Ozone air quality measurement requirements for a geostationary satellite mission

We conduct an Observing System Simulation Experiment (OSSE) to test the ability of geostationary satellite measurements of ozone in different spectral regions to constrain surface ozone concentrations through data assimilation. Our purpose is to define instrument requirements for the NASA GEO-CAPE geostationary air quality mission over North America. We consider instruments using different spectral combinations of UV (290e340 nm), Vis (560e620 nm), and thermal IR (TIR, 9.6 mm). Hourly ozone data from the MOZART global 3-D chemical transport model (CTM) are taken as the “true” atmosphere to be sampled by the instruments for July 2001. The resulting synthetic data are assimilated in the GEOSChem CTM using a Kalman filter. The MOZART and GEOS-Chem CTMs have independent heritages and use different assimilated meteorological data sets for the same period, making for an objective OSSE. We show that hourly observations of ozone from geostationary orbit improve the assimilation considerably relative to daily observation from low earth orbit, and that broad observation over the ocean is unnecessary if the objective is to constrain surface ozone distribution over land. We also show that there is little propagation of ozone information from the free troposphere to the surface, so that instrument sensitivity in the boundary layer is essential. UV þ Vis and UV þ TIR spectral combinations improve greatly the information on surface ozone relative to UV alone. UV þ TIR is preferable under highsensitivity conditions with strong thermal contrast at the surface, but UV þ Vis is preferable under low-sensitivity conditions. Assimilation of data from a UV þ Vis þ TIR instrument reduces the GEOSChem error for surface ozone by a factor of two. Observation in the TIR is critical to obtain ozone information in the upper troposphere relevant to climate forcing.