Logo Lanfrica

NovaHB/floodwatch-africa

Domaine:

climate

Type de record:

dataset
Créateur:
Nov
Hôte:
FloodWatch is a dbt and BigQuery pipeline modeling verified flood events across Africa, 1900 to 2026. # FloodWatch Africa FloodWatch is a dbt and BigQuery pipeline modeling verified flood events across Africa, 1900 to 2026. 1,314 verified flood events across Africa, 1900 to 2026, sourced from EM-DAT | CRED. Each event carries country, region, dates, deaths, people affected, homeless counts, and economic damage in USD, along with flood subtype (flash flood, riverine, coastal) and river basin where recorded. ## The pipeline Architecture ``` EM-DAT export -> BigQuery (raw) -> dbt (staging, intermediate, marts) -> React frontend -> Netlify ``` ## limitation This is a historical dashboard. It tells you what flooded, not what's flooding right now. I built the whole pipeline before I fully answered the question of who actually needs this. A researcher looking at 20 years of flood patterns gets real value here. Someone trying to find out if their town is flooding today does not. I'm leaving this project as is for now, and being upfront about that gap rather than dressing it up as something it isn't. ## Tech stack EM-DAT (source data), Google BigQuery, dbt Core, React, Recharts, Netlify.