View from the Top: Climate change's impact on tropical storms

The consensus in the 2012 IPCC and 2013 IPCC AR5 report is that tropical storms are likely to increase in intensity and rainfall but decrease in frequency with a warming climate. The opinion was that there would be fewer storms globally with more uncertainty in individual basins; an increase in average wind speeds globally but not in all basins; more frequent intense storms; higher rainfall rates in tropical cyclones; and sea level rise to exacerbate storm surge impacts. (IPCC Special Report 2012).

Increased activity and intensity
This study is based on a paper published in the Proceedings of the National Academy of Sciences. Dr Kerry Emanuel from the Massachusetts Institure of Technology has suggested that there will not only be an increase in more severe storms but an overall increase in global tropical cyclone activity, mostly in the North Pacific, but also in the North Atlantic and South Indian Basins. The intensity of storms, as measured by their maximum wind speeds, also increased which agrees with previous studies.

The analysis was based on simulations from six of the Coupled Model Intercomparison Project models and used downscaling which embeds high resolution regional or local models within the general circulation models. An increase in overall frequency of storms was observed when the models were run using historical conditions and representative greenhouse gas concentrations as input.

There are problems with the analysis including mismatch in the physics between the local and global models and the lack of feedback to the global models, but the approach does produce higher spatial resolution of the tropical cyclones.

Coastal populations
This study investigated potential increase in losses due to climate change for three countries; China, Japan, and Australia, as these countries have large populations exposed to tropical cyclone risk. In China, close to 60% of the population live in 12 coastal communities, along the Yangtze River valley, and in two coastal municipalities - Shanghai and Tainjin. Nearly 80% of Japan's population is considered coastal and no one lives more than 120 km from the sea. Australia's coastal population is expected to increase by over 1m over the next 15 years.

To simulate the impact that an increase in frequency and severity of storms would have on potential losses, one of the third party catastrophe models was used. Standard industry numbers were taken from the aggregate model for the three countries. The assumption was made that these countries had a similar increase in intensity and frequency. The time horizon was assumed to be the end of the 21st century with worst case scenario being overall frequency increased by 40% and the severe storms (SS Cat 3-5) also increased by 40%.

Ranges were also calculated by assuming that when the severe storms are increased, there are either more or less of the larger loss producing storms. It should be noted that an increase in storm intensity doesn't necessarily mean expected losses are higher as less intense storms usually have larger radii and affect a much larger area for a longer period of time, hence leading to potentially higher losses.

Differing potential losses
The results of the simulations indicate that losses could potentially double with an increase of 40% in both overall frequency and severe storms, but there is high variability, between 100 and 250%, depending on many factors including characteristics of and locations where the storms make landfall.

With climate change, there is a likely change in cyclone tracks with pole-ward shifts; they will start to impact the southern parts of Australia and the northern sections of China which are heavily populated and therefore vulnerable areas. The major increase in losses, not modelled, will be due to the increase in building stock and many other factors including, increased penetration rate, more precipitation, higher coastal surge, etc.

A recent paper in Nature (Future flood losses in major coastal cities) suggests that without an upgrade of the present coastal protections the coastal flood losses worldwide could reach $1trn annually.

Dr Graham Cook is head of analytics at Hong Kong-based reinsurer Peak Re