Drought in the Pacific Northwest, 1920-2013

Wednesday, 17 December 2014: 12:05 PM
Mu Xiao1, Bart Nijssen1, Yixin Mao1 and Dennis P Lettenmaier2, (1)University of Washington, seattle, WA, United States, (2)University of California, Los Angeles (effective Nov., 2014), Dept. of Geography, Los Angeles, CA, United States
Drought, which is usually defined as an extended period of water supply deficit, is one of the costliest natural hazards. Agricultural, fisheries, hydropower generation and water resource management are all directly affected by droughts. Droughts are usually classified as meteorological, agricultural, or hydrological. The first type is defined based on precipitation and the latter two are connected with hydrological variables (soil moisture and runoff). Because observations of soil moisture and (spatially distributed) runoff are sparse, agricultural and hydrological drought is more difficult to quantify, and hydrological models are often used in their analysis. Nonetheless, the temporal and spatial characteristics of these types of drought are not well understood. Based on hydrological model simulation results for a period of nearly a century, we evaluate the historical characteristics of drought over the Pacific Northwest (PNW) region, defined as the Columbia River basin and coastal drainages. Major droughts in the 1930s, 1977 and early 2000s caused substantial economic losses to the region. We used the VIC macroscale hydrology model to simulate 94 years (1920-2013) of runoff and soil moisture over the region. In order to minimize the effects of model error, we interpret results as percentiles that are calculated according to simulation outputs distribution. We explore in particular the characteristics of drought in the PNW. The main results are: 1, the dry time period are more likely to be included in a drought event while the onset are usually earlier; 2, little snow-melting can lead to hydrological drought (below-normal streamflow) in that year; 3, the spatial pattern of agricultural drought frequency is highly related with precipitation and soil moisture persistence.