How a data scientist can help to deal with the climate change

Feeling under the weather?

If you follow road cycling races, you might have heard about disastrous weather accompanying the recent Vuelta a España competition. Just before the second stage some riders, including a veteran rider Geraint Thomas, called for neutralizing parts of the road in Barcelona because it was still wet after recent flooding. The race’s organizers eventually did so, however it didn’t prevent crashes on the slippery roads. The heavy rain put its mark on some later stages as well, causing chaos in the overall competition. And for a good reason: on September 3rd a rainfall of 90 litres per square meter was registered near Toledo, breaking the previous record of 87 litres from 19721, 2. At the same time media shared photos of water covered streets and damaged households and cars.

It is very hard for climate scientists to attribute a specific event to the climate change. Sceptics might point out that floods, as well as other bad weather events, happened occasionally in the past as well. But there is no doubt that those events are occurring with an increasing frequency and intensity. A study published in Nature found that the percentage of the global population at risk from flooding has risen by almost a quarter since 20003. It is posing a real danger to property markets as well as people’s wellbeing, financial stability, and life in general.

As those dangers has become more apparent, the need for action across the whole of corporate, government & civil society is clear and urgent. But in this blog post let’s focus on one question: what can a data scientist specifically do?

Facing the uncertainty

The extreme weather events are already skewing the financial markets. It was reported recently that two American insurers, State Farm and Allstate Corporation stopped selling new home insurance policies in California, due to the soaring wildfire and construction costs4. Undoubtedly, it is worth embedding climate related risks into already existing risks schemas. And that is the first area where there is a flood of work waiting for data scientists.

The risks come in two separate flavours: one is physical, and one is transition. The physical risk is the weather causing the damage itself. For example, in NatWest we try to understand the possibility of flooding and coastal erosion. Not only do we model the current situation but more importantly we try to predict the future. And the future depends largely on how fast the humanity takes some action, hence how fast the global temperature will raise. This is encapsulated by Representative Concentration Pathway (RCP), which describes possible evolution of greenhouse gas (GHG) concentration in the years to come. A data scientist’s report on climate risks usually models the risk in relation to different RCP scenarios. If you would like to see a glimpse of this work, please refer to the page 70 of the NatWest’s 2022 Climate-related Disclosures Report5.

Knowing about the most hazardous RCP scenarios, governments might introduce policies aiming to reduce GHG emissions. This in turn introduces transition risks, which include decline of certain industries, either due to regulations or customers preferring sustainable solutions, rising costs of energy from traditional sources, and others. Data scientists try to understand which sectors are most exposed to transition risks and how it depends on different possible scenarios (early action, late action, and no action).

But transition also means increasing demand for low carbon goods, hence new opportunities are arising as well. One of the interesting examples is offering green lending to retail customers. This will enable customers to reduce their emissions and, in a long term, reduce their energy cost. With the help of data we might, for example, identify the customers who could benefit the most from using certain green products, like solar panels, and explain to them measurable impact of the product on their financial wellbeing.

We’re on the right path but still have far to go

While we need to think about the climate risks to operate in increasingly uncertain world, we should not forget that the overwhelmingly major cause and acceleration of this affair is because of humans, and it is our own responsibility to mitigate the climate change. Apart from modelling the impact of climate on our own business we would like to understand the impact that NatWest has on the changing of the climate. If we want to attempt to reduce our GHG emissions, firstly we must measure and benchmark their levels in our environment today.

You might think that the bank is in a safe position and there is not much work to be done, as office work produces marginal emissions, when compared to other industries. But one couldn’t be more wrong, because through our lending and investments we finance a lot of other’s emissions. However, calculating those is not straightforward at all, because a lot of parties don’t disclose their emissions or do it differently than others. Estimating financed emissions requires ingesting data from different sources and modelling emissions specific for various industry sectors. The vital part of this is data science.

What will you do?

There is more and more data processed in almost every field and the climate change is not an exception. There is a role for a data scientist in predicting future risks as the global temperature increases, societies try to adapt, and new opportunities arise. Contributing to reduction of greenhouse gases requires measuring the amount of financed emissions, which in turn requires a lot of data, estimation, and extrapolation.

This article is written from the perspective of a person working in a financial institution but undoubtedly it doesn’t exhaust every possible application of data science in climate. If you like this blog post, I encourage you to become curious and to search for other ways that data science can help to fight the climate crisis. Or, equally important, what you can do as a human being to fight for a better future.

Author: Piotr Grabysz

Photo source: Unsplash

Bibliography

[1] Spain floods: Three dead and three missing after torrential rain | Reuters

[2] Record rainfall in Spain leads to deaths and travel chaos | Euronews

[3] Tellman B, Sullivan JA, Kuhn C, Kettner AJ, Doyle CS, Brakenridge GR, Erickson TA, Slayback DA. Satellite imaging reveals increased proportion of population exposed to floods. Nature. 2021 Aug;596(7870):80-86. doi: 10.1038/s41586-021-03695-w. Epub 2021 Aug 4. PMID: 34349288.

[4] Allstate, State Farm stop selling Californians new home insurance - Los Angeles Times (latimes.com)

AI Act is here – there is time to explain, get into the car.

Artificial Intelligence (AI) has become an integral part of our lives, from virtual assistants to self-driving cars. As AI systems become more sophisticated, so do the challenges associated with their deployment. In response, the European Union (EU) has introduced the AI Act, a comprehensive regulatory framework aimed at ensuring the safe and ethical use of AI technologies.