Understanding Large-scale Plant Water Relations Using Remote Sensing and Models

Plant photosynthesis plays a major role in ecosystem-climate feedback. Drought-induced declines in photosynthesis and associated forest die-off events have increased considerably in the past decades due to climate warming. Currently, there is substantial uncertainty in understanding ecosystem vulnerability to drought. This impedes an accurate projection of climate change impact on ecosystem's capability to store carbon and to provide services to humans. Many ecosystems such as forests and shrubs tend to have deep root systems. During droughts, they could access water beyond the groundwater table. Therefore, to understand the drought response in an ecosystem would require information of moisture supply in the deep soil. However, existing observations of underground water supply over large areas are limited to a shallow soil depth (<5 cm). In this dissertation, we establish a framework that uses terrestrial water storage (TWS) from the Gravity Recovery and Climate Experiment (GRACE) to complement other atmospheric and shallow surface drought indictors (e.g., precipitation, temperature, surface soil moisture) by providing information about both root zone soil moisture and groundwater changes. We use this framework to study the correspondence between vegetation status and drought intensification as it moves from early-stage precipitation shortage to shallow and deep soil layers. In addition to droughts, we evaluate how ecosystems respond to shifts in long-term mean moisture supply and demand conditions. Finally, we evaluate human ecological restoration impact on total terrestrial water storage in northern China. This work demonstrates that GRACE TWS is a reliable proxy for total plant-available water supply and advances our understanding of large-scale plant water relations from space