CoAuthors

Shailen Desai (JPL, USA); Nicholas Lahaye (JPL, USA)

The Jason-1 sensor provided more than a decade of sea level measurements and with the conclusion of the mission, the job is now to ensure the final mission product contains the highest quality climate calibration. In this paper, we describe the end of mission calibration for the Jason Microwave Radiometer (JMR). The first step in the process is to ensure the long term stability of the brightness temperatures. To maintain the radiometer brightness temperature calibration on long time scales requires that it be referenced to stable sources external to the radiometer, where the long term calibration is then dependent upon the stability of these external sources. Using several independent external references and demonstrating consistency between them increases the confidence in the resulting long term WPD record and hence GMSL record.
To stabilize the long term calibration, an inter-satellite calibration approach was applied. This approach essentially transfers the long term calibration from other stable externally calibrated satellite microwave radiometers to the altimeter radiometers. The calibration standard in this case was chosen to be the SSM/I TB fundamental climate data record (FCDR) (Kummerow et al., 2010). The SSM/I FCDR was developed by Colorado State University (CSU) for NOAA and extends from 1987 to the present, covering the altimeter time period. The inter-satellite calibration provides a second independent reference (in addition to the natural reference method) and the demonstrated agreement between the two gives confidence in the observed long term TB drift.
In addition to the long term trend, JMR also exhibited time variable changes in calibration instrument temperature dependency. The effect is to introduce transient 60-day variations which cannot be easily detected using the TB references, but is very apparent when comparing the geophysical retrievals to other references. A process was developed to infer the TB instrument temperature dependence over a 120-day moving window from comparisons of the path delay to the ECMWF model and the AMR wind speed to the altimeter wind speed. This correction was applied along with the long term drift.
Finally, the JMR algorithms were updated to the Jason-2 GDR-D standard using the same processing software as is used for Jason-2 processing, but with coefficients specific to the JMR. The JMR processing will be described and estimates of the overall performance improvement will be given.