Articles | Volume 8, issue 10
Geosci. Model Dev., 8, 3021–3031, 2015
Geosci. Model Dev., 8, 3021–3031, 2015

Development and technical paper 02 Oct 2015

Development and technical paper | 02 Oct 2015

Using satellite-based estimates of evapotranspiration and groundwater changes to determine anthropogenic water fluxes in land surface models

R. G. Anderson1,*, M.-H. Lo2,*, S. Swenson3, J. S. Famiglietti4, Q. Tang5, T. H. Skaggs1, Y.-H. Lin2, and R.-J. Wu3 R. G. Anderson et al.
  • 1USDA, Agricultural Research Service, US Salinity Laboratory, Contaminant Fate and Transport Unit, Riverside, CA, USA
  • 2National Taiwan University, Department of Atmospheric Sciences, Taipei, Taiwan
  • 3National Center for Atmospheric Research, Advanced Study Program, Boulder, CO, USA
  • 4Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Dr., Pasadena, CA 91109, USA
  • 5Key Laboratory of Water Cycle and Related Land Surface Processes, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China
  • *These authors contributed equally to this work.

Abstract. Irrigation is a widely used water management practice that is often poorly parameterized in land surface and climate models. Previous studies have addressed this issue via use of irrigation area, applied water inventory data, or soil moisture content. These approaches have a variety of drawbacks including data latency, accurately prescribing irrigation intensity, and a lack of conservation of water volume for models using a prescribed soil moisture approach. In this study, we parameterize irrigation fluxes using satellite observations of evapotranspiration (ET) compared to ET from a suite of land surface models without irrigation. We then incorporate the irrigation flux into the Community Land Model (CLM) and use a systematic trial-and-error procedure to determine the ground- and surface-water withdrawals that are necessary to balance the new irrigation flux. The resulting CLM simulation with irrigation produces ET that matches the magnitude and seasonality of observed satellite ET well, with a mean difference of 6.3 mm month−1 and a correlation of 0.95. Differences between the new CLM ET values and satellite-observed ET values are always less than 30 mm month−1 and the differences show no pattern with respect to seasonality. The results reinforce the importance of accurately parameterizing anthropogenic hydrologic fluxes into land surface and climate models to assess environmental change under current and future climates and land management regimes.

Short summary
Current land surface models (LSMs) poorly represent irrigation impacts on regional hydrology. Approaches to include irrigation in LSMs are based on either potentially outdated irrigation inventory data or soil moisture curves that are not constrained by regional water balances. We use satellite remote sensing of actual ET and groundwater depletion to develop recent estimates of regional irrigation data. Remote sensing parameterizations of irrigation improve model performance.