par Es-sayeh, Maël;Bauduin, Sophie
;Clarisse, Lieven
;Franco, Bruno
;Smith, Michael;Giuranna, Marco
Référence Icarus, 449, 116951
Publication Publié, 2026-04-01
;Clarisse, Lieven
;Franco, Bruno
;Smith, Michael;Giuranna, MarcoRéférence Icarus, 449, 116951
Publication Publié, 2026-04-01
Article révisé par les pairs
| Résumé : | This paper introduces a framework scheme for the detection and quantification of trace species in planetary hyperspectral datasets. The method, only used for terrestrial purposes yet, relies on a Hyperspectral Range Index (HRI) to quantify the target gas spectral signature with respect to the climatological background. The HRI is subsequently converted to a total column using an artificial feed-forward neural network built from radiative transfer (RT) model simulations. Importantly, we also provide an appropriate uncertainties characterization. The method has been applied to the Planetary Fourier Spectrometer (PFS) instrument onboard Mars Express (MEX) to retrieve H2O vapor column amount. Nine Martian years of PFS/LWC (Long Wavelength Channel) measurements have been processed to acquire a data set of H2O vapor total columns, analyzed, and compared to similar products from other retrieval methods. The results are comparable to those from standard retrieval methods but were achieved at a significantly reduced computational cost, demonstrating the method's suitability for analyzing large observational datasets. |



