One of the most alarming projections associated with human-induced climate change is the potential for an increase in the intensity and frequency of rainfall extremes. But how much do we really understand about the likely changes to extreme rainfall patterns over the coming decades?
We know that the amount of moisture the atmosphere can hold (the saturation level, when the relative humidity reaches 100%) increases with temperature, so that a warmer atmosphere can hold more moisture. This is why when we take a cool bottle of water out of the fridge on a warm and humid day, we get water droplets forming on the outside of the bottle; in this case the air near the bottle cools down and becomes super-saturated, and results in the water vapour condensing into liquid water on the bottle surface.
It turns out that the moisture-holding capacity of the atmosphere will increase at roughly 7% per degree Celsius, following what is known as the Clausius-Clapeyron relationship. Assuming that the global average surface temperature will increase by between 1.1 and 6.4°C by the end of the century (with the actual amount of warming depending to a large extent on how much we choose to reduce our greenhouse gas emissions), and that the amount of water vapour in the atmosphere will increase in proportion to the capacity of the atmosphere to hold that vapour, then we can expect much more moisture in the atmosphere in the future.
In an important 2003 paper, Kevin Trenberth and colleagues suggested that the intensity of the most extreme rainfall events will scale with the amount of moisture available in the atmosphere, so that extreme rainfall intensity could be expected to scale with atmospheric temperature at that same 7% per degree rate. He went on to suggest, however, that much higher scaling rates are also possible because the ‘latent heat release’ (the energy released by condensing water vapour into liquid water) inside the cloud can further invigorate the storm system, increasing the vertical motion of air through the storm and thereby further increasing the intensity of extreme rainfall. Such a possibility was also indicated in a recent Dutch study, where it was shown that historical rainfall extremes could scale up to double the Clausius-Clapeyron rate; in other words, by up to 14% per degree change in temperature.
Unfortunately using such fundamental relationships to derive projections of rainfall extremes for any individual region is much more complex. For example, although the amount of water the atmosphere can hold scales at a rate of 7% per degree, the actual increase in atmospheric water content may be less than this rate at some locations, particularly in dry continents such as Australia where access to moisture for evaporation is limited. Similarly, because of projected changes in large scale atmospheric circulation patterns, some areas might have much stronger increases than other areas, and in a limited number of locations extreme rainfall may even decrease. Furthermore, in a recent study, it was shown that the tropics appear to have a higher sensitivity to increasing temperature compared with the extra-tropics.
To complicate things further, the types of rainfall that we can expect might also change, going from a dominance of large scale rainfall at cooler temperatures to an increase in convective systems at warmer temperatures. In our own research, my colleagues and I have been looking at changes in Australian rainfall at different timescales, and we a found much greater sensitivity and larger trends at shorter (sub-hourly) durations compared with longer (daily or multi-day) timescales.
So although we are confident that extreme rainfall will increase on average globally, there is such a diversity of factors at play that projections at any location are still highly uncertain. To better understand what changes can be expected at any given location, global and regional climate models are often employed as they represent one of the most powerful tools that we have available to develop such projections. In a future post I will talk more about how these models are being used to inform our understanding of likely future changes in precipitation extremes.