Directing ML towards pure hazard mitigation by collaboration – Google AI Weblog


Floods are the commonest sort of pure catastrophe, affecting greater than 250 million folks globally annually. As a part of Google’s Disaster Response and our efforts to handle the local weather disaster, we’re utilizing machine studying (ML) fashions for Flood Forecasting to alert folks in areas which are impacted earlier than catastrophe strikes.

Collaboration between researchers within the business and academia is important for accelerating progress in direction of mutual targets in ML-related analysis. Certainly, Google’s present ML-based flood forecasting strategy was developed in collaboration with researchers (1, 2) on the Johannes Kepler College in Vienna, Austria, the College of Alabama, and the Hebrew College of Jerusalem, amongst others.

In the present day we talk about our current Machine Studying Meets Flood Forecasting Workshop, which highlights efforts to deliver collectively researchers from Google and different universities and organizations to advance our understanding of flood conduct and prediction, and construct extra sturdy options for early detection and warning. We additionally talk about the Caravan undertaking, which helps to create an open-source repository for world streamflow knowledge, and is itself an instance of a collaboration that developed from the earlier Flood Forecasting Meets Machine Studying Workshop.

2023 Machine Studying Meets Flood Forecasting Workshop

The fourth annual Google Machine Studying Meets Flood Forecasting Workshop was held in January. This 2-day digital workshop hosted over 100 members from 32 universities, 20 governmental and non-governmental businesses, and 11 personal corporations. This discussion board offered a possibility for hydrologists, pc scientists, and help employees to debate challenges and efforts towards enhancing world flood forecasts, to maintain up with state-of-the-art know-how advances, and to combine area data into ML-based forecasting approaches.

The occasion included talks from six invited audio system, a sequence of small-group dialogue periods targeted on hydrological modeling, inundation mapping, and hazard alerting–associated subjects, in addition to a presentation by Google on the FloodHub, which supplies free, public entry to Google’s flood forecasts, as much as 7 days prematurely.

Invited audio system on the workshop included:

The shows could be seen on YouTube:

2023 Flood Forecasting Meets Machine Studying Talks Day 1

2023 Flood Forecasting Meets Machine Studying Talks Day 2

A few of the high challenges highlighted in the course of the workshop had been associated to the combination of bodily and hydrological science with ML to assist construct belief and reliability; filling gaps in observations of inundated areas with fashions and satellite tv for pc knowledge; measuring the talent and reliability of flood warning programs; and enhancing the communication of flood warnings to various, world populations. As well as, members careworn that addressing these and different challenges would require collaboration between plenty of totally different organizations and scientific disciplines.

The Caravan undertaking

One of many important challenges in conducting profitable ML analysis and creating superior instruments for flood forecasting is the necessity for giant quantities of knowledge for computationally costly coaching and analysis. In the present day, many nations and organizations acquire streamflow knowledge (usually both water ranges or stream charges), however it isn’t standardized or held in a central repository, which makes it troublesome for researchers to entry.

In the course of the 2019 Machine Studying Meets Flood Forecasting Workshop, a gaggle of researchers recognized the necessity for an open supply, world streamflow knowledge repository, and developed concepts round leveraging free computational sources from Google Earth Engine to deal with the flood forecasting neighborhood’s problem of knowledge assortment and accessibility. Following two years of collaborative work between researchers from Google, the college of Geography on the College of Exeter, the Institute for Machine Studying at Johannes Kepler College, and the Institute for Atmospheric and Local weather Science at ETH Zurich, the Caravan undertaking was created.

In “Caravan – A worldwide neighborhood dataset for large-sample hydrology”, revealed in Nature Scientific Information, we describe the undertaking in additional element. Based mostly on a worldwide dataset for the event and coaching of hydrological fashions (see determine beneath), Caravan supplies open-source Python scripts that leverage important climate and geographical knowledge that was beforehand made public on Google Earth Engine to match streamflow knowledge that customers add to the repository. This repository initially contained knowledge from greater than 13,000 watersheds in Central Europe, Brazil, Chile, Australia, the US, Canada, and Mexico. It has additional benefited from neighborhood contributions from the Geological Survey of Denmark and Greenland that features streamflow knowledge from many of the watersheds in Denmark. The aim is to proceed to develop and develop this repository to allow researchers to entry many of the world’s streamflow knowledge. For extra info relating to contributing to the Caravan dataset, attain out to

Places of the 13,000 streamflow gauges within the Caravan dataset and the distribution of these gauges in GEnS world local weather zones.

The trail ahead

Google plans to proceed to host these workshops to assist broaden and deepen collaboration between business and academia within the growth of environmental AI fashions. We’re trying ahead to seeing what advances may come out of the newest workshop. Hydrologists and researchers fascinated by collaborating in future workshops are inspired to contact


Please enter your comment!
Please enter your name here