THE MOVEMENT AND EVAPORATION OF SOIL WATER IN RELATION TO pF1

C. M. WOODRUFF2

r I AHIS paper represents a portion of a more com--»• plete study on which the author is engaged that

is designed to establish ( i ) the nature and influenceof the mechanisms that are responsible for the re-tention and movement of water in the soil at differ-ent moisture potentials; (2) the state, whether freeliquid, adsorbed liquid, or vapor, in which watermoves through the soil at different moisture poten-tials; and (3) to determine whether or not theremight be points of fundamental significance on themoisture potential curve.

The results suggest that most, if not all, of thephenomena of soil water relations may be accountedfor by an integration of the effects of surface tensionand adsorption in pores of different sizes, with re-spect to the vapor pressure of water and to the pres-sure of the atmosphere. This integration indicatesthat in the soil there are three possible groups ofpore sizes that affect the water relations of the soilin different ways, as follows:

1. Pores with effective radii in excess of 1.5 \i.Within pores of this size the water beneath theair water interface is always under a positivepressure in excess of the vapor pressure of thewater. It will flow as a liquid from regions ofhigh pressure to regions of lower pressure. Itobeys the familiar laws of rise in capillarytubes. It adjusts itself rapidly to changes inthe moisture potential.

2. Pores with effective radii of 0.2 [i to 1.5 H-. The removal of a trace of water from pores of

this size reduces the pressure on the remainingwater to the vapor pressure of the water at thecurved interface. This water is at its boilingpoint. It is unstable. It can be removed fromthe soil only as a vapor. It will move throughthe soil only as a vapor. Consequently, mois-ture potential adjustments are slow.

3. Pores with effective radii less than 0.2 \i.These pores are small enough that their con-tents are affected by the adsorptive pressuresacting at the solid liquid interface. The sum ofthe pressures of the atmosphere and of adsorp-tion exceeds the negative pressure produced •by surface tension at the air-water interface.Hence, the total pressure on the water exceedsthe vapor pressure of the water, and the wateris stable. It may move through the soil as aliquid, but it may be extracted from the soilonly as a vapor.

The energy required to extract water from the poresof group i is used in overcoming surface tensionforces. The amount of energy required depends on thesize of the pore from which the water is withdrawn.The energy required to extract water from most of thepores of group 2 is used in overcoming surface ten-sion forces. The amount of energy required is con-stant because the radius of curvature of the air-waterinterface is limited by the stability of the water ir-respective of the size of the pores. However, adsorp-tive forces which begin to influence the smallest poresin this group increase the amount of energy requiredto remove water from the pores so affected. Theenergy required to extract water from the pores ofgroup 3 is used in overcoming both surface tensionforces and adsorptive forces, and the amount re-quired depends upon the size of the pores. Thetransition between the pores of groups i and 2 occursat a moisture potential of pF 3. The transition be-tween the pores of groups 2 and 3 occurs at a mois-ture potential of pF 4.1.

Most soils contain pores of all sizes. Water with-in the pores of group i moves rapidly under stressand is lost rapidly by percolation or by evaporation.Water retained in pores of group 2 escapes only bydiffusion through the vapor state. Hence, the rate ofloss is negligible. This water is .utilized by plants.Water retained in the pores of group 3 moves at asignificant rate. In soils of good tilth these pores arenot continuous, so that very few of them are effectiveconductors. Water cannot be withdrawn from thesepores by plants. Therefore, an average soil aftersaturation loses water rapidly until it attains fieldcapacity at pF 3. If the soil is bare of vegetation, thesurface becomes dry and the dry layer becomesthicker with time. If the soil is in good tilth the poresof group 3 are ineffective, and no moisture is lostfrom the interior of the moist soil. Evaporation oc-curs only at a sharp boundary between the moistsoil and the dry layer. If the soil is vegetated themoisture diffuses to the roots through the vaporstate, and the entire mass of soil dries to the wiltingpoint at pF 4.1, leaving water only in the pores ofgroup 3.

The behavior of some sandy soils is conditionedby a dominant group of pores larger than 1.5 \a

'Contribution from the Department of Soils, Missouri Agricultural Experiment Station, Columbia, Mo. Journal Series No. 791."Instructor in Soils.

I2O

WOODRUFF: MOVEMENT AND EVAPORATION OF SOIL WATER 121

These pores are emptied at a moisture potential be-low pF 3. When these pores are emptied by drainageor by evaporation at the surface of the soil, moisturemovement through the liquid phase ceases, and themoisture potential of the soil corresponds to themoisture potential required to withdraw water fromthe relatively large pores. Plants growing on suchsoils must feed on the water retained at the points ofcontact between particles after the pores have beenemptied.

Puddled soils of fine texture, such as fine silt loamsand clay loams, sometimes contain very few poreslarger than 0.2 \i. These soils continue to lose waterby evaporation at the surface until the moisturecontent of the entire profile is near the wilting point.These soils do not exhibit a field capacity at pF 3.This condition is typical of many of our over-culti-vated gray prairie soils.

These conclusions are supported by data on therate of evaporation from soils of different texturesat different moisture potentials, by data on the thick-ness of the dry surface layer as drying progresses,and by measurements of the force with which wateris adsorbed at different distances from the walls ofthe soil pores.

Theoretical considerations leading to the evalua-tion of the pore sizes at the transition between groupsi and 2 and between groups 3 and 4 will be presentedin a later paper. The purpose of the present paper isto present that phase of the study which bears onthe phenomena of moisture movements when a shortcolumn of soil dries in the air.

RATES OF EVAPORATION AS A MEANS OFCHARACTERIZING THE DYNAMICS OF

THE SOIL MOISTURE SYSTEMPROCEDURE FOR STUDYING RATE OF EVAPORATION

Small jelly tumblers were filled with soil, saturatedwith water, and set in the laboratory to dry. One setof samples was exposed to the still air of the labora-tory, the remainder were placed before an electricfan. The relative humidity was near 50% and thetemperature fluctuated between 24° and 28 °C. Tum-blers of water were intermingled with the tumblersof soil. All were weighed at intervals to observe thequantity of water lost from the soil in relation to thequantity lost from the free water surface. The lossfrom the free water surface was used as a functionalmeasure of time. The soils studied included whitesand uniformly 0.8 mm in diameter, sandy loams,

silt loams, and clay loams. The latter two were gran-ulated and of good tilth. A porous clay tensiometercup was inserted in the sample of sand in order thatthe changes in moisture potential might also beobserved.

EVAPORATION FROM SAND OF UNIFORM PARTICLE SIZEAT MOISTURE CONTENTS ABOVE FIELD CAPACITY

Two hundred seventy grams of the sand of uni-form particle size were covered with 66 cc of waterand set before an electric fan to dry. The quantityof water evaporated, the tension, and the thicknessof the dry layer of sand that developed at the surfaceduring the final stages of drying are presented inFig. i.

The rate of water loss was almost constant until50 cc had been evaporated. During this stage of theexperiment capillarity maintained a uniformly moistsurface over the sand. There was then a short tran-sitional period during which the sand lost 7 cc ofwater and the tension rapidly increased to a valueof 25 cm of water. At this tension the surface of thesand began to dry. The tension remained constantand the thickness of the dry layer increased as addi-tional water was lost by evaporation. The rate ofevaporation dropped abruptly with the appearance ofthe dry surface.

Calculations show that to remove water from thesmallest pores in a system of spheres 0.8 mm in di-ameter will require a tension of 24 cm of water. Theabrupt break in the evaporation curve at a tension of25 cm of water indicates that at this point the capil-lary pores of the system were emptied. Furthermore,at this point only 1.1% by weight of water remainedin the sand. Keen4 has shown that a uniform systemof closely packed spheres will hold 0.8% by weightof water when water has been. withdrawn from thecapillary pores of the systenj. Hence, we may con-clude that the abrupt transition in the evaporationcurve at a tension of 25 cm of water represents thetermination of capillarity in the system.

The moisture content of 1.1% at the terminationof the capillarity is the "field capacity" of the system.In this instance, field capacity occurred at a moisturepotential of 25 cm of water, or pF 1.4; and fieldcapacity was attained through evaporation ratherthan drainage. The evidence indicates definitely thatbelow field capacity the movement of water to thesurface of the soil was through the liquid phase underthe driving force initiated by surface tension. Of

4KEEN, B. A. The Physical Properties of the Soils. New York: Longmans Green & Co. 1931. (Page 120.)

122 SOIL SCIENCE SOCIETY PROCEEDINGS 1941

Evaj,

%oc

50

40

. 30

20

10

0C

>oration

surf age dry v~ water added'

surfac

Xx̂ /

///7

/

surf

:e i

•^

t

\

acei

noist

Isf* •f- tens:large

submerged

.

/»

0*

»•/

/

tjS +Q

.on requirer pores m

1

25 50

v̂ > •"•la

tension- en

••̂*

v

"•

.ension recfrom small

nsion requirtrger pores n

d to withe,ar top of

75X

Iraw watercontainer

. „ .J_ ————— ,

juired to v=st pores.

withdraw ws

2d to withdraw water;ar bottom of contair

from»

Jter

fromler.

Water evaporated • ———— •Tension o ——— o

Depth of drylayer * ——— *

^~-*—— -~~——— K—— —————

— *— ———

100 125 150 175 20

30

25

20

15

10

c.c. from free water surface - f(t).FIG. I.—Quantity of evaporation from, and moisture tension and depth of drying of, 270 grams of saturated sand of

uniform particle size with respect to time.

particular significance is the fact that the attainmentof field capacity was evidenced by the formation ofa dry layer at the surface of the sand.

EVAPORATION FROM SAND OF UNIFORM PARTICLE SIZE

AT MOISTURE CONTENTS BELOW FIELD CAPACITY

There is a question as to whether or not watermoves through the adsorbed water films that sur-round the grains of sand after the water has beenwithdrawn from the capillary pores. If such move-ment occurs, there should be a progressive decreasein the moisture content of the moist sand remainingbeneath the dry surface as the dry surface layer be-comes thicker.

Fig. 2 presents the quantity of water lost with re-spect to the thickness of the dry layer at the sur-face. It is apparent that the thickness of the dry layerwas directly related to the quantity of water lost over

the entire range of values studied. Hence, one mayconclude that the rate of water movement through thefilms surrounding the particles was negligible at fieldcapacity.

According to the laws of physical chemistry, therate of diffusion of a vapor through a porous,mediumis directly proportional to the vapor pressure gradi-ent. This may be expressed mathematically by theequation R = KG, where R is the rate of diffusion,G is the pressure gradient, and K is the constant ofproportionality that depends upon the porosity of themedium, the density of the vapor, etc. The rate Ris equal to AQ/At where AQ is the increment ofquantity lost in increment of time At.

The previous study relates thickness T of the drylayer to quantity Q of water evaporated by the equa-tion T = CQ, where C is the constant of propor-tionality. This may also be written Q = cT, where c

WOODRUFF: MOVEMENT AND EVAPORATION OF SOIL WATER 123

Depth-cm.

depth of container

= KG.

0 1 2 •Evaporation - c.c.

FIG. 2.—Depth of drying of sand with respect to quantity ofwater evaporated after the inception of. drying at thesurface.

is the reciprocal of C. By substitution, the originalequation becomes

cTt

Fig. i shows that the moisture potential of themoist sand beneath the dry layer remained con-stant as evaporation progressed. The vapor pres-sure of the air flowing over the system was alsoheld constant. Hence, the vapor pressure gradient

G = 2~ 1 may be written G = 7=, where D is a

constant representing the difference in vapor pres-sure, ¥2 of the moist sand, and Vj. of the dry air.Substituting this for the value of G in the preceding

equation gives — = -=- which simplifies to T2 = kt.

Values of T2 and of t are plotted in Fig. 3. Thedirect relation between these quantities' suggests thatthe rate of loss of water from the sand at field ca-pacity obeys the established laws for rate of vapordiffusion. The fact that the relation holds is evidencethat there was a transition from movement throughthe liquid phase to movement through the vaporphase at field capacity.

100

FIG. 3.—Relation between the square of the thickness of thedry layer of sand and the time.

EVAPORATION FROM SOILS INITIALLY SATURATEDA typical curve expressing the quantity of water

evaporated from the soil with respect to time is pre-sented in Fig. 4. The curve possesses two distinctlydifferent components. The rate of evaporation fromthe soil was constant and slightly less than the evap-oration from the free water surface during the initialstage. This stage was followed abruptly by a muchslower rate of evaporation that decreased regularlywith time. The transition from the initial stage toEvaporationc.c. from soil100

from free water surfacefrom soil into still airfrom soil into moving air

SO 100 150 200c.c. from free water surface- f(t). 300

FIG. 4.—Quantity of water evaporated from a saturated siltloam at different intervals of time.

124 SOIL SCIENCE SOCIETY PROCEEDINGS IQ4I

the final stage occurred at a moisture content thatcoincided with a moisture potential of pF 3 deter-mined by the pressure method described earlier bythe author.5 The surface of the soil was moist in ap-pearance before the transition. It began to dry at thepoint of transition and as evaporation continued, anabrupt dry moist boundary moved deeper into thesoil. Similar phenomena were observed when the soilwas.exposed to stagnant air. However, the rate ofevaporation in still air was one tenth of that in mov-ing air and the transition between the two portionsof the curve was less abrupt.

Similar results were obtained for other soils. Theoutstanding features of this study were the forma-tion of a dry crust on the surface of the soil and thetransition in the shape of the evaporation curve atpF 3 for fine sandy loams, silt loams, and granulatedclay loams. Interpreted in terms of the previous re-sults for sand of uniform particle size, the resultsindicate that in the graded system of pore sizes,water will move as a liquid by capillarity below pF 3and as a vapor by diffusion above pF 3. The mois-ture potential at pF 3 appears to be the upper limiton the moisture potential scale for the movement ofwater as a liquid; for it seems unlikely that all ofthe soils studied should possess a minimum poresize that would produce a moisture potential ofexactly pF 3 upon the removal of capillary waterfrom the system.

RELATION BETWEEN FIELD CAPACITY, FORMATION OF

A DRY SURFACE CRUST, AND pF 3

The observations of the preceding sections wereextended by similar methods to a larger group ofsoils by relating the moisture content of the soilbeneath the dry surface crust to the moisture contentat pF 3. The results are presented in Fig. 5.

Two coarse sandy soils, contained more waterbeneath the dry surface crust than at pF 3. Thisis in agreement with the results for the homogenoussized sand. Three granulated clay soils also containedmore water beneath the surface crust than they con-tained at pF 3. This may be attributed to the factthat conduction from granule to granule was re-stricted to the points of contact between granules.

Six highly weathered soils, identified by name,puddled upon wetting. These soils failed to show awell-defined boundary between the dry surface andthe moist soil beneath. Moisture continued to movein the liquid phase from these soils at moisture po-tentials above pF 3. Movement of water as a liquid

6WooDRUFF, C. M. Soil moisture and plant growth in relation

Field CapacityPercent.40

30

20

10

'•Decatur.»s . Lebanon

/ • OswegoClarksville

Sieved Shelby

10 20 30pF-3 - percent

40

FIG. 5.—Field capacity or the moisture content beneath thedry surface layer of representative soils in relation to themoisture content at pF 3.

at moisture potentials above pF 3 may occur in soilsdominated by pore sizes that are small enough forthe adsorptive forces to stabilize the contents of thepores at high curvatures of the air water interface.

The moisture content of the soil beneath the drysurface crust corresponded closely with that, at pF 3in 16 of the 27 soils studied. Results comparable tothese are to be expected for the majority of themedium-textured soils of good tilth, because thedominant group of pore sizes in these soils falls with-in the range where surface tension forces impose acondition of instability on the water within the pores.

SUMMARY AND CONCLUSION

The breaks in the evaporation curves of the soilsstudied suggest a limiting soil moisture content atwhich moisture movement through the liquid phaseceases. Water may be lost rapidly by evaporation, orby drainage, at moisture contents above the limitingmoisture content. But water may move only throughthe vapor phase at moisture contents below the limit-ing moisture content. Therefore it is lost at a muchslower rate than when movement through the liquidphase occurs.

Movement of water through the soil in the liquidphase ceases when the surface of a short column ofsoil becomes dry. Water loss may then occur only

to pF. Soil Sci. Soc. Amer. Proc., 5:36-41. 1940.

WOODRUFF : MOVEMENT AND EVAPORATION OF SOIL WATER

by evaporation beneath the surface, and the rate ofevaporation is related inversely to the square of thethickness of the dry layer of soil through which thevapor diffuses. Practically, this means that it willrequire four times as long to dry out the second inchof soil as it does to dry out the first inch.

The limiting moisture content corresponds to thefield capacity. It may occur at any moisture poten-tial. In coarse-textured soils, and in very fine, tex-tured structureless soils, it corresponds to the mois-

ture potential required to empty the pores. Thismoisture potential is a function of the size of thedominant group of pores. In soils of medium texture,such as in fine sandy loams, in silt loams, and ingranulated clay loams, the limiting moisture contentis determined by the instability of water in the liquidphase at a moisture potential of pF 3. Consequently,most agricultural soils which fall in this range oftextures exhibit a moisture potential of pF 3 at fieldcapacity.