Development and application of downscaled hydroclimatic predictor variables for use in climate vulnerability and assessment studies.
Abstract
This paper outlines the production of 270 meter grid‐scale maps for 14 climate and derivative
hydrologic variables for a region that encompasses the State of California and all the streams
that flow into it. The paper describes the Basin Characterization Model (BCM), a map‐based,
mechanistic model used to process the hydrological variables. Three historic and three future
time periods of 30 years (1911–1940, 1941–1970, 1971–2000, 2010–2039, 2040–2069, and 2070–
2099) were developed that summarize 180 years of monthly historic and future climate values.
These comprise a standardized set of fine‐scale climate data that were shared with 14 research
groups, including the U.S. National Park Service and several University of California groups as
part of this project. The paper presents three analyses done with the outputs from the Basin
Characterization Model: trends in hydrologic variables over baseline, the most recent 30‐year
period; a calibration and validation effort that uses measured discharge values from 139
streamgages and compares those to Basin Characterization Model‐derived projections of
discharge for the same basins; and an assessment of the trends of specific hydrological variables
that links historical trend to projected future change under four future climate projections.
Overall, increases in potential evapotranspiration dominate other influences in future
hydrologic cycles. Increased potential evapotranspiration drives decreasing runoff even under
forecasts with increased precipitation, and drives increased climatic water deficit, which may
lead to conversion of dominant vegetation types across large parts of the study region, as well
as have implications for rain‐fed agriculture. The potential evapotranspiration is driven by air
temperatures, and the Basin Characterization Model permits it to be integrated with a water
balance model that can be derived for landscapes and summarized by watershed. These results
show the utility of using a process‐based model with modules representing different
hydrological pathways that can be interlinked.