Heterogeneity of latent heat fluxes over the ARM SGP site from satellite-based and in situ measurements



Sullivan, Ryan - Argonne National Laboratory

Area of research:

Surface Properties

Journal Reference:

Sullivan R, D Cook, V Ghate, V Kotamarthi, and Y Feng. 2019. "Improved Spatiotemporal Representativeness and Bias Reduction of Satellite-Based Evapotranspiration Retrievals via Use of In Situ Meteorology and Constrained Canopy Surface Resistance." Journal of Geophysical Research: Biogeosciences, , 10.1029/2018JG004744. ONLINE.


Evapotranspiration (ET), the cumulative quantities of evaporation and plant transpiration, is a leading mechanism, second only to precipitation, for exchange of water between land and the atmosphere, but measurements of ET are limited. NASA’s operational MODIS ET product uses satellite-based surface measurements and reanalysis meteorology to estimate ET, but is substantially biased and fails to resolve spatial heterogeneities in ET over the ARM Southern Great Plains (SGP) site. We diagnose the cause of and propose corrections to resolve these deficiencies.


Improved estimates of ET are critical for agricultural, water resource, and wildfire management; improved understanding of land-atmosphere interactions that influence planetary boundary-layer and convective processes; and processed-based diagnostics of atmospheric models.


Large bias (-38%) in annual ET from the MODIS product is found relative to in situ measurements from ARM eddy correlation flux measurements systems (ECOR) and energy balance Bowen ratio systems (EBBR) across the SGP site. This is attributed to overestimation in the dry canopy surface resistance (rs) parameterization; by constraining rs to the range of typically measured values, the bias in the annual ET can be substantially reduced (+1%). Further, use of coarseā€resolution reanalysis meteorology in the MODIS ET retrievals results in inadequate characterization of the spatial heterogeneity in measured surface fluxes on scales of tens to hundreds of kilometers, such as those observed over the SGP. When derived using high-resolution meteorology, our ET retrievals show substantial improvement in resolving the spatial distribution of the fluxes. The methodology presented here can be used to estimate latent heat fluxes for use in model evaluation, and estimate annual ET for water budget analyses anywhere spatiotemporally resolved meteorological measurements are available.