Four- dimensional climate data sets of the AMMA Special Observing Period #3 Leonard M. Druyan, Matthew Fulakeza, Patrick Lonergan and Erik Noble The African Monsoon Multidisciplinary Analysis (AMMA) is a concerted effort to improve our understanding of the climate of West Africa, especially the variability of precipitation systems. The goal is for this increased understanding to improve techniques of seasonal prediction of rainfall anomalies, including droughts and floods, which have a devastating socio-economic impact on very vulnerable populations of the Sahel. In addition, there is special interest in studying disturbances entering the North Atlantic from West Africa, since they are often precursors of tropical storms. The AMMA Special Observing Period (SOP-3), September 2006, employed extensive networks of ground-based and aircraft observations, supplemented by measurements from satellite-borne sensors. However, despite its ample scope, the collected SOP-3 observations cannot specify the complete space-time distribution of any of the climate variables. The realism of a four-dimensional matrix of the time evolution of atmospheric conditions generated by a regional climate model (RCM) can be  validated against SOP-3 data. Eventually the analysis might be improved by 4-D assimilation of observations into the integrating model output. However, the appropriate model must be capable of simulating the important climate components of the West African monsoon (WAM). The presentation will show climatological data generated by two regional climate models (RCMs). The RCM of GISS/CCSR (RM3) has achieved a documented and rather unique success in simulating the important mesoscale and synoptic features of the West African monsoon, including daily precipitation rates that are highly correlated in time and place with TRMM daily estimates. The Weather Research and Forecasting model (WRF) is another regional climate model that was developed at NCAR. One WRF version is currently producing operational weather forecasts over North America, but WAM applications have yet to be demonstrated.  The presentation will show mesoscale (0.5° grid) four-dimensional data sets of evolving climate variables over West Africa for the AMMA Special Observing Period-3 (SOP-3), September 2006, from continuous RM3 and WRF nested climate simulations on a grid with 0.5° spacing. Lateral boundary conditions (LBC) to drive RM3 and WRF simulations are taken from NCEP and FNL global analyses, respectively. FNL refers to the 1°x1° observational analyses used for initial conditions in the Global Forecast System (GFS), except that they contain additional late data. Model outputs are compared with relevant meteorological observations from satellite platforms wherever possible, as well as with other observational data collected during the SOP-3. RM3 and WRF 4-D data sets are used to study storm structure and precipitation processes. The data sets provide continuous monitoring of the horizontal and vertical distributions of temperature, moisture, circulation, vorticity and divergence that characterize the storms, as well as all surface energy fluxes. The presentation will show examples of how RCM data sets are used to investigate the characteristics of precipitation processes in the West African region and the adjacent eastern Atlantic. Particular attention is given to the pre-Debbie storm of September 10-13, 2007, especially its transition from continental to maritime structure, before and after crossing the Atlantic coast of West Africa. Differences between RM3 and WRF results are one measure of the uncertainty in the analysis products. In addition, validation against ground-based AMMA data also highlights analysis limitations. The presentation will also show examples of 3- day RCM weather forecasts and perhaps some preliminary hindcasts of September 2006 climate anomalies. LBC for forecast experiments will come from GFS forecasts.

Figs. 1-3 show some characteristics of the structure of RM3 simulated African easterly wave disturbances. Shown is a composite of seven similar events occurring during August and September from the years 1999, 2001, 2002 and 2003. Fig. 1 shows the circulation at two levels, Fig. 2 shows precipitation anomalies and Fig. 3 shows a vertical transect of zonal wind anomalies.   Fig. 1. Composite streamlines of seven easterly wave disturbances simulated by the RM3 driven by reanalysis. Top: 700 mb, Bottom: 925 mb. Note that the low  level vortex  is west of the upper trough.        

Fig. 2. Precipitation anomalies (mm day^-1) for the same composite of seven easterly wave disturbances, simulated by the RM3. Anomalies are relative to August-September 1998-2003 means. Note the enhancement of precipitation SW of the vortex.  
 

Fig. 3. Cross-section of zonal wind anomalies (ms^-1) along 10°W for the same composite of seven easterly wave disturbances, simulated by the RM3. Anomalies are relative to August-September 1998-2003 means. Note the northward tilt with altitude of the cyclonic circulation in the lower levels.

Acknowledgements: This research was supported by National Science Foundation grant ATM-0652518, NASA grant NNX07A193G and by the NASA Climate and Earth Observing System Programs. E. Noble is a NASA Graduate Student Research Program Fellow at NASA/GISS.