DOI Services Persistent Identifiers for Scholarly Research
You are viewing the preserved copy. Open the article on the journal's website →

Contemporary Journal of Social Science Review • pp. 1-54

INCOMING AIR-MASS PATHWAYS, FIRE ACTIVITY AND PM2.5 IN LAHORE, PAKISTAN: ASSOCIATIONS AND WEEKLY FORECASTING

  • Mian Shayaan Akbar Lahore Grammar School, Senior Boys Campus, Johar Town, Lahore, Pakistan
  • Sardar Zayaan Akhtar Nakai Lahore Grammar School, Senior Boys Campus, Johar Town, Lahore, Pakistan
  • Robia Shaheen Lecturer & PhD Scholar Institute of Geography, University of the Punjab

Abstract

Every autumn and winter, Lahore’s air turns thick with fine particulate matter (PM 2.5 ), and the explanation depends on who you ask. Some blame smoke from crop fires in Punjab, on both sides of the border. Others blame the city’s own traffic and industry, or the cold, still air that lets pollution pile up. This study asks one question with three parts: which incoming air-mass pathways and weather conditions go with high PM 2.5 in Lahore, how much of that is linked to ventilation and upwind fire activity, and whether this information helps forecast the following week’s mean PM 2.5 . We rebuilt almost six years of hourly PM 2.5 from the

U.S. Department of State monitor in Lahore (May 2019 to February 2025; preliminary AirNow data) and paired it with 2,161 daily HYSPLIT back-trajectories, ERA5 meteorology, NASA FIRMS fire detections and simple ventilation features from a one-box model. Three recurring pathways emerged, most often slow air from the south-east (60% of days). Within seasons, PM 2.5 was up to 30% higher on days with fast north-westerly air, but these differences largely disappeared once weather was taken into account. Poor ventilation was the clearest link: in winter and October to November, a 10% drop in ventilation went with about 5% higher PM 2.5 , whatever the pathway. In October and November, more fire activity along the air’s path went with modestly higher PM 2.5 (about 6% per doubling of exposure), beyond what seasonal timing alone explains. For forecasting, every week was predicted using only information available before it began. Over 138 test weeks, a Random Forest had a mean absolute error of 28.3 µg m −3 , beating persistence (35.5) and seasonal climatology (32.6) but not reliably better than a simple model of past PM 2.5 (29.7). Each added group of inputs (weather, fire, ventilation and trajectories) changed the error by less than 1 µg m −3 ; only the fire inputs gave a gain narrowly larger than chance, and every model underestimated the worst smog weeks. When the data were assumed to arrive a week late, errors rose by 3 to 5 µg m −3 and the Random Forest was no better than climatology. The extra inputs told the model little that last week’s PM 2.5 and the time of year did not already. When one year of trajectories was recomputed with finer (1°) meteorology, the paths and several pathway results changed, but the fire association did not, so the pathway findings depend on the trajectory data used. Because the study uses one monitor and no chemical data, it describes associations and forecasting skill, not where Lahore’s pollution comes from.

Record

Identifier
https://doi-services.com/1102/3540
DOI
10.66857/z67v2s84
Published
2026-03-18
Version
1
License
https://creativecommons.org/licenses/by/4.0
Original
https://studentsconsultancyhome.com/contemporary-journal/index.php/cjssr/article/view/3540