ISLAMABAD: The Asian Development Bank (ADB) has flagged a critical data gap in Pakistan’s census, warning that the absence of employment status information hampers any meaningful assessment of how climate shocks affect workers.
The Bank in its latest report “heat stress, air pollution risk, and population exposure evidence from selected Asian countries”, stated that in the case of Pakistan, employment status is not reported in the available Census data, limiting us to study the population exposure to climate shocks while distinguishing the workers.
In Pakistan, industrial and traffic-related emissions such as smog are more prevalent throughout the year and show less seasonal variability, leading to smaller fluctuations in overall PM2.5 concentrations.
READ MROE: ADB approves $540mn financing to advance SOE reforms, coastal resilience in Pakistan
Bangladesh, a low-lying delta enclosed on three sides, has limited air circulation except from the Bay of Bengal, allowing pollutants to accumulate under calm winter conditions.
Pakistan, on the other hand, is exposed to westerly winds and desert airflows, with topographic ventilation in some regions—particularly in northern valleys and arid western areas. While the Indus Basin can trap pollutants, it does so less efficiently and less uniformly than the Bengal Plain.
Pakistan demonstrates a positive correlation between heat stress and pollution, with PM2.5 levels increasing significantly from 2000 to 2020.
The PM2.5 levels in Pakistan exhibit great variation, similar to Bangladesh, with values ranging between 0 µg/m3 and 80 µg/m3 in 2000 and between 0 µg/m3 and 100 µg/m3 in 2020. There is a distributional shift in PM2.5 levels.
More district-months have PM2.5 levels exceeding 80 in 2020 compared to 2000, indicating an increase in the extreme values of the distribution. Yet, for PM2.5 levels below 80 µg/m3 , the average concentration is lower in 2020 compared to 2000, suggesting a shift in the central or lower part of the distribution toward lower values.
The observed correlation between temperature and PM2.5 concentration is positive across years, with Pearson correlation coefficients decreasing from 0.39 in 2000 to 0.20 in 2020. This indicates for certain districts, they are exposed to heat stress and air pollution shocks simultaneously. There is no clear seasonality in PM2.5 levels across different months in Pakistan, as both cold and hot months exhibit similar patterns of air pollution.
The Bank stated that there are several possible explanations for the contrasting seasonal pollution patterns observed in Bangladesh and Pakistan. From a meteorological and climatic perspective, Bangladesh experiences cool, dry, and stable atmospheric conditions during the winter months (November–February), when temperature inversions trap pollutants near the surface.
In contrast, during the summer and particularly throughout the monsoon season (June–September), heavy rainfall and strong southerly winds from the Bay of Bengal effectively cleanse the air through wet deposition and enhanced ventilation.
Pakistan, by comparison, spans diverse climatic zones—including arid deserts, high mountains, and fertile plains—resulting in weaker uniformity in meteorological cycles.
Rainfall is less abundant and more irregular than in Bangladesh, even during the summer. Dust storms and dry heat during this period contribute to high natural PM2.5 concentrations (dust aerosols), offsetting potential declines in anthropogenic PM2.5. Consequently, summer PM2.5 levels in Pakistan do not drop sharply, producing a flatter seasonal pattern.
In the case of Pakistan, employment status is not reported in the available Census data, limiting us to study the population exposure to climate shocks while distinguishing the workers. Data limitation also comes from geo-information. For example, for Viet Nam only the province is observed in Census 1989, while district information is recorded in Census 2019.
Unlike natural disasters, the influence from temperature and air pollution on human-being is more granular, hence the measurement error can be much higher if we aggregate the climate factors at the province level and assign them to the individuals living in such locations.
Copyright Business Recorder, 2026
Comments