Session: Government Agency Student Posters

Paper Number: 173856

Integrating Satellite and Ground Data to Extend the Local Spatiotemporal Resolution of Urban Air Phenomena 

As urban areas evolve into smarter, more connected systems, they generate and rely on increasingly diverse streams of environmental data. From satellite based sensors orbiting the Earth to ground-level environmental monitors in neighborhoods, modern cities generate an unprecedented volume of data. Yet, synthesizing these heterogeneous datasets into actionable insights remains a core challenge for renewable energy development, public health, and smart city decision-making.

A critical limitation lies in the spatiotemporal resolution gap between coarse satellite observations and sparse ground-based measurements. Platforms like NASA TEMPO (Tropospheric Emissions: Monitoring of Pollution) and ESA TROPOMI (Sentinel-5P) offer valuable coverage of atmospheric trace gases, but at spatial footprints of approximately 3×5 km², insufficient to resolve block by block variations in urban air quality, wind behavior, or heat island intensity. Local-scale phenomena often demand resolutions on the order of 10–100 meters. Moreover, deploying dense sensor networks remains costly, and integrating multi-source data presents significant analytical complexity.

To address these challenges, we present a data-driven framework that fuses satellite-derived tropospheric column densities (such as NO₂, HCHO, O₃) with ground-level meteorological (such as temperature, wind speed, wind direction) and air quality observations from AirNow, Pandora, and Meteostat. Case studies in Hampton Roads, VA and Cleveland, OH span a full annual cycle at hourly to daily intervals.

 

The preprocessing pipeline uses open-source tools, including the Copernicus Data Space Ecosystem, ESA Atmospheric Toolbox, and EPA pyrsig API, to filter, reproject, and align Level 2/3 satellite products onto a unified spatial grid. Statistical fusion techniques, such as kriging, regression modeling, and correlation-based temporal alignment, integrate datasets while accounting for uncertainty and lag effects.

Preliminary results show strong correlations, using NO₂ as one example, between satellite-derived values and ground-based measurements during specific hourly intervals, particularly under low wind conditions. Overlaying TEMPO and Pandora datasets revealed localized NO₂ hotspots in both cities. Python-based workflows using Pandas, NumPy, Matplotlib, and Seaborn enabled time-series plots, correlation heatmaps, and seasonal trend visualizations that highlight pollutant transport and its interplay with meteorology.

While demonstrated in the context of air quality, the framework serves as a generalizable platform for integrating heterogeneous environmental data. It supports future applications in local wind estimation, urban heat island analysis, and microclimate modeling, to name a few. Additionally, it can inform wind energy density estimation and turbine siting, bridging environmental monitoring with renewable energy planning in smart cities.

 

Keywords

Statistical Data Fusion; Satellite-Ground Integration; Spatiotemporal Environmental Modeling; Urban Wind; Pollutant and Wind Transport Dynamics

Presenting Author: Nate Jackson Lehigh University

Presenting Author Biography: Nate served as an Undergraduate Research Assistant under the NSF REU in the M3TFluiD lab at Cleveland State University in Summer 2025.

Authors:

Navid Goudarzi Cleveland State University
Yash K. Sanap George Mason University
Sridhar Katragadda City of Virginia Beach
Nate Jackson Lehigh University