The study applies multiple satellite-based fog identification products to fog cases that occurred between July 1, 2007 and June 30, 2008 across the US and Canada. Conditions that can lead to non-identification or false identification include high cloud contamination or snow cover. Each algorithm uses different property tests to refine the identification decision and improve Probability of Detection (POD) and False Alarm Ratio (FAR) scores. Surface data are used to evaluate properties such as ceiling height, microphysical properties, and albedo as well as to confirm a positive identification. Using the same cases to test every algorithm will enable us to compare the merits of each algorithm in relation to the others. The study intends to establish the strength of individual algorithms under specific operational conditions.
|Presenter:||Caitlin Hart (Undergraduate Student)|
|Time:||2:30 pm (Session IV)|