Trends in snowfall versus rainfall in the Western United States--Revisited

Abstract:

Knowles et al. (J. Climate, 2006) documented long-term (1949-2004) trends in precipitation form, with a smaller fraction of precipitation falling, in recent decades, on days with reported snow compared to days when no snow was reported (and when precipitation was presumably rain). This precipitation-amount-corrected trend was found at three-quarters of 261 cooperative weather stations across the region. The trends correlated with corresponding trends towards warmer winter air temperatures at the weather stations involved.

An update of those analyses through the more recent period indicates that the overall swing towards less precipitation fraction occurring on snowy days has continued through the intervening years, with 21^st Century rain/snow fractions remaining significantly higher than historical norms at most stations. The same data have also been used to develop site-specific statistical relations between precipitation form (snowy-day precipitation vs purely rainy day) and air temperatures by logistical regressions at over 200 stations across the West, to determine whether the general temperature trends mentioned above have, in fact, been large enough to explain the trending precipitation forms. That is, were the warming trends detected across the West large enough to actually raise temperatures above the local snow-rain thresholds at most stations? The regression relations show that the temperature at which half of the wet days have been snowy historically varies, from station to station, across a range from -2ºC to +4ºC. Thus at some stations winter storm temperatures would have to rise above about -2ºC to markedly impact precipitation forms, while at other stations, temperature had to rise above +4ºC. Nonetheless, observed temperature trends since 1950 have been sufficient to explain the observed regional precipitation-form trends. The fitted precipitation form-temperature relations also provide a basis for estimating precipitation forms in hydrological models and in climate-change projections across the region, allowing—for example—more geographically informed projections of precipitation-form changes under future climates. On the whole, though, the expected relations between warming trends and precipitation-form trends found by Knowles et al. (2006) continue to hold.