Researcher from UT developed an algorithm to predict rainfall in rural Africa

A researcher from the University of Twente developed an algorithm to help detect rainfall in rural Africa. As revealed in a press release, Kingsley Kumah designed the technology utilizing satellite data and training it with information from phone signals. 

On a daily basis, people living in Europe easily find information about the weather on their phones. Radars of high resolution are responsible for these predictions, which are unavailable in some parts of the world, including Africa.

AI and cloud seeding could protect against extreme weather

The hurricane that swept over Florida last week, the summer drought in Europe, floods in Pakistan, forest fires and the melting of glaciers in the Alps: these were all topics that came up last week at the 12th ExtremWetterKongressin Hamburg, Germany.

Purpose of the study

Kumah’s research has the purpose of solving that issue, providing weather predictions by combining satellite data with phone signals. The satellites cannot support small farmers by themselves since they cover a larger territory in scale. 

However, attaching the data from the satellites to the information from the phones makes it possible to train the algorithm to make predictions. Even considering potential challenges, like the instability of mobile operations in the area, the device provides enough accuracy, and the algorithm is in constant training, according to Kumah.  

In the future, Kumah dreams of projecting a product that can be easily accessed through smartphones. From this product, farmers and people who live in those regions can become aware of the weather forecast like everyone else.