Precipitation in Africa: the future of observation, nowcasting and forecasting
| On the world’s fastest urbanizing continent, Africa, urban floods are a big and growing problem. Early warning is the first important step in flood risk management. This requires continuous and reliable precipitation measurements and forecasts, which are not always available in African cities. Four experts form a panel for this session. All of them have their own innovative way to overcome the precipitation information challenge in Africa. Their solutions include the use of machine learning combined with earth observation data, the design of low-cost sensors, and the application of commercial microwave links from cellular telecommunication networks for precipitation observations. Frank Annor will share how the Trans-African Hydro-Meteorological Observatory (TAHMO) pursues its aim to build a station every 30km in sub-Saharan Africa, requiring 20,000 stations to be placed. They developed a rain gauge that can be produced for 100 euros, about 10% of the costs of other rain gauges. Aart Overeem, researcher at the Royal Netherlands Meteorological Institute (KNMI), derives precipitation amounts using commercial microwave links from cellular telecommunication networks. He works on upscaling this rainfall measurement technique in low- and middle-income countries. He conducted pilot studies in Nigeria and Sri Lanka. Bashiru Yahaya, from Ghana Meteorological Agency (GMet) knows from first-hand experience the importance of rainfall information in Ghana. He will elaborate on their efforts to improve precipitation observations and forecasts for Ghana. He is working on nowcasting models for early warning systems. Dorien Lugt will present how she applies machine learning to generate nowcasts from satellite observations. HKV discloses precipitation observations derived from the geostationary satellite MSG in a web- and mobile phone application. These are available every 15 minutes for the entire African continent, with smaller latency and higher spatial resolution than other free available products. She uses machine learning models to forecast precipitation 3 hours ahead. The panel members will share their stories and ideas in a short pitch. Afterwards, they will discuss with the audience the future of precipitation monitoring, nowcasting and forecasting in Africa. The session will be facilitated by Carolien Wegman from HKV. |
Organizer: HKV, GMet, KNMI, NADMO, TAHMO