Researchers in southern Africa are exploring how artificial-intelligence weather models could dramatically improve early-warning systems for storms and floods across the continent. A new forecasting architecture based on advanced machine-learning models is being tested to generate long-range atmospheric forecasts at far lower cost than traditional radar networks. Early warning systems remain limited across many African countries because weather radar installations are expensive and require extensive maintenance. AI-driven forecasting models offer a potential alternative by using global atmospheric data and machine-learning algorithms to predict weather patterns. Scientists involved in the research say the technology could generate 15-day forecasts and distribute alerts through mobile platforms such as WhatsApp, which has extremely high adoption rates across Africa. If implemented at scale, the system could significantly improve disaster preparedness across southern Africa, where severe flooding and storms regularly threaten vulnerable communities.
Read more here
