One of the objectives of the GOES-R Algorithm Intercomparison Test Plan is to objectively intercompare the performance of different nowcasting algorithms under consideration by the Algorithm Working Group (AWG) Hydrologic Algorithm Team (AT) for the implementation on the Advanced Baseline Imager (ABI) onboard GOES-R.

This study shows the performance of a particular nowcasting technique (ForTrACC, Vila et al, 2008) during January, April, July and October 2005 using SEVIRI thermal IR channel (10.8μm) for several validation regions in South America, Europe and Africa. In this case, following the guidelines of AWG, the forecast is performed every 3 hours beginning at 00:00Z and covers every 15 minutes out to a lead time of 3 hours.

The statistics are performed based on contingency tables. This methodology shows the frequency of 'yes' and 'no' forecast and observations (pixels whose brightness temperature are below certain threshold) and several statistical parameters are performed using this information like Probability of Detection (POD), False Alarm Ratio (FAR), among others.