B O C Z N IK I G L E B O Z N A W C Z E T . X X X V I N R 4, S. 5—15, W A R SZ A W A 1985
JERZY U P IE C
COMPARISON OF SOME METHODS FOR DETERMINING OF THE UNSATURATED HYDRAULIC CONDUCTIVITY OF SOILS
Institute o f Agrophysics, Polish A cadem y of Sciences, Lublin
W ater m ovem ent in soil profile tak es place m ainly in th e u nsatu ra te d zone. K nowledge of diffusivity and hydraulic conductivity as a function of soil m oisture content is required for the q u an titativ e evaluation of flow processes such an infiltration, évapotranspiration and w ater uptake by plants [6, 8]. Changes of hydraulic conductivity a t various m oisture contents can be also an index of soil pore-size d istri bution [2].
Methods for determ ination of the diffusivity and un saturated con ductivity, p articu larly in field conditions, are laborious, v ery tim e con sum ing and dem and complicated instrum entation.
A r y a et al. [1] developed and tested a new laboratory procedure called th e hot-air or evaporation method, for determ ination of diffusi vity and un satu rated hydraulic conductivity in undisturbed soil cores. The m ethod seems to be attractive, as i t is quick, cheap andl simple. E h l e r s [7] found th e m ethod was effective in dem onstrating th e hyd raulic differences bew een tilled and untilled loess soil. В o u m a [3] also reported th a t th e first result obtained w ith the m ethod w ere en couraging, b u t m ore testing was necessary before a final judgem ent could be made.
The objective of the work as reported h ere is to compare the unsa- tu rate d hydraulic conductivity m easurem ents by th e hot-air m ethod w ith those from th e in situ cru st-te st procedure a t the low* soil m oisture tensions and w ith the calculation m ethod in a higher range of soil m oisture tension.
MATERIAL AND METHODS
Experim ental plots w ere located on the chernozem, cambisol, ren- dzina and luvisol w ith te x tu re of silt loam, m edium -heavy silty loam,, clay loam and coarse sand, respectively (Table 1).
The diffusivity and u nsaturated conductivity was determ ined by th e laboratory m ethod proposed by A r y a et al. [1]. Soil cores w ith undisturbed stru ctu re of 200 m l and 5.1 cm in diam eter were used. The soil cores were taken in 6 replications from the de'pth of 5—15 and 25—35 cm from each soil. The soil samples w ere saturated by capilary rise and equilibrated in a horizontal position for some days. Therefore w arm air was blown over the exposed of the core to m easure evaporation rates. The w arm air was blown from an electric drier, equipped w ith an additional heating coil and autotransform er allowing to m aintain a constant tem p erature of th e blowing air in th e range of 80-200°C. The soil cores w ere w eighted every m inute. A fter drying the soil was divided into 12—14 segm ents to determ ine soil m oisture distribution. The soil was pushed out from the cylinder w ith a piston of the same diam eter.
P relim in ary tests w ere m ade to find the proper tem p eratu re and tim e of drying to satisfy the following presupposed conditions: cum u lative evaporation m ust be proportional to th e square root of tim e and the original w ater content of the soil at the closed end unchanged during the evaporation procedure.
In the chernozem, cambisol and rendzina these conditions w ere m et by drying tim es of 15, 12 and 10 min., respectively, a t the tem perature of 120— 130°C (Fig. 1). Only in the luvisql w ith th e coarsest tex tu re th e original w ater content a t the closed end was changed.
The diffusivity was calculated from the m oisture distribution curves (Fig 1) obtained after evaporation as a function of volum etric w ater content 0 [cmVcm3] a t 0.02 0 in terv als according to th e equation of B r u c e and К 1 u t e [5] :
ö ftere D (0 X) is th e diffusivity [cm2/min] as a function of w ater content 0 , t is the tim e [min], x is the distance from the evaporation surface [cm] and 0 / is the initial w ater content.
The boundary conditions are:
w here 0 O is the air-d ry soil m oisture content.
The geom etric m ean of replicate D values and 95Vo confidence in terval of the geom etric m ean were calculated at the U niversity of Geettingen. Institute of Agronomy and P lant Breeding (FRG) w ith the m ethod used by E h l e r s [7]. The all diffusivities determ ined w ere
Comparison of som e methods.., 7
Fig. 1. Exam ples of m oisture distribution in cores o f the soils in vestigated in the 5—15 cm layer after evaporation. The inserts show the relationship betw een cum u
lative evaporation and the square root of tim e
used for com putation of th e best fitting D (0] curves. These curves w ere used to calculate the unsaturated hydraulic conductivity к as a function of w ater content (©) according to the equation:
dd
d e
w here is the slope of the m oisture reten tio n curve. F or comparison w ith the data obtained in the cru st-test procedure and calculation m e thod к cm /day w ill be p reten ted as a function of W.
The second m ethod of determ ining к is th e in situ cru st-te st p ro cedure described and tested by В o u m a and D e n n i n g [4]. The me
thod consists af a series of infiltration ru n s through crusts of decreasing hydraulic resistance, each of which gives one point on the hydraulic conductivity curve к (Ï7). The crusts were formed from a pre-w etted m ixture composed of gypsum and sand. Each ru n was made in a soil p it using an excavated cylindrical column of soil w ith th e height and diam eter of 28 cm. A ring steel infiltrom eter of the same diam eter and 10 cm high was fitted onto the column. The w ater w as fed from a b u re tte w ith a M ariotte device m aintaining a constant pressure of about 3 mm w ater over the crust. Soil m oisture tensions w ere m easured by tensiom eters. The infiltration rate constant for a period of at least one hour, equals the hydraulic conductivity w hen the m oisture tension
gradient is zero. According to D arcy’s low:
V
J t* - ri
w here v is the infiltration rate and i is the hydraulic gradient, which is ussually close to unity, as the flow occurs only by gravity [3]. W ith this m ethod the к values w ere m easured to the soil m oisture tension of 60 hPa.
The hydraulic conductivity in a higher range of the tension was calculated according to G reen and Corey’s m ethod as modified by L u x m o o r e [9]. In the m ethod the calculated hydraulic conductivity is obtained from soil pore-size frequency distribution and the statistical probability of continuity of pores across any plane in the soil. Two experim ental values (e.g. at satu ration and a t some unsatu rated w ater content) obtained w ith the crust-test procedure were included in com puting for m atching the calculated conductivity w ith the use of series of weighted m atching factors.
RESULTS AND DISCUSSION
Figure 1 shows th at w ith the hot-air m ethod the required linearity between cum ulative evaporation and the square root of tim e was reached in chernozem and cambisol after 2—4 min, in rendzina after 1—2 min, and in the luvisol after 4—6 min. A r y a et al. [1], E h l e r s [7] and O p a r a - N a d i [10] achieved linearity after 2 min, 2,5 to 4 m in and 3 do 6 min, respectively. These authors used 80 m l H apludoll loam soil cores w ith the diam eter of 3-65 cm and the drying tem perature of about 90—100°C, 200 m l loess soil cores w ith the diam eter of 5,4 cm and the tem perature of 130°C and 500 m l soil loess cores w ith the dia m eter of 8 cm and th e tem perature betw een 185—230°C. The different external conditions could by responsible, at least partly, for the various
T a b 1 • 1 oom* c h a r r . c t o r i r t i c a o f th e o o i l o ir . v o s t i c a t c d S o l l e D epth СП F a r t i o l e o iz e / i n гш / Роге л:.т.е /1 n j u n / К aa t . c a /d a y D e n s ity k g /n 3 ю - з O rg an io m a tte r • o n t ë n t % S p e c i f i c s u r f a c e n 2/ g CaC03 % 1 -0 ,1 0 ,1 - 0 ,0 2 0 ,0 2 < 0 ,0 0 2 > 30 3 0 - з < 3 - - - * % C hernocon 5 -1 5 6 50 44 14 1 0 ,0 2 1 ,1 2 1 ,0 12?,1 1,30 3,06 8 7 ,7 2 ,1 5 2 5-35 6 49 45 13 9 ,6 20,-i 1 9 ,2 8 5 ,2 1, 33 2 ,7 6 9 6 ,3 0 ,6 0 C am biool 5 -1 5 21 31 4ö Q 9 ,3 2 0 ,0 1 9 ,3 4 1 ,0 1 ,40 1,M3 5 3 ,7 0 ,1 5 25-35 18 35 47 10 7 ,1 1 9 ,0 1 5 ,4 GO, 1 1 ,5 2 и , 42 4 6 ,8 0 ,1 3 F c n d z in a 5 -15 10 19 63 27 9 ,7 1 ^ ,3 2 7 ,5 З - .З 1 ,1 3 3 ,95 2 1 9 ,7 2 4 ,0 25-35 17 21 62 29 11,1 i 5 , o 2 6 ,4 '< 2, 3 1,13 2 ,8 7 2 1 5 ,0 3 3 ,7 Lu tIboI 5-15 76 1б 0 2 1 7 ,6 1 3 ,2 10 ,5 2 0 1 ,2 1.51 1 ,4 2 2 9 ,2 0 ,0 8 25-35 73 18 9 3 2 0 ,1 1 3 ,3 7 ,2 135,1 1,56 0 ,2 1 1 6 ,5 0 ,0 6
tim es needed for achiem ent linearity in the investigations mentioned. In this w ork 200 ml cores w ith the diam eter of 5,1 cm and the tem pe ratu re betw een 120— 130°C were used for all soils. Therefore, the various time periods for reaching linearity in p articular soils m ight be explained by th eir indyvidual hydraulic properties and initial w etness ra th e r than by external conditions.
The various hydraulic properties of the soils are shown in Figs. 2 and 3 and typical w ater distribution curves are shown in Fig- 1. The w ater content in the sealed ends of chernozem, cambisol and rendzina did not deviate significantly. However, in the luvisol w ith the coarsest te x tu re and largest contribution of pores > in equivalent dia m eter (Table 1) the w ater content was much lower. Consequently, one
of the theoretical conditions of this m ethod was not satisfied. This failure w ith the luvisol m ay by explained by the relatively high hydraulic conductivity of the sandy soil a t a m oisture tension near saturation and by the rapid upw ard m ovem ent of the w ater. In the luvisol the average value of к at a m oisture tension of 20 hPa was 20.1 cm /day and in the rem aining soils — below 4.9 cm /day. E h l e r s [7] reported earlier th at high hydraulic conductivity m ight be responsible for this effect in the 30—40 cm layer of untilled loess soil. Futherm ore, after evaporation the w ater content range in the luvisol was much sm aller than in the rem aining soils and the w ater contents versus distance did not give a smooth curve (Fig. 1). P erhaps oth er sizes of core and e x te r nal conditions could be sufficient to overcome these difficulties [7].
It is w orth noticing th a t the w ater losses during evaporation from all soils are lim ited to the first two centim eters at the open end (Fig. 1). The high surface drying of the soils resulted in very steep gradients of the w ater content w ith distance near evaporating surface.
Figure 2 shows the calculated diffusivities as a function of soil moi sture content. Mean values and the confidence in terv al w ere calculated for d iffrent w ater contents as observed in the soil columns after drying. The lowest diffusivities in the whole range of corresponding w ater contents w ere in rendzina and the highest diffusivities appeared in the w et range of cambisol.
H ydraulic conductivity as a function of m oisture tension is presented in Fig. 3. In both layers of all soils investigated the hydraulic condu ctivities m easured by the hot-air m ethod w ere lower th an those deter mined by the cru st-test procedure to the soil m oisture tension of 60 hPa. The degree of agreem ent was different in p articular soils. In th e layer of 5—15 cm the lowest differences were in chernozem, somewhat larger in cambisol and much higher in rendzina. At the depth of
Fig. 3. Hydraulic conductivity as a function of m oisture tension, in tw o layers of the soils investigated
25—35 cm the differences in к results, as in the upper layer, w ere lowest in the chernozem and largest in rendzina. However, in cam bisol th e results obtained by th e hot-air m ethod were barely from soil m oisture tension of 70 hPa, thus beyond the tension range, for w hich the к values w ere determ ined by the cru st-test procedure.
Comparison of some methods.., 13 A t the tension range > 60 h P a agreem ent betw een к values by the hot-air m ethod and calculated values was also different in different soils (Fig. 3). The best agreem ent in the m oisture tension range corres- pondig to useful effective retention was found for the chernozem: In the cambisol at the same tension range th e results obtained by the h o t-air m ethod w ere slightly lower compared to th e calculated values. However, in th e rendzina the к values from the hot-air m ethod w ere significantly low er compared to those from the calculated procedure uip to tension of about 300 hPa. However, in the higher tension range the k( T) curves rapidly close together.
It may be concluded th a t degree of agreem ent of k ( ï/ ) obtained w ith the m ethods tested depends on the kind of soil and its properties. In chernozem and cambisol soils showing ra th e r uniform pore-size dis tribution, the results obtained by the laboratory hot-air m ethod com p are w ell w ith those from th e cru st-test procedure and the calculation m ethod. However, in th e rendzina, showing the lowest diffusivity and th e largest contribution of pores < 3 ^ m in equivalent diameter., the
к results from the hot-air m ethod w ere m arkedly low er th an those obtained by the cru st-test and calculation m ethods (to the tension of 300 hPa). In the luvisol, in which th e contribution of coarse pores ■< 30 /л п was large the initial w ater content at the closed end of soil column was changed w hen evaporation started.
Consequently, it precluded application of the h o t-air m ethod to th a t soil. The application of this m ethod w as also not possible on quartz pow der w ith non-uniform pore-size distribution [10].
In judging th e usefulness of the m ethods in term s of labour con sum ption and the apparatus required, it m ight be stated th a t for the d e term ination of к values in a wide tension range the h ot-air m ethod is most suitable, as it is rapid, cheap and dem ands no special in stru m en ta tion.
*
The author is indebted to Ir. P. van !der Sluijs from ;£he Netherlands
Survey Institute in Wageningen and to Dr. W. Ehlers from the Uni versity of Goettingen, Institute of Agronomy and Plant Breeding (RFG) for their helpful suggestions
.
REFERENCES
£1} A r y a L.M., F a r r e l l D A ., B l a k e GJt.: A field study o f soil w ater depletion patterns in presence o f growing soybean roots, I. D eterm ination of hydraulic properties of the soil. S oil Sei. Soc. Amer. Proc. 39, 1973, 424—430.
12] В o u m a J,: Use of physical m ethode to expand soil survey interpretation of .soil drainage conditions. Soil Sei. Soc. Amer. Proc. 37, 1973, 413—420. [3] В o u m a J.: Soil survey and the study of w ater in unsaturated soil. Soil
Survey Paper No 13, 1977, 56 pp, W ageningen.
[4] B o u m a J., D e n n i n g J.L.: Field m easurem ents of unsaturated hydraulic conductivity by infiltration through gypsum crusts. Soil Sei. Soc. Amer. Proc. 36, 1972, 846—847.
[5] B r u c e R.R., K ł u t e A.: The m easurem ent of soil m oisture d iffu sivity. Soil Sei. Soc. Amer. Proc. 20, 1956, 458—462.
[6] D e c h n i k I., L i p i e c J.: Przew odnictw o wodne gleby w strefie n ien asy conej i metody jej pomiaru. Probl. Agrofiz. 32, 1980, 4—34.
[7] E h l e r s W.: Rapid determ ination of unsaturated hydraulic conductivity in tilled and untilled loess soil. Soil Sei. Soc. Amer. Proc. 40, 1976, 837—840. [8] F e d d e s R.A., K o w a l i k P., M a l i n k a K.K., Z a r a d n y H.: Sim ulation
of field w ater uptake using a soil w ater dependent root extraction function. J. Hydrol. 31, 1976, 13—26.
[9] L u x m o o r e R. J.: Application of the Greeen and Corey method for com puting hydraulic conductivity in hydrological m odelling. Univ. of W isconsin, Madison, 1973, No 53706, 1—22.
[10] O p a r a - N a d i O.A.: A com parison of some m ethods for determ ining the hydraulic conductivity of unsaturated soils in the low suction range. G öttin ger Bodenkundlische Barichte 57, ,1979, 1—104.
E. ЛИПЕЦ СРАВНЕНИЕ НЕКОТОРЫХ МЕТОДОВ ОПРЕДЕЛЕНИЯ ВОДОПРОВОДИМОСТЙ ПОЧВ ПРИ НЕПОЛНОМ НАСЫЩЕНИИ Институт агрофизики ПАН в Люблине Р езю м е Сравнивали величины коэффициента водопроводимости к при неполном насыщении почв полученные при использовании нового и быстрого лабораторного метода эвапораций с результатами этой водопроводимости полученными по полевому методу при применении гипсово-песчаных корок с сосущим давлением до 60 гПа и по расчетному методу с высшими значениями сосущего давления. Исследования проводились на черноземе, бурой почве, рендзине и палевой почве. Наиболее сходные значения к полученные с помощью сравни ваемых методов были установлены в черноземе и бурой почве характеризующихся срав нительно однородным распределением пор в соответствии с их размерами. В рендзине ха рактеризующейся самой малой зернистостью и самым высоким учатием малых пор с экви валентным диаметром < 3 /гм, значения полученные по методу эвапорации были заметно меньше значений полученных с помощью остальных методов. В пакевой почве характери зующейся крупнозернистой гранулометрией и самым высоким участием крупных пор >30 дм, не было соблюдено предварительное теоретическое условие эвапотранспирационного ме тода, что сделало невозможным ее применение к этой почве.
Comparison of some méthode.., 15 J. LIPIEC
PORÓWNANIE NIEKTÓRYCH METOD OZNACZANIA PRZEWODNICTWA WODNEGO GLEB PRZY NIEPEŁNYM NASYCENIU
Zakład Agrofizyki PAN w Lublinie
S t r e s z c z e n i e
Porównano w artości w spółczynnika przew odnictw a w odnego к przy niepełnym nasyceniu gleb, uzyskane nową i szybką metodą laboratoryjną ew aporacji z w y nikam i tego przewodnictwa otrzym anym i metodami: połową z zastosowaniem skorup gipsow o-piaskow ych przy ciśnieniu ssącym do 60 hPa oraz obliczeniową przy w yższych w artościach ciśnienia ssącego. Badania przeprowadzono na na stępujących glebach: czarnoziemie, glebie brunatnej, rędzinie i glebie płowej. Najlepszą zgodność w artości к uzyskanych porów nyw anym i metodami otrzymano w czarnoziem ie i glebie brunatnej, charakteryzujących się dość jednorodnym roz kładem porów w edług ich w ym iarów . W rędzinie, w ykazującej najdrobniejsze uziarnienie i najw iększy udział porów m ałych o średnicy równoważnej < 3 {im, wartości к uzyskane m etodą ewaporacji były w yraźnie m niejsze niż w artości uzy skane pozostałym i m etodam i. W glebie płowej, odznaczającej się gruboziarnistym rozkł.adem granulome;trycznym i najwięksizym udziałem porów dużych > 30 wm, nie został spełniony w stęp ny warunek teoretyczny m etody ew aporacyjnej, co u nie m ożliwiło jej zastosow anie do tej gleby.
Dr J erzy Lipiec Z akład A g ro fizyk i PA N
Lublin, ul. K ra k o w sk ie P rzedm ieście 39
