Geoscience Reference
In-Depth Information
southerly flow to the west and northerly flow to the east. The induced meridional flow
influences rainfall via moisture advection, triggering a westward shift in rain, soil moisture,
and low-level temperature. Thus, maxima in cold cloud (indicating deep convection) tend
to follow maxima in southerly flow with a lag of the order of 1 day (Fig.
4
a), and negative
surface heating anomalies due to increased convection, with associated negative low-level
temperature anomalies, also propagate westwards (Fig.
4
b). Douville et al. (
2001
) dem-
onstrated that increased soil moisture over the Sahel region is also associated with
increased monsoon rainfall on the seasonal timescale. However, they also showed that the
South Asian monsoon does not respond in the same way. This is discussed further in Sect.
2.3.5
.
Koster et al. (
2004
) queried whether there exist specific locations (''hot spots'') on the
Earth's surface for which soil moisture anomalies have a substantial impact on precipi-
tation. In the Global Land Atmosphere Coupling Experiment (GLACE), a dozen climate
modelling groups performed the same highly controlled numerical experiment as part of a
coordinated comparison project, allowing a multi-model estimation of the regions on Earth
where precipitation is affected by soil moisture anomalies during Northern Hemisphere
summer. These included the central Great Plains of North America, the Sahel, equatorial
Africa, and India, and, to a lesser extent, South America, central Asia, and China. How-
ever, there was extensive inter-model variability in the strength and positioning of the ''hot
spots'', a reflection of ongoing uncertainty in the proper way to represent in models the
physical processes defining land-atmosphere coupling strength.
Comer and Best (
2012
) applied the GLACE approach to a recent climate configuration
of the MetUM and showed a significant strengthening of the atmosphere's response to soil
moisture compared with the earlier version, HadAM3 (the third Hadley Centre Atmo-
spheric Model), which was one of the weakest of the models assessed in the GLACE
intercomparison project. This strengthening was particularly evident over the Sahel region
of Africa, which was identified by the GLACE project as a key ''hot-spot'' region for land-
atmosphere coupling (Koster et al.
2004
). Comer and Best (
2012
) could not attribute this
improvement to developments in the land surface scheme; instead, it appeared to be mainly
due to improvements in the atmospheric parametrisations and possibly also the model's
vertical resolution. In addition to testing the coupling strength in the updated model, Comer
and Best (
2012
) also demonstrated a dependency of the coupling strength on the way in
which soil parameters are specified.
Understanding the nature and timescale of land-atmosphere coupling is an area of
active research around the world. Various studies have shown that there may be both
positive and negative feedbacks between soil moisture and precipitation depending on the
time and space scales considered (e.g., Cook et al.
2006
; Taylor
2008
; Hohenegger et al.
2009
; Schlemmer et al.
2012
; and references therein). Observations and idealised mod-
elling studies are being used to improve our understanding of land-atmosphere interac-
tions. For example, Schlemmer et al. (
2012
) investigated the influence of soil moisture and
atmospheric stability on mid-latitude diurnal convection and land-atmosphere exchange
using an idealised cloud-resolving modelling framework and demonstrated a positive soil
moisture-precipitation feedback on a scale of 10-50 km. In contrast, Cook et al. (
2006
)
showed a negative soil moisture-precipitation feedback in southern Africa that was related
to stronger atmospheric stratification over wet soils and to the formation of anticyclonic
circulations that induced subsidence and divergence at the surface. Hohenegger et al.
(
2009
) found a similar stabilization of the atmospheric profile over wet soil in summertime
Europe. Furthermore, they found different signs of the soil moisture-precipitation feedback
