Does Distance from the Forecast Site Affect Skill?

Abstract

An outstanding issue in the assessment of forecast skill (and value) is whether any advantage that can be obtained through regional knowledge not readily available to distant forecasters supersedes the leveling effect of information obtained by all forecasters (through the interpretation of numerical weather predictions). An analysis of 1 yr of data from the National Collegiate Weather Forecasting Contest was conducted in order to evaluate whether physical separation from the forecast site (defined by distances outside of and within 1000 km, with a minimum separation of 100 km) has a measurable effect on skill. The results indicate that regional effects (on the meso-α scale) are manifested in forecasts of both temperature (maximum and minimum) and precipitation amount (by category). Furthermore, these effects are a function of the experience level of the forecaster. Specifically, experienced forecasters are able to use regional knowledge to their advantage in forecasting temperature and precipitation amount, while their less-experienced counterparts cannot advantageously use such information for either type of forecast. The implication of these results with respect to the allocation of National Weather Service resources is also addressed.