Environmental Engineering Reference
In-Depth Information
evapotranspiration during the storm; therefore, a storm with a given structure
(with given vertical motion in particular), will generate more precipitation
in a warmer climate, roughly following the increase in saturation vapor
pressure, to the extent that the circulation within the storm does not change
significantly. Additionally, as discussed in Section 4.2, evapotranspiration on
average increases more slowly than does the water-holding capacity of the
atmosphere; therefore, if precipitation intensity increases, then the frequency
of precipitation must decrease to preserve a global balance of precipitation
and evapotranspiration.
That climate models behave in this fashion, to first approximation, is
documented in a number of papers (Emori and Brown, 2005; Pall et al.,
2007; Allan and Soden, 2008; Sugiyama et al., 2010). Although there are
some changes in the intensity of vertical motion, increases in extreme pre-
cipitation are dominated by the changes in atmospheric moisture. Even in
the subtropical semi-arid zones in which mean precipitation is projected
to decrease, the highest percentile daily precipitation events increase in
frequency, as shown by Pall et al. (2007).
O'Gorman and Schneider (2009) have analyzed an idealized atmo-
spheric model and point to the importance of the vertical gradient of the
saturation mixing ratio in the lower troposphere rather than the water vapor
mixing ratio itself as the key quantity whose increase controls intensity
changes. The percentage change in this gradient is somewhat smaller than
that of the mixing ratio but is of the same magnitude for typical atmospheric
conditions, roughly 5% per degree warming. Limitations of this simple ther-
modynamic perspective are most likely to occur in the tropics, where latent
heating is essential in driving circulations and where guidance from climate
models is more suspect than in extratropical latitudes.
Allan and Soden (2008) compare precipitation estimates over the tropi-
cal oceans and GCM simulations to investigate the dependence of extreme
precipitation intensities on surface temperature. They find the greatest rate
of increase at the highest (most intense) precipitation percentiles in both
models and observations. Furthermore, the inferred increases from the
satellite precipitation data were considerably larger—in some cases over
double—those predicted by the GCM simulations for the most extreme
precipitation rates investigated.
The most extreme storms in mid-latitudes are also likely to be affected
in structure by the increased latent heating, possibly causing changes in ex-
treme precipitation larger than suggested by the thermodynamic arguments.
For example, Lenderink and Van Meijgaard (2008) analyzed a 99-year
hourly precipitation record from the Netherlands along with a regional cli-
