Python Climate Predictability Tool (PyCPT) training for improved seasonal climate prediction over Ethiopia

Seid J, Teshome A, Demissie T. 2021. Python Climate Predictability Tool (PyCPT) training for improved seasonal climate prediction over Ethiopia. CCAFS Workshop Report. Addis, Ababa: CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS).

Training on weather forecasting tools and techniques is a fundamental requirement for meteorological services to improve the accuracy and reliability of weather and climate forecasts. These tools greatly support the generation and packaging of forecasts that are destined for private and public consumption. Ethiopia's National Meteorological Agency (NMA), under the support of the International Research Institute for Climate and Society (IRI), through the project Adapting Agriculture to Climate Today, for Tomorrow (ACToday), is working together with the CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS) - East Africa (EA) to address the needs and demands of different stakeholders including governmental, non-governmental organizations and other non-state actors by conducting staff training to improve the generation of reliable, timely and accurate weather and seasonal forecasts. With the support of the IRI and CCAFS - EA, training on the Next Generation (NextGen) seasonal forecasting was given from January 11-15, 2021, to 26 participants from the National Metrological Agency of Ethiopia (NMA). Participants were selected from NMA's Regional Meteorological Service Centers (RMSC's) and NMA head office. The Next Generation (NextGen) multi-model approach is a general systematic approach for designing, implementing, producing, and verifying objective climate forecasts. It involves identifying decision-relevant variables by stakeholders and analyzing the physical mechanisms, sources of predictability, and suitable candidate predictors (in models and observations) for key relevant variables. When prediction skill is high enough, NextGen helps select the best dynamic models for the region of interest through a process-based evaluation and automizes the generation and verification of tailored multi-model, statistically calibrated predictions at seasonal and sub-seasonal timescales.