Asymmetry in hydrologic response to climate change in Western North America – A Land-Atmosphere Interactions Perspective

Abstract:

A number of studies have found mixed hydrologic trends in North America, e.g., increasing stream flow in the Mississippi river basin and decreasing stream flow in the Pacific Northwest. This study aims to bring out an important detail about the asymmetry in hydrologic response by applying the Budyko framework which characterizes watersheds as energy or water limited using a ratio of climatological mean evaporative power to precipitation. We studied long-term hydrologic changes in 46 large unimpaired watersheds (drainage area > 1000 km²) in the western United States and equivalent watersheds in western Canada. Daily stream flow observations from the United States Geological Survey (USGS) and Water Survey of Canada, and PRISM based high resolution (4-km) monthly temperature and precipitation data are employed. We studied long-term trends (1965 to 2010) in 18 hydroclimatic indicators that include annual and seasonal stream flow, precipitation, temperature, and timing of center-of-volume. The trends’ statistical significance is assessed using Mann-Kendall test with long-term persistence consideration.

We found the highest signal in winter (Jan.-Feb.-Mar.) minimum temperature where 48% of all watersheds in western United States show significant increasing trends. Among hydrologic indicators, we found highest signal in summer stream flow (normalized by annual precipitation) where 17% of all watersheds show significant decreasing trends. Further, we group these watersheds in two groups using the Budyko framework: (1) energy limited (15 watersheds), e.g. Pacific Northwest, and (2) water limited (31 watersheds), e.g. southwestern United States. We found 33% of energy limited watersheds and 10% of water limited watersheds show a significant decreasing trend in summer water availability. Averaged across all watersheds in the respective energy and water limited regions, we found a significant decreasing trend in the energy limited regions and no trend in water limited regions. A theoretical basis or a hypothesis for the asymmetry in hydrologic response is proposed. We plan to test the hypothesis using downscaled climate model data and hydrologic modeling approaches. We are also updating the analysis to include unimpaired stream flow stations from western Canada and more stations from the western United States.