The Cloud Readers

The science of local rain monitoring may be more accurate in the near future, thanks to a new ICTP study that is the first to focus on the sensitivities of cloud and precipitation sensors at the regional scale.

The study, titled "Using CloudSat cloud retrievals to differentiate satellite-derived rainfall products over West Africa," has been published in the American Meteorological Society's Journal of Hydrometeorology. The authors are Adrian M. Tompkins of ICTP's Earth System Physics section and Adeyemi A. Adebiyi, an alumnus of ICTP's Postgraduate Diploma Programme who is now a PhD student at the University of Miami, Florida, USA.

According to Tompkins, knowledge of precipitation is particularly important in Africa, where much agriculture is rain-fed (without irrigation) and yields depend not only on seasonal rain totals but also on the sub-seasonal variability. "A dry spell at a critical stage in the growing season can lead to poor yields," he says.

However, having an accurate knowledge of rainfall in Africa can be difficult, because of a sparsity of available ground-based observations. "This means that one has to rely on satellite-derived products," explains Tompkins.  

There is now a wide choice of these products using different combinations of sensors and a diversity of mathematical algorithms to combine data into a final rainfall amount.  To date, these satellite products have been validated in a series of studies on a local scale where ground-based rain gauge data exists.  But these studies are often contradictory, according to Tompkins, and highlight a different product as "best" depending on the region in question.  "This is because the cloud systems producing the rain change from region to region, and each satellite product may be more accurate for certain kinds of cloud systems," he says.

In order to tackle this uncertainty, Tompkins and Adebiyi have taken data from a new satellite launched in 2006 called "CloudSat" that detects ice and liquid cloud crystals. They used the data to study how the rainfall amounts detected by each satellite rainfall algorithm changes with the associated cloud structure detected by CloudSat.  Explains Tompkins, "This is the first study to have combined this array of cloud and precipitation sensors, and it allowed us to understand the sensitivity of each algorithm and why one product may perform better than another for a given region."

Tompkins says that the research allows users of satellite precipitation products to better understand the potential pitfalls of each respective product, and to make a more informed decision of which to employ for their particular application and region of interest.  Moreover, the research indicates where to focus efforts to improve future generations of rainfall retrieval algorithms.