Tracing PM2.5 Transport from Likely Sources across Nablus Using AirAware: A Low-Cost Sensor Network Study

Description

This project investigates whether short-term increases in fine particulate matter (PM2.5) measured by the AirAware low-cost sensor network in Nablus can be associated with known pollution sources and subsequently observed at sensors located downwind. AirAware is a real-time monitoring platform developed and maintained by Saeed Hersheh, with the project coordinated by Tawfiq Saleh, who has authorization to access and use the platform data for this research.The study will focus on industrial sites, quarries, major roads, and other recurring particulate-pollution sources around the Nablus sensors. Source locations, types, and elevations will be mapped. Wind speed and direction at source and sensor locations will be retrieved from Open-Meteo at 15-minute intervals and aligned with the AirAware PM2.5 record. The primary analysis will identify short-term PM2.5 rise events, test whether the wind was directed from a known source toward a sensor, and assess whether the distance and estimated travel time are consistent with the observed delay. The analysis will examine differences between day and night, wind conditions, and source–sensor elevation differences.As a secondary analysis, the project will test whether a rise associated with a source is followed by a rise at further downwind sensors. A simple 30–60 minute prediction model will then be evaluated against a persistence baseline in which the next reading is assumed to remain unchanged. The study will focus primarily on Nablus; selected Tulkarem data may be used as a limited comparison where data quality and coverage permit.Saeed Hersheh will perform data quality checks, define and detect PM2.5 events, calculate source–sensor directions and travel-time relationships, implement the analysis in Python, evaluate the prediction baseline and model, prepare figures, and document the results. The expected outputs are a curated research dataset subject to AirAware’s data-access policy, reusable Python analysis tools, a map of likely source–sensor associations, an evaluated short-term prediction model, and a research report. The project is designed as an observational data-science study; its conclusions will describe likely associations and transport patterns rather than claim definitive causal attribution.