The question of how rainfall intensity will change with global warming is an important one, and confidence is building within the scientific community that rainfall extremes will on average become more intense and/or more frequent as global temperatures increase. While this may be true on average, however, what is perhaps less well appreciated is that there are different types of weather systems that produce rainfall, and that these might change in different ways and sometimes even in opposing directions as the climate warms. This substantially complicates efforts to provide robust projections of likely changes to rainfall extremes at any specific location, and therefore hampers efforts to adapt to future flood risk.
At the largest scale, global atmospheric circulation patterns such as the Hadley cell, which are responsible for transporting huge volumes of moisture over large distances, play the leading role in determining the global distribution of rainfall. An increasing number of scientists are finding that these circulation patterns are changing, with the consequence that the distribution of global precipitation might also change. This means that while on balance the total global volume of rainfall is expected to increase with climate change, changing circulation patterns will cause certain regions, and particularly the arid and semi-arid zones which are already relatively dry, to become drier still.
In contrast to these global and regional averages, the intensity of precipitation extremes is expected to increase by a much larger margin. An heuristic reason for this is that the rainfall from the most intense storms is thought to be proportional to the amount of moisture that the air can hold, with the moisture-holding capacity of the air in turn increasing as a function of the temperature of the atmosphere. In short, a warmer atmosphere can hold much more moisture, and this can lead to much more intense bursts of rainfall. Such increases in the most extreme rainfall events would occur even in areas where average rainfall intensity decreases, so that by implication, the distribution of precipitation must change in time. This will lead to more intense bursts of rainfall occurring less frequently or for shorter durations, as described in a recent paper by Kevin Trenberth and illustrated on the schematic below. One of the challenges, however, is quantifying this change: by how much will extreme precipitation increase?
To this end, our research group developed a technique to estimate how rainfall might change in a future, warmer climate, by looking for locations or seasons which are ‘analogous’ to the climate that will occur in the future. Say, for example, that we are interested in the future rainfall patterns that might occur in Sydney in 2050. We could then use climate models to calculate the expected atmospheric temperature and other variables that might occur in Sydney at this time, and then find locations that currently have such climate features. Following through with the Sydney example, this might involve looking further north since more northerly latitudes are generally warmer and more humid.
The outcomes of this research were that in almost all the cases examined, it was the shortest bursts of rainfall – often lasting for only an hour or less – which appeared to be most susceptible to an increase in intensity as the climate warms. This will have profound implications on flood risk, particularly in urban catchments and steep mountainous catchments, which respond very quickly to rainfall. Methods that can calculate the magnitude of these future changes are therefore essential if we are to successfully adapt to a future of shorter but more intense rainfall events.
Westra, S., Evans, J., Mehrotra, R. & Sharma, A., 2013, “A conditional disaggregation algorithm for generating fine time-scale rainfall data in a warmer climate”, Journal of Hydrology 479, DOI: 10.1016/j.jhydrol.2012.11.033