Special Issue Paper
Estimating evapotranspiration using an observation based terrestrial water budget
Version of Record online: 7 DEC 2011
Copyright © 2011 John Wiley & Sons, Ltd.
Special Issue: The role of remote sensing observations and models in hydrology: the science of evapotranspiration
Volume 25, Issue 26, pages 4082–4092, 30 December 2011
How to Cite
Rodell, M., McWilliams, E. B., Famiglietti, J. S., Beaudoing, H. K. and Nigro, J. (2011), Estimating evapotranspiration using an observation based terrestrial water budget. Hydrol. Process., 25: 4082–4092. doi: 10.1002/hyp.8369
- Issue online: 22 DEC 2011
- Version of Record online: 7 DEC 2011
- Accepted manuscript online: 21 OCT 2011 05:39PM EST
- Manuscript Accepted: 17 OCT 2011
- Manuscript Received: 10 OCT 2011
- remote sensing;
- water budget
Evapotranspiration (ET) is difficult to measure at the scales of climate models and climate variability. While satellite retrieval algorithms do exist, their accuracy is limited by the sparseness of in situ observations available for calibration and validation, which themselves may be unrepresentative of 500 m and larger scale satellite footprints and grid pixels. Here, we use a combination of satellite and ground-based observations to close the water budgets of seven continental scale river basins (Mackenzie, Fraser, Nelson, Mississippi, Tocantins, Danube, and Ubangi), estimating mean ET as a residual. For any river basin, ET must equal total precipitation minus net runoff minus the change in total terrestrial water storage (TWS), in order for mass to be conserved. We make use of precipitation from two global observation-based products, archived runoff data, and TWS changes from the Gravity Recovery and Climate Experiment (GRACE) satellite mission. We demonstrate that while uncertainty in the water budget-based estimates of monthly ET is often too large for those estimates to be useful, the uncertainty in the mean annual cycle is small enough that it is practical for evaluating other ET products. Here, we evaluate five land surface model simulations, two operational atmospheric analyses, and a recent global reanalysis product based on our results. An important outcome is that the water budget-based ET time series in two tropical river basins, one in Brazil and the other in central Africa, exhibit a weak annual cycle, which may help to resolve debate about the strength of the annual cycle of ET in such regions and how ET is constrained throughout the year. The methods described will be useful for water and energy budget studies, weather and climate model assessments, and satellite-based ET retrieval optimization. Copyright © 2011 John Wiley & Sons, Ltd.