A method for convective storm detection using satellite data

Abstract

Moisture and instability, along with a triggering mechanism, are the main keys of deep convective storms initiation and evolution. Satellite data can provide indirect measurements of instability and moisture of a wide area in short periods of time. This paper studies the use of an objective method based on a blended use of multiple satellite-based convection estimation techniques. This method is based on different techniques arranged in a several layers approach of different convective features, aiming to stratify a cloud shield. Meteosat Second Generation (MSG) infrared (IR) 10.8μm and water vapor (WV) 6.2μm channels are explored together with tropopause temperature information provided by a numerical model. Threshold, brightness temperature differences (BTD), and time trends are applied to the information available resulting in a five layers product, highlighting areas of different convective activities. This cloud shield stratification method showed a great ability to better evaluate strong convection when compared with simpler techniques such as IR false color, and was especially useful to better identify the strongest convective cell in a large area with several convective outbreaks. A validation analysis was conducted using radar and lightning data, showing that this approach is very helpful in distinguishing very strong cases from weaker ones by pointing out subtle convective patterns only present in severe storms. Also, small changes in storm evolution were more pronounced in the method output. Besides some uncertainties that were observed, likely due to the large viewing angle, techniques derived from MSG spectral bands displayed good accuracy in studying large convective systems in the South America southern region.