CR 10:217-227 (1998)  -  doi:10.3354/cr010217

ABSTRACT: Analyses of freezing rain in the United States are restricted to the use of a sparse network of first-order stations that provide hourly reports of weather type. From this network, spatial analyses of freezing rain are limited to broad scales. Terrain features, however, can strongly modify these patterns at finer spatial scales. This is especially the case immediately east of the Appalachian Mountains, where cold air wedges commonly set up in conjunction with the approach of cyclonic disturbances and precipitation. In this study, an empirical approach is developed whereby daily climatological variables are used to predict mean annual freezing rain attributes over a dense network of cooperative observation stations in the Appalachian region. Since freezing rain occurs when temperatures are at or below the freezing point, it is hypothesized that mean annual frequencies of daily high and low temperatures in a given range near 0°C on days in which rain is observed (wet days) provide a surrogate for various annual attributes of freezing rain. In order to test this hypothesis, these mean annual temperature characteristics are correlated with the mean annual number of freezing rain events (NFR) and hours of freezing rain (HFR) from a sample of first-order weather stations that observe freezing rain. Using the strongest identified relationships, regression models are developed to predict HFR and NFR. Over 80% of the variance in HFR and NFR is explained by the mean annual frequencies of daily minimum temperature less than 0 and -2°C, respectively, on wet days. These models are applied to daily climatological (e.g. temperature, precipitation, and snowfall) summaries at selected cooperative observer sites in order to generate estimates of freezing rain occurrence. These estimations provide input for maps that better delineate spatial patterns of freezing rain.

KEY WORDS: Freezing rain · Icing · Appalachian