Member Spotlight: JBA Risk Management
Dave, what does your role at JBA involve?
I joined JBA Risk Management in 2014 after completing my PhD and working at various research posts. These mostly focused on flood risk management and computational science, where I used computer modelling to investigate environmental phenomena. At JBA I work on a range of projects, using my hydrology knowledge to provide ideas, support and advice. Another part of my role is to keep on top of the latest scientific developments in hydrology and modelling. I also work on my own specialist projects which tend to centre on rainfall-runoff modelling and flood frequency estimation.
How does snowmelt cause flooding?
Snow will accumulate during winter in mountainous locations and in all regions above 40° of latitude. This includes all of Canada, a lot of the northern US, the Nordic countries, New Zealand and the European Alps among other locations. During winter, freezing temperatures plus precipitation results in the build-up of a snowpack. As temperature changes from freeze to thaw during spring the snowpack melts, releasing its water to the river. This can cause fluvial flooding should the river burst its banks.
Can you give any examples of past flooding events that have been caused by snowmelt?
The classic example would be the Red River, which runs across the US-Canada border in Minnesota. The river flows through Grand Forks, a town with a population of around 60,000. In early April 1997, all the ingredients for a disaster were in place: following a winter of extreme snowfall, large amounts of snow remained in the river catchment and spring was arriving quickly. Flood planners predicted a serious snowmelt flood event. Volunteers and military personnel built extensive sandbag dikes, but by April 17th the river level reached over 16 metres – the highest levels since 1826. This overtopped all defences, inundating the entire town. The population was evacuated, with deep waters remaining in the area and preventing people from inspecting their damaged homes for 2 weeks.
The event was so severe that town officials, working with FEMA, instigated a buy-back scheme for the most exposed properties to prevent a similar disaster in the future. The financial damage from the flood is estimated at $3.5 billion.
What are the challenges in predicting snowmelt?
Snowmelt is a complicated process with a lot of physics going on. The snowpack undergoes complex transformations as it builds up, matures and eventually melts. The process is affected by the condition of the underlying ground, the direction of the slope of the ground, the angle of the sun, cloud cover and air temperature. One approach to predicting snowmelt involves trying to represent all this complexity in a computer model, but this is only possible for small areas where a lot of information can be gathered.
An alternative approach is to make a much simpler model. Altitude, latitude and a daily time-series of temperature and precipitation can be used to build a concise but useful model of how a snowpack is growing and melting throughout the year. Coupled with a rainfall-runoff model this can be used to simulate snow-affected rivers.
How is climate change affecting snowmelt and related flooding?
Climate change is (very generally) increasing both global temperatures and precipitation. Increased rainfall suggests more flooding while higher temperatures suggest less flooding due to a general drying out of soils. Whether flooding at a location increases or decreases with climate change becomes a factor of which of these two processes “wins out”. When snow is included, things get interesting. For example, studies in Alaska show that the onset of the spring snowmelt season is now up to eight days earlier. However, because hours of daylight are shorter earlier in the year, the melt takes place more gradually resulting in less dramatic flow and flooding.
Modelling studies show that under climate change scenarios, warmer winter temperatures make it much more difficult for large snowpacks to build up. These are the type of snowpacks that typically cause severe flooding the following spring. The implication is that despite climate change bringing a lot of negative consequences, spring snowmelt flooding may actually be reduced in the future. This could be good news for flood management, but it may cause challenges for water resource planning, hydropower and local ecology.
What work is JBA doing to help insurers predict future flood risk?
JBA produces flood data and models for every country in the world down to an individual property level for different climate change scenarios.
We are also working on a range of bespoke climate change projects, developing data to meet specific needs and requirements – for example, for disaster risk reduction in low-income countries and helping insurers meet regulatory requirements, both existing and pre-emptive.
How does JBA communicate uncertainty in its climate change projections?
JBA believes that uncertainty estimation is going to play a significant role in risk management in the near future. It’s possible to provide clients with uncertainty estimates using a combination of statistics and computational methods. The challenge is how the insurance industry makes use of this information. At JBA, we believe we should be open with uncertainty information and we want to lead on providing guidelines and frameworks on how, as an industry, we work with uncertainty. JBA has formed a working group to bring together those from the industry and academia to help risk managers better understand and use uncertainty.
How do insurers use the climate change insights that JBA provides?
One of the main ways insurers use our models and data is for risk pricing, exposure accumulation management, and geographical hotspot identification. Insurers want to ensure that no singular catastrophic event can disproportionately impact their losses. Our insights can also be used for future portfolio planning, capital management and meeting regulatory requirements.
What else is JBA working on over 2023?
Building on the launch of our climate change data and models for every country in the world last year, we are adding new science-backed features and assimilating the growing body of model output data produced by the world’s climate change science community so that we can continue to represent standardised climate change scenarios and identify coherent storylines within this data.
An exciting development for JBA is the move towards supplying data via API. Data sets would traditionally be delivered by a one-off upload of files. Whilst this remains popular for our clients, it is now possible for them to embed JBA’s data into their workflows via our API.
Our latest UK flood map release now takes into account the changes in flood frequency that have resulted from the existing impacts of climate and land-use change. This was made possible using bespoke statistical methods applied to the most up-to-date UK river flow data sets.
What should readers do if they want to learn more?
Our website is the best place to go – it has a range of case studies and blog posts describing some of our work. Contact details can also be found from there.