ECMWF helps to probe impact of aerosols in West Africa

Angela Benedetti, Frédéric Vitart (both ECMWF), Peter Knippertz (Karlsruhe Institute of Technology, Germany)


Southern West Africa (SWA) is becoming a focus of interest for climate change and air quality studies due to its rapid population growth and economic expansion. Much of this population will be concentrated in urban centres, and atmospheric emissions of chemical compounds and aerosols from industrial, transport, and energy sectors in the area are likely to increase. These changes are expected to affect human health, ecosystems and biodiversity, and regional climate.

ECMWF is a partner in the DACCIWA (Dynamics–Aerosol–Chemistry–Cloud Interactions in West Africa) project, which aims to provide a comprehensive assessment of these impacts and to provide scientific support for the region’s sustainable development. This EU-funded project runs from 1 December 2013 until 30 November 2018. It involves 16 partners (universities, research institutes, and operational weather and climate centres) from European and West African countries and is co-ordinated by the Karlsruhe Institute of Technology (KIT). DACCIWA builds on a number of past and existing projects and networks in West Africa, including the African Monsoon Multidisciplinary Analysis (AMMA).

Field campaign

A lack of observations is a major obstacle to achieving DACCIWA’s objectives. To alleviate this, DACCIWA will carry out a major field campaign in SWA during June and July 2016, involving co-ordinated flights with three research aircraft and a wide range of surface-based instrumentation at Kumasi, Ghana; Savé, Benin; and Ile-Ife, Nigeria. June–July marks the onset of the West African Monsoon. The increased cloudiness this brings is susceptible to aerosol effects and important for radiation.

ECMWF has an interest in studying West Africa since model errors are particularly evident in the tropics and have implications for predictability in the mid-latitudes. ECMWF actively participated in AMMA by assessing how additional observations can improve analyses and forecasts in the region. It was found that extra soundings had a significant impact on the ECMWF analysis, particularly for low-level temperature over the Sahel and the structure of the African easterly jet. However, the impact on forecasts disappeared after 24 hours. Large model biases in boundary layer temperature over the northern and eastern Sahel were found, consistent with well-known model biases in cloud, rainfall and radiation.

Impact of aerosols

ECMWF’s contribution to DACCIWA will include providing forecasts for the June–July field campaign in collaboration with the Copernicus Atmosphere Monitoring Service (CAMS) campaign support team (Luke Jones and Miha Razinger). ECMWF will also design and run ad hoc experiments for West Africa. These will include medium-range runs with interactive aerosols using the Composition-Integrated Forecasting System (C-IFS) and seasonal runs based on the operational monthly ensemble forecasts. The experiments will serve to assess the importance of aerosols for the West African Monsoon onset and development. Work towards this goal has already started and has led to a multi-year run (2003–2015) with interactive aerosols, which is currently being compared with the operational monthly system. Preliminary results show that including aerosols in the model has a positive impact on weather predictions by reducing the wet bias in precipitation over land in SWA. Forecasts for other areas, such as India and the West Atlantic, also appear to improve in the interactive aerosol run. More cases are needed to assess the impact of including aerosols on the skill of the monthly forecast and to fully evaluate aerosol fields over seasonal scales.

For more details on DACCIWA, visit

Dacciwa field campaign
DACCIWA field campaign. Black stars mark the three DACCIWA supersites at Kumasi, Ghana; Savé, Benin; and Ile-Ife, Nigeria. Red dots mark synoptic weather stations. The size of the dots is proportional to the available number of reports in the WMO Global Telecommunication System from 1998 to 2012. (Source: P. Knippertz et al., Bull. Am. Meteorol. Soc., September 2015, 1451–1460)

Reduced rainfall bias
Reduced rainfall bias. The charts show the bias in rainfall forecasts for the month of June in the control run (top) and the interactive aerosol run (bottom). The bias is calculated over the period 2003-2015 with respect to the ERA-Interim climate reanalysis.