Infrastructure resilience to climate change

Author: Dr Elco Koks, Assistant Professor within the Water and Climate Risk group at the Institute for Environmental Studies (Vrije Universiteit Amsterdam)

When we evaluate the risks climate change poses to infrastructure, there are two aspects to consider. The first is the changing nature and frequencies of the climate hazards themselves – using attribution studies to track the likelihood of extreme events.

My specialism is in the other area: understanding where and how infrastructure is exposed and vulnerable to climate risks. In my work at both Vrije Universiteit Amsterdam and Oxford University, I look for the weak spots in the chain; bottlenecks where a single point of failure can undermine resilience overall.

The field is developing rapidly – moving from stress testing infrastructure in potential scenarios to examining the cascading effects of a climate hazard, and even consecutive and concurrent events. If a tropical cyclone in Houston coincided with flooding in Rotterdam, how would global shipping respond?

Often, we find insurers and asset managers look at the wrong things when assessing climate risk – or rather, they take a narrow view.

Focus on more than assets

By limiting resilience to assets alone, we risk missing the bigger picture

One of the main reasons infrastructure can be vulnerable to climate events is that risk mitigation focuses on the physical asset alone. This is understandable – an insurer or asset manager needs to protect the investment. But frequently, a risk is less about the asset itself, and more to do with how it’s used.

For example, there have been cases where extreme weather conditions offshore have prevented ships from docking, without damaging the port itself. The physical asset is safe from damage, but nonetheless, it can be rendered unusable by disruption to navigation routes.

Ultimately, the changing climate might sometimes require a new way of working. For example, lean and just-in-time manufacturing practices – along with distributed, global supply chains – have dramatically improved efficiency, and accelerated time to market. But they’re also vulnerable to disruption, as individual issues have a multiplied, knock-on effect.

In 2011, serious flooding swept across Thailand, causing widespread devastation. As well as the tragic local consequences, the floods caused prolonged chaos in the global computer industry – because a few factories, in a single location, produced half the world’s magnetic hard drives.

Even now, ten years later, many of the systems we rely upon to support society and trade have deliberately prioritised efficiency at the expense of resilience. But as the climate changes, that balance may need to change too.

Risk exposure is getting worse

Despite all we know, new developments are increasing vulnerability to climate risk

As an academic working to understand how infrastructure is exposed to climate risk, it can be frustrating to see how often decisions continue to increase that vulnerability.

We know for certain that flooding will increase, and that sea levels are going to rise. But still new homes and facilities are built in flood-prone areas, increasing the exposure that we already have.

Here in the Netherlands, there are discussions about building a whole new city at four or five metres below sea level. That’s worrying – perhaps there’s a lack of understanding about the risk.

Even though we’re in 2021, there’s lots we don’t know about the potential impact of natural hazards on our infrastructure assets and systems. This field has an entire community around decision making under deep uncertainty (DMDU), because we have to make choices without fully understanding how everything will play out.

That uncertainty makes it harder to communicate the issues and influence key decisions – especially where there are few people outside academia whose job is to focus on the broader, interconnected impacts a hazard can bring. But if we can acknowledge the network may not be fully resilient, that’s a start.

Collaboration can unlock meaningful change

New technology and datasets have the potential to give us a clearer, more granular view. For instance, we’re about to commence work on using machine learning to identify infrastructure assets from high-resolution satellite imagery.

But to see a real step change in the level of certainty in our modelling, we need greater access to asset data. Companies are understandably reticent to share sensitive information on locations and vulnerabilities – while other data, like shipping information, is a commercial asset worth thousands or millions of Euros.

As a result, my work involves a lot of assumptions about numbers and locations of infrastructure assets, and how they’re used. We can still draw sensible conclusions, but within a defined bandwidth of uncertainty. The more and better data we have to work with, the more certain we can be.

Some new developments – like OpenStreetMap – are opening up new avenues of research on networks and asset-level risk. Communities are working together to create the mapping data they need, and it makes a real difference.

In a way, that’s a model for how we should collaborate to make our infrastructure resilient to climate risk. Academic institutions can partner with corporations – as well as public bodies – to interpret data and better understand cascading, interconnected effects across entire networks and systems. Because the economic and societal impacts of climate change will affect us all.