Estimating Soil Hydraulic Properties from Infrared Measurements of Soil Surface Temperatures and TDR Data

摘要

The spatiotemporal development of soil surface temperatures (SST) depends on water availability in the near-surface soil layer. Because the soil loses latent heat during evaporation and water available for evaporation depends on soil hydraulic properties (SHP), the temporal variability of SST should contain information about the near-surface SHP. The objective of this study was to investigate the uncertainties of SHP derived from SST. The HYDRUS-1D code coupled with a global optimizer (DREAM) was used to inversely estimate van Genuchten-Mualem parameters from infrared-measured SST and time domain reflectometry (TDR)-measured water contents. This approach was tested using synthetic and real data, collected during September 2008 from a harrowed silty loam field plot in Selhausen, Germany. The synthetic data illustrated that SHP can be derived from SST and that additional soil water content measurements reduce the uncertainty of the estimated SHP. Unlike for the synthetic experiment with a vertically homogeneous soil profile, a layered soil profile had to be assumed to derive SHP from the real data. Therefore, the uncertainty of SHP derived from real data was considerably larger. Water retention curves of undisturbed soil cores were similar to those estimated from SST and TDR data for the deeper undisturbed soil. The retention curves derived from SST and TDR data for the harrowed topsoil layer were typical for a coarse-textured soil and deviated considerably from the retention curves of soil cores, which were typical for a fine-textured soil and similar to those from the subsoil.