Evaluating the temperature difference (dT) parameter in the SSEBop model with satellite observed land surface temperature data

Authors: Lei Ji*, ASRC InuTeq, contractor to USGS EROS Center, Gabriel B. Senay2, USGS EROS Center, Naga Manohar Velpuri, ASRC InuTeq, contractor to USGS EROS Center, Stefanie Kagone, SGT, contractor to USGS EROS Center
Topics: Remote Sensing, Water Resources and Hydrology, Climatology and Meteorology
Keywords: actual evapotranspiration, land surface temperature, MODIS, SSEBop model, temperature difference
Session Type: Paper
Day: 4/3/2019
Start / End Time: 4:30 PM / 6:10 PM
Room: Stones Throw 3 - Mica, Marriott, Lobby Level
Presentation File: No File Uploaded


The Operational Simplified Surface Energy Balance (SSEBop) model is an energy balance model for generating spatially explicit actual evapotranspiration (ETa) using remotely sensed thermal data and model-simulated weather data. The temperature difference (dT) in the SSEBop model is a predefined parameter quantifying the difference between surface temperature at bare soil and air temperature at canopy level. Because dT is derived from the average-sky net radiation based primarily on climate data, validation of the dT estimation is critical for assuring a high-quality ETa product. In this study, we used the Moderate Resolution Imaging Spectroradiometer (MODIS) land surface temperature (LST) data to evaluate the SSEBop dT estimation for the conterminous United States. MODIS data (2008-2017) were processed to compute the10-year average LST and normalized difference vegetation index (NDVI) at 1-km resolution and 8-day interval. The observed dT (dTo) was computed from the LST difference between hot (bare soil) pixels (NDVI < 0.25) and cold (canopy level) pixels (NDVI > 0.7). We created 2x2 degrees blocks as the sampling unit, and with each block calculated dTo from the hot and cold pixels. There were sufficient numbers of hot and cold pixels within sampling blocks for creating seasonal dTo timeseries in the Pacific-western and South-central regions. The comparison of MODIS-derived dTo and modeled dT (dTm) showed high agreement, with a bias of 1.5 K and a correlation coefficient of 0.93 on average. This study concludes that the dTm estimation from the SSEBop model is reliable, which further assures the accuracy of the ETa estimation.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login