EVALUATION OF lNSTRUMENTATiON FOR mmcmtou 0F VAPOR. new m MUCK son Thesis for the Degree of M. S.. MICHIGAN STATE UNIVERSITY . JAMES LEROY DRURY. 1969 LIBRARY Michigan Stat: University .. ‘ ”‘ ' magma 31+" '- Ii HUME & SBNS’ , mm mm 191;. , ll“ IIIIIIIII ABSTRACT EVALUATION OF INSTRUMENTATION FOR INDICATION OF VAPOR FLOW IN MUCK SOIL by James Leroy Drury It was desired to know the amount of evaporation which could be expected from a Houghton muck soil with a dry sur- face layer of varying depth. The feasibility of using in- strumentation to give an instantaneous indication of vapor flow in the soil was investigated. An extensive review of literature about soil water movement and previous attempts to develop sensors of soil water flow is presented. A vapor flow apparatus made of acrylic plastic was developed which could fix the various conditions causing vapor movement across a soil column up to ten inches in length and six inches in diameter. Provision was made to directly weigh the amount of vapor moving through the soil column in response to the applied conditions. Instrumentation tested for indicating vapor flow in- cluded a series of thermocouple psychrometers in the soil profile to measure the water potential gradient which drives the vapor flow and a ceramic plate to provide a layer of known conductivity in the soil profile. It was found that the vapor flow characteristics and calculated diffusivities in muck soil did not differ greatly from values previously found in loam soils. The measured rate of evaporation from a muck soil with a dry layer four inches deep was very low with a range of .Ol-.02 cm/day. The instrumentation tested gave an indication of vapor flow in muck soil where the water potential was within the limits of 2-60 atmospheres with volumetric water contents of.10-.25 cc/cc. This range should be adequate for any con- ditions found in the field. The ceramic plate used in this study did not offer the desired layer of known conductivity needed to accurately indicate the magnitude of vapor flow in the soil column. Variation of plate conductivity was the major limitation found to the use of the instrumentation described herein. Approved Major Profe sor Approved &1 M Department Chairman “\ EVALUATION OF INSTRUMENTATION FOR INDICATION OF VAPOR FLOW IN MUCK SOIL By James Leroy Drury A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Engineering 1959 i’ ‘9 4% 65% ACKNOWLEDGMENTS The author wishes to express his sincere gratitude to Professor E.H. Kidder and Dr. J.B. Harrington for their guidance throughout this study. Appreciation is also expressed to Dr. R. J. Kunze (Soil Science) for serving on the guidance committee. A special note of thanks is offered to Mr. John B. Gerrish (Agricultural Engineering) for his interest and valuable assistance with this project. To my wife, Jennie, my infinite appreciation for her personal sacrifices and valuable assistance without which this project would have been impossible. TABLE OF CONTENTS INTRODUCTION . . . . . . . . . . . . REVIEW OF LITERATURE . . . . . . . . I. 8011 water flow . . . . . . A. Saturated flow. . . . . B. Unsaturated flow. . . . C. Effects of thermal gradients. D. Vapor flow. . . . . . . E. Diffusion studies . . . 1. Liquid diffusion. 2. Vapor diffusion . F. Evaporation. . . . . . II. Sensors of soil water flow. III. Soil water potential. . . . IV. Thermocouple psychrometry . V. Ceramic plates . . . . . . VI. Organic soils . . . . . . . DESIGN OF EXPERIMENT . . . . . . . . 1. Sources of error. . . . . . II. Mathematical modeling . . . APPARATUS AND EXPERIMENTAL PROCEDURE DISCUSSION OF RESULTS. . . . . . . . 1. Calibration of thermocouple 111 NUTJrUJLAJI-d psychrometersSA II. Vapor flow through muck soil. . . . . . .56 III. Evaluation of the thermocouple psychrometer and ceramic plate apparatus as an indicator of vapor movement . . . . . . . . . . . .68 CONCLUSIONS 0 o o o o o o o o e o o o o o o o o o 075 RECOMMENDATIONS FOR FUTURE STUDY . . . . . . . . .77 REFERENCES 0 O O O O O O O O O O O O O O O O O O O 78 APPENDICES O O O O O O O O O O O O O O O O O O O O 87 iv LIST OF TABLES Table Page 1. Calculated constants for thermocouple psychrometer calibration curves . . . . . . . . . . . . . . . . . 55 2. Data from vapor flow test run 1 . . . . . . . . . . . 58 3. Water content determinations - test run 1 . . . . . . 61 A. Calculated water vapor diffusion coefficients- test run 1 O O O O O O O C C O O O O O O O O 62 5. Data from vapor flow test run 2 . . . . . . . . . . . 63 6. Water content determinations - test run 1 . . . . . . 66 7. Calculated water vapor diffusion coefficients— test run 2 . . . . . . . . . . . . . . . . . 66 8. Thermocouple psychrometer readings— test run 2 . . . . . . . . . . . . . . . . . 72 9. Soil water potentials from psychrometer calibration curves - test run 2 . . . . . . . . . . . 73 10. Thermocouple psychrometer calibration data .‘. . . . 87 Figure l. 10. Apparatus for vapor flow measurement Weight gain of salt solution vs. elapsed time, test run 1 Weight gain of salt solution vs. elapsed time, test run 2 Depth of soil column vs. water content and LIST OF FIGURES soil water potential . . . . . . Calibration curve of thermocouple number 2 Calibration number A Calibration number 6 Calibration number 8 Calibration number 10 Calibration number 13 curve curve curve curve curve of thermocouple of thermocouple of thermocouple of thermocouple of thermocouple vi psychrometer psychrometer psychrometer psychrometer psychrometer psychrometer Page 59 64 7O 88 88 89 89 90 90 INTRODUCTION The study of microclimate in the immediate environment of agricultural crops has become increasingly important in recent years due to the prospects of a degree of localized control over atmospheric stress conditions. One such study, “Microclimatic Modification by Water Spray", has been coopera- tively undertaken by the Agricultural Engineering Department, Michigan State University, East Lansing and the Michigan Agricultural Experiment Station. One important boundary condition of this study is the amount of water vapor contri- buted to the atmosphere from the soil. The precise measurement of evaporation from soil in the field has long been a problem for researchers in Soil Science, Microclimatology and related areas. Several theoretical methods of estimating evaporation from a soil surface have been developed. These methods rely on the interrelationship of several measurable meteorological factors such as temper- ature, wind velocity and relative humidity as well as assump- tions of homogeneity of soil and atmospheric conditions over relatively large areas. The shortcomings of estimates ob- tained by such methods make it desirable to have an instru- ment which will directly indicate the upward component of water flux near the soil surface. The specific objectives of the study described herein were (1) to carry out a thorough review of literature about water movement in soil; (2) to study vapor movement through a column of muck soil; and the primary objective was (3) to evaluate the use of instrumentation to indicate vapor move- ment in the muck soil. REVIEW OF LITERATURE Preliminary to an extensive review of literature in- volving soil-plant-water systems it was necessary to learn the "language". Standard definitions and terminology for soil-water relationships and processes were obtained from "Soil Physics" by Baver (1956). "Plant-Water Relationships" by Slatyer (1967) contains an excellent review of soil physics and the various theories of water movement in soils. Information from Slatyer's book was very valuable in the planning and analysis of this re- search project. His discussion of vapor movement in soil was particularly useful. Reference will be made to the specific information used in later sections of this review. I. Soil water flow Franzini 33 31, (1967) made a thorough review of lit- erature published about 3011 water. They pointed out that although numerous laboratory studies including all phases of inater movement have been carried out, "controlled eXperiments or systematic field observations concerning water movement under field conditions are still regretably rare". 4 The excellent bibliographys of Slatyer (1967) and Franzini §t_al. (1967) led to many of the articles on the various types of soil water flow reviewed for this study. The information reviewed has been grouped into categories of water flow. Each category has been identified by a sub-title in the following discussion. A. Saturated flow The theory that soil water movement is a relatively simple hydraulic problem obeying Darcy's law has been accept- ed by most soil scientists for years. The work of Richards, Gardner and Ogata (1956) supported this theory with field observations. A pressure-membrane apparatus, described by Richards (1947), has been used by many workers, including Gardner (1956) and Rijtema (1959), to study retention of soil water against applied pressures in saturated and unsaturated conditions. Gardner (1956) developed a method to calculate capillary conductivity from the pressure membrane outflow data and Rijtema (1959) refined the method to account for membrane impedance. Zazlavsky (1935) introduced an improved and more direct method of measuring hydraulic conductivity in the laboratory by use of the moisture moment. Boelter (1964) studied the hydraulic conductivity of pasta with standard laboratory methods and field methods. He found that field methods (cavity refill and piezometers) yielded significantly lower results than laboratory methods. ifie also found a tremendous variation in the conductivity of peat soils related to the degree of decomposition, the lowest conductivities being associated with well decomposed soils. 5 A number of workers have raised questions about the universal applicability of Darcy's law. An example is the work of Swartzendruber (1962) who pointed out that Darcy's law is not applicable in certain cases for either high or low water flow velocities in soil. B. Unsaturated flow Most of the soil water flow of agricultural importance takes place in the unsaturated state with some combination of liquid and vapor prevailing. Many of the studies of un- saturated flow have been limited to liquid phase flow because it is by far the most significant at the water potentials which could previously be measured with tensiometers. Moore (1939) studied water movement through soil tubes 3 to 4 feet long having a water table at one end. The work concerned unsaturated liquid phase water movement but vapor movement was also observed. No measurements or estimates of the quantity of vapor movement were made because it was in- significant in comparison to the liquid flow. Water poten- tial was measured with tensiometers and samples for oven drying were taken through stoppered holes in the tube. A similar apparatus was adopted for the work reported in this thesis. Staple and Lehane (1954) used a slightly different technique to study water distribution in dry soil columns wetted from the top for varying periods of time. Their work demonstrated that capillary conductivity increases with soil water content and soil compaction. Once again vapor movement in the drier areas of the column was noted but no further attention was given to it. 6 Much of the work done with unsaturated soil water flow has been based on the assumption that conductivity is a function of water content and that Darcy's law is valid. Gardner and Gardner (1950) supported both of the above as- sumptions for radial, horizontal water movement. A flow equation based on continuity equations and Darcy 8 law, with a numerical method of solution for various conditions, was developed by Klute (1952). Some steady-state solutions to the flow equation were developed by Gardner (1957) who obser- ved vapor flow but felt it unimportant to his study. New experimental methods for determining the liquid conductivity in soils were developed by Gardner and Miklich (1962) and Watson (1966). Both methods gave results supporting Darcy's law. A field method of determining hydraulic conductivity was described by Rose and Kirshnan (1967). This method requires a series of water-content profiles, a knowledge of the evaporation rate, and several soil characteristics determined in the laboratory. As was the case with saturated water flow, observations of flow behavior which do not conform to Darcy's law have been made and given increasing attention. Swartzendruber (1963 and 1968) has pointed out that the relationship of flow velocity and water content becomes non-linear at low soil water contents due to the presence of clay particles. Several researchers, exemplified by Cary (1967), have studied the phenomenon of hysteresis in wetting and drying soil sam- ples. Cary offers two possible explanations for the unusual behavior observed. One agrees with Swartzendruber that changes occur in the soil matrix itself and the second recognizes the possibility of significant vapor diffusion between the soil and atmOSphere which would not be accounted for in the flow equations. C. Effects of thermal gradients The discussion to this point has been limited to studies of soil water flow in response to pressure or suction gradients where temperature gradients were either absent or ignored. Field conditions where a combination of suction and temperature gradients are present are commonly found, therefore, a great deal of work has been done to determine the effects of temperature gradients on soil water flow. Some of the earliest work with thermal gradients was done by Bouyoucos (1915) and Lebedeff (1927). Both men observed, among other things, that significant quantities of water moved toward the cold side of soil samples sub- jected to temperature gradients. After several years of additional observations of the same phenomena and conflict- ing theories as to how it took place, a group set out to determine exactly what processes were involved. Gurr et_al, (1952) applied thermal gradients to soil columns containing dissolved salts in the water. The salt was to act as a tracer for liquid movement. A net transfer of water from hot to cold was observed but the salt accumulated on the warmer side. The conclusion was that a cyclic condition 8 developed in which water flowed from hot to cold in vapor form and from cold to hot in liquid form. The vapor flow exceeded liquid flow so that there was a gradual increase in water content at the cold side. This observation was confirmed by Rollins g§_gl, (1954) and Hadley and Eisenstadt (1955) while doing very similar work. Rollins 32.21, (1954) used molecular diffusion theory to predict the amount of vapor flow due to thermal gradients. Experimental observations of vapor flow exceeded the calculated rates by six times, indicating diffusion theory was not a complete explanation. Kuzmak and Sereda (1957 a&b) separated liquid and vapor phase movement in an attempt to explain the great disparity between calculated and observed flow rates. The rate of vapor flow across a gap introduced into a saturated porous media depended only on the vapor pressure differences due to temperature, as theory predicts. There was no evidence of liquid phase flow in response to the temperature gradient when the vapor gap was removed from the saturated media. The above discussion of Kuzmak and Sereda (1957 a&b) work failed to explain the high observed rates of vapor transfer. They mention the possibility of some form of series movement in multiple evaporation-condensation steps which would be hard to detect. This concept of series parallel flow through liquid "islands" was expanded on by Philip and DeVries (1957) and DeVries (1958). A new theory was developed which took into consideration the interaction of’vapor, liquid and solid phases of water movement, dif- .ferences in average temperature gradient and actual tempera- ture distribution through a complex soil matrix as well as 9 simple diffusion theory. With these new considerations, theoretical predictions of water flow had the same order of magnitude as experimental results but were still not reliable. A more complex theory was proposed by Cary & Taylor (1962) to explain the entire phenomena of thermally driven soil water movement. This theory gave reasonably good agreement with experimental results but still failed to account for some variations. Cary (1965 b) experimentally separated liquid and vapor flow in response to both tempera- ture and suction gradients. He found that there was signifi- cant movement of liquid in response to a temperature gradient and of vapor in response to suction gradients. Both of these findings were in general disagreement with previous workers. When new theoretical rates for the two new types of soil water movement were added to the previously accepted theoret- ical rates, a prediction of total flow in agreement with experimental observations could be made. This theory was studied further for different soils by Cary (1966) and Weeks §£_al. (1968) and held up with minor modifications. Cary (1966) gave consideration to the upward flow of soil water which occurs in winter. He found that movement of water into the frost zone was much greater than the cal- culations from his newly developed theory. The observation of liquid movement from hot to cold at temperatures below freezing had been made earlier by Hadley and Eisenstadt (1955). This is opposite to the direction assumed in Cary's lO theory and he conceded that further study of the transport mechanisms at sub-freezing temperatures is needed. Weigland and Taylor (1962) and Jensen and Klute (1967) demonstrated that temperature gradients can develop across soil columns in an isothermal environment due to evaporation at a surface which removes both water and heat. This means that soil drying can never be an isothermal process and temperature gradients can not be ignored in field experiments. Slatyer (1967) made the very appropriate observation that both liquid and vapor transfer processes Operate simul- taneously and, depending on conditions, may well be in opposite directions. He also observed that when liquid phase continuity does not exist, substantial amounts of vapor transfer may occur near the soil surface. D. Vapor flow A small amount of work has been done which was limited to the vapor portion of unsaturated water flow. The work of Lebedeff (1927), Kuzmak and Sereda (1957 a) and Cary (1966) has already been mentioned. In each of these cases an effort was made to isolate the vapor portion of unsaturated flow. Each study demonstrated that vapor moves in response to vapor pressure gradients induced by temperature gradients. Cary and Taylor (1962 a) studied vapor movement in soil with an apparatus which allowed only vapor phase movement in reaponse to an applied temperature gradient. The rate of vapor flow was found to be a function of the temperature gradient and thus, of vapor pressure differences. These findings are in agreement with the work of others 11 A laboratory study of the role of soil water tension in water vapor movement in soil was undertaken by Jones and Kohnke (1952). They measured the rate of vapor movement in response to a fixed, temperature induced vapor pressure dif- ference across soil samples of varying water content and con- sequently, varying soil water tensions. It was found that vapor flow increased rapidly as soil water content decreased until it reached a maximum near the wilting point at about 15 atmoSpheres of tension. At greater tensions the rate of water transfer decreased. It should be pointed out that Jones and Kohnke were observing transfer of vapor within a sealed soil sample and rates of transmittal through an umsealed soil sample may not follow the same patterns. As was mentioned earlier, Slatyer (1967) presented a very good discussion of vapor flow. He reviewed the work of others, including those mentioned above, and presented the following general equation of vapor flow: 5122.... EL- .81. P... _ 0""sz DTVdP 82 This equation simply stated means that the quantity of vapor (1) flow depends on an isothermal diffusion coefficient related to water content and a thermal diffusion coefficient related to the temperature gradient. The above equation will be used and the terms defined in a later portion of this study. 12 E. Diffusion studies Several workers have made studies of unsaturated water flow in soils and tried to explain the results with diffusion theory from thermodynamics. As was the case in the discussion of unsaturated flow, workers have concentrated on either the liquid phase and excluded vapor flow or the vapor phase with liquid flow excluded. Very little work was found which con- sidered the simultaneous diffusion of liquid and vapor. 1. Liquid diffusion Soil water diffusivity is the ratio of unsaturated con- ductivity to the specific water capacity of a given soil. Since the conductivity is a function of soil water content, it is evident that diffusivity will be extremely variable between soils and within a given soil sample. Bruce and Klute (1956) measured diffusivity in three different sandy soils and in a media of small glass beads. All samples indicated a maximum rate of diffusion at 75-80 per cent of saturation. Diffusivity increased eXponentially with water content until the maximum was reached, as theory predicts. Calculations of diffusivity were subject to errors of as much as 200 per cent with the equipment used, so only the general trends are of value. Gardner and Mayhugh (1958) also eXperimentally Justified the theory that diffusivity increases exponentially with water content. 13 Bruce and Klute (1963) continued the work of others and calculated diffusivities from tension plate outflow data. They found that the diffusivities calculated from outflow data did not agree with those predicted by theory in the low water tension range. Since the accuracy of their calculations for low water tensions was erratic, no further eXplanation was attempted. Tests of the validity of diffusion theory were performed by Rawlins and Gardner (1963) and Kunze and Kirkham (1964). In both cases it was found that forces other than soil water tension are factors in diffusivity and that existing theory is inadequate, in agreement with Bruce and Klute (1963). Earlier work by Low (1955) and Gardner et 31. (1959) had already shown that diffusivities are significantly re- duced if the soil water contains clay ions or salt electrolytes respectively. Jackson (1963 a&b) studied the effect of porosity and temperature on soil-water diffusivity relations. He found that temperature had a definite, predictable effect on the ratio of surface tension to viscosity which in turn had the effect of increasing diffusivity as temperature increases. Diffusivity was found to be affected in an unpredictable manner by porosity. This discovery was in agreement with the findings of Gardner and Miklich (1962) who developed a faster, constant-flux method of evaluating diffusivities but had considerable variation in values due to non-uniform— ity of soil packing. No answer to the problem of variations 14 with porosity was pr0posed by either Jackson or Gardner and Miklich. Methods of solving the complex equations describing liquid diffusivity in soils were discussed by Gardner (1962) and Scott and Hanks (1963). Some simplifing assumptions were made in each case. The resultant equation of Gardner led to a new one-step method for determining soil water diffusivity which was introduced by Doering (1965). The instantaneous outflow rate and sample geometry were used to compute diffusivity values which were in agreement with the results from conventional pressure plate determinations. The one-step method is about four times as fast as other methods and does away with some of the variations caused by the assumption of constant diffusivity over a range of water content. While the evolution of theory and methods for deter- mining liquid diffusivity is of interest, its importance to this study was not as great as the development of vapor diffusion theory. 2. Vapor diffusion When it first became evident that vapor movement in unsaturated soil was significant, a number of studies con- cerning the general problem of vapor diffusion in porous media were carried out. One of the first was by Penman (1940) who compared rates of diffusion of carbon disulfide and acetone through porous solids to rates for still air. He 15 found that the ratio of steady state diffusivity in porous solids compared to that in dry air was equal to two-thirds of the porosity over a range of porosity from 0 to 0.7. The simple equation for the relation is: D : 0.66 S Do (O43m~a<~u DO {TO The constant ratio of measured to calculated flow values indicates that Penman's value for the tortuosity factor was not large enough. The rate of water diffusion in response to temperature gradients was studied by Cary & Taylor (1962 a) and Cary (1963). Equations develOped from thermodynamics of irrever- sible processes and Onsager's relation satisfactorily pre- dicted diffusion due to thermal gradients within the bounds of experimental error when isolated from other potential fields. Temperature gradients were isolated from this report for simplicity. Water vapor diffusion in relatively dry soils was studied by Jackson (1964 a, b&c) who carried out sorption, desorption and steady-state experiments. Diffusion coef- ficients were plotted against volumetric water content. It was found that diffusion increased to a maximum at about .6 relative pressure and then decreased, presumably because liquid flow commenses. The equation describing steady-state vapor flow is as follows: q---D9V 39 (6) 18 where q = water vapor flow rate 6 = volumetric water content X - distance Dev = combined liquid and vapor diffusivity Dav for Jackson's work became essentially the vapor dif- fusivity because significant liquid diffusion could not take place in the dry soil. The same case applied to the work reported in this thesis. F. Evaporation Many studies of water vapor loss from soil to the at- mOSphere by the process called evaporation have been made. Combinations of all of the preceding types of soil water flow are found in the field and contribute to evaporation, making it a very complex process. Evaporation has been of interest for many years, particu- larly in relatively dry areas where irrigation water is used. One of the more comprehensive early studies was carried out by Fisher (1923). He measured evaporation losses from var- ious materials with water conditions varying from satura- tion to dry. He found that if evaporation rate was plotted against water content for a given material, the resulting curve would have four distinct regions, namely, a constant rate under saturated conditions equal to evaporation from a free water surface, a linearly decreasing rate controlled by vapor pressure differences and water content, a slower linear decreasing rate controlled by capillary transmission to the soil surface and finally, a variable rate at low 19 water contents controlled by rates of water vapor movement in the soil. Fisher's observations were in agreement with earlier workers. The variable evaporation rate observed at low soil water contents was the only region of concern in the research project reported herein. A review of evaporation research by Keen at 31. (1926) revealed that evaporation is controlled by two groups of factors; soil-water relationships and environmental conditions. They also found that evaporation proceeds at an irregular rate over a given sample, and over different samples with identical conditions. Rates of evaporation were also studied by Veihmeyer and Hendrickson (1955). In work very similar to that of Fisher (1923) it was found that evaporation rate was constant over the range of available soil water (saturation to permanent wilting) and then fell off rapidly in dryer soils. The rates for low soil water contents were not of concern in their study. Fukuda (1956) stuuied the quantitative change in dif- fusion of water vapor in relation to its condensation anJ evaporation in soil pores. Field measurements of daily fluctuations of temperature, relative humidity and water content at various depths in loamy and sandy soils were made. Relative humidity in the soil pores, if below 100 per cent, showed fluctuation within a 24 hour periOd similar to soil water. The processes of evaporation and condensation alter- nate back and forth during any 24 hour period. The amount of vapor in soil pores depends on the diffusion exchange 20 between evaporation and condensation. This study demon- strated the complexity of problems encountered when field observations are used in evaporation research problems. Gardner (1957) and Gardner and Fireman (1958) studied the problem of evaporation from soil with a water table near the surface. Both studies deal with unsaturated flow, primarily liquid, but do give some attention to the case where a mulch is applied to the soil surface. A mulch is defined as a medium which tranSports water in the vapor phase only. It was found that evaporation rate through a mulch is inversely proportional to the thickness of the mulch and is described by the equation: I3m (Pl " P2) 3‘ L (7) where E = Evaporation rate Dm = Diffusion coefficient of water vapor in the mulch P1 = Saturation vapor pressure of soil water P2 = Vapor pressure at upper surface of the mulch L = Thickness of the mulch Philip (1957) carried out a comprehensive review of evaporation and its importance in relation to micrometeoro- logical studies. He observed that processes at the soil surface represent boundary conditions of the atmosphere and that transfer processes in the soil must be understood before soil-atmosphere interactions can be pr0perly evaluated. Three distinct regions of evaporation are defined; a con- stant rate from moist soil controlled by atmospheric con— ditions, a falling rate controlled by soil moisture 21 distribution and a dry soil phase where evaporation rate is dependant on heat fields in the soil. The transition between these phases is abrupt. These findings were in general agreement but slightly different than the earlier work of Fisher (1923) and Veihmeyer and Hendrickson (1955). Since the work of Philip (1957) several other studies of the influence of environmental factors on evaporation have been carried out. These are represented by the work of Bahrani and Taylor (1961), Gardner and Hillel (1962), Benoit and Kirkham (1963), Hung (1964) and Hanks g§_§l, (1967). Bahrani and Taylor (1961) defined potential evaporation as a function of many meteorological factors and studied the ratio of actual to potential evaporation with varying soil water potentials. The ratio of actual to potential evap- oration decreased curvilinearly with soil matric potential. This work was carried out at water contents where vapor flow was insignificant. Gardner and Hillel (1962) exposed columns of saturated soil to various rates of potential evaporation for long periods of time. Columns exposed to higher potential evap- oration dried out more quickly, but after a long time, all samples approached a constant low rate of evaporation corres- ponding to vapor movement. Once again, the major interest was in the higher rates of evaporation where liquid flow is significant. The effect of soil surface conditions on evaporation was studied by Benoit and Kirkham (1963). They used columns 22 of mulched and unmulched soil exposed to radiation and/or air movement to compare the relative effects. They found that evaporation rates from mulched columns were lower than unmulched columns. The lowest evaporation rates were associ- ated with the coarsest mulch. Soil water distribution in mulched columns was uniform from top to bottom while unmulch- ed columns dried at the surface and had increasing water con- tent to a depth of about 6 inches where it became uniform to the bottom. Radiation and air movement served to increase evaporation rates from all columns whether mulched or un- mulched. The work with mulches confirmed the assumption made by Gardner (1957) that mulches limit evaporation to the rate of vapor movement. Hung (1964) studied the effect of soil texture and water content on evaporation. He found that evaporation was greater from coarser soils at a given water content and that evaporation increased with water content in a given soil. He also observed that the rate of evaporation from all soils increased with wind velocity, temperature and vapor pressure differences and decreased with increasing soil water tension. All of these observations confirm the work of others discussed earlier. Hanks g£_§l, (1967) studied evaporation from soil columns subjected to wind or radiation. The flow of water in the soil was separated into liquid and vapor components responding to soil water tension and temperature gradients. It was found that under normal field conditions downward 23 vapor flow in response to temperature gradients accounted for only about 10 per cent of the net upward flow in reSponse to tension gradients. When the soil became quite dry, vapor flow in the Upper 5 to 10 cm of a column became significant. A similar study was carried out by Fritton gt_al. (1967). Cylinders of soil were eXposed to various radiation rates for different amounts of time. Chloride tracers were used in the soil water to indicate the type and location of water movement. Evaporation rates were initially faster from cylinders subjected to higher radiation rates until a dry surface layer formed, then evaporation decreased. Chloride distribution changes indicated that vaporization occured at a depth of 7 to 8 cm and thus rates of vapor movement con- trolled evaporation after the dry layer formed. In certain areas of the world, problems of salt accumu- lation on irrigated land is of great concern. Cary (1965) studied the accumulation of salts at the air-water interface in soils with very low evaporation rates. He found that the concentration of salt at the air-water interface was as much as ten times the average concentration in the soil water. This concentration served to decrease the vapor pressure and thus evaporation rates. Weeks gt_al, (1968) also studied salt transfer and accumulation in soils where evaporation was taking place. There was little doubt that the evaporation process caused concentrations of salt but this aspect is not of importance to this research project. 24 Theoretical analysis of the evaporation process is extremely complex and most workers who have attempted it have been forced to make many simplifying assumptions, such as those found in the work of Whistler 33 a1. (1968). Evaluation of these theoretical equations yield results which are questionable at best and usually apply only to steady-state conditions. Slatyer (1967) discussed several methods of estimating evaporation and evapotranspiration in the field. These methods were based on measured climatic factors. They were useful only for long periods of time ranging from a day to a growing season. Many methods have been devised for measuring evaporation under controlled laboratory conditions but none of these can be used effectively in the field. It has long been the desire of research workers to develop a means of accurately measuring evaporation in the field. Some attempts have been made and will be discussed in the next section. II. Sensors of soil water flow One of the earliest recorded attempts to use a sensor to study soil water movement was by Lebedeff (1927). In his study of soil water movement, a hair hygrograph was developed which indicated relative humidity in the soil. A hair which changed length with relative humidity was placed in a porous tube and connected to a recording pen by a fine nickel steel wire. The tube of the hygrograph was placed in soil samples being studied. While the accuracy of this device 25 left much to be desired, it was adequate for Lebedeff's work and did assist him in learnlng several points of importance. He found that the relative humidity in soil air was constant at 100 per cent over a wide range of water content and then fell off rapidly below the hygroscopic limit. Soil air under natural conditions at depths greater than 10 cm was always saturated with water vapor. The surface layer of soil eXper- ienced varying relative humidity. The vapor pressure in soil air was normally greater than in the atmosphere causing a net loss of water vapor (evaporation), but at night there were frequently periods when water vapor entered the soil and con- densed due to reversal of the vapor pressure gradient. Doering (1963) developed a direct method of measuring upward flow of water and, consequently, of evaporation. A monolith of soil was isolated in an evaporimeter located in the field. Conditions inside the evaporimeter were maintain- ed identical to external conditions and all input water care- fully measured. This method gave an excellent measurement of evaporation from a given location but could hardly be con- sidered portable or ineXpensive. Taylor (1968) developed a method of measuring natural evaporation with instrumentation mounted in the atmosphere directly above the soil surface. The instrument uses an anemometer which senses the vertical component of air move- ment, a fast response hygrometer, a multiplier and an inte- grator to detect average evaporation over an area of soil. 26 This device is still in the development stage and, while fairly portable, would still be too expensive for widespread field uses other than research. Gardner and Hanks (1966) evaluated the evaporation zone in the upper soil layer by burying heat flux plates at var— ious depths in the soil profile. Evaporation causes a con- centration of heat immediately above the point of vaporization and thus the location of the evaporation front could easily be traced. It was concluded that most of the evaporation occured in the upper 1 to 3 cm of soil. No estimate of the actual amount of evaporation could be made with this technique. Sensors to detect water flux in soil were developed by Byrne gt_al. (1967 & 1968). Both instruments could detect only liquid flow under saturated or nearly saturated conditions. Each instrument consisted of a heat source, a point source in one, a line source in the other, and very sensitive thermo- meters placed equidistant from the heat source. If the entire instrument is placed in the soil and oriented in the direction in which the rate of water flux is desired, then any water flow will cause a distortion of the heat field which will be detected as a temperature difference between the thermometers. The amount of temperature difference is directly related to the rate of water flow. Byrne 33 21. indicate that the "Point Source" instruments performed more satisfactorily and could detect flow rates as low as 10-4 cm/sec in soil, which is a relatively high liquid transfer rate. Neither of these instru- ments were used in the field. 27 A final instrument for measurement of soil water flow was develOped by Cary (1968). A water transducer, similar in design to heat transducers, was tested in soil columns. Initial data indicate that it may become a useful field re- search tool. The transducer contains a ceramic plate of known conductivity and means of measuring the water potential on each side of the plate. This method offers a good possi- bility for measuring liquid flow at water potentials of less than one atmOSphere. It is evident that there is a need for a reasonably simple instrument which is portable and can accurately detect liquid and/or vapor flow under field conditions. The develop- ment of such an instrument will be the major objective of this research project. III. 8011 water potential The concept of soil water potential has been used for many years by soil physicists to evaluate the forces causing soil water movement. Day (1942) defined soil water potential as the chemical potential of soil water. Several methods of determining water potential are presented and discussed. Most of these methods have since been outdated and will not be discussed here. Jones and Kohnke (1952) in a work discussed earlier, studied the influence of soil water tension on vapor move- ment of soil water. Low (1955) in another work which was discussed earlier studied the effect of osmotic forces arising in soil water containing salt solutes and found that these 28 forces tend to reduce the normal soil water energy status. Peck (1960) studied changes in moisture tension due to temperature and air pressure. He found that temperature has predictable effects on the surface tension of water and thus on soil water potential. The air pressure effects differed for applied suction or pressure. The differences were attrib- uted to air bubbles trapped in the water under pressure, but not present when auction was applied. This factor raised questions about the validity of pressure plate methods for studying soil water retention. Bolt and Miller (1958) updated the theory of total soil water potential and made calculations of both total and com- ponent potentials of water in soils. The component potentials were identified as gravitational, pressure, osmotic and ad- sorption which have the standard definitions given by Bolt and Miller as well as Slatyer (1967). The total amount of water movement in soil depends only on the total potential. This statement was challenged by Corey and Kemper (1961) who argued that total potential is a function of the state of soil water only and any representation of the direction and magnitude of net water flow must also include a function of medium properties. Considerable debate ensued on this par- ticular question but no universal answer was found. In cases involving swelling soils the medium properties have an ob- vious effect while in non-swelling soils, there may be no effect. Chahal (1964 & 1965) studied the effect of temperature and trapped air on the energy status of water and matric 29 suction respectively. He pointed out that water moving under temperature gradients can experience eXpansion of trapped air bubbles. The phenomena, previously neglected, may account for larger observed values of the temperature coefficient of pressure potential than had been predicted by theory. Changes in matric suction (adsorption potential) many times greater than theory predicts can also be caused by these temperature induced effects. Methods of measuring soil water potential in the field have been of great interest. One of the oldest methods was developed by Bouyoucos (1960). Electrical resistance blocks are buried in the soil and come to equilibrium with the sur- roundings. The amount of electrical resistance depends on the water content of the blocks and can be read with a simple voltmeter. The blocks are pre-calibrated so that a given reading indicates a known water potential. This method is not very accurate but is good enough to indicate the need for irrigation which was its original purpose. Another means of determining the energy status of soil water was developed by Richards (1965). His thermistor hygrometer was made up of glass bead thermistors which were calibrated against known relative humidities. The amount of time (heat) required to bring the thermistor to standard con- ditions is dependent on the relative humidity in the soil air. Consequently, the potential of the soil water can be deter- mined. The theory of this device is sound and its results are accurate over a wide range of water contents. 30 A final device used for determining total soil water potential is the thermocouple psychrometer. The following section will be devoted to the theory and development of thermocouple psychrometry. IV. ThermOCOUple psychrometry Thermocouple psychrometers are devices which can measure minute differences between wet and dry bulb temperatures due to the relative humidity of air surrounding a thermocouple. Wet bulb temperature depression is directly related to rel- ative humidity which in turn is related to water potential. Thus the state of water potential at a given point in a soil or plant sample can be determined by measuring the wet bulb temperature depression. While the theory is simple, the instrumentation is quite sophisticated due to the relatively small changes in temperature and relative humidity associated with large changes in water potential. Two types of thermocouple psychrometers have evolved and both are used by researchers. The first and oldest type is the "wet-100p" or Richards psychrometer initially developed by Richards (1938). This type uses a thermocouple junction which is pre-wetted with a dr0p of water giving a "wet-bulb" temperature reading which can be compared to a dry junction exposed to the same conditions. The temperature difference between the two junctions causes a small voltage which is read by a very sensitive potentiometer. The second type of psychrometer is named after its developer, Spanner (1951). 31 This type utilizes the Peltier effect for cooling a junction until water from the surrounding air condenses on it and thus produces the wet-bulb relation in a different way. Both types of psychrometer have peculiar advantages and disadvantages which limit their range of application and warrant further discussion. The earliest recorded attempt to use a "wet-loop" thermocouple psychrometer was by Richards (1938). While this psychrometer was of adequate sensitivity, problems con- nected with calibration and sample handling were not solved until several years later by Richards and Ogata (1958). Calibration required eXposure to known relative humidities at extremely well controlled temperatures. The small volt- ages were read with a microvoltmeter developed by Teele and Schuhmann (1939) which was accurate to 0.01 microvolt. Major problems with extraneous voltages are encountered at such low levels so very careful design and shielding are required. Several major electronic equipment companies now have compact microvoltmeters patterned after the one of Teele and Schuhmann (1939) which are satisfactory for thermocouple psychrometer work. The wet-100p psychrometer was used by Klute and Richards (1962) to study vapor pressure changes due to temperature. Psychrometers were calibrated over salt solutions at differ- ent temperatures. The output of the psychrometers increased linearly with the surrounding temperature for any given vapor pressure. It was found that water potential in soil at a 32 given water content increased only slightly with large temperature increases. The pressure dependence of the relative vapor pressure of soil water was studied by Richards 32 a1, (1954). Psychro- meters were calibrated at a fixed temperature over salt sol- utions and applied pressures were varied from 0 to 15 atmos- pheres. Psychrometer output for a given relative humidity was found to decrease as applied pressure was increased. The effect of applied pressures on the relative vapor pressure of soil water was not significant within the measurement limits of the psychrometers. Rawlins (1964) points out that measurements with wet- loop psychrometers are subject to a systematic error due to the presence of the water drOp in a small sample chamber. The water drop would increase the apparent relative humidity by an amount dependent upon the size of the chamber and the rate of drop evaporation. Suitable correction can be made, providing the problem is recognized. An interesting device similar to the wet-loop psychro- meter is described by Macklon and Weatherley (1965). A drop of water is exposed to a chamber containing a soil or plant sample. The rate of evaporation is a function of relative vapor pressure in the chamber. Accurate measurements of changes in drop size are made with a stOp watch and micro- scope. Calibration is done over salt solutions in a manner similar to psychrometer calibration. The method is limited to laboratory observations for obvious reasons. 33 The psychrometer develOped by Spanner (1951) is of great significance to this study. This psychrometer utilizes the Peltier effect described as follows. When two different metals are placed in contact they generally assume a differ- ence of electrical potential called the Peltier coefficient. Peltier discovered in 1834 that if a current is passed across such a junction, heat is liberated or absorbed at the point of union. The temperature will rise or fall dependent on the direction of current flow across the junction. When current flows two factors must be considered; first, the irreversible heating of conductors by the Joule effect and second, the reversible heating or cooling by the Peltier effect. The degree of Peltier cooling will increase with applied current only up to the point at which Joule heating becomes significant. If the junction of a fine thermocouple is placed in an atmosphere of humid air, and a cooling current passed through it, moisture will condense on the junction and the thermo- COUple will become a delicate "wet-bulb" thermometer provid- ing the amount of cooling is sufficient. The moisture re- evaporates when the circuit is broken and a minute e.m.f. is generated more or less pr0portional to the "wet-bulb" depres- sion associated with the atmosphere surrounding the junction. When a very sensitive voltmeter is connected into the circuit, it is possible to calibrate individual thermocouples over so- lutions of known relative humidities in the same manner as the wet-100p psychrometers discussed earlier. 34 Since the work of Spanner, many workers have used his psychrometer as a research tool. Several design modifica- tions have occured and a variety of metals have been used for the thermocouple wires. Montieth and Owen (1953) point out that one of the major limitations of the psychrometer is the limited amount of Peltier cooling. Accurate determinations of relative humidity are only possible in the range of 95 to 100 per cent. Even with this limitation, most of the range of water potentials of importance in soil and plant systems can be studied with a great improvement in accuracy over Bouyoucos blocks. Additional applications of the Spanner psychrometer are discussed by Korven and Taylor (1959) and Kijne and Taylor (1964). These two studies and all prior ones are dependent upon very precise temperature control of the calibration and sample chambers ( to t 0.001°C). This degree of temperature control is difficult to obtain in practice and restricts workers to laboratory studies of very small samples. A major breakthrough in psychrometer design and applica- tion was made by Rawlins and Dalton (1967). Prior to this work, the temperature difference between the wet junction and a separate reference junction had caused the psychrometer output. With this arrangement any difference in temperature between the two junctions not caused by humidity would cause error in the output reading and thus very precise temperature control was essential. The change made by Rawlins and Dalton was to use a single junction, with the e.m.f. reading immedi- ately before cooling as the reference. If this reading is 35 subtracted from the reading after cooling, the difference represents temperature depression. If the time lapse be- tween these two readings is sufficiently small, then only very rapid changes in temperature around the junction could cause error. Such changes are unlikely in the 15 to 30 seconds required to make a reading. This method also cancels out the effect of any parasitic e.m.f. resulting from temper- ature differences between other junctions in the circuitry. This reduces the necessity for the careful thermal grounding and shielding of earlier methods. Rawlins and Dalton (1967) also made another significant change in psychrometer design. Prior to their work, all soil or plant samples being studied were placed in small chambers and connected to the thermocouple. They found that heat generated by respiration of plant samples and the thermocouple itself was sufficient to raise the temperature within the closed chamber by a small amount. Higher temper- ature increased the water holding capacity of air in the chamber and thus reduced the apparent relative humidity. The solution to this problem was to place the thermocouple in a porous chamber and surround the chamber with a relatively large sample. Water vapor is free to move in and out of the chamber and the effects of minor heating are not significant. A complete theoretical study by Dalton and Rawlins (1968) covers all aspects in the design of thermocouple psychrometers. Chromel-Constantan thermocouples were chosen for reasons of sensitivity and output. 36 After the work of Rawlins and Dalton, the use of their psychrometer or variations on it for specific purposes be- came widespread. Studies of the water potential in plant leaves with the psychrometer is described by Hoffman and Herkelrath (1968). A number of 'in situ' studies of soil and plant systems were carried out by Rawlins §t_§1, (1968), Hoffman and Splinter (1968) and Lang (1968). All of these studies make use of the new portability of the psychrometers and the fact that precise control of temperature is no longer necessary. Several comparisons of the theory and performance of wet-loop and Peltier type psychrometers have been made by Richards and Ogata (1960), Zollinger et 31. (1966) and Rawlins (1966). The major shortcoming of the wet-loop psychrometer is that the water drop must be replaced fre- quently. In most other respects it performed slightly better than Peltier type psychrometers with higher output, better accuracy and greater range. The single limitation of fre- quent replacement of the water drOp with consequent design problems has restricted the wet-loop psychrometer to labor- atory use and it is therefore unsatisfactory for this study. The relation of water potential to relative humidity is given by Rawlins (1966) and is represented by the equation: __ RT P - 2.3 RT P 4:. ~13, - mug-.0 (3) 37 where '6- water potential gas constant Kelvin temperature molar volume of water actual vapor pressure 0 vapor pressure of pure water at same temperature menu-3:0 Psychrometers are calibrated against known P/Po ratios and the relationship of \P to psychrometer output can be shown graphically for each psychrometer. Variations in out- put will occur among identical psychrometers because of minor variations in junction size, shape and wire length. Minor variations in output of a given thermocouple will occur with age so that infrequent recalibration is required. V. Ceramic plates After an exhaustive search, it was found that very little information has been published about the use of ceramic plates of known conductivity for soil water flow studies. Ceramic plates, fritted glass plates and various types of membranes have been used for several years in pressure-plate outflow studies as described by Richards (1947). The rate of water outflow for applied pressures is used to determine soil water conductivity as a function of water con- tent. Theory and experimental procedure for the calculations are described by Gardner (1956). Workers to this point had selected plates with conductivities sufficiently high so that they could be neglected when compared to the much lower con- ductivity rates of the soils being studied. 38 Rijtema (1959) revised the method of Gardner (1956) to account for plate or membrane impedance to flow. A flow resistance was calculated from peculiarities in outflow data and consequently the actual conductivity of the plate was still not known. A final study by Bruce and Klute (1963) compares pre- vious methods of calculating soil conductivities from pres- sure plate outflow data. The outflow data did not agree with theoretical predictions, even when plate conductivity was used. The conclusion from the above mentioned works and from discussions with people who have used ceramic plates is that the conductivity of a given plate is variable with time and conditions. One of the original assumptions of this study was that a ceramic plate of known conductivity could be used as part of the apparatus. This assumption now appears erro- neous and will have to be evaluated from the experimental data. The only good source of information about the character- istics of ceramic plates is manufacturers specifications. These will be referred to in a later section of this study. VI. Organic soils The amount of work that has been done on the water hold- ing capacity and water transfer characteristics of organic soils is unfortunately very limited. Boelter (1964) discusses laboratory techniques for measuring water storage properties of organic soils. It is 39 pointed out that very large differences in bulk density and structure of different types of organic soils make it import- ant to work only with undisturbed samples. Boelter and Blake (1964) discuss the importance of always using volumet- ric water content rather than the usual gravimetric water con- tent when describing the moisture condition of an organic soil. This is necessary because of the low density of organic soil particles and the high water holding capacity. Gravimetric water contents in excess of 100 per cent, dry basis, are fre- quently encountered but are meaningless when comparisons with ordinary soil are attempted. 0n the other hand, volumetric water content of organic soil is very comparable to ordinary soil. Bulk density of organic soils should be expressed in terms of the wet volume because of considerable shrinkage during drying. In a study of the hydraulic conductivity of peat soils by Boelter (1965) it was found that a great deal of varia- bility exists due to the degree of decomposition. Undecom- posed peats had the highest conductivity and well decomposed soils had the lowest. These observations apply for saturated and nearly saturated conditions and liquid flow. No studies of vapor movement in organic soils have been found. Some work by Gardner (1957) and Gardner and Fireman (1958) was discussed earlier. This work involved the effect of mulches on evaporation and gave an equation for vapor flow rate through mulch. Due to the many similarities be- tween the mulch used and organic soils, the equation for vapor flow will be used for preliminary estimates. 40 A final study by Wheaton 22.213 (1958) describes the organic soil of interest in this project as a Houghton muck. It-s profile consists of "a surface layer of 6 to 8 inches of finely divided black muck, a subsurface layer of 8 to 36 inches of reed sedge peat with a few Tamarack roots and some other woody materials, and a substratum layer below 36 inches of coarse fiber material with slight decomposition". DESIGN OF EXPERIMENT The primary objective of the experimental portion of this study was to evaluate the feasibility of using an apparatus of thermocouple psychrometers and a ceramic plate to indicate water vapor movement in soil. The work was done on a Houghton muck soil which is of particular interest in another research project of the Agricultural Engineering Department, Michigan State University. The complete lack of information about vapor movement in muck soils makes it necessary to determine vapor flow characteristics of the Houghton muck soil as a secondary objective of this study. These characteristics must be known for specific conditions before the performance of the psychrometer apparatus can be evaluated. To simplify the study, all work was done in the labora- tory with air dried undisturbed Houghton muck soil cores. An apparatus similar in design to that of Jackson (1964 c) was used to study vapor flow through the muck soil column first without and then with the psychrometer apparatus in place. Details about the entire apparatus are given in a later section. I. Sources of error The variables which must be considered in any evaluation of vapor diffusion in soil columns have been discussed in 41 2+2 the literature review. The driving force causing vapor move- ment is related to the difference in vapor pressure between two points which is primarily a function of temperature. Less significant vapor pressure gradients are caused by variations in soil water content, differences in relative humidity of air in soil pores and in the atmosphere immediately above the soil surface, and by salt accumulations within the soil as pointed out by Cary (1965). All of these causes of vapor pressure gradients may act concurrently within a given soil. The size of soil pores and openings between them as well as the length of the soil column through which vapor is moving limits the quantity of vapor which can move in response to a given vapor pressure gradient. For this study the variables mentioned above were treated as follows. A fixed vapor pressure gradient was created across a soil sample by using a saturated atmosphere on one side and a much reduced relative humidity maintained by a saturated lithium chloride salt solution on the other. Temperature was maintained constant at 250C over the entire apparatus in an air conditioned room. Soil water content was uniformly air dry at the begining of the study and samples were oven dried at the end to find the equilibrium water content dis- tribution. Humidity in the soil pores was measured with thermocouple psychrometers. The effects of salt concentra- tions were neglected. The length of all soil samples used was 4 inches because this is about the maximum depth from which pure vapor flow may be eXpected under field conditions. 43 The soil properties of porosity and tortuosity were determined from the equation of Hanks (1958), shown as (5) in the liter- ature review. The actual water vapor flow rate through the soil columns was measured by direct weighing of the weight gain of the lithium chloride solution. Some unique variables were introduced when the thermo- COUple psychrometers were added to the soil column. The psychrometers had a ceramic sphere which is effectively a large soil pore offering no resistance to vapor flow for a relatively long distance. Several such spheres in a soil column could significantly increase the rate of vapor dif- fusion unless Special precautions are taken. Previous workers have found soil columns of 1-2 inches in diameter to be ad- equate for vapor diffusion studies. Columns used in this study had a diameter of 6 inches so that the area exposed to vapor diffusion would be at least 100 times the area of the 1/2 inch diameter ceramic spheres of the psychrometers. As a further precaution, the psychrometers were carefully placed so they were vertically above one another in the soil column. This prevented the formation of a "path of least resistance" for vapor. These precautions reduced the possible errors resulting from the presence of the psychrometer spheres to less than 1 per cent of the total vapor movement through the soil columns. Values of soil water potential obtained from thermocouple psychrometers are subject to certain limitations. The range of relative humidities which will give an output is limited to 95-100 per cent as discussed in the literature review. 44 This range corresponds to soil water potentials of 0-60 atmospheres. The air dry soil columns used will obviously have much lower humidities and thus will be out of the range of the psychrometers until adequate water vapor is absorbed to increase the water content. A final source of error to be considered was the read- ings obtained from the psychrometers. Reading errors of 1:0.1 microvolt may be expected which correspond to about :10.3 atmOSpheres. For very low humidity gradients this amount of variation may be significant between readings of different psychrometers by as much as 10 per cent. This source of error is minimized by taking all readings enough times to average out the errors. The introduction of a membrane of lower vapor conductiv- ity than the soil column should cause a steeper vapor pressure gradient across the thickness of the membrane than would be evident in an equal thickness of soil for a given vapor flow rate. This discrepancy in gradient should be evident from the outputs of psychrometers on each side of the membrane. It was the goal of this study to demonstrate the effect men- tioned above. 11. Mathematical modeling For design purposes, it was desirable to have an estimate of the amount of vapor flow which may be expected through the muck soil column. Vapor movement may be represented by the equation given by Slatyer (1967) and shown in the literature review as: 45 q... _ 86 _ 3T ( ) 73:2- -DV‘P'ST' Owen-5‘2— 1 where -'flow rate- m/cm2-sec qup - g PW = water density-gm/cm3 Dv : water vapo diffusion coefficient GP in soil-cmE/sec cm3 H20 6 :: volumetric water content- cm soil 2 : soil depth-cm D 2: th rmal moisture diffusion coefficient- TV‘P ems/sec o T : Kelvin temperature- C With the assumption of steady—state conditions and con- stant temperature, equation (1) reduces to the equation used by Jackson (1964 a, b&c) and shown in the literature review as: q=-Dav'§'§— (6) The terms have been previously defined. The water density term PW found in equation (1) has been neglected because it is so close to unity. When the following values are assumed for the evaluation of (6); Dev :expect range of 1:0-5.0 x 10-5cm/sec (from Jackson (1904 a, b&c)) 3% =expect range of l—2.5 x 10'3/cm the expected range of flow rates for the minimum and maximum conditions is q = 1.0 - 12.5 x 10‘7gm/cm2—sec. To put this in a more useful form, the equation: Q=th (9) 46 may be used where Q 2 flow in grams per day A area of diffusing surface = 2.14 x 102 cm2 for a 6 inch dia eter t = time = 8.64 x 10 sec/day An estimated flow rate of Q = 0.2 - 2.3 gm/day is obtain- ed by combining equations (6) and (9). Gardner and Fireman (1958) studied evaporation from a mulch which would permit only vapor movement. Their data indicates an evaporation rate of between 0.01 and 0.02 cm/day if extrapolated to a mulch depth of 10 cm. This evaporation rate corresponds to a flow rate of Q = 2.1 - 4.3 gm/day which is near the rate calculated from Slatyer's equation. Other workers have found that soil columns require from 2 - 4 weeks to reach equilibrium or steady-state conditions. It was assumed that the muck soil core might be in the appara- tus as long as two months with a flow rate as high as 5 gm/day. Thus the salt container must be large enough to hold at least 300 gm of water and sufficient salt to absorb this quantity. A 500 milliliter container was used. The effect of adding a ceramic plate to the column of muck soil is very difficult to model because the relationships of vapor conductivity will not be known until the experimental work is done. A plate of significantly lower vapor conductiv- ity than soil will be used to produce an easily visible effect for this study. A fixed vapor pressure gradient drives vapor through soil columns with and without the ceramic plate. The proportion of this gradient required to force water vapor through the plate is related to the thickness of the plate, 47 and the ratio of the vapor conductivity of the plate to the soil. If soil water content remains constant and a 1/4 inch plate is used in the 4 inch soil column, the total rate of vapor transfer should decrease from the unimpeded rate by 5 per cent for each multiple of soil conductivity to plate conductivity. It is very likely that the obstructing plate will cause a different water content distribution to develop through the soil column and soil vapor conductivity will in- crease as found by Jackson (1964 a, b&c). The relationships of soil vapor conductivity and water content were evaluated from the experimental data and no attempt was made to predict them. APPARATUS AND EXPERIMENTAL PROCEDURE An apparatus which would permit evaluation of vapor flow through soil columns up to ten inches long was designed and constructed of acrylic plastic. Provision was made to allow the use of thermocouple psychrometers and a ceramic plate in the vapor flow apparatus. Figure l is a sketch of the entire vapor flow apparatus showing all dimensions. Holes of 5/8 inch diameter were drilled at 1/2 inch intervals in the center section to allow insertion of thermocouple psychrometers into the soil column and removal of soil samples for water content determination. Number 2 rubber stoppers are used to seal these holes when unused. The hanger rod connected a 500 ml salt and water container to a Metler balance for weight readings. An oil reservoir was built on the rod to form a frictionless vapor seal where the rod passes through the t0p of the apparatus. A steel grid rested on the lower flange of the center section to support the soil column during a test run. A fine wire screen was placed between the soil and the grid to prevent soil particles from falling through. The bottom section was filled with distilled water to provide a saturated atmOSphere at the lower boundary of the soil column. The humidity at the upper boundary was controlled by a saturated salt solution in the container. For this study, a lithium chloride 48 “9 Balance Hanger rod Oil reservoir Top section Container for salt solution Center section 22" Soil column Steel grid soil SUpport Rubber stoppers in 5/8" holes Distilled water reservoir QL——tue Bottom section Figure 1. A aratus for va or flow measurement p 50 solution was used to reduce humidity to a very low level. The actual humidity at the soil boundary was read with an electric hygrometer made by Hygrodynamics Incorporated. An apprOpriate hygrometer probe was positioned immediately above the soil column and the lead wire sealed into one of the holes of the center section. Houghton muck soil cores 6 inches in diameter and 10 to 12 inches deep were collected in aluminum cylinders and allow- ed to dry for several months prior to data runs. A relatively undisturbed soil core was placed in the apparatus by care- fully forcing a 4 inch column of soil out of the aluminum cylinder and into the upper Opening of the center section, cutting off the column between cylinders and then lowering the column gently into position. Some disturbance occured due to the dry, crumbly structure of the soil. With the soil in place, the lower section filled with water and a salt solution suspended in the upper section, the apparatus was ready to be assembled. When the apparatus was assembled for a run, the flanges were greased with Dow Corning high vacuum silicone lubricant, a rubber gasket was placed between the sections and the flanges drawn tightly together with eight 1/4 inch bolts. The entire apparatus was now vapor and air tight. It was then placed in a compartment with temperature controlled to 25 t 1°C for the duration of the experiment. Vapor movement from the water reservoir into the soil core then began. Initially some vapor was absorbed by the soil, but eventually a steady~state condition was reached where the rate of vapor movement through the soil becomes constant. The vapor moving through the soil was absorbed by the salt solution. Daily readings of the weight of the salt solution were made and the rate of weight change plotted until a steady-state condition became evident. At this point, tests of the ceramic plate and thermocouple psychrometer apparatus were carried out and the final equilibrium moisture contents determined. Fifteen thermOCOUple psychrometers of the Rawlins and Dalton (1967) design were obtained from the Logan Instruments Company, Logan, Utah. These psychrometers contain a 1 mil Chromel-Constantan thermocouple in a 1/2 inch diameter ceramic bulb of high vapor conductivity. An electrical control unit capable of providing a cooling current of 0 - 10 milliamperes to a thermocouple was con- structed. The cooling current was fixed at 3 ma for 15 seconds for the thermocouples used in this study as recommended by Dalton and Rawlins (1968). Thermocouple psychrometer output was read with a Kiethly Model 150 B microvolt meter. Calibration of the thermocouple psychrometers was done with potassium chloride salt solutions. The degree of vapor pressure lowering at constant temperature for varying strengths of solution was given by Frazer §t_al, (1968). The psychrometer bulbs were individually wrapped with filter paper which had been soaked in a salt solution of fixed strength. The psychrometers were then sealed into a glass 52 jar containing the same salt solution to make doubly sure that the relative humidity is accurately fixed. The jar was placed in an Aminco liquid bath with the temperature controlled to 25 + o.1°c and allowed to reach equilibrium. The output of each thermocouple was carefully measured two or three times. This procedure was repeated for a total of eight salt solutions of various strengths over the range of the psychrometers. Two soil columns were brought to equilibrium for this study. The first column was used to determine the unimpeded vapor flow characteristics for the Houghton muck soil. The second column was identical to the first except that a ceramic plate 1/4 inch thick was placed at the 2 inch level of the soil column causing an obstruction to vapor flow. The varia- tion of vapor flow characteristics for this column from the first column was dependent on the plate resistance to vapor flow. Six thermocouple psychrometers were inserted into each soil column through the holes in the center section. To accomplish this with minimum disturbance of the soil, a piece of 1/2 inch tubing was used to remove a small horizon- tal section of the soil column. A psychrometer bulb was placed into the soil opening and most of the soil section re- placed behind the bulb without compaction. The thermocouple lead wire was threaded through a one hole rubber stOpper used to seal the hole in the apparatus. The small stopper hole around the lead wire was sealed with rubber cement. Several sets of readings from the thermocouple psychrometers were taken in both soil columns after equilibrium conditions had been reached. Once all necessary data had been collected from the soil columns, the apparatus was dismantled and soil samples taken for moisture content determinations. Samples were taken at 1/2 inch increments of depth and oven dried at 105°C for 48 hours. DISCUSSION OF RESULTS This portion of the study will be presented in three sections to include calibration of the thermocouple psychrom- eters, analysis of vapor flow through the muck soil column, and finally, evaluation of the psychrometer and ceramic plate apparatus as an indicator of vapor movement in soil. I. Calibration of thermocouple psychrometers A total of 15 thermocouple psychrometers were calibrated for use in this study. Psychrometer outputs for water potentials fixed by potassium chloride solutions are shown in Table 10 of the appendix. Each output shown is the aver- age of three readings of the psychrometer at a fixed water potential. The calibration data for each thermocouple psychrometer was analyzed with a digital computor and fitted to the simple linear regression formula: Y=a+bX (10) where Y is the dependent variable of psychrometer output, X the fixed variable of water potential, 3 is an intercept value representing the psychrometer output in a saturated atmosphere and b is the sensitivity of the psychrometer. Table 1 shows the computed values of a and b for each pay- chrometer as well as the 95 per cent confidence limits for 54 55 Table 1. Calibration constants for thermocouple psychrometer calibration curves Thermocouple a 6 atmosphere standardfi psychrometer—'_' '4gégzconfidence limit error number jgv calculated low high Liv 1 .848 -.405 -.385 -.424 .408 2* .438 —.369 -.337 -.401 .669 3 .133 -.365 -.334 -.396 .655 4* .345 -.356 -.335 -.376 .424 5 .405 -.310 -.288 —.332 .454 6* .110 -.411 -.393 -.429 .383 7 .536 -.380 -.352 -.408 .593 8* .927 -.377 -.357 -.398 .425 9 .439 -.339 -.308 -.369 .634 10* .191 -.373 -.343 -.403 .624 11 .058 -.399 -.359 —.440 .851 12 .328 -.415 -.381 -.449 .715 13* .751 -.396 -.366 —.425 .614 14 1.009 -.360 -.341 -.378 .390 15 .805 -.366 -.354 -.377 .238 * Indicates thermocouple psychrometers chosen for further use 56 the values of b and the standard error of observation. Analysis of the calibration data indicated that the measured psychrometer output for the -l7.97 atmosphere salt solution was significantly lower than the calculated output for all but one of the psychrometers, strongly indicating that this salt solution was weaker than intended. When this point is disregarded the remaining points are very close to linearity for most of the psychrometers. Six of the thermocouple psychrometers were chosen for further use in the project. These are indicated by an aster- isk in Table 1 and Table 10 of the appendix. Calibration curves for the six psychrometers used are included as Figure 5 through Figure 10 of the appendix. Theoretically, thermocouple psychrometers should have no output in a saturated atmosphere. However, all of the psy- chrometers calibrated for this study had a small output in the saturated atmosphere. This output is probably caused by the presence of small amounts of salt particles in the ceramic sphere surrounding the thermocouple which effectively makes it impossible to have a saturated atmosphere at the thermo— couple. This effect should be uniform for all readings of a given psychrometer and should not influence the sensitivity at all. For this reason, the intercept has been left on all of the calibration curves and is assumed to be representative. II. Vapor flow through muck soil Two columns of Houghton muck soil 4 inches in depth were brought to steady-state conditions in the apparatus described 57 previously. The first column contained six thermocouple psychrometers and the second had a 1/2 inch thick ceramic plate across the center with three of the psychrometers on each side. Test run 1 was started on May 4, 1969. The assembled apparatus was placed in an air conditioned compartment with the temperature set at 25°C. The salt container was filled with 300 grams of lithium chloride salt and connected to a Metler balance mounted above the apparatus. The balance read accurately to 0.1 gram over a range of 1000 grams. Table 2 shows the recorded weight gain of the lithium chloride salt solution during test run 1. Figure 2 shows the weight gain of the salt solution vs. elapsed time. A linear rate of weight gain was achieved at about 300 hours indicating equil- ibrium conditions existed for the remainder of the test run. Humidity measurements at the upper soil boundary were made periodically during test run 1 and are shown in Table 2. The relative humidity was steady between 23 and 24 per cent through- out the steady-state period of the run. This value of humidity is higher than the 13 per cent which the lithium chloride solution should maintain at this temperature and indicates that a significant humidity gradient developed across the air space in the Upper section of the apparatus. This is accept- able as long as the humidity value at the soil surface is known. The temperature of the compartment was recorded during each data reading and is shown in Table 2. The temperature within the compartment was quite constant throughout test run 58 Table 2. Data from vapor flow test run 1 :Eiapsed Cumulative Temperatu;e Relative *_‘ Time From Weight Gain In Humidity at Start of of Salt Compartment Upper $011 41533.- 3.833%?" -°c- 3238231- 0 O 25.0 - 25.3 6.3 25.6 22 99.0 2.5 24.4 - 141.0 17.2 25.0 - 165.0 14.2 24.5 - 205.5 22.4 26.1 23 243.0 27.8 25.0 23 269.0 33.3 25.0 - 309-0 35-5 25.0 - 376.0 42.7 25.0 23 474.5 53.4 25.0 24 499.5 56.1 25.0 - 518.5 58.1 25.5 24 594.5 65.8 24.4 - 644.0 24.7 24 70.4 59 Weight Gain of Salt Solution - Grams 100 O O O In . \o :r l j I ' I I O N l 700 1 1 600 1 J 500 I l 400 I 1 300 I l 200 l i 100 Figure 2. Weight gain of salt solution vs. elapsed time, Test run 1 Elapsed Time - Hours 60 1 and the minor variations noted should have negligible effect on the rate of vapor flow. A number of attempts were made to take psychrometer readings across the profile of the soil column after steady- state flow was achieved. No meaningfull output was obtained from any of the psychrometers, indicating that the entire muck soil column was dry enough to be out of the soil water potential range of the psychrometers. The lowest psychrometer, near the saturated atmosphere of the lower soil boundary, did give a brief indication of a very high reading on 2 or 3 occasions, but it dissipated so quickly that no reading could be made. This means that the lower soil boundary was nearly wet enough to fall within the limits of the psychrometer and probably had a potential of 80—100 atmospheres. Test run 1 was ended on May 31, 1969 and the apparatus dismantled. Samples of the soil column were carefully collected, weighed, oven dried and reweighed to determine the water content distribution. Table 3 shows the gravimetric and volumetric water contents of the soil samples. values shown are averages of three soil samples from each level. For this table and others to follow, the lower soil boundary was used as a reference plane and distances shown were measured from that plane. 61 Table 3. Water content determinations - test run 1 Wavlmaric W Bottom of Soil Water Content Water Content Column - cm % (Dry Basis) :12- 0- 2.54 25.33 .071 2.54- 3.81 ‘ 23.08 .065 3.81- 5.08 21.39 .060 5.08- 6.35 20.01 .056 6.35- 7.63 18.67 .053 7.63-10.02 18.01 .051 With the vapor flow rate through the soil column and the volumetric water content distribution known, it was possible to calculate the water vapor diffusion coefficients for increments of the muck soil column using the equations: -- .53. (6) q” Devax and Q = q A t ( 9 ) Q was determined from the slope of the steady-state portion of Figure 2 and was equal to 0.105 gm/hr or 2.515 gm/day. The values of A and t were given previously as 2.14 x 102 cm2 and 8.64 x 10“ sec/day respectively. A value of q - 1.36 x 107'7 cm/sec was the rate of vapor flow through the muck soil column of test run 1. Table 4 shows calculated values of the water content gradient 86/8X and vapor diffusion coefficient Dev using data from Table 3 and q = 1.36 x 10"7 cm/sec from equation ( 9 ). 62 Table 4. Calculated water vapor diffusion coefficients- test run 1 D s ,_ From P Diffusion Coefficient Basis 532:,th .2... o- 3.17 4.72x10'3 2.88x10'5 3.17- 4.44 3.94x10'3 3.45:10'5 4.44- 5.71 3.15::10‘”3 4.32x10'5 5.71- 7.04 2.26x10'3 6.01::10"5 7.04-10.02 1.57x10'3 8.72x10'5 The calculated values of the vapor diffusion coefficient in Houghton muck were in excellent agreement with values found by Jackson (1964c) except that they occur at higher volumetric water contents than in the loam soils used by Jackson. The different composition of muck soil could easily account for the minor variation from other soils. Test run 2 was started on June 2, 1969. All experimental conditions were identical to those described for test run 1 except that 1/4 inch thick ceramic plate was inserted across the soil column at the 2 inch level. Table 5 shows the data collected during test run 2 with the ceramic plate in place. Figure 3 shows the rate of weight gain of the salt solution vs. elapsed time for test run 2. A good linear rate of weight gain was not reached during test run 2 but a reason— ably stable rate was obtained after 350 hours. Temperature was reasonably steady during test run 2 with an average slightly greater than the intended 25°C. This should have no significant effect on the data obtained. The measured 63 Table 5. Data from vapor flow test run 2 Elapsed Cumulative Temperature Relative Time From Weight Gain In Humidity At Start of of Salt Compartment Upper Soil 2153?... 336323.23 -°c- 323-8285.. 0 O 25.0 - 49.0 20.2 24.4 - 89.5 34.3 25.6 - 161.0 45.6 25.0 31 185.0 49.3 24.5 32 257.0 59.6 24.4 - 310.5 66.6 25.6 33 338.0 71.3 25.0 - 362.3 74.3 25.5 33 405.5 79.4 25.0 - 459.0 86.4 26.6 33 496 .7 92 .6 25 .o - 521.0 96.0 25.9 34 578.8 103.7 26.5 - 618.5 108.3 25.6 34 648.5 111.8 25.0 - 698.7 117.5 25.5 - 738.0 122.2 25.3 34 837.8 133.5 25.3 ‘ 887.9 139.1 25.0 - 905.0 141.3 25.0 34 140 Weight Gain of Salt Solution - Grams O (I) 64 O \O 0 d O N -112O -100 I I l Figure 200 300 400 500 600 700 800 900 1000 100 0 Weight gain of salt solution vs. elapsed time, Test run 2 Elapsed Time - Hours 65 humidity at the upper soil surface stabilized at 34 per cent, significantly higher than in test run 1. This is probably a function of the higher soil water content reached in test run 2. Successful readings of the thermocouple psychrometers were made during the 450 to 650 hour interval of the test run. These readings will be presented and discussed in the next section of this study. Test run 2 was ended on July 10, 1969 after more than 900 hours. At this point the vapor flow rate was still declining slightly, as shown on Figure 3, indicating that a perfect steady-state had not quite been reached. It was felt that conditions had been stable enough to serve the intended purposes of this study and the apparatus was dismantled. Soil samples were carefully collected, weighed, oven dried and re- weighed for water content determinations. Table 6 shows the average gravimetric and volumetric water contents of the soil samples. The values shown are averages of three soil samples from each level. The vapor flow rate through the soil column in test run 2 taken from the slope of Figure 3 was Q = .125 gm/hr Z 3.00 gm/day. Using equation (9), a value of q = 1.62 x 10'7 cm/sec was calculated. Table 7 shows calculated values of the water con- tent gradient 86/8X and vapor diffusion coefficient Dav using equation (6) and data from Table 6. 66 Table 6. Water content determinations - run 2 Distance From Gravimetric Volumetric Bottom of Soil Water Content Water Content Column - cm_ 5 (Dry Basis) _ -6- O- 1.27 87.30 .242 1.27- 2.54 85.61 .238 2.54- 3.81 80.96 .225 3.81- 5.08 66.79 .185 5.08- 5.71(Plate) 1.55 .029 5.71- 6.99 38.22 .106 6.99- 8.25 33.45 .093 8.25- 9.52 30.81 .086 9.52-10.80 26.57 .074 Table 7. Calculated water vapor diffusion coefficients- test run 2 Distance From Water Content Water'VapSr Boggggmgf-82;1 Gr25132t Diffusiogm ?:§§1C1ent 0- 1.91 2.10::10"3 7.730x10'5 1.91- 3.18 10.22x1o"3 1.590x10"5 3.18- 5.08 21.05::10'3 0.770x10"5 5.08- 5.71(Plate) 125.50x10“3 0.129x10-5 5.71- 7.62 6.84::10"3 2.380x10"5 7.62- 8.89 5.51x10‘3 2.940xio‘5 8.89-10.80 6.29x10’3 2.580x10'5 67 Once again the values of the diffusion coefficient were within the range found by Jackson (l964a,b&c) with the excep- tion of the value across the ceramic plate and immediately below it. It was evident that the plate acted as a barrier to vapor movement and reduced the diffusion coefficient by a factor of 10. This was the desired effect of the plate. The vapor flow rate q = 1.62 x 10"7 cm/sec observed in test run 1 in spite of the presence of the ceramic plate. The possibility of this effect was mentioned in the section on Design of Experiment but could not be predicted due to the many unknown properties of muck soil. Jackson (l964a,b&c) and others have observed increases of vapor conductivity with water content within certain limits in other soils. This effect was apparently true for the Houghton muck soil used in this experiment. The increase in soil conductivity has more than offset the reduced conductivity of the ceramic plate in the soil column. The daily evaporation rate corresponding to the observed vapor flow rate may be calculated from the simple equation: Q A x FM" E (10) where evaporation rate in cm/day 2. 15 gm/day observed in test run 1 21 cm PW . 99663 gm/cm3 A value of E = .012 cm/day was obtained from equation (10). >10!!! a «H I: This was slightly lower than values obtained for comparable 68 mulches by Gardner and Fireman (1958). Gardner (1957) demonstrated that evaporation is inversely proportional to the depth of the mulch layer. Therefore, the evaporation rates for any desired depth of dry soil layer of the Houghton muck soil used in this study may be predicted. III. Evaluation of the thermocouple psychrometer and ceramic plate apparatus as an indicator of vapor movement The primary objective of the experimental portion of this study was to demonstrate that an apparatus of thermo- couple psychrometers and a ceramic plate can be used to indicate the magnitude of upward water vapor movement in soil. When the vapor conductivity of a soil is known, it is only necessary to measure the soil water potential gradient to determine the rate of vapor movement with the equation used by Cary (1968); sz'Kv¢ (11) where Jw: flow of water K= soil conductivity Vcb: soil water potential gradient The conductivity of soil is quite variable with soil water content and is usually unknown. Thus, to use equation (11) for soil in the field, it was necessary to add a medium of known, constant conductivity and to measure the potential gradient across the medium. Cary (1968) introduced such an instrument but it was limited to unsaturated liquid movement 69 with very low potential gradients of less than 1 atmosphere. The use of thermocOUple psychrometers could extend the range of such an instrument and make it possible to measure vapor flow under relatively dry soil conditions as well as unsat— urated liquid flow in a moist soil. Cary (1968) has already shown that the concept works in moist soil. This study demonstrated that the concept may be extended to drier soil by the use of thermocouple psychrometers to indicate soil water potential gradients. The previous portions of this study have established conditions of vapor flow in a relatively dry muck soil column intended to demonstrate that an apparatus of thermo- couple psychrometers and a ceramic plate can indicate vapor movement under such conditions. The ceramic plate used had a very low conductivity to emphasize the effects which indicate vapor movement. For actual field use, a plate with conductivity closer to that expected of the soil would be used. The water content distributions developing in the muck soil columns with and without a ceramic plate were given in Tables 3 and 6 and are shown on Figure 4. Thermocouple psy- chrometer readings were taken in both soil columns after a steady rate of flow became evident. No psychrometer output was obtained in the column without a plate for reasons already given. The column with a plate reached a higher water content and meaningful readings were obtained. Table 8 shows the psychrometer outputs obtained during test run 2. 70 Soil Water Potential - Atmospheres 0 IO 20 30 40 50 60 s . r r n-—r--I+ . T . I -—I2 D 4’ .— .c 8 d s 4" ‘1IO g . 3 7 Potential Curve 0 Run 2 only .4 8 2 3h- 0 a) h - ° 6 g _ P1 t a 2 a e - un o In a _ “ 4 2 e. Water Content Curves - ” Run 1 c: '* Run 2 c - Z a t: I ‘ «a W \ ‘ _ Q 'l \ ‘ O l 1, 1‘ l l l 1 l 1, 'l 14 O o .05 0'0 0'5 .20 .25 .30 Volumetric Water Content - 6 - cc H2C. cc Soil Figure 4. Depth of soil column vs. water content and soil water potential saaqawtquao - umntoo IIOS JO moqqog mos; acueqstq 71 The calibration curves shown as Figure 5 through Figure 10 of the appendix were used to convert the psychrometer out- puts to the soil water potentials shown in Table 9. The average values of the water potential at each point in the soil column were plotted on Figure 4 to show the relation- ship to soil water content. Figure 4 offers proof of the change in the soil water potential gradient caused by the ceramic plate. This change is related to the ratio of plate conductivity to soil con- ductivity. The disparity in conductivities was intention- ally overemphasized for this study and has had the undesired effect of significantly changing the water content distri- bution through the soil column used in test run 2 from that of test run 1 as shown in Figure 4. The steep potential gradient observed across the plate with the psychrometers is a function of the water content distribution. If such a device were used in the field, it would be undesirable to disrupt the normal water content to such an extent. A plate of much higher vapor conductivity should be used to have as little effect as possible on normal soil water characteristics. The change of potential gradient in such a case would be much smaller. So long as the gradient across the plate is large enough to be detected by the psychrometers and the conductivity of the plate is known, the rate of flow could be calculated. 72 Table 8. Thermocouple Psychrometer readings - test run 2 Thermocouple Psychrometer Number Time From 8 6 4 2 l3__ 10 Start of Test Distance From Bottom of Column - cm. Run 2 - Hours 1.91 3.17 4.443 6.35, 7.62 8.89 Psychrometer Output - Microvolts 458.5 .75 1.00 1.30 21.5 o o 476.0 .80 1.20 1.40 21.5 I o 619.0 .85 1.20 1.70 21.5 o 0 621.0 1.00 1.30 1.80 20.5 o 0 624.0 .80 1.00 1.80 18.0 o 0 643.0 .85 1.00 1.70 17.0 0 0 644.5 .95 1.15 1.75 19.0 1 0 646.0 .90 1.20 2.00 17.0 o 0 647.0 1.20 .90 1.65 15.0 1 0 649.0 .70 .90 1.65 15.0 I o Averages _.__8__l_ L09 l._61 l_8__6_ I Q I = Indication of high reading which dissipated too quickly 73 Table 9. Soil water potentials from psychrometer calibration curves — test run 2 Time From Distance from Center of Column - cm Start of Test 1.91 3.17 4.44 6.35 7.62 8.89 Run 2 - Hours Soil Water Potential - Atmospheres 458.5 .93 1.95 2.53 57.2 o 0 476.0 1.06 2.43 2.81 57.2 1 0 619.0 1.19 2.43 3.65 57.2 o 0 621.0 1.59 2.68 3.93 54.5 o 0 624.0 1.06 1.95 3.93 47.7 o 0 643.0 1.19 1.95 3.65 45.0 o 0 644.5 1.46 2.31 3.79 50.4 I 0 646.0 1.33 2.43 4.49 45.0 o 0 647.0 2.12 1.70 3.37 40.9 I 0 649.0 .80 1.70 3.51 39.6 I 0 21232522. 1192. 2219. 2121. 19232. 1. .2 I = Indication of high reading which dissipated too quickly I! 1 74 One of the problems which remained unanswered by this study was the variation in the conductivity of ceramic plates and other types of membranes with time and conditions. A plate of constant pr0perties was assumed in planning this study. Such a plate does not exist although some types are reasonably stable for long periods of time. Laboratory calibration of plates against known flow conditions to determine the range of plate conductivity would be required for field studies. This would be a lengthy procedure and was not judged necessary for this study. The CONCLUSIONS conclusions resulting from this investigation were: The observed rate of vapor flow through a 4 inch column of Houghton muck soil was 1.36 x 10..7 gm/cme-sec. The rate through a soil column containing a ceramic plate was 1.62 x 10"7 gm/ch-sec. Values of the diffusion coefficients calculated for the Houghton muck soil were in the range of 2.38 - 8.72 x 10"‘5 cmg/sec. The expected rate of evaporation from the Houghton muck soil with a dry surface layer 4 inches deep was in the range of l - 2 x 10'”2 cm/day. The calibrated sensitivity of the thermocouple psychrom- eters used in this study was about 25 per cent less than the sensitivity found by others for similar psy- chrometers. The psychrometers used in this study were limited to potentials greater than 2 atmospheres. 75 76 The apparatus evaluated in this study may be used to indicate water movement in muck soil over a range of water potential from 2 - 60 atmospheres. In the Houghton muck soil used, this range included volu- metric water contents of 0.10 - 0.25 cm3/cm3. RECOMMENDATIONS FOR FUTURE STUDY The effect on equilibrium soil moisture content of a soil with conductivity similar to that of an obstruct- ing plate needs further study. The problem of variable membrane conductivity needs more study and a means of providing constant conductiv- ity found. Development of a device to indicate evaporation in the field should be carried out. 77 REFERENCES REFERENCES Bahrani, B. and S.A. Taylor (1961). Influence of soil moisture potential and evaporative demand on the actual evapotranspiration from an alfalfa field. Agron. J., 53; 233-237. Baver, L.D. (1956). Soil Physics (3rd Edition). John Wiley &’Sons, Inc., New York. 89pp. Benoit, G.R. and D. Kirkham (1963). The effect of soil surface conditions on evaporation of soil water. Soil Sci. Soc. Amer. Proc., 21; 495-498. Boelter, D.H. (1964). Laboratory techniques of measuring water storage pr0perties of organic soils. Soil Sci. Soc. Amer. Proc., g§; 823-824. Boelter, D.H. (1965). Hydraulic conductivity of peats. 3011 301., 100; 227-231. Boelter, D.H. and G.R. Blake (1964). Importance of Volumetric expression of water contents of organic soils. Soil Sci. Soc. Amer. Proc., g8; 176-178. Bolt, G.R. and R.D. Miller (1958). Calculation of total and component potentials of water in soil. Trans. Amer. Geophys. Un., 39; 917-928. Bouyoucos, G.J. (1915). Effect of temperature on the move- ment of water vapor and capillary moisture in soils. Jo Age R98 0 y 2; lul-l72 a Bouyoucos, G.J. (1960). Measuring soil moisture tension. Ag. Eng., 41; 40-41. Bruce, R.R. and A. Klute (1956). The measurement of soil moisture diffusivity. Soil Sci. Soc. Amer. Proc., Bruce, R.R. and A. Klute (1963). Measurements of soil dif- fusivity from tension plate outflow data. Soil Sci. Soc. Amer. Proc., 21; 18-21. Byrne, G.F., J.B. Drummond, and C.W. Rose (1967). A sensor for water flux in soil, "Point Source" instrument. Water Resources Res., 3; 1073-1078. 78 79 Byrne, G.F., J.B. Drummond, and C.W. Rose (1968). A sensor for water flux in soil. 2. "Line Source" instru- ment. Water Resources Res., 4; 607-611. Cary, J.W. (1963). Onsager's relation and the non-isother- mal diffusion of water vapor. J. Phys. Chem., 61; 126-129. Cary, J.W. (1965 a). Vapor pressures of an air-water inter- face during evaporation. Agron. J., 51; 607-610. Cary, J.W. (1965 0). Water flux in moist soil: Thermal vs. suction gradients. Soil Sci., 100; 168-175. Cary, J.W. (1966). Soil moisture transport due to thermal gradients: Practical aspects. Soil Sci. Soc. Amer. Proc., 39; 428-433. Cary, J.W. (1967). Experimental measurements of soil-mois- ture hysteresis and entrapped air. Soil Sci., 104; 17u'180 a Cary, J.W. (1968). An instrument for 'in situ' measurements of soil moisture flow and suction. Soil Sci. Soc. Amer. Proc., 3g; 3-5. Cary, J.W. and S.A. Taylor (1962 a). The interaction of the simultaneous diffusion of heat and water vapor. Soil Sci. Soc. Amer. Proc., 26; 413-416. Cary, J.W. and S.A. Taylor (1962 b). Thermally driven liquid and vapor phase transfer of water and energy in 8011. Soil Sci. Soc. Amer. Proc., 26; 417-420. Chahal, R.S. (1964). Effect of temperature and trapped air on the energy status of water in porous media. Chahal, R.S. (1965). Effect of temperature and trapped air on matric suction. Soil Sci., 100; 262-266. Corey, A.T. and W.D. Kemper (1961). Concept of total potential in water and its limitations. Soil Sci., 9;; 299-302. Dalton, F.N. and S.L. Rawlins (1968). Design criteria for Peltier-effect thermocouple psychrometers. Soil Sci., 105; 12-17. Day, P.R. (1942). The moisture potential of soils. Soil Sci., 54; 391-400. DeVries, D.A. (1950). Some remarks on gaseous diffusion in soils. Trans, 4th Int. Cong. Soil Sci., Amsterdam, 2; 41-43. 80 DeVries, D.A. (1958). Simultaneous transfer of heat and moisture in porous media. Trans. Amer. Geophys. Uh a p :2; 909-916 0 Doering, E.J. (1963). A direct method for measuring the Upward flow of water from the water table. Soil Sci., 26; 191—195. Doering, E.J. (1965). Soil-water diffusivity by the one- step method. Soil Sci., '99; 322-326. Domby, C.W. and H. Kohnke (1956). The influence of soil crusts on gaseous diffusion. Soil Sci. Soc. Amer. Proc., (g9; 1-5. Fisher, E.A. (1923). Some factors affecting the evaporation of water from soil. J. Agr. Sci., 13; 121-143. Franzini, J.B., J.N. Luthin, I. Remson, R.W. Stallman, D. Swartzendruber, and C.H.M. VanBavel (1967). Report of the Committee on the Physics f0 Soil Moisture. I.U.G.G. Quadrennial Report (USA), 744-751. Frazer, J.C., W.R.K. Taylor, and A. Grollman (1928); Two phase liquid-vapor isothermal systems, vapor pressure lowering. International Critical Tables, McGraw Hill Book Co. Inc., N.Y., 3; 292-298. Fritton, D.D., D.Kirkham, and R.H. Shaw (1967). Soil water and chloride redistribution under various evaporation potentials. Soil Sci. Soc. Amer. Proc., 31; 599-603. Fukuda, Hitoshi (1956). Diffusion of water vapor and its exchange between condensation and evaporation in soil. Soil Sci., 61; 81-95. Gardner, R.R. and R.J. Hanks (1966). Evaluation of the evaporation zone by measurement of heat flux. Soil Sci. Soc. Amer. Proc., 36; 425-428. Gardner, W.R. (1956). Calculation of capillary conductivity from pressure plate outflow data. Soil Sci. Soc. Amer. Proc., g9; 317-320. Gardner, W.R. (1957). Some steady-state solutions of the unsaturated moisture flow equation with applica- tion to evaporation from a water table. Soil Sci., 66; 228-232. Gardner, W.R. (1962). Note on the separation and solution of diffusion tyfie equations. Soil Sci. Soc. Amer. Proc., g6; 40 . 81 Gardner, W.R. and M. Fireman (1958). Laboratory studies of evaporation from soil columns in the presence of a water table. Soil Sci., 66; 244-249. Gardner, W.R. and W. Gardner (1950). Flow of soil moisture in the unsaturated state. Soil Sci. Soc. Amer. Proc., l6; 42-50. Gardner, W.R. and D.I. Hillel (1962). The relation of ex- ternal evaporative conditions to the drying of soils. J. GeOphys. Res., 61; 4319-4325. Gardner, W.R. and M.S. Mayhugh (1958). Solutions and tests of the diffusion equation for movement of water in soil. Soil Sci. Soc. Amer. Proc., gg; 197-201. Gardner, W.R., M.S. Mayhugh, J.0. Goertzen, and C.A. Bower (1959). Effect of electrolyte concentration and exchangeable-sodium percentage on diffusivity of water in soils. Soil Sci., 66; 270-274. Gardner, W.R. and F.J. Miklich (1962). Unsaturated conductiv- ity and diffusivity measurements by a constant flux method. Soil Sci., 23' 271-274. Gurr, C.G., T.J. Marshall, and J.T. Hutton (1952). Movement of water in soil due to a temperature gradient. Soil Sci., 16; 335-345. Hadley, W.A. and R. Eisenstadt (1955). Thermally actuated moisture migration in granular media. Trans. Amer. Ge0phys. Un., 36; 615-623. Hanks, R.J. (1958). Water vapor transfer in dry soil. Soil Sci. Soc. Amer. Proc., gg; 372-374. Hanks, R.J., R.R. Gardner, and M.L. Fairbourn (1967). Evaporation of water from soils as influenced by drying with wind or radiation. Soil Sci. Soc. Amer. Proc., 3;; 593-598. Hoffman, G.J. and W.N. Herkelrath (1968). Design features of intact leaf thermocouple psychrometers for meas- uring water potential. A.S.A.E. Trans., 1;; 631-634. Hoffman, G.J. and W.E. Splinter (1968). Instrumentation for measuring water potential of an intact lant-soil system. A.S.A.E. Trans., 61; 38-40, 2. Hung, You-Tsai (1964). A Study of Evaporation Losses from Sandy Loam and Sandy Clay Loam Soils. M.S. Thesis, Michigan State University. Jackson, R.D. (1963 a). Porosity and soil-water diffusivity relations. Soil Sci. Soc. Amer. Proc., g1; 123-126. 82 Jackson, R.D. (1963 b). Temperature and soil-water diffusiv- ity relations. Soil Sci. Soc. Amer. Proc., 21; 363 -366 0 Jackson, R.D. (1964 a). Water vapor diffusion in relatively dry soil: 1. Theoretical considerations and sorption experiments. Soil Sci. Soc. Amer. Proc., 26; 172-176. Jackson, R.D. (1964 b). Water vapor diffusion in relatively dry soil: 2. Desorption experiments. Soil Sci. Soc. Amer. Proc., 26; 464-466. Jackson, R.D. (1964 c). Water vapor diffusion in relatively dry soil: 3. Steady-state experiments. Soil Sci. Soc. Amer. Proc., 26; 467-470. Jensen, R.D., and A. Klute (1967). Water flow in an unsat- urated soil with a step-type initial water content distribution. Soil Sci. Soc. Amer. Proc., 6;; 289-296. Jones, H.E. and H. Kohnke (1952). The influence of soil moisture tension on vapor movement of soil water. Soil Sci. Soc. Amer. Proc., 16; 245-248. Keen, B.A., E.M. Crowther, and J.B.H. Coutts (1926). The evaporation of water from soil. III. A critical study of technique. J. Agr. Sci., 16; 105-112. KiJne, J.W. and S.A. Taylor (1964). The temperature depend- ence of soil water vapor pressure. Soil Sci. Soc. Amer. Proc., 26; 595-599. Klute, A. (1952). A numerical method for solving the flow equation for water in unsaturated material. Soil Sci., 13; 105-116. Klute, A. and L.A. Richards (1962). Effect of temperature on relative vapor pressure of water in soil: Apparatus and preliminary measurements. Korven, H.C. and S.A. Taylor (1959). The Peltier effect and its use for determining relative activity of soil water. Can. J. Soil Sci., 32; 76-85. Kunze, R.J. andD. Kirkham (1964). Capillary diffusion and self-diffusion of soil water. Soil Sci., 21; 145-151. Kuzmak, J.M. and F.J. Sereda (1957 a). The mechanism by which water moves through a porous material subject- ed to a temperature gradient: I. Introduction of vapor gap into a saturated system. Soil Sci., 65; 291-299. 83 Kuzmak, J.M. and P.J. Sereda (1957 b). The mechanism by which water moves through a porous material sub- Jected to a temperature gradient: 2. Salt tracer and streaming potential to detect flow in the liquid phase. Soil Sci., 62; 419-422. Lang, A.R.G. (1968). Psychrometric measurement of soil water potential 'in situ' under cotton plants. 3011 Sci., 106; 460—464. Lebedeff, A.F. (1927). The movement of ground and soil waters. Low, P.F. (1955). Effect of osmotic ressure on diffusion rate of water. Soil Sci., _9; 95-100. Macklon, A.E.S. and P.E. Weatherley (1965). A vapour-pressure instrument for the measurement of leaf and soil water potential. J. Exp. Botan., l6; 261-270. Monteith, J.L. and P.C. Owen (1958). A thermocouple method for measuring relative humidity in the range 95- 100%. J. Sci. Inst., 33; 443-446. Moore, R.E. (1939). Water conduction from shallow water tables. -Hilgardia, 32; 383-426. Peck, A.J. (1960). Change of moisture tension with tempera- ture and air pressure: Theoretical. Soil Sci., 62; 303-310. Penman, H.L. (1940). Gas and vapour movements in the soil. 1. The diffusion of vapours through porous solids. J. Agr. Sci., 325 437-462. Philip, J.B. (1957). Evaporation and moisture and heat fields in the soil. J. Meteor., 34; 354-366. Philip, J.B. and D.A. Devries (1957). Moisture movement in porous materials under temperature gradients. Trans. Amer. GeOphys. Un., 36; 222-232. Rawlins, S.L. (1964). A systematic error in leaf water potential measurements with a thermocouple psy- chrometer. Sci., 146; 644-646. Rawlins, S.L. (1966). Theory for thermocouple psychrometers used to measure water potential in soil and plant samples. Agr. Meteorol., 3; 293-310. Rawlins, S.L. and F.N. Dalton (1967). Psychrometric measure- ment of soil water potential without precise temper- ature control. Soil Sci. Soc. Amer. Proc., 1; 297-301. 84 Rawlins, S.L. and W.R. Gardner (1963). A test of the valid- ity of the diffusion equation for unsaturated flow of soil water. Soil Sci. Soc. Amer. Proc., 21; 507- 511. Rawlins, S.L., W.R. Gardner, and F.N. Dalton (1968). 'In situ' measurement of soil and plant leaf water poten- tial. Soil Sci. Soc. Amer. Proc., 32; 468-470. Richards, B.G. (1965). Thermistor hygrometer for determining the free energy of moisture in unsaturated soils. Nature, 208; 608-609. Richards, L.A. (1938). Wet thermoJunctions for measuring relative humidity. Proc. Iowa Acad. Sci., 26; 175-178 s Richards, L.A. (1947). Pressure-membrane apparatus-construc- tion and use. Agr. Engr., 26; 451-454, 460. Richards, L.A., W.R. Gardner, and G. Ogata (1956). Physical processes determining loss from soil. Soil Sci. Soc. Amer. Proc., 29; 310-314. Richards, L.A., P.F. Low, and D.L. Decker (1964). Pressure dependence of the relative vapor pressure of water in soil. Soil Sci. Soc. Amer. Proc., 26; 5-8. Richards, L.A. and G. Ogata (1958). A thermocouple for vapor pressure measurement in biological and soil systems at high humidity. Sci., 128; 1089-1090. Richards, L.A. and G. Ogata (1960). Vapor pressure depres- sion at a tensiometer cup. Trans. 7th Int. Cong. Soil Sci., Madison, Wisconsin, 1; 279-283. Rijtema, P.E. (1959). Calculation of capillary conductivity from pressure plate outflow data with non-negligible membrane impedance. Neth. J. Agr. Sci., 1; 209-215. Rollins, R.L., M.G. Spangler, and D. Kirkham (1954). Move- ment of soil moisture under a thermal gradient. Proc. Highway Res. Bd., 33; 492-508. Rose, C.W. and A. Krishman (1967). A method of determining hydraulic conductivity characteristics for non- swelling soils 'in situ' and of calculatin evapor- ation from bare soil. Soil Sci., 12;; 3 9-373. Rust, R.R., A. Klute, and J.B. Giesekin (1957). Diffusion- porosity measurements using a non-steady state system. Soil Sci., 62; 453-463. 85 Scott, E.J., and R.J. Hanks (1963). Solution of the one- dimensional diffusion equation for exponential and linear diffusivity functions by power series applied to moisture flow in soils. Soil Sci,, 26; 314-322. Slatyer, n.0, (1967). Plant-Water Relationships. Academic Press, New York; 366pps. Spanner, D.C. (1951). The Peltier effect and its use in the measurement of suction pressure. Staple, W.J. and J.J. Lehane (1954). Movement of moisture in unsaturated soils. Can. J. Agric. Sci., 32; 329-342. Swartzendruber, D. (1962). Modification of Darcy's law for the flow of water in soils. Soil Sci., 23; 22-29. Swartzendruber, D. (1963). Non-Darcy behavior and the flow of water in unsaturated soils. Soil Sci. Soc. Amer. Proc., 21; 491-495. Swartzendruber, D. (1968). The applicability of Darcy's law. Soil Sci. Soc. Amer. Proc., 32; 11-18. Taylor, R.J. (1968). The automatic direct measurement of natural evaporation. U.N.E.S.C.O. Arid Zone Research, 11; 42-44. Teele, R.P. and S.Schuhmann (1939). A potentiometer for measuring voltages of 10 microvolts to an accuracy of 0.01 microvolt. J. Res. Nat'l. Bur. Standards, Van Bavel, C.H.M. (1952). Gaseous diffusion and porosity in porous media. Soil Sci., 13; 91-104. Veihmeyer, F.J. and A.H. Hendrickson (1955). Rates of evap- oration from wet and dry soils and their significance. 8011 $01., §_O_; 61-67. Watson, K.K. (1966). An instantaneous profile method for determining the hydraulic conductivity of unsat- urated porous materials. Water Resources Res., 2; 709-715. Weeks, L.V., S.J. Richards, and J. Letey (1968). Water and salt transfer in soil resulting from thermal gradients. Soil Sci. Soc. Amer. Proc., 32; 193-197. Weigland, C.L. and S.A. Taylor (1962). Temperature depression and temperature distribution in drying soil columns. 301]. $01.: $3 75-790 86 Wheaton, R.Z., E.H. Kidder and W.A. Cutler (1958). Stability and durability of underdrains in an organic soil. Quar. Bu1., Mich. Ag. Exper. Station, 49; 555-560. Whistler, R.D., A. Klute, and R.J. Millington (1968). Analysis of steady state evapotranspiration from a soil column. Soil Sci. Soc. Amer. Proc., 32; 167-174. "' Zazlavsky, D. and I. Ravina (1965). Measurement and evalua- tion of hydraulic conductivity throu h the moisture moment method. Soil Sci., 100; 10 -107. Zollinger, W. D., 6.8. Campbell and S.A. Taylor (1966). A comparison of water-potential measurements made using two types of thermocouple psychrometer. Soil Sci., 122; 231-239. APPENDIX 87 Thermocouple psychrometer calibration data Table 10. .mms moapmfim mom momono mpmmeomflommm mamzooommmzp mmpmoflan * 0.0m o.>e 0.:e m.o- m.s a.» m.m m. me 0.0m m_sa 0.xa o.o_ m.» A.” m.m 0. Ar e.mm m.me w.se o.oe e.m s.x m.m m. *me mamm e.me m.:_ o.o_ m.s m.4 J.m m. we e.mm m.w_ m.; o.m m.s e.s e._ m. we m.o o.s_ o.m_ 0.0 m.m w.m m.m a. see e.mr e.me e.me s.m >.m s.m m._ s. m o.em m.>- m.:F m.oe o.m m.s m.m s. *m m.wm o.s_ m.me m.o- m.s :.m m.m w. s m.mm e.me n.3e m.oe m.» m.m m.m m. *o m.mr e.me e.me m.w m.m m.m r.m m. m m.mw 0.0r m.mw o.m N.m m.m m.m :. a: m.om e.me e.mr :.m e.s :.m e.m m. m m.om m.©- 0.:e . o.m m.w m.: a.m :. am m.mm e.mr m.mr . o.m_ m.m 0.x Nam m. - mpHO>Oho§ I. wwdHUme QGHSH .HO mmwhmkwd. .mPSQbSO .HQPQEOHSOKHAM mHQbOOOEmflB .HmQrESZ . mmqmmu mmqmmu mmwmmu . flu. qumu mmqmu [an sopoaosgoama mmmmmmmozea a monHpaom eqam em mmme manezmeom amass 20He5 .= 1 1L, 1 . l 1 l 1 l 1 l 1 A O -10 -2O -3O -40 -50 -60 Water Potential - Atmospheres Figure 5. Calibration curve of thermocouple psychrometer number 2 30 l f I l 1 ] I l l | ‘i’ 20 = .345 - .356x 10 .— Psychrometer Output - Microvolts l 11 1_L 1 L 1 l 1 .1 O -10 -20 -30 -4O -50 -6O Water Potential - Atmospheres Figure 6. Calibration curve of thermocouple psychrometer number 4 30 I .4) H o > o n .3 '3 20 Y Z .110 - I 4..) D 4 a. 4.) 8 A 10 - 0 .p 3 -1 2 .c 2. 1 1 1 1 a: 0 -1o -20 -30 -40 -50 -60 Water Potential - Atmospheres Figure 7. Calibration curve of thermocouple psychrometer number 6 .3 30 I r* I l I *r I 1 I r I '5' > o - a O ”:2 , 20 Y I .927 - .377X ~ .5) D n. .1 49 3 c> h 10 ._ o 4) 3 o d a .C e O i. J 1 1 1 1 1 1 J 1 1 L ' “‘ 0 -1O -20 -30 -4O -50 ~60 Water Potential - Atmospheres Figure 8. Calibration curve of thermocouple psychrometer number 8 30 , 1 1 L 1 1 1 l 1 l 1 Psychrometer Output - Microvolts o -10 -20 -30 -40 -50 -60 Water Potential - Atmospheres Figure 9. Calibration curve of thermocouple psychrometer number 10 3° ..... 11.1.: 20 _ Y z 0751 " 03%x .- 1 l 1 1 1 1 1 1 1 1 1 0 -10 -20 -30 -40 -50 660 Psychrometer Output - Microvolts Water Potential - Atmospheres Figure 10. Calibration curve of thermocouple psychrometer number 13 MICHIGAN STATE UNIVERSITY LIBRARIES 111mm l 30714384 IIIIIIflIIII 1293