H32E-04:
Evaluating ET and Its Components from the CMIP5 Models with New, Global Remote Sensing-Based Estimates
Wednesday, 17 December 2014: 11:05 AM
Manish Verma1, Joshua B Fisher1, Kaniska Mallick2, Youngryel Ryu3, Kevin P Tu4, Hideki Koayashi5, Alexandre Guillaume1, Gregory Moore1, Lavanya Ramakrishnan6 and Val Hendrix6, (1)NASA Jet Propulsion Laboratory, Pasadena, CA, United States, (2)Centre de Recherche Public - Gabriel Lippmann, Belvaux, Luxembourg, (3)Seoul National University, Department of Landscape Architecture and Rural Systems Engineering, Seoul, South Korea, (4)Theiss Research, Davis, CA, United States, (5)Japan Agency for Marine-Earth Science and Technology, Department of Environmental Geochemical Cycle Research, Yokohama, Japan, (6)Lawrence Berkeley National Lab, Berkeley, CA, United States
Abstract:
Accurate and reliable remote sensing-based estimates are required to evaluate the predictive accuracy of Earth System models. We combined MODIS land and atmosphere data and produced global daily evapotranspiration (ET) estimates at 5 kilometer using four algorithms that follow different approaches. The algorithms included a Priestly-Taylor type formulation (PT-JPL), a Penmen-Monteith-based approach (MOD16), ET estimated as a residual (Surface Energy Budget System; SEBS), and a formulation that derives surface and canopy resistances analytically (PMBL). We compared remote sensing-based ET estimates with measurements from more than 100 FLUXNET sites and quantified uncertainty and error in ET estimates from each of the four algorithms. We then used these validated remote sensing-based ET estimates as reference data and evaluated the accuracy and reliability of ET and its three components (soil evaporation, canopy interception evaporation, and transpiration) from the historical and decadal simulations of 20 different Earth System models participating in the Coupled Model Intercomparison Project Phase 5 (CMIP5). We quantified the predictive skill of each model. Finally, we weighted each model based on its predictive skill and analyzed spatiotemporal variations in projected ET and its components for the next 50 years.
