How good are recent satellite based sensors and models for monitoring of air quality in Africa

<p>Satellite observations have become one of the critical tools employed for Earth Observations. The change in tropospheric and stratospheric composition is routinely monitored with improved accuracy using satellite-based sensors in limb, nadir or limb-occultation viewing geometry. However, most of these observations have been constrained due to inherent limitation of limb viewing geometry to observe close to surface processes or poor vertical resolution of nadir observations as well as the complexity of the processes close to surface. As a result, capability of recent satellites have been improved in terms of radiometric, spectral, and spatial resolutions. Moreover,  the radiative transfer models and inverse model algorithms have been improved to overcome these challenges. The performance of these models has been validated using in-situ observations globally. However, the assessment over Africa is limited due to sparse ground observations in Africa. In this study, we assess the performance of the recent satellite observations and model simulations of NO<sub>2</sub>, CO, HCHO, aerosol optical depth (AOD) and dust AOD in capturing the large scale climatology of the criterion pollutants over Africa.</p> <p> </p> <p>Ozone monitoring instrument (OMI) have captured enhanced tropospheric NO<sub>2</sub> column density during December and January months along the location of the ITCZ  during northern summer suggesting that the source of the emission is biomass burning. The MODIS satellite showed enhanced fire power from burning of savanna and C4 grass during December-January period confirming the OMI observations. OMI has also captured enhanced HCHO during summer (July) and winter (December to April) over the same areas. This indicates that source of the two peaks are biogenic and biomass burning respectively. In addition, The Angstrom Exponent (AE) derived from POLDER-PARSOL satellite based on GRASP algorithm indicates AE that exceeded 1.2 implying that the aerosol is generated mostly from biomass burning. Measurement of pollution in the troposphere (MOPITT) instrument observed tropospheric CO VMR which peaks in November to February period over these areas. This seasonal pattern is also simulated by MERRA-2 model. The spatial consistence between MOPITT CO VMR and MERRA-2 is strong during biomass burning season in winter and relatively weak during the rest of the season over Africa. IASI-METOP-A dust AOD (DAOD) and MERRA-2 reanalysis model DAOD are in good agreement over Sahel and Middle East region. Stronger correlation (spatial consistence) is noted during dry months from October to March) than relatively wet months from April to September. DAOD from IASI-METEOP-A and AOD MODIS satellites can also be used in conjunction with surface temperature, relative humidity and wind speed to develop non-linear multiple regression models based on limited in-situ observations of PM2.5 and PM10 over the region to determine aerosol load close to surface. The performance of such models is found to depend on boundary layer height which is a function of atmospheric stability that controls transport of air pollutants.</p>