Blitz Bureau
NEW DELHI: In a blog post last week, Google Research announced that it is expanding its global flood forecasting coverage with the roll-out of flash flood predictions in urban areas.
Using a novel AI training method based on news data, it will provide up to 24 hours advance notice for these rapid-onset events. This expansion, said the post, is a critical step to enhancing global climate resilience and keeping communities safe.
According to the World Meteorological Organisation (WMO), flash floods account for approximately 85 per cent of flood related fatalities worldwide. They typically occur within six hours of heavy rain, turn city streets into gushing rivers, and take more than 5,000 lives annually, making them one of the world’s deadliest disasters.
Early warning systems (EWS) are essential for keeping communities safe and informed. They have been proven to save lives and mitigate damage: even a 12-hour lead time can provide a 60 per cent reduction in flash flood damage.
However, a stark “warning gap” exists between countries. While the most developed nations benefit from robust forecasting, life-saving infrastructure is largely absent across vast regions of the Global South, where less than half of developing countries have access to multi-hazard EWS. This leaves billions of people without the advance notice that makes a critical difference.
To address this, Google Research last week announced the roll-out of Urban Flash Flood forecasts on Flood Hub. So far, its Flood Forecasting Initiative had focused on riverine floods, where rivers overflow their banks over a relatively slow period.
While these forecasts cover over 2 billion people in 150 countries for the most significant riverine floods, urban flash floods present a unique challenge. Unlike riverine floods, flash floods are characterised by their rapid onset, requiring a fundamentally different forecasting approach.
The “invisible” flood
One challenge in forecasting flash floods is a lack of “ground truth” data. Riverine machine learning models are trained on physical stream gauges that measure water levels or streamflow. By training models on historical river gauge measurements, Google can now accurately predict localised water rises and anticipate when a river is likely to exceed its flood banks. The tech giant has also successfully extended these predictions to ungauged locations to provide more global coverage of riverine floods.
Flash floods, however, can happen anywhere and often far from any stream gauge. In urban environments, the complex interaction between intense rainfall, impermeable surfaces, and drainage systems makes traditional physical modelling computationally prohibitive at a global scale.
Google used Gemini to analyse publicly available news reports that mention floods to confirm flood event details (e.g., clear locations and times). These entries were then aggregated to create a dataset of historical flooding events, which were used to train and evaluate the new flash flood model in urban areas.
Also, without a historical record of exactly where and when flash floods have occurred in the past, traditional supervised machine learning (ML) models cannot learn the patterns necessary to predict them.
To address the lack of historical data, Google has used Groundsource, a new AI-powered methodology to extract ground truth from unstructured data with high precision. This has enabled it to create the Groundsource dataset of past flash flood events.
Google used Gemini to analyse publicly available news reports that mention floods to confirm flood event details (e.g., clear locations and times). These entries were then aggregated to create a dataset of historical flooding events, which were used to train and evaluate the new flash flood model in urban areas.
The scaling challenge
Specialised, hyper-local early warning systems have been engineered to address flash floods from rainfall in specific urban settings. These systems typically rely on a network of physical sensors monitoring variables like direct and radar-inferred precipitation, water levels and flow velocities.
While highly accurate for their specific locations, they are difficult to scale because of the high costs of hardware deployment, the need for site-specific calibration algorithms and engineering expertise.
At a broader level, initiatives such as the WMO’s Flash Flood Guidance System (FFGS), the European Runoff Index based on Climatology (ERIC) flash flood indicator, and the US National Weather Service (NWS) Flash Flood Warnings system provide wider coverage through remote sensing and numerical weather models.
These systems, however, encounter significant hurdles regarding global implementation. A primary issue is their dependency on high-resolution hydrological maps and radar-based weather forecasts, resources that are largely unavailable within the Global South. The reliance on professional hydrologists to interpret complex model data and distribute actionable warnings presents a second major challenge.
To help save thousands of lives by predicting imminent danger 24 hours in advance
To achieve near-global reach, the Google model uses only global weather products as well as real-time global weather forecasts and the AI-based medium-range global weather forecasting model by Google DeepMind. The system operates at a 20×20 kilometer spatial resolution, a constraint primarily driven by the resolution of globally available data sources.
Focusing on the city
Trained on Groundsource, the new flash flood model is designed to answer a specific question: Given the forecasted weather and local conditions, is a flash flood likely to occur in this area in the next 24 hours?
The model utilises a recurrent neural network (RNN) architecture constructed with a long short-term memory (LSTM) unit that is specifically suited for processing time-series data. In addition to the meteorological time-series inputs, it also incorporates static geographic, geophysical, and anthropogenic attributes, such as urbanisation density, topography, and soil absorption rates.
Google is focusing its initial launch on urban areas, because the training data — news reports — is more dense in these locations. At present, the model predicts impact in areas with population densities greater than 100 people per square kilometer.