Estimating Annual Groundwater Evapotranspiration from Hydrographic Areas in the Great Basin Using Remote Sensing and Evapotranspiration Data Measured by Flux Tower Systems
Loading...
Authors
Minor, Blake
Issue Date
2019
Type
Thesis
Language
Keywords
Evapotranspiration , Groundwater , Phreatophytes , Satellites , Vegetation , Water
Alternative Title
Abstract
Rising concerns about water availability in the Great Basin in the face of increasing population and water demand have prompted multiple groundwater studies focused on estimating groundwater recharge and discharge. Groundwater is the primary water supply for many hydrographic areas (HAs) in the Great Basin, thus a thorough understanding of the groundwater budget is fundamental for making meaningful predictions as groundwater is developed and ultimately consumed for beneficial use. Estimating groundwater recharge from precipitation (PPT) is challenging and has substantial uncertainty; therefore, much focus is placed on estimating the groundwater discharge in many HAs since estimating discharge is much more constrained and therefore less uncertain. Groundwater discharge in the Great Basin primarily occurs via bare soil evaporation and evapotranspiration (ET) from phreatophyte vegetation within the valley lowlands. Many recent studies have estimated annual groundwater ET by relating ET estimates obtained from Eddy Covariance (EC) and Bowen Ratio (BR) micrometeorological stations, to remotely sensed vegetation indices (VIs) derived from mid-summer Landsat imagery. The objective of this work was to develop and assess the uncertainty of a statistical relationship between annual ET estimates derived from EC and BR stations located in phreatophyte areas, and remotely sensed VIs. A normalized ET index, ET*, was developed at each study site using energy balance closure corrected ET normalized by evaporative demand (ETo) and PPT derived from GRIDMET data. ET* values at each site were related to the source area Normalized Difference Vegetation Index (NDVI), calculated from Landsat Collection 1 Surface Reflectance data. ET* values for 54 site-years of data correlated well with source area NDVI (R2=0.84). The statistical model was applied to the groundwater discharge areas of five HAs in the Great Basin using the mid- to late-summer images from the Landsat data archive (1984-2018) in order to develop estimates of median annual groundwater ET (ETg). Median annual ETg estimates compare relatively well with the estimates developed in previous studies of the five HAs. While uncertainty of the model and predictions of groundwater discharge can be large in some cases (e.g. ± 18% of the mean value estimated for Crescent Valley), uncertainty in groundwater discharge is argued to be substantially less than groundwater recharge. Results from this work will enable the estimation of groundwater discharge over large areas and time periods for which in-situ data does not exist.