of change in bulk density with freezing at saturation (total pore space) and at maximum observed mOi Sture content . . . . . 0 C . . O C . C C . C U C O . 138 22 Inches of water in snowfall, snOWpack, and available to soil, with soil moisture changes from previous sample day to day of peak moisture contents . . . . . . . . . . 1H2 23 Mean shear strengths and moisture contents, averaged for two samplings, before December 21, Kent County . . . 150 2h Mean shear strengths and moisture contents in Clinton County before December 18 (averaged for two samplings). . 151 25 Mean shear strengths and moisture contents in Clinton County for December 31 . . . . . . . . . . . . . . . . . 152 26 Mean shear strengths and moisture contents when various percentages of plot were frozen, Kent County . . . . . . 15h Al Basal area and number of stems by species (based on 50% cruise), Kent County Hardwood plot . . . . . . . . . 17h A2 Basal area and number of stems by species (based on 50% cruise), Clinton County Hardwood plot . . . . . . . . 175 A3 Daily mean air temperatures, deg F . . . . . . . . . . . 176 AA Daily precipitation in inches of water . . . . . . . . . 177 A5 Mean moisture wg£i'three inches of soil . . . . . . . . . 178 A6 Mean bulk density . . . . . . . . . . . . . . . . . . . . 179 A7 Mean daily shear strength readings, frozen depths from surface . . . . . . . . . . . . . . . . . . . . . . 180 Tables 4? LA) CDNCDU'I \O 10 ll l2 13 1h 15 l6 17 18 19 LIST OF TABLES Vegetation of the herbaceous plots Herbaceous species of hardwood plots Tree species on hardwood plots . . . . . . . . . . . Summary of climatological data Soil texture classes of each plot . . . . . . . . . . . . Organic matter contents . Specific gravity . . . . . . . . . . . . . . . . . Atterberg limits for each plot Degree-days for each plot grouped by freezing periods . Minimum.and maximum temperatures in frozen soil; number of frozen cores; and minimum temperature of unfrozen soil, deg F . . . . . . . . . . . . . . . . Mean frost depths, Kent and Clinton Counties Frost occurrence and frequency . . . . . . . . . . . . Per cent of time when 0, 1, 2, 3, and A cores were frozen (based on all samples) . . . . . . . . . . Mean snow depth, density, and water equivalent . . . Minimum and maximum available moisture readings (from Bouyoucos blocks) . . . . . . . . . . . . . . . . Mean moisture contents, standard deviations, and coeffi- cients of variation for the upper 15 in. of soil 60-cm tension values and total pore space before freezing period . Means, standard deviations, and confidence intervals for bulk density before freezing . . . . . . . . . . . . . Statistics for the regressions of bulk density (Y) Hg: frost depth (X). . . . . . . . . . . . . . . Statistics for regression of bulk density (Y) on moisture (X) . . . . . . . . . . . . . . . . iv 76 78 81 82 86 97 114 122 l32 133 137 LIST OF FIGURES Figure Frontispiece: "Neither snow, nor rain, nor heat, nor gloom of night stays these couriers from the swift completion of their appointed rounds." 1 Relative location of study plots in lower Michigan . . 2 Location of bare plot in Kent County . . . . . . . . 3 Location of herbaceous and hardwood plots in Kent County A Location of plots in Clinton County . . . . . . . . . . 5 Surface conditions of bare plots . . . . . . . . . . 6 Surface conditions of herbaceous plots . . . . . . . . 7 Surface conditions of hardwood plots . . . . . . . . . 8 Detailed soil maps of Kent County plots . . . . . . . . 9 Detailed soil maps of Clinton County plots . . . . . . lO Plot diagram and sampling sequence for all plots . . . 11 Diagram of sample square . . . . . . . . . . . . . . . 12 Core cutting board and Hvorslev soil sampler . . . . . 13 Mechanical sampler . . . . . . . . . . . . . . . . . . 1h Mean air temperatures, Kent and Clinton Counties . . . 15. Mean snow and frost depths, Kent County . . . . . . . . 16 Mean snow and frost depths, Clinton County . . . . . . 17 Frost depth versus eight degree-day sum, Kent County plots . . . . . . . . . . . . . . . . . . . . . . . 18 Frost depth versus eight degree-day sum, Clinton County plots . . . . . . . . . . . . . . . . . . . . . . . . . 19 Frozen cores, Kent County Bare (left), Clinton County Hardwood (right) . . . . . . . . . . . . . . . . . . . 20 Lower 1-1/h in. of 3-1/h-in.-deep frozen core, Clinton County Bare plot . . . . . . . . . . . . . . . . . . vi Page 31 32 33 3h 36 37 39 an .50 56 57 63 6S 73 79 8h 91 92 99 . lOO LIST OF FIGURES (Continued) Figure Page 21 Frozen core, Kent County Bare plot . . . . . . . . . . . 101 22 Frozen core, Clinton County Bare plot . . . . . . . . . . 102 23 Frozen core, Clinton County Bare plot . . . . . . . . . . 103 2h 2-1/2-in. frozen core, Kent County Herbaceous plot . . . 10A 25 3-1/2-in. frozen core, Clinton County Herbaceous plot . . 105 26 l-l/8-in. frozen bottom of core, Kent County Herbaceous plot . . . . . . . . . . . . . . . . . . . . . . . . . . 106 27 3-1/h-in. frozen core, Kent County Hardwood plot . . . . 107 28 Frozen upper portion of core, Clinton County Hardwood plot . . . . . . . . . . . . . . . . . . . . . . . . . . 108 29 Looking down on two Clinton County Hardwood cores . . . . 109 30 Mean moisture content, in./15 in. of soil, Kent County . . . . . . . . . . . . . . . . . . . . . . . 115 31 Mean moisture content, in./15 in. of soil, Clinton County . . . . . . . . . . . . . . . . . . . . . 117 32 Mean moisture contents by 3-in. depths, Kent County Bare plot . . . . . . . . . . . . . . . . . . . . . . . 119 33 Mean moisture contents by 3-in. depths, Kent County Herbaceous plot . . . . . . . . . . . . . . . . . . . . . 120 3h Mean moisture contents by 3-in. depths, Kent County Hardwood plot . . . . . . . . . . . . . . . . . . . . . . 121 35 Mean moisture contents by 3-in. depths, Clinton County Bare plot . . . . . . . . . . . . . . . . . . . . . . . . 123 36 Mean moisture contents by 3-in. depths, Clinton County Herbaceous plot . . . . . . . . . . . . . . . . . . . . . 12h 37 Mean moisture contents by 3-in. depths, Clinton County Hardwood plot . . . . . . . . . . . . . . . . . . . . . . 125 38 Mean bulk density in frozen depths (surface to 3-in. depth), Kent County . . . . . . . . . . . . . . . . . . . 128 39 Mean bulk density by 3-in. depths, Clinton County Bare plot 0 O O O O O O O O O I O O O O O O I O O O O O 0 O O 129 vii Figure LtO A1 #2 1+3 an 1&5 #6 1+7 #8 1+9 Bl B2 LIST OF FIGURES (Continued) Mean bulk density in frozen depths, Clinton County Herbaceous and Hardwood plots . . . . . . . . . . . Regressions of bulk density versus frost depth . . . Regressions of bulk density on moisture during freezing . Mean shear strengths, Kent County Bare plot . . . . . . Mean shear strengths, Kent County Herbaceous plot . . . Mean shear strengths, Kent County Hardwood plot . . . . Mean shear strengths, Clinton County Bare plot . . . . Mean shear strengths, Clinton County Herbaceous plot Mean shear strengths, Clinton County Hardwood plot Shear strength pattern for each depth during final thaw, Clinton County Bare plot . . . . . . . . . . Cone penetrometer in use Detailed drawing of cone penetrometer . . . . . . . . viii Page 130 13h 136 lhu 1h5 1A6 1117 1&8 1H9 158 183 184 INTRODUCTION Ground-water reCharge during winter months, heaving of soil, and runoff from melting snow are of vital interest to man. In addition, during many a spring thaw, farmers and loggers suffer serious delay in their operations, whether it be land preparation, planting, or hauling logs. I These processes and delays are directly associated with the condi- tion of the soil with respect to freezing or thawing and with concomit- tant climatic conditions. The amount and kind of precipitation that has fallen, the contribution of air temperature to rapid snowmelt, the soil conditions, including infiltration and storage capabilities of the soil, are all involved. Finally, interactions of climatic and soil variables are greatly influenced by the vegetation present. Consideration of the status of the ground, whether frozen or thawed, has been limited primarily to descriptions of the occurrence, depth, and type of frost (e.g. Scholz, 1938, Pierce, Lull, and Storey, 1958). Much information has also been assembled on the effects of vegetation, snow, and climate on frost depth and occurrence (e.g. Pearson, 1920, Potter, 1956, Sartz, 1957). The effects of excessive meltwaters flowing over frozen ground have been examined (Storey, 1955, Trimble, 1959, and others), and foresters in particular have noted the benefits of vegetative cover in reducing excessive runoff and erosion (e.g. Scholz, 1938, Kienholz, 19A0, Hale, 1950). Studies have been made on the physical phenomena associated with soil freezing (Post and Dreibelbis, l9hl, and others); and, especially 2 from the engineering standpoint, on frost heaving in disturbed soils (e.g. Johnson, 1952). However, detailed experiences with soil moisture and soil physical properties, particularly knowledge of chronologic changes in these properties into, through, and beyond the winter sea- sons, are conspicuously lacking. Information is still needed concern- ing the behavior of soil moisture and physical properties during the winter freezing and thawing periods as related to ingress of water into the soil, runoff during thaw and water available to plants. Soil strength and compaction characteristics in the spring also need elucidation. When.more is known about dynamic processes which take place during soil freezing and thawing and about the actual effects of vegetation on these processes, a better understanding of the end effects, ground- water recharge, runoff, etc., will be possible, and the magnitude and duration of these effects may be predicted. This study was designed primarily to provide information on the soil moisture, bulk density, and shear strength regimes under frozen and nonfrozen conditions. The relation of these regimes to weather, vegetation, and other soil characteristics was to be evaluated. The investigation was conducted in southern Michigan on medium textured soils and included three cover conditions. LITERATURE Many investigators, including Beskow (19u7), Siple (1952), Black, Croney, and Jacobs (1958), and Domby and Kohnke (1953), have reported that the soil moisture content of the upper layers of soil can increase well above normal moisture contents as a result of freezing and thawing action. Conversely, little information has been published concerning soil physical changes. Theoretical relationships have been worked out for the maximum amount of water a soil can hold, and for movement of water into frozen or freezing layers, but changes in soil pore space have rarely been measured. While much information exists on construction of roads, runways, and other structures on disturbed soils, no studies were found which considered trafficability during thawing (or freezing), periods on natural soils. 1 Johnson (1952), summarized the state of knowledge of the soil during thaw as follows: ”For some reason, the literature contains little in- formation on the physical process of soil melting and the subsequent changes in moisture distribution which take place as the soil adjusts itself to an unfrozen environment," and Serova (1959), presumably speaking of Russian and European litera— ture stated, "The object of most investigations has been an elucida- tion on the depth of a soil's freezing through (congelation) with respect to a region geographically, and to meteorological conditions (of that region)." Specific references will be considered under the following headings: 1. Temperature Conditions Necessary for Soil Freezing. 3 h 2. Soil Moisture During Freezing Periods. 3. Influence of Precipitation and Vegetation on Soil Freezing and Thawing. A. Influence of Soil Color, Evaporation, and Microrelief During Freezing and Thawing on Soil Moisture. Temperature Conditions Required for Soil Freezing Air Temperature The magnitude and duration of air temperature below freezing is very important in influencing depth of freezing in the soil. Heat loss must be toward the air above the soil; thus, before soil freezing can begin, the air temperature must be below freezing. Russell (19h3), in reporting on the analysis of 10 million temper- ature readings between 191A and 1931, concluded that an air temperature of 28 F is enough to cause soil surfaces to start freezing, while a rise to 32 F would start thawing. Anderson (l9h7), working at North Fork, California, on gravelly sandy clay loam soils, found that a minimum air temperature of 31.1 F was necessary before soil freezing began on a bare plot, 29.0 F was re- quired on a grass-covered plot, and lh.1 F on a brush-covered plot. Franklin (1919) found that a temperature of -9.2 C at the soil surface would allow frost to penetrate h in. in 12 hr. At a surface temperature of .1.1 C, frost took four days to penetrate the same depth. This illustrates the importance of both magnitude and duration of temperature on frost formation. Domby and Kohnke (1955) showed some effects of diurnal air tempera- ture fluctuations on a Russell silt loam in Indiana. They found that 5 when temperatures dropped from above freezing during the day to a range of 20 to 25 F at night, soil would freeze to a depth of about 1 in. Hale (1950) found that l to 2 in. of frost could form under ponderosa pine stands with overnight freezing, but that it would disappear in the afternoon. Schneider (1957) noted an important relationship between air tem- perature, freezing, and ground water. He found that water tables in Minnesota, which had dropped during freezing weather, rose after a few days of above 32 F temperature. The water table rise occurred too soon to be accounted for by snowmelt. He concluded, "When air temperature rises above 32oF, as it would in the spring, heat continues to move upward from the zone of saturation (below the frost); however, instead of moving through the frozen layer (as it would when air temperatures were below freezing) it starts thawing the bottom of the layer because an opposing thermal gradient (from the atmos- phere) now causes heat to move downward from the atmosphere to the frozen soil.” Thus, water moved downward to the water table as a result of thawing of the frozen soil from below, and not from snowmelt. Another important relationship between air temperature and frozen soil involves the availability of water to plants. Wilner (1955), in studying physiological winter drought in the Canadian prairies, found that for two years November to April air temperatures ranged from -2 to -33 F, and during this period no water was available for plant use in the upper 18 in. of soil, this depth being continually frozen. Post and Dreibelbis (19A2) studied freezing of silt loam soils under pasture, woodland, red clover, alfalfa, and winter wheat. They found that no freezing occurred 1 in. below the soil surface. Air temperatures have been used to predict the depth of frost 6 penetration by Casagrande (1931), Shannon (l9h5), Anderson (19h7), Wilkins and Dujay (195A), Crawford (1952), and Aldrich and Paynter (1953). Siple (1952) developed semiemperical systems for predicting maximum expected frost depth anywhere in the United States based on the number of hours of temperatures below 30 F during an average year. In.most cases, however, air temperature alone does not suffice for frost depth prediction. Soil thermal properties, snow cover, and vegetation each influence depth of freezing. For any one area the relationship may change from year to year and relations developed for one area may not be applicable elsewhere. Soil Temperature Post and Dreibelbis (19h2) noted considerable variation in the soil temperature at which freezing would occur. For all plots, the temperature at which freezing took place was 18 to 26 F at 1/2 in. below the soil surface; 23 to 27 F at 3 in.; 22 to 25 F at 6 in.; and 22 to 26 F at 9 in. They state that, in general in the freezing layers, the deeper the freezing, the higher the temperatures in the frozen layer. To extend the depth of freezing or to begin freezing, not only must soil temperatures be below 32 F but heat must be expended (Chang, 1957). Freezing of water requires the release of 79.63 g/cal per gram of ice. Bouyoucos (1921) found that temperatures required to freeze soil water vary. Free water may freeze at -1.5 C (29.3 F), capillary water at -A C (2A.8 F), and hygroscopic water may not freeze until a tempera— ture of -78 C (-108.A F) is reached. These findings agree with those of Wintermeyer (1925). Presumably the more free water in a soil, the faster the soil will 7 freeze at a given temperature. Belotelkin (19h1) noted that the in- tensity and duration of temperatures required to induce soil freezing increased as the amount of capillary and hygroscopic water decreased. Freezing should be more easily induced in coarse-grained soils than in fine-grained soils at moisture contents at or near saturation as there would be more free water in the coarse—grained soil. Beskow (l9h7) verified this; he found that lower temperatures were required to freeze fine-grained than coarse—grained soils. As Crawford (1952) stated, . ”Moisture content is by far the most important intrinsic factor affecting soil temperatures, and any transfer of water in the soil will not only carry heat, but will alter thermal properties by its movement." Thompson (193A) earlier clarified these relationships when he found heat conductivity of soil to be increased by addition of water produc- ing better thermal attraction between soil grains. As the specific heat of the soil-water mass increases with increasing water content, more energy must be released to produce freezing. Thus, more water may permit faster heat conduction, but at the same time, greater heat loss is necessary to induce freezing. The effect of soil moisture on variation in soil temperatures necessary to induce freezing has been observed by many investigators, including Johnson (1952), Pearson (1920), and Atkinson and Bay (l9h0). Diurnal change (Dcmby and Kohnke, 1955) and soil color (Bouyoucos, 1916) also cause variation in temperature in respect to freezing. A phenomenon observed by Potter (1956) deserves mention. Compar- ing spring and fall temperatures he found an overturn similar to the thermal overturn in bodies of water. He said, 8 “In sail there is no actual mixing as in water, but there is similarity in having the lower levels warmer in the winter, a short period of rather uniform temperature throughout, and then warmer upper levels in the summer. On examination of the temperature curves for the five sites, and the two-year period, the writer was greatly surprised to find a distinct brief period in the spring and fall when the temperature lines con- verged, coincident with the reversal of the order of tempera- ture gradient." Potter observed this phenomenon in the 6-in. to 6-ft depth. He found for five sites the two-year average fall overturn date to be October 3, and the spring date to be May 5. Temperature range between all depths was 2.h and 2.1 F for the two dates, respectively. The relationship of soil density to soil temperature and frozen soil has been neglected. Crawford (1952) in his review noted that, ”Practically all observers of soil temperature fail to record any effect of density." Soil Moisture During Freezing Periods Frost Types and Distribution Post and Dreibelbis (19h2) described three types of frost: (l) concrete-~having very dense structure in which very thin ice lenses are formed, along with fine ice crystals; (2) honeycomb--which has a loose porous structure allowing free water vapor movement; (3) stalactite-- which consists of small, vertical icicles that join heaved surface particles to the main body of soil below. Hale (1950) recognized another frost type which he named "granularfl' Granular frost has scattered granules of ice binding the litter and F layers of humus. That the physical structure of frost is important insofar as move- ment of water into and within the soil is concerned has been brought 9 out by Sartz (1958), Bay (1958), Trimble, Sartz, and Pierce (1958), and others. All agree that concrete frost is most important in that it may reduce infiltration to zero. Other types of frost, which are gen- erally'more porous, may allow water movement. Trimble, Sartz, and Pierce (1958) observed that, in the north- eastern United States, granular frost under hardwood and white pine stands may actually increase infiltration capacity over that of unfrozen soil. Occurrence of frost types was discussed by Storey (1955). "Concrete frost has been observed most frequently in cultivated fields. Honeycomb and stalactite frost, which usually occur during shallow freezing, are found most fre- quently in meadows and pastures. In forested areas frost of the granular type is found oftenest. Honeycomb frost is found next in frequency." He said further, "A concrete type of frost structure is formed practi- cally in all soils which have been largely depleted of humus. "In the presence of humus, frost in the soil is usually of a porous structure. "A concrete type of frost structure is formed in heavily compacted soils, irrespective of the humus content. "A concrete type of frost structure frequently forms when frost penetrates below the humus layer. In lightly com- pacted pastures and meadows this occurs usually at depths below 3 or 4 inches." Concerning concrete frost, Trimble, Sartz, and Pierce (1958) stated, "While remaining unchanged in appearance, at least to the naked eye, it gradually loses its hard rocklike con- sistency and.may easily be broken by hand. In the case of pasture soils with a great number of grass roots, it be- comes pliable and can be bent. Concrete frost eventually becomes quite pliable before it melts completely; At some point in this degenerative process, frozen ground becomes 10 permeable to water. Of course, once it loses its concrete- like hardness, the definition, concrete frost is no longer applicable. "During times of thaw the melt pattern in some areas of concrete frost was very erratic, resulting in a close inter- mingling of frozen and unfrozen ground. This condition ap- peared to be associated with difference in micro-relief, which permitted certain micro-aspects and slopes to receive more solar radiation and thus to melt sooner." Moisture Movement and Physical State Bouyoucos (1921) was earlier quoted as finding that free, capil- lary, and hygroscopic water freeze at different degrees of temperature. Beskow (19h7) found water in narrow pores to freeze below 0 C (32 F). Moreover, Anderson, Fletcher, and Edlefson (19h2) found with below freezing temperatures that water farthest from the surface of the soil particles freezes first. This finding agrees with that of Grim (1952) who stated, "Directly adjacent to the adsorbing soil solidly ad- sorbed water is to be found, the center of a pore space is occupied by ordinary water, freezing at about 00C, and be- tween the ordinary water and the solidly adsorbed water there is a zone of liquid water possessing a melting point down to 2200 which serves as a passageway for the conduction of water to freezing centers." Much frozen soil research has been concerned with whether or not massive (concrete) soil freezing has occurred, particularly in relation to frost heaving, Beskow (19u7), Casagrande (1931), Taber (1932), Black, Croney, and Jacobs (1958). Beskow (19h7) reported that massive freezing always occurs when soil moisture content is below capillary saturation, in contrast, when moisture is greater than capillary saturation, quick freezing will yield massive ice, and slow freezing will develop stratified ice. Ice layers do not form when soil particle sizes are greater than 0.06-0.1 mm in 11 diameter; instead, homogeneous ice formation takes place. In fine-grained soils Beskow (19A7) noted that needlelike crystals were formed.perpendicular to the surface (stalactite). If fine layers of silt or clay occur in a coarse soil, thick ice layers may form and frost heave take place, provided there is saturated soil below the frost layers. When soil is saturated below the frost layer, ice layers can form at discontinuities of soil texture, i.e. sand over silt, etc. Beskow (19A7) found increases in water up to 120% of original volume of the unfrozen soil under roads in Sweden. He also quotes Runeberg (1765) who found that so much moisture had moved into a clay layer during freezing that the ratio of water to soil by volume was A to 1. Data of Post and Dreibelbis (19A2) indicate that moisture content may increase from two to nine times that of the prefreezing content. Their data also revealed that moisture content in the frozen portion of the surface part of soil was generally a little over twice that of the nonfrozen part. Many investigators have observed that moisture migrates to frozen layers. Anderson (19A6) reported that moisture was drawn from 36 in. below the surface into the surface to 3—in. layer. He also noted that Lochhead (192A) found the moisture content of the 10-in. depth in a sandy soil dropped from l2.A to 6.A% (by weight) as freezing progressed down to 8 in. Bouyoucos and McCool (1916) stated that water moves from wetter layers below up to the frost line. However, Domby and Kohnke (1955) found that soil just below a frozen layer became wetter when the frozen layer thawed. Moisture movement to the frozen layer can occur by capillary and 12 by vapor movement, Bouyoucos (1915), Siple (1952), Beskow (19A7), and Penner (1958). Hadley and Eisenstadt (1953) reported on a laboratory study of vapor movement during freezing as follows: "A simulated soil made of glass beads 0.01 in. in diameter and a radioactive tracer technique were used to determine whether the moisture transfer was in liquid or vapor form. A critical moisture content of about A% moisture by weight of the total dry weight separates liquid from vapor movement in the soil. The water moves from the hot to the cold point in the form of liquid water in wet soils. The increase in water con- tent coincides closely with the ice point and no moisture move- ment is apparent when the temperature is kept above 32oF. The moisture moves in the form of vapor in soils containing less than A% water and is not associated with freezing." Explanation of capillary movement is summed up by Penner (1958) c; as follows: "Ice lensing occurs only in soils with small pores, so when water freezes it produces an effect similar to that of drying at that point. The liquid water moves from wetter to drier, i.e. frozen, areas. Freezing water re- leases heat which permits the soil to retain its temperature even though more heat is being lost. This permits lenses to grow until the water supply is exhausted, or rather until the force holding water to particle surfaces exceeds the forces involved in drying and formation of ice." The process of increasing moisture-holding capacity was described in general terms by Siple (1952), as follows: "The freeze-thaw action of free water accumulated in soil has a 'jacking' or ratchet action which deforms the soil about it, and leaves voids during the thaw cycle which are replaced by free water." He notes that water can move into the frozen layer by capillary action during freezing. Studies have been conducted on the relation between soil freezing and aggregation. Jung (19A2), in his experiments of induced freezing of soils with liquid air, found that the aggregating effect of frost decreases th‘ 7 C free21n~ (I i nlv slow fr ment, the 11 a .n R 5 ,. rid the more a vy- .4 n- f‘ u 0 fl‘“ «V*-s..'u \" blag ’5‘ nth? . .4 + w r- - Stalk vya C CI“, A 7 Yrs kr" O“ A. s “‘ M e_\_ ‘ 13 decreases the faster a soil is frozen. With slow freezing, a few large crystals are formed resulting in large aggregates. A repetition of freezing causes a negligible decrease in the degree of dispersion, and the more free water in a soil the more ice crystals are formed. With slow freezing, Jung observed that soil particle dispersion de— creased with increasing water content to full water content, then in- creased. Fast freezing caused dispersion to increase up to full water content, then to decrease. Baver (19A8) explains part of Jung's observations as follows: Slow cooling causes ice crystals to form in the tension-free pores. The crystals in turn draw water from surrounding particles (which would increase the size of the voids), resulting in dehydration of the parti- cles. Dehydration in turn allows more intimate contact of the soil grains (aggregation). The large crystals melt, and form nuclei for more crystals. Quick freezing, on the other hand, causes many small crystals to form, resulting in breakup of the aggregates. Decreases of up to 30% were observed in bulk density supporting the statement of Bouyoucos and McCool (1916) that freezing processes ”...in a soil saturated with water account for not more than 5% bulk density changes." The Frost Effects Laboratory (1951) report stated, "If the total volume of unsaturated soil contained 1/3 water, freezing would yield expansion of only 3%. But, ex- pansion of 60% has been observed.” Moisture must be moving into the frozen zone to account for large den- sity decreases. Domby and Kohnke (1955) reported changes in bulk density and moisture during freezing and thawing in the surface inch of a bare Russell silt loam. Night temperatures were 20 to 25 F, and day 1A temperatures above freezing. An inverse relationship between bulk den- sity and moisture content was noted. For instance a 30% decrease in bulk density was accompanied by over 100% increase in soil moisture. Domby and Kohnke noted that: "The most rapid decrease in water stability (of aggre- gates) occurred where soil was bare, but by the end of two winters most of the water-stable aggregates in the mulched soil also were broken down. Soil which was loose in the fall became compact during the winter, even where a surface mulch of two tons of straw per acre was present.” Slater and Hopp (1951) studied pore size and aggregation in silt loam soils planted to (1) year-round corn, (2) spring-manured corn, (3) corn-wheat-unmowed timothy-red clover, and (A) sod. Their results were similar to those of Domby and Kohnke (1955). They also reported that after 10 years under sod, structure and large pores remained un- changed. But, as to the cropland, they reported that those soils under continuous crop had greater aggregate stability and more large pore space than fields under continuous corn. During thawing periods decreases in bulk density were observed in . . by the upper 6 in. of 5011, Krumbac . In the Lake States 50% decreases in density were found on well-drained, silt loam soils during thaw with associated moisture contents as high as 63% by weight. These occurred on well-drained soils. There were indications that similar although smaller changes also occurred in the 6- to l2-in. layer. Post and Dreibelbis (19A2) reported, "During the freezing process, the water holding forces l/ Krumbach, A. W. 1959. Report on Freeze-Thaw Survey in the Lake States. (Unpublished data.) 15 of the soil were overcome and ice became the soil carrier rather than the soil being the carrier of ice. The average volume weight of the frozen surface soil of Keene and Muskingum Silt Loams were found to be 0.63 and 0.93 respec- tively, while the volume of these same soils when nonfrozen were 1.27 and 1.36 respectively." This is the equivalent of decreases in density of 50.A%, and 31.6% for the two soils. Influence of Precipitation and Vegetation on Soil Freezing and Thawing Snow Snow can prevent frozen soil freezing layers from forming or can modify frost depth by insulating against freezing air temperatures. Likewise, snow may extend the duration of frost in the ground by in- sulating against warm air. The influence of depth of snow on the amount of solar energy reach- ing the soil varies. Church (19Al) stated, "The use of white and black bulb thermometers in snow indicates that radiation from the sun is effective on dark Objects to a depth of 18 inches. Is this heat too slight to melt the ground until the snow cover has completely disappeared?" Kunmin (1957) found that a layer of dry snow 10 cm (A in.) deep absorbs 65% of the solar radiation, and that wet snow of the same depth absorbs 97% of the solar radiation. He indicated that 8 in. (20 cm) of wet snow, or 16 to 20 in. (A0 to 50 cm) of dry snow would effectively protect the soil against thawing from above. Mail (1936) found that 8 to 15 in. of snow maintained frost depth at 3 ft for 23 days when the average air temperature was A0 F. Storey (1955) noted that: 16 "A number of observations have shown that even though frost penetration has started before the first snow, when snow depths have reached 18 to 2A inches further frost penetration is stopped." Eighteen inches of snow in northern Sweden and 8 to 12 in. in southern Sweden prevented soil freezing (Beskow, 19A7). Atkinson and.Bay (19A0) found, on wooded and bare sites in Wisconsin, that with 10 in. of snow, frost depth gradually decreased on 15 of 23 plots. In general, when snow was less than 10 in. deep, frost depth increased. In bare fields with 12 to 2A in. of snow, frost actually disappeared from the ground. Potter (1956) studied frost depths in open corn land as opposed to a corn plot in the lee of a shelterbelt (in eastern North Dakota). In the winters of 1952-53 and l953—5A the open plot had A.5 and 9 in. of snow, respectively, and the sheltered plot, 36 and 120 in. The first winter Potter found that the limited snow cover on the Open plot al- lowed frost to increase, from 12-in. thickness at the first snow, down to as deep as 3 ft. On the plot in the lee of the shelterbelt, the deep snow resulted in the frost line staying at about 1 ft, where it had been at the time of first snowfall. With 10 ft of snow the second winter, frost disappeared from the ground on the plot in the lee of the shelterbelt, and temperatures at the l-in. level rose to several degrees above freezing. The open field with a 9-in. snow cover again froze as deep as 3 ft. Potter attributed the lack of increased frost penetration to the fact that resident heat from the ground can thaw the frost layer from below a heavy snow mantle. Other writers attribute the lack of frost penetration, or the actual melting of frost layers under heavy snow 17 cover to this same phenomenon; e.g. Belotelkin (19Al), Holmes and Robertson (1960). Snow on the ground prior to freezing may delay or prevent the formation of frost. Diebold (1938) working on predominantly sandy loam soils found that with air temperatures continually below freezing: l. Freezing occurred only on an area bare of snow or vegetation (as compared to hardwood stands). 2. With 17 to 31 in. of snow, forest sites had no frozen ground. 3. Fifty-eight inches of snow prevented freezing on another area without vegetative cover. 0n west slopes in the Cascade Mountains, Hale (1950) found that snow cover prevented freezing, while on east slopes with patchy snow cover freezing occurred (in Douglas fir and Lodgepole pine stands). Tigerman and Rosa (19A9) found in mountain soils of Utah that 18 in. of snow was sufficient to prevent soil freezing. The effect of snow cover on temperature fluctuations is illustrated by the findings of Atkinson and Bay (19u0). "Under 5 inches of snow cover there was less change in soil temperature at -2OOF than there was under the 2-inch snow cover at -89F, indicating the insulating effect of snow." They also found under 6 in. of snow fluctuations amounting to 2 F at the soil surface, while under 12 to 2A in. of snow there were no fluctuations. A snow mantle may either speed up the rate of soil thawing by the addition of meltwater to the frozen ground, or it may delay thawing by insulating the frozen ground from high air temperatures. Belotelkin (19Al), studying spruce flat, spruce swamp, fir flat, northern hardwoods, and open areas in New Hampshire, found thawing from 18 the soil surface did not begin until all snow had disappeared. How- ever, thawing of frost occurred from below. In Ponderosa pine plots, north slopes with snow tended to thaw three weeks later than south slopes with no snow, Hale (1950). It is well known that snow meltwater contributes to the moisture content of the soil; however, its contribution can vary. Diebold (1938) found the rate of disappearance of snow was equal in forest (chestnut-oak, aspen, beech—birch—maple) and Open areas. Snow lasted longer in the forest since there was a greater depth at the start of the thaw. Storey (1955) says that l in. of concrete frost may prevent snow meltwater from entering the ground, and Mosolov (1926) found that in- H creases in soil moisture from snowmelt ..varied widely depending upon relief, depth of soil freezing and soils structure." No quantitative information was found concerning the addition of meltwater to the soil. Trimble (1959) states that snowpacks must reach a certain density before melt ("ripening") begins. In the eastern United States this den— sity is near 30%, in the West it is to to 50%. He points out that these are averages, and snOWpack density varies from tOp to bottom. In summarizing Russian work Kuzmin says, ”Fresh snow has the greatest water holding capacity; it can hold 55—35% of water of the total weight of wet snow at its initial density 0,13 - 0,21 (0.13 g/cc - 0.21 g/cc). Coarse-grained snow has the least water holding capacity; at its initial density 0,39 - 0,h5 (0.39 g/cc - o.h5 g/cc) holds 25-15% of waterfl' Rain Rain may hasten the thawing process. Atkinson and Bay (19A0) 19 found this to be true in bare plots. In 1937 at the beginning of the thawing period there was 18 in. of frost in the ground, which required 27 days to thaw without rain. In 1938, with 1.8 in. of rain 3A.5 in. of frost disappeared in 16 days. Though air temperatures during thaw had some influence, the effect of rain in removing frost from these bare plots was clearly evident. Bay (1958) found that spring rains greatly hastened frost removal. Chang (19h7) said, "Thawing is a rapid process in soils which permit the free passage of rain water. Partly because of the easy percola- tion of spring rain, and partly because of their low content of frozen water, the rate of thawing is greater in light than in heavy soils." Goodell (1939) reported that rain may occasionally result in the formation of an ice coat on the soil surface and in the litter. Bay, Wunneke, and Hays (1952) found that rain during thaw resulted in greater runoff and soil losses than did thawing alone. Information on the amount of rain entering frozen soil is scanty; USDA workers report that, "...approximately 80% of an one-half inch rainfall was retained in a Pulaski County (Virginia) watershed...the ground was frozen to a depth of at least 0.25 of a foot before and dur- ing the storm.“g/ Vegetation Vegetation is important in part by its effect on snow cover and, thus, on freezing and thawing. Depth, duration, and type of frost are all affected. MacKinney (1929) compared freezing on a plot with litter and 2/ Unpublished data, U. S. Department of Agriculture. 20 a plot from which litter was removed, both in a mixed red and white pine plantation in Connecticut. He found that litter retarded the first date of frost penetration by one month; once freezing began, there was no difference in freezing rates. Litter decreased frost penetration over A0%; the litter-covered plot showed 5 to 8 in. of frost as 0p- posed to 8 in. in the bare plot. He stated: "The character of the frozen soil was influenced markedly by the litter. The soil on the bare plot froze solidly, and the air spaces were practically filled with ice. 0n the other hand, the frozen soil beneath the litter cover was porous and loose, at no time being frozen too hard to allow the insertion of a shovel. In the litter covered soil the ice formed around the soil particles leaving the spaces between the soil parti— cles open." Presumably MacKinney had observed concrete and honeycomb frost. He found also that rain could penetrate the frost under the litter, but not under the bare plot. Studying differences between an ungraced woodlot and a close- cropped bluegrass pasture, Scholz (1938) reported that frost was deeper in the pasture throughout the freezing season; a maximum depth of 10 in. in the pasture, and A in. in the woodlot. Snow depths were the same over both plots at about 10 to 11 in. Scholz reported, "The manner in which the frost left the ground is also of interest. In the open pasture, the direct rays of the sun and above-freezing air temperatures progressively thawed out the soil, beginning at the surface, and working down into the subsoil until the frost had completely disappeared. In the woodlot, however, at those points where frost still occurred, thawing evidently took place from the bottom up, for in no case did the ice crystals disappear in the sur- face soil prior to the thawing of lower layers. Yet, all frost had disappeared in the woods two days before the pasture soil was completely thawed." 21 Similar observations were made by Krumbaché/ in Wisconsin. Compar- ing a pasture, a woodlot, and a plowed field, it was found that the plowed area started thawing before the woodlot, but finished later. The woodlot thawed in patches. The pasture began and ended thawing after both the woodlot and plowed field. Goodell (1939) studied sites with oak and oak-hickory on silt loam over clay soils, grassed areas on clay, and an open area on silt loam. He found that no soil freezing occurred in the woodlots; a maximum of 1.5 in. occurred under the pasture; while in contrast the Open (corn) plot had maximum of 5.5 in. of freezing. There was essentially no snow cover. In comparing corn stubble, wheat mulch, small grain stubble, and grass sod sites, Potter (1956) reported that the grass sod had higher minimum temperatures. Variation of soil temperature during the freez- ing season was less under grass than under the other types of cover. Kienholz (19A0) studied six forested areas, an open area, a light sod area, and a heavy sod area. He found that soil freezing in the forested areas did not begin until 15 to 35 days after the plowed area started to freeze. There were 125 days when frost occurred in the open plot compared to 9A days in white pine. The duration of freezing on the heavy sod site approximated that of the forest; frost duration in light sod was shmilar to the open site. Maximum frost depths in the winter of 1938-39 were 8.5 in. in the Open compared to 2.9 in. under white pine. The average depth of freezing was also deeper in the Open site. 3/ Krumbach, A. W. 1959. Report on Freeze-Thaw Survey in the Lake States. (Unpub- lished data.) 22 Pine and hemlock sites intercepted light snowfalls, had less drifting of snow, less surface crust formation, longer duration of snow on the ground, and less and lighter snow than the open or the sod areas. Thus snow disappeared from the pine and hemlock sites before it left the open sites. Although snow depth varied little from site to site, differences in freezing did occur. White pine was more effective in lessening frost penetration than oak on a ridge site; the latter more effective than red maple, which, in turn, was more effective than mixed hardwood. Winding up the list in decreasing order of effectiveness of preventing frost penetration were the heavy sod, light sod, and bare cover types. Kienholz (19A0) found that leaves in depressions were very effec— tive in reducing frost in the ground, and suggests perhaps organic mat- ter thickness, not type, determines over-all effectiveness. The thicker the organic matter, the less frost in the ground. When snow differences are considered, relationships are somewhat similar. Comparing northern hardwoods, chestnut oak, aspen, and bare sites, Diebold (1938) reported that the snow depth was equal in the forested areas, while bare areas on the flat had none. The northern hardwood plots had 11 in. of snow on March 18, 1936, while seven of eight open areas were bare. Diebold noted that only bare plots showed runoff from a 7.9-in. rainfall, indicating that frost caused the bare plots to be impermeable, while the forest plots were able to absorb water._ Belotelkin (19A1) in studying spruce, fir, northern hardwoods, and Open areas, found that forest cover and its interrelationships with snow cover had a profound effect on time of soil freezing, and rapidity 23 of thawing. Penetration of frost was least in the hardwood stand, and greatest in the spruce swamp. Thawing began and finished earlier in the hardwood stand. He attributed this to the insulation provided by more snow and thicker litter in the hardwood stand. He found, also, that even though coniferous areas had frost in the ground later than hardwood areas, thawing patches in these plots before all freezing had gone soon permitted infiltration. In a five-year study on a sandy clay loam near Northfork, Cali- fornia, Anderson (19A6) found no freezing under brush sites, and less under grass than under bare areas. 0n the average, soil freezing oc- curred 17 days earlier and ceased A5 days later on bare than grass sites. Sartz (1957) studied frost for two years in several timber types in southern Maine, northern New York, south central New York, north- western Massachusetts, and northeastern Pennsylvania. He found that concrete frost began to form 3 to A7 days after the ground began to freeze in open land, and 26 to A8 days after freeze in hardwoods. Two—year averages for last frost ranged O to 26 days later in soft— woods, and from 22 days earlier to A days later in hardwoods than in open land. Pierce, Lull, and Storey (1958) noted that frost was found more frequently and to greater depth in open lands as opposed to hardwood or conifer. They found average frost depth in hardwood stands to be about one-half that in conifer stands. Open areas were found to have twice as much snow on the ground as forested areas; snow depth averaged about 2.7 in. more in hardwood than conifer stands; hardwood forests and reproduction areas averaged 1.75 times deeper snow than open areas; and 2A snow depths were 1.5 times greater in conifer forest than in hardwood. The relationship between forest types and depth to permafrost (and soil moisture content) was noted in Alaska by Lutz and Caporaso (1958). They found that the minimum.depth to permafrost under black spruce stands was 12 to 20 in.; under white spruce 2A to 26 in. (if frost is present); under paper birch 36 to A8 in.; under quaking aspen at least A8 in.; and under balsam poplar at least 6 ft (if permafrost was present at all). Thus in an area where conditions suitable for frost formation are present throughout the year, forest types could be used as indicators of depth to frost. Influence of Soil Color, Evaporation, and Microrelief on Soil Moisture During Freezing and Thawing Soil Color Dark colored surfaces absorb more heat than light colored surfaces. Bouyoucos (1916, 1913) reported on studies of five sandy soils, and one peat soil. After four years of investigation, he stated, "This investigation goes to prove that soils with white colors and low moisture content, or with black color and high water content have lower average temperatures during the spring and summer than soils possessing these properties in medium proportion. In other words, the white color of a soil reflects so much of the sun‘s rays that it prevents the soil from at- taining a high temperature, in spite of its low water content and small amount of evaporation, while the excess water of a black soil, such as peat, consumes so much of the heat in its evaporation process, that it keeps the temperature of the soil low in spite of its black color and hence its great heat absorbing power.” Everson and Weaver (19A9) worked with carbon black in Merrimac fine sandy loam during the spring and summer of 19AA and 19A5. Carbon black was mixed to a depth of 2 in. at the rate of A,000 lb per acre. 25 They found that maximum temperatures at the surface and 2 in. below the surface averaged 2 and 3.A F higher, respectively, than for un- treated soils. Minimum daily temperatures were 0.8 and 0.5 F higher on the treated than untreated plots for the soil surface and 2-in. depth, respectively. These authors continued studies into 19A5 and 19A6 on Agawan fine sandy loam soils. In 19A5 carbon treated soil thawed two days earlier than untreated, in 19A6 one week earlier; and it could be plowed two weeks before the untreated soil. On January 27, 19A5, they found the carbon treated soil thawed to 2 in. under snow, while the untreated did not. In the winter of 19AA treated soil remained unfrozen until Decem- ber 22, while the untreated soil froze December 3. Darker soils may result in higher energy intake, with increased soil temperature, delay in soil freezing, and advancement of thawing. Evaporation No quantitative information was found on evaporation in relation to freezing and thawing. Bouyoucos (1916) inferred that evaporation may keep a soil cool. Anderson (19A6) worked out an equation for predicting evaporation when the soil was freezing. In reference to the upper 3 in. of soil he stated, "Freezing of a soil kept bare of vegetation greatly in- creased the evaporation loss from the soil. The evaporation during freezing periods was three times as great as during similar nonfreezing periods, nearly four times as great as combined transpiration-evaporation from brush-covered soil during same periods and about 12 times as great as evapora- tion from a free-water surface during the same periods." Daily measurements were made, and no internal drainage was in- dicated. Freezing and thawing would bring water to the surface in 26 amounts greater than by movement during nonfrozen periods which would account for the greater differences than on a bare soil. The brush- covered soil had not frozen during these periods, and probably the free-water surface had frozen over the night, greatly reducing evaporation. Microrelief Trimble, Sartz, and Pierce (1950) reporting on a study in New Hampshire stated: "During times of thaw the melt pattern in some areas of concrete frost was very erratic, resulting in a close inter- mingling of frozen and unfrozen ground. This condition ap- peared to be associated with differences in micro-relief, which permitted certain micro-aspects and slopes to receive more solar radiation and thus to melt sooner.” This was the sole reference to microrelief. Perhaps mocrorelief effects may well be the explanation for spotty freezing or thawing patches observed in many investigations. Conclusions from Literature Temperature Air temperatures must be below freezing to induce soil freezing; a temperature of 32 F is not low enough as sufficient heat loss does not take place. Overnight air temperatures in the temperature range of about 20 to 25 F will cause freezing in the upper l to 2 in. of soil. Soil temperature must also be well below 32 F before the soil freezes. In one instance, a temperature of 18 F was recorded before the upper 1/2 in. of soil froze. Investigators agree that the lower the soil moisture content, for a given soil, the lower the soil temper- atures necessary to induce freezing. Further, at or near saturation, 27 the finer the soil texture, the lower the soil temperature necessary for freezing. The tension at which moisture is held in the soil is also related to its ease of freezing. The more tension on the water, the lower the temperatures necessary to freeze it. Moisture Movement and Physical State In general, slower freezing results in greater soil aggregation, and vice versa. Considering changes in soil volume on a strictly theoretical basis even a saturated soil should not increase more than 5% upon freezing. However, increases in volume up to 60% over the un- frozen state have been found. This indicates moisture movement into the frozen or freezing soil layer. Soil Moisture During Freezing Periods Studies of soil moisture under the frozen layer indicate that moisture may be drawn from as deep as 36 in. up into the frozen layer. water entering the frozen soil from.above is affected by the frost type; concrete frost presumably prevents snow meltwater, or rain water from entering the soil. However, stalactite or honeycomb frost may in- crease the infiltration capacity above the level expected when the soil is not frozen. Moisture contents in a frozen layer were reported to increase to as much as nine times the moisture content of the layer just before freezing. Investigators have suggested that moisture may move into the frozen layer both through vapor and capillary action, and at least one found evaporation rates increased when the soil surface was freezing. 28 Precipitation Eight to twenty inches of snow have been shown to effectively in- sulate the soil from the effect of air temperature change. The effec- tiveness of a given depth of snow increases with its wetness. Snow in its insulating role, may result in a soil freezing later in winter than a soil without snow, or with a thinner snow mantle. In the spring the same relationship may exist with the time of thaw. Snow may also hasten soil thawing in the spring through the release of melt- water. Spring rains generally hasten soil thawing. Vegetation Open areas freeze before forested areas. Conifer-covered soils freeze before hardwood soils. Grass-covered soils freeze before forested but after bare soils. Open soils generally thaw before conifer-covered soils, but after hardwood soils. Grass-covered soils thaw after bare soils. Snow accumulation is generally greater in hard- wood than coniferous areas, and greater in Open areas than coniferous areas. Deficiencies in Knowledge Information was minimal or lacking on the following points: 1. The moisture content of the soil during freezing and thawing periods. (Though some information was available on compara- tive moisture contents, there was no investigation found which followed the moisture regime in detail through continuous periods of freezing or thawing including comparative effects of soils and vegetation.) 2. The number, magnitude, and duration of soil thawing periods. 29 Changes in soil moisture-holding capacities during freezing and thawing periods. Information presented on soil pore space was not for sequential freezing or thawing periods and did not measure the type of pore volume (i.e. 60-cm tension levels, field capacity, total pore space, wilting point, or inter- mediate values). Duration of soil moisture contents, or moisture-holding ca- pacities after the final spring thaw. LOCATION AND DESCRIPTION OF STUDY AREAS General Six plots were selected on medium—textured, level (less than 6% slope) soils. One set of three plots was established in a region of high snowfall, and one set in a lower snowfall area. One plot in each set was bare, one under herbaceous cover, and one in hardwood forest. Location One set of plots was located in Kent County, Michigan, a high snow- belt area, and one set in Clinton County (Figure l). The plots were about 70 miles apart. Kent Countnglots Kent County plots were located in Gratton Township (Figures 2 and 3). The hardwood plot was about l/A mile north of the herbaceous plot, and the bare plot about 2—1/2 miles northwest of the hardwood plot as indicated below: Plot Location Bare sw l/A of NW 1/A Sec. 6, T8N, R9W Herbaceous SW l/A of NW l/A Sec. 17, T8N, R9W Hardwood sw 1/A of sw 1/A Sec. 17, T8N, R9W Clinton County Plots Plots in Clinton County were located in Dewitt Township (Figure A). The bare plot was A0 ft north of the herbaceous, and the hardwood plot was about 1/8 mile southwest of the other two. Locations were: Plot Location Bare and herbaceous NE l/A of sw l/A Sec. 2, T5N, R9W Hardwood sw l/A of sw 1/A Sec. 2, T5N, R9W 3O 31 LOWER PENINSULA /( olr marl/CAN, / l CLINTON KENT. COUNTY COIUNTY e I l l/ _ \/ , pLOTS IO 0 I0 czaL so 40 m ____J tn Figure 1. Relative location of study plots in Lower Michigan 32 I0 MILE R04 0 FARM HOUSE TIFFANY ROAD MILE ———>- Row OF 12”- lei/HICKORY TREES PLOT JO'FROM ROAD /00'FROM BA RN 00 0 MILES ro M-ll—h-l l i: l l a i.— Figure 2. Location of bare plot in Kent County 33 ABANDONED FIE L0 u: IQI uxfl 5 awAE'PLOTfiS: 1M”%C/~ ;: ZéM-NJmflo: 8 ~ \‘ «B 3 I:h <3 “ w "' wooos '93 \\ 0° IE: \\\§ \ LAKE BELD/NG ROAD (IV-44) “3M3 =x=x=x= FENCE LINES WW: 300’ V HE RBA C E OUS PLOT GAVIN LAKE ROAD x/ RA INCA GE L—gu TO 6RATTON—->- NOR TH Figure 3. Location of herbaceous and hardwood plots in Kent County 31+ __II II 1L pouMozuuarAwmo HOU$E I Houas Q I Q 3 K 2% q Q m . z 420, R 3 I E \l i, BAREPLOT- \ lg g: hmmanuvaa 2 F"""“. I anooaa); | PLOT ‘I‘ Lorr I I sfizrx' iwooas} I l l l ,— Hpfi+~gy—»+ howc'RaAD IAIFO u527 I HOUSE Figure ’4. Location of plots in Clinton County 35 Land Use and Vegetation Bare Plots (Figure 5) Corn stubble was present on the Kent County plot throughout the study. The plot had been manured the preceding winter (1958-59), but no raw manure was observed. The Clinton County Bare plot had been in the third year of an alfalfa-red clover-alfalfa rotation. Three weeks before the first sampling the plot had been plowed to 6 in. and packed. Herbaceous Plots (Figure 6) The Kent County plot was in the fourth winter since cultivation, and had not been grazed. Predominant cover, based on visual estimate, consisted of 30% Ladino Clover, 30% quack grass, and 20% each alfalfa and wild mustard. Many other species were present (Table l). The Clinton County plot was in the third year of an alfalfa- red clover-alfalfa rotation, and had not been grazed. Alfalfa made up almost 100% of the cover but some red clover was present from the previous year. Other species were scattered in the plot (Table l). Hardwood Plots (Figure 7) Neither hardwood plot had been grazed, nor was there evidence of recent cutting. The Kent County plot showed some evidence of fire with scars on the base of trees over 5 or 6 in. in diameter. Herbaceous species were generally similar and.well distributed in the two woods (Table 2). In the spring the Clinton County plot had a profuse cover of dogtooth violets not noted in Kent County. There were noticeable differences in both the number and variety of tree species (Table 3). In Kent County the predominant overstory Species were red oak, soft maple, and black cherry, with understory 36 Facing southeast from northwest corner of Kent County Bare plot. (Thermocouple stack in center of photo) Facing east across south half of Clinton County Bare plot Figure 5. Surface conditions of bare plots 37 Facing southwest across Kent County Herbaceous plot. Thennocouple stack, left center; resistance stack in background of right elbow of man Facing north on Clinton County Herbaceous plot. Hvorslev sampler shown with plunger extended Figure 6. Surface conditions of herbaceous plots 38 Table l. Vegetation of the herbaceous plots Plant Name Common Scientific Kent County and Clinton County Plots Ladino clover Red clover Alfalfa Quack grass Dandelion Mullein Kent Trifolium.sp; (commercial hybrid) ‘Trifolium pratense L. Medicago sativa L. Agropyron repens (L.) Beauv. Taraxacum sp. Verbascum thapsus L. County Plot Only Wild mustard Corn cockle Curled dock Goldenrod Common plantain Wild carrot Brassica sp. Agrostemma Githago L. Rumex crispus L. Solidago Sp. Plantago Major L. Daucus Carota L. 39 Facing southwest across center of north edge of Kent County Hardwood plot Note water Facing south from center of Clinton County Hardwood plot. tree-throw depression to left front of man in Surface conditions of hardwood plots Figure 7. ho Table 2. Herbaceous species of hardwood plots Common Name Scientific Name Kent County and Clinton County Plots Poison ivy Wild geranium May apple Violets Trillium False Solomon‘s seal Bracken fern Raspberry Honeysuckle Bellwort Dogtooth violet Bittersweet Strawberry Rhus radicans L. Geranium.maculatum L. Podophyllum peltatum L. Viola EB; Trillium grandiflorum (Michx.) Salisb. Smilacina racemosa (L.) Desf. Pteridium aquilinum (L.) Kuhn. Rubus occidentalis L. Kent County Plot Lonicera sp. Uvularia sessilifolia L. Clinton County Plot Erythronium americanum Ker. Solanium dulcamara L. Fragaria virginiana Duchesne Al Table 3. Tree species on hardwood plots Common Name Scientific Name Kent County and Clinton County Plots Black cherry Sugar maple Soft maple White ash Red oak White oak Sassafras Chinquapin oak Swamp white oak Green ash Shagbark hickory Pignut hickory Slippery elm American elm Large-toothed aspen Prunus serotina Erhr. Acer saccharum Marsh. Acer rubrum L. Fraxinus americana L. Cuercus rubra L. Cuercus alba L. Sassafras albidium Nutt. Quercus prinoides Uilld. Clinton County Plot Guercus bicolor Uilld. Fraxinus pennsylvanica var lanceolata (Borkh.) Sarg. Carya ovata (Mill.) K. Carya glabra (Mill.) Ulmus rubra Muhl. Ulmus americana L. Populus grandidentata Michx. A2 and reproduction composed mainly of soft maple. The average d.b.h. was 8.3 in. and the basal area 102.1 Sq ft per acre. (See Appendix Table Al.) 6. b. In Clinton County the average,bmdflht was 5.h in. and the basal area 76.0 sq ft per acre. There were 479 stems (2 in. and above) per acre as compared to 281 on the Kent County plot. Red oak made up the majority of the overstory with scattered slippery elm, soft maple, and white ash. Understory trees on the Clinton County plot included the Species above plus shagbark hickory, basswood, green ash, swamp white oak, and pignut hickory. The organic horizon on the Kent County plot varied from a thin duff mull to a coarse mull; the humus layer ranged in thickness from none to about 1/2 in. thick. A surface humus layer was never found in the Clinton County plot, hence the organic horizon was listed as a coarse mull.E/ Soil Descriptions Unless otherwise noted the soils in each county were developed from.glacial till. In the main, the land surface is gently rolling.2/ Kent County Bare Plot This plot drains toward the low area of Twining Loam.in its center. Tile has been laid through this low area and the spoil from the ditch Hased on the classification presented by Hoover and Lunt 1952). For detailed information on the physiography and geology of each county see Veatch (1953), Wildermuth and Kraft (1926), and Johnsgard, et a1. (19h2). Q'i #3 thrown up to the south, resulting in a Spoil deposit in soil area C, Figure 8. . ‘nn ¢F1¢C+Er' Twining fine sandy loam (soil area A, Figure 8). This wig}.- drained soil develOped from loamy parent material on ground moraine. It has a l to 2% SlOpe with a southern aspect. Horizon Depth, in. A 0-8 p Agg 8-16 B 16—l8 m B2g 18-34 c 3u—6O+ Description Fine sandy loam; very dark grayish brown (10 YR 3/2, moist); weak, medium, subangular blocky structure; friable; abrupt smooth boundary. Loamy fine sand; brown (l0 YR 5/3 moist), with few, medium, distinct, dark yellowish brown mot- tles (10 YR h/h); weak, coarse, subangular blocky structure; very friable; clear, wavy boundary. Sandy clay loam; brown (7.5 YR h/2 moist), with many, fine, distinct dark yellowish brown (10 YR h/h) mottles; moderate, medium, subangular blocky structure; very firm; fragipan. Clay loam; brown (7.5 YR h/2 moist), with many fine, distinct, dark yellowish brown (10 YR h/h) mottles; moderate, medium, subangular blocky structure; firm; clear, wavy boundary. Loam; h0% brown (7.5 YR h/2 moist). h0% dark yellowish brown (10 YR h/h), 20% gray; friable; slightly effervescent. Twining loam, deeply leached (soil area B, Figure 8). This im- perfectly drained soil is slightly eroded and has a 2%, south-facing SlOpe. Ground water was 65 in. below the surface at the time of sampling. Horizon Depth) in. A 0-8 P Description Loam; very dark brown (10 YR 2/2 moist); weak, medium, granular structure; friable; abrupt, smooth boundary. (Continued) 3on 5.5300 pamM .Ho made ..How Umaflopon .w oaswwm ML kka m. m ulbOUO‘QMIk 0 950.9% a $on ..m ox .o 8.3me «m 53 $33 \E 53.8 5093 EVE skis: o 2336: f . 33.3.? #03 “Sam. ES: ..m on ..v :33 23 $6 o . 8:33 \$§$k U ..m E ..V 36.3 22 $6 .m . 35.64 . 20N\Q0I xv. so.» XQZXW M29. 20k >65 .m \GQNNQ 3V0.» ..u\<\>§§k N ..V 0k :0 SGGQ QOKIQQ .‘ 33.8 \RVZYM. W>§K YdNflQRfl .v‘ 3‘04 XQEVM. M>§K 0\<\>\\§k .v. L .2: J , Q Q . a v. U m V v. GOOiQQ v.1 Whomuvfitml NQ‘Q Ix Q02 , 09/ AS Horizon Depth, in. Description A2 8-18 Sandy loam; brown (10 YR 5/3 moist) with common, medium.distinct, dark brown (10 YR 4/3 moist) mottles; weak, medium, subangular blocky struc~ ture; very friable; clear, wavy boundary. B21 l8-h0 Clay loam; brown (7.5 YR 5/h moist, 10 YR 5/3 moist, 7.5 YR 5/2 moist); moderate, medium, subangular blocky structure; firm; gradual boundary. 322 ho-6o+ Sandy clay loam; brown (7.5 YR 5/u moist, 10 YR 5/3 moist, 7.5 YR 5/2 moist); weak, medium, sub- angular blocky structure; friable. Twining loam (soil area D, Figure 8). The till parent material shows indications of added local alluvium. This is the lowest area of the plot. The lepe is less than l%. No erosion was apparent. Drainage is imperfect. Horizon Depth, in. Description A 0-8 Loam; very dark brown (10 YR 3/1 moist); weak, p coarse, subangular blocky structure over fine, moderate, crumb structure; friable; abrupt, smooth boundary. A 8-11 Loam; dark gray (10 YR h/l moist); weak, medium, subangular blocky structure; friable; clear, wavy boundary. B 11-20 Sandy clay loam; extremely mottled, dark grayish 21g brown (10 YR 4/2 moist), to gray brown (10 YR 5/2 moist), to strong brown (7.5 YR 5/6 moist); moderate, medium, subangular blocky structure; friable; gradual boundary. B22 20-30 Clay loam; extremely mottled, dark grayish brown g (10 YR h/2 moist), to gray brown (10 YR 5/2 moist), to strong brown (7.5 YR 5/6 moist); weak, medium, subangular blocky structure; friable; clear, wavy boundary. (Continued) A6 Horizon Depth, in. Description 30—h0 Clay loam; strongly mottled, h0% gray (N 5/0 moist), h0% strong brown (7.5 YR 5/6 moist), 20% pinkish gray (7.5 YR 6/2 moist); fine, weak, subangular blocky structure; friable; calcareous; gradual boundary. C21 C22 h0—60+ Loam, friable, calcareous. There is a distinct hardpan in the bottom of the A2 horizon in area A and not in area B. Depth to lime is 30 to #0 in. in area A; greater than 5 ft in area B. There may be a relationship between the hardpan and the depth to lime. The hardpan is not readily visible, consistence being its most outstanding characteristic. It is very ef- fective in causing a perched water table. Deep plowing (probably to combat the hardpan) has destroyed all evidence of the podzol sequm. Kent Herbaceous Plot The soil is generally uniform and the entire plot is well drained. The plot has a west facing lepe of 2%. Isabella fine sandy loam (area A, Figure 8). The parent mate- rial is glacial till with some local alluvium. Erosion is slight. The soil is well drained with a deep water table. Horizon Depth, in. Description A.p 0-9 Fine sandy loam; dark brown (10 YR 3/3 moist); weak, coarse, subangular blocky structure; very friable; abrupt, smooth boundary. B21 9-lh Sandy clay loam; dark brown (7.5 YR h/h moist); weak, fine subangular blocky structure; friable; gradual boundary. lh-28 Sandy clay loam; dark brown (7.5 YR 4/4 moist); weak, medium, subangular blocky structure; friable; gradual boundary. B 22 (Continued) #7 Horizon Depth, in. Description B -01 28-3u Loam; dark yellowish brown (10 YR LL/LL moist); 3 weak, fine, subangular blocky structure; fri- able; gradual boundary. C 34-h8 Sandy clay loam; dark brown (10 YR h/3 moist); friable; gradual. Dg h8-60+ Medium sand; yellowish brown (10 YR 5/h moist), with common, medium to coarse distinct dark brown (7.5 YR l/h moist) mottles; loose. Dighton fine sandy loam (soil area B, Figure 8). The parent mate- rial is local alluvium. Physiographically the area appears to be a minor drainageway in the tillplain. Slope is 1% with a southwest ex- posure. The soil is well drained with a deep water table. Horizon Depth, in. Description AP O-lO Fine sandy loam; very dark gray-brown (10 YR 3/2 moist); weak, medium, subangular blocky struc- ture; very friable; abrupt, smooth boundary. A 10-20 Fine sandy loam; dark brown (10 YR 4/3 moist); very weak to weak, fine, subangular blocky struc- ture; very friable; clear, wavy boundary. B 20-32 Sandy clay loam; dark brown (7.5 YR h/h moist); weak, medium, subangular blocky structure; friable; clear, wavy boundary. C 32-46 Loamy sand; dark brown (10 YR h/3 moist); single grain structure; loose consistency; clear, wavy boundary; many rounded pebbles. c 16.60 Sandy loam; dark yellowish brown (10 YR t/u moist); ' very friable. Kent County Hardwood Plot General. Soils here are classified as Dighton, rather than c Isabella, because of the sand substratum.at 3-1/2 to h ft. Soil area 3' (Figure 8) is Slightly lower than the rest of the plot and has a thicker h8 Al horizon. Lime content is low throughout. The podzol A2 horizon is discontinuous and poorly defined. Kent County Hardwood Plot Soils have substratum at #2 to #8 in. Soil area C (Figure 8) is slightly lower than the rest of the plot and has a thicker Al horizon. Lime content is low throughout. The podzol A is discontinuous and 2 poorly defined. Physiography is that of ground moraine. Dighton loam (Figure 8). In this area slope is less than 2% and area is well drained with a deep water table. Horizon Depth, in. Description Al 0-3 Loam; black (10 YR 2/l moist); moderate, medium, granular structure; very friable; clear, wavy boundary. A2 3-h Sandy loam; dark gray (10 YR h/l moist); weak, fine, granular structure; very friable; discon- tinuous boundary. B21 h-6 Loam; dark brown (10 YR h/3 moist); weak, fine, granular structure; friable; clear, irregular boundary. A2 6-13 Loam; brown (10 YR 5/3 moist); weak, fine, sub- angular blocky structure; friable; clear, ir- regular boundary. Bt 13-30 Sandy clay loam; dark brown (7.5 YR h/h moist); moderate, fine, subangular blocky structure; friable; clear, irregular boundary. C 30-50 Loamy sand; dark brown (10 YR h/3 moist); few bands sandy loam, dark brown (7.5 YR h/h); very weak, medium, subangular blocky structure; very friable to loose; gradual boundary. D 50-60+ Clean medium sand; pale brown (10 YR 6/3 moist), with few bands of loamy sand, dark brown (7.5 YR h/h); Single grain; loose. #9 Clinton County Bare and Herbaceous Plots General. Both plots are located on the tOp of a broad knoll. Parent material is a loam till. Celina loam (area A; Figure 9). The slope is 1% and the general aspect is south, southeast on the herbaceous plot, and north, northwest on the bare plot. nnodondcbr Erosion has been slight. The soil is'well drained with a deep water table. Horizon A P 2ltg B22tg ls 2s Depth, in. 0-9 9-13 13-25 25-33 .3 31-hh hh-60+ Description Loam; very dark gray brown (10 YR 3/2 moist); weak, fine, crumb structure; friable; abrupt, smooth boundary. Loam; brown (10 YR 5/3 moist); weak, fine, platy structure; friable; clear, irregular boundary. Clay loam; dark brown (10 YR 5/3 moist), with many coarse, faint brown (10 YR 5/3 moist) mottles; moderate, medium, subangular blocky structure; friable, Slightly sticky and plastic; gradual boundary. Clay loam; dark brown (10 YR 5/3 moist), with common, coarse, distinct, dark brown (7.5 YR h/h moist) mottles; moderate, medium, subangular blocky structure; slightly plastic; clear, wavy boundary. Clay loam; dark brown (10 YR h/3 moist), with common, medium to coarse, faint, dark, yellowish brown (10 YR h/h moist) mottles; weak, medium, platy structure; slightly sticky to plastic; slightly effervescent; gradual boundary. Loam; brown (10 YR 5/3 moist), with many coarse, distinct, yellowish brown (10 YR 5/6 moist), and common, medium, distinct gray (10 YR 5/1 moist) mottles; friable; nonsticky to Slightly plastic; strongly effervescent. Conover loam (area C, Figure 9). Conover areas are minor drainageways in the plots, with SlOpes which are generally less $39 580 sowfiao mo mama don 308.3 .m 8% $3 ‘2me .0 .930th MMQK $00 @500 .0 3‘00 $3200 .v. NWVS‘Q *000‘tfi a: V0d QM\_0>\00 .0 dd0>Cs 0NO0QM ..si‘0d 3S 4N0 Q :VOQ szdhb .v. 000§0Qv$ RUSK” Haggis 0 QEQNQ WWVtQ §0n~u~$fiw a: V0.» QN§0>x00 .0 39.2 smootm $.64 35 H3 .m 3‘00 vesqmb .V ~00\ ‘ EVEN: 5358‘ 51 than 1%. Internal drainage is moderate, and moderate erosion is present from water action. Mottling in soil is partially a result of a drainage deficiency. Horizon Depth, in. Description A 0-12 Loam; very dark brown (10 YR 3/2 moist); weak, p fine, crumb structure; friable; abrupt, smooth boundary. A 12-lu Loam; brown (10 YR 5/3 moist); weak, medium, sub- angular blocky structure; friable; discontinuous boundary. ":‘7 \. B2lg lh-26 Clay loam; dark brown to brown (10 YR h/3-5/3), with many, medium, distinct, dark yellowish brown (10 YR h/h moist) mottles, and common, coarse, prominent, strong brown (7.5 YR h/6 moist) mottles; moderate, medium, subangular blocky structure; firm, sticky to slightly plastic; gradual boundary. far“..— B22g 26-36 Clay loam; brown to dark brown (10 YR 5/3—h/3 moist), with black manganese concretions; mod- erate, medium, subangular blocky structure; firm, sticky to Slightly plastic; gradual boundary. Cg 36+ Loam; brown (10 YR 5/3 moist), many, medium, dis- tinct, dark yellowish brown (10 YR u/u moist) mottles, and common, fine, distinct light brownish gray (10 YR 6/2 moist) mottles; fine weak, platy structure; firm; slightly effervescent. Considerable material has been removed from the smaller knolls (area B, Figure 9), and deposited in the lower areas (area C, Figure 9). In the shallower Celina areas, the plow layer extends into the upper B horizon; in the Conover areas, local deposition produces an Al horizon lO-lh in. thick. The majority of the area is well developed Celina soil. Mottles in the Celina B horizon indicate difference in material, not drainage restriction. 52 Clinton County Hardwood Plot General. Most of the plot appears imperfectly drained. The best drained area is Celina loam (area C, Figure 9) in the northwest corner. The tree-throw depressions (area B, Figure 9) are low areas of imper- fect drainage. Conover loam. The slope is about 1% over the plot. This is a naturally imperfectly drained soil, and at the time the soil profile " w ‘0 LL description was written water was 2 ft below the soil surface. Horizon Dgpth, in. Description ! is. Al 0-6 Loam; very dark gray (10 YR 3/1 moist); strong, fine, crumb structure; slightly sticky to plastic; clear, wavy boundary. A2 6-10 Loam; dark grayish brown (2.5 YR h/2 moist); weak, medium, subangular blocky structure; nonplastic to slightly plastic; clear, wavy boundary. B lO-lh Clay loam; gray brown (10 YR h/2 moist) with many coarse, faint, dark brown (10 YR h/3 moist) mottles, weak, medium, subangular blocky struc- ture; sticky to plastic; clear boundary. B21g lh-2h Clay loam; dark grayish brown (2.5 YR h/2 moist) with many, medium to coarse, distinct, dark, yellowish brown (10 YR h/h moist) mottles; mod- erate, medium, subangular blocky structure; sticky to plastic; clear boundary. B22g 21-32 Clay loam; gray brown (10 YR 5/2 moist) with many coarse, distinct, dark brown to strong brown (7.5 YR h/h-5/6 moist) mottles; moderate, medium, subangular blocky structure; very firm hardpan at 2h- to 26-in. depth, with rest of ' horizon sticky to plastic; clear boundary. Cl 32-h0 Sandy clay loam; grayish brown to dark yellowish g brown (10 YR 5/2-h/h moist) to dark brown (7.5 YR h/h moist) mottled horizon; nonsticky to very slightly plastic; gradual boundary. (Continued) 53 Horizon Depth, in. Description 02 h0-60+ Loam; dark yellowish brown (10 YR h/h moist), 8 with common, coarse, rominent, light brownish gray (10 YR 6/2 moist and many, coarse, dis: tinct, strong brown (7.5 YR 5/6 moist) mottles; slightly sticky to plastic, slightly effervescent. The majority of the area is Conover loam (soil area A, Figure 9). It is interesting to Observe that more xeric species, such as shagbark hickory and white oak, occur in common with "wet site" species like American elm and ash. Climate The climate of Kent and Clinton Counties varies in a few respects (Table h). Most important Kent County has a growing season averaging Table h. Summary of climatological data x'int County ——Clinton 0' "oun" ' Ty Plots Station of record Greenville* East Lansing** Distance and direction 3 from plot! 12s6 ' ENE 705 " south Years or record 22 22 Tapperature factors ' 1952-60 1252-60 January mean temperature (OF) 22.2 25.7 25.5 26.0 June mean temperature ( 66.h 69.3 Maximum temperature th.O 99.0 Minimum temperature -25.0 -2.0 -9.0 -5.0 Last spring frost 26 May 28 Apr 16 May 8 May First fall frost 27 Sept 28 Oct 16 Sept 16 Sept Mean growing season (days) 1&8 157 Precipitation factors Mean annual precipitation (inches) 30-5 29-9 Mean snowfall (inches) h3.2 39.9 3 In MontcalmCounty, 11 miles east of the Kent County plots. ** Moved from Lansing (Capitol City Airport) June 195%, and back in May 1959. Airport is 5 miles southwest of the Clinton County plots. 51+ 9 days longer than in Clinton County, with attendant later last Spring frost dates and earlier first fall frost dates. Further, the average snowfall is 3.3 in. more in the Kent County area. In 1960 the mean January temperature was 25.6 and 26.0 F in Kent and Clinton Counties, 3.5 and 0.5 F higher than their respective ‘fi. averages. An interesting feature was that the first and last frosts occurred almost a month earlier and a month later than normal in Kent County, F EL. while the last frost was the same date as the average in Clinton County, and the first frost 8 days earlier. EXPERIMENTAL PROCEDURE Design of Sampling Plots An important consideration in designing plots was the need to avoid trampling the snow on areas prior to sampling. Various authors had warned that compaction of snow cover might affect depth of freezing of soil buried beneath it or near it. Plots 10 ft square were selected ’ as large enough to provide space for maneuvering within the sample ! square, yet small enough to permit layout of many squares in a reason- ably sized sample plot. Each of the six sample plots was designed to contain 256 10- by lO-ft squares (arranged in a 160— by iéo-rt plot), divided into 4 quadrants of 6% plots each. Figure 10 is a diagram of a sample plot. Installation of Plots One corner of the plot was established and boundaries were run in cardinal directions using a hand compass and tape. Stakes were placed 10 ft apart along opposite sides of the plots to guide sampling. One stack of thermocouple units and two stacks of plaster-of-Paris resist- ance blocks were placed at points on the plot boundary, Figure 10. Resistance blocks were installed in a l-ft-diameter hole about 18 in. deep. At the designated depths a jackknife was used to cut a rectangular hole just large enough to insert the resistance block with some pressure. Extra space around the block was packed with soil from the same depth. The resistance blocks were offset 90 degrees from.one block to the 55 56 .lv. 1J82753347€32l63774 653371336858/528 464/]455/62/8345 M2888g87/478NUH433 l4/286248563 862 2g I. .0 7. at t: I. .d a. .4 I“ 2. .4 so .I go 7. .J :4 .4. 1w .1 n. .4 <4 .I 7. c. .4 I. .3 I. .04572657867337628“ ”1366/7/62/2788873r 5.434438428462128 [847522834/2/244 77/6383746833335 8/7345/5552/AH7662 9‘ {a .6 (a .I to 1. I. 91 7. .J (a .4 ,4 I. .1 4422844381576586 6353675£r53645757 l2345678fi8765432l s°\T|l| ‘o > b \ V U ABCDEFGHHGFEDCBA LEGEND o RESISTANCE BLOCKS [J THERMOOOUPLE STACK Plot diagram and sampling sequence for all plots Figure 10. IO’ 5! B 5.5 \Mg/ RI \ I S I | I l ~49 5’ =i LEGEND ( ) SNOW SAMPLE (S) \— o MOISTURE-DENSITY(MD) o REMOLDING INDEX (RI) 0 SH EAR STRENGTH ($5) I __y‘~ Figure 11. Diagram of sample square w . . . .. le 58 block above it. The block was installed in its hole, the wire led across the pit, the soil was packed into the level at which the next block was to be installed. The mechanics of thermocouple installation were similar. A.pit for the thermocouples was dug about 2 ft deep; a post hole auger (5- in. diameter) was used to extend the hole to MB in. Thermocouples were installed in the hole by bending the end of the wire 90 degrees to the lead-in wire, and lowering this portion to the proper depth. Between depths the soil was tamped by a 2-be 2—uL by 2-i33 stake, or a round (h-in. diameter) iron tamp, to the level of the next depth. In the pit, the end of the wire was pushed about 3 in. into the side. Sampling Procedures Modified Scheme of Randomization The requirement to eliminate trampling on unsampled squares pre- cluded complete randomization of sampling. Instead, a modified scheme of sampling was used to insure that trampling would occur only on plots already sampled. The scheme provided that an outermost row be sampled before the next inner row was begun. Details of the procedure are discussed in the next section. Selecting Sample Squares Each sampling quadrant contained eight columns of sample squares labeled A through H, and eight rows numbered 1 through 8 (Figure 10). The nine squares within each row were assigned random numbers. Each of the 256 squares could be identified by a combination of letters and a number to indicate quadrant, row, and column, in that order. For 1:4L‘9W I 59 example, the four squares occupying the four corners of the sample plot are designated as AlA, BlA, ClA, and DlA. A sample schedule was prepared in advance. The numbers 1 to 8 were drawn at random for each square of each quadrant, and assigned to the columns in the order they were drawn. Thus, on the first drawing the numbers 7, 6, l, 2, 5, A, 8, and 3 were drawn in that order, and assigned to sample squares AlA, AlB, AlC, AlD, AlE, AlF, AlG, and AlH, , respectively. The numbers 6, h, 2, 8, 7, l, 5, and 3 were drawn for I-HW the respective columns in row 1, quadrant B. FT‘ On the first sampling day the squares used were AlC, BlF, ClE, and DlG. On the eighth sampling day the squares were AlG, BlD, ClA, and DlB. On the ninth sampling day the sample squares were A20, B2G, 02A, and D2E. For a given visit, the same designated squares were sampled in all three plots first in one county, then in the other. Adequacy of Number of Samples Four squares, one from.each quadrant, comprised a sample. This number was chosen for two reasons. At least four observations per sample are necessary to permit a "t" test sensitive enough to detect differences between means in the parameters estimated. Four plots per sampling were all that could reasonably be scheduled for a two-man crew. Each days sampling normally required #8 soil temperature read- ings, 36 resistance readings, 12 snow samples, 60 moisture-density samples, 12 series of shear strength readings, plus travel time. Collection of Data Properties may be grouped under the general headings of 60 vegetation, weather, static soil properties, and dynamic soil properties. Vegetation Herbaceous plots. Vegetation was enumerated by species, and the relative proportion of each species over the plot was estimated. Hardwood plots. Basal area of trees 2 in. in diameter or over was measured and totaled as an estimate of the cover. Herbaceous vegetation was listed. Weather Factors Precipitation and air temperature. Precipitation was measured with a recording rain gage on the Kent County Herbaceous plot. Standard rain gages were used on the other plots. Records from the weather sta- tion nearest each set of plots were used for air temperature data. Snow mantle. Snow was measured both for its effect as an insulator, and as a potential source of soil moisture. Four snow samples were taken on each sample day (one in each sample square). Snow depth, den- sity, and water equivalent were determined from these samples. Each h—in.-diameter core was taken in the center of a sample square, directly over the position for the moisture-density core (Figure ll). Static Soil Propertiesé/ Static soil properties are those which would not be expected to change over a short period. They include soil texture, organic matter content, specific gravity, and Atterberg limits. Texture. After the last freeze of 1960, bulk samples were taken 6/ Soil texture, specific gravity, Atterberg limits, and organic matter (loss-on-ignition) were determined by procedures described in Soil Laboratory Manual, Lower Mississippi Valley Division, U. S. Army Corps of Engineers, Waterways Experiment Station, Vicksburg, Mississippi. 61 on three occasions. The three randam samples from each depth of each quadrant were composited. Mechanical analyses were run on the composite samples by quadrant and depth. . . o H 0 Organic matter. Organic matter was determined by the loss-92C ignition" method (Wilde and Voigyi, 1955). A composite sample of the plot for each depth was used. Specific gravity. To help characterize the soil on each plot, and to provide a basis for total pore space determinations, specific gravity determinations were made for each depth on a composite of the spring samples. Dynamic Soil Properties Dynamic soil properties are those that would be expected to change from day to day such as soil temperature, depth of freezing, bulk den- sity, and soil moisture. Soil temperature records for Michigan indicated that soil could freeze to a depth of A2 in. (USDA, 19u1). In this study, physical observations of soil freezing were made to 15 in. It was hoped that by correlating observed temperatures with observed freezing in the upper 15 in., greater depth of freezing could be estimated through temperature observations alone. Thermocouples were installed at the specific depths below the soil surface to represent the soil layers designated in the tabulation below. Depth Depth Depth Depth Layer of Unit Layer of Unit Layer of Unit Layer of Unit in. in. in. in. in. in. in. in. o 0 9—12 10.5 21-2u 22.5 33—36 34.5 0—3 1.5 12-15 13.5 24-27 25.5 36—39 37.5 3—6 4.5 15-18 16.5 27-30 28.5 39-u2 no.5 6—9 7.5 18—21 19.5 30—33 31.5 u2-u8 u5.o 62 Frost depth. Frozen depths were determined from.moisture—density cores. It was found that the difference between a frozen and unfrozen portion of a core could readily be determined using the point of a jackknife. The frozen portion of a core was always extremely hard, and the unfrozen portion soft. Bouyoucos resistance blocks were placed at 1.5, A.5, 7.5, 10.5, 13.5, and 16.5 in. below the soil surface to indicate the successive 3-in. intervals to 18 in. A meter reading of 10% or less of available moisture was a strong indication that the soil around the block was frozen. Soil moisture and bulk density. The Hvorslev samplerZ/and a mechanical boring tool were used to take moisture—density cores. The Hvorslev sampler, used when the soil was frozen less than 2 in. from the surface, took cores up to 6 in. long. With mechanical boring tool frozen cores to 18 in. long could be obtained. In either case, the cores were cut into 3-in. lengths. Both samplers used a stainless steel sample tube with an inside diameter of 1.87 in. A.moisture-density core was taken in the center of the sample square (Figure 11). The soil cores were extruded, by the piston from the Hvorslev sampler, and by a wooden block fitted into the tube from the mechanical sampler, into a special cutting box 18 in. long (Figure 12). The cores were then cut into 3-in. lengths with a thin butcher knife, or a '7/ For complete details on construction and use of the Hvorslev sampler see Waterways Experiment Station (l9h8). 63 Frozen core in cutting box. Box divided into 3-in. sections on each side to pennit cutting to even lengths Hvorslev sampler in ground under 10 in. of snow, Kent County Herbaceous plot Figure 12. Core cutting board and Hvorslev soil sampler 6h piano-wire cutter, or, in the case of frozen cores, a hack saw. When the line of separation between the frozen and unfrozen por- tion of the core did not fall at a 3-in. division line, the 3-in. core was divided into frozen and unfrozen portions. Each portion was meas- ured and moisture and density determined separately. Frozen portions less than 1/8 in. long were noted as frozen, but were not separated from the unfrozen portion of the sample. [The soil tube on the mechanical sampler (Figure 13) was connected to a flexible drive shaft. The shaft in turn was driven by a 2-l/2—hp, 2-cycle gasoline engine. The soil tube turned at approximately 300 rpm when not under load. When the tube was inserted into the soil the turn- ing rate varied from.a dead stop to a maximum of 300 rpm. Frequently a good sample could not be obtained with the first bor- ing with the mechanical sampler. However, cores were extracted until a 15-in. core was obtained. The three major causes for difficulty are listed in order of frequency of occurrence: (1) stones in the profile, (2) excessive melting of the frozen core during drilling, (3) breaking of the core in the sample tube. 60-cm tension values. Samples for tension values were taken on one occasion before freezing. Tension cores were taken in lieu of the moisture-density core for that day. Shearing strength. Measurements were made at frequent intervals even while the soil was frozen whenever, with reasonable effort, the instruments could be inserted into the ground. Measurements were made at 3-in. intervals from the surface to 18 in. . OJ " ‘0 Taking frozen core under 3 in. of snow on Clinton County Bare plot. One man holds sled to prevent jump- ing while engine is running ’0“! vb. Machine at rest. Note flexible drive shaft leading to sample shaft housing. Reduc- tion gear connects flexible shaft to sam- ple shaft. Mercury clutch connects flex- ible shaft to engine Taking frozen core under 1h in. of snow, Clinton County Herbaceous plot Figure 13. Mechanical sampler 66 Shear strength readings were taken 1 ft to the left of the moisture-density core. A l2-in. core was removed 1 ft to the right ‘ of the moisture-density sample for a special remold test (Figure 11). RESULTS Static Soil PrOpertieS Detailed knowledge of static soil properties is essential to in- telligent interpretation of the phenomena associated with moisture, bulk density, and Shear strength. These properties are described here for each plot. Soil Texture Moisture—density sampling revealed considerable variation in tex- ture over each plot. To obtain an accurate estimate of soil properties composite samples were obtained in the spring for each quadrant of each plot. Each depth was represented by a sample composited from three points in the quadrant; 12 samples to a depth. Soil texture classes are Shown in Table 5. Table 5. Soil texture classes of each plot Kent County Clinton County Depth Plot Quadrant Plot Quadrant in. Avg A _E_ _E‘ .;2 Avg A;. _B C D Bare 0-3 SL SL SL SL _SL L L SL L SiL 3-6 SL SL SL SL SL SL L SL SL SL 6-9 SL SL SL SL SL L L L SL L 9-12 SL SCL SL SL SL L L L L L l2-15 SL SCL SL SL SL L L L L L Herbaceous 0-3 SL SL SL SL SL SL L L SL SL 3-6 SL SL SL SL SL L L L SL L 6-9 SL L SL SL SL L L L L SL 9-12 SL L SL SL SL L L L SL L 12-15 SL L SL L SL L L L L L Note: SL — sandy loam; L - loam; SiL - silt loam; SCL - sandy clay loam; CL - clay loam. (Continued) 67 68 Table 5. (Concluded) JKEnt County L Clint on County Depth Plot guaggant Plot Quadrant .__1n- ._s_Av .5. .12. .2 .2 .aAv .5. .13 _c_ i Hardwood 0-3 SL SL SL SL SL L SL L SiL SiL 3-6 L SL L L L L L L SiL SiL 6-9 SL SL SL L ‘L L L L SiL CL 9-12 SL SL SL SL SL L L L CL CL 12-15 SL SL SL SL SL L L L CL CL Kent Countyr plots. These plots were all sandy loans with the ex- ception of the 3— to 6-in. depth of the hardwood plot. This depth was classified as loam. Clinton County plots. Except for 3- to 6-in. depth in the bare plot and the 3.. to 6-in. depth in the herbaceous plot, the soils in Clinton County were classified as loams. Organic Matter As the modified potassium dichromate method may give low values with high organic matter contents (Wilde and Voigt, 1955), this method was not used in the upper two depths of the hardwood plots. The loss- on-ignition method was used. Thus the surface to 3- and 3- to 6-in. depths in the wooded plots could not legitimately be compared with other depths . The surface 3 in. of both hardwood plots proved to contain over 8% organic matter by weight (Table 6). The surface to 3-in. depth of the Kent County Bare plot had 2.87%, probably as a result of the manure applied the previous winter. The other surface to 3-in. layers fell between 1 and 2% organic matter content. 69 Table 6. Organic matter contents by modified potassium dichromate method (loss-on-ignition method used in top two depths of hardwood plots) 535th, in. Kent County Clinton County Bare 0—3 2.87 1.88 3-6 2.23 1.93 6-9 0.91 0.62 9-12 0.55 0.51 12-15 0.51 0.62 Herbaceous 0»3 1.h5 .98 3-6 l-hS .88 6-9 0.35 .78 9-12 0.u2 .70 12-15 1.05 0.59 Hardwood 0-3 8.26 8.70 3-6 2.35 n.56 6-9 0.70 1.29 9-12 0.59 0.62 12-15 0-h2 0.59 The lower three depths were below 1% in all but the Clinton County Hardwood plot, while the 3- to 6-in. depths were above 2% in the Kent County Hardwood and Bare plots, and between 1 and 2% in the other plots. Specific Gravity Specific gravity determinations (Table 7) generally fall within 70 Table 7. Specific gravity g/cc Depth Kent County Clinton County in. Bare Herbaceous Hardwood Bare Herbaceous Hardwood 0-3 2.63 2.63 2.55 2.65 2.65 2.57 3—6 2.63 2.65 2.6M 2.6a 2.65 2.65 6-9 2.65 2.65 2.65 2.66 2.67 2.68 9-12 2.66 2.66 2.66 2.68 2.68 2.69 12-15 2.66 2.67 2.66 2.69 2.70 2.69 the range expected for mineral soils, i.e. 2.6 to 2.7 g/cc (Lutz, 1951). Organic matter was not removed from the soil in the analysis, and is reflected in the relatively low specific gravities of the surface to 3-in. layers of the Kent and Clinton County Hardwood plots, 2.55 and 2.57 g/cc, respectively. Atterberg Limits Atterberg limits are shown in Table 8. Sample treatment, soil colloid content, mineral composition of the colloids, and organic matter all affect Atterberg limits (Bayer, 1956). The liquid and.p1astic limits in the surface to 3-in. depth of the bare plots and the surface to 3— and 3- to 6-in. depth of the hardwood plots were higher in Clinton than in Kent County. This was to be ex- pected from the higher organic matter content of these depths in Clinton County. Other variation in values is not susceptible of simple explanation. 71 Table 8. Atterberg limits for each plot Per Cent Moisture by Weight Depth Liquid Plastic Plasticity County in. Limit Limit Index Bare Kent 0-3 23 18 3 3-6 21 17 t 6-9 1h 1h 0 9—12 11+ 13 1 12-15 22 12 10 Clinton 0-3 20 17 3 3-6 21 17 h 6.9 21 13 8 9-12 2h 13 11 12-15 28 1h 1% Herbaceous Kent 0-3 18 17 1 3-6 16 15 1 6-9 21 13 8 9-12 18 13 5 12-15 18 lb h Clinton 0-3 22 17 5 3-6 21 17 t 6-9 20 13 7 9-12 19 13 6 12-15 2h 13 ll Hardwood Kent 0-3 hl 31 10 3-6 20 18 2 6-9 18 15 3 9-12 17 1h 3 12-15 20 1h 6 Clinton 0-3 51 36 15 3-6 MO 29 11 6-9 2h 17 7 9-12 21 15 9 12-15 26 15 11 72 Soil Freezing and Thawing Temperature Conditions In- fluencing Freezing Opportunity Air temperature. The daily plot of air temperature for both counties is shown in Figure 1h. Temperature changes follow a similar pattern with the high and low extremes usually greatest for Clinton County. Periods above and below 32 F are of about the same duration for each weather station. Monthly mean air temperatures for the two counties summarize the major differences. Data were taken from Greenville Weather Station, Montcalm County, and Lansing (Capitol City) Airport, Clinton County; stations nearest the Kent and Clinton County plots, respectively. 1959—60 Monthly Mean Air Temperatures, Nov Dec Jan Feb Mar Apr Mean Kent County 33.8 32.3 25.7 25.5 2u.8 h9.8 32.0 Clinton County 32.7 32.7 26.0 23.0 22.0 A8.7 30.8 On the average the Clinton County area was 1.2 F colder than the Kent County area. In Clinton County February and March were 2.5 and 2.8 degrees colder, respectively, than in Kent County. The majority of soil freezing occurred during these months. Degree-days. Table 9 lists the periods when air temperatures were below 32 F. The period in one county may have been one to three days longer or Shorter than a period in the other, but in every case, one overlapped the other. There were 10h days in which air temperatures were below freezing in Kent County, and 110 days in Clinton County. Degree-days were calculated using the mean temperature in relation to a base of 32 F. AS degree-days are considered here as a measure of the freezing Opportunity, those days on which the air temperature 73 moflpgoo sovcfluo 9.8,, pcoM ammhfinaomamo. 8.8 can: .12” .8me ¢&( 1(2 owu 2(fi Uwo .----..-. 20.530 Ill Eu: OZuUmJ can. JJBHNBUHVJ $338930 7h .0m >02 a npfls msquHmop «m mm koaon mo so pomsm * m.w m>< s.w+ ma< o.:sm+ Hspoe o.asm+ Hanoe OHH anoe HOH anoa o.o m.wam+ m.mpw+ m.H+ o.aws+ m.mms+ H na< m na< m m has 0H na< m mH m.a+ m.msm+ o.mac+ m.o+ o.mms+ m.sms+ N has 6 naa m H has m naa m sH o.m0m+ o.ome+ o.Hmm+ m.msm+ m.mow- o.mmm+ on as: mm pom SH mm as: mm ppm mH mH o.ma+ m.mmm+ m.saH+ m.mm+ o.smm+ m.mmH+ OH boa mH nos 6 a pot mH not a «H m.mmH+ o.maH+ m.mm+ o.Hs+ m.O~H+ m.mmH+ Hm mom a can mH m pot a can Hm HH m.sHH+ m.HMH+ o.aH+ mH can am can mH 0H m.~+ m.Hm+ 0.:H+ m can mH ssh HH m o.am+ o.mm+ o.mm- o.sc+ o.mH+ o.ws- a she HH can m m can m can a w m.ma+ o.am- m.ow- m.mm+ m.sa- o.aw- a can H can mm m can u one mm a m.ms+ o.ms- m.Hm- m.ws+ 0.66- m.aHH- a one am one wH m con mm com pH 6 o.H+ 0.0a- o.H>- o.o o.mm- o.mm- m can SH one mH H can MH com mH m m.m+ o.mm- m.mm- o.m+ m.me- m.mm- m one a one 6 m com a com S a o.wm+ o.mm- o.mm- o.mH+ o.mm- o.sa- m >oz om >oz mm m soz om >oz mm m o.mo+ o.sm- o.mm- m.mm+ m.mm- o.mHH- OH >oz Hm >oz NH m >oz om >oz NH m m.H+ o.mc- m.mm- o.H- o.se- o.ma- m >oz a >oz m H UOHHom ham mam doahom ham mam mama ham awn mama ham awn .oz not pmsH pnnHa not panH pnnHm anoe panH annHa anoa pang pnnHm ooHnom wasp wasp apesoo sochHo hpsdoo pang nomnmma immawom A0>HdeoqutmmpsQ oOHumm mGHNooMm hpgsoo songHo hpQSOO puma whatnomnmom o>Hpnasaso *mcoHsom msfluooam hp meSOHm poam £088 90% manpuomamom .m manna 75 averaged over 32 F were given negative values, and those with an average below 32 F, plus values. Thus, a day with a mean temperature of ho F would have a degree-day value of -8, and a day with a mean of 2h F would have a value of +8. Comparison of periods of equal duration revealed that the Clinton County plots had slightly higher freezing opportunity than Kent County. Only during the period from.February 17 to March 26 was Opportunity ap- preciably different (Period number 13, Table 9). Clinton County had a total of 509 degree-days, and Kent County, 376.5. Average degree-day values were 8.h and 8.9 for Kent and Clinton Counties, respectively. Soil temperature. Thermocouples were installed to indicate soil temperature and suggest when the soil should be sampled. Since frozen cores could not be taken in the immediate neighborhood of the thermo- couples, the supposed utility of temperature readings to indicate soil condition (i.e. frozen or unfrozen) soon proved to be invalid.§/ However, comparison of frost and soil temperature data suggests several relationships (Table 10). The highest recorded soil temperatures when the soil was frozen were no.0 and h5.u F in the surface (0 to 3-in.) layers of the Kent County Bare and Clinton County Hardwood plots (Table 10). Other maximum readings were below 37.7 F (3- to 6-in. reading in the Kent County Bare plot). Thus, though frozen ground is found on a given plot, tempera- tures well above freezing can be found elsewhere in the same plot. 8/ Soil temperature data is on file (vs 3h.2) with the Trafficability Section of the Army Mobility Research Center, U. S. Army Engineer Waterways Experiment Station, Vickaurg, Mississippi. 76 Table 10. Minimum and maximum temperatures in frozen soil; number of frozen cores; and minimum temperature of unfrozen soil, deg F Depth, in. Reading 0 to 3 3 to 6 6 to 9 9 to 12 12 to 15 Kent Bare Maximum u5.0 (1)* 37.7 (2) Minimum 22.1 (3) 30.1 (1) Minimum unfrozen 37.3 25.9 Kent Herbaceous Maximum, 37.0 (h) 36.5 (1) Minimum 29.0 (t) 32.0 (2) Minimum unfrozen 31.2 29.2 Kent Hardwood Maximum 3h.0 (1) Minimum 29.6 (1) Minimum unfrozen 29.3 Clinton Bare Maximum 33.9 (t) 33.6 (h) 3h.8 (A) 35.2 (1) 3u.t (1) Minimum 19.3 t) 22.h (h) 20.7 (h) 27.7 (t) 29.3 (3) Minimum unfrozen 32.3 28.8 28.0 21.6 22.2 Clinton Herbaceous Maximum 33.9 (t) 36.0 (1) 33.2 (1) Minimum 22.3 (h) 23.u (2) 23.1 (2) Minimum unfrozen 28.5 2h.l 26.6 Maximum 45-8 (2) Minimum 28.8 (2) Minimum unfrozen 22.5 * Numbers in parentheses are numbers of frozen cores in four soil sam- ples. Depths are shown only if some freezing was noted. 77 In each county, the lowest temperatures when the soil was frozen were in the bare plots, the next lowest in the herbaceous, and the highest in the woods. For plots with the same cover, depth by depth, the lower tempera- tures occurred in Clinton County. These plots had appreciably less snow cover than those in Kent County. The surface layer on each plot had the lowest minimum.temperature, reflecting the fact that this layer is more subject to daily tempera- ture drops. Deeper layers would reflect the daily temperature fluctua- tion to a lesser degree. This agrees with the findings of Serova (1958). Minimum.temperatures occurring when the soil was not frozen are also revealing. In every case, other than the surface to 3-in. layers in the bare and herbaceous plots, the temperatures range from 2.8 to 10.4 degrees lower than 32 F, ample, it would seem, for soil freezing. Again this indicates that soil temperature at one point on a plot does not necessarily reflect the temperature elsewhere on a plot, nor can it take into account the solute concentration of the soil solution or the moisture tension levels prevailing elsewhere. Soil Regime (Frozen and Thawed) The first sampling of all three plots was made December 2 in Kent County, and December 9 in Clinton County. Kent County Bare plot (Table 11). Frozen ground was first found on this plot December 10. There was frozen ground thereafter to March 30 (Figure 15), though there were as many as five days in a row when freezing air temperatures did not occur (e.g. December 13 to 18, Figure 1h). Heat increments during these periods were not enough to remove all frost from.the ground. 78 Table 11. Mean frost depths, Kent and Clinton Counties Kent County Frost Depth, in. Clinton County Frost Depth, in. Date Bare Herbaceous Hardwood Date Bare Herbaceous Hardwood 1959 1959 12 Nov 0.00 No data 10 Nov 0.00 No data 17 Nov 0.00 No data 11 Nov No data 0.00 No data 27 Nov No data 0.00 No data 19 Nov No data 1.82 30 Nov No data 0.00 25 Nov 0.00 No data 2 Dec 0.00 0.00 0.00 9 Dec 0.00 0.00 0.00 A Dec 0.00 0.00 0.00 in Dec 1.55 0.00 0.00 10 Dec 1.17 0.00 0.00 18 Dec 0.72 0.00 0.00 16 Dec 0.2a 0.00 0.00 23 Dec 9.00* 1.5u 0.6h 21 Dec 2.18 1.56 0.60 28 Dec 0.00 0.00 0.00 30 Dec 0.60 0.00 0.00 31 Dec 0.2M 0.00 0.00 1960 1960 2 Jan 0.81 0.30 0.00 5 Jan h.37 1.31 0.36 7 Jan 1.20 1.32 0.00 9 Jan 8.h2 h.12 0.60 11 Jan 1.62 2.3a 0.36 16 Jan n.0h 1.11 0.00 1h Jan 1.80 1.98 0.00 20 Jan 8.32 3.5M 0.00 19 Jan 2.0h No data 23 Jan 6.85 2.8% 0.00 21 Jan 0.36 0.57 0.00 28 Jan 7.95 3.02 0.00 26 Jan 0.93 0.72 0.00 2 Feb 8.u0 u.38 0.00 30 Jan 1.23 0.61 0.00 h Feb 9.12 2.25 0.00 3 Feb 0.30 0.2M 0.00 9 Feb 9.2h 2.h0 0.00 6 Feb 1.12 0.00 0.00 13 Feb 10.92 1.71 0.00 12 Feb 0.99 0.12 0.00 17 Feb 8.71 2.0u 0.00 16 Feb 1.32 0.h8 0.12 2h Feb 11.0u 3.00 0.h8 20 Feb 0.93 1.h1 0.00 1 Mar 10.50 3.00 0.h8 25 Feb 0.39 1.17 0.80 8 Mar 15.00+ h.17 No data 5 Mar 0.5h 1.32 0.00 11 Mar 12.50 h.05 1.h7 10 Mar 1.17 1.50 0.8h 15 Mar 10.18 5.u9 0.17 1h Mar 3.00 1.56 0.8h 21 Mar l5.00+ 2.69 0.2M 17 Mar 1.51 2.11 0.00 25 Mar 10.20 3.90 0.72 23 Mar 2.19 1.96 0.36 29 Mar 5.76 3.03 0.00 28 Mar 0.2h 3.12 0.60 31 Mar 5.h0 3.2h 0.00 30 Mar 0.72 2.21 0.72 3 Apr h.08 0.00 0.00 1 Apr 0.00 0.00 0.00 6 Apr 1.92 0.00 0.00 9 Apr 1.32 0.00 0.00 * Only two cores taken. 79 H3560 psmvH Hmnpmod chomp was 39.6 saw: .mH ossmHa me<¢oom< 262m .952 one. one. «(x mun zOn< 302m. ”wFOZ one. coo. ma< «<3 mu... 2!. one o. ..n on 2 cu aw o. .n ow o. 5 cu o. . i 4* . 1 . q a q a o q 4 J a a 4 eVVa V» V. Va. 3 Va \ ‘ NV.“ M? w I . c _ _. .56 oz .56 oz\. 0003015... - a J — u q u — q u D _ q q o q d 4 q W \VVV V NV V n VV VV V V V V V V V — b —— — p n — P— p m — s M _ rye ‘P:— am _ _9< .56 9<\ 3035mm: — 4 q - 4 d + V w V V? M V V O. EWJUOS ONOOUS W083 $3HONI NI 33NVJSIO 85 illustrate the relationship of snow on the ground to frost. On these plots vegetation directly affected the relative amount of snow present. In Kent County snow accumulation did not begin until January 18 (Figure 15). The maximum snow accumulated prior to this time was 2.61 in., and on the last sampling day before January 18, all snow had dis- appeared (Table lh). Mean snow depths per plot are shown below (based on the average snow depths from January 19 to March 30). Mean Measured Snow Depth in Inches, Kent County January February yggggg Bare 7.2h 8.03 9.96 Herbaceous 8.51 9.35 10.6h Hardwood 6.24 7.42 11.51 Mean snow depth increased on each plot from.January to March. Ex- cept for March when snow in the woods was deepest, snow depth decreased from herbaceous to bare and bare to hardwood plots (Table 1H). Snow had disappeared from.the bare and wooded plots on March 29, one to two days after it left the herbaceous plot. In Clinton County snow accumulation began on the 16th of January on the herbaceous and wooded plots, and on the 20th of January for the bare plot, Figure 16, Table 1h. Mean monthly snow depths are shown below. Mean.Monthly Snow Depths in Inches, Clinton County December January February .EEEEE Bare 2.15 2.99 2.53 3.36 Herbaceous .2.61 5.17 5.66 10.22 Hardwood 2.uo 3.22 n.92 9.71 Snow accumulation increased from January to March; the herbaceous 86 Mean snow depth, density, and water equivalent Table 1h. T Hardwood water Den- Herbaceous water Den- Depth sity Equiv Bare water Den- Depth sity Equiv chc V in. .flg Depth sity in. g[cc in. in. in. in. Date Kent Count 8 98 6815 1 1136825 01 h3h6mm&0my 9657276 eeCeeeeeeeeeeeeeeeee 00 m00000011111111210 7 35“ 531 h 9h Rm nwowmlllmm11121m11 eeieeeeeeeeeeeeeeeee % 00 .00000000000000000 h m C l h3 77 h8915 8 t %2% mm9owm33 3388722m m mLme&&&&i&%%&mm&..k%i r 1 111221 0 W nt89 8319 hh6h6 85 h a000 8h“? 89909 09&%®6 lee heeeeeeeeeeeeeeee mpoo “0000001211212221 w w w h 76 5689 58267577 20 /&00 n0000nm11211111m2n lneeieeeeeeeeeeeeeeees <000 .00000000000000000 V n W0h91159h6w26h6 hh3878N mmd532808£.3062.33368£ 8 ll .7 977999 h 7 o 01 79 mum21m tN n .m 39 7 9 335 8 8 h237 n 1o oatmeaouauoueaona m 00 00010011121212220 m w I 86 HT? 3 877888 77lh5 00 800m1m111111m11222 ee eeeeeeeeeeeeeeeee 00 m00000000000000000 .Mw enammmaauawnmnmma 1.11nv vuvuzzvunuhwnunuqunumwmgmwwummny.1 mm mflfi®®$$$®$$®$$®®®$®®®$$ mmmmmmmmmmmmmmmmmmmmmmm HEaTumwflfim36mwm55mmflfiw Clinton County < 1/h in. snow Rain 2.h0 0.03 0.09 Thin snow crust No snow 0.69 0.06 0.0h 3.09 0.07 0.23 h.h0 0.06 0.25 h.72 0.09 0.h3 3.79 0.09 0.37 3.06 0.16 0.50 5.61 0.1h 0.77 h.h7 0.19 0.85 7.69 0.1h 1.10 9.98 0.11 1.08 8.61 0.13 1.1h 9.h3 0.15 0.h2 9.28 0.16 1.h8 10.81 0.1a 1.52 10.15 0.15 1.56 h.23 0.18 0.76 0.07 Rain No snow 03 03 05 .01 09 ll 10 10 11 09 ll 12 15 13 13 < l/h in. snow 2.61 0.03 00 5% 5 #531733mm 6 65 oeatsao.. 1/2 to 3/h in. snow 5. 1 5 a amesewunsam O O O I O O O O O O O O I 0 00000000000000 0. 7hl65533l§1 11211111112 eeeeeeeeeee 00000000000 2771 3 n 2131M3umm lee eee e e eee e 1333h33333 < l/h in. snow Rain 2.15 0.06 No snow No snow No snow 2.68 0.0h 3.23 0.08 .05 0.02 3 aaaaaaaaaaaaaa mmmmmmmmmmmmmm aau9nna1nwnns 87 plot showed.more snow than the woods or bare plot each.month. Table 1h reveals that this same relationship held true on a sample to sample basis. Snow disappeared on the Clinton County Bare and Herbaceous plots between the 25th and 29th of March, and from the Clinton County Hard- wood plot between the 29th and 3lst. The month to month relationship of snow depths between the two counties is summarized below. ‘With the exception of December when no snow was measured in Kent County, snow on the ground was generally higher, plot for plot, in Kent than in Clinton County. Difference in Mean Monthly Snow Depth, Kent County over Clinton County Plot December January February' E9523 [Mean Bare -2.15 n.25 5.50 6.60 3.55 Herbaceous -2.61 3.34 3.69 0.h2 1.21 Hardwood ~2.h0 3.02 2.50 1.80 1.23 Daily air temperature and degree-day data revealed that the freez- - ing opportunity was somewhat greater in Clinton than Kent County. In addition, the Kent County plots generally had deeper snow on the ground than the Clinton County plots. Frost regime before snowfall. From.the standpoint of its effect on soil freezing, it is assumed that there was no snow until the samplings of January 19 in Kent County, and January 20 in Clinton County. Cumula— tive degree-days from November 1 were 51.5 and 52.5 for these areas, respectively--essentially the same. All three plots in the county were sampled on the same day; nine times in Clinton County, and eight times in Kent County. Mean frost depths up to the beginning of continuous snowfall are shown on the next page. 88 Mean.Frost Depth Before Snowfall in Inches Bare Herbaceous Hardwood Kent County 2.98 0.9h 0.16 Clinton County 2. 62 0. 85 0. 12 The mean depths do not vary greatly for plots under similar cover. However, in each county the effects of cover are evident; the bare plots had the deepest frost; the herbaceous plots the next deepest; and the hardwood plots the least. Frost regime under continuous snow cover. Considerable effort was spent in an attempt to develop quantitative relationships between meas- ured factors (snow depth, snow density or water equivalent, air tempera- ture, degree-days) and frost depth. Systems which have been evolved previously require detailed knowledge of many factors such as volu- metric heat capacity and latent heat of soil, thermal gradients of frozen and unfrozen soil, etc. (Aldrech and Paynter, 1953, Beskow, l9h7). Any system in which readily measured factors could be used would have great utility. The following independent variables were tested with frost depth: 9/ l. Average degree-day— from Nov 1 + average degree-day from Nov 1 sample water equivalent 2. Same as l. but with average degree-day from first day of frost. 3. Same as 1. but degree-days from first snowfall. h. Total degree-days since last sample + total degree-days since last sample change in water equivalent from last sample ° 2/ An average degree-day is the sum or the daily degrees above or below 32 F on a sample date, divided by the total days from Nov 1, or from first day of frost, etc., to the sample date. 89 5. Average daily change in degree-days from last sample - (average change water equivalent) x (average daily change in degree- days since last sample). 6 Average degree-days from first frost ' Average water equivalent from first frost 7. (Average degree-day from first snow) x (average water equiva- lent from first snow). 8. Total degree-days since last sample - (total degree-days . change in water equiva- s1nce laSt sample) X: (VGent since last sample )' 30-day accumulated degree-days + 3-day accum. degree-day values 30-day accumulated snow depths + 3-day accum. snow depth 10 30-day accumulated degree-days + h-day accum. degree-day values ' 30-day accumulated snow depths + 3-day accum. snow depth 11. h-day accumulated degree-days (including sample day). 12. 8-day accumulated degree-days (including sample day). The last four independent variables were tested on the basis that there could be a long-term heat effect and a short-term heat effect re- sulting in the frost depth on any given day. As can be readily realized, an indeterminable number of long- and short-day combinations could be used. The eight-day accumulated degree- day value proved significant in relation to frost depth for the Clinton County plots. All other combinations could not be related. This variable did not correlate with frost in the Kent County plots, pri- marily because of the greater amounts of snow on these plots. As stated earlier, degree-day values are positive if the accumu- lated temperatures were below 32 F (indicating freezing opportunity) and negative if above 32 F (indicating thawing opportunity). Frost depths for each sample day in Clinton County are plotted 90 against the eight degree-day sums in Figure 18. Statistics for the regressions of degree-days on frost depth are: Standard No. of Regression Correlation Deviation Plot Samples Coefficient Coefficient y on x Bare 19 Y = b.7700 + 0.0511X 0.66 2.18 Herbaceous 19 Y = 1.2210 + 0.0236X 0.72 0.93 Hardwood 19 Y = 0.2102 + 0.0068X 0.67 0.03 These regressions were developed only for the time when snow was on the ground and were all highly significant. 0n the bare plot, samples 18 and 21 (number in symbols in Figure 18) were not included in the regressions. A frost depth of 15 in. was measured but may have been deeper, and so did not truly represent the actual depth of frozen ground. Vegetation-snow interactions are reflected in slopes of the regression lines. The slope for the bare plot was 2.2 times that for the herbaceous plot, and that for the herbaceous plot 3.5 times that for the hardwood plot. In other words, in reducing frost depth the combination of snow and vegetation (herbaceous plot) was more ef- fective than snow alone (bare plot), and the combination of snow, vegetation, and litter (hardwood plot) was most effective. The variation in frost depths (standard deviation of y on x) was greater for the bare than the herbaceous plot, and greater for the herbaceous than the hardwood.plot. This is paralleled by cor- responding variation in snow depths observed on the bare plot. The relationships for Kent County were not so clear as for Clinton County (Figure 17). However, the hardwood plot always had less frost than either of the other two plots on every sample day. Here again mpoam zydeco poem «Ham hmduomhwme ..Ewfio mamnocr named pmofim .5” opswwm 91 mr<0 .. wmmowo SBHDNI NI HLdSO .LSOUJ 0.0.. o: 00. ca ow on em 0. o o... ow: a a e n. . _ @ Vi. NV V ex fie t mun O. . d i. he . e. I I gr \\ V.” a. me a two .an .l. : x. a 2 m e! E 2 r].— .4 ol.1 m. \,N\ .N OK mg m a. \ ‘4‘ l|.(®’.v||‘|@¢ 1.“.b.“ #1 I'lf || A n ONx w meld. m. 1 k.“ 0 T ‘JW . l' I.) N6 'l LT @ d 59 E J 1|... 7 1... w .1; Iii!) - IJP.T.- , .11 20.22330 59.... 51.“. 20¢“. .28 02.4.3.3 ® Educ 3ozm are 30392: a 38365: O mm Synod pmomh .ma ohswwm m>¢mmoO Fmom... PmEu _ _ 2°C.... »(0 OZ_JQI(m Itwo 302m ‘8 11 .17-. 000265: 0 N. e. mDOwUonm mmhwoo «hymdnSOQ GmNonmqsudwNomh pmonw hoasumaw was opflpowadpm xoflgp-.qfi-ma\m mgmhoa m and a an pmone pesosmnon .nfl gm>oo w\m-H op mH\m seem nflgp Magoo Mo QUE P::',4-“.'-' 1:32; -1 . .ma masmfim lOO poam opdm hpasoo nopnflao Kmnoo cwaonm mmmdn.nflud\aum yo .cfl :\HIH hwzoq .om mhsmflm noflpwhmmom Hflom yo omdmo mehm mwfiposadpm mafi :mp manmponm mfl dnflnwn pmonm mpohonoo ncfiondw dam definmn pmonm gpfik Adouoaoo pnwflav mehm opflpomadpm mpmnoqoo was momma Madam mama 66H xofinp'.nfl-m\a afiosmm 101 Light, opaque areas are granular frost Dark, irregular areas are open spaces in soil Horizontally oriented dark lines (< 1/16 in. thick) are ice lenses Note: Area in rectangle has beginnings of stalactite frost. Figure 21. Frozen core, Kent County Bare plot 102 Pencil points to root in cross section Porous channels with stalactite and concrete frost. Fine lines in core paralleling these channels are ice lenses Root in cross section Figure 22. Frozen core, Clinton County Bare plot 103 < 1/64 in. ice lens Examples of small horizontal ice lenses throughout core giving laminated appearance to core Examples of open pores (dark areas) in core Figure 23. Frozen core, Clinton County Bare plot 104 Area of stalac- tite frost (below) grading into concrete frost above Dark areas are open channels in soil Ice coat around root. Open Area of concrete- Solid ice over areas surround root-ice area granular-stalactite core frost in root zone; cause of much soil separation Figure 24. 2-l/2-in. frozen core, Kent County Herbaceous plot 3/8— to l/2-in.— thick ice lens with stems and leaves within. Lighter areas within are granular-concrete Open pore above root Pencil points to ice lens across core Examples of ice lenses (< 1/64 in.) through core Open pore below root. This and other rounded areas are open Figure 25. 3-1/2—in. frozen core, Clinton County Herbaceous plot 106 poam msomomnpmm hpnsoo puma nwhoa mo Eoppon ammonw .cfilw\HnH .mm masmwm poom mo pnmflh Hozoa cflnpfls.pmoaw oponocoo Op swam Como onpmsosoo1Mmadcmhmv mmflzflpwcoa upmadcmhm 080m Sufi: mampmhpo 00H Sufi: pmanoahfiw poom dopw>om poom r.poos BOHmQ moan Como o‘up-ru..quu .. . \ .. . .. .n........... a. at. a \. a. a... a .1. 0900 Op mondhmmm law dmpdqflawa m>flm mwmnma mmnwa 00H Op mpzflom Hwoqmm 107 jn’fi‘griiia‘r,’ Granular and granular- honeycomb frost below litter, and in Al horizon Examples of granular and granular-honeycomb frost around pores and root channels Solid frozen mineral soil; no evidence of ice lenses Note: General admixture of porous, crumblike soil into solid concrete frost. Figure 27. 3-1/4-in. frozen core, Kent County Hardwood plot C C Q 3.; C hr I] u #UCLL .Ja . H; Li. Tar 108 poam doospnnm hpndoo sopnfiao «choc mo coflpnom nomad nonopm .mm onswfim mpoam Has Ga momsma 00H A.zfi w\H AV MUHAP nonpo was mfinp nfl endow unspodhpm Ampflpodampmv “00H anode wo mowpomm macadaoo mo Cowpmowmsm psfiwm Apamfiav mom was mama nwmzpwp “Apgwav mnOHpHom mom dHHom nmwspmn mumOhm mpfipuwadpm pmoym nEOohmconuhdeqdnw Mo mdmh< mmhoo dooapnmm handoo nOPGHHo 03p no stop mdfixooq .mm ohsmflh wasp yo coflppom mmopSO op mpnflom aflonmm .mofi mo wasps: manhom mpoz .mehM neoohononnhwadumgw dmmsm wnflkoMm o0>oamh poked Hmppfla mo HH< nsonm 00H neoozmsonuhwadnmhw dam mo>m0H Hm>o 00H Human mo .QH m\H 0p ms man ohoo .nohda magnfl opnfl oofl amazcdmm wnflSOQm «60>oema moppfia wo pmoz 110 the jeep. As soon as surface melting began, the core could be ex- truded into the cutting board. This technique was used extensively from.the middle of January to the last week.in.March. After March 28 the manual Hvorslev sampler was adequate to remove the frozen cores. HOwever, it was often necessary to hammer the sampler into the frozen ground. Porosity of Frozen Cores and Frost Type Porosity. The literature implies that concrete frost is imperme- able. However, in the concrete frost zone of almost every core, few to many open pores were observed. These varied in size from.barely visible up to about l/2 in. in diameter. Many study cores were broken Open to scrutinize these pores. It was concluded that most cores were porous and that there was a reasonable possibility that moisture movement could occur through concrete frost. Frost type. Each photograph is annotated with specific comments on frost type. Frequently several frost types occurred in close as- sociation in the cores. This was particularly noticeable in the case of concrete and granular types. In some cases granular cleavage lines were noted in areas of "solid” ice; in these instances the frost is termed "granular-concrete," Figures 25 and 26. There were evidences of columnar structure in some areas of solid ice, and in some cores, stalactite frost was noted contiguous to con- crete frost, Figure 20. A combined term was not given to these frost types, but there seems to be ample evidence that concrete frost may develop from stalactite. Concrete frost predominated in the Clinton County Bare plot. 111 any horizontal ice lenses of varying length and up to 1/32 in. wide were present (Figure 22). Frequently solid ice up to 1/2 in. thick was found across cores between 3 to 7 in. below the surface. Vegetative material was sometimes included in the ice band (Figure 20). On the Kent County Bare plot ice lenses were generally thicker, to about 3/16 in. No ice bands were found (Figure 21). However, granular frost mixed with concrete was frequently found in the upper 1-1/2 in. of the cores. In the herbaceous plots concrete frost was general from 1 in. below the soil downward. A definite line separated this from the more porous frost above (Figures 24 and 25). Frost in the upper portion was a mixture of granular and honeycomb with some concrete frost. Open pores were noted in the lower concrete zone as in the bare plots. The upper porous zone generally coincided with the dense root mass, but the few roots in the concrete zone (Figure 26) were often sur- rounded by granular-concrete, granular, or concrete ice with contiguous open spaces. As with the bare plots, ice lenses were thinner in Clinton than in Kent County. Ice lenses in concrete frost may be important from the standpoint of percolation observed during freezing (page 141). Impermeable when frozen, these lenses might melt during periods of rising temperature allowing meltwater from the snOWpack, as well as from the lenses them- selves, to percolate to lower horizons. Under undisturbed snow on the hardwood plots, freezing was not sufficient to produce a long frozen core. Hence pictures of frozen cores in the hardwood plots are of cores taken where the snow has been trampled, and the soil was frozen more deeply. ll2 Frozen soil of the crumb type was found in the upper 1/2 to l in. of the hardwood cores. This layer generally contained granular frost, and often some honeycomb and stalactite frost (Figures 27 and 28). Cores from the hardwood plots were more porous than cores from the herbaceous or bare plots. In hardwood cores the lower concrete zone was not distinctly divided from the upper as in the herbaceous cores. There was a greater admixture of porous, crumblike soil grading and intruding into the lower concrete zone (Figure 27). The concrete frost zone itself was more porous in cores from the wooded cores than elsewhere. There was a geater root mass in this zone, and the area around visible roots, and wirm channels were lined with granular-concrete, concrete, or granular frost. Two pictures have been included to show the frozen litter that could be found at almost any time during freezing periods in the hard- wood.plots. Frost was of all types and combinations of types. Pure ice frequently held this mixture together, though most often the frost was granular or granular-concrete (Figure 29). Soil.Moisture and Bulk Density There was a tendency toward an increase in moisture with increasing frost depth and a concurrent decrease in bulk density, Figures 32 through 37. As a preliminary step, the possibility that significant 10 differences occurred from one sample day to the next was investigatedu—/ y This analysis is on file (VB34.2) with the Trafficability Section of the Army Mbbility Research Center, U. S. Army Engineer Water- ways EXperiment Station, Vicksburg, Mississippi. 113 Comparisons made from one sample day to the next proved incon- clusive and contradictory; increases in moisture during the freezing period could not be consistently explained by moisture movement during freezing, by associated changes in bulk density, or even by a shift from the frozen to the nonfrozen state. The comparisons did indicate that: 1. Significant day to day changes in bulk density and moisture could not be accounted for because of the relatively long interval (2 to 6 days) between samples. 2. Inherent variations in bulk density frequently masked daily changes in both moisture and bulk density. 3. Some soil moisture increases were attributable to losses of water from the snowpack. ’ h. A rise in air temperature to near or above freezing often preceded increase in soil moisture. Likewise examination of the air temperature regime (i.e., were air temperatures directly related to freezing or thawing, to moisture release, or density changes?) provided no logical and consistent ex- planation for the changes in moisture and bulk density. The analysis that follows developed from these preliminary in- vestigations. A.general.picture of moisture movement was obtained from the entire upper 15 in. of the soil and relationships of’moisture and bulk density with frost depth were explored for each 3-in. layer. Meaningful comparison of moisture and bulk density with frost depth required determination of the inherent variation in bulk density and in water holding capacity. Inherent variation in maximum moisture content determined when the soil was not frozen. The variation in 114 bulk density that might be expected to occur in the nonfrozen soil was determined from samples obtained prior to the first freeze. Direct relationships between moisture and bulk density were also investigated; as was the possibility that snow meltwaters con- tributed to moisture in the frozen soil. Moisture Movement in the 15-in. Depth Kent County plots. Moisture contents were shown in Figure 30. Starting with initial.moisture values of 4.68, 4.13, and 4.77 in. of water per 15 in. of soil, respectively, for the bare, herbaceous, and hardwood plots, moisture remained relatively constant during the study period. From the first freeze until the end of the freezing period, March 30, the variation from day to day was generally less than 0.5 in. of water. Standard deviations of 0.36, 0.23, and 0.21 in. of moisture per 15 in. of soil for the bare, herbaceous, and wooded plots respec- tively, substantiate the graphic evidence (Table 16). Table 16. Mean moisture contents, standard deviations, and coeffi- cients of variation for the upper 15 in. of soil Standard Coefficient No. Mean Deviation of Variation Plot Samples in. in. % Kent Bare 25 4.71 0.36 7.60 Herbaceous 24 4.25 0.23 5.40 Hardwood 24 4.73 0.21 4.40 Clinton Bare 19 4.97 0.53 10.70 Herbaceous 23 4.86 0.44 9.10 Hardwood 22 5.58 0.33 5.90 Moisture contents on the first and last day of freezing with 115 @3300 pnmm «flow wo .nH mQéfl «pnopnoo 0.25.308 9mm: .om magma“ ”3(> uchmai 20:8 PnOKL hW(I an? mg mm... 2(fi Umo >02 OOOIO¢ 33.5.0; 20 .8 III EOE .523 oz< 5x: 0 oz.|.J.u3 82 .3. ix F2 «(.2 ~ 3.1 25. 03 >02 M D H 3 S d 3 a m. _ M s D 1 118 Bare Herbaceous Hardwood December 14 4.78 4.58 5.27 Last day of freezing 3.47 4.15 6.05 (April 3) (March 31) (March 25) Field maximum 5.14 4.79 4.58 Gain in moisture by the hardwood plot in Clinton County may be ex- plained.by its poor drainage resulting (in many sectors) from an im- pending horizon about 20 in. below the soil surface. The herbaceous and hardwood plots had been covered with an ap- preciable snow mantle in each county. 0n the Clinton County Bare plot which had very little snow (none on March 29), moisture loss was ap- preciable. In Kent County, the bare plot had 1.94 in. of snow on the last day of freezing and moisture loss was less than in the Clinton County Bare plot on the last day of freezing. Again, in Clinton County, linear relationships between moisture and frost depth were tested without Significant results. Relation of Soil Moisture Con- tent to Frost Depth and Pore Space Soil moisture plotted with frost depth, Figures 32 through 37, indicated that in some plots there was a general increase in moisture in the frozen layers as freezing progressed (see Appendix A). Frost depth also tended to increase during the freezing season. Kent County plots. Mbisture content, in inches per three inches of soil, was plotted with frost depth (Figures 32 to 34). On all three plots, moisture contents remained within the range of 0.9 to 1.5 in. of moisture per 3 in. of soil regardless of frost depth (Figures 32 to 34). poam 0.8mm 43560 920M «mgpmod .cflum an. mpcopcoo 935308 coo: .mm oaswfim 119 use}; 2305 0 use)!” 23052: O mus.» #6.. 450... I'll OZuouJ 000. 3.: >34 «.2 «(1 on... 2.}. uuo >Oz 2.. 0. on o. .n on 0. ON ON 9 .n ow o. .n ON 0. on om 9nd Ihnwo .Z_I0. OF IN. a . . a 4 Puma .Z_la OP IO {4111,1147 I»- l-. 7 a. |I.. .. .. 1 to ad m.— 0.0 Ikamo .Z_In OF IO . - l 7..-, 3 ad ..— n; n.— "IIOS 'Nl-E/SBHDNI NI BUDLSIOW poam msooomoraom 55.550 p23. «mnpmop .GHnm ho. mpcmpsoo 0.35308 Goo... .mm magmas. 120 u4m2(1 ¢1( ((3 Gun 2(5 Uuo >02 ON 0. On ON O. _n ON O. ON ON O. .n ON 0. .n ON 0. On ON 0. . Ihnwo .2.Io. OF IN. .Z.IN. 0... IO .2. IO 0... lo .270 O... In .Z.Ifl OP IO 'IIOS 'Nl-E/S3H3NI NI BUOLSIOW 121 poam doogpamm aOCSOO paom smflpmoo .QHIm an mpsopnoo endpmfloa smog .sm enemas ON >(1 O. on ma( ON 0. .n. e vs \m.< HNU‘QH wQOl 0850.» ON ((1 0. ON ON cum 0. qu}(n ZuNOmu 0 qu3(0 ZuNOKuZD 0 U0(ln w¢0m ..(FOF Illll OZwOm... 00o. Z(... Umo >02 ON O. _m ON 0. on ON 0. 0.0 1..le .270. OF IN. .Z.IN. 0». IO .Z.IO 0... IO 0.0 .27” 0... IO 5.0 O6 WIOS 'Nl-E/S3H'JNI NI BUHLSIOW 122 On the herbaceous and hardwood plots all moisture contents were between the 60-in. tension value and the total pore Space of the soil before freezing. Moisture contents during freezing in the bare plot showed a wide range, some falling both above and below total pore space and 60-cm values, Figure 35, Table 17. Clinton County plots. Moisture was again plotted with frost depth (Figures 35 through 37). The bare plot data are presented for all five depths, as freezing occurred in all. Freezing occurred in the herbaceous plot in the surface to 3-in. and 3- to 6-in. depths and hardwood plot data in the surface depth only. Moisture was significantly related to frost depth only in the surface to 3—in. depth of the bare plot. The regression equation, Y = 0.0322X + 0.14, where Y is moisture content and X is frost depth, evidenced a significant correlation coefficient of 0.470. Table 17. 60-cm tension values and total pore space before freezing period Depth 60-cm Tension Total Pore 60-cm.Tension Total Pore Kent County Clinton County Plot in. Values, in. Space, in. Values, in. Space, in. Bare 0—3 1.03 1.22 1.09 1.20 3—6 0.93 1.15 0.98 1.16 6—9 0.93 1.04 0.99 1.05 9-12 0.90 1.00 0.99 1.05 12-15 0.93 1.12 1.09 1.02 Herbaceous 0-3 0.83 1.36 0.83 1.25 3-6 0.94 1.11 0.98 1.18 6-9 0.86 1.18 0.86 1.10 9-12 0.91 1.11 1.01 1.19 12-15 0.87 0.96 0.97 1.07 Hardwood 0-3 0.87 1.87 0.67 1.73 3—6 0.80 1.66 0.84 1.57 6-9 0.89 1.31 1.00 1.28 9-12 0.83 1.28 1.03 1.28 12-15 0.97 1.06 1.04 1.17 poaou whom .5560 sowswfio smflpmop $3.7m an. mpcmpqoo 0.25308 new: .mm mhsmflm 123 and)!» 2353 o and}; 23052: o 36... #6.. agoe III 950m; 03. one. :3 ma... «(I own 21:. one >oz om 0. on ow o. .n o~ o. 3 o~ o. _n o~ o. _n o~ o. on om IPQNO .Z.In. 0P IEUO .Z.IN. OP Ihluo .Z.IO OP IleO .2. IO 0». thuuo .270. 0... WIOS‘ NI-Q/SSHDNI NI aanlsnow 124 poam mdomompamm 43350 sopcflao «managed .cflnm he. mpampqoo magpmwoa coo: .mm madman. mama}. 2:05 o whiz}. $3523 0 uooz om 9 on cm 2 _n o~ 0. am o~ o. 3. ON 9 .n o~ 0. on ow o. I ma .Z.Ifl. OF IN. IENO .Z.IN. 0... I0 Ituo .270 0... Io Ituo .Z.IO 0.? In V0 .0 'IIOS 'NI-C/S3H3NI NI 3UI'ILSION I I 125 poam doosoamm 43.8500 GOpGHHo smashed .sflnm 59 3809800 0.85308 8802 ..Lm 0.3mm”. U403(l It( 5: Out 2(5 _ ONO >02 ON O. on ON O. .0 ON O. ON ON 0. .n ON O. .0 ON O. on ON .270. OF .: \n .s IINUVQW Ngfi QVROK as .\ INU‘QM. Nat 'IIOS 'NI-C/S3H3NI NI 3801.510” In t} total porI EL layer deeper the lower the; On 6 Space (ca increased I moisture Exax 37, revea a decrea; This Ole; soil def‘ 0n that of Other th and Pea} Moj the free 126 In the Clinton County plots moisture during freezing exceeded the total pore space value except in the hardwood plot and the 12- to 15- in. layer of the bare plot. In these two exceptions the frost was never deeper than the layer for which the moisture contents were plotted; i.e. lower than 3 in. in the hardwood plot, or 15 in. in the bare plot. On 6 of the 12 days when the moisture contents exceeded total pore space (calculated before the soil was frozen) the entire 15-in. layer increased in moisture content. On three of these six days the lS-in. moisture content had not changed from the last sampling date. Examination of the moisture patterns on each plot, Figures 32 to 37, revealed that when the moisture content in a given depth increased, a. decrease in moisture did not necessarily occur in the depths below. This clearly infers that moisture need not have moved up from the lower soil depths to cause the increase. On certain dates the moisture content in each depth increased over that of the previous sampling, again suggesting a source of moisture other than upward movement. The source of this moisture may have been the snowpack. This possibility is considered in the section "Snowfall and Peak Moisture Contents ," page 140. Moisture content varies for a given frost depth. For example, when the frost line was about 1.2 in. deep in the Kent County Bare plot, moisture contents of 0.92, 0.96, 1.07, and 1.28 in. were recorded. (These occurred on December 10, January 7 and 30, and March 10.) In every plot more than one moisture. content per frost depth was noted. Discussion. There are ample indications of moisture movement into the frozen layer, a phenomenon previously noted by other workers (e. g. DOmby and Kohnke, 1955). Moisture movement due to capillarity (N A '4 tu tel mo 12‘ TC 1r. ‘2 k. 127 undoubtedly takes place; the presence of ice induces very high tension gradients between it and the unfrozen soil. Though there is much disagreement as to the quantities of water that can be moved by vapor transfer, such movement must also be consid- ered as a potential source. No published work on direct experimenta- tion with moisture movement from.unfrozen to frozen soil layers was found. Several investigators, notably Bouyoucos (1915) and Smith (1943), have reported.moisture movement in the vapor phase when tempera- ture gradients were established in unfrozen soil. With the extreme temperature differentials between frozen soil and unfrozen soil, vapor movement into the frozen zone would be expected. Vapor movement, plus water added by capillary rise might well explain some of the higher moisture content. Water movement in the liquid or gaseous phase may also explain why moisture contents could not be related, linearly, to frost depth. Only moisture present as ice influences frost depth as observed in the field. That is, the frozen core determined visually depends on frozen soil moisture, while liquid water or water vapor may or may not be present in this core. Bulk Density Versus Frost Depth and Limit of Bulk Density Variation In testing relationships between bulk density and frost depth, consideration was limited to depths with frequent freezing. These data are plotted in Figures 38, 39, and 40 (see Appendix Table A6 for bulk densities of each depth). Confidence intervals were developed for bulk density based on the number of unfrozen cores obtained in each depth to the time of first hpnsoo pamv. «Aflpmod 8.7m Op 000..”ng mppmofio meson.“ 8H hpwmaoo Mash. 9mm: .wm shamans. 128 Jm>w4 U023...ZOU 1N0... II I I 3136 23.9... o 3...?» 23052: 0 020004 08. coo. >3. ¢c< .21 3.. 21. one >oz _n .0 15.18 .Z.In IO 0.0 5.0 0.0 Ituo .Z.In 05. IO some ‘Allswao x108 1:11:11147144:.ij: .... 129 poem whom .358 8830 68.8% .510. .3 banana 8.8. not: .8 2:88 g INN? 0 303 SN: 0 JU>NJ 002830 (a; I'll .Z.IN. OFIO 2......00 .270 O... IFQUO .270 Oh 03".) Nl AMSNN NW 130 30.3 dookdndm and mSomompnmm 59550 GOPGHHU «mnpmmd smack,“ 5. hpfimcww Mia, «80: .0: mhswwm JU>UJ uguezu (U304 I'll all; ZUNOCu Ufa!“ 5N2; §UJ .0 a. as. g 3 E 2‘. 08 >9 E8 .176 Oh IO Ihtuo .270 In ‘2‘“. -. O v~ m .‘ .0 DD/M 'AJJSNBO ”1'19 E8 .270 0° freezing. confidence freezing. test (Tab: Inte: bulk dens: County Ha; intercept with the ' tions fro no TEIati The dEpth and Canny pl Plot, no: level: 8.1“. In a linear re (Figure L determine lEVel) Fj D'c {E bulk denc mOi sture time dur 8011 moi 131 freezing. Bulk densities during freezing which were below the lower confidence level were assumed to be a result of moisture movement and freezing. The 99% confidence level was used to provide a stringent test (Table 18). Intercept values of the regression equations generally reflect bulk density values of the several soil depths, Table 19. The Kent County Hardwood plot, which had the lowest bulk density, had the lowest intercept value; while the Clinton County Bare plot, 9- to 12—in. depth with the highest mean bulk density, had the highest intercept. Devia- tions from the regression varied from O.lh g/cc to 0.10 g/cc and showed no relation to depth, vegetation, or mean bulk density. The hardwood plots showed no significant relation between frost depth and bulk density. There were only 10 frozen samples in the Kent County plot, and 9 in the Clinton County plot. In the Clinton County plot, none of the bulk density data fell below the lower confidence level, and only four points did so in the Kent County plot. In all other frozen depths a significant or highly significant linear regression was developed between bulk density and frost depth (Figure #1). In all but two depths where significant regressions were determined, a,majority of the points were below the lower confidence level, Figures 38, 39, and ho. Discussion. Results of the bulk density-frost analysis suggest that bulk density of specific soils can be predicted from frost depth. If moisture contents could be determined from.bulk density, a useful pro- cedure might be evolved for estimating soil moisture contents at any time during the winter period, and.a prognosis might be made as to the soil.moisture status during thaw. 132 Table 18. Means, standard deviations, and confidence intervals for bulk density before freezing* . Bulk Density, g/cc Depth No. of __ Standard Confidence in. Samples X Deviation Interval (99 %) Kent Bare 0-3 2h 1.62 0.11 1.31 - 1.93 3-6 32 1.59 0.1M 1.20 - 1.98 6-9 87 1.65 0.30 0.82 - 2.h8 9-12 167 1.73 0.10 1.h6 - 1.99 12-15 133 1.75 0.05 1.62 - 1.88 Kent Herbaceous 0-3 12 1.52 0.09 1.25 - 1.78 3-6 23 1.69 0.01 1.06 - 2.33 6-9 132 1.73 0.01 1.70 - 1.77 9-12 132 1.77 0.01 1.73 - 1.81 12-15 128 1.81 0.002 1.80 - 1.82 Kent Hardwood 0—3 12 0.96 0.07 0.73 - 1.18 3-6 132 1.38 0.01 1.31 — 1.u2 6-9 129 1.57 0.007 1.55 - 1.59 9-12 129 1.68 0.02 1.63 - 1.72 12—15 121 1.76 0.01 1.73 - 1.79 Clinton Bare 0-3 26 1.61 0.00 1.35 - 1.87 3-6 10 1.71 0.11 1.36 - 2.06 6-9 10 1.82 0.06 1.6h - 2.00 9-12 35 1.79 0.003 1.78 - 1.79 12-15 M3 1.75 0.09 1.50 - 2.01 Clinton Herbaceous 0-3 12 1.5u 0.09 1.29 - 1.80 3-6 22 1.61 0.10 1.31 - 1.90 6-9 38 1.70 0.17 1.24 - 2.16 9-12 35 1.77 0.18 1.28 - 2.25 12-15 126 1.75 0.08 1.5M - 1.97 Clinton Hardwood 0-3 12 1.08 0.18 0.52 — 1.66 3-6 131 1.38 0.02 1.33 - 1.h3 6—9 130 1.61 0.01 1J58 - 1.6h 9-12 128 1.65 0.01 1.62 - 1.68 12-15 116 1.6u 0.009 1.62 - 1.67 * Based on number of frozen cores before freezing. 133 .pcmoflMficmfim 902 + .H®>®H Ram DUN PQwOflmHHCWHm ** .Hm>ma $mm pd paw0fi%flawflm * NH :m.o ma.o + comm.o: xwa.o :O.H M muo @003Upmm NH ::.H :H.o * omow.o: xmo.o H~.H w mum mSOmompymm om mm . o S . o t. 0me . o- fie. o om . H w m..o msomomfimm ma mm 4 ma . o ** 8m» . 0. $0. 0 :0. m w NTm 6.8m ma mm . a i . o * 8mm . o- 56. o we . a up To 26m 9H mmé 2.0 * 88.0.. xmod 34 x 9m 83 ma NH.H ma.o ** ooww.o: x:o.o m:.H w mno mnmm hpczoo copsfiao m mm.o mo.o + mmma.01 xoa.o mm.o w muo voozuamm om mm.H oa.o ** Omm~.01 xma.o m:.H w mso m:omomenmm mm wH.H mH.o * OmOm.On xmo.o OJ.H M muo chem hpQSOO psmx mmamedm AMV cOflmmmhwmm evacuauficmfim pcmwofimmmoo coapmawm .cw poam mo .02 zpflmcwm scam mo Hm>mq cofipmamhnoo nvmmm Masm new: soapmfi>mm cnwusmpm CS 5me pmoum a AHV hpflmcod £39 .wo mcoflmmmhmmh 93 MPH moepmflpmpm .9” manna Show pmoum mSmnmb 23628.0 “4.5.5. Mo mQOHmmmhwmm A: mhsmflm 131+ IPQUO FmOm... m> >._._me0 XJDm uO mZO_mmwmomm ”UIOZ. Z. Ihlmo Fmozu 0. Q. N. 0. O o V I. x 20 > do 20.533 o¢ _ _ {.2 OZMOuJ 0 H [[1 2.5....8 2:16 o». -o _ mar—02. Z. IPtUO Pugh 0 N O. V N m: 0. O O Q .. O 2E3 .27o OJ um [I - \er *[ [j a T I “Qf . \fl * III _ 502 2.5.3 .270 02 -m _ 69— 6.2.x _ _ _ _ _ u \\.‘l*\ 1,! _ . 3.! [I r T 031:; M8 “Haydn! W fl .2 2...an .24» oJuo xtwo .zqn 0H6 _ -_.2 1:3 .270 B IO ..PZM! 389652 29.230 wm.H 0H.H 01m me.a mo.a 1:.oI RH.H H0.H Hm.H mIo beem zpqsoo copnwao HN.H 0H.H >:.0I m0.H mm.H ma.H mlo muomomphmm N. 0m.H ma.H 03.0: mm.H m0.a mm.H mno chem 292560 babe .OH OO\w OO\m mommm OO\m .OH .OH poam mpspmaoz ONSpmHoz om>pmmpo OONORMOD_SORM Omom Hmpoa zpflmcmm mommm gumma cw>ummno asaaxmz Essflxmz pm hpfimcmn hpamcmm Masm Ommopm Rmpw< Masm whom Amv masm dmpwazoamo ca OOCOROMMHQ Rpfimqmm Mazm m>< Hmpoe 3 3 80.3836 80 3 «m0 pnmpnoo endpmfloandm>hmmpo azaflxma pm 020 Amommm whom propv ROHmeSpmm pm mnflwmmnm epH3_thmnmd Mann RH mmnm£o Mo GORpmaHpmm .Hm edema 139 Frozen bulk density in the surface to 3-in. layer of the herbaceous plots was calculated to be 1.05 g/cc in Kent County and 1.06 g/cc in Clinton County. Before freezing bulk densities were 1.52 and 1.54, respectively, for the two plots. The greater reduction in bulk density of the surface depths of the herbaceous plots than that of the bare plots reflects the greater in- fluence of honeycomb and granular frost, as Opposed to the concrete frost, on soil expansion. Differences in bulk density at saturation both before and after freezing were computed, column A, Table 21. The differences ranged from 0.39 to 0.u9 g/cc. Theoretical bulk densities were computed from the highest moisture contents observed in the field (column 5, Table 21). With one exception, the surface to 3-in. depth of the Kent County Herbaceous plot, the com- puted bulk densities were lower than those which would have occurred at saturation (total pore space) with freezing. This substantiates the fact that bulk density during freezing can and does go below the normal where the ground is thawed. In every case where a significant regression was determined, some of the moisture contents were above the estimated total pore space (Figures 32 and 33, surface to 3-in. depth; Figures 35 and 36, surface to 3- and 3— to 6-in. depths; and Figure 37, surface to 3- and 3- to 6-in. depths). In most cases the values fell above the estimated field maximum moisture content (60-cm moisture tension value). In those depths where moisture was not related to bulk density, moisture contents were near or below the field maximum moisture content, or bulk densities were within the range of variation. In all the 1H0 depths in which significant moisture-bulk density relations were evolved, except the 9- to 12—in. depth of the Clinton County Bare plot, all or most of the bulk densities were outside the range of normal variation. Discussion. The relationship of moisture to bulk density brings together the findings on moisture and bulk density versus frost depth. As noted, moisture was not related to frost depth, presumably be- cause all the moisture in a frozen soil depth need not be frozen. How- ever, bulk density changes during freezing were a result of very high moisture contents, which in turn caused an increase in pore space-- moisture freezing to ice caused bulk density reductions. Thus bulk density could be related to frost depth. When regressions of bulk density on moisture were significant some of the moisture contents were above total pore space. This indicates that the normal pore volume was exceeded and hence prior freezing must have resulted in a density reduction. In the 12- to 15-in. depth of the Clinton County Bare plot and in the 3— to 6-in. depth of the Clinton County Herbaceous plot, most mois- ture contents during freezing were below the 60-cm tension level, and significant correlations with bulk density were not found (Figures 35 and 36). In the other plots, where most moisture contents during freezing were above the OO-cm tension level, significant relations were developed. It is possible that moisture in the noncapillary pores froze and expanded causing a reduction in bulk density, and that the total pore space value need not be exceeded. Snowfall and Peak Moisture Contents Soil moisture contents during "peak" days and water available 1M1 to the soil from snowfall and the snOWpack appeared related. "Peak" days were defined as those days when soil moisture content rose above the total pore space value indicated for each plot, Figures 32, 33, 35, and 36- Table 22 lists the days when moisture contents of the surface layers of the bare and herbaceous plots were above total pore space. The sum of the moisture from snowfall during the period from the last sample plus the water in the snOWpack at the sampling preceding the peak day is shown in column 2. Column 3 lists the measured water content of the snowPack on the peak day, and column h the difference between columns 2 and 3, i.e. the water unaccounted for. The last five columns are the cumulative changes in water content to 3, 6, 9, 12, and 15 in. below the soil surface since the last sampling before the peak day. In every instance when water was lost from the snOWpack the mois- ture content of the soil increased. Though the moisture lost from above the soil and that gained by the soil do not necessarily balance, the consistency of the loss-gain relationship may be considered strong evidence that moisture is moving down through the concrete frost. The number of soil cores frozen, out of four possible per sample, was used as an indication of the areal extent of freezing (Table 13). The Clinton County Bare plot appeared frozen solid (concrete frost) on all but two occasions (February 17 and March 25), when peak moistures occurred. The Clinton County Herbaceous plot was also frozen com- pletely on each peak moisture occasion, and the single "peak" moisture listed for the Kent County Herbaceous plot also coincided with com- plete freezing. The Kent County Bare plot was completely frozen only on January 11. 1&2 30.0 m0.0 30.0 00.0 00.0 0m.0 00.3 00.3 002 mm 002 3m 03.0- m3.0- 30.0 00.0 03.0 03.0 33.3 mm.3 002 33 002 3 30.0 00.0 0m.0 03.0 30.0 h>.0 03.0 mm.3 900 0 900 3 00.0 00.0 00.0 30.0 mm.0 3m.0 00.0 33.0 00b om 00h 03 msomo0npmm 300500 0000330 30.0 03.0 03.0 30.0- 33.0 00.0 0m.0 00.0 002 mm 002 3m - 33.0 - 30.0 00.0 00.3 0m.0 30.3 002 3 000 30 00.0- 00.0- 33.0 00.0 03.0 03.0 30.0 00.0 900 NH 90m 03 m0.0 00.0 00.0 3m.0 mm.0 00.0 03.0 00.3 000 0 000 3 pm.0 00.0 00.0 03.0 00.0 mm.0 00.0 00.0 000 0m 000 mm 00.0 00.0 30.0 03.0 00.0 30.3 00.0 30.3 000 0H 000 0 0000 000500 0000330 03.0 03.0 m3.0 33.0 03.0 0N.0 00.3 Hm.m 900 cm £00 03 mzomo0npmm 000300 000M 30.0 00.0 00.0 30.0 03.0 33.0 00.0 Fm.m 00: mm 002 03 03.0 0m.0 mm.0 03.0 03.0 00.0 03.0 mm.m 00: 03 002 m 03.0 00.0 30.0 0m.0 03.0 03.0 m3.m 00.m 90m mm 900 om 03.0 03.0 00.0 03.0 mm.0 00.0 00.0 00.0 000 0m 00h 30 00.0- 00.0 03.0 03.0 00.0 00.0 00.0 00.0 000 33 00h N 03.0 03.0 m3.0 00.0 00.0 00.0 00.0 00.0 000 00 000 3m 0000 000500 000M m3 op 0 m3 op 0 0 op 0 0 00 0 m op 0 .03 «0600 .03 0300 .03 «3600 030000 030000 .03 «00000 -3000 0530 3000 00 Izosm 0530 M000 0003 .03 «3300 03 00500302 03 000030 3000 8030 30093000 33003000 00000 0000 «000: 00 90003 03 00003 330 “m0 300 A30 00000000 00500308 3000 mo 500 00 000 030800 mdofl>0nm 8000 0000030 00300300 3300 3033 «3300 on 0390330>0 000 «30093000 «33003000 :3 00003 00 mmnocH .mm 0390B 133 Of equal importance is the decline in soil moisture between the peak days. On those plots where frost was of the concrete type, the bare and herbaceous plots, the obvious inference is that moisture is moving through the frozen soil. That this is possible was pointed out in reference to frozen cores in which it was evident that concrete frost was not necessarily nonporous. Soil Shear Strength Before, During, and After Soil Freezing Soil shear strength was measured with the cone penetrometer in pounds per square inch (Appendix B). To make a measurement, the penetrometer was pushed to the t0p of each depth (i.e. O, 3, 6, 9, and 12 in. below the soil surface), and the force required to push the cone past that level was recorded as shear strength. Concurrent moisture contents for the surface reading are those occurring in the surface to 3-in. depth; for the 3-in. strength reading, those in the 3- t0 6-in. depth; etc. The range of the penetrometer was 0 to 300 psi. Readings of 300 may be over 300 psi, and are lumped as 300 psi or over. Readings were taken whenever time permitted and where the penetrometer could be pushed into the ground with reasonable effort. The degree of freezing determined whether or not readings were obtained. Strength readings for each 3—in. depth are plotted in Figures #3 through #8. On the Kent County plots and the Clinton County Hard- wood plot readings were relatively frequent. However, in the Clinton County Bare and Herbaceous plots, heavy frost prevented cone readings from January to the end of March. 00333 0300 03003300 000M «030000.300 300330 0002 .03 0.30-033.3 11M w4t2<fl ZmNocu O u4l2<fl ZUNOCKZD O OZNONJ 000. 600. >(I JEl< ¢(1 0mm 2(0 ONO >02 O. on ON 0. _n ON 0. ON ON O. _n ON 0. .m ON 0. on ON O. OO. OON oon OO. OON OOn OOQ lSd'HL9N3ULS avaus 115 0039 050000000m_%00500 000M «000000300 30000 0002 .33 000030 >(1 O _ on C: O. Ono. 1(1 nun 2(3- u4t3<fl Zu Nan O u40202 ON 0. on ON O. OO. OON Don thauo .2_I _ 00¢ 00. OON Ikmuo .Z.Ia 00‘ OO— OON OOn IPn NO .270 00' OO— OON con 00' OO— OON con 00' lSd ‘Hlemaus uvaus 1A6 90%. 009.690.. 59560 PQ0M «mnpwnwsm 90030 8002 .m: masmflm Ono. >(Z J.¢&( 1(3. mun. 2(fi 0_ On ON 0. .n ON O. oN ON O. .m ON 0. .n ON m4¢102 O. On ON 0. Ikmmo .Z.IN. Ikumo .Z_ I o IPluo .2. IO 5&3 ..zT :Fl 00 .2. I OO. OON OOM 00¢ 00 . OON COM 00' OO. OON OOH OO. OON com 00! OO. OON OO 0 use ‘Hmuams avaus pod... 080m 5.9980 GOPSHHU «mapwuaoh 800nm G002 .9. 0.39% 11+? IPauo ZUNOCK 0 Its ZUNOCbZD 0 an... 0:. 000. ><1 ¢&( 3 Gun 15: UUO >02 IPtUO .Z.IN. IPIUO .Z.Ia tPtUO .2. IO ISd NI HDN3ULS EVBHS lh8 poam 050000932 29:80 sepuflao «mflpwcgpm 90020 G002 ..L: 0.39%. >(2 ¢&( 000. «3‘! nu... Iktuo ZUNOKL O rhmuo ZuNommza O OZUOUJ an! 2(1 Uuo >02 ZPQUO .Z.IO Ibluo .Z.IO IhtwO .Z.In oo. OON 00m lSd NI HLQNBBLS BVBHS ll+9 poam doozdp0m “$560 Gopzfiao «mnpwcgp0 H0020 :002 .w: 09%.“... than. ZuNOKu 0 1:8 ZUNOEZD C an; OOO. 30. >(} 1....( g nun Z(... UUO >02 ON O. on ON O. .n ON O. ON ON O. .n ON O. .n O. thtuo .Z.IN. IPQNO .2. Io IFQUO .Z.Io OON OOM Ihtuo .Z.In 9.th .Z.IO I‘d NI HJONBULS UV3HS 150 Rush and Kennedyll/ both found significant inverse linear rela- tionships between shear strength (the measured average of readings in the 6-, 9-, and l2-in. depths) and moisture content for every soil textural class listed in the USDA classification. These relationships were inversely pr0portional. In these studies, as in the analysis that follows, moisture content changes were necessarily considered when soil strength changed. Shear Strength Before Freezinglg/ ' Kent County plots. Soil freezing was first encountered on December 21 in the herbaceous and bare plots. Freezing was not encountered in the woods until January 1h (Figures 43 and #5, mean data in Appendix A). Mean soil strengths and corresponding moisture contents are listed in Table 23. Table 23. Mean shear strengths and.moisture contents, averaged for two samplings, before December 21, Kent County Plot Bare Herbaceous Hardwood Depth Strength Moisture Strength. Moisture Strength Moisture in. psi in./3 in. psi in./3 in. psi in./3 in. 0 69.70 1.00 105.00 0.79 16.80 1.03 3 126.h0 0.95 226.30 0.77 uu.30 0.89 6 185.h0 0.93 271.30 0.78 50.60 1.12 9 187.10 0.85 220.00 0.7a 67.50 0.76 12 190.20 0.90 159.60 0.85 82.50 0.83 E In separate office reports to the Army Mobility Research Center, U. S. Army Engineer waterways Experiment Station, Vicksburg, Mississippi. lg/ Frozen ground was not always found at the points of the shear strength observations on the same sample occasion that it may have been found in a moisture-density core. 151 Depth for depth the hardwood plot had the lowest strengths, but not necessarily the highest moisture contents; the herbaceous plot, the highest strengths, and except for the 12—in. depth, the lowest moisture contents. When soil layers are of the same texture, and moisture contents are nearly equal, similar shear strengths would be expected. A comparison of plots in which the depths were alike in moisture content indicated that the hardwood plot always had the lowest strength. This relation- ship has occurred in other studies at Vicksburg, Mississippi. In gen- eral, wooded soils are subject to less use by man and animals, show less compaction, and hence have lower strengths than cultivated soils of the same texture. Clinton County plots. First freezing in this area was recorded December 18 on the bare plot. The bare, herbaceous, and hardwood plots had been sampled on November 10, 11, and 19, respectively. Correspond- ing strength and moisture contents appear in Table 2H. Table 2h. Mean shear strengths and moisture contents in Clinton County before December 18 (averaged for two samplings) . Plot Bare Herbaceous Hardwood Depth Strength Moisture Strength Moisture Strength Moisture in. psi in./3 in. psi in./3 in. psi in./3 in. 0 33.70 1.08 87.50 0.81 h3.70 1.17 3 45.00 1.00 108.70 0.87 108.70 1.08 6 62.50 0.9M 110.00 0.8h 125.00 1.02 9 207.50 0.91 187.50 0.77 158.70 0.91 12 2h3.70 —- 190.00 -- 228.70 -- The inverse moisture-strength relation appeared in all plots with the exception of the surface depth of the herbaceous plot. On December 31 all quadrants were thawed. Extended freezing in 152 Clinton County did not begin until January. Moisture contents and shear strength for December 31 are listed in Table 25. Table 25. Mean shear strengths and moisture contents in Clinton County for December 31 Plot Bare Herbaceous Hardwood Depth Strength Moisture Strength Moisture Strength Moisture in. psi in./3 in. psi in./3 in. psi in./3 in. 0 88.70 1.15 31.20 1.00 27.50 1.25 3 72.50 0.95 80.00 0.9h 62.50 1.09 6 85.00 0.88 60.00 0.83 196.20 1.11 9 170.00 0.78 168.70 0.82 1h0.00 0.79 12 223.70 0.90 180.00 0.85 201.20 0.85 In comparison with the readings before December 18, there were generally higher shear strengths associated with lower moisture con- tents. Both strength readings at 0, 3, and 6 in. and moisture contents of the bare plot had increased from the last thaw period (ex- cept for the 6-in. depth which had dropped only 0.05 in. of moisture). The bare plot had been plowed three weeks before the first readings, and was undoubtedly less compact on December 18 than when the December 3lst measurements were made. Normal soil settling plus the fact that the soil was frozen to 9 in. on December 18 may have caused an in- crease in compaction yielding higher strengths. Shear Strength Dur- ing the Freezing Period Frozen and thawed conditions did not necessarily coincide on all three plots in either county. Strength data were examined on the basis of the per cent of the plot which was frozen (five classes were used: 0, 25, 50, 75, or 100% frozen), based on the number of unfrozen 153 samples out of four taken on a sampling day. Kent County plots. Frozen soil occurred from December 31 to March 30. 0f 23 moisture-density samplings 15 strength samplings were made in this period (Table 26). It was assumed that with deep freezing strength readings could not be taken, and occurrence of frozen condi- tions was based on the 23 samples. The surface was unfrozen 13.0, 8.7, and 78.3% of the time on the bare, herbaceous, and hardwood plots, respectively (Table 26). Below the surface layer, strength readings could be made about 50% of the time in the herbaceous and bare plots, and on every sampling in the hardwood plot. Complete freezing occurred 30.5% of the time on the bare plot, 83.6% of the time on the herbaceous plot, and never in the hardwood plot. Generally as the proportion of the frozen area on each plot de- creased the mean strength readings did also (Table 26). Tilled soils (bare plot) and wooded soils are generally less com- pact in the upper layers than pasture and meadow soils with herbaceous cover. This is revealed in the higher strength readings of the herba- ceous plot throughout the frozen season. When 75% of the area was frozen, strength ranged from 233.7 psi for the surface depth in the hardwood plot to 256.5 psi in the 3-in. depth of the herbaceous plot. An equalizing effect of frost on soil strength is evident. Although Kent County plots were similar in texture, moisture in- creased as strength decreased only in the hardwood plot when the soil was unfrozen. In the hardwood plot the effects of frost on soil compaction, .000000000 000800 mm 00 00000 * 0.00H mm.o o.mHH NH o.ooa m®.o w.OOH m 0.00H mm.o m.m® m 0.00H 00.H mw.mm m m.®h NO.H 0.0: b.w NH.H N.mm v.0 NO.H 0.000 m.: 00.0 h.mmm 0 00030900 0.»: 00.0 w.NHN o.ma NN.O w.owm m.: 00.0 o.mwm NH 0.9: 90.0 0.090 o.m9 mm.o 0.Mmm m.: 00.0 0.000 m m.>: 00.0 N.HNN o.m0 00.0 H.NHN m.: 00.0 o.owm w 0.0m 00.0 m.me o.MH m~.o b.0mm m.0 mm.o F.0mm 0.NH N®.o m.mmm m b.0 mm.o 0.0: v.0 00.0 m.::0 m.: mz.a m.mmm 0.0m 00.0 0.0mm m.m: mo.0 00m 0 0000000900 m.v: mm.o P.00H m.: Nw.o o.mbm m.: mh.o o.mbm v.0 Pm.o N.mmm NA 0.9: 00.0 m.mma m.: 09.o 0.090 m.: m0.o 0.090 9.0 00.0 9.000 m m.m: mm.o m.0mH v.0 mp.o 0.0mm m.: 00.0 N.me h.® :m.o 0.000 m H.mm HO.H >.mm v.0 Nb.o w.me o.mH mm.o m.>ma v.0 00.0 m.~:m m o.mH OO.H m.mm :.~H mo.H N.mm 0.NH mo.H N.PNH b.0m MH.H v.000 m.om 0H.H 00m 0 09mm 0 .09 m 900 m .00 m 900 u .00 m 9007 mi .cH m 9007 w .09 m 900 .09 #00009 \.09 00000990 #00009 \.09 09000990 #00009 \.09 09000990 #00009 \.09 09000900 #00009 \.00 09000990 00000 n90000 .90002 :90000 .90902 n90000 .90902 -90000 .9000: :90000 .0090: 0 mm Om m» 000 0000mm‘009< no 0000 900 090000 0000 .000090 0903 9000 no 00000000900 000990> 000: 09009000 09090000 000 009000990 90000 0002 .00 00003 155 aggregation, and consistency undoubtedly influenced the strength rela- tions. This is brought out in the surface depth, where strengths were lower during the freezing period (when no frost was in the ground) than before freezing began, although no large differences in moisture were evident. Clinton County plots. The bare and herbaceous plots show consid- erable contrast to their Kent County counterparts. Because of frozen ground, no measurements were taken on the bare plot from December 31 to March 25 or on the herbaceous plot from January 28 to March 25. However, it was possible to make measurements through the winter on the hardwood plot. Samplings were made January 16 and 23 in the herbaceous plot. Mean shear strengths and their respective moisture contents were: Depth Strength Moisture in. psi in.[3 in. 0 300.00 1.21 3 iou.uo 0.9A 6 72.60 1.09 9 l3h.lO 0.92 12 185.00 0.82 All quadrants were frozen in the surface depth, accounting for the 300—psi readings. In the hardwood plot soil strength changed but little over the winter (Figure #8) and average strength and.moisture content for January 16 and 23 were: Depth Strength Moisture in. psi in./3 in. o h2.50 1.30 3 96.70 1.36 6 103.10 0.95 9 1h3.7o 0.97 12 185.00 0.97 156 Every depth had increased in strength from December 31, with at- tendant moisture increase, except in the 9-in. depth where both moisture and strength decreased appreciably. No explanation was found for these anomalies. The hardwood and herbaceous plots had similar strengths, ex- cept for the frozen surface of the herbaceous plot. Six additional strength samplings were made when no soil was frozen in the woodlot, but the bare and herbaceous plots were frozen. On one of these occasions there was less than 1 in. of frost in one quadrant, March 21; on a second occasion about 2 in. of frost were found in two quadrants, March 25. On these consecutive sampling days strength and moisture readings were: March 21 Reading March 25 Reading Depth Strength Moisture Strength Moisture in. psi in.[3 in. psi in./3 in. 0 107.50 1.31 168.70 1.3A 3 77.50 1.05 70.00 1.30 6 183.70 0.9h 110.00 0.96 9 212.50 0.99 211.20 1.0h 12 212.50 1.05 2h3.70 1.h1 Again moisture strength relations were obscure. However, the in- fluence of one additional frozen quadrant (on March 25) in the surface depth raised the average shear strength from 107.5 to 168.7 psi. Shear Strength After the Freezing Period The rate at which the soil returns to the strength condition ap- proaching that before continuous freezing is of particular interest. Kent County plots. In these plots the last day of observed frozen ground was on March 30. Strength values after March 30 to the final sampling dates are shown in Figures nu to M6. 157 The bare plot reached its highest strength in the surface and 12- in. depths on.May 13, the 3-in. depth on April 20, and the next two depths on April 22 and 28, respectively. Depth Strength in. psi 0 53.70 3 133.70 6 201.20 9 203.70 12 192.50 With the exception of the surface reading, the strengths were above that before freezing (before December 21). As noted, most of the high strengths did not occur until 20 days or more after frost left the ground. The low surface strength.may have been a result of reduction in consistency. Moisture had dropped below prefreeze level on April 8 (Figure 32), but the soil strength remained low until.May 13. In the other depths strength increase to that before freezing and.moisture re- duction were closely related. In the herbaceous plot soil strength never approached that before freezing in the surface depth but soil moisture dropped'below prefreeze on April 20. The 3-in. strength value approached the before-freezing level on April A, but the moisture did not drOp below prefreeze con- tent until April 9. In these upper two depths reduction in consistency may have delayed strength return. The strength pattern in the hardwood plot was fairly uniform during the entire study period, and it is doubtful if soil freezing had any effect. Clinton County plots. Figure #9 shows the shear strength on the 158 no.3 0.3m hpgoo £09236 «Swap Hana.“ magafi 9.506 :98 90% €3de npmcmnpw Macaw .9. 095.3% mh< I Elmo .279 n 7 Emma .27.. c 1 Shun .270 o 2 Emma ”27... p .Illi llllll .T IIIIII 1, llllll llllllll IIIIIH. Ill; time 270 c «at 53¢ Emma. 2T~T\< ozmomn. .. ~¢ w LII; can 159 bare plot from March 29 to April 13 as well as the strength levels be- fore the first freeze. The effect of freezing can be noted depth for depth. The surface and 3-in. readings dropped well below their original values (before freezing) until the freeze of April 6 when both went above the original strength. On April 9, when the 3-in. depth had thawed again, and on April 13, when the surface depth had thawed, each depth dropped to approximately the original value. This indicated that the short-term freezing was not as effective in reducing shear strength (soil consistency) as was the extended winter period. Shear strength at 6, 9, and 12 in. in the bare plot behaved similarly to that at O and 3 in.; as soon as most frost disap- peared they dropped well below the original strengths. On the bare plot the return to original strength was almost simultaneous with the return to original moisture. The respective dates of return to prefreeze moisture and strength were: Depth, in. Strength Moisture 0 Apr 13 Apr 18 3 Apr 6 Apr 3 6 Apr 6 Apr 3 9 Apr 23 Apr 23 (No moisture data were available for the l2-in. depth before freezing.) Thus, in the 3- and 6-in. depths there may have been some reduc- tion in consistency, but in the surface and l2—in. depths the rela- tionship between strength and moisture appeared strong. In the herbaceous plot the recovery of strength in the surface, 3-in., and 6-in. depths was preceded by the loss of moisture, and con- sistency was probably the answer. In the 9—in. depth both strength 160 and moisture returned to prefreeze levels on April 3. Again, no moisture data were available before freezing for the l2-in. depth. In the hardwood plot there had been no frost since March 25. On this date the strength of each depth was appreciably greater than it was on December 31 (Figure #8). However, the snowpack had not disap- peared with the beginning of thaw as it had on the other plots (Table 16). The addition of water from the snowpack (0.76 in. of water on March 29) plus the rain (April 11-20 and subsequently) had not yet permitted return of strength equal to that before freezing when the last sampling was made May 20. Discussion. The lag in return of strength to the prefreeze level, though moisture contents had become low, has been ascribed to reduc- tion in soil consistency. However, initially after thaw begins the distribution of soil moisture influences strength. During freezing, water is drawn to ice crystals resulting in de- hydration of the soil aggregates, and intimate contact of the soil particles, i.e. increased aggregation. With thaw, moisture is not immediately distributed through the soil mass, but coats the surface of the aggregates lowering adhesion between them. This greatly reduces shear strength even at the low moisture contents observed. SUMMARY AND CONCLUSIONS Vegetation, snow cover, and soil type are known to influence occur- rence and depth of frost penetration. In this investigation soil mois- ture, bulk density, and shear strength regimes were studied, with particular reference to changes occurring in these parameters during freezing and thawing periods. Regimes were studied on wooded, herbaceous, and bare plots 160 ft square. All plots were located on medium textured soils derived from glacial till. One set of three plots was located in Kent County and one set in Clinton County. Kent County lies in a higher snowbelt. Influence of Vegetation and Snow on Frozen Soil Freezing opportunity based on air temperature was similar in both counties. The ground froze on the bare plot 11 days before the wooded and herbaceous plots in Kent County, and similarly 9 days before in Clinton County. From the beginning to the end of the winter period, frost was found approximately 90% of the time on the bare plots, 72% of the time on the herbaceous plots, and 36% of the time in the woods. The combina- tion of tree cover plus litter in the woods proved to be more effective in reducing frost occurrence than did the grass and legume cover on the herbaceous plots. When frost was present in Clinton County, it was deeper in the bare than the herbaceous plot, and in the herbaceous than the wooded plot. Snow depths averaged 2.2 times deeper on the herbaceous than the bare plot. In Kent County there was less than 1.5 in. difference in snow 161 162 depth between the bare and herbaceous plots. In Kent County snow aver- aged 3.55 in. deeper on the bare plot, 1.21 in. deeper on the herbaceous plot, and 1.23 in. deeper on the hardwood plot than on the respective plots in Clinton County. Before the period of continuous snow cover, there was less than O.h in. difference in frost depths between plots under similar cover. Snow effects were manifest in the relations developed between freezing opportunity and frost depth. In Clinton County frost depth was closely correlated with freezing opportunity. Maximum snow depth on the Clinton County bare plot was h.ll in., on the herbaceous plot 12.12 in., and in the woods 10.81 in. In Clinton County the herbaceous cover-snow combination was 2.2 times more effective in reducing frost depth than was snow alone, while the vegetation-litter-snow combination was 3.5 times more effective than herbaceous cover and snow. An over-all comparison of the Kent to Clinton County frost data indicated the effect of snow on the Kent County plots was similar to vegetation alone in Clinton County. Soil Moisture in the 15-in. Depth The moisture regime in Kent County did not change appreciably over the winter. Maximum freezing was generally less than 3 in. in any plot. The same relationships held for the Clinton County Hardwood plot. More freezing occurred in the Clinton County Bare and Herbaceous plots and a general increase in moisture was noted over the freezing period. On the day when frost left each plot, the moisture content 163 was highest on those plots with the most snow on the ground. Soil Moisture and Bulk Density Moisture-bulk density relationships were quantified and tested by linear regression technique. Bulk density was found to show a signifi- cant inverse relationship to frost depth while moisture was not signifi- cantly related to frost. However, significant inverse relationships were found between bulk density and moisture during freezing. This indicated a possibility of predicting soil moisture during the freezing period, and up to the time of soil thawing. Study of soil and snow moisture regimes indicated that moisture moved into the frozen soil profile from the snOWpack. Other investi- gators have demonstrated that water vapor may also move in the soil from warmer to colder regions, and that capillary water moves from unfrozen to frozen regions. Thus, the movement of moisture into the frozen soil without freez- ing helps explain why moisture content bears no necessary relation to frost depth. Bulk density reductions are a product of moisture movement to the frozen zone, freezing and expanding, allowing for more moisture move- ment, with subsequent freezing and expanding, etc. Reductions in bulk density occurred only in frozen soil, allowing the development of quantitative expressions between bulk density and moisture content, and bulk density and frost depth. Shear Strength In the wooded plots little variation occurred in soil shear 161+ strength, presumably because no appreciable soil freezing occurred. In plots where freezing and thawing occurred, the soil showed lower strength during the thaws than had been measured before initial freezing set in. Explanation for these low strengths is associated with soil mois- ture distribution. During freezing, water is drawn to ice crystals re- sulting in dehydration of the soil aggregates, and intimate contact of the soil particles, i.e. increased aggregation. With thaw, moisture is not immediately distributed through the soil mass, but coats the surface of the aggregates lowering adhesion between them. This greatly reduces shear strength even at the low moisture contents observed. LITERATURE CITED Aldrech, Jr., H. P. and H. M. Paynter 1953 Alway, F. 1933 Anderson, 19h2 Anderson, 19u6 1957 Analytical Studies of Freezing and Thawing of Soils. Corps of Engineers. 66 pp. J. and P. R. McMiller The floor of some northern Minnesota forests. Amer. Soil Survey Assoc. Bull. 1h: ll-A3. A. B. C., J. E. Fletcher, and N. E. Edlefson Soil moisture conditions and phenomena in frozen soils. Trans. Amer. Geophys. Union 23, Pt. II: 356- 371. Henry W. The effect of freezing on soil moisture and on evaporation from bare soil. Trans. Amer. Geophys. Union 27: Soil freezing and thawing as related to some vegetation, climatic, and soil variables. Jour. For. AS: 9h-lOl. Andrews, L. A. and W. M. Broadfoot 1958 Atkinson, 19u0 Baver, L. 19h8 1955 Bay, 0. E. 1952 Bay, R. R. 1958 The San Dimas soil sampler. Soil Sci. 85: 297-301. H. B. and C. E. Bay Some factors affecting frost penetration. Trans. Amer. Geophys. Union III: 935-951. D. Soil Physics. 2nd ed. John Wiley and Sons, Inc. New York. 398 pp- Soil Physics. 3rd ed. John Wiley and Sons, Inc. New York. pp. 111, 113, l39-lhh. , G. W. wunneke, and O. E. Hays Frost penetration into soil as influenced by depth of snow, vegetative cover, and air temperatures. Trans. Amer. Geophys. Union 33: 5hl-5h6. Occurrence and depth of frozen soil. Jour. Soil and Water Conserv. 13: 232-233. Belotelkin, K. T. 19u1 Soil freezing and forest cover. Trans. Amer. Geophys. Union 22: 173-175. 165 166 Beskow, Gunnar l9h7 Soil freezing and frost heaving with special application to roads and railroads. 26th Yearbook No. 3, Series C, No. 375. Published by Tech. Institute Northwestern Univ., Evanston, Ill. 2A2 p. Black, M. P. M., D. Croney, and J. G. Jacobs 1958 Field studies of the movement of soil moisture. Road Res. Tech. Paper No. A1. Dept. of Sci. and Ind. Res., Her Majesty's Stationery Office. 72 pp. Bouyoucos, G. J. 1913 An Investigation of Soil Temperature and Some of the Most Important Factors Influencing It. Mich. Agr. Coll. Exp. Sta. Tech. Bull. 17. 196 p. 1915 Effects of temperature on movement of water vapor and capillary moisture in soils. Jour. Agr. Res. Vol. V, No. A: 1h1-172. 1916 Soil temperature. Mich. Agr. Coll. Exp. Sta. Tech. Bull. 26: 1—133. 1921 A new classification of soil moisture. Soil Sci. 11: 33-h7. and M. M. McCool 1916 Further Studies on Freezing Point Lowering of Soils. Mich. Agr. Coll. Exp. Sta. Tech. Bull. 31. 51 pp. and 1928 The correct explanation for the heaving of soils, plants, and pavements. Jour. Amer. Soc. Agron. 20: h80-h91. Burke, H. D. and A. w. Krumbach, Jr. 1959 The use of liquid nitrogen for wet soil sampling. Jour. Geophys. Res. Vol 6h: lO39-th1. Burns, G. P. l9hh Studies in tolerance of New England forest trees. XV. Soil temperature as influenced by the density of white pine cover. Vermont Agr. Exp. Sta. Bull. 513: 1-35. Casagrande, A. 1931 Discussion on frost heaving. Proc. Highway Res. Board. Pt. 1, Vol. x1: 168-172. Chang, Jen-hu. 1957 Frost in the Ground. Chap. VII. Study of Ground Temperatures, Harvard University. Blue Hill Meteorological Observatory. Vol. I. 300 pp. 167 Chirrkov, F. w. and A. Malagin 1926 The moisture of the "podzol” in freezing and thawing in year 1923-24. Nauch, Agron, Zhur. (Jour. f Landw. Wiss.) 3: (40-51. - Church, J. E. 19h1 Discussion of "soil—freezing and forest cover" by K. T. Belo- telkin. Trans. Amer. Geophys. Union 22: 173-175. Crawford, C. B. 1952 Soil temperatures, a review of published records. Highway Res. Bu'rd Special Report No. 2: l7-hl. Diebold, C. H. 1938 The effect of vegetation upon snow cover and frost penetration during March, 1936, floods. Jour. For. 36: 626—629. Domby, C. w. and H. Kohnke 1955 The effect of freezing and thawing on structure of the soil surface. Agron. Jour. 47: 175-177. Everson, John N. and James B. Weaver 19h9 Effect of carbon black on properties of soil. Ind. and Engin. Chem. M1: 1798-1801. and 1950 Effects of carbon black on the properties of soils. II. Effects on humid soils. Franklin, T. B. 1919 The effect of weather changes on soil temperatures. Proc. of the Royal Soc. of Edinburgh XL: 56-79. Frost Effects Lab. 1951 Summary Report of Frost Investigations. Frost Effects Lab., Corps of Engineers, U. S. Army, New England Div., Boston, Mass., Processed Paper. Goodell, B. C. 1939 Soil freezing as affected by vegetation and slope aspect. Jour. For. 37: 626-629. Grim, Ralph E. 1952 Relation of frost action to the clay mineral composition of soil materials. Highway Res. Board Special Report No. 2: 167-172. Hadley, w. A. and R. Eisenstadt 1953 Moisture movement in soils due to temperature differences. Heating, Piping, and Air Conditioning 25: lll-llh. 168 Hale, C. E. 1950 Some observations on soil freezing in forest and range lands of the Pacific Northwest. U. S. Dept. Agr., For. Serv., Pac. Northwest For. and Range Exp. Sta. Paper 66. 17 p. 1951 Further observations on soil freezing in the Pacific North- west. Pac. Northwest For. and Range Exp. Sta. Res. Note 7h. Holmes, R. M. and G. w. Robertson 1960 Soil heaving in alfalfa plots in relation to soil and air temperature. Canadian Jour. Soil Sci. ho: 212-218. Hoover, M. D. and H. A. Lunt 1952 A key for the classification of forest humus types. Soil Sci. Soc. Amer. Proc. 16: 368. Hopper, T. H. 1933 Combustion train for determination of total carbon in soils. Ind. Engin. Chem. Anal. Ed. 5: lh2-1h3. Johnsgard, G. A‘:.EE.§l' .19h2 Soil Survey, Clinton County, Michigan. U. S. Dept. Agr., Bur. Plant Industry, Series 1936, No. 12. 71 pp. Map. Johnson, H. W. 1952 Frozen Action in Roads and Airfields. A review of the litera- ture. Highway Res. Board Special Report No. l. 287 pp. Jumikis, Alfreds R. 1958 Some concepts pertaining to the freezing soil systems. Highway Res. Board Special Report No. MO: 178-189. Jung, Erhard 19h2 New experimental examinations of the aggregating influence of frost on soil. Translation 5668, Soil Conserv. Serv., Coshocton, Ohio, by Moshne Kallner. 63 pp. Kienholz, Raymond 19h0 Frost depth in forest and open in Connecticut. Jour. For. 38: 3h6-350. Kittredge, Joseph 19h8 Forest Influences. McGraw-Hill, Inc. New York. 39h pp. Kuzmin, P. P. 1957 Hydrophysical investigations of land waters. Internat. Union of Geodesy and Geophysics. Internat. Assoc. of Scientific Hydrology, Vol. III: h7h. Lochhead, A. G. 192% Hydrobiological studies of frozen soil. Trans. Royal Soc. of Canada. 3rd Series, Sec. V: 75-96. 169 Lutz, H. J. 1951 Forest Soils. John Wiley and Sons, Inc. New York. 236 pp. and A. P. Caporaso 1958 Indicators of forest land classes in air-photo interpretation of the Alaska interior. Alaska For. Res. Center, For. Serv., U. S. Dept. Agr., Sta. Paper No. 10. 31 pp. illus. and R. F. Chandler 1951 Forest Soils. John Wiley and Sons, Inc. New York. 51h pp. illus. MacKinney, A. L. 1929 Effect of forest litter on soil temperature and soil freezing in autumn and winter. Ecology 3: 312-321. Mail, G. A. 1936 Soil temperatures at Bozman, Montana, during subzero weather. Science LXXXIII, No. 2163: 57h. Meyer, Adolf F. 1960 Effect of temperature on ground water levels. Jour. Geophys. Res. 65: 17u7-1752. Mosolov, V. P. 1925 Soil thawing and accumulation of snow meltwater. Nauchno- Agronomicheskii Zhurnal (Moscow) 2: 681-699. Razmerzanie pochvy i ispol'zovonie snegovoi vody; (Text in Russian) SIPRE Biblio. 8082 6: 172. 1954. Pearson, G. A. 1920 Factors controlling the distribution of forest types, Pt. I. Ecology 1: 139-159. Penner, E. 1958 The nature of frost action. Jour. Roads and Engin. Construc- tion. pp. 61-660 Pierce, R. S., H. W. Lull, and H. C. Storey 1958 Influence of land use and forest condition on soil freezing and snow depth. For. Sci. h: 2h6-263. Post, F. A. and F. R. Dreibelbis 19h2 Some influence of frost penetration and microclimate on the water relationships of woodland, pasture, and cultivated soils. Soil Sci. Soc. Amer. Proc. 7: 95-10h. Potter, L. D. 1956 Yearly soil temperatures in eastern North Dakota. Ecology 37: 62-70. 170 Rowland, E. F., I. H. Lewis, and G. A. Crabb, Jr. 1955 Frost determination by electrical resistance. Highway Re- search Board, Bull. 100. pp 17-21. Runeberg, E. 0. 1765 Anmarkninger om nagra ffirandringer pa jord-ytan i allmanhet, och under de kalla climat i synnerhet (Observations on some changes in the earths surface in general, and under cold climate in particular), Kungl. Vetensk. Akad. Haol. (Cited by Beskow). Russel, R. J. l9h3 Freeze and thaw frequencies in the United States. Trans. Amer. Geophys. Union 2h: 125-133. Sartz, R. S. 1957 Influence of land use on time of soil freezing and thawing in the Northeast. Jour. For. 55: 716-718. Schneider, R. 1957 Correlation of ground-water levels and air temperatures in the winter and Spring in Minnesota. Internat. Assoc. of Scien- tific Hydrology Pub. No. #3: 219-228 illus. Scholz, H. F. 1938 Forest cover keeps frost line at shallow depth. Jour. For. 36: 78-79. Serova, N. V. 1958 An investigation of the heat regime in soil during winter. (Issledovanie Teplovogo Rezhima Pochvy v Zimnee Urimia, Meteorologia i Gidro-logiia i 1958, No. 2: 2h-27). Trans- lated by George S. Mitchell, Foreign Area Sec., Office of Climatology, Wea. Bur., U. S. Dept. Commerce. Shannon, W. L. 1945 Prediction of frost penetration. Jour. New England Water Works Assoc. 59: 356-363. Siple, Paul A. 1952 Ice-blocked drainage as a principal factor in frost heave, slump, and solifluction. Highway Res. Board Report No. 2: 72-175. 1952 Climatic aspects of frost heave and related ground frost phenomena. Highway Res. Board, Special Report No. 2, Frost Action in Soils. pp. 10-16. Slater, C. S. and H. Hopp 1951 Winter decline of soil structure in clean-tilled soils. Agron. Jour. A3: l-h. Smith, w. 19h3 171 0. Thermal transfer of moisture in soils. Trans. Amer. GeOphys. Union, Pt. II: 356-371. Snedcor, G. W. 1956 Statistical Methods. Iowa State Coll. Press. 53h pp. illus. Storey, H. C. 1955 Taber, S. 1932 Thompson, 193A Tigerman, 1949 Frozen soil and Spring and winter floods. In Water. Year- book, U. S. Dept. Agr. pp. 179-18h. Discussion of frost heaving. Proc. Highway Res. Board II: 173-177- w. A. Soil temperatures at Winnepeg, Manitoba. Sci. Agr., Canada 15: 209-217. M. H. and V. M. Rosa Erosion from melting snow on frozen ground. Jour. For. #7: 807-809. Trimble, G. R., Jr. 1959 A problem analysis and program for watershed management re- search in the White Mountains of New Hampshire. Sta. Paper 116, Northeast For. Exp. Sta., For. Serv., U. S. Dept. Agr. #6 pp. illus. , R. S. Sartz, and R. S. Pierce 1958 How type of soil frost affects infiltration. Jour. Soil and Water Conserv. 13: 81-82. U. S. Dept. Agr. 1941 Climate and Man. U. S. Dept. Agr. Yearbook. l2h8 pp. illus. Veatch, J. O. 1953 Soils and land of Michigan. Mich. State Coll. Press. 2M1 pp. illus. Waterways Experiment Station l9h8 Wilde, S. 1955 Trafficability of Soils -- Development of Testing Instruments. Tech. Memo. 3-2AO, 3rd Supplement. Waterways Exp. Sta., Corps of Engineers, U. S. Army. 100 pp. illus. A. and G. K. Voight Analysis of Soils and Plants. J. W. Edwards, Publisher, Inc. 117 pp. illus. Wildermuth, Robert and L. Kraft 1926 Soil Survey of Kent County, Michigan. U. S. Dept. Agr., Bur. Chem. and Soils, Series 1926, No. 10. 37 pp. Map. 172 Wilkins, F. B. and W. C. Dujay 195h Freezing index data influencing frost action. Proc. Seventh Canadian Soil Mechanics Conference. Tech. Memo. 33. Associ- ate Committee on Soil and Snow Mechanics, National Res. Council. pp. 36-37. Wilner, J. 1955 The effect of low temperatures on available soil moisture during winters on Canadian prairies. Agron. Jour. M7: hll-hl2. Winterkorn, H. 19h3 The condition of water in porous systems. Soil Sci. 56: 109-115. Wintermeyer, A. M. 1925 Percentage of water freezable in soils. Public Roads 5: 5-8. APPENDIX A: TABLES 173 17h .cH m.w u Hmpoadfic mmmnm>< me : when use macaw Mopeds Hence pm dm mmo.m0H u when Mom swam Hanan Hmpoe mmm.m m mom.m m Hmm.o e mmw.mH NH mso.mm mHH mas.me He Hem.mH mm osoc\Hceoe QHOsH H msm.o H mmH.o m mmo.e m mee.m mm www.mH Hm mmm.m om Hence mom.H H o.mH-H.mH - - o.mH-H.eH mmH.H m o.eH-H.mH mam.H H - - o.mH-H.mH 0mm.H H - - meH.H H o.mH-H.sH oeo.H H mmo.H H mem.o H omm.H m o.sH-H.mH HHm.o H - - mHm.o H o.mH-H.mH mee.o H mee.o H smm.o H o.mH-H.HH mew.o H - - - - mHm.o H o.HH-H.OH mwm.o H mHo.H m mOH.H m HmH.o H o.OH-H.m - - - - Hme.o m o.m-H.w omm.o H som.o m mmm.o H o.m-H.e mes.o m mHH.o m u - 0.»-H.m omH.o H - - mmm.o m o.w-H.m OHH.o H HHH.o s omH.o H mom.o m o.m-H.H mem.o m me.o m ems.o e o.H-H.m emo.o H ose.o mH mmo.o m mmH.o m o.m-H.m mso.o H o.m-H.H o.H-o p9 mm mavpm uh mm mawpm pa mm. mawpm 9% dm mawpw pm dm macaw up an mamvm pm dm msmpm .nH am .oz _¢m .oz _¢m .oz am .oz am .oz am .02 oml “my. .Hm< .82 90%. ash. own >oz Nmm 399010189qu hpgoouflumu 1. tops: .8 08qu 5 00333893 .038 . .2 0309 x3033. to 15 12 9 to 12 Depth, in. Kent County Hardwood Depth) in. 178 Kent County Herbaceous hJ in. \4 .4. Deg‘ Mean moisture I” three inches of soil Kent Count Bare Date Appendix Table A5. 50) 85 93 8995»... 66137 391 5 .&8fl& 78.W98%£fi887nwm778 .88ufiflfiufi. . oooooooooooooooooo 00000000 OOOOOOOOOOOOOOOOOO 000 0 57.7 18 51456792351593 3201 2 788%88 .8887989888888 8818. 7 _%% ccccccccc - o 0 000.00. a o can -000- 00 000000000 0000000000.000000.10 100 000 5761147 7 73 97.9255 15.4 1 1 7- 2990/98 _ 8%Mw99m 888988 999.9 9 O I 0. 011000000 l00000100000000000 000 000 32 39963.4 9:) 72569255 14714 65 7:0 ZOO/29.0.90 92008000®2%W~190 .08 . OOOOOOOOO - OIIOOOOIOOOIICIIlo- OII- .- 010110101 01001110111000.1010 100 111 %w%% Bummmummmm%MMmfiwwmww59w7 1%m.m ........ O ........... O no.1.U O 1.1111.1.1 1.1.1111n.111.111.111.11111.1.111.11 11nv1. 1.1111 999w9wmmwm®mmwwmwwmmwwwwwmmw&wmmw&ww 555 cccccmmwmmmmmbbbbbbr rrrrrrmmmmmmmm mmmmmJJJJJmJJFFFFFmemMMWMAAAAAAAA mm2w 11w2%w 6mm 2 mM122w mmmfiee —9.7688778 _98 88 807 7888 7777: .8 u oooooooo _ 000000000000000000000 o. 0 00000000 00 00000000100000000000 0 3 559:038 8029!“. 2272h1255 55 7 ............................... . 000000000 0000000000010000000000 o meo HEW_%&mfiw%&mwwwn%%%%%fimfiafi.ww 00.0100000 0000000000000000000000 00 nmnmwmwwm.wwmmwwmmmmummomman sn_nv 000000100 0000000010000000000000 00 m&% %%mm%.manm%%m%B®m 20%%&%%%1_w& 000000111 0.110nw00101111110000000 00 09000000000000wwwwwwmwmmmmmmmmwwmww cccccwmwmmwmmwwawwwrrrrr mppppppp “WWWMMJJJJJJJJFFFFFF MMMMWMAAAAAAAA 1mmmwzvnmmmx0360mmesmuuawmlammmxmmn "wwm0.m&&&mm&.mmwmmn&&n&&&mwm&m9wfi 000000000000000001000 0000000000000 0%mmmmw&w%g w0000mm%&0ww w:%9wm% 0000000000000000000000000101000000 wmwmm&MM%mmm%%m @:m%9 m003n.wmamvss 000ll00001010010.0w010.00000000000000 ama&.m0mmwm. %m0 .9...mm9%m mw%%xa%m 0000000llll0.10.01001000000nU.1Am0.0-0000 o9mw8 %ommnfim%mmmmm%%W%fimmn%W 5%%9 .1.Ol1.1.10112111011111011011010000000 $559 flmmmwwmmmmwwm&&m&wwwmmmmwwm&mw c c c c ma ma w m.b.0.be b b.b r r r mmmmmJJJJJJJJPPFPthmmmmmm mmmmmmm 1 mmNQTHhEm%$36RMmSSmMH88918RMann 0.91 Clinton County Hardwood 19 Nov 59 1.17 1.08 1.02 0.85 0.77 Clinton County Herbaceous 11 Nov 59 0.81 0.87 0.8h 0.60 Clinton County Bare 0.98 1.00 0.65 00.0»m009.mm000011m00 n01.nonunv11nv11nvnv1‘1.111.1.1.11 11nua1 TW®W%wm%%%9%wm%m 5%% 101001100010111111100 m%%n%%%%mfimwam :W %9 001100010111111011000 %mwo m%®B%W82fi9W%%®fiM® 011111111111111110111 1.1.:121 :17: :21; .4 <1247|11 mwohou9_owmwau29 ncLuM¥mw2 1u1u9.1u2u 110111111111111111111 0500000000000000000000000000000000000 mmmm mMMJJmmmmmmmmmmmmmmmm 123112 0%08 mamm 3&nw mamaammm. .mo .mmummM&mn&mmu 0000000010000.00.nw0.nw0nw0 QchoaRvO/nvcj czaJéu a) (0:111 9&88980818&8018%779 OOOOOOIOIOonvO-llnmooomooo %%m%&w%w9 3%%m%mwmw%%m. oooooooooomomoolllLOnml mmfimwmo m1 lawfi®a 06 000100111111111111111 $$$9m&wwwwmwwwwmwwwmwwwwwmwwwwmwmwwmw “wunbbbbbb MRRRJJJJJhHFFFFPmmmmmmmmHAA0mm 8315 382b937|¢18151h5 Gamma/.2 .112 11222 123 nwOOQOOOOOOOOOOO 00 00 %3 w8m&WQW%@hun &%%. .%.%. .mawfi. 0000000000001101100 .0 37 1 335 77327 51 1 coco-cocooooooo-ooooo 000000010011111 11111 6W%m%m%%finwm 9?” owm 010010100011111011101 m%5m 22m? Mw33hah 9m3wa 100110111111111111101 fifiwflx”aw&&mm&wwwmmw®wmw&&w&w@&wmwwwmwaw mmmmmmmmmmmm mmmmmmmuumuummmmmmmmmmmmmmm m91m2959mmfim2h9817a18uwaafiafl369nmaafl$523 222 5.W%$%.w&nmw. .wfiefl.8&&8 8&369Bmmflfl85mm ”0.0%m0m01&flflfiuwnnwmn&fimfiwnufi 9.07W.&fiw ..fiDB. .68 179 Mean bulk density Appendix Table A6. in. Depth 12 9 to 12 to 15 é to 9 3 to 6 Berth, in. 0 w 12 9 to 12 to 15 6 to 9 3 to 6 O 12 9 to 12 to 15 3 Kent County Hardwood Date to 3 Date Date Kent County Herbaceous Bare Kent Count .ssanawvs.m%mmnwn9wvmvwnmumm w&finm&m 111 1 11.111111 1.112111111111111qu 813 29 9 965938223 87213 259 ‘41 M.ééfl..55"6&8h6577788nm66787u6fl6“6nm 111111111 llllllllllllllllll 111 111 E) 9623 l 7221...) I4 3141752 2 ST 752 IIIIIIIIIIIIIIIIIIIIIIIIIII I 111111111 11111111111111.1111 1L1 111 78 3.45 75 l 327. 3783 3 2 17. ......... - IIIIIOIIIIIIIIIIII 0| 111111111 llllllllllllllllll 111 111 858OI4514 5.49 897789 1.11416th 971 7 &9071090 .900 0/88098 8980091.00.1u.nN/M%o ooooooooo - III II I II I III I II II I. - 001011010 01100001000000.1101 111 010 11111111 llllllllllllllllllllll ll :JC/czz/cJ cccccnnnnnwnmwbwbbbrrrrr PWWWWWPPY mmmmmhhhththflFFFFMMMMMMMAAAAAAAAMM 1mmaw27nmwaxw 36mm 55mnm8wwlemwmmmmn .&fim&%nn& &nmvmnwfi%%&%WM&me&mwfi"&& 1158331 9957%172h917721w&&85 5 h9 T 5888h77“5775 8887838588 88 THJJ .............................. 111111111 llllllllll2lllllllllll ll 625%lww59 08w2381 2151131322915 11 OOOOOOOOOOOOOOOOOOOOOOOOOOOOO 0 111111111 llllllllllllllllllllll 11 7 7151.4 391 7187 8 5199293h2 22 3 6 5677756.5766W&7m575657567w77%m.6 I. 111111111... llllllllllllllllllllll ll ”firm/14.45373 15629969429 966237.. 57- 7 79 23hh192_u33hu3u323 1310..“6 56 5.55 ......... - I I I I I I I o I I I I I I I I I I I I I I- I I 111111101 llllllllllllllllllllll ll 999%9w@mwwwwmm&wwwwww&wwwwwmwmwwwww 5C)....) cccccmmmmwwmwmwwwwwrrrrrrrmmmmwmmmy mmmkawaJJJJFFFFFFmMMMMMMAAAAAAAAm l l 2 l.“ l 3 2 55 “7.3 .l. 7 3 mMzw Tllwemw 61mm2 m112$fi BMmezwl 7315 15153552 Slhl 7 9&9862 61 71 92 "6676876778757 TTTBWfiuTJ6778WP7£m8£H££ OOOOOOOOOOOOOO 0.... I 0...... I. llllllllllllllllllll 1111111111111 11 wwhuuam6%_&wnm an we u9 fin IIIIIIIIIIIIIIII .IIOIIIOIOIIIO 33573 315855 6 529 39h 77 979 3 Q0107568rw6/O6555/ww6m738m."BS/hm.68 87 687u%7 .......................... lllllllllllllllllllll 1111111111111 2 95391517575 2 5991292314 w5mfi5k5h6575766fl7"6%6mhhh665676.®% IIIIIIIIIIIIIIIIIII I O lllllllllllllllllllll 1111111111111 11 %3 %:wfimwwmmmfih %@%m%.9 m1 0M9%w%%%wm 5a IIIIIIIIIIIIIIIII O — lllllllllllllllllllll 1111111111111 11“ $9»999mmmmmmwmmww@mwmwmwwmmmmawmwmmwww mmmmmmmmmmmmnhmmmmmmmmmmmmmmmmwwmwwmmm lummawevnmwmsteme ssBmMHawwluaRMmamwB Clinton County Herbaceous Clinton County Hardwood Clinton County Bare 5335h3193$893539 921 5339 Inns 3 5 3 ..593949353.5559£..535153351353.5.5wnm ..1111111111111111-111 1111 11111 1111 9.3999..99999999999999.9999"9999916699 l.11111111111111111111-1111 11111 1111 9.99999999999999999.9919999_.9999.999. 1.11111111111111111111 1111-11111—1111 8.x8%8fl@$flflfifi%flfifi%58&w.%$.M.M.MM..8%8£ .h . 1.1.nw1M1“1.1.1M1u1.1.1.u.1.1.1“1.1.1.1.. 1.1.1.1.. 1.1.1.1.11 .1u1u1.1“ %.3wwM8mmBmU 8:W8 20“.mwfiufi%mmmulm9% nw .1.1.nv1.1.1.1.1.1.1.1.1.1.nvn.n.1.nv1.1. nv1.1.1. 1.1.1.1.1. 1.1.11nv 8%8889®®$®$$®$$®®$$$®®&&m&&®&&wwmw&www VCCCC n unbbbbb rrr yyy mmmmm.MJJJJJ99999mmmmmmam.pppmwwwwpmmm 89HE8$596 3829BH$18H18%583369nwm88852m ..9999 99999999 .99.99 9:9.999 919mm. o.111111111111111111111.1111 11111 111- fl.3 6%?&%8W%&®&fi8fl%w3 .m988.8flflmfiflfimm. 1ulllllllllllllllllllll-1111.11111 111- 9.999999.99999..9 .999.M.9mfim.mn9mm..9m. l.11L1L1111111111L11111.1111 11111-111- 9.9 999n999999999999m99 9999 99 9:.99._ 1.111111111111111111111-llll.12111-111- 9.999.mm 999 999999999 .99 99.99999.999. 1.111111111001011100100.0011-11111-111- 888885m®mw&ww&®®wwwwwwwwwwwwwwwwmwwwwwm meccccmmmuwnbbbbbbrrrr rrrpp ppmmmmyyy HummmmJJJJJJFFFFPmemmmmmmmAAAAAAAAAmmm u9flfi8fl59fim8£2h9BN$18nHmflfi8fl369flfim8fi852m ..9999999m999999.. 991mm. 99 9999. 99m.99m. ..111111111111111-11 lLllll .111-111- $.88 5W%B@HW &%%M688$%.m.mufl88humfifi..mfi. 1-1111111111111111111-1-1 1111 111-111— .&.&m8fl&%%8%fi%@&88.3 &%.w.$%.&%mufiwflufi.fi. 1.111111111111111.11101.111111 111 111- $unfififiw. ehfi $82 xflfifi 3&8 .M"21w%5%.&.$“$$fl. 1 llllllllllllllllLllll 111111-111 111- fiuflwwum8 aflM%N%mm9%ll .mu73%&$ 3uflflfiufi$flu 1 111111111111011010101 011111 111 111 $$$$$wmw&wmmmwwww&wwmwmwmmww&wwwwwwwwww mmmmmumMMMMJmmmmnmmmmmmmmmmmmmmmmmmmmmm m9n 3159mmseau9nnalen mans mas93ma stmm Iona daily'ohoer Itrength readings, frozen depths from surface Appendix Thble A7. Frozen Depths from Frozen Depths from Surface, in. Surface, in. Degthl 1n. Depthirin. 9| m m <| De the from Surflce 1n. mmm1m __§C. Date mu Kent County Hordwood Kent County Herbaceous 180 H NW H M H m H H C)C)C>V\C)C)U\U\C>U‘OJU\U\C)CDCUC)C3b—U\OJC)F-C)U\U\V‘ In mpm NN b-CDNNanr-tom CONKOUN NtLN ngghgwogodd F01 01.300"):ng b—Nxo OOOU‘ONOOU‘OWOWU‘NV‘OOOPNWI‘OOWO genhgfloa 8:0“t8§é$ 385% 85: FPWNNU‘V‘OONNONNV‘OFFU‘OU‘OFFOON mmF-Hr;NNU\ HOJOQCT) NWCXSCD O Oer-WOCDQ 09:0BECDFOfld)—3 “OFF- Pg r—r—cguxb—U\r-b—CJC)h—n:u\c>b—C)o:c>c>o:b-c>b-U\o1uxb— CDMLHN (*1me "\OCDKDNWJ médmxoooommorhoo Ont:mmmcohaxoxowONcoxothmr—hxomwxou‘m QIO:U\b-b-U\U\(V 3 n:n:u\cu<3 W\r—b-r—<3 “\U)r:rg<2(y can} JQFHHNmem mongammgmmhmmw NHHSNBMMNHNN: mHNNHH gR3338888888888838888888888 >>ooucc:nunuuuuukuaaaaaaahh figfiaggggfififi §§§§§£<<<<<<<§§ :aaaazanwa “Si28%8“®288$%98 m m HNH : H n ma H Hmmaaam H a m N N mHmmwoxggaul In N N gamma: a m PWONOtj-FOPO-V‘U‘OOMOV‘NOV‘NU‘ONOF Gill. 0 I It... fiNOWOtDClRRm mgmwxomto Nan-manna) O\ONHN Hf‘cQgEFm Fb—HJF— NmmNmm4 HHMNNHN NNHNI“ HNNHNNNN OOmeFmOObNONOFONNOOOFOFV‘U‘ 3 (DNNMNO 06mm H\DU\ OmU‘MFtl; fixomoxcuxomr—i CDWQCI)‘ NOHIAJ H 4.1!“: N NNHHNNN NNN HNHNN N NNHN moomonNOOmONOmONb-WNOWU‘NU‘O 0.0.00... I‘D... FWWNWOFVDOWNOOCMSQ N\OONNHNU\ mgr): )mmm e—cmxor—F Ounxo )Nmmmwxg NNN HNN NNNNNNNHH NN HNNNN O“"V‘ONr—ONOOONOmNh-Ob-hlfiommNNm “\FFBQD WHOWOCDUWNCDIIJOMMPONNVDQF NNH H mJFHbNb-PMMJWNJMWONHNN NNNHHHHHNNNNNNNNNHNHHHHNHN “\U‘OFONU‘OmmI-ANONOOOWOU‘FNNU‘OU‘ NF (“Or-4F NNN\D H ONWNMWCONWN 008:: 00013.: N: MRMRSHMMJ mmg munxo HHM MHH NNNNNNNMH O\O\O\O\ 0 mmnm$m88888888888888888888 z°°°sgsafie~~~~~h“aaaaaaaan zaaabhbhhh§§§£§§§<<<<<<<<§ “HOP-48.:ngmRV‘OdmeRH—amNOSNCflM HHN HN H HHHNN HH NNH O\ N Hmmm H HHH (bu-1mm NN N 3 NH mm Nd N? V‘ \O H '\ M m N 3N mmmm H H u a Pb—OOFFFOONNU‘U‘OONNOthb-NmI—m 8 m8 mmom mmhhmm HWMHNSH Rm Hgggmg mkHHHHQb—F :b-lndko an mg HHHHNNNNNNNHH HHHHHHHHH mkmppokmou‘mmoommooCOP-ONOI-V‘ "“0“"? Fm NNb-ln N‘O §§mzznamnanmmmn 2§§§§s§a§ CTFONNNOFW““9°N°QQQCCQCQNO RMR HSMHFNU‘ H088 Nb—U‘NH oéac: U\r—a> r- «x Eguannr—UN HHNHH HHHHH Wm H H HHH “if".‘OU‘ONOb-PNUNU‘V‘ONFNOFFU‘b-NNU‘ §§333m an 0:MOPOFmmehoomonthNNoNNor- sagsoamaagagmaasanaxanasam naaassssssssssg38888888888 fififiiéiéfiififiiiifiiiEEEEEEZE “as :savaa“ £5888”‘°fi£ an? Clinton County Hardwood Clinton County Herbaceous Clinton County Bare H HN FFNOOONOFFmfi-OOF—NNV‘V‘FMNNMI— «wagos swnautaeags g5 3358 MHNHH HHH NOOOOV‘FMFOWNU‘OFMNWNNNNWNN 11‘ mCXShr-rlm oENNI—mmmNNFOWt—NmNNfiNfi-o‘qI—‘N mm ma H mmH mm CDCDFFNO ONM MN 0‘88wa m maamdxoowocxmo BHHO‘Q HHHHH FU‘WF‘U‘NNmV‘FV‘OOmNNmb—t—BFPOMM CDNNMNQ)HFNCDPOU‘NQHFCDMNQMMP-t- OU)\OO\wOO\\OFCOF-FU\—1 mxoaam33m [~01mommmbNOU‘b-FU‘NU‘ONFFOPWPO O m In mh— memm mgCOU; 5?;i55633¥5¥:;;;flgz g_:sr4r4!irtrtririfll quarooa 8888888888888888888888 O\O\0\ Inmtn >uocccnp H35 Hkhhaahhh o a a a 0 o 9.9.9.0.Q zafinnnmm§§£§£ <<<<<<<§££ Hm mamamam Hmwmmmam n figm 2N HHHNN m HHNNg 38 mNN M NMNu N mm mm OOOOONNPWNWNNWFOO 0 mm H HN b mm m8mm§a§mam§mmam8 NHHH NHNH HNNHH WMNmoooommmbthNh thhmm mN FQMQF m mmwmahgamwmadmmad HHHHH N HHHHHNH omooomONONoNNthm .00... N Hém gm88 slassmmfimad H FNOOFFNONFOV‘FNNOU‘ $882 MMJaasgggsgg "\U‘NOOU‘OU‘U‘b-b-OF-FOU‘F- O O FéMCO (taco COM gags 888888888888 500° “agaaaaaahh 288’!) £<<<<<<<.§ IO tn: .2 5311.; G «2.3.02. 450 . OH! a Jan; .1 .0 @ Jami: ouzlldiz o: -.IIIII . hiauk . ktguk . _ §~ u 3.3..» a 5.5.1 I... _ a... _ 3.. . duh 33225 @ uza . a 1598. Iqu 1.. 355.5. .- 2 .fl .5303... .58 x animusriséo “2.02 8.53... 283 3.. a 82.29. ..qu do at... h ® 00m. 29m. zlu hlxu. m In. M... .b . k... 20‘ hknak . . VITA Arthur M. Krumbach, Jr. Candidate for the Degree of Doctor of Philosophy Final Examination: December 9, 1960 Dissertation: The Effect of Soil Freezing and Thawing on Soil Moisture, Bulk Density, and Shear Strength Under Open and Wooded Cover Outline of Studies: Major subject: Forestry Minor subject: Soil Science Biographical Items: Born, March 15, 1930, Detroit, Michigan Undergraduate Studies: Michigan College of Mining and Technology, Forestry, 19h8-52. B. S. 1952. Graduate Studies: Michigan State University, Dept. of Forestry, l95h-57. M. S. 1957. Experience: Forester, Nekoosa Edwards Paper Co., 1952; Lieutenant, U. S. Air Force 1952-5h; Forestry Aide, U. S. Forest Service, Summer 1955; Forestry Research Assistant, Michigan State University, Summer 1956; Research Forester, U. 8. Forest Service, 1957 to date. Member: Xi Sigma Pi Society of American Foresters 185 4’; .v. "