Climate information services and farm-level decision-making in the Pra River Basin of Ghana

Abstract Agriculture is the backbone of many economies and provides livelihood to many nations in Africa. However, the sector is largely rainfed and faces considerable challenges due to the impacts of climate variability and unpredictable rainfall patterns. Despite advances in climate information services, farmers have limited access to reliable climate information to make adaptive decisions. Studies have shown that farmers with limited access hardly rely on the information they get access to. Using the Pra River basin as a case, this study assessed climate information’s influence on farmers' farm-level decisions by interviewing 382 farmers and evaluated the performance of seasonal forecasts from the European Centre for Medium-Range Weather Forecast (ECMWF) in meeting farmers’ needs. Rainfall onset, amount and cessation were the three most prioritised needs ranked in descending order. Most farmers preferred to receive climate information at a 2-month lead time. However, ECMWF-S5 only provided the second-ranked need (rainfall amount) at a 1-month lead time with the highest skill in forecasting rainfall in the study area at correlation coefficient, RMSE (ranging between 40.31–52.84, 68.85–75.82, 92.17-101.07), and KGE within the ranges of 0.7–0.9, 40–53, and 0.6–0.8, respectively. The skill of ECMWF-S5 decreases with increasing lead times. ECMWF-S5 rainfall forecast information at a 1-month lead time is advisable and likely to positively impact the decision-making process for farming in the Pra River Basin. It is recommended that other forecast models, including indigenous techniques be combined to improve the forecast's lead time and determine the onset and cessation for adaptive farming.