There are two algorithms for TROPOMI tropospheric ozone retrieval:

• S5P_TROPOZ_CCD: The convective-cloud differential algorithm (CCD) derives tropospheric ozone columns (TCO) in the tropics by taking differences between total columns under clear-sky conditions and above-cloud ozone columns.
• S5P_TROPOZ_CSA: The cloud slicing algorithm (CSA) retrieves mean upper tropospheric ozone volume mixing ratios above clouds.

The TROPOMI tropospheric ozone product is a climatology Level-2C product. It is the output of the S5P_TROPOZ_CCD and S5P_TROPOZ_CSA algorithms, which use TROPOMI total ozone and cloud Level-2 products as input.

S5P_TROPOZ_CCD

The tropical tropospheric ozone column is retrieved with the convective-cloud-differential method (CCD) using both ozone column and cloud measurement Level-2 products from TROPOMI.

In the first step, cloudy TROPOMI measurements with cloud fraction f_c ?0.8, cloud top albedo a_ct ? 0.75, and cloud top pressure p_ct ? 300 hPa, are used to determine the above-cloud ozone column (above ~200 hPa, including the ozone column in the stratosphere and the Tropical Transition Layer (TTL)), as shown in Figure 1. The cloudy TROPOMI measurements are selected from tropical measurements over the highly convective eastern Indian Ocean and the western Pacific (70°E – 170°W), where the greatest frequency of high level and high albedo clouds is found. The above-cloud ozone column is determined using TROPOMI Level-2 total ozone data and cloud data. The stratospheric (above 200 hPa) ozone columns are then averaged for 1.25° latitude bands between 20°N and 20°S. It is assumed that the stratospheric ozone column is independent of longitude in a given latitude band.

In the second step, cloud-free TROPOMI measurements (f_c ? 0.1) are used to determine the total ozone column. In the case of cloud-free pixels, TROPOMI is able to detect both ozone in the stratosphere and troposphere. The total ozone columns are averaged for 10-day periods on a 1.25° by 2.5° latitude-longitude grid between 20°N and 20°S.

In a last step, the zonal mean stratospheric ozone column is subtracted from the gridded total ozone values, resulting in the 10-day averaged tropospheric ozone column.

Figure 1: Schematic Illustration of Above-Cloud Ozone Column Measurements from TROPOMI for the Tropics As Used in the CCD Technique

S5P_TROPOZ_CSA

Calculation of the upper tropospheric volume mixing ratios at heights between a defined lower limit of cloud top heights and maximum top heights is simple. It uses the correlation between Cloud Top Pressures (CTP) and the ozone columns above those clouds (ACCO). ACCO is calculated from the total ozone, the ghost vertical column and the cloud fraction. For the Cloud Slicing Algorithm (CSA) a set of pairs of above-cloud column ozone and cloud top height pressures is used to determine tropospheric ozone information at heights between the lowest and highest cloud tops. Figure 2 shows a scheme of the CSA.  Data is sampled within one grid cell. When applying a linear regression, the volume mixing ratio can be determined from the slope of the above-cloud column ozone against the cloud top pressure (see example for SCIAMACHY data in Figure 3).

Figure 2: Scheme of the Cloud Slicing Technique

Figure 3 shows how the Tropospheric Ozone Volume Mixing Ratio is calculated from the slope of the linear regression between above-cloud ozone and cloud-top-pressure dataset (see Figure 15). The linear regression here was limited to ozone data where the cloud top pressure was below 425 hPa (black solid symbols).

Figure 3: SCIAMACHY Upper Tropospheric Ozone Volume Mixing Ratio [ppbv] calculation

The algorithm input data are:

• TROPOMI Level-2 total ozone and its associated ghost column to infer the above-cloud column of ozone.
• TROPOMI cloud-top-pressure and the cloud fraction both from TROPOMI Level-2 cloud product are needed for the same geolocation.