CONTEXTE : While the heat wave impacts on public health have been widely addressed in developed countries especially after the intense event over West Europe during summer 2003, no effort has been made so far to detect them and to study their impacts in least developed countries. However, in the Sahel, the climate is much warmer already and adaptation capacities are low. Indeed, a few pioneer climate and epidemiologic analyses show that this problem is emerging. Moreover climate projections in this area indicate that such events could increase in frequency and intensity in the near future. However models display huge temperature and radiative biases in this region. In order to provide robust information on the future evolution of heat waves, it is necessary to first reduce these biases.

Alerte aux Canicules Au Sahel et à leurs Impacts sur la Santé,
Sahelian heat waves and their health impacts warning

ACASIS represents the first pluridisciplinary initiative which aims at setting-up a pre-operational heat wave warning system over West Africa tailored to health risks of the population living in this region, and relies on an pluridisciplinary consortium.

It focusses in particular on two countries, Senegal and Burkina Faso, where national weather services are already involved in developing products dedicated to weather-climate and health relationships, where several populations observatories have been operating since several decades. This allows to characterize the impact of extreme temperatures on morbidity and mortality, to evaluate the perception and vulnerability of the population to these events, and to define tailored bio-meteorological indicators to be used in the warning system.

If the results are conclusive, they could be integrated into an operational system, and further serve as an example and a motivation to set-up such a system in other countries in West Africa.

ACASIS relies on dedicated studies of West African heat waves, from their detection, climatology, predictability and forecast, to the improvement of the underlying physical processes and analyses of their evolution (frequency, intensity...) with climatic change.

ACASIS also aims to characterize the occurrences of temperature extremes in historical data sets, meteorological forecast and climate models. CNRM and LMD CMIP5 and CORDEX simulations will be evaluated and the processes responsible for low-level temperature and radiative biases over this region in the models, including aerosols, will be revisited.

A final major goal of the project is to reduce the biases, still very large large, affecting climate modelling, via improved parametrizations. In that way, it will both contributes to the French DEPHY (Développement et Evaluation des PHYsiques des modèles de climat et de prévision du temps). Project and takes advantage of the results and data from the AMMA project.

Coordinateur Serge Janicot
Correspondant CNRM-GAME Françoise Guichard
Site Internet du projet ACASIS
Financement ANR
Début Jan 2014
Duration 4 ans

  Climat Sahélien : température

Climate change and its rapid emergence in the past decades are a major challenge to public health together with health inequity and weakening of health systems, especially in the Sahel. There, the temperature regularly reaches very high values (35°C is a typical monthly-mean value of daily average temperature in May). Temperature reaches even higher values during specific events such as during the 2010 heat wave (see figure below, where a time series of 24-h mean temperatures acquired during this period is shown).

Furthermore, the temperature has indeed largely increased in the Sahel since 1950, especially in Spring when the temperature is already very high as illustrated below.


The main objective of ACASIS is to demonstrate that a pre-operational heat wave warning system can be set-up over West Africa tailored to health risks of the population living in this region. It is developed with Senegal and Burkina national weather services that are active in this field, with several health and demographic surveillance systems (HDSS) operating for up to several decades.

After constituting qualified databases, we want first to better understand the dynamics of the heat wave and their atmospheric patterns as well as their evolution over the last decades. Then their predictability at short and medium ranges will also interested us, looking at ensembles of multi-models forecasts, including Meteo-France operational deterministic and ensemble forecast systems.

On the longer timescale, we will also evaluate control simulations of the CMIP5/AR5 database in terms of heat wave events, and examine the simulated future evolution of heat waves in climate scenarios and their associated uncertainty. Within these climate models we want to understand the origins of their huge thermodynamic and radiative biases and we will work to reduce them as much as possible.

Epidemiologic studies associated with interviews will be conducted in the health and demographic sites in Senegal and Burkina in order to better evaluate the physiologic and social vulnerability of the African population to high temperature extremes. It will allow us to define tailored bio-meteorological indicators to be used in the warning system.

From these outcomes, we will set-up a demonstration warning system on a “testbed” platform named MISVA. Based on the interviews, and with the setting of several workshops with stakeholders and public institutions, we will be able to provide specific recommendations associated to these warnings. An implementation in the Meteo-France operational system at the end of the project or after might be possible.


The main challenges of this project are linked to :

  • The biases in the modelling of the low-level temperature and in particular the daily minimum of temperature that is important for health issues but particularly difficult to simulate in this region, due to strong vertical gradients near the surface, and uncertainties in the boundary layer turbulence, the distribution of humidity and aerosols, mid-level clouds, and their coupling with radiative processes. However we think being able to reduce significantly these biases.
  • The predictability of environmental conditions linked to heat waves and the skill of models to forecast them from a few days range to future projections. The project will enable to evaluate the available model products and to assess the skills of downscaling processes.
  • The actual links between environmental conditions and local conditions. These links are probably complex. For instance we don’t know whether the impact of moderate but persistent heat waves are more or less or equally severe than those of short but intense heat waves. In order to identify such links a robust monitoring and tracking platform is needed. Other important questions arise : How are these links predictable ? What will be the range of uncertainty when using ensemble of forecasts ? Will this information be useful for stakeholders ? The project is expected to provide valuable information on these issues.

  Description du Projet

The project is organized in 7 Tasks summarized in the schematic below

Task 1 is in charge of the coordination and animation, of the links with the African partners. The meteorological phénomenon ’Heat Wave’ is at the centre of the project.
Task 2 extracts, stores and evaluates the quality of the different datasets necessary for the other Tasks of the project.
Task 3 characterizes the variability of West African heat waves and analyses the associated large-scale atmospheric configurations, it also documents their low-frequency component and assesses their potential predictability, and documents the skill of current state-of-the-art weather forecasting models for predicting such events.
Evaluation of forecasts will also be done over the health and demographic surveillance systems (HDSS) sites used in Task 5 through downscaling procedures.
Task 4 investigates the physics and couplings of the heat wave events, in particular through specific case studies identified in Task 3, and work more generally at the understanding and reduction of the climate model biases critical for heat wave modelling and forecast.
Task 5 will analyse the heat wave events in terms of physiological, societal and environmental vulnerability, and establish as precisely as possible the statistical links between critical variables as temperature and humidity and morbidity/mortality occurrences through data analysis and ensemble of interviews in four HDSS sites, two in Burkina and two in Senegal. Adaptative capacity of the populations of these sites will be evaluate too.
The results from Tasks 2 to 5 will be capitalized into Task 6 in order to set-up a pre-operational warning system tailored to the population needs. information about heat wave occurrences and related health risks. Task 7 supports the dissemination of these results.

 Contribution of CNRM-GAME

Our contribution relies on our :

  • Expertise on West African weather and climate, built from the AMMA project
  • Expertise in atmospheric processes, weather forecast and climate modelling, parametrizations of physical processes.

It involves some dedicated work to :

  • the analyses of observations and the development of physically-based diagnostics ;
  • the utilization of these diagnostics to evaluate models and guide their development ;
  • the simulation of heat-wave events with mesoscale and climate models ;
  • the improvement of the simulation of heat waves and more broadly spring temperatures in the Sahel by weather forecast and climate models ;
  • the assessment of the utilization of these models to heat waves and health related issues.


in France :

  • CNRM-GAME, Toulouse
  • CRC, Dijon
  • LPED, Marseille
  • IPSL(LOCEAN plus LMD and LATMOS), Paris
  • SEDOO, OMP, Toulouse

In Sénégal :

  • ANACIM, Dakar
  • CSE, Dakar
  • LPAOSF, Dakar
  • UGB, Saint Louis

In Burkina Faso :

  • CRSN, Nouna
  • DGM, Ouagadougou
  • ISSP, Ouagadougou