Validation Contribution

A first validation against NDACC and WOUDC ground-based data confirms that the TROPOMI/S5p NRTI total O3 (L2_O3) product meets mission requirements.
Initial Sentinel-5p TROPOMI L2_O3 O3 column data retrieved with the PDGS NRTI processor (v1.0.0) have been compared to reference measurements acquired by instruments contributing to WMO’s Global Atmosphere Watch. Of the order of 20 to 100 co-locations have been identified at about 20 Brewer and Dobson sites and at 11 SAOZ sites, sampling many latitudes from the Arctic to the Antarctic. The following preliminary conclusions could be drawn:
Short scale temporal variations in the ozone column as captured by ground-based instruments are reproduced very similarly by S5P. This overall good agreement is corroborated by Pearson correlation coefficients well above 0.9.The bias between S5P and reference data at individual stations rarely exceeds 3%,and the network-averaged bias is about +1%.The 1-sigma spread of the differences (between S5P and reference data) rarely exceeds 3-4% for the comparisons with direct-sun instruments.The analysis of potential dependence of the S5P bias and spread on key influence quantities, namely, the Solar Zenith Angle (SZA), Air Mass Factor (AMF) and cloud fraction of the S5P measurement, does not reveal any variation of the bias larger than 2% over the range of those influence quantities. The spread of the differences increases up to 7% with increasing SZA.Overall, the quality of the initial L2_O3 PDGS NRTI v1.0.0 data product appears to comply with the S5P mission requirements. Some patterns in S5P vs. GOME-2/MetOp-A comparison maps, seemingly related to cloud cover, require further investigation.
For details, please read the report in pdf format.
Download PDFfile
Created by
Bavo Langerock
2018-07-27 15:44:46
Involved Targets