Premium
Climatic and physiographic controls of spatial variability in surface water balance over the contiguous U nited S tates using the B udyko relationship
Author(s) -
Abatzoglou John T.,
Ficklin Darren L.
Publication year - 2017
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1002/2017wr020843
Subject(s) - precipitation , evapotranspiration , water balance , environmental science , spatial variability , surface runoff , snow , hydrology (agriculture) , aridity index , arid , surface water , vegetation (pathology) , atmospheric sciences , geology , geography , ecology , meteorology , mathematics , statistics , medicine , paleontology , geotechnical engineering , pathology , environmental engineering , biology
The geographic variability in the partitioning of precipitation into surface runoff (Q) and evapotranspiration (ET) is fundamental to understanding regional water availability. The Budyko equation suggests this partitioning is strictly a function of aridity, yet observed deviations from this relationship for individual watersheds impede using the framework to model surface water balance in ungauged catchments and under future climate and land use scenarios. A set of climatic, physiographic, and vegetation metrics were used to model the spatial variability in the partitioning of precipitation for 211 watersheds across the contiguous United States (CONUS) within Budyko's framework through the free parameter ω . A generalized additive model found that four widely available variables, precipitation seasonality, the ratio of soil water holding capacity to precipitation, topographic slope, and the fraction of precipitation falling as snow, explained 81.2% of the variability in ω . The ω model applied to the Budyko equation explained 97% of the spatial variability in long‐term Q for an independent set of watersheds. The ω model was also applied to estimate the long‐term water balance across the CONUS for both contemporary and mid‐21st century conditions. The modeled partitioning of observed precipitation to Q and ET compared favorably across the CONUS with estimates from more sophisticated land‐surface modeling efforts. For mid‐21st century conditions, the model simulated an increase in the fraction of precipitation used by ET across the CONUS with declines in Q for much of the eastern CONUS and mountainous watersheds across the western United States.