Using Satellite Observations of Soil Moisture to Improve Modeling of Terrestrial Water Cycles

Terrestrial evapotranspiration (ET) describes the flux of water from the Earth’s surface to the atmosphere, calculated as the sum of evaporation from soil and leaf surfaces, and transpiration through plant stomata. ET is the largest terrestrial water flux, returning over half of the precipitation that falls on land back to the atmosphere, annually. Additionally, ET plays a key role in Earth’s carbon, water, and energy cycles, linking them together via the movement of water and CO2 through plant stomata. Because of its important role in these Earth system processes, it is essential that existing methods of measuring and modeling ET are accurate. A common method for estimating and monitoring ET at global scales is through satellite remote sensing. The remote sensing-based models use a combination of satellite observed vegetation and surface meteorology to estimate ET. Although these models can be effective at representing global patterns of ET, a common shortfall is that few use soil moisture as a direct model input. The lack of soil moisture information in these models can significantly degrade ET estimates, as soil moisture is tightly linked to both soil evaporation and plant transpiration. This thesis addresses this gap by introducing a satellite observed soil moisture control to an existing operational remote sensing-based ET model, MOD16. The results show that introducing a soil moisture control to MOD16 improves estimates of ET across a wide range of climates and vegetation types within the contiguous United States study area. This research provides an improved regional representation of ET and clarifies the role of soil moisture in regulating terrestrial ET and the water cycle. These results can be used to better understand and predict shifts in the regional water cycle induced by drought and climate change