Geoscience Reference
In-Depth Information
simulations with parametrised convection, although they found that this did not necessarily
result in a better MJO simulation.
Comparing precipitation from a 1.5 km resolution regional model simulation over the
UK with gridded hourly radar rainfall, Kendon et al. (
2012
) showed that, although the
1.5 km model appears to have a tendency for heavy rain to be too intense, it gives a much
better representation of its spatial and temporal structure. With parametrised convection,
heavy rain events are not heavy enough and tend to be too persistent and widespread. There
are also not enough short-duration high-peak intensity events. These biases are signifi-
cantly reduced in the 1.5 km model. In addition, the 1.5 km simulation shows a much
better representation of the occurrence of dry days.
Similar discussions regarding model horizontal resolution concern the storage, move-
ment, and quality of water at and near the land surface. A recent opinion paper by Wood
et al. (
2011
) discusses the need for much higher horizontal resolution in continental- or
even global-scale models in order to include detailed information about these processes.
They suggest that ''Adequately addressing critical water cycle science questions and
applications requires systems that are implemented globally at much higher resolutions, on
the order of 1 km, resolutions referred to as hyperresolution in the context of global land
surface models''. However, they note the need for improvements in satellite remote sensor
resolution and the development of advanced downscaling methodologies, as well as
improvements to in situ observation networks, in order to support such a modelling effort.
We have described research which indicates that improving model horizontal resolution
can have significant benefits for modelling the hydrological cycle. However, improved
representation of hydrological processes through better physical parametrisations is also
crucial. This is discussed in the next section.
2.3 Representing Hydrological Processes and Phenomena
As mentioned in the introduction, although great advances have been made in global
modelling in recent decades, there are still large uncertainties in many processes, such as
clouds, convection, and coupling to oceans and the land surface. In the previous section,
we discussed the influence of model resolution on these uncertainties. In the next sub-
sections, we highlight the role of model physical parametrisations in the representation of
cloud microphysics, convection and precipitation processes, in the interactions between the
atmosphere and the land, the oceans and the cryosphere, and in the representation of one of
the largest seasonal phenomena in the hydrological cycle: the monsoons.
2.3.1 Clouds, Humidity and Radiation
The surface radiation balance is a key determinant of the intensity of the GWC. Decadal
variations in the surface radiation balance during the 20th century are reflected in the
variations in intensity of the GWC, but these are not captured in models. The simulation
of cloud processes and feedbacks by GCMs remains one of the most critical aspects of
climate modelling. In particular, cloud radiative feedbacks remain the primary source of
uncertainty for transient and equilibrium climate sensitivity estimates (e.g., Randall et al.
2007
; Dufresne and Bony
2008
) and play a critical role in anthropogenic aerosol-induced
climate forcing (Lohmann and Feichter
2005
).
In addition, clouds play a key role in the hydrological cycle and in the large-scale
atmospheric circulation, at both planetary and regional scales. By affecting precipitation
and atmospheric dynamics, uncertainties in cloud and moist processes remain a major
