Decision support system for farmers and fishermen in Malawi

BRLI's mission was to work with WIMES to develop a weather and climate warning system that would enable farmers and fishermen to take decisions adapted to the situation (outings on the lake, cultivation, etc.). The ultimate aim is to save lives and preserve the livelihoods of farmers and fishermen. These livelihoods are increasingly affected by extreme weather events, and Malawi's farmers and fishermen are trapped in poverty by alternating severe floods and droughts, as well as annual losses due to intense storms. These weather phenomena are sustained by rising global temperatures, and there is evidence that climate change is leading to more frequent and extreme storms, droughts and floods in Malawi.

Services provided by BRLI :

• Identification of stakeholders.
• Literature review on the information needs of fishermen and farmers, and assessment of global and local data sources that can be used for product generation.
• Design of the alert production system and identification of the institutions responsible for each stage.
• Development of software routines and interfaces to automate data processing and alert production in formats for different media.
• Installation of code and software in each institution and development of Standard Operating Procedures (SOPs) to clearly identify the responsibilities of each institution and the timetable/scheduling of tasks.
• Design of a Monitoring Evaluation and Learning (ME&L) system to gather information on the use of information products, collect feedback from users and use this information to improve alerts.
• Training users on software/code development, how it can be improved in the future, and SOPs for system operation and continuous improvement.

• Technical context: Web application based on the Wimes BRLi solution - final platform deployed and hosted by DCCMS. Data source: GPM, GFS, COSMO climate prediction model operated by DCCMS, Lake Malawi buoy tracking system, lightning strike point. observation system operated by DCCMS. Databases : PostgreSQL/PostGIS/TimescaleDB, MongoDB. Alert interoperability standard : Common Alerting Protocol (CAP). UI development: React/Javascript/Typescript, responsive. Web mapping: MapLibre/WebGL. Backend development: Java/Scala. Infrastructure: Kong Gateway, microservices, RabbitMQ, Docker.