How AI flood forecasting can help communities in need

  • Published
  • Posted in Google
  • 2 mins read
  • Tagged as

Floods are the most common and deadliest of natural disasters putting communities at risk. For the past few years Google Research has invested in developing and applying AI modeling to advance the field of flood forecasting, as part of our efforts to help communities around the world address the effects of climate change with AI-based early warnings for natural disasters.

The Google flood forecasts are provided to the public in multiple ways, in partnership with organizations, governments and with the WMO. Global flood information is publicly available on the Flood Hub (g.co/floodhub) for the use of the public, NGOs, governments and researchers. As part of our efforts, we actively collaborate with local humanitarian organizations that utilize our flood information to accelerate their ongoing anticipatory actions supporting local communities. Here are four examples from the past year:

Google and Give Directly in Mozambique — supporting anticipatory action

In early 2023, we partnered with GiveDirectly, an aid organization supporting communities in need via unconditional cash transfers, in the poorest areas of the world. Together, we tested how riverine flood forecasts and digital cash infrastructure could enable communities to receive payments within a day of a flood warning, to help them better prepare before a potential disaster. We piloted our collaboration in Mozambique during Cyclone Freddy in March 2023, where GiveDirectly provided cash transfers to over 4,100 households with the support from Google.org, based on the Google flood forecasts. This pilot provided us with meaningful insights regarding the level of geospatial detail and lead time aid organizations need to distribute cash to the communities expected to be affected. As Vera Lummis, Senior Manager of Digital Innovation at GiveDirectly, says: “Through our partnership with Google Research, we can combine new insights from AI-based forecasts with local early warning systems to better reach people at the frontlines of the climate crisis.” This collaboration is planned to continue in 2024 with enhanced tools and larger coverage in Bangladesh and Nigeria.

News Article Courtesy Of »