Occurrence of ozone anomalies over cloudy areas in TOMS version-7 level-2 data

Abstract. This study investigates anomalous ozone distributions over cloudy areas in Nimbus-7 (N7) and Earth-Probe (EP) TOMS version-7 data and analyzes the causes for ozone anomaly formation. A 5°-longitude by 5°-latitude region is defined to contain a Positive Ozone Anomaly (POA) or Negative Ozone Anomaly (NOA) if the correlation coefficient between total ozone and reflectivity is > 0.5 or < -0.5. The average fractions of ozone anomalies among all cloud fields are 31.8 ± 7.7% and 35.8 ± 7.7% in the N7 and EP TOMS data, respectively. Some ozone anomalies are caused by ozone retrieval errors, and others are caused by actual geophysical phenomena. Large cloud-height errors are found in the TOMS version-7 algorithm in comparison to the Temperature Humidity Infrared Radiometer (THIR) cloud data. On average, cloud-top pressures are overestimated by ~200 hPa (THIR cloud-top pressure < 200 hPa) for high-altitude clouds and underestimated by ~150 hPa for low-altitude clouds (THIR cloud-top pressure > 750 hPa). Most tropical NOAs result from negative errors induced by large cloud-height errors, and most tropical POAs are caused by positive errors due to intra-cloud ozone absorption enhancement. However, positive and negative errors offset each other, reducing the ozone anomaly occurrence in TOMS data. Large ozone/reflectivity slopes for mid-latitude POAs show seasonal variation consistent with total ozone fluctuation, indicating that they result mainly from synoptic and planetary wave disturbances. POAs with an occurrence fraction of 30--60% occur in regions of marine stratocumulus off the west coast of South Africa and off the west coast of South America. Both fractions and ozone/reflectivity slopes of these POAs show seasonal variations consistent with that in the tropospheric ozone. About half the ozone/reflectivity slope can be explained by ozone retrieval errors over clear and cloudy areas. The remaining slope may result from there being more ozone production because of rich ozone precursors and higher photolysis rates over high-frequency, low-altitude clouds than in clear areas. Ozone anomalies due to ozone retrieval errors have important implications in TOMS applications such as tropospheric ozone derivation and analysis of ozone seasonal variation.