Assessing rooting depths of an austrian pine stand by inverse modeling soil water content maps

Abstract

Abstract. Rooting depths in a forested stand (0.11 ha) were estimated indirectly by inverse modeling maps of soil water contents from time domain reflectometry (TDR) measurements at 150 points. These maps were described with a calibrated onedimensional soil water flow model, with specific values for the rooting depth, van Genuchten's [1980] a parameter, and throughfall fraction at each point. At about one third of the 150 points, modeled rooting depths did not fall within the a priori likely range of rooting depths; these values were discarded when kriging the point estimates to a map of rooting depths. The resulting rooting depths range from 0.8 m to 3.5 m within the stand in good agreement with root observations (R 2 = 0.85). Several factors contributed positively to this good match, such as the fairly uniform soil hydraulic properties and the high sensitivity of soil water dynamics to root water uptake. Overall, the results demonstrate the suitability of soil water content maps based on TDR measurements to quantify spatial variability in soil water dynamics. IntroductionRoot water uptake in ecosystems is often studied with soilvegetation-atmosphere (SVAT) models that integrate the simulation of the different elements of the hydrological cycle. SVAT models offer a tool to estimate plant water uptake and its distribution over the soil as plant uptake fluxes cannot be measured directly. From a hydrological perspective, water fluxes to the atmosphere and groundwater are largely controlled by root systems of trees Schaap et al., 1995]. Rooting depths are therefore key parameters in SVAT models but are difficult and tedious to collect by classical methods such as root excavation and soil coring [Atkinson, 1991]. As a consequence, rooting depths are often only known by estimation; hence results of SVAT models are difficult to interpret.Potentially, root distribution parameters like the rooting depth can be estimated by inverse modeling the soil water dynamics. This indirect estimation is preferable to conventional root observation methods, mainly because root information is not directly related to root activity [Gardner, 1991]. Inverse modeling will succeed only if root water uptake significantly influences the soil water dynamics; its use is therefore restricted to shallow root systems (<4 m) and to dry periods when sensitivity of soil water dynamics to root water uptake is largest. Also, root parameters cannot be estimated indirectly without accurately describing the soil water fluxes. Thus inverse modeling requires a calibrated soil water flow model and consequently a monitoring program to provide the necessary data for calibrating a soil water flow model. In addition, roots never grow uniformly through a soil, and root parameters should therefore be studied on a relatively large scale (>100 The use of indirect methods to estimate root parameters has for long time been frustrated by the inability to measure The aim of this study was to estimate rooting depths in a forested pine stand by inverse modeling the soil ...

Estimating hydraulic conductivity of internal drainage for layered soils in situ

“…The gentle slope of the K functions from the Swartland was consistent with the low clay content (< 22 %) and the presence of saprolite rock in the C horizon. Similar observations of clay soils were made by Freankel (2008) and Nhlabatsi (2011). Given the poor drainage properties of Tukulu and Sepane, it was proposed that a zero drainage flux be assigned at the bottom of these soil profiles for soil water balance studies (Hensley et al, 2000).…”

“…The hydraulic parameters are usually based on the SWRC, because it is easily measured and can be estimated using a parameter optimisation technique (Kool et al, 1987;Hopmans and Simunek, 1999). Several studies applied inverse modelling and parameter estimation of hydraulic functions directly from in situ drainage transient experiments (Dane and Hruska, 1983;Romano, 1993;Zijilstra and Dane, 1996;Musters and Bouten, 1999;Dikinya, 2005). Agreement between in situ and predicted hydraulic functions was generally satisfactory, even though the K function was highly variable.…”

“…In variably structured soils, −10 kPa is often used as a hypothetical boundary for separating drained structural pores and water-filled micro-pores (Marshall, 1959;Kutilek, 2004). Various work from local and international drainage experiments recorded suctions around −10 kPa from variably structured soils (Hensley et al, 2000;Sonnleitner et al, 2003;Nhlabatsi, 2011;Adhanom et al, 2012). The use of undisturbed core samples three times larger than the area sensitive to tensiometer ceramic cup qualifies this procedure, even though estimates were made from parameters based on SWRC.…”

A Linking Test that establishes if groundwater recharge can be determined by optimising vegetation parameters against soil moisture

“…It can be concluded that no unique vegetation parameter can be obtained by matching observed and simulated θ and the vegetation parameters are falsely linked parameters and not sets of partially linked parameters since ΔE > 10% and ΔINT > 10%. These finding are similar to Hupet et al (2002;2003), , Musters and Bouten (1999;2000) that found that the root water uptake parameters are not sensitive enough to be calibrated against θ.…”

“…In the present case study we assume the hydraulic parameters are known and we want to know whether accurate groundwater recharge could be determined by optimising the vegetation parameters against θ. The root water uptake parameters are assumed to be known, since Hupet et al (2002;2003), , Musters and Bouten (1999;2000) showed that the root water uptake parameters are not sensitive enough to be optimised against θ. In this case study the vegetation parameters consists of 3 parameters: 2 interception parameters and 1 crop factor parameter.…”