Convective Initiation Overview
The convective initiation (CI) product is a nowcasting tool created by researchers at the University of Alabama Huntsville that uses the NOAA Geostationary Satellite (GOES) to track and monitor growing cumulus clouds. The CI algorithm exploits the five infrared channels on GOES and assesses various aspects of cumulus including cloud-top cooling rates, cumulus depth, and the size of the cloud updraft. The algorithm is called the SATellite Convection AnalySis and Tracking (SATCAST) System.
For more information on SATCAST, visit John Mecikalski's UAH webpage.
Figure 1. Example of the CI product centered
over the domain of the NYC TRACON.
Since 2004, SATCAST has been running every half hour over the Southeastern U.S. In addition, a subsection of this domain is processed for the Huntsville WFO. These data are fed directly into AWIPS and are viewable by the forecasters at 30 minute intervals. Specific products being sent to the NWS are 30-min cloud-top cooling rates and the CI "scores" (1-8, with each "score" a particular infrared interest field that is within a favorable range for CI).
During the summers of 2006 and 2007, CI assessment periods were held with the NWS WFO Huntsville. Several training exercises were held to train the forecasters of the usefulness of the algorithm along with the advantages and disadvantages. However, because both of these summers were convectively quiet, very few surveys were received. Currently, the CI product is undergoing some improvements that will easily allow additional WFOs to have access to the CI nowcasts for their individual county warning areas. The CI nowcast product was also used in support of the FAA NYC TRACON project to assist in identifying areas of potential convective activity.
Future work with the CI algorithm will be to incorporate satellite-based radar from Cloudsat and Calipso to help identify differences with the various indicators currently used to make the appropriate adjustments to the CI algorithm. These added tools will help improve probability of detection and minimize false alarm rates.
