Abstract We describe a Geometry‐dependent surface Lambertian‐Equivalent Reflectivity (GLER) climatology developed to support operational retrievals for the Tropospheric Emissions: Monitoring of Pollution (TEMPO) mission. This data set provides monthly and hourly surface reflectance for snow‐free land, snow‐covered land, and oceanic regions. The set is constructed by integrating MODIS MCD43C1/C2 bidirectional reflectance distribution function (BRDF) parameters, snow cover information, and radiative transfer simulations. A Factor‐Analysis spectral reconstruction method, constrained by spectral priors from the USGS spectral library and the SCIAMACHY LER data set, was applied to extend discrete MODIS four‐band reflectance over land to a near‐continuous spectral range (335–900 nm), covering most of the spectral range of TEMPO. The ocean climatology GLER was simulated using a Cox–Munk surface slope distribution model coupled with the VLIDORT model. Preliminary comparisons with the TROPOMI directionally dependent LER (DLER) empirical data set demonstrate strong seasonal consistency at a local solar time of 13:30. Summer mean differences generally remain within ±0.01, while larger winter biases (up to 0.03) are primarily attributed to extreme solar geometries, sub‐pixel residual snow, and adjacency pixel effects. This GLER climatology currently serves as a crucial input for TEMPO Level 2 operational algorithms, supporting retrievals for nitrogen dioxide, formaldehyde, and ozone, as well as cloud parameter estimations.

On 15 January 2022, a highly explosive eruption of the submarine Hunga volcano (Kingdom of Tonga) generated the largest stratospheric hydration event ever observed and the largest aerosol perturbation since the 1991 Pinatubo eruption. Here, we develop a novel method for satellite retrieval of stratospheric aerosol optical depth (AOD) and layer peak height (zp) using solar backscattered ultraviolet (BUV) radiation; this is made possible by the exceptional mid-stratospheric altitude of the Hunga aerosols. We analyze BUV observations of the Hunga stratospheric aerosol cloud on 17 January 2022 (47 h after the eruption), using BUV band 1 measurements from the TROPOspheric Monitoring Instrument (TROPOMI) on board the ESA/Copernicus Sentinel-5 precursor (S5P) satellite, and the Ozone Mapping and Profiling Suite- Nadir Profiler (OMPS-NP) on board the National Oceanic and Atmospheric Administration (NOAA)-20 satellite. We retrieve AOD and zp by fitting hyperspectral BUV radiance ratios in a narrow spectral window restricted to 289–296 nm, chosen in order to reduce interference from tropospheric clouds while highly sensitive to stratospheric aerosols located above ozone peak altitude. The retrieval employs radiative transfer calculations from the Vector Linearized Discrete Ordinate Radiative Transfer (VLIDORT) forward model. We assume a single Hunga aerosol layer composed of polydisperse sulfuric acid spherical particles embedded in a Rayleigh atmosphere with a known ozone profile. The ozone profile is supplied from a version of the Modern-Era Retrospective analysis for Research and Applications, version 2 (MERRA-2) Stratospheric Composition Reanalysis of the Microwave Limb Sounder (MLS) on board NASA Earth Observing System-chemistry (EOS Aura) satellite – produced by NASA's Global Modeling and Assimilation Office using a stratospheric chemistry model and MERRA-2 meteorology. We also include a sulfur dioxide SO2 layer, which coincides spatially with the retrieved aerosol vertical profile, and with the total loading normalized to the stratospheric SO2 vertical column density from the operational TROPOMI SO2 product. We validate our AOD retrievals against ground-based AErosol RObotic NETwork (AERONET) direct-sun AOD measurements, and zp retrievals against Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) overpasses using Lagrangian trajectory modeling. We estimate the total Hunga stratospheric wet aerosol mass (sulfuric acid solution droplets, including water uptake) to be Maer∼0.5±0.05 Tg. This value is consistent with our previous BUV estimates of Hunga SO2 emissions (∼ 0.4–0.5 Tg SO2) and rapid conversion of SO2 to sulfate aerosol. Based on these BUV retrievals we can also estimate the sulfuric acid (H2SO4) mass fraction w∼0.4 and H2SO4/H2O solution density: ρ∼1.34 g cm−3. These new values represent an extreme departure from the stratospheric background sulfate aerosol (Junge layer), which is typified by values of w∼0.75 and ρ∼1.7 g cm−3 supported by decades of observations of the lower stratosphere during both quiescent and volcanically impacted periods. The new low values, inferred from BUV observations and backed up by microphysical modeling, are a result of the uniquely water-rich conditions in the early Hunga plume. Relative humidity in the plume, as modeled by the NASA Goddard Earth Observing System Chemistry-Climate Model with the Community Aerosol and Radiation Model for Atmospheres (CARMA), reached values as high as 60 %, compared to background values closer to 1 %. These findings are unique in the long-term observational record of the stratosphere; similar relative humidities only otherwise occur in overshooting clouds or cold winter hemisphere vortices.

Retrievals of ocean color (OC) properties from space are important for better understanding the ocean ecosystem and carbon cycle. The launch of atmospheric hyperspectral instruments such as the geostationary Tropospheric Emissions: Monitoring of Pollution (TEMPO) and GEMS, provide a unique opportunity to examine the diurnal variability in ocean ecology across various waters in North America and prepare for the future suite of hyperspectral OC sensors. While TEMPO does not have as high spatial resolution or full spectral coverage as planned coastal ocean sensors such as the Geosynchronous Littoral Imaging and Monitoring Radiometer (GLIMR) or GeoXO OC instrument (OCX), it provides hourly coverage of US coastal regions and great lakes, such as Lake Erie and the Gulf of Mexico at spatial scales of approximately 5 km. We will apply our newly developed machine learning (ML) based atmospheric correction approach for OC retrievals to TEMPO data. Our approach begins by decomposing measured hyperspectral radiances into spectral features that explain the variability in atmospheric scattering and absorption as well as the underlying surface reflectance. The coefficients of the principal components are then used to train a neural network to predict OC properties such as chlorophyll concentration derived from collocated MODIS/VIIRS physically-based retrievals. This ML approach compliments the standard radiative transfer-based OC retrievals by providing gap-filling over cloudy regions where the standard algorithms are limited. Previously, we applied our approach using blue and UV wavelengths with the Ozone Monitoring Instrument (OMI) and TROPOspheric Monitoring Instrument (TROPOMI) to show that it can estimate OC properties in less-than-ideal conditions such as lightly to moderately cloudy conditions as well as sun glint and thus improve the spatial coverage of ocean color measurements. TEMPO provides an opportunity to improve on this approach since it provides extended spectral measurements at green and red wavelengths which are important particularly for coastal waters. Additionally, our ML technique can be applied to provisional data early in the mission and has potential to demonstrate the value of near real time OC products that are important for monitoring of harmful algae blooms and transient oceanic phenomena.

We present the first sulfur dioxide (SO 2 ) retrievals from Tropospheric Emissions: Monitoring of Pollution (TEMPO), the first geostationary atmospheric composition sensor to cover North America, along with some potential applications of TEMPO SO 2 data. We show that high resolution (∼10 km 2 ) TEMPO measurements can be used to produce good quality SO 2 retrievals with relatively small noise and biases. We demonstrate that hourly TEMPO data are useful for monitoring volcanic hazards, by providing frequent updates on the plume location and additional information on the plume height or winds. With the large number of measurements from TEMPO, it is also feasible to monitor diurnal changes in SO 2 for relatively large sources such as the Cantarell oil field. We also show that high‐cadence TEMPO measurements allow estimates of SO 2 degassing from Popocatépetl volcano on sub‐daily timescales. Overall, our results suggest that TEMPO can significantly enhance space‐based SO 2 detection and monitoring over North America.

Ultraviolet (UV) radiation effects on Earth’s ecosystems on a global scale can be assessed on a basis of satellite estimates of hyperspectral irradiance on the surface and in ocean waters and the spectral biological weighting functions. The satellite UV surface irradiance algorithms combine satellite retrievals of extraterrestrial solar irradiance, cloud/surface reflectivity, aerosol optical depth, and total column ozone with radiative transfer computations. The assessment of in-water irradiance requires additional information on inherent optical properties (IOPs) of ocean water. Our Ozone Monitoring Instrument (OMI) surface hyperspectral irradiance algorithm is updated by implementing a new absorbing aerosol correction based on OMI daily retrievals of UV aerosol absorption optical depth (AAOD). To provide insight into the temporal and spatial variability of absorbing aerosols, we consider a monthly global AAOD climatology derived from the OMI UV aerosol algorithm. Hyperspectral underwater irradiance is computed using Hydrolight radiative transfer calculations along with a Case I water model of IOPs extended into UV. Both planar and scalar irradiances are computed on the Earth’s surface and propagated underwater. The output surface products include the UV index. The output underwater products include the hyperspectral diffuse attenuation coefficients of the planar and scalar irradiances. Effects of the seasonal variability of AAOD on the UV index and the deoxyribonucleic acid (DNA) damage dose rates are considered. The reduction in the UV index and DNA damage dose rate due to the presence of absorbing aerosols can be as large as 30–40%.

Nitrogen oxides are key gas components of emissions from fossil-fuel combustion, are known to degrade air quality and have adverse health effects. Diurnal NO2 observations are crucial for enhancing our understanding of NOx emissions, lifetime, and chemistry. Geostationary Environment Monitoring Spectrometer (GEMS) has been providing hourly observations NO2 columns over Asia since November 2020. The latest NO2 version 3 products have significantly improved with updated air mass factors (AMFs) and the separation of stratospheric and tropospheric columns. To identify the dependency of the distribution on the time of the day, we investigated hourly tropospheric NO2 cycles of cities over Asia using GEMS measurements for the first time. The cities show similar diurnal concentration patterns with peaks in the morning and troughs in the afternoon, although the amplitude and specific times vary by city. The reduction rate of NO2 was influenced by the temporal dependence of the spatial distribution within and around cities. We also observed distinct NO2 diurnal patterns in certain industrial areas and cities where NOx emissions are thought to be controlled. To explain the location-dependent variations of the tropospheric NO2 columns, we compared the diurnal NO2 cycles obtained from the GEMS measurement with WRF-Chem models for some cities. In addition, estimated top-down NOx emissions from GEMS measurements are presented in comparison with bottom-up emission inventory, showing a smaller difference compared to the top-down emission from TROPOMI measurements. It is expected that hourly top-down NOx emissions using GEMS measurements can provide a useful information in improving the future performance of air quality modeling.

The oil sands in Alberta, Canada, are a significant source of air pollution. Observations from the Ozone Monitoring Instrument (OMI) on the NASA Aura satellite have been used to quantify NOx emissions from the surface mining region of the oil sands. Two related emissions methods were utilized, one for point and one for area sources, where OMI vertical column densities of NO2 were combined with winds from a meteorological reanalysis and a two-dimensional exponentially modified Gaussian (EMG) plume model. This work better connects the two (point and area) emissions methods and discusses the interpretation of fit parameters and the ability of OMI (and other sensors) to resolve emissions between neighbouring sources. The two methods employed, in good agreement with each other, indicated an increase in emissions from about 55 to 80 kt [NO2] yr−1 between 2005–2011 and a flat trend thereafter. Reported emissions were within 15 % of reported emissions, consistent to within uncertainties. In an extension of this methodology, OMI observations were combined with reported point source emissions to derive the more uncertain emissions component from the large off-road mining fleet. These were found to make up about 60 % of total NOx emissions, also consistent with reported emissions. The OMI-derived 0.9 % yr−1 increase in fleet emissions and the 5.5 % yr−1 increase in bitumen mined, generally a good proxy for fleet emissions, can be reconciled by considering the evolution of the mine fleet over this period. OMI is therefore able to track the transition from US EPA Tier 1 standards, through Tier 4 standards, to the present and in doing so demonstrates the efficacy of this policy. Furthermore, this analysis shows that had the fleet remained at Tier 1, this source would currently be emitting an additional 40 kt [NO2] yr−1.

Satellite measurements of nitrogen dioxide have been used to infer nitrogen oxide emissions, a critical component in tropospheric chemistry and pollution. New observations from the Geostationary Environmental Monitoring Spectrometer offer a breakthrough by providing a daytime record of nitrogen dioxide over Asia. Here we present the summertime diurnal patterns of nitrogen dioxide at major cities, power plant regions, and the Strait of Malacca. The Geostationary Environmental Monitoring Spectrometer data across various regions show high nitrogen dioxide in the morning which decrease in the afternoon, with varying hourly peaks, troughs, and amplitudes reflecting diurnal characteristics of local emissions and chemistry. Nitrogen oxide emissions inferred from Geostationary Environmental Monitoring Spectrometer and the Weather Research and Forecasting model coupled with Chemistry also show distinct patterns among regions: early morning peaks occur over Hanoi, Guangzhou, and Bangkok; mid-to-late morning peaks appear over Seoul and Beijing; and late afternoon peaks are noted in the Yangtze River Delta region. Top-down emissions incorporating temporal changes in the Geostationary Environmental Monitoring Spectrometer nitrogen dioxide yield the most accurate nitrogen dioxide simulations.

For nearly 2 decades, the Ozone Monitoring Instrument (OMI) aboard the NASA Aura spacecraft (launched in 2004) and the Ozone Mapping and Profiler Suite (OMPS) aboard the NASA/NOAA Suomi National Polar-orbiting Partnership (SNPP) satellite (launched in 2011) have been providing global monitoring of SO2 column densities from both anthropogenic and volcanic activities. Here, we describe the version-1 NOAA-20 (N20)/OMPS SO2 product (https://doi.org/10.5067/OMPS/OMPS_N20_NMSO2_PCA_L2_Step1.1, Li et al., 2023), aimed at extending the long-term climate data record. To achieve this goal, we apply a principal component analysis (PCA) retrieval technique, also used for the OMI and SNPP/OMPS SO2 products, to N20/OMPS. For volcanic SO2 retrievals, the algorithm is identical between N20 and SNPP/OMPS and produces consistent retrievals for eruptions such as Kilauea in 2018 and Raikoke in 2019. For anthropogenic SO2 retrievals, the algorithm has been customized for N20/OMPS, considering its greater spatial resolution and reduced signal-to-noise ratio as compared with SNPP/OMPS. Over background areas, N20/OMPS SO2 slant column densities (SCDs) show relatively small biases, comparable retrieval noise with SNPP/OMPS (after aggregation to the same spatial resolution), and remarkable stability with essentially no drift during 2018–2023. Over major anthropogenic source areas, the two OMPS retrievals are generally well-correlated, but N20/OMPS SO2 is biased low, especially for India and the Middle East, where the differences reach ∼ 20 % on average. The reasons for these differences are not fully understood but are partly due to algorithmic differences. Better agreement (typical differences of ∼ 10 %–15 %) is found over degassing volcanoes. SO2 emissions (https://doi.org/10.5067/MEASURES/SO2/DATA406, Fioletov et al., 2022) from large point sources, inferred from N20/OMPS retrievals, agree well with those based on OMI, SNPP/OMPS, and the TROPOspheric Monitoring Instrument (TROPOMI), with correlation coefficients >0.98 and overall differences <10 %. The ratios between the estimated emissions and their uncertainties offer insights into the ability of different satellite instruments to detect and quantify SO2 sources. While TROPOMI has the highest ratios of all four sensors, the ratios from N20/OMPS are slightly greater than OMI and substantially greater than SNPP/OMPS. Overall, our results suggest that the version-1 N20/OMPS SO2 product will successfully continue the long-term OMI and SNPP/OMPS SO2 data records. Efforts currently underway will further enhance the consistency of retrievals between different instruments, facilitating the development of multi-decade, coherent global SO2 datasets across multiple satellites.

The TROPOspheric Monitoring Instrument (TROPOMI) observed unprecedented declines in NO2 vertical column densities (VCD) over the world's most densely populated cities during the 2020 COVID-19 lockdowns. These favorable changes in NO2 air quality were correlated with sharp reductions in traffic volume and economic activity during the lockdowns. In this comprehensive global study, we provide revised estimates of the declines in anthropogenic emissions for 36 megacities using a novel methodology for disentangling the anthropogenic emissions from the meteorological transport and natural variability. We further quantify the uncertainty associated with changes in the a priori profile shape information during the lockdowns due to reduced emissions. Satellite NO2 retrieval techniques calculate an air mass factor that requires a priori NO2 profile shape information representative of the local atmosphere. This information, which is typically obtained from a chemical transport model (CTM), was not available for the early studies. This study also accounts for the satellite sampling errors resulting from the selective sampling of non-cloudy scenes during the study period. For our analysis, we used CTM simulations that were generated with and without COVID-impacted emissions. We perform retrievals of tropospheric NO2 columns with the NASA NO2 algorithm, and then use observed and simulated data to disentangle the meteorological transport from the contribution due anthropogenic emissions. We found that the meteorological transport was most significant source of variability ranging between −35% and 22% of the change total tropospheric VCD. We also find that not accounting for changes in the a priori NO2 profile shape information during the lockdowns resulted in systematic retrieval errors that were up to 12% of the estimated decline, and the elimination of cloud contaminated scenes resulted in sampling errors that in general ranged between varied ±15%.

Volcanic eruptions can be catastrophic events, particularly when they occur in inhabited coastal environments. They also play important roles in climate and biogeochemical cycles, including through nutrient deposition in the ocean. Volcanic ash studies in the ocean have focused on the phytoplankton response, generally quantifying changes in chlorophyll-a concentration. Many gaps remain in addressing fundamental questions regarding why volcanic ash deposition may enhance or limit both phytoplankton growth and/or drive community composition shifts. Here we outline a wide, multidisciplinary vision for monitoring volcanic eruptions near ocean ecosystems from satellites, including considerations for characteristics of airborne volcanic ash and ash geochemistry in seawater. Ultimately, observations beyond chlorophyll-a are needed to quantify phytoplankton communities (including harmful algal blooms) and possible impacts across higher trophic levels. We synthesize relevant research from volcanic studies as well as atmospheric and ocean sciences to identify the 'known unknowns' in ash-ecosystem studies. Our goal is to move toward an improved understanding of how real-time and near-realtime monitoring of volcanic eruptions can help address societally relevant questions.

Early versions of satellite nadir-viewing UV SO2 data products did not explicitly account for the effects of snow/ice on retrievals. Snow-covered terrain, with its high reflectance in the UV, typically enhances satellite sensitivity to boundary layer pollution. However, a significant fraction of high-quality cloud-free measurements over snow is currently excluded from analyses. This leads to increased uncertainties of satellite emission estimates and potential seasonal biases due to the lack of data in winter months for some high-latitudinal sources. In this study, we investigated how Ozone Monitoring Instrument (OMI) and TROPOspheric Monitoring Instrument (TROPOMI) satellite SO2 measurements over snow-covered surfaces can be used to improve the annual emissions reported in our SO2 emissions catalogue (version 2; Fioletov et al., 2023). Only 100 out of 759 sources listed in the catalogue have 10 % or more of the observations over snow. However, for 40 high-latitude sources, more than 30 % of measurements suitable for emission calculations were made over snow-covered surfaces. For example, in the case of Norilsk, the world's largest SO2 point-source, annual emission estimates in the SO2 catalogue were based only on 3–4 summer months, while the addition of data for snow conditions extends that period to 7 months. Emissions in the SO2 catalogue were based on satellite measurements of SO2 slant column densities (SCDs) that were converted to vertical column densities (VCDs) using site-specific clear-sky air mass factors (AMFs), calculated for snow-free conditions. The same approach was applied to measurements with snow on the ground whereby a new set of constant, site-specific, clear sky with snow AMFs was created, and these were applied to the measured SCDs. Annual emissions were then estimated for each source considering (i) only clear-sky and snow-free days, (ii) only clear-sky with snow days, and (iii) a merged dataset (snow and snow-free conditions). For individual sources, the difference between emissions estimated for snow and snow-free conditions is within ±20 % for three-quarters of smelters and oil and gas sources and with practically no systematic bias. This is excellent consistency given that there is typically a factor of 3–5 difference between AMFs for snow and snow-free conditions. For coal-fired power plants, however, emissions estimated for snow conditions are on average 25 % higher than for snow-free conditions; this difference is likely real and due to larger production (consumption of coal) and emissions in wintertime.

The way volcanic clouds evolve is very sensitive to the initial spatial 3D distributions of volcanic materials, which are often unknown. In this study, we conducted inverse modeling of the Mt. Pinatubo cloud using total ozone mapping spectrometer 2D mapping of Aerosol Index and SO 2 loading during the first three post‐eruption days to estimate the time‐dependent emissions profiles and initial 3D spatial distributions of volcanic ash and SO 2 . We account for aerosol radiative feedback and dynamic lofting of volcanic ash in the inversion calculations for the first time. This resulted in a lower ash injection height (by 1.5 km for ash) than without ash radiative feedback. The Pinatubo eruption ejected ≈77% of fine ash at 12–23 km, ≈65% of SO 2 at 18–25 km. In contrast with previous studies, which suggested that all volcanic materials were emitted above the tropopause, a significant fraction of SO 2 (5.1 of 15.5 Mt) and fine ash (37.2 of 66.5 Mt) were ejected in the troposphere, where SO 2 quickly oxidized into sulfate aerosol that is short‐lived in the troposphere. This explains the early presence of sulfate aerosols in the plume and why the models can reproduce the observed volcanic aerosols' optical depth (AOD), assuming lower‐than‐observed SO 2 emission in the stratosphere. Despite the quicker than in observations build‐up of sulfate AOD, in a month after the eruption, the evolution of the Pinatubo AOD simulated using the obtained ash and SO 2 initial distributions converges with the available stratospheric aerosol and gas experiment observations.

&lt;p&gt;As part of its Climate Change Initiative extension program (CCI+), the European Space Agency started a new activity aiming to develop space-based long-term climate data records for precursors gases involved in the formation of aerosols and ozone (https://climate.esa.int/en/projects/precursors-for-aerosols-and-ozone/about/). The targeted trace gases are NO&lt;sub&gt;2&lt;/sub&gt;, SO&lt;sub&gt;2&lt;/sub&gt;, HCHO, CHOCHO, CO and NH&lt;sub&gt;3&lt;/sub&gt;.&lt;/p&gt; &lt;p&gt;For SO&lt;sub&gt;2 &lt;/sub&gt;in particular the project relies on recent developments for Sentinel-5 Precursor/TROPOMI based the Covariance-Based Retrieval Algorithm (COBRA). This algorithm will be used to build consistent column retrievals from a series of four satellite sensors (GOME, SCIAMACHY, OMI and TROPOMI), covering nearly three decades of observations.&lt;/p&gt; &lt;p&gt;Here, we give an overview of the SO&lt;sub&gt;2&lt;/sub&gt; development activities in the Precursors CCI+ project. To achieve the project objectives, an extensive comparison of results is performed in the form of a round Robin exercise. This includes comparison of different algorithms, but also testing of various processing options for ancillary data such as cloud and surface reflectance parameters. First results from the round Robin exercise are presented with a focus on OMI, in particular the comparison between COBRA and Principal Component Analysis (PCA) column retrievals. Preliminary results using COBRA for the historical sensors GOME and SCIAMACHY are also shown.&lt;/p&gt; &lt;p&gt;&amp;#160;&lt;/p&gt;

Nitrogen dioxide (NO2) and nitrogen oxide (NO), usually referred to as nitrogen oxides (NOx), are emitted into the atmosphere by anthropogenic and natural sources. The detection and monitoring of NO2 plays a key role in air quality managements because of its effects on health and of its contribution to the increase of tropospheric ozone and nitrate aerosols. Unfortunately, up to now, observations were possible only once a day based on satellites in low-earth orbit (GOME, GOME-2, OMI and TROPOMI). However, from now on, it is possible to observe the diurnal variations over the Asia based on Geostationary Environment Monitoring Spectrometer (GEMS) in geostationary earth orbit. Here, we present results of tropospheric nitrogen dioxide column observations with high temporal (hourly) and spatial resolutions over major cities in Asia. In addition, we evaluate the GEMS NO2 operational algorithm by comparing GEMS total and tropospheric NO2 columns with independent observations from ground-based Pandora (total column) and MAX-DOAS (tropospheric column). Additionally, we retrieved the GEMS tropospheric NO2 columns by subtracting the stratospheric NO2 columns, which are assumed based on SLIMCAT model data and then scalded with the real GEMS observations, from the total NO2 columns. Then we also compared the tropospheric NO2 columns that are retrieved based on GEMS NO2 operational algorithm and SLIMCAT model, respectively. Lastly, we compared the GEMS NO2 with other low-earth orbit satellite instruments that include OMI and TROPOMI.

Sulfur dioxide (SO2) measurements from the Ozone Monitoring Instrument (OMI), Ozone Mapping and Profiler Suite (OMPS), and TROPOspheric Monitoring Instrument (TROPOMI) satellite spectrometers were used to update and extend the previously developed global catalogue of large SO2 emission sources. This version 2 of the global catalogue covers the period of 2005–2021 and includes a total of 759 continuously emitting point sources releasing from about 10 kt yr−1 to more than 4000 kt yr−1 of SO2, that have been identified and grouped by country and primary source origin: volcanoes (106 sources); power plants (477); smelters (74); and sources related to the oil and gas industry (102). There are several major improvements compared to the original catalogue: it combines emissions estimates from three satellite instruments instead of just OMI, uses a new version 2 of the OMI and OMPS SO2 dataset, and updated consistent site-specific air mass factors (AMFs) are used to calculate SO2 vertical column densities (VCDs). The newest TROPOMI SO2 data processed with the Covariance-Based Retrieval Algorithm (COBRA), used in the catalogue, can detect sources with emissions as low as 8 kt yr−1 (in 2018–2021) compared to the 30 kt yr−1 limit for OMI. In general, there is an overall agreement within ±12 % in total emissions estimated from the three satellite instruments for large regions. For individual emission sources, the spread is larger: the annual emissions estimated from OMI and TROPOMI agree within ±13 % in 50 % of cases and within ±28 % in 90 % of cases. The version 2 catalogue emissions were calculated as a weighted average of emission estimates from the three satellite instruments using an inverse-variance weighting method. OMI, OMPS, and TROPOMI data contribute 7 %, 5 %, and 88 % to the average, respectively, for small (<30 kt yr−1) sources and 33 %, 20 %, and 47 %, respectively, for large (>300 kt yr−1) sources. The catalogue data show an approximate 50 % decline in global SO2 emissions between 2005 and 2021, although emissions were relatively stable during the last 3 years. The version 2 of the global catalogue has been posted at the NASA global SO2 monitoring website (https://doi.org/10.5067/MEASURES/SO2/DATA406, Fioletov et al., 2022).

Nitrogen dioxide (NO2) is an important trace-gas pollutant and climate agent whose presence also leads to spectral interference in ocean color retrievals. NO2 column densities have been retrieved with satellite UV–Vis spectrometers such as the Ozone Monitoring Instrument (OMI) and the Tropospheric Monitoring Instrument (TROPOMI) that typically have spectral resolutions of the order of 0.5 nm or better and spatial footprints as small as 3.6 km × 5.6 km. These NO2 observations are used to estimate emissions, monitor pollution trends, and study effects on human health. Here, we investigate whether it is possible to retrieve NO2 amounts with lower-spectral-resolution hyperspectral imagers such as the Ocean Color Instrument (OCI) that will fly on the Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) satellite set for launch in early 2024. OCI will have a spectral resolution of 5 nm and a spatial resolution of ∼ 1 km with global coverage in 1–2 d. At this spectral resolution, small-scale spectral structure from NO2 absorption is still present. We use real spectra from the OMI to simulate OCI spectra that are in turn used to estimate NO2 slant column densities (SCDs) with an artificial neural network (NN) trained on target OMI retrievals. While we obtain good results with no noise added to the OCI simulated spectra, we find that the expected instrumental noise substantially degrades the OCI NO2 retrievals. Nevertheless, the NO2 information from OCI may be of value for ocean color retrievals. OCI retrievals can also be temporally averaged over timescales of the order of months to reduce noise and provide higher-spatial-resolution maps that may be useful for downscaling lower-spatial-resolution data provided by instruments such as OMI and TROPOMI; this downscaling could potentially enable higher-resolution emissions estimates and be useful for other applications. In addition, we show that NNs that use coefficients of leading modes of a principal component analysis of radiance spectra as inputs appear to enable noise reduction in NO2 retrievals. Once trained, NNs can also substantially speed up NO2 spectral fitting algorithms as applied to OMI, TROPOMI, and similar instruments that are flying or will soon fly in geostationary orbit.

Most volcanism on Earth is submarine, but volcanic gas emissions by submarine eruptions are rarely observed and hence largely unquantified. On 15 January 2022 a submarine eruption of Hunga Tonga-Hunga Ha’apai (HTHH) volcano (Tonga) generated an explosion of historic magnitude, and was preceded by ∼1 month of Surtseyan eruptive activity and two precursory explosive eruptions. We present an analysis of ultraviolet (UV) satellite measurements of volcanic sulfur dioxide (SO2) between December 2021 and the climactic 15 January 2022 eruption, comprising an unprecedented record of Surtseyan eruptive emissions. UV measurements from the Ozone Monitoring Instrument (OMI) on NASA’s Aura satellite, the Ozone Mapping and Profiler Suite (OMPS) on Suomi-NPP, the Tropospheric Monitoring Instrument (TROPOMI) on ESA’s Sentinel-5P, and the Earth Polychromatic Imaging Camera (EPIC) aboard the Deep Space Climate Observatory (DSCOVR) are combined to yield a consistent multi-sensor record of eruptive degassing. We estimate SO2 emissions during the eruption’s key phases: the initial 19 December 2021 eruption (∼0.01 Tg SO2); continuous SO2 emissions from 20 December 2021—early January 2022 (∼0.12 Tg SO2); the 13 January 2022 stratospheric eruption (0.06 Tg SO2); and the paroxysmal 15 January 2022 eruption (∼0.4–0.5 Tg SO2); yielding a total SO2 emission of ∼0.6–0.7 Tg SO2 for the eruptive episode. We interpret the vigorous SO2 emissions observed prior to the January 2022 eruptions, which were significantly higher than measured in the 2009 and 2014 HTHH eruptions, as strong evidence for a rejuvenated magmatic system. High cadence DSCOVR/EPIC SO2 imagery permits the first UV-based analysis of umbrella cloud spreading and volume flux in the 13 January 2022 eruption, and also tracks early dispersion of the stratospheric SO2 cloud injected on January 15. The ∼0.4–0.5 Tg SO2 discharged by the paroxysmal 15 January 2022 HTHH eruption is low relative to other eruptions of similar magnitude, and a review of other submarine eruptions in the satellite era indicates that modest SO2 yields may be characteristic of submarine volcanism, with the emissions and atmospheric impacts likely dominated by water vapor. The origin of the low SO2 loading awaits further investigation but scrubbing of SO2 in the water-rich eruption plumes and rapid conversion to sulfate aerosol are plausible, given the exceptional water emission by the 15 January 2022 HTHH eruption.

Abstract. Nitrogen dioxide (NO2) is an important trace-gas pollutant and climate agent whose presence also leads to spectral interference in ocean color retrievals. NO2 column densities have been retrieved with satellite UV-Vis spectrometers such as the Ozone Monitoring Instrument (OMI) and Tropospheric Monitoring Instrument (TROPOMI) that typically have spectral resolutions of the order of 0.5 nm or better and spatial footprints as small as 3.5 km × 5 km. These NO2 observations are used to estimate emissions, monitor pollution trends, and study effects on human health. Here, we investigate whether it is possible to retrieve NO2 amounts with lower spectral resolution hyper-spectral imagers such as the Ocean Color Instrument (OCI) that will fly on the Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) satellite set for launch in early 2024. OCI will have a spectral resolution of 5 nm and a spatial resolution of ∼1 km with global coverage in 1–2 days. At this spectral resolution, small scale spectral structure from NO2 absorption is still present. We use real spectra from the OMI to simulate OCI spectra that are in turn used to estimate NO2 slant column densities (SCDs) with an artificial neural network trained on target OMI retrievals. While we obtain good results with no noise added to the OCI simulated spectra, we find that the expected instrumental noise substantially degrades the OCI NO2 retrievals. Nevertheless, the NO2 information from OCI may be of value for ocean color retrievals, as our simulations suggest that it will be of similar or slightly better quality as compared with TROPOMI NO2 data at TROPOMI spatial resolution on a daily basis and will be available simultaneously. OCI retrievals can also be temporally averaged over time-scales of the order months to reduce noise and provide higher spatial resolution maps that may be useful for downscaling information provided by lower spatial resolution instruments such as OMI and TROPOMI, for high resolution emissions estimates, and other applications. In addition, we explore the possibility of using an extended fitting window for NO2 retrievals as compared with traditional approaches. We demonstrate that the use of an extended spectral fitting window can reduce random errors in a current state-of-the-art OMI NO2 SCD product. Machine learning approaches with extended fitting windows, once trained, can also substantially speed up NO2 spectral fitting algorithms as applied to OMI, TROPOMI, and similar instruments that are flying or will soon fly in geostationary orbit.

Space-based quantitative passive optical remote sensing of the Earth’s surface typically involves the detection and elimination of cloud-contaminated pixels as an initial processing step. We explore a fundamentally different approach; we use machine learning with cloud contaminated satellite hyper-spectral data to estimate underlying terrestrial surface reflectances at red, green, and blue (RGB) wavelengths. An artificial neural network (NN) reproduces land RGB reflectances with high fidelity, even in scenes with moderate to high cloud optical thicknesses. This implies that spectral features of the Earth’s surface can be detected and distinguished in the presence of clouds, even when they are partially and visibly obscured by clouds; the NN is able to separate the spectral fingerprint of the Earth’s surface from that of the clouds, aerosols, gaseous absorption, and Rayleigh scattering, provided that there are adequately different spectral features and that the clouds are not completely opaque. Once trained, the NN enables rapid estimates of RGB reflectances with little computational cost. Aside from the training data, there is no requirement of prior information regarding the land surface spectral reflectance, nor is there need for radiative transfer calculations. We test different wavelength windows and instrument configurations for reconstruction of surface reflectances. This work provides an initial example of a general approach that has many potential applications in land and ocean remote sensing as well as other practical uses such as in search and rescue, precision agriculture, and change detection.