Saharan - West African Monsoon Multi-scale Analysis (SWAMMA)

The West African Monsoon is critical to the livelihoods of millions, but remains a major challenge for weather and climate prediction.
The monsoon provides moisture for large systems of thunderstorms that provide the annual rainfall in the Sahel. The monsoon is driven by the summertime heating over the continent. The Sahara exhibits the highest summertime temperatures and it is known that the monsoon is sensitive to the heating in the Sahara and the Sahel. The heating in the Sahara is controlled by dust, cloud and ventilation by colder air from the Sahara's margins. It is known that models exhibit major differences in this region, which is for example important for monsoon onset, but historically there has been an absence of data with which to evaluate models.

Recent research has shown that the monsoon, dust and the storm-systems form a fully coupled system and that global models exhibit major biases in the couplings within this system. Operational models fail to capture the organisation, location and timing of the monsoon rains and this leads to incorrect coupling between the storm systems and the monsoon. Operational models also fail to represent the haboob dust storms generated by the storm systems, which are likely the main driver of dust emission in the summer. This means that even operational models which include dust fail to accurately capture the impacts of the dust on the meteorology, since a major dust emission mechanism is missing in the models. Recent research has also shown the existence of larger dust particles over the Sahara than previously observed, which will affect heating rates, feeding back on to the meteorology.

SWAMMA will bring together a recently-developed model that can for the first time capture these couplings with recently-acquired data from this highly data-sparse region (collected during the NERC-funded AMMA and Fennec projects). SWAMMA will therefore combine for the first time a model that should be able to represent the key features of the summertime energy budget of the monsoon system with the first data sufficient to evaluate such a model.

SWAMMA will evaluate the new model, quantifying the winds that generate dust uplift in the Sahara and Sahel. It will use the new data to analyse the impacts of variations in the sizes of dust particles. SWAMMA will evaluate the energy budgets of the monsoon system using both models and observations from AMMA and Fennec. SWAMMA will quantify how the coupled processes of the monsoon system respond to the seasonal evolution of the monsoon's boundary conditions to control the seasonal stability of the monsoon rainfall and its variability. By bringing together the recently-acquired data and the new model with existing weather and climate models the project will evaluate how process errors in operational models lead to systematic errors in predictions, both over West Africa and globally. The project will provide metrics for the processes that must be captured by operational models for them to capture the key processes and couplings in the West African monsoon system.

SWAMMA research will benefit operational weather and climate prediction centres worldwide, indirectly benefiting all users of their predictions, most obviously in the West African Monsoon (WAM) region, but also globally. Errors in tropical convection are known to be a major source of error in all global models, particularly in monsoon regions. Dust is a significant source of error in weather prediction and an important Earth-system component. Dust is increasingly being included in operational weather and climate models. SWAMMA will therefore benefit Earth-system science, as well as weather and climate.

Forecasting the annual monsoon rains, and particularly their onset or breaks during the season, is the most urgent requirement of weather prediction systems for the region. SWAMMA will quantify the balance of processes that determine the WAM state as it evolves through the seasonal cycle. Studying the three years with sufficient data will allow an assessment of inter-annual variability. SWAMMA will generate metrics for the processes that global models must represent to capture the evolution of the monsoon. SWAMMA will use the Met Office Unified Model (UM), but the process coupling errors identified in the UM are common across global models so the lessons learnt will be applicable to all models.

A strength of SWAMMA is its seamless modelling framework. The UM is both an operational weather and climate model. It is known that errors in low-resolution forecast models correspond closely to errors in climate models, since they are often caused by the same 'fast processes'. SWAMMA will confront new model runs, with a range of resolutions and complexities, with unique field-campaign data, which will for the first time allow detailed model evaluation in the Sahel and Sahara. We will use the process errors diagnosed in the new UM simulations run within SWAMMA to diagnose the sources of error in operational UM weather and climate simulations and determine the global impact of errors in the WAM system. This will provide a robust chain of assessment from observations, through process models to weather and climate predictions. This will inform priorities for operational model development and interpretation of model predictions.

SWAMMA complements the Met Office strategy for developing its convection and dust schemes. Communication with the Met Office will be facilitated by the formal partnerships that exist between the Met Office, Leeds and Reading (with Met Office staff based at both Leeds and Reading) and annual visits of SWAMMA staff to the Met Office. This will allow exchange with scientists working on model development (convection, dynamics and dust) and weather, seasonal, climate and Earth-system predictions.

The main funded phase of AMMA has finished, but AMMA continues to provide international coordination of WAM science. Prof Doug Parker was PI of UK AMMA and SWAMMA will make full use of AMMA to maximise its impact (e.g. attending the AMMA conference). Prof Doug Parker is leading the production of the first WAM forecasters handbook, which will further facilitate exchanges with forecasters and scientists based in West Africa.

Communication with other operational centres will take place through journal papers, conferences and direct communication. John Marsham and Cathryn Birch (a Met Office employee at Leeds) regularly attend (and have given recently plenary talks at, or convened) GASS (Global Atmospheric System Studies), WGNE (Working Group on Numerical Experimentation) and JWCRP (Joint Weather and Climate Research Programme) meetings, which are well attended by scientists in the operational communities. Their ongoing involvement in these fora will facilitate communication with prediction centres worldwide.