Deep learning bias correction of GEMS tropospheric NO: A comparative validation of NO from GEMS and TROPOMI using Pandora observations.

Despite advancements in satellite instruments, such as those in geostationary orbit, biases continue to affect the accuracy of satellite data. This research pioneers the use of a deep convolutional neural network to correct bias in tropospheric column density of NO (TCDNO) from the Geostationary Environment Monitoring Spectrometer (GEMS) during 2021-2023. Initially, we validate GEMS TCDNO against Pandora observations and compare its accuracy with measurements from the TROPOspheric Monitoring Instrument (TROPOMI). GEMS displays acceptable accuracy in TCDNO measurements, with a correlation coefficient (R) of 0.68, an index of agreement (IOA) of 0.79, and a mean absolute bias (MAB) of 5.73321 × 10 molecules/cm, though it is not highly accurate. The evaluation showcases moderate to high accuracy of GEMS TCDNO across all Pandora stations, with R values spanning from 0.46 to 0.80. Comparing TCDNO from GEMS and TROPOMI at TROPOMI overpass time shows satisfactory performance of GEMS TCDNO measurements, achieving R, IOA, and MAB values of 0.71, 0.78, and 6.82182 × 10 molecules/cm, respectively. However, these figures are overshadowed by TROPOMI's superior accuracy, which reports R, IOA, and MAB values of 0.81, 0.89, and 3.26769 × 10 molecules/cm, respectively. While GEMS overestimates TCDNO by 52 % at TROPOMI overpass time, TROPOMI underestimates it by 9 %. The deep learning bias corrected GEMS TCDNO (GEMS-DL) demonstrates a marked enhancement in the accuracy of original GEMS TCDNO measurements. The GEMS-DL product improves R from 0.68 to 0.88, IOA from 0.79 to 0.93, MAB from 5.73321 × 10 to 2.67659 × 10 molecules/cm, and reduces MAB percentage (MABP) from 64 % to 30 %. This represents a significant reduction in bias, exceeding 50 %. Although the original GEMS product overestimates TCDNO by 28 %, the GEMS-DL product remarkably minimizes this error, underestimating TCDNO by a mere 1 %. Spatial cross-validation across Pandora stations shows a significant reduction in MABP, from a range of 45 %-105.6 % in original GEMS data to 24 %-59 % in GEMS-DL.