Soil Sampling and Methods of Analysis - Part 7 pot

69.4 SOIL WATER DESORPTION AND IMBIBITIONSoil water desor ption and imbib ition curves char acterize the rel ationship betwee n soilvolumet ric water cont ent, uv [L 3 L 3 ] Chapter 72 t

VII SOIL WATER ANALYSES Section Editors: W.D Reynolds and G Clarke Topp

Chapter 69 Soil Water Analyses: Principles

and Parameters

W.D ReynoldsAgriculture and Agri-Food Canada Harrow, Ontario, Canada

G Clarke ToppAgriculture and Agri-Food Canada Ottawa, Ontario, Canada

Soil water analyses can be organized into two main groups: (i) analysis of storage propertiesand (ii) analysis of hydraulic properties Storage properties refer to the soil’s ability to absorband hold water, and these properties include water content, water potential, and waterdesorption and imbibition characteristics Hydraulic properties, on the other hand, refer tothe soil’s ability to transmit or conduct water, and these include saturated hydraulic conduct-ivity, unsaturated hydraulic conductivity, and various associated capillarity parameters such

as sorptivity, flux potential, sorptive number, and flow-weighted mean (FWM) pore eter These properties and their interrelationships are discussed in the following sections

diam-69.2 SOIL WATER CONTENTSoil water content can be defined on a gravimetric basis (mass of water per unit mass of drysoil) or on a volumetric basis (volume of water per unit bulk volume of dry soil), and it isexpressed either as a dimensionless ratio or as a percentage These two definitions are notequivalent, however, and it is consequently essential to specify the definition used whenreporting water content values It should also be noted that ‘‘bulk volume’’ of dry soil refers

to the dimensions of the soil sample just before the water volume determination and before

any soil disturbance Gravi metric water cont ent (i.e., mass water =mass dry soil) is related tovolumetric water cont ent (i e., volume water =bulk volume dry soil) via soil dry bulk density ,

rb (mg m  3 ) and pore water density , r w (mg m 3 ), accordi ng to the formul a

to a max imum value of 1 (pore space comple tely water -filled) When the soil pore space iscompletely water-fille d (relati ve saturati on ¼ 1), the soil volumet ric water cont ent is equal tothe soil porosity, n (por osity is defined as the total volume of soil pore space per unit bulkvolume of soil) Relativ e saturati on is frequently expre ssed as an ‘‘effect ive saturati on,’’ Se ,which include s the residua l soil water content , ur , that is the ‘‘im mobile’’ water remainin g

in air-dry soil and retained in small isolat ed pores Effective sat uration is defi ned

as Se ¼ ( uv  ur ) =( us  ur ) and ranges from a minimum v alue o f 0 at residua l saturati on(i.e., uv ¼ ur ) to a max imum valu e of 1 at complete saturation (i.e., uv ¼ us )

Water content measur ement tech niques are often classified as ‘‘direct ’’ or ‘‘indirect.’ ’ Directmethods usually alter the sample irrevocabl y by changing its water content and physi calcharacterist ics (i.e., they are ‘‘de structive ’’ methods ); and these methods involv e som e form

of remo val or separation of water from the soil matrix with a dir ect measur ement of theamount of water removed Separation of the water from the soil matri x may be achieved byheating (wate r vapor ization), by water repl acement with a solvent (w ater absor ption), or bychemical reac tion (water d isassociati on) Th e amo unt of water removed is then determin ed

by measuring the change in soil mass after heating, by collecting and condensing emittedwater vapor, by chemical or physical analysis of the extracting solvent, or by quantitativemeasurement of chemical reaction products The removal of water by heating is commonlyreferred to as the thermogravimetric technique (see Topp and Ferre´ 2002 for details) and it is

by far the most com mon of the dir ect methods (see Chapter 70) Indirec t methods measur esome physical or chemical property of soil that depends on its water content Theseproperties include the relative permittivity (dielectric constant), electrical conductivity, heatcapacity, hydrogen content, and magnetic susceptibility The indirect methods usually alter thesample minimally (or not at all) in that the water content and physical characteristics of thesample are not changed appreciably by the measurement (i.e., they are ‘‘nondestructive’’methods) However, the accuracy and precision of indirect methods depends to a large extent

on the accuracy and precision of the relationship between the measured property (e.g., tivity) and uv In Chapter 70, we limit discussion to the indirect methods that are based on relativebulk soil dielectric permittivity, as they are the most highly developed and versatile

permit-The electromagnetic (EM) methods discussed in Chapter 70 all arise from analyses based in

EM wave propagation or radio frequency (RF) circuits Measurement of soil water content

by these methods involves using the soil as an EM wave-propagating medium or as a resistor

or capacitor in a circuit The time-domain reflectometry (TDR), ground-penetrating radar(GPR), and remote radar (remote sensing) methods use the EM wave-propagation properties

of the soil, whereas the capacitanc e and impeda nce methods use the soil as a resist or o rcapac itor in a circuit.

The unique electrical properties of water (both pure water and soil pore water) form the basis

of soil water content measurements by EM wave propagation The relative dielectricpermittivity of water is generally more than an order of magnitude larger than that of othersoil components As a result, the bulk relative dielectric permittivity of soil ( «ra ) is almostentirely a function of the soil’s volumetric water content (uv ), with only a slight dependence onthe volume fraction of soil solids and the bulk soil electrical conductivity (Topp et al 1980).For each of the EM methods presented in Chapter 70, a measurement of «rais used to infer uv

A single relationship between «ra and uv for all soils does not yet exist because of thecomplex interactions among EM waves and soil components Many soils have very similarrelationships, however, and thus sufficient accuracy can usually be attained using only a few

‘‘quasigeneral’’ relationships, for example, mineral soils, organic soils, saline soils, etc.(Topp et al 1980; Topp and Ferre´ 2002) It has further been established that assuming alinear relationship between uv and ffiffiffiffiffiffi«

rap

is appropriate for most soil materials (Topp andReynolds 1998), and that this relationship is predicted by dielectric models employing athree-component mixture of soil solids, soil water, and soil air (Robinson et al 2005) Thus

EM methods are now considered highly reliable for measuring soil volumetric water content

Water content methods are described in Chapter 70 and include thermogravimetry (ovendrying), TDR, GPR, and a general description of the impedance and capacitance techniques.Thermogravimetry based on oven drying is usually considered the ‘‘benchmark’’ ofaccuracy and relevance against which other methods are assessed

69.3 SOIL WATER POTENTIALTotal water potential (ct) is classically defined as the amount of work (force distance)required to transport, isothermally and reversibly, an infinitesimal quantity of water from aspecified reference condition (pool of pure water at specified pressure and elevation) to thesystem under consideration (Or and Wraith 2002) It is usually more convenient for naturalporous materials, however, to consider ctas the amount of work required to transport wateraway from the material (i.e., remove water rather than add water), as most natural materialsare hydrophilic and thereby tend to absorb and retain water in a manner similar to that of apaper towel (nonswelling materials) or sponge (swelling materials)

Water potential is commonly expressed in units of energy per unit mass,Um (J kg1), energyper unit volume,Uv(Pa), or energy per unit weight,Uwt(m), with the latter two being by farthe most prevalent Conversion amongst the units is achieved using

where cm is the matric pote ntial, cp is the osm otic pote ntial, cp is the pressure pote ntial, and

cg is the gravit ational pote ntial Th e matri c potential is negativ e (c m  0) and arises from thevarious electrostat ic forc es in the soil matri x that attract water when the soil is u nsaturated The osmoti c potential is als o negativ e ( cp 0) and resu lts from disso lved mater ials (salts)and colloids, which lower the pore water’s activit y (free energy) below that of pure water The pres sure potential is posi tive ( cp  0), and is cause d by the hydros tatic pressure of thepore water overlying the measur ement point whe n the soil is sat urated The gravitationa lpotential ( cg ) arises from the action of the earth’ s g ravitational forc e field on the pore waterand can be either positive or negative depending on whet her the datum is b el ow or abovethe measuring point, r espect ively Campbe ll (1987) reviews the various techniques formeasuring matric potential and t he type of sensors e mployed; and t he readers arerecommended t o r efer to Passioura (1980) for a more detailed discussion of the m eaning

of matri c potent ial A reaso nable estim at e o f o smotic po tential c an be derived f rommeasurements of electrical conductivi ty corrected f or water c ontent (Gupta and H anks1972); howeve r, more reliable measures can b e obtained by extracting soil pore waterand m easuring the osmotic potential directly in a thermocouple psychrometer or

by usi ng the combine d pr es sur e c h ambe r a nd thermocouple psychrometer system ofCampbell (1987) W he n the matric, pressure, and gravitational p ot enti al s a re expressed

in unit s o f e nergy p er unit w ei ght (Uwt ), they are generally called ‘‘heads’’ rather thanpotentials, a nd they are e quival ent to t he vertical distance between the m ea surement

p o i n t ( e g , piezometer intake, tensiometer cup, etc.) and either the free surface water level(for matric and pressure heads) or the selected reference elevation or datum (for gravitationalheads) (Figure 69.1) Water flow can be induced by gradients in all four water potentials,although a gradient in osmotic potential requires the presence of a membrane that ispermeable to water but impermeable to selected solutes and colloids (Or and Wraith 2002).Methods for measur ing water pote ntial are describ ed in Chapter 71 and include thepiezometer method, the tensiometer method, resistance block methods, and selectedthermocouple psychrometer methods

(a)

Datum

(water table) Piezometric surface Piezometer riser pipe Manometer

Soil surface

Selectively permeable cup (to water, not air)

Tensiometer point of measurement

Well screen

Piezometer point of measurement

FIGURE 69.1 The operating principles of a piezometer (a) and a tensiometer (b) The piezometer

measures pressure potential (cp), and the tensiometer measures matric potential (cm)

69.4 SOIL WATER DESORPTION AND IMBIBITION

Soil water desor ption and imbib ition curves char acterize the rel ationship betwee n soilvolumet ric water cont ent, uv [L 3 L 3 ] (Chapter 72 through Cha pter 74), and pore watermatric head, cm [L] (Chapt er 71) The desorption curve (also know n as the water releasecharacteristic, water retention curve, and soil moisture characteristic) describes the decrease

in uvfrom saturation as cmdecreases from zero, whereas the imbibition curve describes theincrease in uv from dryness as cm increases from a large negative value (see Figure 69.2).The two curves generally have different shapes because of hysteretic effects (Hillel 1980);and when a partially drained soil is rewetted, or when a partially wetted soil is redrained, therelationship between uvand cmusually follows an intermediate and nonunique path betweenthe desorption and imbibition curves (see Figure 69.2) For this reason, the desorption curve

is often referred to as the ‘‘main drainage curve’’; the imbibition curve as the ‘‘main wettingcurve’’; and the intermediate curves as ‘‘scanning curves’’ (see Figure 69.2) When the soilhas a relatively uniform and narrow pore size distribution (e.g., structureless sandy soil),distinct ‘‘air-entry’’ and ‘‘water-entry’’ matric heads can occur on the desorption andimbibition curves, respectively (Figure 69.2) The air-entry head or value, ca [L], is thepore water matric head where the saturated soil (i.e., uv constant and maximum) suddenlystarts to desaturate as a result of decreasing cm; and the water-entry head or value, cw[L], isthe pore water matric head where an unsaturated soil suddenly saturates as a result ofincreasing cm Both ca and cw are negative, and typically, jcaj  2jcwj (Bouwer 1978).Also note that in Figure 69.2 that the saturated volumetric water content on the imbibitioncurve (i.e., ufs at cm¼ 0) is less than the saturated volumetric water content on thedesorption curve (i.e., us at cm¼ 0), which is a consequence of air entrapment in soilpores during the wetting process (Bouwer 1978) As implied above, soil water desorption

Imbibition or main wetting curve

Scanning curves

Desorption or main drainage curve

FIGURE 69.2 Desorption, imbibition, and scanning curves, u(c), for a hysteretic soil The arrows

indicate the direction of the drainage and wetting processes Note that the ated volumetric water content for the imbibition curve, ufs, is less than that for thedesorption curve, us, due to air entrapment upon rewetting Note also that thewater-entry matric head, cw[L], is greater (less negative) than the air-entry matrichead, ca[L]

satur-and imbib ition is a com plicated proce ss that is difficul t satur-and time-con suming to char acterize

in d etail Fort unately, it is usually not necessar y to measur e the scann ing curve s and Cha pter

72 through Chapter 74 conse quent ly focus on dete rmination of only the desor ption (maindrainage) curve and the imbibition (main wetting) curve

69.4.1 APPLICATION OFDESORPTION ANDIMBIBITION CURVES

The shape and magnitude of desorption and imbibition curves depends on the number andsize distribution of the soil pores, which in turn depends on texture, porosity, structure,organic matter content, and clay mineralogy Figure 69.3 gives schematic examples ofdesorption curves for a representative coarse-textured, unstructured soil (e.g., uniformsandy soil), and for a representative fine-textured soil (e.g., clayey soil) with and withoutstructure, where ‘‘structure’’ refers to the presence of aggregates, peds, and macropores(i.e., large cracks, root channels, worm holes, etc.) Note that the coarse-textured (sandy)soil retains less water than the fine-textured (clayey) soil (i.e., lower uv values), and it alsoreleases its water in a different manner (i.e., different curve shape) Note also thatsoil structure can increase the saturated water content (if the bulk density decreases) and itcan cause the wet-end of the desorption curve to be very steep relative to a structurelesscondition when aggregates, peds, and macropores are not present

Soil water desorption and imbibition curves are important for determining soil pore sizedistribution, for interpreting soil strength data, and for determining the transmission and

Structured clayey soil

Unstructured clayey soil Unstructured

FIGURE 69.3 Soil water desorption curves for a ‘‘representative’’ unstructured sandy soil, and

a representative clayey soil with and without structure us[L3L3] is the ated volumetric water content and c [L] is pore water matric head Note thatthe increase in us for the structured clayey soil relative to the unstructuredclayey soil implies a decrease in soil bulk density If bulk density remainsconstant, the presence of structure changes only the shape of the curve andnot the value of us

satur-storage of flu ids (liquids, gases ) in the soil profile The sizes of soil pores relevan t to thestorage and transmis sion of fluids are determin ed from desorption and imbib ition curves viathe Kelvin or ‘‘capi llary rise’’ equat ion Soil strengt h relat ionships , such as cone penet rationresistanc e and vane shear, are h ighly depend ent on the antecede nt soil water cont ent at thetime o f the mea sureme nt, and must therefore be related to the desorption and imbib itioncurve s befor e deta iled anal yses can be conduc ted With resp ect to water and solu tetransm ission, the desor ption and imbibition curve s are require d for defin ing the watercapacity relationship in the water transport (Richards) equation, and various solute sorption–desorption relationships in the solute transport (convection–dispersion) equation With respect

to water and air storage , the desorption and imb ibition curves are used to determ ine saturatedand field- saturated soil water content s, field capacity water cont ent, permanent wiltin g pointwater content , air capacity, and plant- available water capac ity These water =air storagepara meters and othe r quantitie s derived from these parameter s are defined and brieflydiscusse d in the follow ing sectio ns

69.4.2 WATER ANDAIRSTORAGEPARAMETERS

The volumetric water content, uv [L3L3], for a rigid soil (i.e., no shrinkage or swelling) isdefined by

whereVw [L 3 ] is the volume of soil water per unit bulk volume of dry soil, V b [L 3 ] (seeSect ion 6 9.2) When the soil is com pletely saturated (i.e., no entrappe d air), Vw¼ volume ofpore space and thus uv¼ us¼ soil porosity When the soil is ‘‘field-saturated’’ (entrappedair present), Vw< volume of pore space and uv¼ ufs< soil porosity, usually by 2–5percentage points (Bouwer 1978) For most field applications where wetting and dryingare involved, ufsis a more relevant measure of the maximum soil volumetric water contentthan us or porosity because entrapped air is almost always present

Field water capacity (more commonly known as field capacity, FC) is formally defined asthe amount of water retained in an initially saturated or near-saturated soil after 2–3 days offree gravity drainage without evaporative loss (Hillel 1980; Townend et al 2001) Forapplication purposes, however, FC is usually defined as the equilibrium volumetric watercontent, uFC, at a specified matric head, cFC For intact soil containing normal field structure,

cFC¼ 1 m is most often used, although values as high as cFC¼ 0:5 m have beenrecommended for wet soils with a shallow water table, and as low as cFC¼ 5 m for drysoils with a very deep water table (Cassel and Nielsen 1986) If the soil has been disturbedand repacked, use of cFC¼ 3:3 m is usually considered to provide uFC values that arecomparable to intact soil values

The permanent wilting point (PWP) is defined as the soil water content at which growingplants wilt and do not recover when the evapotranspirative demand is eliminated byproviding a water vapor–saturated atmosphere for at least 12 h (Hillel 1980; Romano andSantini 2002) Once the soil water decreases to the PWP value, plants are permanentlydamaged and may even die if water is not added quickly In this respect, the PWP watercontent also represents the amount of ‘‘plant-unavailable’’ water; i.e., water that is toostrongly held by the soil to be extracted by plant roots Although the true PWP can varywidely with plant species, plant growth stage, and soil type, it has been found that theequilibrium volumetric water content, uPWP, at the matric head, cPWP¼ 150 m, is a

suitable working definition (Soil Science Society of America 1997) This is because watercontent becomes relatively insensitive to matric head (i.e., water content is nearly constant)

in the cm 150 m range for most agricultural soils (Romano and Santini 2002)

Plant growth and performance is critically dependent on adequate supplies of air and water inthe root zone Convenient and popular measures of the soil’s ability to store and provide airand water for plant use are the so-called air capacity and plant-available water capacity Aircapacity (AC) is defined as

and proposed minimum values for adequate root-zone aeration are 0:10 m3 m3 for loamysoils (Grable and Siemer 1968), 0:15 m3 m3 for clayey soils (Cockroft and Olsson 1997),and about 0:20 m3 m3 for horticultural substrates (Verdonck et al 1983; Bilderback

et al 2005) Field soils that have AC values appreciably below these minimums aresusceptible to periodic and damaging root-zone aeration deficits Plant-available watercapacity (PAWC) is defined as

and it represents the maximum amount of water that a fully recharged soil can provide toplant roots This definition is based on the concept that soil water at cm> cFC drainsaway too quickly to be captured by plant roots, whereas water at cm< cPWP is held toostrongly by the soil to be extracted by the roots (compare PWP discussion) The proposedminimum PAWC for optimum plant growth and minimum susceptibility to droughtiness

is 0.20–0.30 m3 m3 (Verdonck et al 1983; Cockroft and Olsson 1997; Bilderback

These criteria are based on the finding that maximum production of crop-available nitrogen

by aerobic microbial mineralization of organic matter occurs when about 66% of the soilpore space in the root zone is water-filled, or alternatively, when 34% of the pore space is air-filled (Skopp et al 1990) The rationale for applying Equation 69.7 and Equation 69.8 torain-fed crops is that root-zone soils with these ratios are likely to have desirable water andair contents (for good microbial production of nitrogen) more frequently and for longerperiods of time (especially during the critical early growing season) than root-zone soils thathave larger or smaller ratios

69.4.3 DETERMINATION OFDESORPTION AND IMBIBITION CURVES

The generally accepted ‘‘ideal’’ for obtaining soil water desorption and imbibition curves is

to collect simultaneous field-based measurements of volumetric water content, uv, and

matric head, cm , in an undisturbed vertical profile under conditions of ste ady drainage(des orption) or steady wett ing (imbibi tion) Sever al approache s are availa ble for achievingthis (e.g., Bruce and Lu xmoore 1986), with the most p opular appro ach being the

‘‘instant aneous profile’’ method (see Chapter 83) Sever al factor s inhibit or complicatethe field-bas ed methods , however , includi ng com plex and poorly control led bou ndarycondi tions (e.g., varying water table dept h, strong and v arying temper ature gradient s);limit ed instrume ntation for determ ining cm (e.g., tensiom eters h ave a narr ow opera tingrange and often fail after a period of time ); diffi culty in maint aining cont inuous wett ing ordrainage throughout the soil profile (e.g., periodic rainfa lls can induc e hyst eretic effects);com plicated and labor-int ensive experiment al setu ps (e.g., installa tion of many pairs of uvand cm senso rs over a subst antial d epth range with minimum soil distur bance, equipme nt forappl ying large volumes of water to saturate the soil prof ile, com plex elect ronics and datalogging equipme nt for simul taneous and long-term monit oring of uv and cm , limit ed abili tyfor spat ial replication) ; and potential ly very long mea surement times (it can take severalweeks to months to obtain adequa te desorption or imbib ition curve ov er the require d soildept h because of slow wetting and drainage rates) As a result, exper imental ly dete rmineddesor ption and imbibition curves are usually ob tained in the labo ratory on relatively smallsoil cores or columns whe re uv and c m senso rs are more easily instal led and maintai ned, andwhe re initial and boundar y conditions can be prec isely defined and cont rolled Desorpt ionand imbibition curves can also be estimated from basi c soil data via pedotran sfer functions(see Chapter 84); from flo w experiment s, such as the evaporat ion method (see Chapter 8 1)and the instantane ous prof ile method (see Cha pter 83); or from inverse mode ling proce dures(Hopm ans et al 2002)

Labora tory determin ation of desorption and imbib ition curve s that are represen tative of fieldcondi tions require s (i) the collec tion of soil core s or colu mns that are large enough toadequa tely sample the ante cedent soil structure and (ii) use of collection, handl ing, andanalysi s proced ures that maint ain the soil structure intact Bouma (1983, 1985) sugges ts thatthe volume encompas sed by the core =colu mn shoul d include at least 20 soil structura l units(e.g., peds, worm hole s, abando ned root channe ls, etc.), which is especia lly important forthe cm > 3: 3 m range and if saturated hy draulic conduc tivity (see Chapter 75) is to bedeterm ined on the sam e sample For relative ly structure less sandy soils, the minimumreco mmended core =column inside diameter and length is on the order of 7.6 cm, whereasstruct ured loamy and clayey soils should use a core leng th and d iameter of at least 10 cm(Mc Intyre 1974) Th e sample s shoul d be collecte d whe n the soil is near its fie ld capacitywater content, uFC , whi ch gener ally makes the soil strong enough to resist comp action andstructural collapse during core=column insertion, but still plastic enough to prevent shatteringand breakage of peds Recommended procedures for the collection of minimally disturbed soilsamples are given in McIntyre (1974) and Chapter 80 Excavated soil cores should be trimmedflush with the ends of the sampling cylinder, capped to prevent damage of the core ends,wrapped in plastic to prevent evaporation, and transported to the laboratory in cushionedcoolers to minimize vibration-induced damage and large temperature-changes Sample storagebefore analysis should be in darkened facilities maintained at 0  C  4  C , w h ic h i s c ol d e no ug h

to inhibit faunal–bacterial–fungal–algal activity, but not so cold as to cause freezing and icelens formation

Soil water desorption–i mbibi tion methods are descr ibed in Cha pter 72 through Cha pter 74and include the tension table, tension plate, and pressure extractor methods (Chapt er 72), thelong colu mn method (Chapter 73), and the dew point psychrom eter method (Chapter 7 4).The appro ximate matric head ranges of thes e methods are com pared in Figure 69.4

69.5 SATUR ATED HYDRAULI C PROPERTIESThe saturated hydra ulic prope rties are used to descr ibe and predict water move ment inpermeab le porous materia l (e.g., soil, building fill, sand, rock, etc ) when the pore waterpressure (or matric) head in the materia l is grea ter than or equal to the water-en try valu e orair-entry value (see Sect ion 69.4 for explanat ion of water -entry and air -entry valu es).The saturated soil hydraulic properties of greates t relevan ce include sat urated hydrauli cconducti vity, field-sat urated hydra ulic conduc tivity, and the so-called capillarity parame terssuch as matric flux pote ntial, sorpti vity, sorpti ve number, Green–Am pt wetting frontpressure head, and FWM pore size and pore numb er Saturat ed hydraulic conductivit y,

Ks [LT  1 ], and field-sat urated h ydraulic conduc tivity, Kfs [LT  1 ], are measur es of the

‘‘ease’’ or ‘‘ability’ ’ of a permea ble porous medium to transm it water The Ks parame terapplies whe n the water -cond ucting pores in the porous med ium are comple tely water -filled(saturated), and the Kfs parameter a pplies when t he water-conducting por es containentrappe d or encapsul ated bubb les of air or gas (fiel d-saturat ed) Th e capillar ity parame tersmeasure various aspec ts of the suct ion or ‘‘capillary pull’’ that u nsaturated soil exer ts oninfiltrat ing water ; and measur ement or estimation of the soil’s capillar ity is usually require dwhen Ks or Kfs are measur ed in initially unsatura ted soil (e.g., soil above the water table) The Ks and K fs para meters are disc ussed belo w and the capillar ity parameter s are discusse d

in Section 69.6 (unsa turated hydra ulic properties )

The Ks and K fs parame ters are defined by Darcy’ s law, which may be written in the form

where q is the water flu x density thro ugh the porous medium (volume of water flo wingthrough a unit cros s-sectiona l area of porous medium per unit time ), i is the hydra ulic headgradient in the porous medium (di mensionle ss), and Ksat ¼ Ks or Kfs , depend ing on whe ther

Matric head, (m)

Dew point psychrometer

Pressure plate extractor

High tension table/plate

Low tension table/plate Long column

FIGURE 69.4 Approximate matric head ranges of the long column, tension table, tension plate,

pressure extractor, and dew point psychrometer methods for measuring tion and imbibition curves usis the saturated water content and uris the residualwater content These methods are described in Chapter 72 through Chapter 74

desorp-the porous med ium is comple tely saturated or field-satur ated, respect ively As implied byEquat ion 69.9, the dimens ions of Ksatare the same as those forq (i.e., volume of water perunit cross-sectional area of flow per unit time); however, these dimensions are usuallysimplified to length per unit time so that Ksat may be expressed in the more convenient(but physically incorrect) units of velocity (i.e., cm s1, cm s1, cm h1, m days1, etc.).TheKsatvalue is a constant when the porous medium is rigid, homogenous, isotropic, andstable; when in-situ biological activity such as earthworm burrowing and algal=fungalgrowth are negligible; and when the flowing water maintains constant physical and chemicalproperties (e.g., temperature, viscosity, dissolved air content, dissolved salt content, etc.) anddoes not chemically or physically interact with the porous medium The primary factorsdetermining the magnitude ofKsatinclude the physical characteristics of the porous mediumand the physical and chemical characteristics of the flowing water (discussed further below).

The physical characteristics of the porous medium affectingKsatinclude the size distribution,roughness, tortuosity, shape, and degree of interconnectedness of the water-conductingpores For soils,Ksat increases greatly with coarser texture (larger grain sizes), increasingnumbers of biopores (e.g., worm holes, root channels), and increasing structure (e.g.,aggregates, interpedal spaces, shrinkage cracks), as these factors increase the number ofwater-conducting pores that are relatively large, straight (i.e., low tortuosity), smooth,rounded, and interconnected Soils and other porous media that are coarse-textured,structured, and bioporous consequently tend to have largerKsat values than those that arefine-textured, structureless, and devoid of biopores In addition, texture, structure, andbiopores can interact in such a way that it is not uncommon for a fine-textured materialwith structure or biopores (e.g., a clay soil with shrinkage cracks or worm holes) to have asubstantially larger Ksat than a coarse-textured material that is devoid of structure andbiopores (e.g., single-grain sandy soil) An important implication of this texture–structure–biopore interaction is that the physical condition of the porous medium must be preserved bythe measuring technique in order for the measured Ksat value to be representative of theporous medium in its ‘‘natural’’ orin-situ condition

Hydraulic conductivity is inversely related to water viscosity, which is inversely related totemperature (Bouwer 1978, p 43) Consequently, the measured value ofKsat will increasewith the temperature of the water used; and an increase in water temperature from 10C to

25C will result in a 45% increase inKsat, all other factors remaining equal Temperatureeffects can be important if the water used in a field measurement differs greatly intemperature from that of the resident soil water or groundwater, or if laboratory measurements

of field samples (e.g., intact cores) are conducted at temperatures that differ greatly fromthe field temperature Precise measurements and comparisons of Ksat values shouldtherefore always be referenced to a specific water temperature, which is usually 20C (Bouwer

1978, p 43), as it yields a water viscosity of nearly 1 cP Note in passing that thetemperature of ‘‘deep’’ soil water and shallow groundwater is fairly constant and close tothe local mean annual air temperature, for example, about 10C at 408N–458N latitude(Bouwer 1978, p 378)

The concentration and speciation of dissolved salts in the water can affect Ksat throughswelling, flocculation, or dispersion of silt and clay within the porous medium, and throughthe creation or dissolution of precipitates TheKsatvalue will usually increase if silt and clayparticles are flocculated, or if precipitates are dissolved, as this tends to increase the size andinterconnectedness of water-conducting pores Alternatively, formation of precipitatesand swelling=dispersion of silt and clay particles will usually decreaseKsatthrough narrowingand plugging of pores Reduction inKsatmost commonly occurs in silt- and clay-rich soils

when the cat ionic speciat ion is change d o r the conce ntration of the residen t soil water

is diluted by incoming rainfa ll, irrigation water, or g roundwater The rel ative conce ntrations

of sodium, cal cium, and mag nesium in solu tion and sorbed onto the porous mediumexchange sites are partic ularly import ant in this respect (Bouw er 1978, p 44 ) In extrem ecases, such as when water low in dissolv ed sal ts (e.g., rainw ater) is introdu ced into salinesoil, the resu lting silt and clay dispersion can reduc e Ksat to virtually zero The water used formeasuring the Ksat of a natu ral porous medium should theref ore be either ‘‘na tive’’ waterextracted from the porous medium , or a laboratory ‘‘approx imation, ’’ which has about thesame major ion com position and conce ntrations as the native water Local munici pal tapwater is often an adequa te appro ximat ion to native soil water , although this shoul d always bechecked as som e muni cipal water treat ment facil ities can change major ion chem istryradicall y Distille d or deioniz ed water should never be used for measur ing the Ksat of anatural porous medium, as it will almost always induc e cla y swellin g or dispers ion of silt andclay particle s

Entrapped bubbl es tend to const rict or block the water-co nducting pores in a porous med ium

As a resu lt, Kfs (i.e., field-satur ated K sat ) is usually less than Ks (i.e., comple tely saturated

Ksat ) with the degree of reduc tion largely depend ent on the mec hanism responsi ble forbubble formati on Bub bles can becom e encap sulated in pores thr ough physi cal entrapment

of residen t air during wett ing of an initially unsatura ted porous medium (Bouw er 196 6); byaccumulat ion of bioga ses (e.g., methan e) as a result of micro bial activit y (R eynolds et al.1992); and by ‘‘ex solution’’ of dissolv ed air as a resu lt of change s in the temper ature orchemistr y of the pore water (B ouwer 1978, p 4 5) Air encap sulation as a resu lt of rapidwetting (e.g., p onded infiltration) often cause s Kfs to be on the order of 0: 5 Ks (Bouw er 1966;Stephens et al 1987; Con stantz et al 1988), while gradu al accumul ation of bioga ses andexsolved air can cause muc h grea ter reductions (Bouwer 1978; Reynol ds et al 1992)

Further inf ormation concerni ng the theoretic al basis and othe r aspec ts of Ks , Kfs , and theirassociated capillar ity parame ters can be obtaine d from Bouwer (1978) , Kooreva ar et al.(1983), Smith (2002) , Re ynolds and Elrick (2005), and references cont ained therein.Saturated hydraulic propert y methods are describ ed in Cha pter 75 through Cha pter 79 andChapter 84; and they include the const ant and falling head core methods (Chapter 75),selected const ant and falling h ead well permeame ter methods (Chapter 76), sel ected constantand falling head ring infiltr ometer methods (Chapt er 77), the auger hole method (Chapter

78), the piez ometer method (Chapter 79), and sel ected estimation methods (Chapt er 84)

69.6 UNSATURATED HYDRAULI C PROPERT IES

Unsaturat ed hydra ulic prope rties are used to descr ibe and predict water movemen t inpermeab le porous mater ial (e.g., soil, buildi ng fill, sand, rock, etc.) that is only partia llysaturated and has a pore water matric head that is less than the materia l’s air-en try valu e orwater-en try valu e (see Section 69.4 for expl anation of air-en try and water-en try valu es) Theunsaturated hydraulic properties of greatest relevance include unsaturated hydraulicconductivity, K(c) or K(u) [LT1], sorptivity, S(c) [LT1=2], sorptive number,a*(c) [L1], flux potential, f(c) [L2T1], FWM pore diameter, PD(c) [L], and the number

of FWM pores per unit area, NP(c) [L2] The K(c) or K(u) parameter quantifies theability of an unsaturated porous material to transmit water as a result of a hydraulic headgradient, while S(c) measures the ability of the material to imbibe water as a result ofcapillarity forces (Philip 1957) The a*(c) parameter, on the other hand, indicates therelative magnitudes of gravity and capillarity forces during unsaturated flow (Raats 1976),

while the f ( c) para meter relates to the ‘‘po tential’’ for water flow (G ardner 1958) ThePD( c) parameter represe nts the effective equival ent mean pore size conduc ting waterduring const ant head infiltrat ion, and NP( c) indicat es the number of PD( c) poresthat are active (Phi lip 1987) These parame ters and their interr elationshi ps are disc ussedbriefly belo w.

Vertica l water flo w in rigid, homo geneous, v ariably saturat ed porous mater ial (e.g., soil) can

be describ ed by (Rich ards 1931)

as a functi on of either pore water matric head, c [L], or volumetric water cont ent, u [L 3 L 3 ]

The K ( c) and K ( u) relat ionships depend strongl y o n the magnitude and shape of the porewater desorption–i mbibiti on relations hip, u( c) [L3 L 3 ], which itself describ es the change involumet ric water content with changi ng pore water matric head (Sect ion 69.4) As a result,the K ( c) and K ( u) relations hips decr ease from the Ksat max imum (Sect ion 69.5) as c and udecr ease from their respective max imum valu es at porous med ium saturati on (i e., c¼ 0 and

u¼ us ) Through their connec tion with the u( c) relations hip, K (c) and K ( u) depend on thenumb er and size distribut ion of the porous medium pores , which in turn depend on porosity,struct ure, textur e, organ ic matter content , and clay mineralo gy Unlik e u( c), however , K (c )and K ( u) also depend on pore mor phology parame ters such as tor tuosity, roughn ess,connec tivity, and continuit y Th ese various depend encies cause K (c ) and K (u) tochange by man y orders of magnitude over the range in c appl icable to plan t grow th(i.e.,  150 m  c  0)

Due to the extreme sensitivity of u nsaturated hydra ulic conduc tivity to pore size and poremorpho logy, the mag nitude and shape of the K ( c) and K ( u) relations hips change subst an-tial ly with the textur e and struct ure of the porous med ium Figur e 69.5 gives schemat icexampl es of K ( c) and K ( u) relat ionships for a repr esentativ e ‘‘sandy’ ’ soil, and for areprese ntative ‘‘loamy’ ’ soil with and without struct ure, where struct ure refers to thepres ence of aggre gates, peds, cracks, root channe ls, wor m hole s, etc For conven ience,the structure d loam was assumed to h ave the same u( c) rel ationship as the unst ructuredloam Note in these figures that for a rigid (nonswe lling) porous mater ial, K ( c) and K (u) aremax imum and constant whe n the mater ial is saturated, i.e.,

K ( c) ¼ K ( u) ¼ constant ¼ Ksat ; c c e , u¼ usat (69 : 11)

whe re Ksat [LT  1 ] is the saturated or field-sat urated hydraulic conduc tivity, ce [L] is the entry or water-entry matric head, and usat[L3L3] is saturated or field-saturated volumetricwater cont ent (see Sect ion69.4 and Sect ion 69.5) Note also that the near-sa turated hydraulic

air-conductivity relationship in a structured porous medium can change very rapidly (by orders

of magnitude) with only small changes in c or u, and that the hydraulic conductivity of afine-textured material with structure can be either greater than or less than the hydraulicconductivity in a coarse-textured material, depending on the value of c or u Textureand structure effects are also illustrated in theKsat values, where it is seen that theKsat ofthe sandy soil is two orders of magnitude greater than the Ksat of the unstructuredloam (texture effect), but two orders of magnitude less than theKsatof the structured loam(structure effect)

(a) Pore water pressure head, (cm)

(b) Volumetric water content, q (cm3 cm − 3 )

FIGURE 69.5 (a) Hydraulic conductivity, K (c), versus pore water matric (or pressure) head, c

and (b) hydraulic conductivity, K (u), versus volumetric water content, u, for arepresentative sandy soil (Sand), and a representative loamy soil with structure(structured loam) and without structure (loam)

The sorptiv ity para meter, S(c ) [LT ], is related to K ( c) and f( c) by (Phi lip 1957; Whi teand Sully 1987)

is thus controlled by the shape and magnitude of the K(c) relationship, as well as themagnitude of ci Figure 69.6 gives the S( c0 ) and f( c0 ) relations hips correspo nding to theK(c) (and u(c)) relationships for our three representative soils, and it is seen that S(c0) andf(c0) are essentially ‘‘subdued replicas’’ ofK(c) Note from Equation 69.12 and Equation69.13, however, thatS(c0)¼ f(c0)¼ 0 when u(c0)¼ u(ci) or when c0¼ ci; and thatS(c0)and f(c0) do not exist for positive pore water pressure heads (i.e., cp> 0)

If theK(c) relationship is represented by the Gardner (1958) exponential function

then Equation 69.13 becomes

Figure 69.7 a; and general ly speak ing, a( c0 ) increases as c0 increases, indicating an increas e

in the import ance of the gravit y com ponent of infiltrat ion rel ative to the capi llarity ent as the soil gets wetter Note, however , that the a (c0 ) relat ionships have com plex slopes,and the sand and unst ructured loam produce curve s with loca l maxima and min ima Thisoccurs because a( c0 ) is based on the expone ntial K ( c) functi on (i.e., Equation 69.14) ,whereas the actual K ( c) relationship s were not expone ntial , especia lly those for the sandand unstructur ed loam (see Figur e 69.5a) Generally speaking, the closer the K ( c) relation-ship is to a mono tonic expone ntial functi on (i.e., Equation 69.14), the closer the a(c0 )relations hip is to a single constant valu e Figure 69.7b compare s a*( c0 ) to a( c0 ) for the

compon-(a) Pore water pressure head, (cm)

(b) Pore water pressure head, (cm)

FIGURE 69.6 (a) Sorptivity, S(c), versus pore water matric (or pressure) head, c, and (b) matric

flux potential, f(c), versus pore water matric head, c, for a representative sandysoil (sand), and a representative loamy soil with structure (structured loam) andwithout structure (loam)

structured loam, where it is seen that a*(c0) diverges progressively for c0<50 cm.This occurred because K(ci)¼ K(260 cm) in this scenario, and the assumptionK(ci) K(c0) became progressively more incorrect as c0decreased, resulting in increasingerror in a*(c0) with smaller (more negative) c0 values The a*(c0) parameter (and rela-tionships based on the a*(c0) parameter) must consequently be used with caution whenK(ci) is not substantially less thanK(c0), such as might occur in very wet porous materials,

or in fine-textured materials whereK(c) does not decrease rapidly with decreasing c

(a)

(b)

Pore water pressure head, (cm)

Pore water pressure head, (cm)

a( )

a∗( )

FIGURE 69.7 (a) Alpha parameter, a(c), versus pore water matric (or pressure) head, c, for a

representative sandy soil (sand) and a representative loamy soil with structure(structured loam) and without structure (loam) and (b) alpha parameter, a(c), andsorptive number, a*(c), versus pore water pressure head, c, for the structuredloamy soil

Substituting Equat ion 69.16 into Equat ion 69.12 produces

The FW M pore diameter, PD( c0 ) [L], is defined as (Phi lip 1 987)

an inde x parameter that represe nts the mea n ‘‘wat er-conduc tiveness’’ o f the hydra ulicall yactive pores, rather than an actual pore size This is becau se the PD( c0 ) parameter is derivedfrom a flow measur ement (ass ociated with the mea surement of K ( c0 ); Eq uation 69.18 ), andmust conseq uently reflect in some way the combine d sizes, tortuos ities, roughn esse s, andconnectivi ties of all water -conduct ing pores at c¼ c0 (Reynol ds et al 199 7) Asso ciated withPD( c0 ) is the ‘‘concen tration’ ’ of pore sizes, NP( c0 ) (number of pores L 2 ), which may bederived from Po iseuille’s law for flow in smo oth, cylindr ical capillar y tubes (Philip 1987):

NP( c0 ) ¼ 128 mK (c0 )

where m [ML  1 T 1 ] is the dy namic viscosi ty of water and the othe r parame ters are asdefined above The NP( c0 ) parameter is an indicator of the number of hydra ulicall y activepores per unit area of infiltr ation surface , which have FWM d iameter, PD( c0 ) The relat ion-ships amo ng PD( c0 ), NP( c 0 ), and K (c0 ) for the struct ured loam soil are il lustrated in Figur e69.8, whe re it is seen that a two-order of mag nitude increase in flow-we ighted mea n porediameter , PD( c0 ), correspo nded to about a six- order of mag nitude increas e in K ( c0 ), andabout a four -order of magn itude decr ease in NP( c0 )

Equation 69.14 through Equation 69.19 also apply whe n mea suring saturat ed flow para eters in unsatura ted porous materia ls (see Se ction 6 9.5) In this case, c0 is at its maximumvalue in the equatio ns (i.e., c0 ¼ 0), and conse quently the K ( c), f( c0 ), a*( c0 ), u( c0 ), S(c 0 ),PD( c0 ), and NP( c0 ) relat ionships becom e max imum-val ued const ants, which are indi cated

m-by Ksat (i.e., Ks or K fs ), fm , a*, usat (i e., us or ufs ), S, PD, and NP, resp ective ly As men tioned

in Sect ion 69.5, the matri c flu x potential ( f m ), sorpti ve number ( a*), and sorpti vity ( S) aremeasures of the capillary suction =p ull or ‘‘c apillarit y’’ that unsat urated hydrop hilic porousmaterials exert on infiltrating water Mathematically, fm is the area under theK(c) curvebetween c¼ c0¼ 0 and c ¼ ci (Equatio n 69 13); and as a result, the magnitude of a

mater ial’s capill arity depend s on the shape and magnitude of the K ( c) curve, and on theantecede nt pore water matric head, ci Porous media that are coarse-tex tured, struct ured,biopo rous, or wet conse quently tend to have low er capillar ity (i.e., smaller area under the

K (c) curve) than porous media that are fine-text ured, struct ureless, dry, or devoi d ofbiopo res Furthermor e, all porous med ia (regard less of texture or struct ure) have zerocapillar ity (i e., fm ¼ 0) when they are saturated or field-satur ated b ecause under thatcondi tion, c0 ¼ ci ¼ 0 in Equation 69.13 If the K (c ) functi on is represe nted by Equat ion69.14, it can be shown that for porous materials at field capac ity or drier (Me in and Farrell1974; Scotter et al 1982; Reynolds et al 1985; see also Section 69.4):

a sat =f m  c 1f ; cf < 0 < a * (69 : 20)

whe re a* [L 1 ] is the maxim um sorptive number (for the mater ial in quest ion) and cf [L] isthe Green–Am pt wetting fron t matri c head (negative quantity) Nea r-zero cf (large a *)occur s prima rily in porous materia ls that are coarse-tex tured and =or highly struct ured and =orhighly biopo rous, while large negat ive cf (small a*) occurs pri marily in mater ials that arefine-text ured or structure less or devoi d of biopores When c0 ¼ 0, the S, fm , K sat , a* , and cfpara meters are related by

S¼ [g(ufs ui)fm]1=2¼

g(ufs ui)Ksat

a*

1=2

¼ [g(ui ufs)Ksatcf]1=2 (69:21)

where ufs [L3L3] is the field-saturated volumetric water content (Section 69.4), ui [L3L3]

is the initial or antecedent volumetric water content, and the other parameters are aspreviously defined Note that in Equation 69.21,S decreases to zero as ui increases to ufs,indicating (as expected) that field-saturated porous material has no ability to absorb or store

Flow-weighted mean pore diameter, PD( ) (mm)

FIGURE 69.8 Hydraulic conductivity, K (c), and number of FWM pores per unit area, NP(c),

versus FWM pore diameter, PD(c), for the structured loamy soil

additional water The PD and NP para meters (Equati on 69.18 and Equat ion 69.19, resp ively) are often used to quant ify tempor al and managem ent-induced change s in porousmedium structure as they relat e to water flow (e.g., Whi te et al 1992; Reynol ds et al 1995).

ect-In structure d porous mater ials, it is oft en important to disti nguish betwee n ‘‘matrix’’ flowparameter s and ‘‘mac ropore’’ flow para meters, given that macropo res (e.g., large cracks,worm hole s, abando ned root channels , large intera ggregate space s, etc.) can have a substan-tial eff ect on near-satur ated water flow and solu te transport Matri x pores are d efined as allpores that are small enough to remain water-f illed at a specified pore water matric head,

cmat [L], whe reas macropo res are pores that are too lar ge to remai n water -filled at cmat Thevalue of cmat is not yet agreed upon (i e., various values have b een propos ed such as  3,  5,

10 cm); howe ver, growin g exper imental evidence sugges ts that cmat ¼ 10 cm isappropri ate (Jarvis et al 2002), which corr esponds to an equi valent p ore diameter of0.3 mm accor ding to classical capillar y rise theo ry (Or and Wra ith 2002) Using thi scriterion, all p ores with equival ent diameter s  0:3 mm (c  cmat ¼ 10 cm) are matrixpores, whe reas thos e with equival ent diamet ers > 0: 3 mm (c > cmat ¼ 10 cm) are macro-pores The various ‘‘tot al porous medium’’ flow para meters described above (i.e., Equation69.11 through Eq uation 69.21 that appl y to all pore siz es) can be recast as matrix flowparameter s by simply restrictin g c0 to the range, ci < c0 < cmat Macr opore flow para m-eters can be simi larly defined by restrictin g c0 to the range , cmat < c0 < 0; however , thehydraulic conduc tivity relations hips must be rewritten as

Kp (c) ¼ K ( c)  K (c mat ); cmat  c  0 (69: 22)

Kp ( u) ¼ K ( u)  K [ u( cmat )]; u( cmat )  u  us (69: 23)

where the subsc ript ‘‘p’ ’ denot es the macropo re flo w doma in, and K (c ) and K (u) refer to thetotal porous med ium (i.e., both matrix pores and macropo res) As a result of thes edefinitions, the flow para meters in the matrix domain are at their max imum valu es whe n

c0 ¼ cmat ; wherea s the flow parameters in the macropo re doma in are either zero(Kp ( c) ¼ Kp ( u) ¼ f( c0 ) ¼ S( c0 ) ¼ 0) or undef ined (PD( c0 ) and NP( c0 )) when c0 ¼ cmat .Figure 69 9 and Figur e 69.10 illustr ate sel ected flow parameter relations hips for the matrix,macropo re, and total porous med ium flow doma ins in our repr esentativ e struct ured loam soil.Note in these figures that the matrix and total porous medium flow parame ters are coin cidentwhen c0  cmat because the macropo res are emp ty, and thus only the matrix pores are water-conducti ng Note also that the macropo re relat ionships produc e com plex patterns and mayhave values that are greater than, equal to, or less than the corr esponding matrix andtotal porous med ium v alues, depend ing on the v alue of c0

The p rimary physical and chemical factors affecting the above unsat urated flo w parame tersinclude porous med ium textur e and structure , pore water viscosi ty, the conce ntration andspeciatio n of dissolv ed salts in the pore water , and porous medium hydrop hobicity All ofthe unsatura ted flow parameter s are h ighly sensi tive to porous med ium texture and structure(compare Figur e 69.5 thr ough Figure 69.7), and hence measur ing tec hniques must preservethe porous med ium in its natural =in-situ =ante cedent condition to as great an exte nt aspossible The effects of pore water viscosity and dissolv ed salts on the u nsaturated flowparameter s are similar to thos e describ ed for saturated and field-satur ated hydra ulicconducti vity (see Sect ion 69.5) A hydrop hobic soil is nonw etting (i.e., it partially orcompletely repels water rather than attracts water), and this in turn impedes infiltrationbecause of reduced (or even negative) capillarity Soil hydrophobicity can be caused byaccumulation of certain naturally water-repelling organic constituents (such as pine tree

needles), or by extreme or prolonged drying (such as after a long drought or after a forestfire), which causes certain organic materials and mineral oxides lining the soil pores

to become partly or completely water-repellent Hydrophobicity reduces the capillarityparameters (i.e., f(c0), a(c0), a*(c0),S(c0), PD(c0), NP(c0)) relative to a hydrophilic(water-wetting) situation, all other factors remaining equal Although soil hydrophobicitycan be initially strong enough to prevent infiltration of even shallow-ponded water, it usuallybreaks down over time, allowing normal soil capillarity to eventually return Furtherinformation on soil hydrophobicity and its impacts on soil hydraulic processes and properties

FIGURE 69.9 For the structured loamy soil: (a) hydraulic conductivity, K(c), versus pore water

matric (or pressure) head, c, in the total soil (Kt(c)), matrix flow domain, (Km(c)),and macropore flow domain (Kp(c)) and (b) sorptivity, S(c), versus pore waterpressure head, c, for the total soil (St(c)), matrix flow domain (Sm(c)), and macro-pore flow domain (Sp(c))

can be found in Bauters et al (1998, 2000), Nieber et al (2000), and refere nces contain edtherein.

Unsaturat ed hy draulic prope rty methods are descr ibed in Chapter 80 through Cha pter 84 andinclude the laboratory tension infiltrometer (Chapter 80), the evaporation method (Chapter 81),the field tension infiltrometer (Chapter 82), the instantaneous profile method (Chapter 83), andselected estimation methods (Chapter 84)

1.2

PDp( ) PDm( ) PDt( )

Pore water pressure head, (cm)

20000

30000

40000

NPp( ) NPm( ) NPt( )

FIGURE 69.10 For the structured loamy soil: (a) FWM pore diameter (PD), versus pore water

matric (or pressure) head, c, in the total soil (PDt(c)), matrix flow domain(PDm(c)), and macropore flow domain (PDp(c)) and (b) number of FWM poresper unit area, NP, versus pore water pressure head, c, in the total soil (NPt(c)),matrix flow domain (NPm(c)), and macropore flow domain (NPp(c))

Bauters, T.W.J., DiCarlo, D.A., Steenhuis, T.S.,

and Parlange, J.-Y 1998 Preferential flow in

water repellent sands Soil Sci Soc Am J 62:

1185–1190

Bauters, T.W.J., Steenhuis, T.S., DiCarlo, D.A.,

Nieber, J.L., Dekker, L.W., Ritsema, C.J., Parlange,

J.-Y., and Haverkamp, R 2000 Physics of water

repellent soils.J Hydrol 231–232: 233–243

Bilderback, T.E., Warren, S.L., Owen, J.S.,

and Albano, J.P 2005 Healthy substrates need

physicals too!HortTechnology 15: 747–751

Bouma, J 1983 Use of soil survey data to

select measurement techniques for hydraulic

conductivity.Agric Water Manage 6: 177–190

Bouma, J 1985 Soil variability and soil survey

In J Bouma and D.R Nielsen, Eds.Proceedings

of Soil Spatial Variability Workshop PUDOC,

Wageningen, The Netherlands, pp 130–149

Bouwer, H 1966 Rapid field measurement of

air-entry value and hydraulic conductivity of soil

as significant parameters in flow system analysis

Water Resour Res 2: 729–738

Bouwer, H 1978 Groundwater Hydrology

McGraw-Hill, Toronto, ON, Canada

Bruce, R.R and Luxmoore, R.J 1986 Water

retention: Field methods In A Klute, Ed.,

Methods of Soil Analysis, Part I—Physical and

Mineralogical Methods 2nd ed American

Soci-ety of Agronomy, Madison, WI, pp 663–683

Campbell, G.S 1987 Soil water potential

meas-urement In R.J Hanks and R.W Brown, Eds.,

Proceedings of International Conference on

Measurement of Soil and Plant Water Status,

Vol 1 Logan, UT, July 1987, pp 115–119

Cassel, D.K and Nielsen, D.R 1986 Field

cap-acity and available water capcap-acity In A Klute,

Ed.,Methods of Soil Analysis, Part 1—Physical

and Mineralogical Methods 2nd ed American

Society of Agronomy, Madison, WI, pp 901–926

Cockroft, B and Olsson, K.A 1997 Case study

of soil quality in south-eastern Austrialia:

management of structure for roots in duplexsoils In E.G Gregorich and M.R Carter, Eds.,Soil Quality for Crop Production and EcosystemHealth Developments in Soil Science, Vol 25.Elsevier, New York, pp 339–350

Constantz, J., Herkelrath, W.N., and Murphy, F

1988 Air encapsulation during infiltration SoilSci Soc Am J 52: 10–16

Gardner, W.R 1958 Some steady-state solutions

of the unsaturated moisture flow equation withapplication to evaporation from a water table.Soil Sci 85: 228–232

Grable, A.R and Siemer, E.G 1968 Effects

of bulk density, aggregate size, and soil watersuction on oxygen diffusion, redox potentials,and elongation of corn roots Soil Sci Soc Am.Proc 32: 180–186

Gupta, S.C and Hanks, R.J 1972 Influence ofwater content of electrical conductivity of thesoil.Soil Sci Soc Am Proc 36: 855–857.Hillel, D 1980 Applications of Soil Physics.Academic Press, Toronto, ON, Canada

Hopmans, J.W., Simunek, J., Romano, N., andDurner, W 2002 Simultaneous determination

of water transmission and retention properties:inverse methods In J.H Dane and G.C Topp,Eds., Methods of Soil Analysis, Part 4—Physical Methods, Soil Science Society ofAmerica, Madison, WI, pp 963–1004

Jarvis, N.J., Zavattaro, L., Rajkai, K., Reynolds,W.D., Olsen, P.-A., McGechan, M., Mecke, M.,Mohanty, B., Leeds-Harrison, P.B., and Jacques, D

2002 Indirect estimation of near-saturatedhydraulic conductivity from readily available soilinformation.Geoderma 108: 1–17

Koorevaar, P., Menelik, G., and Dirksen, C 1983.Elements of Soil Physics Elsevier, New York,

228 pp

McIntyre, D.S 1974 Soil sampling techniquesfor physical measurements In J Loveday, Ed.,Methods for Analysis of Irrigated Soils Tech

Commun No 54, Commonwealth Agricultural

Bureau, Australia, pp 12–20

Mein, R.G and Farrell, D.A 1974 Determination

of wetting front suction in the Green–Ampt

equation.Proc Soil Sci Soc Am 38: 872–876

Nieber, J.L., Bauters, T.W.J., Steenhuis, T.S.,

and Parlange, J.-Y 2000 Numerical simulation of

experimental gravity-driven unstable flow in water

repellent sand.J Hydrol 231–232: 295–307

Olness, A., Clapp, C.E., Liu, R., and Palazzo, A.J

1998 Biosolids and their effects on soil properties

In A Wallace and R.E Terry, Eds., Handbook

of Soil Conditioners Marcel Dekker, New York,

pp 141–165

Or, D and Wraith, J.M 2002 Soil water content

and water potential relationships In A.W Warrick,

Ed.,Soil Physics Companion CRC Press, Boca

Raton, FL, pp 49–84

Passioura, J.B 1980 The transport of water from

soil to shoot in wheat seedlings.J Exp Bot 3:

1161–1169

Philip, J.R 1957 The theory of infiltration 4:

Sorptivity and algebraic infiltration equations

Soil Sci 84: 257–264

Philip, J.R 1987 The quasilinear analysis, the

scattering analog, and other aspects of infiltration

and seepage In Y.S Fok, Ed., Infiltration,

Development and Application Water Resources

Research Centre, Honolulu, HI, pp 1–27

Raats, P.A.C 1976 Analytical solutions of a

simplified flow equation Trans ASAE 19:

683–689

Reynolds, W.D., Bowman, B.T., Drury, C.F.,

Tan, C.S., and Lu, X 2002 Indicators of

good soil physical quality: density and storage

parameters.Geoderma 110: 131–146

Reynolds, W.D., Brown, D.A., Mathur, S.P., and

Overend, R.P 1992 Effect of in-situ gas

accumu-lation on the hydraulic conductivity of peat.Soil

Sci 153: 397–408

Reynolds, W.D., Bowman, B.T., and Tomlin,

A.D 1997 Comparison of selected water and

air properties in soil under forest, no-tillage,

and conventional tillage In J Caron, D.A

Angers, and G.C Topp, Eds., Proceedings of3rd Eastern Canada Soil Structure Workshop.Universite´ Laval, Sainte-Foy, Quebec, Canada,

pp 235–248

Reynolds, W.D and Elrick, D.E 2005 ment and characterization of soil hydraulicproperties In J Alvarez-Benedi and R Munoz-Carpena, Eds., Soil–Water–Solute ProcessCharacterization: An Integrated Approach CRCPress, Boca Raton, FL, pp 197–252

Measure-Reynolds, W.D., Elrick, D.E., and Clothier, B.E

1985 The constant head well permeameter: effect

of unsaturated flow.Soil Sci 139: 172–180.Reynolds, W.D., Gregorich, E.G., andCurnoe, W.E 1995 Characterization of watertransmission properties in tilled and untilledsoils using tension infiltrometers Soil Till Res.33: 117–131

Richards, L.A 1931 Capillary conduction ofliquids in porous mediums.Physics 1: 318–333.Robinson, D.A., Jones, S.B., Blonquist, J.M Jr.,and Friedman, S.P 2005 A physically derivedwater content=permittivity calibration model forcoarse-textured, layered soils.Soil Sci Soc Am J.69: 1372–1378

Romano, N and Santini, A 2002 3.3 WaterRetention and Storage, 3.3.3 Field In J.H Daneand G.C Topp, Eds., Methods of Soil Analysis,Part 4—Physical Methods Soil Science Society

of America, Madison, WI, pp 721–738.Scotter, D.R., Clothier, B.E., and Harper, E.R

1982 Measuring saturated hydraulic conductivityand sorptivity using twin rings.Aust J Soil Res.20: 295–304

Skopp, J., Jawson, M.D., and Doran, J.W 1990.Steady-state aerobic microbial activity as afunction of soil water content.Soil Sci Soc Am J.54: 1619–1625

Smith, R.E 2002.Infiltration Theory for logic Applications Water Resources Monograph

Hydro-15, American Geophysical Union, Washington,

DC, 212 pp

Soil Science Society of America 1997.Glossary

of Soil Science Terms Soil Science Society ofAmerica, Madison, WI

Stephens, D.B., Lambert, K., and Watson, D 1987.

Regression models for hydraulic conductivity

and field test of the borehole permeameter.Water

Resour Res 23: 2207–2214

Topp, G.C., Davis, J.L., and Annan, A.P 1980

Electromagnetic determination of soil–water

content: Measurement in coaxial transmission

lines.Water Resour Res 16: 574–582

Topp, G.C and Ferre´, Ty P.A 2002 3.1 Water

content In J.H Dane and G.C Topp, Eds.,

Methods of Soil Analysis, Part 4—Physical

Methods Soil Science Society of America,

Madison, WI, pp 417–545

Topp, G.C and Reynolds, W.D 1998 Time

domain reflectometry: a seminal technique for

measuring mass and energy in soil Soil Till

Res 47: 125–132

Townend, J., Reeve, M.J., and Carter, A 2001

Water release characteristic In K.A Smith and

C.E Mullins, Eds., Soil and EnvironmentalAnalysis: Physical Methods, 2nd ed MarcelDekker, New York, pp 95–140

Verdonck, O., Penninck, R., and De Boodt, M

1983 Physical properties of different horticulturalsubstrates.Acta Hortic 150: 155–160

White, I and Sully, M.J 1987 Macroscopicand microscopic capillary length and time scalesfrom field infiltration Water Resour Res 23:1514–1522

White, I., Sully, M.J., and Perroux, K.M 1992.Measurement of surface-soil hydraulic properties:disk permeameters, tension infiltrometers, andother techniques In G.C Topp, W.D Reynolds,and R.E Green, Eds.,Advances in Measurement

of Soil Physical Properties: Bringing Theory intoPractice SSSA Special Publication No 30, SoilScience Society of America, Madison, WI, pp.69–103

Chapter 70 Soil Water Content

G Clarke ToppAgriculture and Agri-Food Canada Ottawa, Ontario, CanadaG.W ParkinUniversity of Guelph Guelph, Ontario, Canada

Ty P.A Ferre´

University of Arizona Tucson, Arizona, United States

70.1 INTRODUCTION

Water in soil is a vital link in the hydrological cycle that controls exchange with theatmosphere above and with the groundwater below Water in soil acts both as a lubricantand as a binding agent among the soil particulate materials, thereby influencing the structuralstability and strength of soil and geologic materials The high heat capacity of water causes

a moderation of diurnal and seasonal temperature cycles at the soil surface Chemically,water serves as the transport agent for the dissolved inorganic chemicals and suspendedbiological components that are involved in the processes of soil development and degrad-ation Biological production from soil, either as forest products or agricultural crops, isinfluenced primarily by water availability The measurement of soil water content then

is important directly for quantifying water balance, for estimates of plant water status, andfor characterizing most soil physical, chemical, and biological processes

The measurement of soil water content has undergone revolutionary advancements in the last

20 years From having gravimetric sampling and neutron moderation as the primary fieldmethods in the early 1980s, we now have numerous options, such as time-domain reflecto-metry (TDR), capacitance (and impedance) devices, ground penetrating radar (GPR), air-borne=satellite active radar, and passive microwave methods (Gardner et al 2001; Topp andFerre´ 2002) These five newer methods are all based on electromagnetic (EM) measure-ments Information on EM properties of soil and their use in soil water content measurementscan be found in Topp et al (1980), Ferre´ and Topp (2002), and Topp and Reynolds (1998).All of the EM methods make use of the high relative permittivity (dielectric constant) of the

water (80) in soil com pared with the permi ttivities of the other soil component s, whi ch rangefrom one for air to 3–5 for typical soil solids Due to this contras t, methods that mea sure thebulk dielec tric permi ttivity o f soil are effect ive for the measurem ent o f volum etric watercontent A selection of EM methods is the focus of this chapt er, as these offer a variet y ofsample geom etries and spat ial coverage, are minim ally site disr uptive, collect data digita llyallowing real- or near real-time information, and measure on volumetric basis directly InSection 69.2 the basic water cont ent parame ters and expre ssions for water cont ent aredefined, such as volumetric, gravimetric, and degree of saturation In addition, the principlesbehind the use of EM methods, including how dielectric permittivity relates to water contentappear in Section 69.2.

70.2 GRAVIMETRIC WITH OVEN DRYINGThe thermogravimetric method is conceptually simple Initially, a moist soil sample isweighed The sample is then oven dried at 105C and reweighed The gravimetric watercontent is defined as the ratio of the mass lost, attributed to water initially present in thesample, to total mass of the fully dried soil The method is apparently straightforward and iscommonly thought to yield absolute results In fact, this is not so for several reasons Water isretained by the components of the soil at a wide range of energy levels and there is noabsolute time at which the soil reaches a ‘‘dry’’ state when maintained at 105C Soilsamples continue to decrease in mass slowly at 105C for many days (Gardner 1986) Inaddition, many soil samples contain organic materials, some of which are volatile at 105C,

so some of the decrease in mass may be due to volatilization of components other than water.Finally, there is the problem of temperature control Although the drying ovens in commonuse in most soil laboratories can maintain temperatures in the range of 100C–110C withcareful adjustment, temperatures within the oven vary depending on the location in the ovenchamber Given that the actual temperature of the soil sample is not measured, this variabi-lity can lead to differential heating among soils placed in the same oven for the same amount

of time In spite of these imperfections, however, the oven-drying method is a commonlyused and convenient method to obtain a good estimate of soil water content The use ofmicrowave ovens is not as rigorously standardized, as in the case of incandescent heatingovens A more complete discussion of the procedures and limitations of the gravimetricmethod are given in many standard texts (e.g., Topp and Ferre´ 2002)

70.3 TIME-DOMAIN REFLECTOMETRY

It was just 30 years ago that TDR was first applied to measurement in soil and earth materials(Davis and Chudobiak 1975) Since those first measurements, TDR has been used to measurewater content at many scales and under a broad range of conditions (Topp and Reynolds1998; Robinson et al 2003a), and has become a standard method of water content measure-ment The popularity of the method for soil=environmental monitoring and research arisesfrom a combination of its accuracy in a wide range of soils and its relative ease of usecompared with many other available techniques TDR provides real-time,in-situ soil watercontent measurements Measurement systems can be multiplexed and data-logged, allowingfor remote automated monitoring For most soils, the accuracy of measurements of volu-metric water content change is within 0:02 m3 m3 without the need for soil-specificcalibration, and better absolute water contents can be achieved with calibration There isconsiderable flexibility in the design and placement of TDR probes, allowing users to modifywater content measurement networks to conform to the requirements of any specific study.Finally, because TDR determines the volumetric water content, the data are directly applicable

to hydrolo gic water bala nce analyses with no need for the measur ement of suppor ting soilpara meters such as bulk density

70.3.1 M ATERIAL AND INSTRUMENTS

TDR instru mentati on consists of four basi c com ponents: a timing circuit, a pulse generat or, asamplin g receiver, and a display or recording device It is the pulse gener ator, whichlaun ches a pulse or wave whose travel is analyzed Most com mercial instru ments have allcom ponents in a sing le unit, which also perform s anal yses of the TDR traces , display ing andreco rding the interp reted water contents Any of these instru ments can be used to measurewater cont ent as long as they provide stable low noise read ings with high time base accur acy,typicall y with a pulse transition time of 0:2 ns (Hook and Livi ngston 1995)

For custom analyses invol ving highly prec ise applications, the capabil ity to reco rd the entirewav eform is neces sary An addi tional u seful feature is the ability to provide auto mated watercont ent analysis The capab ility of display ing the actual TDR trace, and manual ly interp ret-ing it, is very usefu l for assuring that the instru ment is opera ting prope rly and that theauto mated interp retation is reas onable Con nection of the TDR instru ment to a multipl exerfor seque ntial mea sureme nt at a numb er of locations increases greatly the efficienc y o f datacollec tion possibili ties

The initial and still widel y used TDR instru ment is the po rtable cabl e tester (Model 1502

B or C, Tektroni x) In this instrume nt, the trace is display ed and anal yses may be performed,and recorded manual ly, or data may be recorded and anal yzed digi tally on a PC (Or et al.2003) The cable tester and a PC were incorporated into a number of custom systems designed

to achieve automated TDR trace analysis, and multiplexing (Ferre´ and Topp 2002) Mostcom mercial instru ments now offer auto mated analysi s as a p art of the basic instrume nt withthe multipl exing capab ility as an o ption So me of the featur es of commerci ally availa bleinstru ments are li sted in Table 70.1

The basi c elements of a TDR probe are conduc tive component s, often para llel metal lic rods,which act as wav e-guides , and the soil materia l in whi ch the wav e or signal propa gates(Fig ure 70.1) Currentl y, the most common soil probe s are of the balanced pair transm issionline, consist ing o f two para llel rods, with rods that vary in length, depend ing on themeasur ement require ment, from 0.1 to 1.0 m and with probe separat ions from 0.01 to0.1 m (Topp and Dav is 1985) Th e minim um practical probe length for standar d equipme nt

is 0.1 m The upp er limit on length of probe is largely dete rmined by elect rical conduc tivity,clay content, and maximum water content expected Altho ugh no firm guide can be offe red,Dalt on (1992) show ed that probe lengths will have to b e reduc ed to 0.2 m in clayey soil of

EC > 0:1 S m 1 Coa ted probe s, disc ussed later, overcom e this limitation to some exte nt.Zegel in et al (1989) introduced multipr onged probe s where one prong or wire is cent rallylocate d and variable numbers of prongs are loca ted circumfe rentially around the central wire.These confi gurations, even with only two oute r prongs , act electri cally to emulate a coaxialtransmission line and result in a marginally improved TDR reflection The extra rods,however, make for greater installation difficulty and associated soil disturbance than from

a parallel pair The configuration of the wave-guide or probe determines the extent and shape

of the measured soil sample Earlier experimental and theoretical analyses have strated that the distribution along the length of probes has an effect, which is represented by alinear-weighted average (Hook and Livingston 1995) Specific refinements may be requiredfor layered soils (Robinson et al 2003b)

demon-TABLE 70.1 TDR Instruments, Listing Options and Capabilities

with shorting diodes

For the lateral distribution, the situation is more complex Knight (1992) and Knight et al.(1994) examined theoretically the spatial weighting function for parallel pair and multiwireacoaxial probes inserted in a medium of nearly uniform permittivity The analytical expres-sions and approaches from Knight (1992) form the basis for probe design specifications andfor evaluation of probe performance For example, Knight (1992) showed that the ratio of thewire or prong spacing to the wire diameter in a soil probe is an important geometricdescriptor of all TDR probes and should be considered for design and installation purposes.Knight (1992) proposed that the ratio of wire spacing to wire diameter should not exceed 10.

It is reasonable that the wire diameter should be at least 10 times the representative pore size

or particle diameter to provide sensible averages One important finding of this analyticalinvestigation is that the sample area of TDR is independent of the water content of themedium We have found that 6 mm diameter rods spaced at 50 mm have worked well in avariety of studies in tilled and untilled agricultural soil Many other probe configurationshave come into current use and these can be used successfully with due consideration of thelimitation applying to each probe type

70.3.2 PROCEDURE

The TDR method is straightforward but varies for different types of applications such aslaboratory or field; surficial or at depth; point specific or spatially referenced; and so on

Insert Soil Probe or Transmission Line into the Soil Sample

The installation of TDR probes is also important for high-quality measurements Airgaps around the probes can cause erroneously low water content measurements.However, Knight et al (1997) and Ferre´ et al (1998) applied a numerical analysis to show

FIGURE 70.1 A collection of TDR from a limited number of suppliers The numbers indicate the

sources of those shown as: 1 and 2 are from Soilmoisture Equipment Corp.; 3, 4,and 5 are from Environmental Sensors Inc (ESI); 6 are custom design developed inour laboratory

that partia l air gaps, surroundi ng only a fraction of the probe perim eter, will not adversel yaffect the mea sured relative permi ttivity Rods of the probe should be installe d in para llel;however , mi nor deviations from parallel alignm ent will not lea d to significa nt errorsunless the rods com e into cont act with each other Installa tion b y inse rtion of nonparal lel

or poorl y aligned rods or probe s may lead to air gaps along the rods; this shoul d be avoi ded

at all times Care shoul d be take n to minimize disturbance of the soil whe n inserting therods, especia lly in compres sible media In laboratory and near -surface field measur ements ,

it is import ant that the cross-s ection and length of the probe be chose n so that the EMfield associa ted with the TDR signal is contain ed within the soil sample (Knigh t 1992) Inaddition, the max imum probe leng th is limit ed by exces sive conducti ve loss in the soil

Connec t Prob e to TDR Instrume nt Usin g Coax ial Cable and Initia te Signal

Transm ission an d Recovery of the TDR Wavefor m

The length o f cable connec ting probe, and instru ment is best limit ed to 25 m to achieveaccepta ble signal-to -noise ratio and prevent excessi ve sign al attenua tion Some instru mentshave introdu ced compensa tion for sign al loss due to cable leng th allowing the use of grea tercable lengths Cho ice of acce ptable cable length should be based on signal qual ity from themeasurem ents in the wettest , most conduc tive condi tions The use of multipl exers introdu cesadditional signal deteri oration and may restrict additional ly the separ ation betwee n probeand instru ment

Analyze the Wavefo rm to Determine the Time of Travel of the Signal in the Soil,Which Serves to Determine the Relative Permittivity

Of interest for water content determination is the two-way travel time of the TDR signal inthe soil in and surrounding the probe Two times are measured; the time of arrival of signalreflected from the probe-to-soil interface (t1 in Figure 70.2) and the time of arriva l of thesignal reflected from the end of the probe (t2in Figure 70.2) The TDR waveform in Figure70.2 shows the recommended way of estimating the two times The intersection of tangentiallines on either side of the identifying signal reflection is the most precise indication of thedesired times The time difference (t2 t1) is a measure of the two-way travel time for thepulse or wave along the length of the rods For some probes, the choice of where to pickt1may be difficult under some conditions Robinson et al (2003b) present a method for probecalibration using only water and air, claiming this to be highly accurate

Periodic Measurements in a Reference Liquid to Detect Instrument Drift

and Malfunction

Reference liquids of known dielectric permittivity are useful to check measurement ability and instrument drift We have used repeated measurements with the TDR probeimmersed in isopropyl alcohol or water and recorded at hourly intervals during fieldmeasurement It is important to ensure that the container is sufficiently large to contain thesignal entirely within the reference fluid

repeat-70.3.3 CALCULATIONS

For many instruments calculations of volumetric water content are made within the ment The simple calculation sequence given here applies to those instruments where traveltime measurement is made explicit

instru-Convert Travel Time to Relative Permittivity

The travel time (t2 t1) from second section, p 944, is converted to propagation velocityand then to apparent relative permittivity as follows:

(t2 t1)¼2L

v ¼

ffiffiffiffiffiffi

«rap

30 35 Time (ns)

40 45 25

1.2 1.1 1.0 0.9 0.8

Probe specific dip

Vf = 0.78 V0

20 40 60 80 100 120 Time (ns)

t1

t2

45 50 Time (ns)

55 60 40

1.2 1.1 1.0 0.9 0.8

V0

0.7 0.6

FIGURE 70.2 Two TDR curves from 20 cm probes in silty clay loam soil The soils are at similar

water content but the soil solution is more conductive in (b), giving a smaller returnreflection and resulting lower Vf In (a) uv¼ 0:304 m3m3and s0¼ 57 mS m1and in (b) uv¼ 0:271 m3m3and s0¼ 95 mS m1 (From Topp, G.C and Ferre´,

Ty P.A., in D Hillel et al (Eds.), Encyclopedia of Soils in the Environment, Vol 4,Elsevier, Oxford, UK, 2004, 174–181 With permission.)

uv ¼ 0:115 ffiffiffiffiffiffi«

rap

Soils where it may become advisable to develop a specific calibration include those high inclay and=or salt and having organic matter above 0.05 kg kg1 The effect of high claycannot be made specific as its effect depends on grain size and mineralogy of the clay Thehigh clay, salt, and organic matter alter the slope (0.115) of Equation 70.2 and mayintroduce curvature as well (Topp et al 2000) Dense soils may influence the intercept(0:176) in Equation 70.2, as the magnitude of that term is dependent on the soil solidscomposition

70.3.4 COMMENTS

Measuring Water Content Profiles in the Field

For profiles near to (approximately 1.5 m depth) and extending to the surface, three generalapproaches have been used (Ferre´ and Topp 2002) Each offers certain advantages alongwith some limitations

With a series of differing length probes, vertically installed from the surface, it is possible tosegregate the water into layers in the profile Water in each layer is assumed evenlydistributed over the appropriate length interval Spatial variability laterally contributes tothe uncertainty or error associated with this type of profile determination, which can be aslarge as 0:03 m3 m3 (Topp 1987) The longest probes are useful for water balancecalculation, where a single measurement gives the total water quantity over the depthspanned and is not dependent on the depth distribution of the water Vertically installedrods tend to generate cracks in the soil between them at the surface and=or gaps around theindividual rods These soil openings each affect the infiltration of rainfall or irrigation andalso affect the TDR reading Vertical rods will tend to be moved vertically out of the soilduring winter by the processes of frost-heave

A second method involves installation of a number of horizontally oriented probes, one ateach measured depth These provide a more precise profile of water content that is notinfluenced strongly by lateral spatial variations, but these cannot compensate for majordiscontinuities in the vertical water content distribution The total profile storage for waterbalance estimates involves sums of values measured at each depth, being less precise thanfrom a single vertical probe Horizontally installed probes generally require opening a pit orhole into the soil, creating the possibility of disturbance to the region to be measured.Additionally the cable and probe connection must be hermetically sealed

An optimized profiling option uses parallel rods installed from the soil surface but 45off thevertical These can be placed so that the resulting water content profile is a single verticalprofile, and affected less by lateral variability and each depth increment provides for equalmagnitude lateral and vertical integration Schwartz and Evett (2003) give an evaluation of

30 installations for wetting front evaluation in a soil column The two disadvantages ofangled installations are the greater difficulty of making installations at an angle with therequired precision to know the actual depth at the end of the installed rods The increasedprobe length to achieve an angled installation decreases the total vertical depth that can bemeasured in clayey soils

Hook et al (1992) describe the use of diode shorting to segment probes, which areconstructed as profiling probes for use with model MP917 TDR instrument from Environ-mental Sensors, Inc These probes allow for determination of water content profiles havingthe same accuracy for each segment of0:01 m3 m3 The EM field for TDR measurementpropagates both in the resin comprising the probe, and in the surrounding soil, with more ofthe field in soil in wet than in dry soil, meaning that the sampling volume changes with watercontent Installation of this type of probe is of critical importance as any disturbance of thesoil adjacent to the probe is in the most sensitive of the measured region These edge effectfactors have not been adequately evaluated to allow specific quantification.

Recently, a number of attempts have been made to determine the water content profile usingwaveform analysis (Todoroff and Luk 2001; Heimovaara et al 2004) With additionalresearch and development, this approach may become the method of choice to overcomelimitations cited above

Coated-Rod Probes

Applying electrically resistive coatings to the rods can minimize the signal attenuation andloss of signal in conductive soils Ferre´ et al (1996) extended the work of Annan (1977) toshow that coated rods do not measure the arithmetic average of the dielectric permittivities ofthe coatings and the surrounding medium Because common coating materials have lowdielectric permittivities, coated rods are more sensitive to lower water contents than to higherwater contents One result of this variable sensitivity is that, unlike uncoated rods, coated rodprobes do not measure the correct length-weighted average water content along their length ifthe water content varies along their length Therefore, probes that measure the water contentthrough coatings should be installed in a manner that minimizes water content differencesthroughout their sample volume In addition, the reduced sensitivity of coated rods to conductivelosses reduces the usefulness of these probes for electrical conductivity measurement

70.4 GROUND PENETRATING RADARGPR has been widely applied in the geosciences (Neal 2004) and methods have recentlybeen developed for measuring soil volumetric water content (Davis and Annan 2002; Huis-man et al 2003) GPR methods offer the advantage of providing data from larger spatialregions than for TDR Another significant advantage of surface and airborne GPR methodsover TDR is that both methods are nonintrusive The borehole GPR method is intrusiverequiring installation of GPR transmitter and receiver in horizontal or vertical boreholes(Parkin et al 2000; Rucker and Ferre´ 2003) Surface and airborne GPR methods are mostappropriate for root zone investigations, whereas borehole methods are more appropriate fordeeper vadose zone applications This discussion is limited to above ground methods, whichoffer greater spatial coverage than downhole methods However, many of the conceptspresented are equally applicable to borehole GPR

The physics of the GPR method is identical to TDR (Weiler et al 1998) Both methods rely

on measuring the travel time or amplitude of EM wave fields Energy emitted from the GPRtransmitter travels through air and soil to the receiver Depending on the method used, the

travel time or amplitud e of ener gy from reflec ted or direct pathwa ys is measured andconverte d to the soil relat ive permi ttivity Equat ion 70.2 (or a soil-specifi c cal ibration) isthen used to convert the mea sured relative permi ttivity to volumetri c water content Moredetails on the principle s of usin g GPR to measur e soil water cont ent are found in Davis andAnnan (2002) and Huisman et al (2003)

Figure 70.3 shows two of the GPR ante nna configurat ions that have been used to measur esoil water content of which the methods are descr ibed herein The air -launched surfacereflectivi ty method has the advantage over the surface method in that the antenna can besuspended above the land surf ace, with the ener gy dir ected down wards The chosen antennafrequenc y and heig ht above ground will depend on the desired siz e of the energy footprint(area sam pled) on the ground For exampl e, Huism an et al (2003) show that when usin g

1 GHz and 2 25 MHz antenna system s elevated 1 m above the ground , approximat e footprintareas are 0.79 m by 0.79 m and 1.76 m by 1.76 m, resp ectively The followi ng equatio n(Davis and Annan 2002; Redma n et al 2002) calcul ates the soil relative permi ttivity fromthe amplitud e of the reflecte d energy, Ar , relative to a maximu m amplitud e, Am , from aperfect reflector such as a metal sheet placed on the ground, which has a reflectioncoefficient of1:

FIGURE 70.3 A schematic diagram showing two different GPR antennae configurations for

measuring soil water content (a) Surface GPR, direct ground wave, (b) launched surface reflectivity Tx and Rx are GPR transmitter and receiver anten-nae, respectively

Air-Active microwave remote sensing (synthetic aperture radar or SAR) operates on the sameprinciple as air-launched surface reflectivity GPR As SAR operates at a higher frequency(above 1 GHz) there are additional complications caused by soil surface roughness andgrowing plants One of the primary motivations behind SAR satellites is to be able toestimate surface soil conditions, including water content Mapping soil water content withSAR has been extensively researched and several projects have demonstrated the feasibility

of deriving water content using SAR As the methods continue to be under developmentmicrowave remote sensing is not given detailed coverage in this manual McNairn et al.(2002) prepared a state of the art summary of the SAR approach as a method for themeasurement of soil water content

The surface GPR method differs from the air-launched in that the antennae are placed indirect contact with the ground surface and the relative permittivity is determined bymeasuring the GPR wave velocity The velocity of direct or reflected waves between theGPR transmitter and receiver is converted to relative permittivity and water content as shownabove for TDR For either method, fixed or multiple antennae separation distances (offsets)can be used to collect velocity data Multiple offset methods include wide-angle reflec-tion and refraction (WARR) and common midpoint (CMP) A WARR survey involveskeeping the receiver antenna at a fixed location and moving the transmitter antenna awayfrom the receiver a set increment, measuring the ground wave velocity at each offset A CMPbegins with the transmitter and receiver placed very close together, and then incrementallymoving them in opposite directions, again measuring the ground wave velocity at eachoffset The ground wave velocity is more straightforward to measure when using the twomultiple offset methods as opposed to the fixed offset (FO) method The FO method requirespicking the arrival time of the direct ground wave, whereas the WARR and CMP methodscan use ground wave peak amplitude arrival times at different offsets to measure groundwave velocity

70.4.1 MATERIAL AND INSTRUMENTS

There are no commercially available GPR instruments designed specifically for measuringsoil water content Davis and Annan (2002) give GPR manufacturers whose equipment can

be adapted for measuring soil water content The procedures described forthwith assume thatusers are familiar with the basic operational methods, licensing requirements, and potentialhealth and safety issues of their GPR equipment and will therefore not be repeated

70.4.2 PROCEDURE

Air-Launched Surface Reflectivity

1 Position the GPR transmitter and receiver about 1 m above the ground surfaceusing a cart or vehicle (Davis and Annan 2002; Huisman et al 2003) Use at least

a 10 MHz system so that the electrical conductivity of the ground does notsubstantially influence the electrical current flow (Davis and Annan 2002)

2 Place a metal sheet, larger than the energy footprint on the ground, under the GPRantennae Measure Am, the maximum amplitude of the energy reflected from themetal sheet

3 Measur e A r over the soil of interest and then use Equation 70.3 to calcula te «r .Final ly, Eq uation 70.2 or a so il-specific calibrati on can be used to co nvert «r tosoil volum etric wat er content

Surface GPR (Dire ct Ground Wave, Fixed Offse t)

The followi ng proce dure is adapted after the stud ies by Grote et al (2003) , and Galagedar a

et al (2003, 200 5), and the review by Huisman et al (2003) :

1 Pe rform a W ARR survey to de termine the GPR system airwave velocity calibration(time zero, t0 ), to identify clearly the ground wave on the GPR output of energyversus time, and to select the best antennae offset distance for the FO survey Timezero is defined as the start of the transmitter pulse, which may vary due to thermaldrift and flexing of the fiber optic cables of the GPR system It is critical todetermine, using Equation (70.4), an accurate t0 c a l i b r at io n u si ng t h e a ir wa vevelocity measurement from the WAR R survey as all direct ground wave arrivaltimes are meas ured relative to it:

wher e t ab is the absolute ground wave arrival time, and t gwis the meas ured groun dwave arrival time (lead ing ed ge of the groun d wave) Fo r more inform ation ontime z ero, an d other issu es related to this method see Galagedar a et al (2003) As

a general recom mendat ion, Gal agedara et al (2005) suggest using seven offset s(from 0.5 to 2.0 m) for the WARR survey

2 Afte r sele cting the best offset (one for which the groun d wave is clearly separa tedfrom reflected wave s), perfo rm the FO survey by keeping the GPR antennae at thesele cted offset (Galage dara et al 2005 recom mend 1.5–2.0 m) and movi ng alongthe survey line (Huisman et al 2003) Measur ement s can be taken at a very smalltime incr ement (few seco nds), dep ending on the speed at which the antennae aremoving, a nd the de sired measu rement resolution Synch ronizing the GPR mea-surem ents wi th a GPS system facilit ates analys is of spatial variabili ty of soil water

de termine the soil water content

Surface GPR (Direct Ground Wave, Multiple Offsets)

Either the WARR or CMP methods can be used to gather data on the velocity of the directground wave Huisman et al (2001) found that soil water contents measured with the WARRmethod were more accurate than those measured with the FO method The WARR and CMPmethods do not rely on an accurate measurement oft0or picking of the leading edge of theground wave; they only depend on the slope of the peak arrival time versus antenna offsetrelationship: