Exploring the Potential of the Future Geostationary Ocean Color Instruments GLIMR and GeoXO OCX through the Eyes of TEMPO

Retrievals of ocean color (OC) properties from space are important for better understanding the ocean ecosystem and carbon cycle. Traditional OC imagers in polar orbit only provide information at most once a day at mid and low latitudes and have incomplete daily spatial coverage limited by clouds, aerosols, and sun glint. Unlike polar orbiting instruments, the future Geostationary OC Instruments such as the Geostationary Littoral and Monitoring Radiometer (GLIMR) and Geostationary Extended Observations Ocean Color Instrument (GeoXO OCX) will be capable of taking hourly measurements over North America at latitudes <60N. As a result, there will be improved spatial coverage of OC measurements as the cloud, aerosol, and sunglint features are not stationary through the day. Additionally, these measurements will provide a much better understanding of the sub-daily to hourly variations in coastal ecosystems and ocean dynamics. While the launch of GLIMR and GeoXO is years away, the geostationary Tropospheric Emissions: Monitoring of Pollution (TEMPO) instrument launched in 2023 provides a unique opportunity to examine the diurnal variability in ocean ecology across various waters in North America and prepare for the future suite of hyperspectral OC sensors. While TEMPO does not have as high spatial resolution or full spectral coverage as planned coastal ocean sensors such as the GLIMR or GeoXO OCX, it provides hourly coverage of US coastal regions such as the Great Lakes and the Gulf of Mexico at spatial scales of approximately 2x4.5 km at the center of its area of observations. We have applied our machine learning (ML) based atmospheric correction approach for OC retrievals to TEMPO data to begin evaluating the potential of geostationary OC measurements. Our approach begins by decomposing TEMPO hyperspectral L1b solar normalized spectra into principal components, which statistically explain the observed spectral variability in atmospheric scattering and absorption as well as the underlying surface reflectance. The coefficients of the principal components are then used to train a neural network to predict OC properties such as chlorophyll concentration derived from collocated MODIS/VIIRS/OLCI physically-based retrievals. The ML approach provides an efficient way to develop OC retrievals for the TEMPO instrument that provide insight into the potential capabilities of geostationary OC instruments. This approach also allows for retrievals under some low to moderate optical thick cloud and aerosol conditions as well for bright reflectance features such as sunglint.