Estimates of regional evapotranspiration (ET) are critical for streamflow forecasts, global climate studies, as well as short and long-term water supply management. There have been numerous advances in monitoring and quantifying ET and the increasingly available arrays of remotely-sensed variables have allowed significant developments in estimating ET from space. The current study focuses on developing relatively simple, near real-time algorithms to estimate ET utilizing a range of remotely-sensed products. As a first step, a satellite based Potential ET (PET) product is formulated at the daily time-step for all-sky (clear and cloudy) conditions at a moderate resolution (∼1km). Since PET provides an upper limit of actual ET under the assumption of a well-watered surface, ET can be then be derived as a function of PET by taking into account water availability (i.e. soil moisture). After development of the PET product, we utilize a physically based model (Noah LSM) to account for soil moisture availability and derive an ET product as a function of NOAH model products and the MODIS-based PET. We also explore an ET methodology based on development of an evaporative fraction (EF) through an empirical model using remotely-sensed land surface temperature and vegetation inde). Both approaches provide a ratio of ET to PET. By integrating with previously derived estimates of PET, an actual ET is established. Finally, representation of soil moisture (∼the ratio of ET/PET) is also explored and a downscaling algorithm to provide critical high-resolution soil moisture information is derived. In general, validation has shown that the developed ET model provides reasonable estimation of the temporal and spatial dynamics of vegetation and soil water loss. The soil moisture derived from the developed downscaling approach shows significant improvement over several existing methods and provides insight on the potential usefulness of soil moisture estimation in the ET model. In conclusion, results from the current study indicate that the best hope for providing robust estimates of the terrestrial dynamics of ET is accomplished by integrating a combination of visible/infrared and microwave observations. The methods developed in this study are a critical step in providing more reliable and efficient spatial and temporal estimates of ET for water resource and ecosystem management.
展开▼
