Cloud-top temperatures for precipitating winter clouds

A winter study of precipitating clouds using remotely sensed brightness temperatures from the Geostationary Operational Environmental Satellite (GOES) 8/12 Imager was conducted during the second half of the winter of 2003 (February through March) and during the winter of 2004 (December through March). The study assumes that 1) the remotely sensed cloud top is the cloud that produced the hydrometeor, 2) the cloud top temperature is the coldest temperature in the cloud, and 3) the cloud that produced the hydrometeor behaves as a blackbody. Data was collected using an acquisition algorithm to retrieve cloud top temperatures above every surface station reporting precipitation throughout the contiguous United States (CONUS). Cloud top temperatures were determined using a 2 x 2 pixel box surrounding the station with pixel resolution at satellite subpoint of 2.3 km x 4.0 km (Menzel and Purdom 1994, Schmitt et al. 2001). This method was chosen to alleviate parallax and navigational errors in satellite data. Data was included in the sample if the standard deviation was less than or equal to 1°C. The data was then separated into four categories (rain, snow, freezing rain and snow) and graphed in a frequency distribution. Results of the study indicate that liquid precipitation tends to exhibit a bi-modal distribution with modes near -45°C and -14°C. Similar frequency distributions between the rain and freezing rain graph suggest that the only difference between these processes is the near surface temperature profile. Snowfall reports tend to exhibit a Gaussian distribution that is skewed heavily towards warmer temperatures with a very pronounced mode at -16°C. A very sharp drop in the frequency distribution warmer than -12°C is indicative of the depletion of ice nuclei with progressively warmer temperatures. Sleet reports were too few to draw any significant conclusions.