The climate of Chad varies from arid in the north to tropical in the south, with irregular rainy seasons (May to October) and a long dry season (November to April) [1][14][22]. Climate variations influence the dispersion and concentrations of aerosols [14]. The following cities were selected based on Chad’s climate zones.

Table 1. Geographical coordinates for the study area.

Figure 1. Study areas.

These cities are chosen according to climatic zones and are also represented in Figure 1, including isohyets, which provide more detail for their analysis, as well as their geographical characteristics indicated in Table 1. Aerosol-carrying winds from different countries in the Sahara pass through the Sahel, such as the Harmattan, whose corridor is formed by the Tibesti-Ennedi-Lake Chad triangle [23]. It is predominant in winter (December to February) in Chad. It strongly influences the climate of the Sahel and some of the northern regions of Chad. The monsoon, on the other hand, is a system of periodic winds in tropical regions around the world [10]. Coming from the southwest, it brings rainfall and balances temperatures. It has a significant impact on the climate of the Sahel and southern Chad and promotes the development of agriculture and ecosystems [1].

2.2.1. Instrumentation

The instrumentation is based on a set of satellite sensors whose characteristics are summarized in Table 2. These sensors are used by some authors in their aerosol studies. Some of them are also referenced in Table 2.

Table 2. Sensors and their characteristics.

2.2.2. Methodologies for Analysis and Validation of Observations

The validations concern monthly and interannual average data from satellite sensors with the following spatial resolutions: MODIS-Terra (1˚), SeaWIFS (1˚), MISR (0.5˚), OMI (1˚), TOMS (1.0 × 1.25˚), and MERRA-2 reanalysis (0.5 × 0.625˚), measured at wavelengths of 500 nm, 550 nm, and 555 nm. The optical thickness, Ångström exponent, aerosol index, and main types of aerosols were studied in Chad. We analyzed the data using Python. The data were collected in the reflective range [0.4; 2.5 μm], and preferably in the visible and near-infrared range [0.4; 1.0 μm], where aerosols have the strongest optical impact. The choice of MERRA-2 as a reference for the study of the intercomparison of AOD is justified by the fact that it integrates observations from various satellites, thus providing a coherent and homogeneous dataset over an extended period. Furthermore, it uses a general circulation model that simulates atmospheric processes and the interactions between aerosols and climate, allowing for a better estimation of AOD.

2.2.3. Optical Properties of Atmospheric Aerosols

The optical thickness of an aerosol (AOD for aerosol optical depth or AOT for aerosol optical thickness) is a dimensionless quantity that characterizes the transparency of the atmosphere to solar radiation [35]. It corresponds to the amount of light that is absorbed or scattered when light passes through a medium, such as a layer of the atmosphere, a cloud, or any other transparent or translucent material [36]. It is a fundamental and macroscopic parameter that is most commonly used in the characterization of atmospheric aerosols [9]. AOD is determined according to the filters of the available detectors and according to the devices, as there may be more or fewer wavelengths depending on the sensors. However, for a sensor such as the AERONET photometer, there are always at least four: 440, 675, 870, and 1020 nm. Using direct solar irradiance measurements from the ground, the aerosol optical depth (AOD) is calculated using Beer-Lambert’s law (1852), solved for atmospheric boundary conditions:

where I λ is the irradiance measured at ground level; I λ , 0 , extraterrestrial irradiance at wavelength λ ; m is the optical air mass along the line of sight connecting the observation point and the sun; R , is the distance between the Sun and Earth in astronomical units, normalized to variations around the average distance[35][37]. The optical thickness is extracted by eliminating all other extinguishers, including traces of gas and physical processes such as Rayleigh scattering [36]. All these assumptions together allow us to deduce that:

A high optical thickness value of one or more indicates a sky heavily laden with aerosols (or clouds, if it is not possible to distinguish between the optical thickness of aerosols and that of clouds) and therefore not very transparent, while a low optical thickness indicates a clearer sky. For this reason, optical thickness is particularly important near source regions [38]. The spectral dependence of optical thickness provides information on particle size via the Angström exponent [39]. They are linked by the Angström formula (1929):

where A O D λ 1 and A O D λ 2 are optical thicknesses at respective wavelengths λ 1 and λ 2 . α varies very slowly with wavelength, both in the visible and infrared ranges. It is unique in a spectral band [40]. In general, if:

- α < 1 → presence of coarse particles (desert dust, sea salt);

- α ≈ 1 → mixture of particles of various sizes;

- α > 1 → presence of fine particles (urban pollution, smoke, anthropogenic aerosols).

In practical terms, these two parameters are linked by the empirical formula [21][29]:

Optical thickness and the Ångström exponent are considered key parameters in the study of aerosol climatology [10].

In Equation (4), α , A O D λ 1 and A O D λ 2 are defined at wavelengths λ 1 and λ 2 . Subsequently, a linear regression analysis is performed based on Equation (5), where l is the slope of the line and B is the point of intersection between the line and the y-axis[25][29].

The quantities ( l , B ) and the correlation coefficient R of the intercomparison serve as an indication of the local and spatial characteristics of AOD over time [41]. The slope l of the linear regression equation (5) reveals the extent to which the assumed aerosol model represents that of the observed area, while B relates the error caused by surface reflection [42][43]. Thus, the regression equation provides information on the factors affecting the correlation [44] and, in the case of a good correlation, B must be zero and l must tend towards unity [42].

The parameter that distinguishes absorbing aerosols from non-absorbing aerosols in the ultraviolet and visible ranges (e.g., desert dust aerosols, smoke aerosols, and volcanic ash) is the aerosol index ( A I ) [35]. This is a qualitative, unitless measure that expresses relative intensity. Absorbing aerosols can be characterized by the A I , which emphasizes the optical properties, concentration, and altitude of the aerosol. It is used to detect and monitor aerosol plumes on a global scale. It is based on spectral differences between light scattered and absorbed in the atmosphere:

where I m e s u r e d ( λ 2 ) is the measured intensity of backscattered light at a wavelength λ 2 by a satellite sensor and I c a l c u l e d ( λ 1 ) is the calculated intensity of backscattered light at a wavelength λ 1 in pure Rayleigh atmosphere [35]. Wavelengths λ 1 and λ 2 are generally ultraviolet (for example, between 340 and 380 nm). In general, if:

- A I > 0 → presence of absorbing aerosols (desert dust, forest fire smoke, volcanic ash, etc.).

- A I ≈ 0 → clear atmosphere or presence of non-absorbent aerosols.

- A I < 0 → possibility of error or specific atmospheric conditions (example of erroneous corrections).

The analysis focuses on time series of monthly AOD averages from the MERRA-2 reanalysis and MODIS, MISR, OMI, and SeaWiFS satellite sensors. Figure 2 shows that AOD values vary between 0.31 and 0.81 depending on the sensor, for different periods. The MERRA-2 AOD values (0.46 to 0.69) for N’Djamena are more accurate, indicating an underestimation compared to satellite observations. The monthly peaks for MODIS (0.73), SeaWiFS (0.81), OMI (0.74), and MERRA-2 (0.69) are recorded in March or April, while that of MISR (0.75) appears in May. These maxima correspond to the dry season (spring), which is the period when soil warming promotes the uplift of particles. A second peak is observed in October (autumn), reinforcing the idea that particle load is higher outside the rainy season. In the post-monsoon period and in winter, the thinner atmospheric boundary layer traps pollutants near the surface, increasing the AOD. The seasonal variability of AOD reflects changes in sources: Saharan dust, biomass combustion, and other anthropogenic emissions, with a clear contrast between the dry season (Harmattan) and the wet season (monsoon), which is not conducive to the suspension of atmospheric aerosols [45].

Figure 2. Time series of monthly AOD averages for N’Djamena.

The linear regression analysis clearly shows the correlation between the AODs measured by SeaWiFS, MISR, OMI, and MODIS at 500 nm, 550 nm, and 555 nm, respectively. This is illustrated in Figures 3(a)-(d). It indicates a good correlation between MODIS and MERRA-2 measurements with a linear regression coefficient R = 0.62 (Figure 3(d)). Unlike the other sensors, SeaWiFS (R = 0.55), MISR (R = 0.61), and OMI (R = 0.41) are the most representative satellite models for long-term aerosol characterization. The accuracy of airborne sensors is also evident in the slopes of the regression lines: 0.6314, 1.0275, 1.1129, and 0.4853 obtained respectively between SeaWiFS and MERRA-2 (Figure 3(a)), MISR and MERRA-2 (Figure 3(b)), OMI and MERRA-2 (Figure 3(c)), and MODIS and MERRA-2 (Figure 3(d)). However, the smallest intersection point is given by the intercomparison between MISR and MERRA, which is 0.0069. This shows that this sensor is more accurate than the others. However, there is a very large inconsistency in the OMI measurements. In addition to its relatively high correlation coefficient (R = 0.41) and slope (l = 1.113), its intercept constant is virtually zero (B = 0.0919). Thus, the OMI model shows a very large deviation from the MERRA-2 measurements, justifying the overestimation revealed by the time series of monthly AOD averages and the shift in the aerosol peak in spring, mostly in April.

Figure 3. Linear regression curves illustrating the correlations of MODIS, MISR, SeaWIFS, IMO TOMS, and MERRA-2 data.

Each of these sensors uses different retrieval algorithms that may vary in their approach to estimating AOD, leading to different estimations, as the algorithms often rely on assumptions about the composition of aerosols and the properties of the surface, which, if incorrect, can result in errors. Additionally, these sensors often filter out data affected by clouds, which can lead to the underestimation of AOD in cloudy regions, as clouds can interfere with aerosol measurements, making it difficult to distinguish between the signals from aerosols and those from clouds. Consequently, they can lead to discrepancies in their AOD estimates.

3.3.1. Annual Cycles of Optical Thicknesses

Figure 4. MODIS-TERRA monthly averages of AOD for the period: 2000-2024 in Chad.

Chad experiences significant spatial and temporal variability in dust emissions. Figure 4 shows an interannual and monthly climatology of AOD, with a maximum observed in winter (December to March), mainly in the northeast (17˚N - 18˚E), under the influence of the Bodélé epression, the world’s main source of desert dust [2][16]. Between January and February, AOD peaks reflect intense transport of Saharan dust. In spring (April-May), AOD gradually decreases but remains high in the north, indicating persistent activity. In summer (June to September), AOD becomes moderate (0.4 - 0.5) and spreads across the entire country, due to the effect of convective systems and humidity, which limits dust uplift [39]. This period is also marked by emissions from bushfires and biomass combustion in the south. In autumn (October to November), an increase in AOD is observed, linked to the return of dry, arid winds that promote dust resuspension.

3.3.2. Spatial and Seasonal Distribution of Aerosol Shape

The Angström exponent helps identify the nature of aerosols: values below 1 indicate coarse particles (mineral dust), while values above 1 signal fine particles (such as pollutant particles). Figure 5reveals a coexistence of these two types throughout the year, with varied seasonal variations. In winter (DJF), coarse particles dominate in the north, reflecting the Saharan influence of mineral dust from Bodélé. In the south and southwest, fine particles thrive due to several causes, including lower temperatures and increased humidity that favor the formation and persistence of fine particles such as organic aerosols and sulfates. Human activities, like biomass combustion for heating or cooking, can also increase the concentration of fine particles in these regions. Additionally, precipitation can clean the atmosphere of larger particles, resulting in a higher concentration of fine particles during this season in Chad.

In spring (MAM), larger particles are concentrated in the north due to the prevailing influence of the Bodélé depression during this season, which is consistent with the AOD peaks observed in Figure 3. A low concentration of fine particles around Emi Koussi and in the southeast is linked to the influx of dust from predominant desert regions, which can dilute the presence of fine particles often associated with organic aerosols or combustion products from the Gulf of Guinea. In summer (JJA), fine particles are located in the north, center, and east of Chad. They are brought in by the northeastern wind (Harmattan) and the positioning of the monsoon (southwest). Summer is often characterized by higher temperatures and increased relative humidity, favoring the formation of fine aerosols, such as organic aerosols and sulfates. Coarser particles are more present in the south and around Bodélé, coming from strong winds lifting dry sediments and spreading combustion residues northward. In autumn (SON), both the south and north experience a diversity of aerosols linked to the transition between the monsoon and the Harmattan, marked by the movement of the intertropical convergence zone (ITCZ) [16][38]. Hydric erosion followed by rapid drying also promotes sediment mobilization. The low AE values observed in Figure 5 in the north of Chad are closely related to the nature of mineral dust transported from the Bodélé depression, corresponding to larger particles that affect the light diffusion less in the observed wavelengths. This highlights the predominant role of mineral dust in the optical characteristics of aerosols in this region.

Figure 5. Spatial and seasonal distribution of Angström exponent by MODIS-TERRA for the period: 2000-2024 in Chad.

Figure 6 shows the time series of interannual and monthly AOD averages obtained by MODIS-Terra inverted between 2000 and 2024 for Chad in general and for N’Djamena in particular. These time series are compared with those produced by the MERRA-2 reanalysis of N’Djamena. MERRA-2 (blue curve) shows the maximum aerosol concentration in spring, clearly highlighted by the AOD peaks in April. The same is true for MODIS (red curve), which is in good agreement with the AOD maxima, also in spring (April). However, in addition to spring, MODIS reproduces the AOD peak in summer (June). This evolution of atmospheric particles is corroborated by the seasonality of aerosols, as illustrated in Figure 5 and Figure 6. Aerosols can be distributed vertically and horizontally. Studies conducted by [15][24] show that the vertical distribution of aerosols allows the aerosol layer to be located in summer between 3 and 6 km in West Africa. Thus, the presence of particles in the upper troposphere could well explain the observation of the summer AOD peak at MODIS, which has a much wider field of view than MERRA-2, whose data-providing device may be limited by precipitation and cloud cover. The particles observed in summer are mainly mineral dust from the long-range transport of aerosols contained in the Saharan air layer, and also due to convective systems. On the other hand, the AOD minima in August (blue curve), November (red curve), and December (green curve) are probably due to the elimination of particles by sedimentation, following water recovery or atmospheric leaching due to heavy precipitation. However, the maximum aerosol levels observed in Chad from October to December are probably linked to the transition between the monsoon and the Harmattan, which is very turbulent during this period [15].

Figure 6. Representation of the annual AOD cycle for Chad (green) and for N’Djamena observed by MERRA-2 (blue) and MODIS (red).

Figure 7(a) illustrates the monthly averages of aerosol optical depth (AOD) at 550 nm for several cities in Chad. Maximum concentrations occur in summer, particularly in August, except in Pala and N’Djamena, where peaks occur in spring (April). This difference can be explained by the persistence of the Harmattan, which is active from winter to spring, before the transition to the monsoon. This dynamic reflects the spatial and temporal variability of aerosols in Africa. The maximum AOD values in Mongo (0.47) and Salal (0.53) can be explained by the combined influence of the monsoon and the Harmattan, which transports Saharan dust [21][7]. In N’Djamena (0.62), Pala (0.52), and Goré (0.50), high AOD values are linked to both Bodélé dust (0.47) and biomass burning in the Gulf of Guinea (October to May). Local emissions from human activities and long-range transport contribute to elevated AOD levels in large urban areas such as N’Djamena (AOD = 0.62), Ouagadougou (AOD = 0.64) [21], and Banizoumbou (AOD = 0.62) [30]. Figure 7(b) highlights the predominance of aerosols in spring and summer, while the Angström exponent maxima occur in autumn (September) and winter (December), and the minima occur in summer (June), in connection with the action of the monsoon, which stirs up sediments. In August, fine desert particles dominate due to long-range transport. In autumn, peaks in the Angström exponent at Pala and Goré reflect the impact of aerosols from the transition between the monsoon and the Harmattan, characterized by intense convective systems and easterly winds that generate local dust emissions.

Figure 7. Interannual monthly averages of optical depth (AOD) (a) and Angström exponent (AE); (b) were observed at Bodélé (orange), Salal (pink), Mongo (dark blue), N’Djamena (green), Pala (red), and Goré (purple).

To analyze the influence of large-scale meteorological factors on the mobilization of absorbing aerosols, two satellite datasets were examined: OMI (2004-2024) and TOMS-EP (1996-2005), as shown in Figure 8. These sensors measure the aerosol index (AI), an indicator of the presence of absorbing aerosols in the atmosphere. TOMS-AI evaluates the difference in ultraviolet backscattering between an atmosphere containing aerosols and a pure Rayleigh atmosphere [16], allowing the distinction between absorbing aerosols, such as soot (positive AI), and non-absorbing aerosols, such as sulfates (negative AI) [46]. An AI value of zero indicates the absence of aerosols in the atmosphere. Both OMI-AI and TOMS-AI are used to detect large dust plumes from deserts, such as those crossing the Sahel, or smoke emissions from forest and bush fires. The OMI-AI and TOMS-AI indices reach their respective peaks in May (1.2 and 2.4). Chad records particularly high concentrations of absorbing aerosols in spring (March to May), resulting from the superposition of several natural and anthropogenic sources. Among the natural sources, the Bodélé depression, located in northeastern Chad, is one of the world’s main sources of desert dust emissions [17][31]. During this period, easterly and northeasterly winds (the Harmattan) reach maximum intensity and transport large quantities of mineral particles rich in iron oxides, which are highly absorbing in the visible and infrared ranges [47].

At the same time, the dry season at the end of winter and during spring corresponds to an intensification of bush and savanna fires in the Sahelian and Sudanian zones. These biomass fires release significant amounts of black carbon (soot) and organic carbon, whose optical properties enhance the radiative absorption of aerosols [48][49].

Finally, local urban sources, particularly domestic wood and charcoal burning as well as road traffic, contribute significantly to the atmospheric load of absorbing aerosols [50]. Spring weather conditions, characterized by high temperatures and a shallow atmospheric boundary layer, limit vertical dispersion and promote the accumulation of particles over the Chadian atmosphere. The minimum AI values, observed in September (0.5) and November (1.1), are caused by factors such as rain washout, reduced dust emissions due to wet soils and vegetation, changes in air trajectories (supply of cleaner or moister air), and optical effects related to humidity (increased scattering, relative decrease in absorption).

Figure 8. Variability of monthly averages of OMI-AI and TOMS-AI in Chad for the periods 2004-2024 and 1996-2005.

3.6.1. Seasonal Distribution of AE based on AOD

Determining the type of aerosols essentially requires knowledge of their optical and physical properties, which depend on both wavelength and solar and terrestrial radiation. The combined use of the descriptive properties of aerosol load or optical depth (AOD) and Angström exponent (AE) is the most widespread and widely used technique for classifying various types of aerosols [35].

Figure 9shows the seasonal variations in AE as a function of AOD in N’Djamena. For all seasons, the highest values of the monthly averages of the Angström exponent correspond to low values of optical depth. The monthly average data are from MODIS-TERRA. These two parameters provide information on the presence and class of aerosols (coarse mode, accumulation mode, and fine mode). AE values are specific to particle types: AE ≈ 0 for large particles such as desert dust and AE > 1 for fine particles such as those produced by biomass combustion or industrial pollution.

The contribution of accumulation-mode aerosols is particularly significant in summer, fall, and spring, suggesting their predominance in this region during these seasons. The substantial presence of accumulation-mode particles in the atmospheric column above a location is associated with anthropogenic sources, including biomass burning, transportation, and agricultural activities, as well as the presence of the fine mode, which is slightly observed in spring and winter [11]. Throughout all seasons, the atmospheric aerosol content consists of a mixture of both accumulation and coarse modes. Accumulation-mode aerosols are especially dominant in N’Djamena, followed by coarse-mode aerosols. This finding is consistent with the study on lithometeors in N’Djamena conducted by [51]. The predominance of these modes can be explained by the city’s geographical location and locally generated wind patterns, which transport particles from the Sahara Desert.

Figure 9. Seasonal distribution of AE based on AOD in N’Djamena for the period 2000-2024.

3.6.2. Classification of the Main Types of Atmospheric Aerosols in N’Djamena

The main objective is to analyze and evaluate the percentage of atmospheric aerosol types and to identify the most predominant type of aerosol during all seasons in N’Djamena. We followed the methodology as in the study by Ali M. Al-Shalihi (2018), which focused on the characterization of aerosol types based on their optical properties in the city of Baghdad, Iraq. Unlike that study, this classification was based on the concentrations of different aerosol types. To this end, we analyzed the concentrations of these main types of atmospheric aerosols derived from MERRA-2, using the site: https://giovanni.gsfc.nasa.gov/giovanni/. This allowed us to obtain results in percentages, as these different types of aerosols have varying climatic impacts depending on their concentrations. For example, dust and sulfates can have a cooling effect by reflecting sunlight, while carbonaceous particles (like black carbon) tend to warm the atmosphere. Distinct modeling allows for the evaluation of specific contributions to climate feedback.

Figure 10shows the percentage contributions of the main aerosol types over N’Djamena. The results indicate that desert dust is the dominant component in this region. It should be noted that aerosol types are influenced by synoptically variable sources, the strength of sinks, transport pathways, local meteorological conditions, and vertical mixing processes. In addition, particle growth mechanisms such as humidification and hygroscopic growth play an important role in modifying aerosol characteristics [52].

The aerosols affecting N’Djamena originate mainly from the Sahara and Sahel regions. Other sources include agricultural activities such as plowing, construction work, and unpaved roads. However, anthropogenic emissions from traffic, industry, and transportation contribute significantly, especially in capital cities. For example, in Baghdad (Iraq), these activities account for about 38% of total emissions [35].

In winter, their contribution rises to 89.1%. Organic aerosols, originating from biomass combustion (wood for heating) and soil erosion, account for only 5.5%, while sulfate aerosols contribute 3.6%, mainly from volcanic eruptions in East Africa [13] and other regions. Additional minor sources include the oxidation of dimethyl sulfide (DMS), savanna fires, fossil fuel combustion, industrial emissions, agricultural activities, and urban pollution. Carbonaceous aerosols and sea salts, representing small fractions (1.7% and 0.3%, respectively), are primarily produced by agricultural burning (stubble and wood) and industrial processes, as well as by natural atmospheric mechanisms. Sea salts are also transported southward by the Harmattan, which carries dust and salts from the Sahara Desert (Figure 10(a)).

In spring, dust aerosols contribute more significantly (93.5%) than during other seasons. This increase results from a combination of climatic factors, such as strong winds during the dry season, including the Harmattan, and anthropogenic factors such as farming, slash-and-burn practices, overgrazing, and deforestation, which enhance soil erosion, particularly in uncultivated areas. These human activities, combined with drought conditions, release large amounts of particles into the atmosphere, which are subsequently transported by winds during this period. Sulfate aerosols, which make a smaller contribution (2.8%), originate from volcanic eruptions, bush fires, and the combustion of fossil fuels in vehicles and power plants, as well as from industrial emissions. Once emitted, these particles can undergo chemical transformations to form secondary sulfate aerosols. Organic aerosols, with a similar contribution (2.5%), are mainly of natural origin, produced by wind erosion of dry soils and biomass burning, including vegetation fires and agricultural combustion. Additional sources may include secondary organic aerosols formed through chemical reactions in the atmosphere. The relatively low contributions of carbonaceous aerosols (0.9%) and sea salts (0.3%) can be attributed to changes in wind patterns that limit their transport from sources such as desert sand and burned biomass to Chad, reduced biomass burning activity at the beginning of the dry season, and increased rainfall in early spring that removes these aerosols from the atmosphere through wet deposition (Figure 10(b)).

In summer, the contribution of dust aerosols decreases to 89.1% due to increased rainfall, which moistens the soil and limits particle suspension, as well as changes in wind direction and intensity, which reduce the transport of dust from the Sahara. Additional factors, such as increased vegetation cover due to higher humidity and a reduction in surface erosion sources, may also contribute to this decrease (Figure 10(c)). Sulfate aerosols contribute 10%, primarily resulting from the combustion of fossil fuels (coal and oil), which release sulfur dioxide that is subsequently converted into sulfates, and, to a lesser extent, from soil erosion prevalent during this season. Organic aerosols, originating from vegetation decomposition, biogenic emissions, biomass burning (bush and agricultural fires), motorcycle traffic, and fossil fuel combustion, contribute only 4.4%. Finally, carbonaceous aerosols (1.4%) and sea salts (1.1%), whose contributions have increased, are largely associated with fossil fuel combustion and deforestation. Sea salts are also linked to soil salinization caused by climate change (erosion, reduced water availability) and the evaporation of saline waters. In Chad, summer conditions exacerbate these phenomena through higher temperatures and increased evaporation, enhancing the concentration of these substances in the environment, particularly the salinization of soil and surface water (Figure 10(c)).

Figure 10.Percentage contributions of different types of aerosols in N’Djamena for the period 1994-2024 at MERRA-2. (a) Winter; (b) Spring; (c) Summer; (d) Fall.

With the arrival of fall, the contribution of dust aerosols decreases to 86.8%. This reduction is primarily due to the onset of monsoons and rainfall, which clean the atmosphere and stabilize the soil. The intertropical front shifts southward, slowing the transport of desert dust. In addition, Sahelian sources become less active as soils become wetter. Sulfate aerosols contribute 7.3%, a decrease explained by reduced pollution sources such as lower industrial emissions and biomass burning, more efficient atmospheric cleansing due to changes in atmospheric chemistry and precipitation, and variations in wind patterns that either transport aerosols away or promote their deposition. The contributions of organic aerosols (3.7%), carbonaceous aerosols (1.2%), and sea salts (0.9%) are mainly attributable to the reduction in dust sources and in biogenic and anthropogenic emissions. This is particularly evident at the end of the rainy season, when surface winds decrease at the beginning of the dry season and agricultural activities change. Autumn in Chad is therefore characterized by reduced wind erosion of soils and lower dust concentrations, while variations in temperature and humidity also influence natural emissions (Figure 10(d)).