There are a number of regional climate models for different regions. I am interested to downscale global climate models based on regional climate mode for dynamical downscaling. Any suggestion is appreciated.
Just like Liew's suggestion, CORDEX is a good place to start.
http://cordex.dmi.dk/joomla/
has more information. You can also specify countries or shapefiles into ClimateWizard to get predictions and hindcasts downscaled to a 50 km grid. Challenges remain especially in mountain areas where there is a lot of microclimatic variability. The generally decreasing vegetation cover across much of Subsaharan Africa also has implications on both precip and runoff. I wonder how this can be incorporated into downscaled models.
As mentioned earlier, there is CORDEX Africa. Already hear that in South Africa and Tanzania are using the REGCM4 for long simulations using dynamic downscaling. Another possibility is to use the WRF-CORDEX (http://www.meteo.unican.es/wiki/cordexwrf/CordexWrfGroups). I am part of this group (http://www.meteo.unican.es/wiki/cordexwrf/GroupPPGCC), where our study domain is the Equatorial Atlantic from the coast of Brazil to the coast of Africa.
Beside CORDEX, we are also involved in higher resolution simulations for WEST AFRICA (22km, 18km, 11km) with WRF and WRF-HYDRO. If your region of interest is West Africa, you can go to http://www.wascal.org
If you are a beginner in RCM and intend to conduct own simulations related to your question, you need a strong partner who understands what the model is simulating.
Otherwise, I recommend to analyse the data available befor starting own additional simulations. However, for correct interpretation and appropriate selection of data our of the huge amount of data I recommend to cooperate with those who produced the data.
You may use PRECIS model output for African region. As suggested by other CORDEX data is the best choice. We are working on Indian Summer Monsoon Rainfall. If you need more, may write mail to me on
I think CORDEX site is very good way for you. But you know, you can use other RCM for your aim. I mean that you can use RegCM4 with yourself. You can run this model with nested way, I mean that you should run this model in several run and in every run you can give finer and finer outputs with nested. But absolutely this way is very time consuming, need strong cpu with many cores and... . But it is a safe and precise way to receive outputs data.
Of course, there are many models and a broad band of evaluation results exhibiting a substantial spread.
If you are not familiar with a mesoscale model, I cannot recommend conducting own simulations without an experienced cooperation partner. It's easy to produce numbers which don't describe the nature in an acceptable way.
Using the CORDEX results first is strongly recommended.
You find all models and the reference configuration for Africa on the CORDEX-AFRICA webpage. However, there are several attempts to simulate parts of Africa at higher resolutions by different groups. If you are interested in that, you probably need a non-hydrostatic model system like COSMO-CLM or WRF. E.g. at the CLM-Community webpage you may find a contact person who can answer your detailed questions.