Cloud Feedbacks



Other MIPs



Google Calendar

Lab Calendar

Site Map


Privacy & Legal Notice

Thanks to Our Sponsors:

PCMDI > WCRP CMIP3 Model Output > Diagnostic Subprojects Printer Friendly Version
<< Back to WCRP CMIP3 Subprojects

Seasonal rainfall and climate change in African, Asian and Maritime monsoon regions

PI: Dan C. Collins
Institution: Georgia Tech
Additional Investigators: Peter Webster, Thomas Hopson
An analysis of the seasonal rainfall in the African, Asian and Maritime monsoon regions, interannual variability and climate change as simulated by GCMs will be undertaken. Our objectives are to assess: 1. Do GCMs forecast (or hindcast) the interannual variability of seasonal rainfall in monsoon regions that we have been examining using reanalysis, satellite and in-situ data? 2. Is the relationship of simulated interannual precipitation variability to GCM generated sea surface temperature fields the same as in reanalysis or real data? 3. Can this information be used to generate an improved multi-model ensemble forecast of monsoon precipitation under future climate change? We will derive a multi-model ensemble forecast of the probability of changes in the seasonal precipitation and interannual variability during the 21st century (and possibly the 22nd and 23rd) using a combination of subjective and objective skill tests of the GCMs. Objectively, simulations of the 20th century can be compared to seasonal precipitation data, generating a skill score for each model. Recognizing that these climate simulations cannot forecast individual years, we would test the skill of model simulations of the seasonal mean and interannual variability for periods on the order of decades, and we would focus on area average precipitation for each monsoon region. For a 10-year period for example, each model would generate 10 ensemble-member forecasts (each ensemble member from one year of seasonal precipitation), and for example a rank probability skill score or cross-probability density function for each model would be assembled from the complete 20th century. Further analysis will help determine why individual models produce good or poor monsoon precipitation simulations. Correlations of interannual precipitation variability to SST (and lagged SST) would be examined for each model and compared to the correlations using climate data. Models that do not produce a reasonable monsoon simulation could be objectively excluded from a multi-model ensemble forecast of future climate change. Probabilistic forecasts of future monsoon precipitation will be determined using the skill of each model to generate a weighting function. Our primary objective would be to determine if a useful seasonal rainfall hindcast for monsoon regions could be calculated from GCM simulations, and then to apply the same procedure to monsoon precipitation forecasts in future climate change simulations in a form of nearest-neighbor forecast. We would primarily need only model precipitation and sea surface temperature fields for each GCM to carry out this research; however surface temperature, pressure and wind fields would be of interest in assessing the simulation of monsoon dynamics.

    Add Publication

    << Back to WCRP CMIP3 Subprojects
    For questions or comments regarding this website, please contact the Webmaster.
    Lawrence Livermore National Laboratory  |  Physical & Life Sciences Directorate