Validation of the TROPOspheric Monitoring Instrument (TROPOMI) surface UV radiation product

Abstract

Abstract. The TROPOspheric Monitoring Instrument (TROPOMI) onboard the Sentinel-5 Precursor (S5P) satellite was launched on 13 October 2017 to provide the atmospheric composition for atmosphere and climate research. The S5P is a Sun-synchronous polar-orbiting satellite providing global daily coverage. The TROPOMI swath is 2600 km wide, and the ground resolution for most data products is 7.2×3.5 km2 (5.6×3.5 km2 since 6 August 2019) at nadir. The Finnish Meteorological Institute (FMI) is responsible for the development of the TROPOMI UV algorithm and the processing of the TROPOMI surface ultraviolet (UV) radiation product which includes 36 UV parameters in total. Ground-based data from 25 sites located in arctic, subarctic, temperate, equatorial and Antarctic areas were used for validation of the TROPOMI overpass irradiance at 305, 310, 324 and 380 nm, overpass erythemally weighted dose rate/UV index, and erythemally weighted daily dose for the period from 1 January 2018 to 31 August 2019. The validation results showed that for most sites 60 %–80 % of TROPOMI data was within ±20 % of ground-based data for snow-free surface conditions. The median relative differences to ground-based measurements of TROPOMI snow-free surface daily doses were within ±10 % and ±5 % at two-thirds and at half of the sites, respectively. At several sites more than 90 % of cloud-free TROPOMI data was within ±20 % of ground-based measurements. Generally median relative differences between TROPOMI data and ground-based measurements were a little biased towards negative values (i.e. satellite data < ground-based measurement), but at high latitudes where non-homogeneous topography and albedo or snow conditions occurred, the negative bias was exceptionally high: from −30 % to −65 %. Positive biases of 10 %–15 % were also found for mountainous sites due to challenging topography. The TROPOMI surface UV radiation product includes quality flags to detect increased uncertainties in the data due to heterogeneous surface albedo and rough terrain, which can be used to filter the data retrieved under challenging conditions.