干旱区生态环境知识资源中心

High-elevation snowpacks provide critical inputs to the hydrological system of mountainous semi-arid regions where summer precipitation is insufficient to maintain adequate discharges for ecological and economic needs. The Okanagan Basin in Western Canada is an example of such a system, as most of the summer streamflow is derived from snowmelt. To better understand how snowmelt events vary as a result of atmospheric conditions, this study developed statistical models using upper-air atmospheric data for evaluating changes in snowpack ablation. Specifically, radiosonde data were statistically linked with detailed ground-based measurements of snowmelt and associated streamflow. Statistical models were developed based on data from the 2007 ablation season and concurrent data from the 850 hPa geopotential height. These models explained 57-68% of the variance in snowmelt for 2007, and were extended to predict snowmelt for the radiosonde period of record (1972-2012). Time-series analyses showed significant trends toward higher winter and spring temperatures, vertical temperature gradients in the atmospheric boundary layer in spring, and earlier dates for snowmelt and freshet initiation. Significant negative trends were also found towards decreasing spring precipitation. More broadly, ablationseason climatic and hydrological variables were significantly positively correlated with the winter and spring Multivariate El Nil-10 Southern Oscillation and Pacific Decadal Oscillation indices, in which the positive (negative) phase was associated with higher (lower) magnitude and frequency of melt events. This combination of strong correlations and significant temporal trends indicates that with projected air-temperature increases, the magnitude and duration of melt events are likely to increase, particularly during favourable phases of the above teleconnections. Crown Copyright (C) 2014 Published by Elsevier B.V. All rights reserved.