Takao M. Sato

The Venusian atmosphere serves as a natural laboratory for understanding atmospheric chemistry that cannot be observed in the terrestrial atmosphere. Different chemical processes occur at different altitudes: photochemistry driven by solar UV radiation in the upper atmosphere, cloud formation- and dissipation-related chemistry in the cloud layer, and thermodynamic equilibrium controlled by high temperatures in the lower atmosphere. Another categorization can be made in terms of chemical cycles, such as, sulfur oxides (SOx), carbon oxides (COx), and chlorides (Clx). Observational constraints on the spatial and temporal variability of trace gases associated with these cycles can help us better understand atmospheric chemistry.We have observed high-resolution (R ~ 75,000) spectra of the Venusian dayside using iSHELL mounted on the 3.2-m NASA Infrared Telescope Facility (IRTF) on Mauna Kea since 2018. The iSHELL instrument is a cross-dispersed high-resolution echelle spectrograph operating over the wavelength range of 1.06–5.3 μm. Twenty standard observing modes with different wavelength coverages are available by selecting one of the six cross-dispersing gratings (J, H, K, L, L’, and M) and adjusting the tilt position. By fully exploiting the capabilities of iSHELL, we can retrieve the abundances of HDO, CO, HF, HCl, OCS, and their isotopologues near the Venusian cloud tops. To date, we have accumulated 23 days of successful observations over four observing periods (2018, 2020, 2023, and 2025). Among them, our first results focusing on the HCl abundance near the cloud tops have already been published (Sato and Sagawa, 2023).Because the observations were conducted using various observing modes and slit lengths after the publication, we revisited and refined the data calibration methods and radiative transfer model. For example, we developed a method to estimate the slit position on the Venusian disk from guide images of Venus taken with the 3.46-μm filter by fitting the limb of the Venusian dayside disk, independent of the slit length (5”, 15”, or 25”). In addition, we estimated the instrument response function for each observing mode using multiple Ar-Th lamp emission lines fitted with a combination of two Gaussian functions: one symmetric and the other asymmetric. The molecular absorption database used in the radiative transfer model was updated to HITRAN2024. In this presentation, we provide an overview of our observations conducted so far. We will present initial results on the retrieval of CO and its isotopologues near the cloud tops.

Observations of Venus' night-side disc in the transparency windows of the CO2 atmosphere (spectacular at 1.74 and 2.3 μm near-infrared wavelengths) allow us to study various morphology and dynamics in the middle to lower cloud layers of Venus. For such studies, Akatsuki, Japan's Venus Orbiter was equipped with the IR2 2-μm camera (Satoh, et al., 2017). A sharp and large discontinuity of cloud opacities was imaged in the night-side of Venus by IR2 (Peralta et al., 2020). This feature was best observed by IR2 in 4 occasions, 9 August, 18 August, 27August, and 5 September in 2016. Since the faint night-side emission features receive contamination from intense light from the day crescent, we have developed image restoration techniques which successfully recovered the "true" contrast between this enormous cloud cover (ECC) and the adjacent background cloud (BC) regions. From such restored images, the radiance from pixels in BC and ECC from each image were extracted. By conducting a series of radiative transfer computations, we compare the observed brightness changes (from BC to ECC) in two filters of IR2 (1.735 and 2.26 μm). It is found that an increase of Mode 3 particles near the cloud base (∼48 km altitude) can reproduce the decreasing radiance from BC to ECC for 9 August and 27 August data. On the other hand, the 18 August data require both the increase of Mode 3 particles AND decrease of smaller particles at the same time to explain the observation. Finally, the 5 September data do not need the increase of Mode 3 particles but slight increase of smaller particles explains the data. Although these are not a unique explanation of how the cloud structure changes from BC to ECC region, this seems to be a favorable characterization of its lifecycle. Implications to the possible mechanism will also be discussed.

Thermal infrared imaging by Longwave Infrared Camera (LIR) aboard JAXA's Venus orbiter Akatsuki has revealed horizontal structures of large‐scale topographic gravity waves (mountain waves) and thermal tides in the Venusian atmosphere. For quantitative analysis of these waves, we developed a radiative transfer model for an atmosphere perturbed by a gravity wave, which represents the mountain waves and the thermal tides in the equatorial region. Combining the infrared images with temperature profiles from the Akatsuki radio occultation, the cloud particle scale height, as well as the vertical wavelengths and growth rates of mountain waves and the semidiurnal thermal tide were determined. The cloud particle scale height around the cloud top was estimated to be approximately equal to the atmospheric scale height, indicating a vertically well‐mixed layer of the cloud particles near the cloud top in low latitudes. It was demonstrated that the brightness temperature amplitude observed by LIR is approximately half of the atmospheric temperature amplitude at the cloud top. The waves' impact on the mean flow above the cloud top was assessed by estimating their momentum deposition. The results show that mountain waves can induce an intense but localized deceleration of tens of m s ⁻¹ /day or larger above 80 km, while the semidiurnal tide provides a global and persistent deceleration that increases with height to reach the order of 1 m s ⁻¹ /day at 80 km. These results reveal two important but distinct mechanisms that contribute to the momentum budget of the Venusian upper atmosphere.

Planetary‐scale waves are expected to be crucial in driving the Venusian planetary‐scale atmospheric circulation, including the superrotation. To understand the interaction between the waves and the mean flow, we obtained temporal frequency spectra of the cloud‐top brightness temperature using thermal infrared images taken by the Longwave Infrared Camera (LIR) onboard Akatsuki over a period of 10 Venus years. Waves in the equatorial region with periods of around 3.5–4.3 days were identified as Kelvin waves, while waves in the mid‐latitude region with periods of about 5.0–6.0 days were identified as Rossby waves. The mid‐latitude waves with periods 5.0–6.0 days tend to accompany additional local amplitude maxima near the equator, especially when observed at small emission angles. Considering that the contribution function of LIR extends to lower altitudes for smaller emission angles, the result implies that the waves arise from Rossby‐Kelvin instability and the associated Kelvin modes reside below the cloud top. Mid‐latitude peaks are also sometimes seen around periods of 3.5–4.0 days and are coupled with equatorial modes, indicative of Rossby‐Kelvin instability. The coupled Rossby‐Kelvin modes are expected to transport angular momentum equatorward to sustain the superrotation. The mid‐latitude modes decay with altitude. The periods and amplitudes of the waves change with time, and the variations seem to correlate with the background wind in such a way that waves with small intrinsic frequencies are less prominent.

Venus is completely shrouded by optically thick clouds of sulfuric acid that are located between ~47 and 70 km. The cloud tops have been investigated through imaging, spectroscopy, and polarimetry in a broad range of wavelengths, from UV to mid-infrared, as well as in-situ measurements. For example, Ignatiev et al. (2009) studied the cloud top altitude from the depth of CO2 absorption band at 1.6 μm acquired by VIRTIS onboard Venus Express. They found that the cloud tops decreased poleward of 50° and this depression coincided with the eye of the polar vortex. Sato et al. (2020) described the dayside cloud top structure of Venus using the 2.02-μm channel (designed for sensing at a CO2 absorption band) of the 2-μm camera (IR2) onboard Akatsuki. They showed that the latitudinal structure of the cloud top altitude was symmetric with respect to the equator. The average cloud top altitude was 70.5 km in the equatorial region and showed a gradual decrease of ~2 km by the 45° latitude. It rapidly dropped at latitudes of 50–60° and reached 61 km in latitudes of 70–75°.We have investigated the solar phase angle dependence of the reflected sunlight in the low-latitude region (30°S–30°N) using the complete set of 2.02-μm images. A total of 374 images taken from December 11, 2015, to October 29, 2016 were selected by carefully excluding those with saturated pixels and/or defect image tiles. In general, the reflected sunlight gets more intense as solar phase angle increases due to the forward scattering of aerosols in the atmosphere. Interestingly, the curve of the reflected sunlight in solar phase angles higher than 90° acquired in Orbit #29–30 was approximately twice more gradual than that acquired in Orbit #11–12. Because these data were taken under the similar geometry, the difference is likely to be caused by some real variations in the cloud top structure. Such solar phase angle dependence of the reflected sunlight or brightness temperature in the same region was also studied using images taken in the same observation period but at other wavelengths: 283 nm and 365 nm from the Ultraviolet imager (UVI), 0.9 μm from the 1-μm camera (IR1), and 10 μm from the Longwave infrared camera (LIR). This multispectral comparison showed that no significant difference in the curve between Orbit #11–12 and #29–30 was detected at the wavelengths other than 2.02 μm. Wavelengths at 283 nm, 365 nm, and 0.9 μm are sensitive to optical thickness of aerosols but insensitive to their vertical distributions; therefore, vertical distributions of Modes 1 and 2 particles characterized by scale height are key parameters to make the difference in the curve of the reflected sunlight discovered from the 2.02-μm images. To support this hypothesis, we selected the data to make two groups: Group 1 (including Orbit #11–12) and Group 2 (including Orbit #29–30). For each group, the observed solar phase angle dependence and the center-to-limb variation of the reflected sunlight in the low-latitude region were used to retrieve cloud top altitude zc and cloud scale height H by means of radiative transfer calculation. While the best-fit combination for Group 1 is obtained at (zc, H) = (70.4 km, 5.5 km), that for Group B is found at (zc, H) = (69.9 km, 3.5 km). The results quantitatively show that the difference can be explained by the 2-km difference of cloud scale height.In this presentation, we present the solar phase angle dependence of the reflected sunlight or brightness temperature using the five wavelengths of Akatsuki’s instruments. The details of the radiative transfer model and fitting results are shown. Finally, we will discuss whether the best-fit models retrieved from Groups 1 and 2 are compatible with the data at other wavelengths.

The measurement of isotopic ratios on Venus is crucial to track atmospheric loss though time, constrain planetary origin and trace surface-atmosphere interactions [1]. The deuterium to hydrogen ratio (D/H) is found to be 157 times greater than that of Earth's oceans [2, 3], suggesting Venus once had a larger quantity of water. The carbon and oxygen isotopic ratios, ¹²C/¹³C and ¹⁶O/¹⁸O, have been measured by the Pioneer Venus and Venera missions [4, 5], as well as through ground-based infrared spectroscopy [6, 7, 8, 9]. They are consistent with Earth’s ratios within uncertainties (¹²C/¹³C ~ 89, ¹⁶O/¹⁸O ~ 500), suggesting little escape or photochemical alteration of CO2 through time.We present new simultaneous measurements of bulk-averaged ratios of ¹⁶O/¹⁸O, ¹6O/17O and ¹²C/¹³C, measured on CO2, on Venus’ dayside, from high spectral resolution (R ~ 80 000) iSHELL/IRTF observations, using the H3 mode (1.64-1.82 µm), performed on the 2nd February 2024. Venus angular diameter was 12.2'', with an illuminated fraction of 86.1 %. Wavelength calibration, flat fielding, dark correction and sky subtraction were performed using the Spextool data reduction software [10, 11]. Absolute flux calibration could not be performed since the SNR of our standard star was too low.We selected spectral order 314 (1.660–1.667 µm) to estimate the ¹⁶O/¹⁸O ratio, order 316 (1.650–1.657 µm) for the ¹⁶O/17O ratio, and order 311 (1.676–1.683 µm) to determine the ¹²C/¹³C ratio. Pairs of adjacent isotopologue lines were selected, with a minimum telluric and solar contamination. Forward models were simulated using ASIMUT-ALVL [12] for a range of possible isotopic ratios (scalers 0.85-1.15). For each line pair, the line depth ratio method [7, 13, 14] was used to obtain the best fit isotopic ratio, by comparison of the measured line depth ratio with the modeled one through a simple linear interpolation. The following preliminary weighted averaged isotopic ratios were computed: ¹⁶O/¹⁸O ~ 507 ± 50, ¹⁶O/17O ~ 2745 ± 152 and ¹²C/¹³C ~ 90 ± 6, which are consistent with previously reported values [7, 8, 9]. The main limitations of our current measurements arise from uncertainties in line intensities (approximately 5%, based on the HITRAN database), the lack of an absolute flux calibration source with sufficiently high signal-to-noise ratio and uncertainties in the assumed temperature profile in the forward modeling. Ongoing analysis of data from mode H3 (1.64-1.82 µm), and from mode K3 (2.26-2.55 µm), which covers the OCS isotopologues absorption features at 2.42-2.46 µm, will be explored to get an estimate of the ratios of 35Cl/37Cl and 32S/34S, respectively, opening an opportunity to shed some light on current volcanic outgassing processes, photochemistry and sulphur cycle dynamics.References. (1) Avice et al. 2022, Space Science Reviews (2) Donahue et al., 1997, Science (3) Bézard et al. 2007, JGR Planets (4) Hoffman et al. 1980, JGR Space Physics (5) Istomin et al. 1980, 23rd COSPAR meeting, Budapest, Hungary (6) Kuiper et al. 1962, Comm. Lunar Planet. Lab (7) Beźard et al. 1987, Icarus (8) Hedelt et al. 2011, A&A (9) Iwagami et al. 2015, Planetary and Space Science (10) Cushing et al. 2004, PASP 116 (11) Vacca et al. 2003, PASP 115 (12) Vandaele et al. 2006, ESRIN, Italy (13) Krasnopolsky et al. 2010, Icarus (14) Encrenaz et al. 2012, A&AFunding. JAD acknowledges funding through the research grants UIDB/04434/2020 and UIDP/04434/2020 and a fellowship grant 2022.09859.BD. SR acknowledges funding by the Belgian Science Policy Office (BELSPO) with the financial and contractual coordination by the ESA Prodex Office (PEA 4000144206).

Planetary-scale waves are expected to be crucial in driving the Venusian planetary-scale atmospheric circulation. To understand the interaction between the waves and the mean flow, we obtained temporal frequency spectra of the cloud-top brightness temperature using thermal infrared images taken by the Longwave Infrared Camera (LIR) onboard Akatsuki over a period of 10 Venus years. Waves in the equatorial region with periods of around 3.5–4.3 days were identified as Kelvin waves, while waves in the mid-latitude with periods of about 5.0–6.0 were identified as Rossby waves. The mid-latitude waves with periods 5.0–6.0 days tend to accompany additional local amplitude maxima near the equator, especially when observed at small emission angles. Considering the contribution function of LIR extends to lower altitudes for smaller emission angles, the result implies the waves arise from Rossby-Kelvin instability and the associated Kelvin modes reside below the cloud top. Mid-latitude peaks are also sometimes seen around periods of 3.5–4.0 days and coupled with equatorial modes, indicative of Rossby-Kelvin instability. The periods and amplitudes of the waves change with time, and the variations seem to correlate with the background wind in such a way that waves with small intrinsic frequencies are less prominent.Figure: Latitudinal distributions of the wave amplitudes at the period of (a) 5.04 days and (b) 3.71 days for the emission angle ranges of 30–40 (blue solid), 40–50 (orange dashed), and 50–60 (green dotted).

Venus is completely shrouded by optically thick clouds of sulfuric acid that are located between ~47 and 70 km. The cloud tops have been investigated through imaging, spectroscopy, and polarimetry in a broad range of wavelengths, from UV to mid-infrared, as well as in-situ measurements. For example, Ignatiev et al. (2009) studied the cloud top altitude from the depth of CO2 absorption band at 1.6 μm acquired by VIRTIS onboard Venus Express. They found that the cloud tops decreased poleward of 50° and this depression coincided with the eye of the polar vortex.Sato et al. (2020) described the dayside cloud top structure of Venus as retrieved from 93 images acquired at a wide variety of solar phase angles (0–120°) using the 2.02-μm channel of the 2-μm camera (IR2) onboard the Venus orbiter, Akatsuki, from April 4 to May 25, 2016 (Orbit #12–16). Since the 2.02-μm channel is located in a CO2 absorption band, the sunlight reflected from Venus allowed us to determine the cloud top altitude corresponding to a unit aerosol optical depth at 2.02 μm. First, the observed solar phase angle dependence and the center-to-limb variation of the reflected sunlight in the low-latitude region (30°S–30°N) were used to construct a spatially averaged cloud top structure characterized by cloud top altitude zc, Mode 2 particle radius rg,2, and cloud scale height H, which were 70.4 km, 1.06 μm, and 5.3 km, respectively. Second, cloud top altitudes at individual locations were retrieved on a pixel-by-pixel basis with an assumption that rg,2 and H were uniform for the entire planet. The latitudinal structure of the cloud top altitude was symmetric with respect to the equator. The average cloud top altitude was 70.5 km in the equatorial region and showed a gradual decrease of ~2 km by the 45° latitude. It rapidly dropped at latitudes of 50–60° and reached 61 km in latitudes of 70–75°.In this study, we focused on investigating the solar phase angle dependence of the reflected sunlight in the low-latitude region (30°S–30°N) using the complete set of 2.02-μm images: a total of 374 images taken from December 11, 2015, to October 29, 2016 were selected by carefully excluding those with saturated pixels and/or defect image tiles. In general, the reflected sunlight gets more intense as solar phase angle increases due to the forward scattering of aerosols in the atmosphere. Interestingly, the curve of the reflected sunlight in solar phase angles higher than 90° acquired in Orbit # 29–30 was approximately twice more gradual than that acquired in Orbit #11–12. Because these data were taken under the similar geometry, the difference is likely to be caused by some real variations in the cloud top structure. Such solar phase angle dependence of the reflected sunlight or brightness temperature in the same region was also studied using images taken in the same observation period but at other wavelengths: 283 nm and 365 nm from the Ultraviolet imager (UVI), 0.9 μm from the 1-μm camera (IR1), and 10 μm from the Longwave infrared camera (LIR). This multispectral comparison showed that no significant difference in the curve between Orbit #11–12 and #29–30 was detected at the wavelengths other than 2.02 μm. Wavelengths at 283 nm, 365 nm, and 0.9 μm are sensitive to optical thickness of aerosols but insensitive to their vertical distributions; therefore, vertical distributions of Modes 1 and 2 particles characterized by scale height are key parameters to make the difference in the curve of the reflected sunlight discovered from the 2.02-μm images. In addition, wavelength at 10 μm is less sensitive to Mode 1 particles because their size is small compared with wavelength. These facts imply that such difference is caused by vertical distribution of Mode 1 particles.In this presentation, we present the solar phase angle dependence of the reflected sunlight or brightness temperature using the five wavelengths of Akatsuki’s instruments. While Sato et al. (2020) reproduced the solar phase angle dependence and the center-to-limb variation of the reflected sunlight by Mode 2 particle size and its cloud scale height (common with Mode 1 particles), free parameters in this study are the scale height of Mode 1 particles (scale height of Mode 2 particle is fixed to the value used by Haus et al., 2015) and the abundance ratio of Mode 1 and 2. Trial-and-error results to reproduce the difference in the curve only obtained from the 2.02-μm images using radiative transfer calculation will be discussed.

The atmosphere of Venus can be vertically divided into three regions with different chemical conditions. High temperature and pressure and the absence of effective photolysis processes are dominant in the lower atmosphere up to 60 km. The middle atmosphere between 60 and 110 km is controlled by photochemistry driven by solar UV radiation. In the upper atmosphere above 110 km, dissociation, ionization, and ionospheric reactions are important processes.HCl is the primary chlorine reservoir in the Venus’ atmosphere below 110 km. Highly reactive chlorine species (ClOx) is produced by solar UV photolysis of HCl and has been proposed to play an important role in catalysis of CO and O recombination to CO2, thereby stabilizing the CO2 atmosphere. Chlorine chemistry is also linked to sulfur chemistry and its understanding is necessary to explain the observed vertical distribution of SO2.Interestingly, there is a large inconsistency between the HCl abundances measured by spacecraft and ground-based telescopes. The SOIR instrument onboard Venus Express measured its abundance as less than ~50 ppb at the cloud top (~70 km) increasing with altitude, reaching to 1 ppm in the upper atmosphere (~110 km) [Mahieux et al., 2015]. Such a vertical trend conflicts with the results obtained by sub-mm ground-based observations which inferred a vertically constant profile (up to ~80 km) [Sandor and Clancy, 2012]. Near-infrared ground-based observations also showed the HCl abundance at the cloud top as ~500 ppb [Iwagami et al., 2008; Krasnopolsky, 2010], which are nearly one order of magnitude larger than the SOIR results. The reason for this inconsistency has not been understood yet.In order to re-examine HCl abundance at the cloud top, we carried out a high-resolution spectroscopy of Venus’ dayside at wavelengths of 3.580-3.934 μm with IRTF/iSHELL on August 5-7, 2018 and August 18-20, 2020 (UT). Venus was near its greatest eastern and western elongations, respectively, in the observation periods. Taking the full advantages of iSHELL’s high spectral resolution of R ~ 75,000, iSHELL resolved individual HCl lines with sufficient separation from terrestrial lines. We analyzed three cross-dispersed echelle orders (orders 141, 142, and 144). For each order, retrievable lines of HCl35, HCl37, and O16C12O18 were included. With using radiative transfer modeling, HCl35 and HCl37 abundances were derived after cloud top altitude was retrieved from several O16C12O18 lines. Our preliminary results showed that HCl abundance at the cloud top is at least larger than 100 ppb and does not vary with the observation period (i.e., no difference between the morning and evening hemispheres).In this presentation, we show latitudinal distribution of HCl abundance and its isotopic (HCl35/HCl37) ratio at the cloud top, retrieved from the iSHELL spectra and compare them with the previous studies.

The 2-µm infrared camera (IR2) onboard Akatsuki observed a remarkable cloud feature in the Venus' night-side disk, a sharp discontinuity of cloud opacity which subtends latitudinally to some thousands of km (Peralta et al., 2020).  Though obvious and seemingly common in the Venus' atmosphere as similar features can be identified in imagery since the beginning of the night-side observations (Allen and Crawford, 1984), the mechanism of this emormous cloud cover (ECC) has not yet been explained. To characterize this ECC (aerosol size parameters and column numbers), we have analyzed the Akatsuki/IR2 data, as well as the Venus Express/VIRTIS-M data. Six sets of the Akatsuki/IR2 data (MM-DD = 03-27, 07-22, 08-09, 08-18, 08-27, and 09-06) are measurable with varying photometric uncertainties, due to contaminations from the intense day crescent. Seven VEx/VIRTIS data, as tabulated in Peralta et al. (2020), are also measured by the consistent method with that for IR2 data.A reference region, which is just west of the discontinuity and is seemingly not affected by the ECC, is defined as the background cloud (BC) region. Radiances at the BC and the ECC regions are measured for two IR2 filter passbands (1.735 and 2.26 µm). They are plotted in the correlation plot (radiance at 2.26 µm in horizontal axis and radiance at 1.735 µm in vertical axis). The BC-to-ECC slope can be used to infer the aerosol size and abundance that changes the BC region to the ECC region. Comparison of obtained characteristics of the ECC for different observing times will be presented and implication to the possible mechanism of this large-scale phenomenon will be discussed.