{
  "id": 1776145,
  "title": "Ask a Scientist: How can researchers use AI to predict a flood?",
  "url": "https://urgent.news/2026/08/18/ask-a-scientist-how-can-researchers-use-ai-to-predict-a-flood",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-18T16:00:00.000Z",
  "source": {
    "name": "Google Blog",
    "slug": "google-blog",
    "url": "https://blog.google/innovation-and-ai/technology/research/flood-prediction-ai/"
  },
  "original_language": "en",
  "account": "Google employs artificial intelligence to forecast river and urban flash floods, alerting over 2 billion people in 150 countries. Utilizing Flood Hub and an open-source API, researchers can access these predictions to bolster disaster resilience. Initially piloted in India in 2018, the AI-driven methodology has expanded globally, now predicting floods in 150 countries affecting more than 2 billion individuals. In March 2026, they unveiled Groundsource, a new AI-powered technique that transforms disaster data into a high-quality archive focused on flash floods in urban regions.\n\nDeborah Cohen, a senior staff research scientist at Google Research, elucidates how AI facilitates flood forecasting and its potential for predicting other natural disasters. Their Flood Forecasting team aims to protect lives by predicting floods early, leveraging AI to analyze vast global datasets encompassing rainfall, river levels, and ground conditions. This enables predictions of riverine floods up to seven days in advance and urban flash floods within 24 hours.\n\nThe Flood Hub tool aggregates global weather data, employing AI to issue prediction alerts on maps, extending traditional models that require local historical data for calibration. While conventional models necessitate local data, AI technology allows for the incorporation of international data, crucial for regions with insufficient historical data. The Flood Hub initially focused on river floods, but advancements like Groundsource now extend coverage to urban flash flooding.\n\nInitially, integrating urban flash flooding was challenging due to limited global data on such events. Groundsource overcame this by gathering insights from 5 million news reports on floods over two decades, creating a dataset of 2.6 million historical flood events across 150 countries. This data fuels a new urban flash flood model now integrated within Flood Hub. These predictions primarily aid researchers and aid organizations in anticipating flood-affected areas and at-risk populations. By providing early cash assistance to flood-prone communities, initiatives like Give Directly have significantly improved outcomes in Kogi, Nigeria, with measurable improvements in income, food security, and preparedness. Google has open-sourced its hydrology framework and made the Groundsource dataset and Flood Forecasting API publicly available, fostering further research and enabling NMHSs worldwide to tailor forecasts to their specific needs.",
  "summary": "A Google researcher explains how AI technology like Flood Hub and Groundsource can predict floods and help people around the world.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}