Direct-sun versus Sky-Scan Pandora Formaldehyde Retrievals: Implications for OMI Validation in Tropical Southeast Asia

Abstract. Ground-based Pandora spectrometers are widely used for validating satellite formaldehyde (HCHO) retrievals; however, the influence of scanning geometry and spatiotemporal representativeness remains insufficiently quantified in tropical environments. This study evaluates Pandora Level-2 HCHO total vertical columns from five Southeast Asian stations (Bangkok, Bandung, Agam, Pontianak, and Singapore-NUS) over 2021–2025, comparing Direct-sun and Sky-scan retrievals and assessing their consistency with OMI Aura observations. HCHO distributions exhibit strong inter-site variability and pronounced skewness, with Direct-sun retrievals showing higher medians and substantially larger variance than Sky-scan observations. Mean Direct-sun HCHO columns are strongly influenced by episodic enhancements at biomass-burning-affected sites, particularly Agam, whereas Sky-scan retrievals display lower central values and reduced variability, consistent with broader atmospheric sampling and diminished sensitivity to localized plumes. Satellite–ground comparisons are conducted using nine spatiotemporal averaging configurations that vary OMI spatial footprints (nearest grid, 3 × 3, and 5 × 5) and Pandora temporal averaging. Direct-sun comparisons generally yield weak or unstable correlations (R ≈ −0.1 to 0.3) and large errors (RMSE ≈ 8–14 × 10¹⁵ molecules cm^-2). In contrast, Sky-scan retrievals show systematically improved agreement, with optimized configurations achieving RMSE values of ~5 × 10¹⁵ molecules cm⁻², MAE of ~4–7 × 10¹⁵ molecules cm⁻², and moderate positive correlations (R ≈ 0.4–0.6) at several sites. Solar zenith angle–dependent analysis reveals persistent positive biases in Direct-sun retrievals (~10–20 × 10¹⁵ molecules cm^-2), while Sky-scan retrievals exhibit near-zero bias at low to moderate SZAs and substantially reduced extremes. Overall, the results demonstrate that scanning geometry exerts a first-order control on Pandora–OMI consistency in the tropics, with Sky-scan observations providing a more spatially representative reference for satellite validation, although optimal configurations remain site dependent.