Soil hydraulic parameters and surface soil moisture of a tilled bare soil plot inversely derived from l-band brightness temperatures

Authors: DIMITROV, M - Julich Research Center; VANDERBORGHT, J - Julich Research Center; KOSTOV, K - Bulgarian Academy Of Sciences; JADOON, K - King Abdullah University Of Science And Technology; WEIHERMUELLER, LUIS - Julich Research Center; BINDLISH, RAJAT - Science Systems, Inc; Pachepsky, Yakov; SCHWANK, MAX - German Research Center For Environmental Health; VEREECKEN, HARRY - Julich Research Center

Submitted to: Vadose Zone Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/15/2013
Publication Date: 1/13/2014
Publication URL: http://handle.nal.usda.gov/10113/61455
Citation: Dimitrov, M., Vanderborght, J., Kostov, K.G., Jadoon, K.Z., Weihermüller, L., Jackson, T.J., Bindlish, R., Pachepsky, Y., Schwank, M., Vereecken, H. 2014. Soil hydraulic parameters and surface soil moisture of a tilled bare soil plot inversely derived from l¬band brightness temperatures. Vadose Zone Journal, 13 (1), DOI: 10.2136/vzj2013.04.0075.

Interpretive Summary: Soil brightness temperature (the surface temperature measured from a satellite or aircraft) is affected by soil water content and evaporation rate. Evaporation rates are affected by soil water content and soil hydraulic parameters, which control the ability of soil to conduct water upward and to store water near the soil surface. Finally, evaporation and brightness temperature are related via the radiative transfer model. These three relations create an opportunity to back-calculate soil hydraulic parameters from soil brightness temperature dynamics. We developed a method to make this calculation and then evaluated it with two models of soil hydraulic properties and radiative transfer, using data from plowed plots of loess soil. We found that hydraulic properties of the plowed soil should be defined separately for soil aggregates and for interaggregate space. Based on this assumption, soil hydraulic parameters retrieved from the data on brightness temperature were found to be close to the parameters measured in the laboratory. The successful retrieval of hydraulic properties was verified also by comparison between measured and simulated water contents at 2 cm and 5 cm depths. Results of this work will be useful in environmental projects dealing with soil hydraulic properties in that these properties can be estimated for large areas where brightness temperature can be measured with remote sensing technologies.

Technical Abstract: We coupled a radiative transfer approach with a soil hydrological model (HYDRUS 1D) and a global optimization routine SCE-UA to derive soil hydraulic parameters and soil surface roughness from measured brightness temperatures at 1.4 GHz (L-band) and measured rainfall and calculated potential soil evaporation. Long term data sets from a bare soil plot prepared after ploughing for five meteorologically different measurement periods were used in this coupled model inversion routine. We considered two different models for the hydraulic soil properties: the uni-modal model of Mualem van Genuchten and the bi-modal model of Durner, and two reflectivity models: the Fresnel equation based on the depth averaged dielectric permittivity in the top 2 cm thick soil layer and the coherent radiative transfer model (CRTM). The modeled brightness temperatures were fitted well to the measured brightness temperatures using both radiative transfer models. The fitted surface roughness parameter and its direct measurement using a laser profiler were similar. The lab derived water retention curve of the ploughed soil showed a bi-modal behavior which could be retrieved consistently from brightness temperatures for the different periods. Using a uni-modal and bi-modal model resulted in an underestimation of the hydraulic conductivity near saturation. Simulated surface soil moisture using retrieved soil hydraulic parameters was compared with soil moisture, calculated from measured dielectric permittivity at 2 cm and 5 cm using 5TE sensors. The effect of depth averaging by the in-situ moisture sensors and the use of soil and depth-specific calibration relations were evaluated and found to be of importance for the comparison with L-band retrieved soil moisture.