Comparison of Noah-MP and VIC long-term drought predictions over the Western U.S.

Land Surface Models (LSMs) developed to represent the land surface in coupled land-atmosphere weather and climate models are useful in their own right for off-line simulations and reconstructions of land surface hydrology at regional, continental, and global scales. In off-line applications, LSMs have been shown to provide plausible root zone soil moisture simulations for drought monitoring for instance, and are now used as one source of information by the U.S. Drought Monitor. Here we compare the performance of two widely used LSMs, the Variable Infiltration Capacity (VIC) macroscale hydrology model and the Noah with multiparameterization (Noah-MP) LSM, in the context of long-term drought monitoring over the Western U.S. We ran the two models at a spatial resolution of 1/16^th degree, and daily time step for VIC and hourly for Noah-MP for the period 1920-present. We conducted a severity-area-duration (SAD) analysis, in which severity was measured by total moisture percentile with respect to the climatological period 1920-2010, to identify and characterize the most severe events at various durations and spatial extents. In general, both models reconstruct known severe drought periods over the last century, such as in 1930s, 1976-77, etc.. The Noah-MP and VIC showed similar spatial extents and severities of these drought events. Both models have been implemented in the UCLA Western U.S. Drought Monitoring System (http://www.hydro.ucla.edu/SurfaceWaterGroup/forecast/monitor_west/index.shtml) where drought conditions across the West can be viewed in real time.