Merged LEO-GEO Tropospheric NO2 Product from TEMPO and TROPOMI Measurements

A data fusion technique is presented that uses the Kalman Filter algorithm to combine tropospheric nitrogen dioxide (NO₂) column retrievals from NASA’s Tropospheric Emissions: Monitoring of Pollution (TEMPO) satellite sensor and the Sentinel 5-Precursor TROPOspheric Monitoring Instrument (Sentinel-5P TROPOMI) satellite sensor. The approach is similar to data assimilation except that the background model field uses data from a geostationary (GEO) satellite sensor, which in this case are TEMPO measurements that are taken hourly for the entire continental United States during daylight hours. TROPOMI, operating in low Earth orbit (LEO) provides measurements once or twice daily for a given location and may have higher accuracy due to differences in viewing and solar geometry, and instrument characteristics. The Kalman Filter adds TROPOMI measurements to the background data as observations, weighting each dataset based on their respective uncertainties. This process mitigates the uncertainties in each dataset, resulting in a more accurate tropospheric NO₂ product, with the high temporal resolution of TEMPO data. The merged product is compared against the Pandonia Global Network (PGN) ground-based tropospheric NO₂ column retrievals. Urban air quality case studies in Baltimore and New York City are presented to demonstrate the capabilities and quality of the merged product.