Temporal and Spectral Unfiltering of ERBS WFOV Nonscanner Instrument...

Shrestha, A. K., S. Kato, T. M. Wong, P. Stackhouse, and R. P. Loughman (2017), Temporal and Spectral Unfiltering of ERBS WFOV Nonscanner Instrument observations for the period over 1985 to 1998, Journal Of IEEE Tgrs, 1-11.
Abstract: 

Earth Radiation Budget Experiment (ERBE) WideField-of-View (WFOV) nonscanner instrument onboard Earth Radiation Budget Satellite (ERBS) provided critical 15-years outgoing broadband irradiances at the top-of-atmosphere (TOA) from 1985 to 1999 for studying the Earth’s climate. However, earlier studies show that the uncertainty in this radiation dataset (Ed3) is significantly higher after the Mt. Pinatubo eruption in 1991 and satellite battery issue in 1993. Furthermore, [1] showed that the transmission of ERBS WFOV Shortwave (SW) dome degraded due to exposure to direct sunlight. To account for this degradation, a simple time dependent but spectral independent correction model was implemented in the past. This simple spectral-independent model did not completely remove SW sensor artifact as seen in the temporal growth of the tropical mean day-minus-night Longwave (LW) irradiance [2]. A new temporal-spectral dependent correction model of SW dome transmissivity loss similar to that used in the Clouds and the Earth’s Radiant Energy System (CERES) project is developed and applied to the 15-year ERBS WFOV data. This model is constrained by the solar transmission obtained from ERBS WFOV SW nonscanner instrument observations of the Sun during bi-weekly in flight solar calibration events. This new model is able to reduce the reported tropical day-minus-night LW irradiance trend by 34%. In addition, the slope of this new trend is observed to be consistent over different regions. Further, the contrast in the slope of ocean and land observed in the Ed3Rev1 dataset is also minimized with this spectral unfiltering. The remaining trend is accounted using a post-process Ed3Rev1 correction described in [2]. Furthermore, the time series analysis of this data over the Libya-4 desert site showed that the instrument is stable within 0.7%.