Met Office supercomputers in the pipeline for seasonal forecasting

The Met Office has been providing weather forecasts for the BBC since 1923, and even after 88 years, they can still get the weather wrong. MPs have backed the claim that scientific advances are being held back by insufficient computing capacity, and as a result, new supercomputers for the Met Office could not be too far around the corner.


The new supercomputers would be able to handle more data, and process more complex calculations. Not only would this produce seasonal forecasts on top of the daily and weekly, but also give a much better estimation of the probability of each weather event occurring. One of the great advantages would be improved contingency planning for emergencies. More on this story can be found on the BBC website here.

Climate Change and Cyclone Damage

Satellite Image of Cyclone Yasi from Feb 2011 (Image Source)

A very belated Happy New Year - it's been a while since I last blogged! Earlier today I was kindly pointed in the direction of a new paper which was released in Nature this week (the article can be found here).

The paper ties in nicely with some of the latest cyclone research that I blogged about in December. The purpose of the research was to assess future damages caused by cyclones and climate change. They are clear in their introduction that future scenarios with cyclone models are sketchy at best and that despite increases in cyclone numbers in the past 40 years, they are still well within natural variability. The paper is also quick to note that increasing damages in the last few decades are largely down to the increasing population and more assets in harms way rather than as a result of anthropogenic warming. However, they have taken the scientific viewpoint further by modelling cyclones in separate ocean basins, taking into account frequency, intensity and regional location changes (rather than just wind speeds and sea surface temperatures as seen in previous studies and models). The figure below shows the global storm tracks and their intensity (minimum pressure). It clearly indicates that the storms are more frequent over the warmer equatorial waters, and most occur in the Western Pacific.

The present annual global damage from tropical cyclones is US$26 billion and is used as a baseline for scenario comparisons up to 2100. The results in this paper, reveal (as would be expected) that the distribution of climate-change damage is not even across the world. The figure below shows the estimated projected damage caused by climate change in each region. North America has the highest average damage of US$26 billion yr-1, which is half of the global damage. East Asia and Central America Caribbean average damages of US$15 and US$10 billion yr-1 respectively.

There are some limitations with this study and its projections, which largely revolve around economics and countries GDP and growth rate. Damages vary a great deal if a storm hits a city versus a rural area, but this is not yet captured in this analysis. Moreover, increases in income and population along the coast, relative to the rest of the country, will cause more damage. There is also no account of any adaptation policy that could be put in place. Despite these limitations it is a nice to see an initial attempt at quantifying costs of cyclone damage associated with climate change.