Sensitivity of 30-Day Dynamical Forecasts to Continental Snow cover

Abstract

Several series of 30-day simulations with a global simulation model are used to evaluate the sensitivities to continental snow cover over North America and Eurasia. The model is initialized with National Meteorological Center analyses for specific dates during the winter of 1976/77 through 1983/84, and snow cover in each case is prescribed according to 1) the distribution derived from observational data; and 2) the distribution containing a corresponding anomaly of the opposite sign. In ten pairs of midwinter forecasts, the major effect of extensive snow cover in eastern North America is a reduction of the near-surface air temperature in the vicinity of the snow anomaly. When snow cover is extensive, sea level pressures are somewhat lower and precipitation amounts somewhat higher offshore of the East Coast; sea level pressures are generally higher inland. In a set of six March cases, positive anomalies of Eurasian snow cover reduce the air temperatures by at least several degrees ceisius throughout the lower half of the troposphere in the region over and downstream of the snow anomaly. The positive Eurasian snow anomalies also produce systematically lower pressures and upper-air heights in the Aleutian region, higher pressures in the Asian Arctic, and lower pressures over western Europe and the extreme northeastern Atlantic. In the Eurasian experiments, the 30-day forecast pressures for the Eurasian hemisphere vary with snow coverage in a manner consistent with the observed pressure fields of the same months.