PROMISE - Second six-monthly report

Reiner Schnur, Max Planck Institute for Meteorology

 Since the first annual report, the vegetation feedback on Sahelian rainfall variability was further analysed with the simple dynamic vegetation model SVEGE coupled to the ECHAM4 atmospheric model (see WP1300 in annual report). In addition to the earlier coupled runs with fixed mean and fully interactive vegetation, a new experiment was performed using a fixed annual cycle of vegetation. Results support the earlier conclusion that the interactive modeling of dynamic vegetation processes is necessary to substantially improve the simulation of multi-decadal rainfall variability in the Sahel region, as observed during the last 5 decades. The publication "Vegetation Feedback on Sahelian Rainfall Variability in a Coupled Climate Land-Vegetation Model" by Schnitzler, Knorr, Latif, Bader and Zeng has now been submitted to Geophys. Res. Lett.

The analysis of the sensitivity of West African rainfall to SST anomalies (reported in WP1200 in the annual report) has been continued. New experiments with the ECHAM4 GCM confirmed the earlier result that rainfall variability along the Guinea Coast is strongly linked to SST anomalies in the Eastern Tropical Atlantic, while there is no clear relationship with Sahelian precipitation. It was found earlier that there is a high correlation between Sahelian rainfall and SST anomalies near Greenland. A set of (prescribed SST) sensitivity experiments is now in progress to investigate this relationship. An analysis of NCEP reanalysis data showed that rainfall in the Sahel region is mainly determined by changes in wind and latent heat flux over the Atlantic leading to changes in the southwest monsoon.

Modelling work is still in progress to implement an improved land surface scheme and a dynamic vegetation model in the new ECHAM5 climate model. This framework will be available during the PROMISE period to study the impact of land surface conditions on monsoon variability.

We don't foresee problems to meet the year 2 milestones. We will make available data from our model integrations for the PROMISE data archive very soon.