Climate-stratified assessment of PM2.5 using machine learning: Geographic controls dominate meteorological factors
- 1Department of Software Engineering, Faculty of Computing, Universiti Teknologi Malaysia, Johor Bahru, Johor, Malaysia
- 2Department of Software Engineering, Daffodil International University, Dhaka, Bangladesh
Abstract
Fine particulate matter (PM2.5) is a major environmental concern, yet pollution assessment models are rarely tested across fundamentally different climate regimes. We develop a climate-stratified machine-learning framework using 49,997 monthly MERRA-2 satellite-assimilated samples (2019–2023) across five latitude-band climate zones, evaluated through Leave-One-Climate-Out (LOCO) cross-validation with nested hyperparameter tuning to prevent data leakage. Tuned LightGBM achieves a global LOCO R2 = 0.367 (MAE =6.43 μg m-3), but predictability varies substantially by zone: subtropical regions reach R2 = 0.604, while tropical regions achieve only R2 = 0.142, reflecting fundamental climate-dependent limits on meteorology-only models. SHAP analysis shows geographic coordinates—acting as proxies for emission patterns and climatological transport pathways—account for 64% of predictive importance on average across zones, compared to 23% for meteorological variables, indicating that PM2.5 spatial structure is driven more by where emissions occur than by month-to-month weather variability. To verify that coordinates add genuine predictive value beyond spatial autocorrelation, a retrain ablation confirms that removing latitude and longitude collapses model performance across all zones (mean R2 drops from 0.367 to −0.099), while an IDW spatial baseline confirms that the ML model captures meteorological signal beyond simple proximity interpolation (+98% R2 improvement over IDW). Temporal validation (2019–2021 training, 2022–2023 testing) yields R2 = 0.544, confirming LOCO is the more demanding test. Ensemble stacking degrades performance relative to the zone-tuned model, providing evidence of feature-set saturation rather than model weakness. This reproducible, open-source framework offers a useful benchmark for reanalysis-based PM2.5 assessment, particularly in data-sparse regions where ground monitoring is limited.
Related
- Synoptic and regional-scale meteorological controls of stratus altitude in the Namib Desert
- Decoupling Kinetic and Hydrological Controls on Global Silicate Weathering via a Stratified Observational Factorial Framework
- Quantifying meteorological impacts on local landfill methane emissions by using field measurements and machine learning
- Cutting-edge trends in drought prediction and mitigation using ensemble learning and augmented XAI: a meteorological study
- Sentinel-1 SAR time series for forest phenology: multimodal deep learning with meteorological fusion across deciduous broadleaf and evergreen needleleaf forests
BibTeX
@article{utsho2026climatestratified,
title = {{Climate-stratified assessment of PM2.5 using machine learning: Geographic controls dominate meteorological factors}},
author = {Someyo Kamal Utsho and Md Rawjatul Fahim and Ruhaidah Samsudin},
journal = {PLOS One},
year = {2026},
month = sep,
doi = {10.1371/journal.pone.0347591},
url = {https://doi.org/10.1371/journal.pone.0347591},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Climate-stratified assessment of PM2.5 using machine learning: Geographic controls dominate meteorological factors
AU - Someyo Kamal Utsho
AU - Md Rawjatul Fahim
AU - Ruhaidah Samsudin
PY - 2026
DA - 2026/09/11
JO - PLOS One
DO - 10.1371/journal.pone.0347591
UR - https://doi.org/10.1371/journal.pone.0347591
AB - Fine particulate matter (PM2.5) is a major environmental concern, yet pollution assessment models are rarely tested across fundamentally different climate regimes. We develop a climate-stratified machine-learning framework using 49,997 monthly MERRA-2 satellite-assimilated samples (2019–2023) across five latitude-band climate zones, evaluated through Leave-One-Climate-Out (LOCO) cross-validation with nested hyperparameter tuning to prevent data leakage. Tuned LightGBM achieves a global LOCO R2 = 0.367 (MAE =6.43 μg m-3), but predictability varies substantially by zone: subtropical regions reach R2 = 0.604, while tropical regions achieve only R2 = 0.142, reflecting fundamental climate-dependent limits on meteorology-only models. SHAP analysis shows geographic coordinates—acting as proxies for emission patterns and climatological transport pathways—account for 64% of predictive importance on average across zones, compared to 23% for meteorological variables, indicating that PM2.5 spatial structure is driven more by where emissions occur than by month-to-month weather variability. To verify that coordinates add genuine predictive value beyond spatial autocorrelation, a retrain ablation confirms that removing latitude and longitude collapses model performance across all zones (mean R2 drops from 0.367 to −0.099), while an IDW spatial baseline confirms that the ML model captures meteorological signal beyond simple proximity interpolation (+98% R2 improvement over IDW). Temporal validation (2019–2021 training, 2022–2023 testing) yields R2 = 0.544, confirming LOCO is the more demanding test. Ensemble stacking degrades performance relative to the zone-tuned model, providing evidence of feature-set saturation rather than model weakness. This reproducible, open-source framework offers a useful benchmark for reanalysis-based PM2.5 assessment, particularly in data-sparse regions where ground monitoring is limited.
SN - 1932-6203
ER -