Google Flood Hub — Bangladesh Water Development Board's AI-Powered River Flood Alert System
Bangladesh
Since 2020, Google Research's machine-learning flood model — which forecasts river levels even in ungauged basins — has combined with …
Mozambique · Maputo · See the Mozambique profile · See the Maputo profile
Evidence: Descriptive / self-reported Top 88% 33/100 · Ask Evidence Copilot about this practice
On 15 April 2026 LIMCOM — Botswana, Mozambique, South Africa and Zimbabwe's joint commission — launched a hydrological 'Digital Twin' and 'WaterCopilot', an AI assistant for querying basin data in plain language. No usage or outcome data published yet.
LIMCOM, the treaty-based commission through which Botswana, Mozambique, South Africa and Zimbabwe jointly govern the Limpopo river basin, launched the "Limpopo Digital Twin" and "WaterCopilot" on 15 April 2026, built with the International Water Management Institute (IWMI) and Microsoft.
The tools aim to strengthen regional water security and climate-adaptive planning by giving basin water managers a near-real-time hydrological, climate and irrigation data replica and a conversational AI assistant for querying basin conditions in plain-language English or Portuguese.
The launch was paired with a two-day technical training (15–16 April 2026) in Maputo, Mozambique, that onboarded technical staff from all four LIMCOM member states.
This is a real, named, dated deployment with credible technical partners and a formal intergovernmental governance structure (LIMCOM established by treaty in 2003), but no usage statistics, query volumes, decision outcomes, or before/after comparisons have been published since launch, making it an early-stage, evidence-thin case worth tracking as it moves into routine use.
National / regional programmes Philanthropic / foundation funding
Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.
Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.
Where this practice's information was retrieved from, and when.
Bangladesh
Since 2020, Google Research's machine-learning flood model — which forecasts river levels even in ungauged basins — has combined with …
Philippines
A $12.7m Korean-funded project has installed an AI-assisted flood forecasting system and warning centre for the Pasig-Marikina-Tullahan river basins, intended …
France
France's open-data portal uses Albert, the state's open LLM, to auto-draft dataset descriptions and suggest metadata tags, and exposes its …
Nigeria
Nigeria's Hydrological Services Agency says it deployed deep-learning models trained on 200+ years of flood data for its 2026 Annual …
Open full copilot Grounded in cited practices — always check the sources.