An Improved Quality Control for AIRS Total Column Ozone Observations within and around Hurricanes

Abstract

Abstract Atmospheric Infrared Sounder (AIRS) provides twice-daily global observations from which total column ozone data can be retrieved. However, 20% ~ 30% of AIRS ozone data are flagged to be of bad quality. Most of the flagged data were identified to have total precipitable water (PW) errors, defined by the ratio between PW errors and PW retrieval exceeding 35%. It was found that most data within hurricanes were flagged because of extremely low total PW, which is also retrieved from AIRS observations. In this study, a new PW ratio, defined by the AIRS PW error divided by the National Centers for Environmental Prediction (NCEP) zonal average PW, is used to replace the one in AIRS quality control (QC) scheme. Data are removed if the new PW error ratio exceeds 33%. Only 5% ~ 10% of AIRS ozone data are flagged to be of bad quality. Following this step of QC, a linear regression model, which links the total column ozone to the model’s vertical mean potential vorticity (MPV), is established for future data assimilation of AIRS total ozone. Outliers identified by a biweight algorithm are further removed. Numerical results implementing the proposed QC method are compared with those provided by AIRS for Typhoon Sinlaku (2008) in the Pacific Ocean and Hurricane Earl (2010) in the Atlantic Ocean. It is shown that the new scheme works by retaining more of the good data while still removing the bad data.