Evaluation of the WAMME model surface fluxes using results from the AMMA land-surface model intercomparison project
- First Online:
- Cite this article as:
- Boone, A.A., Poccard-Leclercq, I., Xue, Y. et al. Clim Dyn (2010) 35: 127. doi:10.1007/s00382-009-0653-1
- 108 Downloads
The West African monsoon (WAM) circulation and intensity have been shown to be influenced by the land surface in numerous numerical studies using regional scale and global scale atmospheric climate models (RCMs and GCMs, respectively) over the last several decades. The atmosphere–land surface interactions are modulated by the magnitude of the north–south gradient of the low level moist static energy, which is highly correlated with the steep latitudinal gradients of the vegetation characteristics and coverage, land use, and soil properties over this zone. The African Multidisciplinary Monsoon Analysis (AMMA) has organised comprehensive activities in data collection and modelling to further investigate the significance land–atmosphere feedbacks. Surface energy fluxes simulated by an ensemble of land surface models from AMMA Land-surface Model Intercomparison Project (ALMIP) have been used as a proxy for the best estimate of the “real world” values in order to evaluate GCM and RCM simulations under the auspices of the West African Monsoon Modelling Experiment (WAMME) project, since such large-scale observations do not exist. The ALMIP models have been forced in off-line mode using forcing based on a mixture of satellite, observational, and numerical weather prediction data. The ALMIP models were found to agree well over the region where land–atmosphere coupling is deemed to be most important (notably the Sahel), with a high signal to noise ratio (generally from 0.7 to 0.9) in the ensemble and a inter-model coefficient of variation between 5 and 15%. Most of the WAMME models simulated spatially averaged net radiation values over West Africa which were consistent with the ALMIP estimates, however, the partitioning of this energy between sensible and latent heat fluxes was significantly different: WAMME models tended to simulate larger (by nearly a factor of two) monthly latent heat fluxes than ALMIP. This results due to a positive precipitation bias in the WAMME models and a northward displacement of the monsoon in most of the GCMs and RCMs. Another key feature not found in the WAMME models is peak seasonal latent heat fluxes during the monsoon retreat (approximately a month after the peak precipitation rates) from soil water stores. This is likely related to the WAMME northward bias of the latent heat flux gradient during the WAM onset.