The current study was undertaken in the EAC, a consortium of countries forming one of the developing communities of the world. It is a regional intergovernmental organization of six (6) partner states (Figure 1) in the African Great Lakes region of Eastern Africa namely: Uganda, Kenya, Rwanda, Tanzania, Burundi and South Sudan. Its headquarters is in Arusha, Tanzania [9].

As outliers with Burundi among the East African responses to COVID-19 pandemic, Tanzania (Dar es Salaam) was considered in this study as a reference rather than a case study. The two countries (Uganda and Kenya) have been chosen because the former enforced a complete lockdown earlier while the latter started with a partial lockdown followed by a total lockdown. The major determining factors for the choice of these countries were because they were the only countries in the EAC for which nitrogen dioxide (NO₂) emissions satellite data were available from National Aeronautics and Space Administration (NASA) database.

This study used satellite monitoring measurements for air pollutants recorded from the onset of COVID-19 in January 2020 through partial to complete COVID-19 lockdowns until December 2020 in the EAC. The baseline period used were the years 2015-2019 for nitrogen dioxide emissions and 2019 for the absorption aerosol index. We adopted this approach because the faculty of satellite devices in remote sensing of the lower troposphere has improved considerably over the years. Aerosol factors, for instance, are verifiable on spatial scales of a few kilometers using the space-borne radiometers [10]. Further, nitrogen dioxide along with other trace gases can be effectively identified on urban levels using spectrometers [10]. Unlike field monitoring stations that forecasts only vacillations in air quality or assess emissions at discrete points, satellite monitoring offsets this drawback with its faculty to model horizontal transportation of air pollutants for both long- and short periods [11]. This singular aspect, enhances greatly, synoptic and geospatial knowledge of ground-based air quality data [11]. In this context, data from satellite monitoring can be harnessed for establishing regions with the highest concentrations or furnishing warnings of upcoming air quality events such as wildfires and dust storms. Another advantage conferred by satellite systems is supplying air quality information for regions devoid of surface-based monitors [11].

In this study, changes in NO₂ emissions before and during the COVID-19 pandemic year of 2020 were monitored by following Aura satellite data [12]. NASA launched Aura satellite on July 15^th 2004. It is equipped with an ozone monitoring instrument (OMI) to monitor the changes in air quality for various regions worldwide [13]. The OMI satellite instruments precisely detect air pollutants such as NO₂, ozone, sulfur dioxide, carbon monoxide and methane [14]. The tools compute backscattered radiations from the sun in a broad spectral range from ultraviolet (UV) to infrared wavelengths. Advanced retrieval algorithms are thereafter applied to transform the measured radiation to pollutant concentrations, such as a tropospheric column density of NO₂ [15].

The absorbing aerosol index (AAI) data for 2019 and 2020 were obtained from Tropospheric Emission Monitoring Internet Service [16] of the European Space Agency (ESA).

Data retrieved from satellite monitoring assessments during the COVID-19 pandemic lockdown were compared with both statutory and international air quality compliance standards.

Data for NO₂ emissions were retrieved from OMI for January 2020 to December 2020 and the baseline years (2015-2019). Data for the targeted months (March, April and May) in 2020 represents the peak of the pandemic with total or partial lockdowns in Uganda and Kenya, while Tanzania had no such containment measure in the same period. Comparison of the data for each month in 2020 with those of the previous five years (2015 to 2019) aimed at logically identifying whether the implemented containment measures during COVID-19 outbreak had an effect on NO₂ emissions in Uganda and Kenya or not. In Figures 2-4, the darker the color, the higher NO₂ concentration/emission is [15] [17].

Data retrieved indicated that 2020 NO₂ emissions in Kampala and Nairobi were lower than the average levels in the respective cities during the same months in 2015-2019. It is clearly observed that the orange colors in Figure 2 and Figure 3 are darker for 2020 than for 2015-2019, indicating there was a reduction in NO₂ emissions in 2020 during the COVID-19 pandemic.

A percentage difference was calculated inside a 1˚ × 1˚ box over Kampala (Figure 2). In January 2020, NO₂ emissions in Kampala were 16% lower than the baseline. This could be due to the reduced movements during the festive season. In February 2020, NO₂ emissions were only 9% lower than the baseline in Kampala. This is because during February, most schools and tertiary institutions had reopened, which translates into increased vehicular movements. By March 18^th 2020 when total lockdown along with other containment strategies were adopted in Uganda to reduce the spread of COVID-19 [18] [19] [20], NO₂ emissions over Kampala dropped by about 2% compared with the baseline period (Figure 2(c)). From April 1^st 2020 to May 1^st 2020, NO₂ emission over Kampala reduced by 12% and 3%, respectively. The lower percentage reduction in May 2020 could be because the government eased some restrictions despite maintaining lockdown in late April and early May [21] [22]. This also explains the lower percentage reductions in NO₂ emissions (10%, 3%, 17%, 0%, 6% and 8%) lower than the baseline recorded between June and November 2020, as the government allowed vehicular movements with restricted number of passengers (carrying half their capacity), reduced curfew hours and removed total lockdown, and allowed resumption of work and commercial activities to reduce risks of economic recession [23] [24]. In December 2020 (Figure 2(l)), NO₂ emissions in Kampala was 8% greater than the baseline. This could be in addition to the foregoing reasons, increased movements in preparation for the festive season (Christmas celebration on 25^th December 2020 and New Year celebration on 1^st January 2021) as well as increased campaign activities in preparation for elections which occurred from 14^th January 2021 up to early February 2021.

For Nairobi (Kenya), a percentage difference was calculated inside a 1˚ × 1˚ box over it. As depicted in Figure 3, NO₂ emissions were 18% greater than the baseline in January 2020 but this increased to 20% lower than the baseline in

February. This could be due to reduced industrial production post-festive season of 2019. On March 1^st 2020, there was no net increase in NO₂ emissions when compared to 2015-2019 baseline. On 6^th April 2020 when a partial lockdown was considered, NO₂ emissions were 8% lower than the baseline. This could have been due to decreased industrial production and vehicular movements. For May, July, August, September, October, November and December, NO₂ emissions in Nairobi were 31%, 27%, 24%, 13%, 19%, 11% and 7% lower than the baseline. On 1^st June 2020, NO₂ emissions were 13% greater than the baseline. The lower reductions in percentage emissions of NO₂ from May to December 2020 could be due to the gradual easing of COVID-19 restrictions by the government, including allowing resumption of work, industrial activities and public transport i.e. vehicular movements [25]. Within Nairobi metropolitan area for example, movement restrictions in the Eastleigh area were lifted [26]. The November-December lower percentage reductions in NO₂ emissions could also be explained by increased industrial production and vehicular movements ahead of the festive season (Christmas celebrations on 25^th December 2020 and the new year celebration of 1^st January 2021).

Interestingly, the observed reductions in NO₂ emissions in 2020 over Kampala and Nairobi cities have not been the case with Tanzania where there was no lockdown instituted. As seen in Figure 4(a), Dar es Salaam had 4% greater NO₂ emissions than the baseline on January 1^st 2020 which reduced to 4% lower than the baseline in February 2020 (Figure 4(b)). In March when COVID-19 became a reality in the EAC, NO₂ emissions dropped by 9% lower than the baseline (Figure 4(c)). This could be because the country initially tried to follow WHO guidelines in containing the virus and reduced vehicular traffic in its coastal city of Dar es Salaam [27] [28]. In April 2020, there was no net difference between the NO₂ emissions in the city as compared to the 2015-2019 baseline, but this increased to 21% lower than the baseline in May 2020. When the president declared the country “coronavirus-free” in June 2020 and allowed free movements and operation of industries [29] [30], NO₂ emissions increased by 15% and 4% greater than the baseline for June and July 2020 (Figure 4(f) and Figure 4(g)). After this, NO₂ emissions were 4% lower than baseline in August 2020. There was no net difference in NO₂ emissions over Dar es Salaam in September 2020 but this decreased to 4% lower than the baseline in October 2020. November 2020 and December 2020 recorded 16% and 17% increase in NO₂ emissions when compared to the baseline. This indicates that the lockdown containment measure could have had a reductive effect on NO₂ emissions in Dar es Salaam.

Overall, there were about 6.0% and 8.91% reductions in NO₂ emissions over Kampala and Nairobi in 2020 while Dar es Salaam had about 1.16% increase in NO₂ emissions. These findings are in agreement with the percentage reduction in the use of gasoline and diesel from March 2020 for Kampala and Nairobi. Thus, the reductions in NO₂ emissions could mainly be attributable to the reductions in vehicular traffic and industrial activities during the pandemic year (2020). The NO₂ emissions reported were comparable to those retrieved from the Google community mobility reports based on anonymized datasets from users in Kampala and Nairobi cities [31] [32]. High air pollution levels in Tanzanian city of Dar es Salaam is not a new report; it had been reported by previous authors to be caused by its heavy traffic, especially in the rush hours [33] [34] [35].

Reduction in NO₂ levels during the novel COVID-19 pandemic have been previously reported in other countries (cities) including China [17], Cairo and Alexandria governorates of Egypt [15], United States of America [1], India [36] [37], Kingdom of Saudi Arabia [38], Sao Paulo and Rio de Janeiro of Brazil [39] [40], Croatia [41], Barcelona of Spain [42], Southern European cities of Nice, Rome, Valencia and Turin [43] and Italian cities of Florence, Milan, Pisa and Lucca [44] [45].

The concentration of NO₂ in the atmosphere is primarily a function of the magnitude of nitrogen oxide (NO_x) emissions and weather factors such as sun angle, wind speed and temperature, though meteorological variations between years can cause column NO₂ differences of ~15% over monthly timescales [46]. It is clear that the implemented COVID-19 lockdowns (March to May) reduced NO₂ emissions over Kampala and Nairobi, and thus improved air quality [47]. However, these improvements are temporary and unsustainable, with deterioration of air quality anticipated to upsurge as the countries resuscitate their economies post-lockdown or in the post-pandemic period as earlier witnessed in China [48]. It is imperative that stricter laws are enacted in the crucible of the fight against the virus itself so as to protect the environment of the individual countries, and the EAC at large.

As depicted in Figure 5, the 2020 absorbing aerosol index (AAI) in Kampala and Nairobi were lower than the AAI level during the same period in 2019 for February 2020 to May 2020 by about 1.91%, which confirms the positive effect of COVID-19 on air quality in Uganda and Kenya. In the other months, there were insignificant changes in AAI in Kampala and Nairobi. For Dar es Salaam, there were no appreciable increases in the AAI during the pandemic year. Recent reports indicate that much of the particulate matter emissions are from the transport sector (old private cars and taxis) from 6:00 am to 9:00 am in Kampala and the rest of Uganda [49], Nairobi [50] and Dar es Salaam [51]. This explains why lockdowns led to reduced AAI in the former than the latter. This is analogous to the situation reported previously in Egypt [52] [53].

The stringent containment measures adopted by Uganda and Kenyan governments led to less traffic flow and relatively fewer construction activities. Less traffic flow usually leads to less particulate matter (PM) emissions from automobile engines particularly from old taxis and public transportation buses, tyre and brake wears and unmaintained roads [15] [50] [54]. Previous authors reported similar attenuation in PM levels during the COVID-19 in countries (cities) of

Egypt [15], Rio de Janeiro and Sao Paulo, Brazil [39] [40], China [17] [55], India [37], Milan [45], Barcelona [42], Nice, Rome, Valencia and Turin [43]. The PM emissions in Kampala, Nairobi and Dar es Salaam has increasingly surpassed WHO regulatory guidelines in both PM₁₀ and PM_2.5 size ranges [20] [34] [35] [47] [56] - [64]. Thus, there is need to tighten the air quality guidelines to maintain the air quality levels.