Journal cover Journal topic
Hydrology and Earth System Sciences An interactive open-access journal of the European Geosciences Union
doi:10.5194/hess-2017-109
© Author(s) 2017. This work is distributed
under the Creative Commons Attribution 3.0 License.
Research article
02 Mar 2017
Review status
This discussion paper is under review for the journal Hydrology and Earth System Sciences (HESS).
Unrepresented model errors – effect on estimated soil hydraulic material properties
Stefan Jaumann1,2 and Kurt Roth1,3 1Institute of Environmental Physics, Heidelberg University, Im Neuenheimer Feld 229, 69120 Heidelberg, Germany
2HGSMathComp, Heidelberg University, Im Neuenheimer Feld 205, 69120 Heidelberg, Germany
3Interdisciplinary Center for Scientific Computing, Heidelberg University, Im Neuenheimer Feld 205, 69120 Heidelberg, Germany
Abstract. We investigate the quantitative effect of unrepresented (i) sensor position uncertainty, (ii) small scale-heterogeneity, and (iii) 2D flow phenomena on estimated effective soil hydraulic material properties.

Therefore, a complicated 2D subsurface architecture (ASSESS) was forced with a fluctuating groundwater table. Time Domain Reflectometry (TDR), Ground Penetrating Radar (GPR), and hydraulic potential measurement devices monitored the hydraulic state during the experiment. Since the measurement data are analyzed with an inversion method, starting close to the measurement data is key. Therefore, we developed a method which estimates the initial water content distribution by approximating the soil water characteristic on the basis of TDR measurement data and the position of the groundwater table. In order to reduce parameter bias due to unrepresented model errors, we implemented a structural error analysis to identify uncertain model components which have to be included in the parameter estimation. Hence, focussing on TDR and hydraulic potential data, we realized a 1D and a 2D study with increasingly complex setups: Starting with estimating effective hydraulic material properties, we added the estimation of sensor positions, the estimation of small-scale heterogeneity, or both.

The analysis of these studies with a modified Levenberg-Marquardt algorithm demonstrates three main points: (i) The approximated soil water characteristic for the initial water content distribution is reasonably close to inversion results. (ii) Although the material properties resulting from 1D and 2D studies are similar, the 1D studies are likely to yield biased parameters due to unrepresented lateral flow. (iii) Representing and estimating sensor positions as well as small-scale heterogeneity improves the mean absolute error by more than a factor of 2.


Citation: Jaumann, S. and Roth, K.: Unrepresented model errors – effect on estimated soil hydraulic material properties, Hydrol. Earth Syst. Sci. Discuss., doi:10.5194/hess-2017-109, in review, 2017.
Stefan Jaumann and Kurt Roth
Stefan Jaumann and Kurt Roth
Stefan Jaumann and Kurt Roth

Viewed

Total article views: 271 (including HTML, PDF, and XML)

HTML PDF XML Total BibTeX EndNote
247 22 2 271 3 4

Views and downloads (calculated since 02 Mar 2017)

Cumulative views and downloads (calculated since 02 Mar 2017)

Viewed (geographical distribution)

Total article views: 271 (including HTML, PDF, and XML)

Thereof 268 with geography defined and 3 with unknown origin.

Country # Views %
  • 1

Saved

Discussed

Latest update: 26 Mar 2017
Publications Copernicus
Download
Short summary
We investigate the quantitative effect of neglected sensor position, small-scale heterogeneity, and lateral flow on soil hydraulic material properties. Therefore, we analyze a fluctuating water table experiment in a 2D architecture (ASSESS) with increasingly complex studies based on Time Domain Reflectometry and hydraulic potential data. We found that 1D studies may yield biased parameters and that estimating sensor positions as well as small-scale heterogeneity improves the model significantly.
We investigate the quantitative effect of neglected sensor position, small-scale heterogeneity,...
Share