319-2 Soil Hydraulic Properties Spatial Variability and Resolution Effects On Hydrologic Modeling.

See more from this Division: S05 Pedology
See more from this Session: Digital Soil Assessment for Ecosystem Modeling: I
Wednesday, November 3, 2010: 1:15 PM
Long Beach Convention Center, Room 103C, First Floor

Zamir Libohova1, Phillip Owens1, Laura Bowling2, Keith A. Cherkauer3, Bibi Naz4 and Valbona Hobdari5, (1)915 W. State St., Purdue University, West Lafayette, IN
(2)915 W State Street, Purdue University, West Lafayette, IN
(3)Purdue University, Agricultural and Biological Engineering, West Lafayette, IN
(4)Purdue University, West Lafayette, IN
(5)Agricultural University of Tirana, Tirana, Albania
The success of hydrologic modeling in predicting hydrologic processes depends on the accurate representation of major drivers such as weather, land use management, geomorphic surface, and soils, especially, their spatial variability. Recent developments in spatial tools offer an opportunity to represent the spatial variability of soils and their properties in a continuum raster format useful for modeling.  The overall objective of the research is to assess the impact of spatial resolution of raster based predictive soil property maps generated from terrain attribute soil mapping on simulated streamflow response using the Distributed Hydrology Soil Vegetation Model (DHSVM). A USGS gauged 56 km2 watershed in a loess veneered landscape in southern Indiana was used in this study. A continuum raster based soil depth to a restrictive layer map based on terrain attributes derived from 10, 30 and 90 m pixel resolutions was used as an input to the DHSVM. The soil depth maps generated at 90 and 30 m pixel resolution were disaggregated to 30 and 10 m pixel resolution. Measured soil hydraulic properties such as Ksat and available water holding capacity were also used as model inputs. The DHSVM simulated daily streamflow assessed at 30 m pixel resolution agreed well with the measured streamflow (Nash-Sutcliffe efficiency = 0.52) without model calibration for a 5 year simulation. The simulated mean annual streamflow was not sensitive to model resolution changes and soil map disaggregation level (p-value = 0.09). The simulated annual minimum decreased significantly with increasing model resolution while annual maximum increased (p value < 0.001). The simulated mean annual flow, annual minimum and maximum were not significantly different for native 10 m, and 90 m and 30 m soil depth maps disaggregated to 10 m. R-B Flashiness increased significantly (p < 0.001) with increasing model resolution but was not significantly different between the different 10 m disaggregated soil depth maps.