Land surface processes are an important
and difficult aspect of climate modeling. Important, since it affects primarily
the lower part of the atmosphere and therefore the climate variables which
are of interest for impact studies. Difficult, because it rellies on heavely
parameterized processes leading to very different schemes. The international
PILPS
project was therefore launched to provide a benchmark and compare these
various parameterizations. I was involved with this project and led the
Phase
4(b) which aims to couple several land-surface schemes to the regional
forecast model of the BMRC: LAPS.
This work was focused on the impact of such parameterization on short time-scale
relevant to weather forecast; more details can be found here.
An other important aspect of this work and other PILPS sub-projects was
to investigate
rigorous
and homogeneous ways to couple land-surface schemes into atmospheric
models. I am also interested by the role played by soil moisture in climate
predictability
in particular over Australia.
