Assessing the impacts of water management on evapotranspiration in the Colorado River Basin

Stephanie Castle1, John T Reager II2, Brian F Thomas2, Adam J Purdy3, Min-Hui Lo4, Matthew Rodell5 and James S Famiglietti6,7, (1)University of California Irvine, Irvine, CA, United States, (2)NASA Jet Propulsion Laboratory, Pasadena, CA, United States, (3)University of California Irvine, Irvine, United States, (4)NTU National Taiwan University, Taipei, Taiwan, (5)NASA GSFC HSL, Greenbelt, United States, (6)Arizona State University, School of Sustainability, Tempe, United States, (7)University of Saskatchewan, School of Environment and Sustainability, Saskatoon, Canada
Abstract:
The complexity involved in accurate observation and numerical simulation of regional evapotranspiration (ET) can lead to a large range of uncertainty between techniques, usually attributed to methodological deficiencies. In this study, we hypothesize instead that such discrepancies in ET estimates are sometimes warranted, and in extreme cases can be applied to measure the effect of anthropogenic influences on ET fluxes in a developed river basin. We compare an ensemble of monthly ET estimates from land surface models (as a representation of “natural” ET fluxes) with monthly satellite-based estimates (as a representation of “realistic” ET fluxes) in the intensely managed Colorado River Basin (CRB) to contrast human impacts and natural variability. Two remote sensing estimates of ET are evaluated: a water balance approach applying GRACE data and a surface radiation budget approach using MODIS and GOES data. A comparison between model ensemble and satellite-based ET indicates that satellite observations yield larger amplitudes during summer months in the CRB, primarily during the growing season. The CRB has a total observation-based ET flux of 142 ± 5 MAF yr-1, and we estimate approximately 10% of this flux (14.9 +/- 3 MAF yr-1) is due to anthropogenic influences. Our results highlight the relative importance of human impacts on the terrestrial water cycle.