A linear retrieval approach to infrared sounder data product simulation in support of Observing System Simulation Experiments (OSSEs)

Satellite infrared sounders including the Atmospheric Infrared Sounder (AIRS) on NASA's Earth Observing System Aqua satellite acquire temperature and water vapor profiles of Earth's atmosphere with global daily coverage and unprecedented vertical and horizontal resolution and accuracy from space. Simulating the data acquisition and retrieval process of sounders like AIRS applied to modeled atmospheric conditions (e.g. hurricanes) enables prediction of how well these systems should improve forecasts. Simulations using AIRS performed prior to development of the instrument predicted high forecast improvement with infrared sounders and were proven accurate with AIRS having among the highest impact of any observations to the operational forecast. The next generation of forecast models are expected to operate at an order of magnitude higher spatial resolution than those used for AIRS.

A new set of Observing System Simulation Experiments (OSSEs) are underway to assess the impact of higher spatial and temporal resolution sounding on hurricane forecast accuracy. To support these studies, we have developed an AIRS linear retrieval simulation system. The simulation system uses a satellite orbit track simulation to provide sample locations and footprint area of the infrared sounder configuration to be simulated over the region of interest. The OSSE nature run (simulated temperature and water vapor profiles and cloud properties) is sampled at the sounder locations and integrated over the sounder footprint area. The resulting averaged profiles are convolved with the AIRS averaging kernels using a linear retrieval methodology. The sampled and convolved profiles simulate the retrievals we can expect from the infrared sounder configuration under test. The results can be used in sensitivity and assimilation studies to assess the performance of these new sounders on forecast accuracy. We have examined the AIRS baseline case, a 2km sounder from LEO and a 5-10km sounder from GEO. Retrieval simulation results will be presented.