3.1 Optical parameters at Lahore and Karachi
Aerosol optical depth and the Ångstroṁ exponent are two primary columnar aerosol parameters used to characterize aerosols. The AOD, which is a commonly obtained optical measurement through satellites and ground-based devices, indicates the amount of aerosol loading and specifies the column-integrated aerosol abundance in the atmosphere (Goudie., 2006), and the AE is the aerosol particle size indicator (ÅngstrÖm., 1961). The atmospheric aerosol burden and their size distribution are usually described by the AOD and AE.
The AOD and AE were investigated to get more information about aerosols for 1173 (Lahore) and 988 (Karachi) days. At the Lahore site, the daily average value of AOD-440 nm (AE 440 nm-870 nm) was 0.81 ± 0.49 (1.02 ± 0.32), while for Karachi it is 0.49 ± 0.20 (0.66 ± 0.32). Both optical parameters (AOD and AE) have comparatively large mean values at the Lahore site which is attributed to the more aerosol burden due to vehicular, and industrial emissions, and dust (both from local sources and transported from other regions) aerosols at Lahore, especially during the spring and summer seasons.
Figure 3 (a − e) shows the trend of interannual and seasonal variations in AOD and AE values for the Lahore and Karachi sites with standard deviation bars. The AOD values sharply vary between the spring (Mar, Apr) and autumn (Oct, Nov) seasons, indicating the presence of dust activities during the pre-monsoon, while the AE decreases at the same time from Mar to Nov, which is attributed to the greater contribution of coarse mode aerosols, especially dust transported from the Cholistan and Thar Deserts to the atmosphere of Lahore and Karachi. Over Lahore, the AOD (AE) values ranged from 0.68 − 0.92 (1.16 − 1.28) in winter, 0.53 − 0.70 (0.72 − 0.96) in spring, 0.91 − 1.09 (0.75 − 1.02) in summer, and 0.84 − 1.01 (0.97 − 1.23) in autumn. A decrease in the AE value between Apr, and Jun, at both sites was linked mainly to dust loading. At the Lahore and Karachi sites, aerosols are not only produced through local sources but also transported from the transboundary sources, which results in smog production and reduced atmospheric visibility (especially in autumn and winter) that may contribute to deteriorating air quality. Furthermore, with increasing AOD, the PLDR decreases which means that the particle dimensions become smaller. A positive correlation (r = 0.35) was detected for the monthly mean AOD − AE at Lahore, while a negative correlation (r = − 0.30) was detected at the Karachi site. There are more coarse-mode (86% and 99%) aerosols than fine mode (14% and 1%) aerosols at both Lahore and Karachi sites, respectively. In Fig. 4, the variation in AOD-AE is shown with PLDR and SSA at 1020 nm which shows that with an increase in PLDR, AOD decreases as the dimensions of atmospheric particles increase, i.e., more coarse mode aerosols. A strong negative correlation (r = − 0.94) between PLDR (1020) and AE at both sites was recorded. Table 3 shows the statistical results of the direct and inversion products during the entire study period.
Table 3
Monthly mean statistical variations in AOD, AE, SSA, and PLDR for both sites during the study period
| Lahore | Karachi |
Parameters | Max | Min | SD (±) | Max | Min | SD (±) |
AOD (440 nm) | 1.08 | 0.52 | 0.18 | 0.59 | 0.39 | 0.07 |
AE (440–870 nm) | 1.28 | 0.72 | 0.19 | 0.99 | 0.35 | 0.23 |
SSA (1020 nm) | 0.96 | 0.89 | 0.02 | 0.87 | 0.96 | 0.02 |
PLDR (1020 nm) | 0.17 | 0.02 | 0.05 | 0.10 | 0.26 | 0.05 |
3.2. PLDR Correlations with Other Parameters
This study examines the spectrum pattern of VPSD and SSA at different PLDR intervals derived from observations made at the Lahore and Karachi sites. We investigated the sensitivity of these characteristics to alterations in the PLDR and the potential of these parameters to classify atmospheric particles. Figure 5 displays the distinct patterns of spectrum changes in the average SSA for different PLDRs for Karachi and Lahore, throughout the measurement period from 2015 to 2023. For a PLDR ⩾ 0.29 (< 0.10), a positive (negative) linear relationship between the SSA and wavelength shows the presence of anthropogenic pollution and mineral dust from biomass burning (Giles et al., 2012). For Lahore (Karachi), SSA values ranged between 0.86 (0.88) and 0.92 (0.94) for PLDRs < 0.10 and between 0.93 (0.94) and 0.97 (0.97) for higher PLDRs. Furthermore, compared to that in Karachi, where the range of the SSA is comparatively similar at all wavelengths, Lahore indicates the largest range of the SSA for various PLDR intervals at 440 nm. Soot, also known as black carbon, is the primary absorber of light over the whole radiation spectrum and possesses substantial light-absorbing properties in anthropogenic pollution, on the other hand, the iron oxides included in mineral dust cause significant absorption of light at nearby wavelengths (Bergstrom et al., 2002; Derimian et al., 2008).
The variations in the fine (radius ≤ 0.6 µm) and coarse-mode (radius ≥ 0.6 µm) parts of the volume size distribution parameter for the different intervals of the PLDR are shown in Fig. 6. The inversion technique returns the distribution of aerosol sizes in the range of 0.05 µm to 15 µm. Fine-mode aerosols (< 0.07 µm3/µm2) dominate at lower PLDRs (< 0.10), while the concentrations of coarse mode- aerosols (0.21 µm3/µm2 at Karachi, and 0.27 µm3/µm2 at Lahore) increases with on increasing PLDRs. PLDRs higher than 0.3 (lower than 0.08) indicate the presence pure dust (biomass burning, marine, or anthropogenic pollution) aerosols (Burton et al., 2012; Shin et al., 2018). A range of PLDRs from 0.08 to 0.30 indicates the presence of non-spherical particles and mixtures of mineral dust (Groß et al., 2011).
3.3. Characteristics of aerosol types
The spatial distributions of eight types of atmospheric aerosols, considered for Lahore and Karachi for a period of 9 years (2015–2023) are shown in Fig. 7. The incidence rates of dust-free (MA + NA + SA + WA) pollution-type aerosols over Lahore were as follow: 46% weakly absorbing (WA), 25% moderately absorbing (MA), 24% of non-absorbing (NA), 1% strongly absorbing (SA) aerosols, and other types that contain pure dust (PD) and polluted dust (PDM and DDM), accounting for 43% of the PD, 34% of Dust-Dominated Mixture (DDM), 22% of Pollutant-Dominated Mixture, and just 1% of the Pollution aerosols, respectively. The lowest percentage of aerosol were strongly absorbing (SA: <1%) aerosols in the dust-free category and pollution aerosols (PA: 1%) in the polluted-dust category over Lahore, while weakly absorbing (WA: 46%) and pure dust (PD: 43%) aerosols had the highest percentage occurrence. Fewer light-absorbing (NA + WA) aerosols are detected more frequently over Lahore. The occurrence rates of all aerosol types over Lahore in descending order is WA > PD > DDM > MA > NA > PDM > SA > PA.
Similarly over Karachi, the percentages of dust-free aerosols are 44% for WA and NA, 11% for MA, merely 1% for SA, 49% for PDM, 30% for DDM, 13% for PA, and 8% for PD. The Karachi site is mostly affected by the pollutant-dominated mixture (PDM: 49%) aerosol-type among the pure-polluted dust category of aerosols. Similar to Lahore, the Karachi site also has a higher occurrence rate of less light-absorbing aerosols than more light-absorbing (MA + SA) aerosols. Similar to that in Lahore, the occurrence rate in Karachi was ranked as follow: PDM > NA = WA > DDM > PA > MA > PD > SA. Because of the relatively low frequency of PD and the high frequency of PDM and DDM over Karachi, the dust particles mix with different forms of anthropogenic pollution, which will constitute as an optical signature of a mixture in the column-integrated AERONET data.
On a monthly and annual basis, (Fig. 8), the PD aerosol type is dominant over the Lahore in the fall and winter seasons, with the highest occurrence rates occurring in the winter (88% in Dec, 93% in Jan, 74% in Feb) and fall (23% in Sep, 32% in Oct, 86% in Nov); additionally, the PD aerosol-type contributes 46% of the annual average, and has a relatively lower concentration in the spring (Mar: 19%, Apr: 4%) and summer (Jun: 7%, Jul: 31%, Aug: 26%) seasons. Prevailing pure dust aerosols is the major cause of poor visibility during autumn and winter in Lahore. The NA aerosol type sharply increased from May (10%) to Aug (87%), and had highest occurrence in the summer (Jun: 31%, Jul: 64%, Aug: 87%) season and then decreased in the autumn (Sep: 40%, Oct: 7%, Nov: 1%) and spring (Mar: 25%: Apr: 27, May: 10%), followed by the winter (Dec: 5%, Jan: 10%, Feb: 15%) season, and contributed 27% yearly to the total aerosol burden. The spring (Mar: 40%, Apr: 60%, May: 65%) and autumn (Sep: 51%, Oct: 53%, Nov: 51%) seasons had high occurrence rates of WA aerosols, followed by the summer (Jun: 65%, Jul: 33%, Aug: 9%) and winter (Dec: 43%, Jan: 37%, Feb: 28%) seasons. WA aerosols have a 45% annual mean occurrence rate. DDM aerosols vary between 20% and 60% during spring (Mar: 54%, Apr: 45%, May: 47%) and summer (Jun: 40%, Jul: 20%, Aug: 27%) seasons and have relatively higher concentrations during autumn (Sep: 49%, Oct: 59%, Nov: 13%) and lower concentrations during winter (Dec: 12%, Jan: 7%, Feb: 17%). It contributes 33% to the annual mean. The PDM aerosol type varied between 20% and 55% throughout the study period over Lahore, with more concentrations in summer and spring and less concentrations during winter and autumn. The PDM has a 33% annual average aerosol contribution to the atmosphere of Lahore. Other types of aerosol (MA: 23%, SA: 8%, PA: 7%) had relatively low annual mean concentrations. Annual mean occurrence of all aerosol types in descending order over Lahore was PD (46%) > WA (45%) > PDM (33%) > DDM (32%) > NA (27%) > MA (23%) > SA (8%) > PA (3%). Higher concentrations of PD (winter + autumn), PDM (spring + summer), and DDM (spring + summer) indicate that the urban industrial pollutants and dust particles are combined, while among the dust-free pollution particles; aerosols with weak absorption have the largest contribution. Dust-free pollution particles are observed in all seasons, with more occurrences of WA and NA in spring − summer, and MA and SA in autumn − winter.
At the Karachi site, among dust-free aerosol types, WA particles are persistent in winter (Dec: 90%, Jan: 17%, Feb: 21%), NA in spring (Mar: 60%, Apr: 40%, May: 68%) and summer (Jun: 55%, Jul: 50%, Aug: 67%), and WA in all seasons, especially in spring (Mar: 33%, Apr: 57%, May: 30%) and autumn (Sep: 36%, Oct: 48%, Nov: 65%), while SA was persistent during the studied period, which indicates a possible combination of dust, urban/industrial aerosols, and sea salt over the Karachi region. Among the dust-containing categories, PDM aerosol types are largely observed in all seasons, especially in the spring (Mar: 66%, Apr: 86%, May: 55%) and autumn (Sep: 82%, Oct: 33%, Nov: 11%) seasons, followed by the winter season, while PD particles are dominated in winter (Dec: 20%, Jan: 50%, Feb: 7%) and autumn (Sep: 4%, Oct: 2%, Nov: 27%) seasons. On an annual average, the occurrence rates of all aerosol types in Karachi were ranked as follow: NA (47%) > PDM (44%) > WA (40%) > DDM (38%) > PA (22%) > MA (21%) > PD (17%) > SA (3%). Table 4 shows the seasonal percentage occurrence rates of all aerosol types for both sites.
Collectively, the higher occurrence rates of light scattering aerosols, i.e., NA and WA, over Karachi and Lahore, respectively, indicate that both of these sites are mainly affected by sulfates (SOx), nitrates (NOx), anthropogenic particles, dust, organic matter, and sea salt aerosols which originate from industrial emissions, vehicle exhaust, manufacturing, agricultural activities especially during the harvesting season, waste burning, and the use of biomass for cooking or heating purposes. Furthermore, MA and SA aerosols i.e. light absorbing aerosols, are less abundant at both sites relative to polluted aerosols. Sulfates and other secondary inorganic aerosols mostly scatter light, while carbonaceous aerosols have considerable light absorption characteristics (Bergstrom, 2006).
Table 4
Seasonal frequency rates of dust-free and dust-mixed aerosol types over the Lahore and Karachi AERONET sites during the studied period
Aerosol Type | | Winter | Spring | Summer | Autumn |
| | Lahore | Karachi | Lahore | Karachi | Lahore | Karachi | Lahore | Karachi |
MA | | 41.26% | 42.70% | 19.65% | 4.85% | 2.27% | 16.67% | 29.51% | 15.42% |
NA | | 10.35% | 33.04% | 20.44% | 55.10% | 60.94% | 57.22% | 16.28% | 38.62% |
SA | | 12.14% | 4.17% | 6.57% | | 1.82% | | 3.96% | 2.4% |
WA | | 36.25% | 33.89% | 55.53% | 40.05% | 36.18% | 37.22% | 51.56% | 49.46% |
DDM | | 11.59% | 54.29% | 48.88% | 14.15% | 29.11% | 32.96% | 40.49% | 45.55% |
PD | | 85.42% | 25.71% | 11.99% | | 25.04% | 26.67% | 49.16% | 10.92% |
PDM | | 8.96% | 20.00% | 42.60% | 68.61% | 43.01% | 10.83% | 15.01% | 40.95% |
The high occurrence rates of PD (over Lahore) and PDM (over Karachi) mainly between Mar and Sep are due to the large amount of desert and mineral dust particles, including dust from local sources and transport from Thar (India), since the Karachi site is located near the Arabian Peninsula. Another factor is that both sites experience high levels of air pollution because of fog and haze carried by excessive aerosol loading as a result of increased anthropogenic emissions and dense a population. Furthermore, aside from variations in the climatic phenomena, local emissions are also responsible for the seasonality of aerosol characteristics in these areas (Khan et al., 2019).
3.4 Particulate Matter (PM2.5) Variations
Particulate matter exposure can have negative health impacts, such as lung cancer, pulmonary inflammation, and cardiopulmonary death. The components of PM2.5 in the atmosphere come from natural (wind-blown dust, sea salt from seas, and volcanic eruptions) and anthropogenic sources (burning biomass, combustion from cars, and emissions from power plants). Figure 9 shows year-to-year variations in the Air Quality Index (AQI in µgm− 3) from 2019 to 2023 over Lahore. PM2.5 concentrations start increasing in Sep, reach their highest values in Dec/Jan, and then vary during the remaining months, with having the maximum concentrations occurring in Nov and Dec. Almost every year, the autumn season has the highest concentration of PM2.5 followed by the winter, spring and summer seasons. The highest PM2.5 concentrations were observed in Dec 2019, 2020, 2021, and 2022, with an average AQI values of 298.63 µgm− 3, 291.37 µgm− 3, 374.96 µgm− 3, and 305.69 µgm− 3, respectively making the atmosphere “very unhealthy” for humans.
The Lahore division was under the influence of intense smog in the autumn of 2023, with average AQI values of 122.77 µg/m3 in Sep, 185.26 µg/m3 in Oct, and 288.00 µg/m3 in Nov. Such high values of particulate matter (PM2.5) over Lahore and its neighboring districts have very “unhealthy” effects on human health. Every November or December for the past few years, pollution has decreased in Lahore and the surrounding areas. This haze has grown much worse over the past five years due to high levels of pollution from industry and cars, as well as poor air quality. A notification for the implementation of a smart lockdown was issued by the Provincial Disaster Management Authority (PDMA), Punjab (punjab.gov.pk). Ten districts—Lahore, Nankana Sahab, Sheikhupura, Kasur, Gujranwala, Gujrat, Sialkot, Narowal, Hafizabad, and Mandi Bahauddin—remain under lockdown, per order.
Conclusion
The current study applied an approach to investigate and compare aerosol properties and aerosol type classifications over Karachi and Lahore, based on 9-year ground-based AERONET V3 L2 inversion products. Aerosols are classified into dust-free (NA + WA + MA + SA) and dust-containing (PD, PDM, and DDM) types using SSA and PLDR at 1020 nm, respectively. As the PLDR increases, the AOD and AE decrease (increase) during winter and autumn (spring and summer) at both sites, the SSA increases at each wavelength, and the contribution of fine (coarse) mode aerosols to the volume particle size distribution decreases (increases) over Lahore and Karachi. As a result, compared to the more light-absorbing spherical particles, the non-spherical particles added to the atmosphere of Karachi and Lahore, have larger diameters and less absorption. Among the dust-containing type (PD, PDM, and DDM) of aerosols, PD (Lahore) and PDM (Karachi) were detected frequently at both sites, with occurrence rates of 43.16% and 49.14%, respectively, while WA aerosols were more dominant, at both sites, with occurrence rates of 46.38% (Lahore) and 44.09% (Karachi) among dust-free pollution particles. There were more contributions from NA and WA than from SA and MA (Figs. 2 − 7). Annually, the occurrence rates of dust-containing aerosols (28.47% in Lahore and 30.43% in Karachi) are greater than those of dust-free (25.62% in Lahore and 27.87% in Karachi) types of pollution aerosols. The findings of the present study are in agreement with those of other studies (Khokhar et al., 2016; Tariq., 2018; Khan et al., 2020; Bilal et al., 2022) reported for the same region, although the present work utilized dust ratio parameter for aerosol type classification. The annual variation in the AQI indicated that the autumn and winter seasons, especially in Dec, had the highest concentration of PM2.5, with an average value of 374.96µgm− 3 in December 2022. The highest seasonal average AQI values were 300.52 µgm− 3, 288.81 µgm− 3, 262.64 µgm− 3, 220.38 µgm− 3, and 217.47 µgm− 3 in the winters of 2021, 2019, 2022, 2020, and 2023, respectively, which are the consequences of smog and several diseases such as pneumonia and other respiratory disorders, during this season. The findings from the current study can be used to activate space sensors and for the validation of chemical transport models that provide PLDR parameters. In addition, these findings are useful to overcome the uncertainties in climate estimations of aerosol radiative forcing and addressing issues concerning urban health in megacities such Lahore.