Srotex/Lagos-ThermoPollution-Nexus: Initial Release for Manuscript Submission

Complete code for "Unmasking the Thermo-Pollution Nexus: A Multi-Modal Decomposition of LST-AOD Dynamics in a Tropical Megacity" manuscript (Salawu et al., 2026).

Contents:
- Google Earth Engine scripts for MODIS LST (MOD11A1) and MAIAC AOD (MCD19A2) data processing
- Python Jupyter notebooks for time-series decomposition (STL), lag correlation, PCA, and extreme event detection
- Statistical analysis and visualization code

Corresponding manuscript abstract:
Rapid urbanization in tropical sub-Saharan Africa intensifies the thermo-pollution nexus—the interplay between land surface temperature (LST) and aerosol pollution (AOD)—with implications for environmental health and sustainability. This study examines spatiotemporal dynamics and statistical interactions between AOD and LST in the Lagos Metropolitan Area (January 2017–August 2025) using a multi-modal decomposition framework (PCA, STL, lagged correlation). Results show no significant long-term monotonic trends; strong seasonal cycles dominate. AOD peaks during Harmattan (>1.25), LST maxima (~30.5°C) occur about one month later. Lagged positive associations (AOD leads LST by 1-3 months) weaken when Harmattan months are excluded. PCA identifies synchronous seasonal (52%) and contrasting (48%) modes. High-AOD episodes are strongly seasonal; heatwaves are episodic with no co-occurrence.

Keywords: Thermo-Pollution Nexus; Urban Climate; Lagos; Remote Sensing; Time-Series Analysis; Google Earth Engine; STL Decomposition; Principal Component Analysis

Version changes (v1.0.0 → v2.0.0):
- Replaced initial placeholder code with complete, validated analysis pipeline
- Added Jupyter notebooks for all statistical analyses
- Includes Harmattan-excluded sensitivity test and partial correlation controls