Engineering, Science, Technology, and Entrepreneurship (ESTE 2024) Internal Research Symposium, 22–23 July 2024 at IDL Conference Centre in KNUST, Kumasi, Ghana
The difficulties in accessing data in formats compatible with scientific processes have hindered the need to include indigenous knowledge in addressing global challenges, such as climate change. This study addresses this challenge by collecting indigenous weather and seasonal climate forecasts via citizen science and verifying their accuracy or skill in the Pra River Basin of Ghana. A mental model was constructed to link identified Indigenous ecological indicators (IEIs) to types of rainfall (heavy, medium or small) and the level of certainty in forecasts. 25 rain gauges were calibrated and installed in the residential compound of 25 farmers across the river basin. Farmers were trained to use the Sapelli mobile app to send their forecasts and monitor and record observed rainfall. Results show that indigenous forecast (IF) has an established link with IEIs, however, the forecast skill for each IEIs varied geographically. The skill of IF daily weather in 12- and 24-hours was 27–42 % and 31–38%, respectively, thus about one hit in every three rainfall predictions. Also, two out of three correct rejections were recorded during the rainfall season. The most used IEI in the study area was cloud (44 – 54 %), followed by the sun (15 – 20 %). The moon (12 %), heat (11 %) and dew (13 %) were the third most used IEIs in the Central, Eastern and Ashanti regions, respectively. Moon and heat predominated followed clouds in the forecast of heavy, medium and small rainfall. The major IEI for seasonal climate forecast was a tree (different species for different geographical locations), and accuracy depended more on experience. We conclude that it is possible to systematically collect and empirically evaluate indigenous weather and seasonal climate forecasts to enhance climate information services for climate change adaptation.