Satellite-based observations of nitrogen dioxide (NO2) are important for monitoring air quality, estimating nitrogen oxide emissions, and informing chemistry transport models. While advancements have been made in space-based NO2 retrievals, most measurements are still unable to resolve the detailed structure of NO2 plumes. Designed primarily for aerosol and ocean applications, the plankton, aerosol, cloud, ocean ecosystem (PACE) ocean color instrument (OCI) provides a unique opportunity to retrieve NO2 from high spatial resolution (∼1km^2) hyper-spectral measurements. We exploit a machine learning technique to show that OCI with a spectral resolution of 5nm can provide high spatial resolution information about NO2 when trained with tropospheric monitoring instrument (TropOMI) retrievals. This work demonstrates the potential to rapidly spatially downscale NO2 observations from heritage instruments and retrievals. This can potentially enable emissions estimates with reduced uncertainties and higher temporal resolution. Additionally, it could provide higher resolution information for exposure estimates used in health studies.