Daily estimation of NO₂ concentrations using digital tachograph data.

Area of Science:

• Environmental Science
• Big Data Analytics
• Air Quality Monitoring

Background:

• Estimating nitrogen dioxide (NO2) concentrations relies on traffic data, but traditional methods are static and fail to capture dynamic traffic changes.
• Static traffic data limits the accuracy of predicting constantly fluctuating NO2 levels, impacting environmental and health assessments.

Purpose of the Study:

• To develop and evaluate a novel Land Use Regression (LUR) model incorporating real-time spatial big data from digital tachographs (DTG).
• To compare the predictive performance of a DTG-LUR model against a conventional non-DTG-LUR model for NO2 concentration estimation.
• To demonstrate the capability of DTG data for capturing spatial and temporal traffic variations for improved air quality modeling.

Main Methods:

• Utilized real-time spatial big data from digital tachographs (DTG) installed in commercial vehicles.
• Constructed a DTG-based Land Use Regression (LUR) model, integrating dynamic traffic variables like cargo traffic.
• Compared the performance of the DTG-LUR model with a non-DTG-LUR model using explanatory power metrics.

Main Results:

• The DTG-LUR model achieved a superior explanatory power of 0.46, compared to 0.36 for the non-DTG-LUR model.
• Spatially and temporally dynamic DTG variables, such as cargo traffic, significantly enhanced model performance.
• The study demonstrated the advantage of DTG data in predicting daily NO2 fluctuations at a precise 200-m grid resolution.

Conclusions:

• Incorporating DTG data into LUR models offers a novel and effective approach for correlating with NO2 concentrations.
• Real-time spatial big data, specifically DTG data, enables more accurate and granular predictions of air pollution than conventional methods.
• The findings highlight the potential of DTG data for fine-grained air quality analyses, paving the way for hourly NO2 predictions.