Mars almost certainly had a considerable amount of water in its past. Recent observations reveal that during southern summer, when the atmosphere is warmer and dustier, water vapor can reach high altitudes without condensing, leading to water loss to space. Here, by combining infrared, visible, and ultraviolet data from multiple Mars orbiters, we identify a new pathway for water loss, observed for the first time to our knowledge during the opposite season. Our findings show that a strong, localized, and short-lived dust storm in Martian Year 37 (August 2023) drove considerable vertical transport of water vapor in the northern summer season. Just days after the storm, enhanced water vapor concentrations were observed at altitudes over 40 km across northern high latitudes, followed by an increase in escaping hydrogen detected at the exobase. These results suggest that water loss on Mars can be triggered by strong local dust storms at any time of year. Observations compiled from several Mars observation missions suggest a significant but short-lived dust storm during the Northern hemisphere summer of Mars Year 37 drove substantial vertical transport of water vapor into the upper atmosphere.

Abstract. Clouds over the Antarctic Plateau exert a strong influence on the regional radiation budget, yet observations and modelling of their properties remain scarce. Here, nine years (2012–2020) of ground-based high-resolution spectral radiance measurements from the REFIR-PAD spectroradiometer at Concordia Station (Dome C, Antarctica) are analyzed in synergy with co-located lidar observations. A machine-learning Cloud Identification and Classification (CIC) algorithm is applied to discriminate clear sky, ice cloud, and mixed-phase cloud conditions, enabling the construction of a long-term cloud climatology. Cloud optical and microphysical properties are subsequently retrieved using a simultaneous atmospheric and cloud retrieval framework, for cases with reliable cloud boundaries and cloud base heights above 500 m. Results confirm that cloud occurrence over Dome C is dominated by optically thin ice clouds, with approximately 95 % of cases exhibiting optical depths below 1. Median optical depth ranges from 0.11 in summer to 0.32 in winter. The median temperature of the ice layers is approximately 237 K. Mixed-phase clouds are rare and mainly confined to the austral summer, but exhibit larger optical depths (median 1.7) and warmer temperatures (approximately 246 K). Based on the retrieved dataset, a new parameterization of ice crystal effective dimension is derived. Compared with commonly used parameterizations developed for tropical and midlatitude conditions, the proposed scheme predicts systematically smaller particle sizes, highlighting the inadequacy of existing formulations for the Antarctic environments. These results provide new observational constraints on Antarctic cloud microphysics and support improved cloud representation in climate and numerical weather prediction models.

This study presents an optimal estimation retrieval framework for deriving atmospheric thermodynamic state and cloud microphysical properties from infrared radiance measurements. The proposed method exploits a fast radiative transfer model, sigma, capable of computing atmospheric emission spectra in all-sky conditions. The retrieval framework follows the optimal estimation approach, enhanced by principal component analysis state compression. Cloud prior information is introduced through a dedicated covariance matrix designed to generate physically plausible vertical profiles of ice content and effective dimensions and stabilize the inversion. This combination of methods offers a robust and flexible solution for retrieving atmospheric and cloud parameters from spectrally resolved infrared observations. The retrieval algorithm is tested on both synthetic and real Infrared Atmospheric Sounding Interferometer (IASI) acquisitions. The results highlight the capability of the presented scheme in retrieving cloud property profiles, such as optical depth, effective dimension, and cloud position and its feasible application to operational processors.

To retrieve surface and atmospheric temperature profiles, together with trace species concentrations is a fundamental challenge in numerical weather prediction and Earth monitoring. Over the last 20 years, the development of high-resolution infrared sensors on board Earth observation satellites has opened new remote sensing opportunities, providing an unprecedented source of information. However, infrared sensors cannot probe into thick cloud layers, rendering their observations insensitive to surface under cloudy conditions. This results in spatial fields flagged with missing data, disrupting the continuity of inferred information and hindering accurate modeling of energy fluxes between the surface and the atmosphere. Consequently, advanced interpolation techniques and spatial statistics are essential to process the available (very large) data sets and produce satellite products on a regular grid mesh. This paper reviews and presents the physical modeling of radiative transfer in the atmosphere and the related mathematics of inversion, tailored for high spectral-resolution infrared sensors.

Reliable estimation of surface-level nitrogen dioxide (NO2) concentrations is critical for air quality assessment in urban regions, where ground-based monitoring networks often provide limited spatial coverage. Satellite observations from Sentinel-5P offer valuable information on tropospheric NO2 columns, but their translation to surface concentrations remains challenging due to strong spatial heterogeneity and meteorological influences. This study investigates the estimation of daily surface-level NO2 concentrations across the Community of Madrid (Spain) by combining Sentinel-5P observations with routinely available ground-based meteorological data using machine learning approaches. Four modelling paradigms were evaluated: Random Forest, Support Vector Machine, hybrid ensemble models based on Random Forest and XGBoost, and Artificial Neural Networks. The analysis systematically assessed the influence of temporal and spatial preprocessing by comparing two temporal aggregation strategies-satellite overpass conditions alone and a dual-time window including antecedent atmospheric conditions-and four spatial configurations ranging from regional aggregation to environmentally coherent clustering. Results indicate that incorporating historical meteorological information through an extended temporal window consistently improves predictive performance across all models. Spatial stratification was found to be equally critical: grouping monitoring stations into environmentally coherent clusters substantially outperformed both purely geometric grids and region-wide aggregation. The best-performing configuration, an Artificial Neural Network combined with simplified coherent spatial clusters, achieved an RMSE of 2.44 & micro;g/m3, an R2 of 0.87, and a MAE of 1.61 & micro;g/m3. These findings demonstrate that high predictive accuracy can be achieved through informed temporal and spatial design choices without increasing model complexity or relying on auxiliary emission inventories or chemical transport models, providing a transferable framework for urban air quality monitoring.

In this work, we analyse data from the Jovian Infrared Auroral Mapper (JIRAM) imaging spectrometer on board the NASA mission Juno, to investigate the presence of spectrally identifiable ammonia clouds (SIACs). Focusing on the data from the first perijove passage, we found that a white vortex structure near 40 degrees N provides the best candidate. Implementing atmospheric retrieval thanks to the planetary spectrum generator, we fit the JIRAM spectra (in the 2.5-3.1 mu m range) inside and outside the vortex, varying the gaseous ammonia profiles, and the clouds and hazes properties and composition. We found that outside the vortex, the best fit is achieved using main clouds composed of tholins (approximation of an unknown contaminant material). Inside the vortex the best fit is achieved when main cloud decks are composed of pure ammonia ice, or ammonia-coated tholins. We therefore claim the detection of 38 SIACs, all detected over the white vortex structure. With respect to the external regions, the retrieved parameters distributions inside the vortex show: (i) higher altitude hazes and clouds, (ii) smaller haze's effective radii, and (iii) higher gaseous ammonia relative humidity values. Both the detection of pure ammonia ice clouds and the retrieved physical parameters are consistent with the vortex being the result of a moist convection storm that uplifted fresh ammonia from the deep troposphere which in turn either condensed or became a possible source of coating material for existing cloud particles. This work confirms the trend established by space and ground-based observations, for which ammonia clouds on Jupiter are rare and connected to strong convective episodes.

We present observations of the 1.35 ± 0.07 Earth radius planet L 98-59 c, collected using Wide Field Camera 3 on the Hubble Space Telescope (HST). L 98-59 is a nearby (10.6 pc), bright ( H = 7.4 mag) M3V star that harbors three small, transiting planets. As one of the closest known transiting multi-planet systems, L 98-59 offers one of the best opportunities to probe and compare the atmospheres of rocky planets that formed in the same stellar environment. We measured the transmission spectrum of L 98-59 c, and the extracted spectrum showed marginal evidence (2.1 σ ) for wavelength-dependent transit depth variations that could indicate the presence of an atmosphere. We forward-modeled possible atmospheric compositions of the planet based on the transmission spectrum. Although L 98-59 was previously thought to be a fairly quiet star, we have seen evidence for stellar activity, and therefore we assessed a scenario where the source of the signal originates with inhomogeneities on the stellar surface. We also see a correlation between transits of L 98-59 c and L 98-59 b collected 12.5 hr apart, which is suggestive (but at <2 σ confidence) of a contaminating component from the star impacting the exoplanet spectrum. While intriguing, our results are inconclusive and additional data are needed to verify any atmospheric signal. Fortunately, additional data have been collected from both the HST and James Webb Space Telescope. Should this result be confirmed with additional data, L 98-59 c would be the first planet smaller than 2 Earth radii with a detected atmosphere.

We present an evaluation of radiative transfer model performance in all-sky atmospheric conditions through comparisons with infrared radiances measured by the Infrared Atmospheric Sounding Interferometer (IASI). The study focuses on testing different forward modeling approaches within the s-IASI/f2n framework, including a novel linear-in-T approximation. Simulated top-of-atmosphere radiances are produced under a range of atmospheric scenarios (including both clear-sky and cloudy-sky conditions) assuming collocated ECMWF analysis as state vector using various configurations of the s-IASI/f2n code: with and without the linear-in-T approximation, and with either the Chou or Tang cloud treatment. These simulations are compared against collocated IASI observations to assess the spectral accuracy of each configuration. The goal of this study is to quantify how modeling assumptions affect the representation of all-sky infrared spectra, particularly under cloud-contaminated conditions where accurate radiative transfer is most challenging. While the retrieval of cloud properties is supported within the modeling framework, this work focuses primarily on the forward model as a critical step for possible retrieval applications. By systematically evaluating model behavior in all-sky scenes, this study contributes to the refinement of radiative transfer tools for use in climate research, satellite data analysis, and atmospheric remote sensing applications.

Ammonia is historically thought to be the main source of condensable species for Jupiter's main cloud layer (0.5-1 bar level). However, measurements from Galileo first [1] and Juno later [2] showed that the spectral features connected to ammonia clouds are rare (less than 2 % of the entire planetary disk) and not ubiquitous. Using infrared spectra collected by the JIRAM instrument on board the NASA Juno mission we investigated the possible presence of SIACs (spectrally identifiable ammonia clouds) in PJ1 data. As a preliminary step, we used two spectral indicators sensible to the absorption of ammonia ice particles in the 2.97-3.01 micron range and ran a PCA+GMM (Principal Component Analysis + Gaussian Mixture Models) clustering analysis. The two indicators showed higher values in a high-latitude region in which cross-referenced JunoCam images highlight the presence of a Nautilus-shaped cloud, already noticed in PJ14 by previous work [3]. The PCA-GMM analysis identified the spectra in this region as belonging to a specific cluster, different from the surroundings. Performing optimal estimation atmospheric retrievals using the powerful NASA PSG (Planetary Spectrum generator) suite as the forward model, we tried to model all the spectra of this region (considering only the 2.5-3.1 micron range). We first used a toy model with a variable ammonia profile and parametrized pure reflecting hazes (complex refractive index 1.4+0i) and tholin clouds. It is important to stress that Titan’s like tholins must not be intended as a realistic candidate for Jupiter’s aerosol clouds but as an approximation of the real amorphous unknown material that exhibits an evident N-H-bond-like absorption. We found that the described toy model fits well the majority of the spectra outside the Nautilus, whereas the spectra near and inside the Nautilus require more complex assumptions on cloud compositions and so have been re-modeled. As a result, we noticed that a total of 20 spectra are best fitted by a pure ammonia ice cloud model and so have been identified as SIACs. The SIACs are located at the center of the Nautilus-shaped cloud and in correspondence with the nearby swirls. In most cases, the SIACs are surrounded by spectra best fitted by a cloud deck composed of tholin particles coated with ammonia ice. Our results in correspondence with the Nautilus suggest: (I) higher altitude hazes and clouds, (II) higher values of ammonia relative humidity that also reach super-saturation conditions, and (III) smaller effective radii for the haze particles. Such results are compatible with the presence of pure ammonia ice clouds, formed at these latitudes as a consequence of an uplifting event from the lower troposphere that brought a large fraction of fresh ammonia up to reach super-saturation conditions, triggering condensation and/or coating of mixed particles. [1] Baines K. H. et al. (2002) Icarus, 159, 1, 74-94. [2] Grassi D. et al. (2021) MNRAS, 503, 4, 4892-4907. [3] Guillot T. et al. (2023) EGU23, the 25th EGU General Assembly, EGU-17178.

This paper presents advancements in the sigma-IASI/f2n radiative transfer model, developed under two Italian Space Agency projects to support Italy's contribution to ESA's Earth Explorer 9 mission. The model simulates spectrally resolved radiance and its analytical derivatives across 5-3000 cm(-1), addressing key challenges in cloud and aerosol scattering, atmospheric inhomogeneity, and resolution scalability. A novel parameterization represents cloud and aerosol scattering via an apparent optical thickness, significantly accelerating calculations for high-resolution simulations. Additionally, the model introduces an innovative treatment of atmospheric layer inhomogeneity, improving realism in atmospheric representation. Enhanced scalability allows the model to adapt to various instrument resolutions and applications, making it a flexible tool for remote sensing studies involving clouds, atmospheric composition, and climate. These innovations enable accurate, efficient simulations tailored to diverse observational scenarios. The implications for hyperspectral infrared data analysis and remote sensing applications will be discussed.

During March 2025, three intrusions of Saharan dust affected southern Italy, with observable effects on atmospheric composition and, in particular, on greenhouse gases. A recent study conducted by the Institute of Methodologies for Environmental Analysis of the National Research Council of Italy (CNR-IMAA) documented these events through integrated in situ and remote sensing observations. Significant variations in CH4 and CO2 concentrations were detected in correspondence with the dust transport episodes. In this work, we propose an approach based on Physics-Informed Neural Networks (PINNs) to retrieve the vertical profile of CH4. The results are evaluated against high-precision ground-based measurements from CNR-IMAA, in order to assess the model's predictive accuracy and its sensitivity to atmospheric variations associated with the presence of mineral aerosols.

Hydrogen chloride (HCl) was recently discovered in the Mars atmosphere using the ESA's ExoMars Trace Gas Orbiter (TGO) onboard ESA's ExoMars mission. Its discovery is the first confirmation of an active presence of any chlorine-bearing species in the modern Mars atmosphere. TGO permitted investigations of HCl altitude profiles with high precision and showed that water vapor and ice clouds play an important role in the production and temporary loss of HCl. TGO cannot always sample the Martian atmosphere near the surface, and when those measurements are possible, they are highly affected by the increase in dust opacity, nor can TGO observe at equatorial latitudes with high cadence, due to orbital constraints, so its measurements are not suitable to obtain instantaneous global coverage. In this work, we present a methodic investigation of the Martian atmosphere, in support of the ExoMars TGO mission, targeting HCl and water using iSHELL at NASA/InfraRed Telescope Facility. Our observations mapped the Martian atmosphere, exploring three seasons in Martian Year 36. We observed the beginning of an increase in the HCl abundances around L S = 249 degrees-301 degrees, followed by a drop in the abundances around L S = 319 degrees. We confirmed a strong correlation between the spatial distribution of water vapor and HCl-both globally and locally-suggesting that water vapor plays an important role in the production of HCl, in agreement with previous studies. Our observations also suggest the presence of two competing processes involving the dust, one supporting HCl production and another one contributing to its destruction.

The Nadir and Occultation for MArs Discovery (NOMAD) spectrometer has been collecting Mars observations since 2018, providing a massive amount of information regarding its atmospheric composition, its vertical structure and bridging the gap between the previous knowledge of the lower atmosphere and the data from other missions (e.g., MAVEN) regarding atmospheric escape. The capability of the Solar Occultation (SO) channel to map the vertical structure of the atmosphere with a very high (>1000) signal-to-noise ratio, at a very high spectral resolution (>17000) and a high vertical sampling (0.5 to 2 km) is valuable in many contexts, ranging from the search for trace species in the lower atmosphere (10 to 40 km) to mapping the isotopic composition of the main atmospheric constituents (H2O, CO2, CO) or exploring the vertical structure of dust, water ice and CO2 ice clouds. Aerosols are some of the main drivers of the Martian climate, and the study of their spatial distribution and microphysical properties can advance our knowledge of their impact on the climate of the planet and on their formation and dynamics. This work will show the extension of previous investigations focused on dust, water ice and CO2 ice using NOMAD data, by presenting the mapping of these atmospheric components on a global scale over 3 Martian Years (MY34 Ls 160 to MY37 Ls 170). The acquisition by NOMAD of several diffraction orders during a single occultation allows in fact to obtain spectrally broad information that can be used to map dust and water ice vertical distributions and particle sizes. The information content of NOMAD data about particle sizes of water ice has been demonstrated to be particularly high and to give important information about the nucleation processes of water ice. NOMAD data can also be used to look for CO2 ice by combining broad spectral information with localized CO2 ice features at 3600 and 3710 cm-1, which are well identifiable in the NOMAD spectra. Besides presenting the climatology of aerosols, we will illustrate specific features occurring during the Martian Year and their repeatability; more specifically, we will look into the differences between MY 34, characterized by a Global Dust Storm, and following years, to highlight the impact of dust-induced heating over cloud formation. We will also give some insights into CO2 ice cloud formation, which was confirmed to be surprisingly heterogeneous compared to results obtained before TGO operations.

Nadir and Occultation for Mars Discovery (NOMAD) onboard ExoMars Trace Gas Orbiter (TGO) started the science measurements on 21 April, 2018. We present results on the retrievals of water vapor vertical profiles in the Martian atmosphere from the first Mars year measurements of the TGO/NOMAD.NOMAD is a spectrometer operating in the spectral ranges between 0.2 and 4.3 μm onboard ExoMars TGO. NOMAD has 3 spectral channels: a solar occultation channel (SO – Solar Occultation; 2.3–4.3 μm), a second infrared channel capable of nadir, solar occultation, and limb sounding (LNO – Limb Nadir and solar Occultation; 2.3–3.8 μm), and an ultraviolet/visible channel (UVIS – Ultraviolet and Visible Spectrometer, 200–650 nm). The infrared channels (SO and LNO) have high spectral resolution (λ/dλ~10,000–20,000) provided by an echelle grating used in combination with an Acousto Optic Tunable Filter (AOTF) which selects diffraction orders. The concept of the infrared channels are derived from the Solar Occultation in the IR (SOIR) instrument onboard Venus Express (VEx). The sampling rate for the solar occultation measurement is 1 second, which provides better vertical sampling step (~1 km) with higher resolution (~2 km) from the surface to 200 km. Thanks to the instantaneous change of the observing diffraction orders achieved by the AOTF, the SO channel is able to measure five or six different diffraction orders per second in solar occultation mode. In this study, we analyze the solar occultation measurements at diffraction order 134 (3011-3035 cm-1), order 136 (3056-3080 cm-1) and 168 (3775-3805 cm-1) acquired by the SO channel in order to investigate H2O vertical profiles.Knowledge of the water vapor vertical distribution is important to understand the water cycle and escape processes. Solar occultation measurements by the two spectrometers onboard TGO - NOMAD and Atmospheric Chemistry Suite (ACS) - allow us to monitor daily the water vapor vertical profiles through one whole Martian Year and obtain a latitudinal map for every ~20° of Ls. In 2018, for the first time after 2007, a global dust storm occurred on Mars. It lasted for more than two months (from June to August). Moreover, following the global dust storm, a regional dust storm occurred in January 2019. TGO began its science operations on 21 April 2018. NOMAD observations therefore fully cover the period before/during/after the global and regional dust storms and offer a unique opportunity to study the trace gases distributions during such events. We have analyzed those datasets and found a significant increase of water vapor abundance in the middle atmosphere (40-100 km) during the global dust storm from June to mid-September 2018 and the regional dust storm in January 2019. In particular, water vapor reaches very high altitudes, at least 100 km, during the global dust storm (Aoki et al., 2019, Journal of Geophysical Research, Volume124, Issue12, Pages 3482-3497, doi:10.1029/2019JE006109). A GCM simulation explained that dust storm related increases of atmospheric temperatures suppress the hygropause, hence reducing ice cloud formation and so allowing water vapor to extend into the middle atmosphere (Neary et al., 2020, Geophysical Research Letters, 47, e2019GL084354., doi: 10.1029/2019GL084354). The current study presents the results obtained when considering the extended dataset, which covers a full Martian year. The extended dataset includes the recent aphelion season that involves interesting phenomena such as sublimation of water vapor from the northern polar cap and formation of the equatorial cloud belt, and is known as a key period to understand the large north-south hemispheric asymmetries of Mars water vapor. Yet, until now, only few papers reported the water vapor vertical distribution during the aphelion season. The extended dataset also includes the period when the global dust storm occurred the year before; this will allow us to compare the water vapor distributions under global dust storm conditions with those found during non-global dust storm years. In the presentation, we will discuss the H2O vertical profiles as well as the aerosols vertical distribution retrieved from the first full Martian year measurements of the TGO/NOMAD.

The new sigma-IASI/F2N radiative transfer model is an advancement of the sigma-IASI model, introduced in 2002. It enables rapid simulations of Earth-emitted radiance and Jacobians under various sky conditions and geometries, covering the spectral range of 3-100 mu m. Successfully utilized in delta-IASI, the advanced Optimal Estimation tool tailored for the IASI MetOp interferometer, its extension to the Far Infrared (FIR) holds significance for the ESA Earth Explorer FORUM mission, necessitating precise cloud radiative effect treatment, crucial in regions with dense clouds and temperature gradients. The model's update, incorporating the "linear-in-T" correction, addresses these challenges, complementing the "linear-in-tau" approach. Demonstrations highlight its effectiveness in simulating cloud complexities, with the integration of the "linear-in-T" and Tang correction for the computation of cloud radiative effects. The results presented will show that the updated sigma-IASI/F2N can treat the overall complexity of clouds effectively and completely, at the same time minimizing biases.

NOMAD [1] is one of the four instruments on board ESA’s Trace Gas Orbiter and consists of three channels: SO, LNO, and UVIS. The SO channel is dedicated to solar occultation measurements and thus probes the Martian terminator. SO is an infrared spectrometer (2.3-4.3 µm) composed of an echelle grating with an acousto-optic tunable filter for the selection of the diffraction orders. SO has been regularly scanning the atmosphere of Mars from the troposphere to the upper thermosphere since the beginning of the science operations of the Trace Gas Orbiter on April 21, 2018. One diffraction order is ~25 cm-1, and six diffraction orders are scanned at each occultation. The spectral resolution is ~0.15 cm-1, and the signal-to-noise ratio is ~2500. The field of view varies between 1.6 km and 1.85 km, and the vertical sampling varies between 0.1 km and 1 km depending on the beta-angle of TGO. The vertical resolution of the profiles is ~2.5 km. Recently, in 2024, SO started to scan 12 diffraction orders per occultation, dividing the vertical sampling by two but improving the coverage of several species. The resulting vertical resolution is then reduced to a maximum of 50 %. The instrument function of SO was described in ref. [2].The retrieval of CO2 density and temperature was described in ref. [3]. The radiative transfer computations are performed with the ASIMUT software [4]. The regularisation of the profiles is carefully fine-tuned with an iterated Tikhonov method [3, 5]. This fine-tuning of the regularisation helps to better constrain some variabilities in the profiles that are, for instance, produced by tides and gravity waves. The regularisation does not add any a priori information to the retrieved profiles.The diffraction order 132 (2966 to 2930 cm-1) is used to infer the CO2 density and temperature in the troposphere (altitudes below ~50 km), while the CO2 density and temperature are inferred in the mesosphere (~50 to ~100 km) from diffraction orders 148 (3326 – 3353 cm-1) and 149 (3348 – 3375 cm-1). Previously, the retrievals in the mesosphere were done only with order 148 (3138 measurements from 2018 to 2023) for the thermosphere, but diffraction order 149 (2880 measurements from 2018 to 2023) is now added to the retrieval pipeline. Diffraction order 148 is sensitive to CO2 density but weekly sensitive to temperature, while diffraction order 149 is highly sensitive to temperature in addition to CO2 density. Diffraction order 165 (3708 – 3738 cm-1) is used to retrieve CO2 density and temperature in the upper thermosphere (140 – 190 km). The lower bounds of the diffraction orders are due to the saturation of the lines[3]. Nevertheless, a full profile combining GEM-Mars [6, 7] and the retrieved profiles from the diffraction orders 132 and 148 (altitudes below 100 km) is provided for the retrievals of other species whose lines are dependent on temperature.We analysed the longitudinal variations of temperature for some profiles with very close solar longitude, local solar time, and latitude around 60°. Only non-migrating tides (non-synchronous with the relative position of the Sun) can be analysed as the set of profiles corresponds to tight ranges in local times: either 0 h, 9 h, or 15 h. The local times close to 0 h cover the Northern hemisphere in the first half year and the Southern hemisphere in the second half year. The local times close to 9 h and 15 h cover the Southern hemisphere in the first half year and the Northern hemisphere in the second half year. Some preliminary results concerning four sets of profiles in Martian year 35 where longitudinal variations could be inferred were presented in [8]. This analysis was now extended to more than fifty of those sets of profiles from Martian years 34 to 37. Amongst the results, we found an important wavenumber-1 structure at 9 h around LS 60° and 110° in the Southern hemisphere with very similar amplitude and phase for Martian years 35 to 37. Still, we found no structure higher than 1% of the background temperature at 15 h around LS 85° in the Southern hemisphere.Comparisons to simulations from a GCM [6, 7] show some large differences in the amplitude of longitudinal variations in the mesosphere, especially closer to aphelion in the Southern hemisphere, showing that the dynamical processes occurring at that time might still need to be better constrained. Comparisons to the results of measurements from other instruments are ongoing to confirm those results obtained with SO.

A Cloud Identification and Classification algorithm named CIC is illustrated. CIC is a machine learning method used for the classification of far and mid infrared radiances which allows to classify spectral observations by relying on small size training sets. The code is flexible meaning that can be easily set up and can be applied to diverse infrared spectral sensors on multiple platforms. Since its definition in 2019, the CIC has been applied to many observational geometries (airborne, satellite and ground-based) and is currently adopted as the scene classificator of the end-2-end simulator of the next ESA 9th Earth Explorer, the Far-infrared Outgoing Radiation Understanding and Monitoring (FORUM) which will spectrally observe the far infrared part of the spectrum with unprecedent accuracy. The algorithm has been recently improved to enhance its sensitivity to thin clouds (and also to surface features) and to increase the cloud hit rates in challenging conditions such as those characterizing the polar regions. The newly introduced metric is presented in details and the set-up procedures are discussed since they are critical for a correct application of the code. We illustrate the definition of the metric, the calibration process and the code optimization. The issues related to the definition of the reference training sets and to the classification of multiple classes are also presented.

The ExoMars Trace Gas Orbiter (TGO) mission is a joint venture of the space agencies ESA and ROSCOSMOS which was launched in 2016 and carries onboard instruments dedicated to studying the trace gas compositions of the Martian atmosphere. NOMAD (Nadir and Occultation for MArs Discovery) is one such instrument that housed three observing channels named UVIS (the Ultra Violet and Visible Spectrometer), LNO (Limb Nadir Occultation) and SO (Solar Occultation) to scan the Martian atmosphere in nadir and limb geometries [1]. The SO channel of NOMAD operates in the IR (Infra-Red) region of the solar spectrum in the wavelength range 2.3 – 4.3µm. The SO spectrometer contains an echelle grating which can produce diffraction patterns of multiple orders but only one order is allowed to fall onto the detector selected by an AOTF (Acousto Optical Tunable Filter) filter. Spectral region of diffraction orders from 186 – 191 contains well-separated and strong absorption lines of CO. The NOMAD-SO channel is using diverse diffraction orders to monitor the CO due to its importance in understanding the dynamics and chemistry of the Martian atmosphere. CO is produced in the upper Martian atmosphere by the photolysis of CO2 and destroyed by the hydroxyl (OH) radicals in the lower atmosphere. Hydroxyl radicals thus recycle CO into CO2. The study of the CO vertical distribution is important to understand the photo-chemical stability of the atmosphere. CO not only links the chemistry of the carbon and odd hydrogen families but is a long-lived species which also serves as a dynamical tracer. At IAA-CSIC we have developed a preprocessing scheme to clean the NOMAD calibrated data from a number of systematics and prepare them for inversion of different atmospheric species [2,3,4,5]. Those systematics are spectral shift of the absorption lines and spectral bending which occurs due to thermally induced mechanical stress on the detector [6]. The work presented here is in continuation with our previous work on the retrievals of CO [3] wherein the retrieval scheme has been described in detail. Our previous study reveals two crucial factors that need to be considered for a correct CO retrieval, one is the saturation of spectral lines in diffraction orders 186 and 190, those used in our work to derive CO. The second one is the use of observed temperature and pressure in the retrieval rather than the climatological T/P from GCMs (general circulation model). For order 190, the absorption lines become saturated below 70 km while for orders 186, the lines remain unsaturated for most of the atmospheric region below this altitude. In the altitudes above 70 km, the absorptions in 186 are dominated by random noise but the lines in 190, due to their strength remain clear. Due to this fact, an adequate combination of these two diffraction orders is recommended for performing CO inversions from TGO solar occultation data.In this work, we will present the improved CO vertical densities using this strategy and the impact on the CO distribution.References[1] Vandaele, A. C., Lopez-Moreno, J. J., Patel, M. R., Bellucci, G., Daerden, F., Ristic, B., ... & NOMAD Team. (2018). NOMAD, an integrated suite of three spectrometers for the ExoMars trace gas mission: Technical description, science objectives and expected performance. Space Science Reviews, 214, 1-47.[2] López‐Valverde, M. A., Funke, B., Brines, A., Stolzenbach, A., Modak, A., Hill, B., ... & NOMAD team. (2023). Martian atmospheric temperature and density profiles during the first year of NOMAD/TGO solar occultation measurements. Journal of Geophysical Research: Planets, 128(2), e2022JE007278.[3] Modak, A., López‐Valverde, M. A., Brines, A., Stolzenbach, A., Funke, B., González‐Galindo, F., ... & Vandaele, A. C. (2023). Retrieval of Martian atmospheric CO vertical profiles from NOMAD observations during the first year of TGO operations. Journal of Geophysical Research: Planets, 128(3), e2022JE007282.[4] Stolzenbach, A., López Valverde, M. A., Brines, A., Modak, A., Funke, B., González‐Galindo, F., ... & Vandaele, A. C. (2023). Martian atmospheric aerosols composition and distribution retrievals during the first Martian year of NOMAD/TGO solar occultation measurements: 1. Methodology and application to the MY 34 global dust storm. Journal of Geophysical Research: Planets, 128(11), e2022JE007276.[5] Brines, A., López‐Valverde, M. A., Stolzenbach, A., Modak, A., Funke, B., Galindo, F. G., ... & Vandaele, A. C. (2023). Water vapor vertical distribution on Mars during perihelion season of MY 34 and MY 35 with ExoMars‐TGO/NOMAD observations. Journal of Geophysical Research: Planets, 128(11), e2022JE007273.[6] Liuzzi, G., Villanueva, G. L., Mumma, M. J., Smith, M. D., Daerden, F., Ristic, B., ... & Bellucci, G. (2019). Methane on Mars: New insights into the sensitivity of CH4 with the NOMAD/ExoMars spectrometer through its first in-flight calibration. Icarus, 321, 671-690.Acknowledgements:The IAA/CSIC team acknowledges financial support from the Severo Ochoa grant CEX2021-001131-S and by grants PID2022-137579NB-I00, RTI2018-100920-J-I00 and PID2022-141216NB-I00 all funded by MCIN/AEI/ 10.13039/501100011033. A. Brines acknowledges financial support from the grant PRE2019-088355 funded by MCIN/AEI/10.13039/501100011033 and by ’ESF Investing in your future’. ExoMars is a space mission of the European Space Agency (ESA) and Roscosmos. The NOMAD experiment is led by the Royal Belgian Institute for Space Aeronomy (IASB-BIRA), assisted by Co-PI teams from Spain (IAA-CSIC), Italy (INAF-IAPS), and the United Kingdom (Open University).

Nadir and Occultation for Mars Discovery (NOMAD) onboard ExoMars Trace Gas Orbiter (TGO) started science measurements on 21 April, 2018. Here, we present results on the retrievals of water vapor vertical distributions in the Martian atmosphere from three years of TGO/NOMAD science operations. NOMAD is a spectrometer operating in the spectral ranges between 0.2 and 4.3 μm onboard ExoMars TGO. NOMAD has 3 spectral channels: a solar occultation channel (SO – Solar Occultation; 2.3–4.3 μm), a second infrared channel capable of nadir, solar occultation, and limb sounding (LNO – Limb Nadir and solar Occultation; 2.3–3.8 μm), and an ultraviolet/visible channel (UVIS – Ultraviolet and Visible Spectrometer, 200–650 nm). The infrared channels (SO and LNO) have high spectral resolution (λ/dλ~10,000–20,000) provided by an echelle grating used in combination with an Acousto Optic Tunable Filter (AOTF) which selects diffraction orders. The sampling rate for the solar occultation measurement is 1 second, which provides a good vertical sampling step (~1 km) with higher resolution (~2 km) from the surface to 200 km. Thanks to the instantaneous change of the observing diffraction orders achieved by the AOTF, the SO channel is able to measure five or six different diffraction orders per second in solar occultation mode. In this study, we analyze the solar occultation measurements at diffraction order 134 (3011-3035 cm-1), order 136 (3056-3080 cm-1), order 168 (3775-3805 cm-1), and order 169 (3798-3828 cm-1) acquired by the SO channel in order to investigate water vapor vertical distributions. Knowledge of the water vapor vertical profile is important to understand the water cycle and its escape process. Solar occultation measurements by two new spectrometers onboard TGO - NOMAD and Atmospheric Chemistry Suite (ACS) - allows us to daily monitor the water vapor vertical distributions through the whole Martian Year and obtain a good latitudinal coverage for every ~20° of Ls. In 2018, for the first time after 2007, a global dust storm occurred on Mars. It lasted for more than two months (from June to August). Moreover, following the global dust storm, a regional dust storm occurred in January 2019. The NOMAD and ACS observations therefore fully cover the majority of the global and regional dust storms and offer a unique opportunity to study the trace gases distributions during the dust storms. We analyzed those datasets and found a significant increase of water vapor abundances in the middle atmosphere (40-100 km) during the global dust storm from June to mid-September 2018 and the regional dust storm in January 2019. In particular, water vapor reaches very high altitude, at least 100 km, during the global dust storm (Aoki et al., 2019, Journal of Geophysical Research, Volume124, Issue12, Pages 3482-3497, doi:10.1029/2019JE006109). A GCM simulation explained that dust storm related increases of atmospheric temperatures suppress the hygropause, hence reducing ice cloud formation and so allowing water vapor to extend into the middle atmosphere (Neary et al., 2020, Geophysical Research Letters, accepted, Volume47, Issue7, e2019GL084354, doi: 10.1029/2019GL084354). This study presents the results with the extended dataset, which covers a full Mars year. The extended dataset newly includes aphelion season that involves interesting phenomena such as sublimation of water vapor from the northern polar cap and formation of the equatorial cloud belt, which are known as key periods to understand the large north-south hemispheric asymmetries of Mars water vapor. Yet, only a few papers report the water vapor vertical distributions in the aphelion season. The extended dataset also includes the southern summer season (dusty season) in MY 35, which will allow us to compare the water vapor distributions in the global dust storm year with those in the non-global dust storm year. In the presentation, we will discuss the water vapor vertical profiles as well as the aerosols vertical distributions retrieved from the three-year measurements of the TGO/NOMAD.