Spatial air pollution modelling for a West-African town.

Area of Science:

• Environmental Epidemiology
• Geographic Information Systems (GIS)
• Air Quality Modelling

Background:

• Land use regression (LUR) is widely used for air pollution assessment in Europe and North America.
• Limited application of LUR models exists in African settings, necessitating research into their adaptability.
• Urban characteristics and data availability in Africa differ significantly from other continents.

Purpose of the Study:

• To evaluate the applicability and effectiveness of LUR modelling for assessing air pollution in a West African town.
• To develop and validate a parsimonious LUR model for nitrogen dioxide (NO2) in Kaédi.
• To demonstrate the potential of LUR for cost-effective air pollution mapping in African urban environments.

Main Methods:

• Developed a regression model using 48-hour nitrogen dioxide (NO2) concentrations from 40 monitoring sites.
• Incorporated geographic information system (GIS) variables, including road networks and land use characteristics.
• Validated the model using leave-one-out cross-validation and assessed parameter robustness.

Main Results:

• The LUR model explained 68% of the variability in 48-hour NO2 concentrations in Kaédi.
• Road variables and settlement land use characteristics were identified as significant predictors.
• The model demonstrated moderate performance and robust parameters upon internal validation.

Conclusions:

• Land use regression (LUR) modelling is a valid and cost-effective method for air pollution assessment in African towns.
• The developed model provides a spatial map of NO2, useful for understanding air pollution distribution.
• Adoption of LUR by authorities could enhance local air pollution burden assessment and support policy development.