Evaluation of a MODIS Triangle-based Algorithm for Improving ET Estimates in the Northern Sierra Nevada Mountain Range

Evapotranspiration (ET) is a crucial component in the hydrologic cycle and consequently the disciplines of hydrology, meteorology, agriculture, and climate change science. As a controlling factor, ET governs the water cycle and energy transport among the biosphere, atmosphere and hydrosphere, and thus plays a large role in the estimation of regional-scale hydrologic processes and large-scale atmospheric circulation and global climate change. Accurate characterization of the rate and amount of ET across both spatial and temporal scales is therefore very important, especially in regions where water deficiency is beginning to cause economic hardships and constraints on sustainable development. The current study aims to evaluate the effectiveness of the triangle method in ET estimation in the sub-alpine region of the Northern Sierra Nevada Mountain Range. The approach uses a previously developed net radiation model and MODIS-based remote-sensing-based algorithm with added temperature (topographic) correction. This approach is beneficial in that we utilized solely MODIS-based products for the estimation of evaporative fraction (EF), net radiation (Rnet) and ground heat flux (G) during both clear and cloudy skies. This approach, minus temperature correction, has been successfully implemented and evaluated in the semi-arid region of southern Arizona. However, little work has been done on evaluation of this approach in sub-alpine environments, where temperature fluctuations due to elevation inhibit key assumptions made by the triangle method. We compare the proposed MODIS-derived Rnet and 8-day average ET products against ground-based observations at four sites in the Sagehen watershed, located on the eastern slope of the Northern Sierra Nevada Mountain Range. Results show correlation coefficients of 0.74 to 0.80 and root mean square error values of 121 to 167 W/m2 when comparing model simulated Rnet to observed values. These values are slightly improved relative to those reported in southern Arizona. Likewise, simulated 8-day average ET values show relatively good correlation (0.52 to 0.67) and root mean square error (62 to 81 W/m2) when compared to observations. Given few ground-based measurements are available in sub-alpine environments, the proposed MODIS-based algorithm with topographic correction provides a relatively reasonable estimate of ET on an improved spatial and temporal scale. This work is part of ongoing research in developing and implementing advanced products into operational hydrologic forecasting, including the National Weather Service (NWS) and associated River Forecast Centers (NWS-RFC) in ungauged and highly altered (drought, forest fire, beetle kill) western watersheds.