11 December 2023

AI for Climate Change: AI for flood adaptation plans and disaster relief

By Giulia Cirri

If, like me, you feel like the amount we hear about devastating floods has drastically increased in the last few years, know that it is not just a feeling. While I can’t say whether the media is more likely to pick up and report on floods, according to new research from NASA the proportion of people across the world living in flood-prone areas has risen by 20% to 24% since 2000. This is 10 times greater than the number previous models had predicted. 

Floods can be devastating. I am Italian, so I felt very emotional this past May when I followed the news about what has been defined as Italy’s worst flooding in 100 years around the areas where my dad grew up. 14 people died, and the affected areas are still struggling to recover. While the floods in Italy have quite literally hit close to home for me, they are just a small example of what a flood can do. In Pakistan, the floods of 2022 have caused more than 1,700 deaths. Then there’s the aftermath: floods leave a trail of destruction that takes a long time to overcome. Debris and hardened mud need to be cleaned up; houses and infrastructures need to be rebuilt; rivers banks need to be secured; crops might be irreparably damaged. The 2022 flood in Bangladesh is estimated to have cost $1.0 billion and affected 7.3 million people by damaging their homes, infrastructure, croplands, and sanitation facilities.

As for many climate-related events, scientifically identifying a direct causal link between climate change and floods can be challenging, because of the many variables that play a role in flooding events. In its 2012 report on extreme events and disasters, the Intergovernmental Panel on Climate Change (IPCC) noted how, while climate change may not induce floods directly, it exacerbates many of the factors that do, increasing the likelihood and the intensity of floods. For instance, in September 2023 massive flooding in Libya led to two dams bursting and a catastrophic death toll of more than 11,300 people. A study from the World Weather Attribution reported that the amount of rain that fell in Libya was up to 50% higher than it would have in a world where people had not changed the climate, although the study has a high level of uncertainty.


Floods and climate change

There are several reasons that might cause floods: heavy rains, ocean waves coming on shore, or snow melting quickly. As we have mentioned, climate change has impacted many factors connected to floods, increasing the likelihood and the intensity of floods. Moreover, floods are also worsened by our land management practices, which might affect the risk of floods by eliminating many natural features that would have otherwise slowed and absorbed rainwater. Paved roads, for instance, block water from being absorbed, while deforestation increases rainwater runoff and the risk of mudslides

Floods may also be affected by other climate-related events. For instance, flash floods might increasingly follow wildfires in what the New York Times has defined as “a deadly cascade of climate disasters,” as wildfires destroy vegetation, making the soil less permeable. 

While floods can happen within minutes or over a long period of time, flash floods are the ones we are likely to hear about the most, as they are responsible for the greatest number of flood-related fatalities. Flash floods rise quickly, combining the hazard of floods with speed and unpredictability, and are most often caused by heavy rains over a short period (usually six hours or less). Scientists expect that, as the climate warms, flash floods will get “flashier,” meaning that the timing of the floods will get shorter while the magnitude increases. 

With floods likely to continuously increase in likelihood and impact, improving our ability to mitigate and respond to floods is both vital and urgent. AI can lend a (data-based) hand. 


How can AI help?

AI is a good instrument to analyse large amounts of data quickly and relatively cheaply. This ability can be useful in the longer term, to inform adaptation plans to mitigate the impacts of potential floods, but also in the immediate aftermath, for disaster management. 

Before floods: developing adaptation plans

AI can be used to analyse different data sources — such as climate models, satellite imagery and weather patterns — to identify areas with a higher vulnerability to floods. These insights can be used by decision makers and communities to create tailored adaptation plans to improve the resilience of their local area and provide resources to high-vulnerability areas. 

For instance, in the case of coastal communities, AI can be used in flood modelling to help predict the potential impact of sea-level rise or extreme weather events on infrastructures and homes. AI can analyse data on topography, land use, and urban development to identify areas that are more vulnerable to flooding. Decision makers can use this information to plan and implement proactive measures such as building sea walls, reinforcing or relocating infrastructure, or implementing zoning regulations to avoid constructions on the high vulnerability areas, therefore reducing risk.

The use of AI is of course not limited to coastal communities. For example, a 2021 study used AI to assess flood vulnerability assessment in Dire Dawa, Ethiopia. Researchers used several flood-causing factors, including rainfall, slope, land use, elevation, vegetation density, river distance, geomorphology, road distance, and population density to train a model to generate a hazard map of the area. In short, the results were compared with data from historic floods in 2006, and the errors were fed back into the model until the result was appropriately close to the actual historic data. After several rounds of iteration, the model was able to provide an assessment of flood vulnerable zones for this city and its catchment, providing useful information for flood management.

After floods: post damage relief and recovery

AI can also play a role in disaster response. In a comprehensive review of uses of Machine Learning for Climate Change, Kris Sankaran highlights two types of machine learning tasks that have proven useful to support post disaster relief: creating maps from aerial imagery, and retrieving information from social media data.

First, machine learning can create accurate maps from aerial imagery to inform evacuation planning and the delivery of relief. This imagery can also be used to compare scenes pre and post disaster, which can inform damage assessment, and therefore decision making on the allocation of relief resources and reconstruction efforts.  

A practical example of an AI tool for disaster management is the deep-learning model DAHiTrA, used to to classify building damages based on satellite images after natural disasters. The model recognises the geographic features in different locations and compares images of a building, road, or bridge taken before and after a disaster to determine the level of damage. The damage assessment of homes and buildings after a natural disaster can usually take months. Since satellite images are available within 24 hours, and the model is fast, DAHiTrA allows authorities to determine the number of buildings and infrastructures impacted, as well as the extent of the damage the day after the event took place. Faster and more accurate damage assessments can help communities and governments allocate resources more effectively. The DAHiTrA team is working with government agencies, emergency management offices, and international organisations to use their model to assess building damage. 

Machine learning can also be used to analyse information posted on social media. While this might sound slightly dystopic, social media data after a natural event can contain “kernels of insight.”  Information posted by users about places without water, or clinics without supplies which can inform relief efforts. Machine learning can analyse large volumes of these social media data and summarise it into key insights, which can inform action and plans of disaster managers. 

The potential of AI in natural disaster relief has caught the attention of major international organisations. In 2021, the World Meteorological Organization (WMO), the International Telecommunication Union (ITU) and the United Nations Environment Programme (UNEP) formed the expert Focus Group on AI for Natural Disaster Management, to help lay the groundwork for the use of AI for natural disaster management. 

In this blog, we covered applications of AI specifically relating to floods. Even so, the same capabilities can be used for other natural events. The same applications of AI for disaster relief can also apply to tsunamis, or earthquakes. Similarly, AI can inform climate adaptation plans tailored to other climate events. Of course, the variables necessary to measure an area’s vulnerability to heatwaves, for instance, might be different from those to floods, and each face their individual challenges. What doesn’t change is AI’s ability to process large volumes of raw data, which, with the right training datasets, can inform decisions on climate adaptation plans and disaster relief. 

We have seen the use of AI before and after floods. But, as I was writing this blog, I kept having one question in mind. What about during floods? Can AI be used to predict floods, even, and in particular, flash floods? Can AI help inform local communities and notify them of the need to evacuate before it’s too late? The next blog of our series will focus on these questions, as we look at AI during floods: AI for early warning systems.


More insights

21 April 2017

Why Government is ready for AI

12 July 2017

Five levels of AI in public service

26 July 2017

Making it personal: civil service and morality

10 August 2017

AI: Is a robot assistant going to steal your job?

19 September 2017

AI and legitimacy: government in the age of the machine

06 October 2017

More Than The Trees Are Worth? Intangibles, Decision-Making, and the Meares Island Logging Conflict

16 October 2017

The UK Government’s AI review: what’s missing?

23 October 2017

Why unconference? #Reimagine2017

03 November 2017

AI: the ultimate intern

09 November 2017

Motherboard knows best?

23 November 2017

Beyond driverless cars: our take on the UK’s Autumn Budget 2017

05 December 2017

Why Black people don’t start businesses (and how more inclusive innovation could make a difference)

06 December 2017

“The things that make me interesting cannot be digitised”: leadership lessons from the Drucker Forum

23 January 2018

Want to get serious about artificial intelligence? You’ll need an AI strategy

15 February 2018

Economic disruption and runaway AI: what can governments do?

26 April 2018

Ranking governments on AI – it’s time to act

08 May 2018

AI in the UK: are we ‘ready, willing and able’?

24 May 2018

Mexico leads Latin America as one of the first ten countries in the world to launch an artificial intelligence strategy

05 July 2018

Beyond borders: talking at TEDxLondon

13 July 2018

Is the UK ready, willing and able for AI? The Government responds to the Lords’ report

17 July 2018

Suspending or shaping the AI policy frontier: has Germany become part of the AI strategy fallacy?

27 July 2018

From open data to artificial intelligence: the next frontier in anti-corruption

01 August 2018

Why every city needs to take action on AI

09 August 2018

When good intentions go bad: the role of technology in terrorist content online

26 September 2018

Actions speak louder than words: the role of technology in combating terrorist content online

08 February 2019

More than STEM: how teaching human specialties will help prepare kids for AI

02 May 2019

Should we be scared of artificial intelligence?

04 June 2019

Ethics and AI: a crash course

25 July 2019

Dear Boris

01 August 2019

AI: more than human?

06 August 2019

Towards Synthetic Reality: When DeepFakes meet AR/VR

19 September 2019

Predictive Analytics, Public Services and Poverty

10 January 2020

To tackle regional inequality, AI strategies need to go local

20 April 2020

Workshops in an age of COVID and lockdown

10 September 2020

Will automation accelerate what coronavirus started?

10 September 2020

Promoting gender equality and social inclusion through public procurement

21 September 2020

The Social Dilemma: A failed attempt to land a punch on Big Tech

20 October 2020

Data and Power: AI and Development in the Global South

23 December 2020

The ‘Creepiness Test’: When should we worry that AI is making decisions for us?

13 June 2022

Data promises to support climate action. Is it a double-edged sword?

30 September 2022

Towards a human-centred vision for public services: Human-Centred Public Services Index

06 October 2022

Why You Should Know and Care About Algorithmic Transparency

26 October 2022

Harnessing data for the public good: What can governments do?

09 December 2022

Behind the scenes of the Government AI Readiness Index

06 February 2023

Reflections on the Intel® AI for Youth Program

01 May 2023

Canada’s AI Policy: Leading the way in ethics, innovation, and talent

15 May 2023

Day in the life series: Giulia, Consultant

15 May 2023

Day in the life series: Emma, Consultant

17 May 2023

Day in the life series: Kirsty, Head of Programmes

18 May 2023

Day in the life series: Sully, Partnerships Associate/Consultant

19 May 2023

LLMs in Government: Brainstorming Applications

23 May 2023

Bahrain: Becoming a regional R&D Hub

30 May 2023

Driving AI adoption in the public sector: Uruguay’s efforts on capacity-building, trust, and AI ethics

07 June 2023

Jordan’s AI policy journey: Bridging vision and implementation

12 June 2023

Response to the UK’s Global Summit on AI Safety

20 June 2023

 Unlocking the economic potential of AI: Tajikistan’s plans to become more AI-ready

11 July 2023

Government transparency and anti-corruption standards: Reflections from the EITI Global Conference in Dakar, Senegal

31 August 2023

What is quantum technology and why should policymakers care about it?

21 September 2023

Practical tools for designers in government looking to avoid ethical AI nightmares

23 October 2023

Collective Intelligence: exploring ‘wicked problems’ in National Security

23 October 2023

Exploring the concepts of digital twin, digital shadow, and digital model

30 October 2023

How to hire privileged white men

09 November 2023

Inclusive consensus building: Reflections from day 4 of AI Fringe

13 November 2023

AI for Climate Change: Can AI help us improve our home’s energy efficiency?

14 November 2023

Navigating the AI summit boom: Initial reflections

20 November 2023

AI for Climate Change: Improving home energy efficiency by retrofitting

24 November 2023

Will AI kill us all?

27 November 2023

AI for Climate Change: Preventing and predicting wildfires 

28 November 2023

Service Design in Government 2023: conference reflections

04 December 2023

AI for Climate Change: Using artificial and indigenous Intelligence to fight climate change

06 December 2023

Release: 2023 Government AI Readiness Index reveals which governments are most prepared to use AI

18 December 2023

AI for Climate Change: Managing floods using AI Early Warning Systems

28 May 2024

Strengthening AI Governance in Chile: GobLab UAI and the Importance of Collaboration

05 June 2024

Towards a Digital Society: Trinidad and Tobago’s AI Ambitions