In the mesosphere, vertical wind velocities can exceed tens of m/s, driven by gravity wave (GW) processes such as wave breaking and the generation of secondary GWs. These dynamic events play a crucial role in redistributing energy and momentum across atmospheric layers. In this study, we investigate the GW dynamics over the Southern Andes—a prominent hotspot for GW activity—using a combination of satellite and ground-based observations. In particular, we analyze nearly two decades of atmospheric infrared sounder (AIRS) data, along with approximately 5 years of sodium (Na) lidar observations from the Andes Lidar Observatory (ALO). Our analysis reveals a clear seasonal co-variation between lower-stratospheric GW activity and mesospheric perturbations, with both datasets exhibiting pronounced austral-winter enhancements. The mesospheric Na-lidar variance also exhibits secondary maxima during late summer and spring; a simple vertical-wavenumber spectral analysis indicates that these shoulder enhancements result from modest, broadband increases in short-vertical-wavelength power. This variability is consistent with vertical coupling mediated by seasonally varying background winds and mesospheric wave dissipation.

Abstract. Tropical cyclones are prominent sources of atmospheric gravity waves due to their intense and organized deep convection, yet the sensitivity of cyclone-induced gravity waves to model physics choices remains poorly constrained. Here, we examine how different physical parameterizations in the Weather Research and Forecasting (WRF) model influence the generation and characteristics of stratospheric gravity waves during the well-observed case of Typhoon Soudelor (2015). A suite of high-resolution simulations of the tropical cyclone employing multiple microphysics, planetary boundary layer, and cumulus parameterizations is analyzed. Gravity wave diagnostics are derived from vertical velocity variability and a localized S-transform spectral analysis. Although the simulations produce similar tropical cyclone tracks and intensities, they generate markedly different gravity wave responses. Differences in microphysics and boundary layer schemes lead to systematic variations in gravity wave amplitudes and spectral characteristics, while the inclusion of a cumulus parameterization consistently weakens deep convection and gravity wave amplitudes. These differences occur despite comparable background wind and stability conditions, demonstrating that gravity wave sensitivity is largely controlled by differences in diabatic forcing among the simulations. Our results highlight the critical role of model physics in shaping gravity wave generation and stratospheric wave characteristics and demonstrate the importance of carefully evaluating diabatic forcing in mesoscale modeling studies.

Volcanic injections into the upper troposphere-lower stratosphere (UTLS) affect climate by altering Earth's radiation budget and atmospheric chemistry. However, the pathways by which mid-latitude eruptions spread globally remain poorly understood. We combine nighttime Compact Optical Backscatter Aerosol Detector (COBALD) profiles over Lhasa (China) with ERA5-driven Chemical Lagrangian Model of the Stratosphere (CLaMS) backward trajectories and global three-dimensional sulfur dioxide (SO2)-based tracer simulations. With this integrated framework, we track the plume of the Raikoke eruption (21-22 June 2019) during its transport through the Northern Hemisphere and its interaction with the mature Asian Summer Monsoon Anticyclone (ASMA). Balloon-borne measurements capture the plume's arrival, vertical spreading, and dilution by air in the ASMA interior. Trajectories reveal two principal pathways from distinct Raikoke plumes: (i) an upper-level branch within the summertime stratospheric easterly flow (similar to 460-490 K) carrying the trailing filament of the vorticized volcanic plume (VVP), and (ii) a lower-level branch within the subtropical westerly jet (similar to 390-430 K) carrying the main plume. Although the ASMA can act as a transport barrier at certain potential-temperature levels, it admits mixing into the anticyclone along jet-aligned filaments and redistributes aerosols internally. We show that the transport of the volcanic plume is sensitive to parameterized mixing by varying the mixing strength in CLaMS. Portable Optical Particle Spectrometer (POPS) profiles over Boulder (USA) confirm the plume's timing and altitude, providing an independent evaluation of its transport outside the ASMA region.

Atmospheric gravity waves (GWs) are a key driver of vertical energy and momentum transport in the atmosphere, with important implications for large-scale dynamics and chemistry. However, they remain difficult to predict in operational weather and climate models due to their small spatial scales relative to model resolution, and are typically not assimilated into numerical weather prediction (NWP) systems because of the large departures they introduce from model initial conditions.Here we use stratospheric temperature measurements from the Atmospheric Infrared Sounder (AIRS) and the Cross-track Infrared Sounder (CrIS) to evaluate how well archived operational analyses and forecasts from ECMWF’s Integrated Forecast System reproduce observed GW activity over Greenland, a major Northern Hemisphere source region for orographic GWs. The combined AIRS–CrIS sampling at high latitudes provides an unusually high measurement cadence, enabling assessment of forecast performance and time variability at relatively fine temporal resolution.Operational analyses and forecasts with lead times of up to 240 h are sampled at the AIRS and CrIS measurement footprints and regridded to a common resolution to allow consistent spectral analysis. A 2D+1 Stockwell Transform is applied to both synthetic and real observations to characterise GW amplitudes and spatial structure, producing directly comparable GW fields across forecast lead times.Using a Structure–Amplitude–Location (SAL) framework adapted from precipitation forecast verification, we quantify the evolution of GW forecast skill with lead time. We find that model performance exhibits only weak dependence on forecast range: across all lead times, the model systematically produces GWs with smaller horizontal scales and reduced amplitudes relative to observations, while errors in wave location increase only modestly with lead time. This behaviour is unexpected, as shorter lead times are associated with more accurate resolved winds, and would therefore be expected to yield more accurate GW generation. The results suggest that errors in simulated GW characteristics in operational forecasts are dominated by structural and representational limitations rather than by forecast wind errors alone.

Advances in computational power and model development have enabled the generation of global high-resolution models. These new models can resolve a large proportion of gravity waves (GWs) explicitly, reducing reliance on subgrid parametrizations. GWs are vital components of the middle and upper atmosphere, they transport energy and momentum both horizontally and vertically, driving the atmospheric circulation. Evaluating the realism of these resolved waves is a crucial step in advancing future model development.Here we provide the first global multi-model GW observational comparison that accounts for the observational filter. We assess the representation of stratospheric GWs in three high-resolution (3-5 km horizontal resolution) global free-running simulations (ICON, IFS and GEOS), for the period 20 January-29 February 2020, against AIRS satellite observations.Time-mean wave amplitudes are systematically lower in the models than observations, consistent with previous studies. GW occurrence rates are higher in all models than the observations, dominated by low amplitude waves in the models. During the first 10 d spatial patterns of GW occurrence rate, amplitudes and momentum flux agree across the models and observations but subsequently they diverge. Agreement is more consistent in the Northern Hemisphere (where orographic waves dominate) than in the Southern Hemisphere (where convective waves dominate).These results benchmark the current state of high-resolution modelling and demonstrate that whilst there are strengths in models' ability to capture the morphology of GWs (particularly orographically generated waves), there is room for improvement in modelling amplitudes, occurrence rates and zonal-mean flux magnitudes globally, with the largest discrepancies in the tropical convective regions.

The continuous increase in computational power comes with a corresponding demand for storage space. However, the ability to store data has hardly increased in recent years. This makes the demand for efficient storage solutions even more pressing, especially for meteorological reanalysis data. The current European Centre for Medium-Range Weather Forecasts (ECMWF) ERA5 reanalysis data already poses a major challenge for the community, but with the upcoming ERA6 reanalysis data, which will have an even higher resolution, significantly more storage space will be needed. An efficient way to store data is to use either lossy or lossless data compression methods to reduce storage requirements. To compress the meteorological data, we perform a multiresolution-analysis using multiwavelets on a hierarchy of nested grids. Since the local differences become negligibly small in regions where the data are locally smooth, we apply hard thresholding for data compression. This results in a high compression rate while preserving the accuracy of the original data. This strategy has been implemented into the Lagrangian model for Massive-Parallel Trajectory Calculation (MPTRAC) and has been successfully applied to ERA5 reanalysis data. Compression rates ranging from 1.6 to 12.6 can be achieved while at the same time maintaining the accuracy of the data within acceptable error limits. This leads to a reduction in storage of up to 93%, for example, reducing the file size of an ERA5 data file corresponding to a time instant from 4.9 GB to 389 MB. This renders the multiresolution-based grid adaptation a particularly suitable and effective approach for addressing the data storage challenges in atmospheric transport simulations.

Mountain-wave-induced temperature perturbations can locally enable the formation of polar stratospheric clouds (PSCs). We examine a decade-long (2002–2012) record of ice PSCs derived from MIPAS/Envisat measurements. The points with the smallest temperature difference (ΔTice_min) between the frost point temperature (Tice) and the environmental temperature along the line of sight have been proposed and shown to provide a better estimate of the location of ice PSC observation from MIPAS. The temperature for the ice PSC observations is analyzed based on ERA5. Following this, we investigated the temperature history of the ice PSCs detected above Tice at the observation points along 24 h backward trajectories.We find that 52 % of Arctic and 26 % of Antarctic ice PSCs are detected above Tice, with pronounced clustering over mountainous terrain and in downstream regions. The backward trajectories were calculated by using the MPTRAC model, initialized at the ΔTice_min​ locations. Analysis of the temperature evolution along these trajectories shows that the fraction of ice PSCs at a temperature above Tice along the trajectory decreases, with the strongest decrease within the 6 h before observation. Accounting for temperature fluctuations along the air-mass histories, reduces the fractions of too warm ice PSCs at observation to 33 % in the Arctic and 9 % in the Antarctic.These results demonstrate the substantial role of orographic waves in ice PSC formation and provide observational constraints for chemistry–climate model evaluation. This contribution is based on the published analysis of Zou et al. (2024, Atmos. Chem. Phys., 24, 11759–11774, https://doi.org/10.5194/acp-24-11759-2024) .

Satellite observations of the atmosphere are often extremely noisy due to both hardware limitations and the inherent complexity of retrieving and making measurements of the atmosphere. Gravity waves, which are low amplitude signals present in the atmosphere, are hard to resolve in this data due to their relatively low amplitude and small spatial extent. As a result, noise becomes a limiting factor when trying to identify and characterise them in real observed data.Current methods to address this problem often lean upon smoothing approaches; however, such approaches suppress small scale signals and reduce measured amplitude and momentum fluxes significantly. This impedes the process in developing the next generation of models where these waves must be resolved accurately.A novel supervised machine learning approach is introduced which is able to accurately remove small scale noise features from nadir observations of gravity waves. This model was trained on synthetic observations derived from high resolution DYAMOND model runs. This is then applied to 22 years of NASA AIRS data and 12 years of MetOp IASI data and used to produce a new gravity wave climatology to better access small amplitude gravity waves.

Modern numerical modelling simulations of the Earth's atmosphere have developed over the recent decades to ever finer spatial resolutions, allowing for a greater portion the atmospheric gravity wave (GW) spectrum to be resolved. Specialised global simulations with kilometre-scale resolutions have been performed offline that can resolve very large portions of the GW spectrum in the lower stratosphere and, as such, the balance between resolved and parameterised (unresolved) GW forcing in today's numerical simulations of the middle atmosphere is shifting. However, these kilometre-scale simulations are still too computationally costly to perform routinely and can quickly deviate from their initial conditions, which makes validating the resolved gravity waves in these simulations with satellite observations challenging. For this reason, a growing number of studies are using resolved GWs in lower-resolution stratospheric reanalyses as proxies for GWs in the real atmosphere, due to the apparent reliability, long timescale, global coverage and real-date data assimilation of these reanalysis products. However, these resolved GWs in reanalyses have not been widely tested or compared to satellite observations of GWs to assess their realism. One reason why such a comparison has been so challenging is due to the different ranges of GW wavelengths to which any given model or observational instrument is sensitive due to its grid spacing or sampling and resolution limits, an effect known as the observational filter. Therefore, any like-for-like assessment of resolved GWs in reanalysis using satellite observations must be able to sample the model using the exact sampling and resolution of the instrument. Here we use 3-D satellite observations from AIRS/Aqua to evaluate the realism of resolved stratospheric gravity waves in ERA5 reanalysis produced by the European Centre for Medium Range Weather Forecasts (ECMWF). We carefully apply the sampling and resolution limits of AIRS to the model using a full 3-D weighting function for each measurement footprint to create synthetic measurements of the ERA5 stratosphere as if were viewed by AIRS. We then follow identical processing steps to detrend, regrid and spectrally analyse both the real and synthetic measurements to recover localised GW amplitudes, wavelengths and directional momentum fluxes between 25 and 45 km altitude. We investigate the global momentum budget of GWs in reanalysis compared to observations and compare the seasonality and spectral properties of GWs over known stratospheric hot spots. Our preliminary results suggest that AIRS measurements exhibit more frequent large-amplitude wave events at larger horizontal wavelengths (greater than 150km) and larger net momentum fluxes overall than equivalent ERA5 measurements. Our satellite-sampling approach is applicable to any GW-resolving model, and sets out a potential roadmap towards more direct validation and comparison of resolved mesoscale dynamics in numerical models that could help to guide developments in the coming era of high-spatial resolution atmospheric modelling.

Computer performance has increased immensely in recent years, but the ability to store data has hardly increased at all. The current version of meteorological reanalysis data ERA5 provided by the European Centre of Medium-Range Weather Forecasts (ECMWF) has increased by a factor of ∼80 compared to its predecessor ERA-Interim. This presents scientists with major challenges, especially if data covering several decades is to be stored on local computer systems. Accordingly, many compression methods have been developed in recent years with which data can be stored either lossless or lossy. Here we test three of these methods two lossy compression methods ZFP and Layer Packing (PCK) and the lossless compressor ZStandard (ZSTD) and investigate how the use of compressed data affects the results of Lagrangian air parcel trajectory calculations with the Lagrangian model for Massive-Parallel Trajectory Calculations (MPTRAC). We analysed 10-day forward trajectories that were globally distributed over the free troposphere and stratosphere. The largest transport deviations were derived when using ZFP with the largest compression. Using a less strong compression we could reduce the transport deviation and still derive a significant compression. Since ZSTD is a lossless compressor, we derive no transport deviations at all when using these compressed files, but do not loose much disk space using this compressor (reduction of ∼20%). The best result concerning compression efficiency and transport deviations is derived with the layer packing method PCK. The data is compressed by about 50%, but transport deviations do not exceed 40 km in the free troposphere and are even lower in the upper troposphere and stratosphere. Thus, our study shows that the PCK compression method would be valuable for application in atmospheric sciences and that with compression of meteorological reanalyses data files we can overcome the challenges of high demand of disk space from these data sets.

Lagrangian transport models are important tools to study the sources, spread, and lifetime of air pollutants. In order to simulate the transport of reactive atmospheric pollutants, the implementation of efficient chemistry and mixing schemes is necessary to properly represent the lifetime of chemical species. Based on a case study simulating the long-range transport of volcanic sulfur dioxide (SO2) for the 2019 Raikoke eruption, this study compares two chemistry schemes implemented in the Massive-Parallel Trajectory Calculations (MPTRAC) Lagrangian transport model. The explicit scheme represents first-order and pseudo-first-order loss processes of SO2 based on prescribed reaction rates and climatological oxidant fields, i.e., the hydroxyl radical in the gas phase and hydrogen peroxide in the aqueous phase. Furthermore, an implicit scheme with a reduced chemistry mechanism for volcanic SO2 decomposition has been implemented, targeting the upper-troposphere–lower-stratosphere (UT–LS) region. Considering nonlinear effects of the volcanic SO2 chemistry in the UT–LS region, we found that the implicit solution yields a better representation of the volcanic SO2 lifetime, while the first-order explicit solution has better computational efficiency. By analyzing the dependence between the oxidants and SO2 concentrations, correction formulas are derived to adjust the oxidant fields used in the explicit solution, leading to a good trade-off between computational efficiency and accuracy. We consider this work to be an important step forward to support future research on emission source reconstruction involving nonlinear chemical processes.

We present the first-ever global simulation of the full Earth system at 1.25 km grid spacing, achieving highest time compression with an unseen number of degrees of freedom. Our model captures the flow of energy, water, and carbon through key components of the Earth system: atmosphere, ocean, and land. To achieve this landmark simulation, we harness the power of 8192 GPUs on Alps and 20480 GPUs on JUPITER, two of the world's largest GH200 superchip installations. We use both the Grace CPUs and Hopper GPUs by carefully balancing Earth's components in a heterogeneous setup and optimizing acceleration techniques available in ICON's codebase. We show how separation of concerns can reduce the code complexity by half while increasing performance and portability. Our achieved time compression of 145.7 simulated days per day enables long studies including full interactions in the Earth system and even outperforms earlier atmosphere-only simulations at a similar resolution.

This study examines the seasonal distributions of simultaneous stratospheric concentric gravity waves (GWs) observed by the Atmospheric Infrared Sounders and concentric traveling ionospheric disturbances (TIDs) detected by the ground‐based Global Navigation Satellite System Total Electron Content observations over the U.S. in 2022, to illustrate the mesoscale vertical coupling between the lower atmosphere and the ionosphere. We compared epicenters of GWs and TIDs in the stratosphere and ionosphere with tropospheric weather conditions and background winds in the thermosphere. Epicenters of concentric TIDs associated with stratospheric concentric GWs correspond to areas with high convective available potential energy over the central to eastern U.S. (∼60–110W) in summer and over the southern U.S. (south of ∼40) in spring and fall. Conversely, in fall to spring, epicenters over the northern U.S. (north of ∼40) appeared south of regions with high extratropical cyclone activity. These findings suggest that convection was a primary source of concentric GWs driving TIDs over the continental U.S. during all four seasons, although the specific weather phenomena associated with the convection varied by season. Convection over the central to eastern U.S. in summer and the southern U.S. in spring could be linked to thunderstorms. In contrast, convection over the northern U.S. from fall through spring was likely linked to extratropical cyclones. We also found that concentric TIDs were linked to 66% of the stratospheric concentric GW events (195 events in total), underscoring the significant role of convection as a source of TIDs in the lower atmosphere and its contribution to the vertical coupling.

Comparisons between observed and model‐resolved gravity waves (GWs) are crucial for evaluating general circulation model (GCM) simulation accuracy and understanding wave characteristics. However, observational noise often obscures waves, complicating such comparisons. To address this, we have developed a GW detection method using a convolutional neural network (CNN). The CNN is trained on Atmospheric Infrared Sounder (AIRS) temperatures with labels indicating wave presence based on Berthelemy et al. (2025, https://doi.org/10.5194/egusphere‐2025‐455 ). Their method detects noise‐induced pixel‐to‐pixel variations in horizontal wavelengths; in contrast, the CNN robustly identify waves even when applied to smoothly varying model data. Using this method, we compare stratospheric GWs in boreal winters between AIRS observations and a high‐top GW‐permitting GCM, Japanese Atmospheric GCM for Upper Atmosphere Research (JAGUAR). The results agree well and exhibit similar interannual variability, with discrepancies also identified, including a more zonally elongated distribution of tropical GWs in JAGUAR. This method is broadly applicable to the future use of satellites for guiding wave‐resolving atmospheric model development.

Atmospheric gravity waves (GWs) are an important mechanism for vertical transport of energy and momentum through the atmosphere. Their impacts are apparent at all scales, including aviation, weather, and climate. Identifying stratospheric GWs from satellite observations is challenging due to instrument noise and effects of weather processes, but they can be observed from nadir sounders such as the AIRS instrument onboard Aqua. Here, a new method (hereafter “neighbourhood method”) to detect GW information is presented and applied to AIRS data. This uses a variant of the 3D S-transform to calculate the horizontal wavenumbers of temperature perturbations, then find areas of spatially constant horizontal wavenumbers (assumed to be GWs), which allow for creating a binary wave-presence mask. We describe the concept of the neighbourhood method and use it to investigate GW amplitudes, zonal pseudomomentum fluxes, and vertical wavelengths over 5 years of AIRS data. We compare these results to those calculated from GWs detected using another widely used method based on an amplitude cutoff. 35 % of regions of wave activity detected using the neighbourhood method have amplitudes lower than is visible using the amplitude cutoff method. Three regions are studied in greater depth: the Rocky Mountains, North Africa, and New Zealand/Tasmania. GWs detected using the neighbourhood method have wave phase propagation angles consistent with linear theory. Using the neighbourhood method produces new statistics for regional and global GW studies, which compare favourably to the amplitude cutoff GW detection method.

Atmospheric gravity waves (GWs) are one of the most important drivers of the circulation of the middle and upper atmosphere. Usually generated in the lower atmosphere and propagating upwards through the atmospheric layers, the aggregated forcing of these waves drives circulations in the middle atmosphere that are far from that expected under radiative equilibrium. Circulations in the mesosphere and lower thermosphere (MLT) and above, especially in polar regions, have shown extreme sensitivity to GW parameterisations in recent high-top modelling simulations and can exhibit significant and limiting biases compared to observations. This uncertainty in the role of GW dynamics between models has made predictions of how these high-altitude circulations are expected to respond to a changing climate very challenging. This is confounded by a relative scarcity of global observations of GW activity in the middle and upper atmosphere with which to understand these connections over climate timescales. Since the early 2000s, satellite and ground-based instrumentation has provided an unprecedented observational view of middle atmospheric dynamics and composition, especially for the study of GWs. However, due to different instrument capabilities and limited hardware lifetimes, examining long term trends of GW properties observationally has been challenging due to the need to re-establish baselines. Here we examine results from some of the longest known single-instrument records of GW activity in the middle and upper atmosphere spanning more than two decades. We explore changes in GW amplitudes, wavelengths and directional momentum flux in the stratosphere from a 22-year climatology derived from global 3-D satellite observations from the AIRS/Aqua, the longest single-instrument climatology of this type. We also explore changes in wind, temperature and large-scale GW activity in the polar MLT from nearly 20 years of single-station meteor wind radar observations in the Arctic and Antarctic. We compare these trends to equivalent analysis of other long-term satellite GW datasets and resolved GW activity in ERA5 stratospheric reanalysis. Finally, we discuss limitations and best practise for considering observed trends in GW observations, such as how changes in circulation can affect GW propagation and their apparent sensitivity to satellite remote sensing techniques.

The first Tropospheric Ozone Assessment Report (TOAR, 2014–2019) encountered several observational challenges that limited the confidence in estimates of the burden, short-term variability, and long-term changes of ozone in the free troposphere. One of these challenges is the difficulty to interpret the consistency of satellite measurements obtained with different techniques from multiple sensors, leading to differences in spatiotemporal sampling, vertical smoothing, a-priori information, and uncertainty characterisation. This motivated the Committee on Earth Observation Satellites (CEOS) to initiate a coordinated activity VC-20-01 on improving the assessment and harmonisation of tropospheric ozone measured from space. Here, we report on work that contributes to this CEOS activity, as well as to the ongoing second TOAR assessment (TOAR-II, 2020–2025). Our objective is to harmonise the spatiotemporal perspective of (sixteen) satellite ozone data records, thereby accounting as much as possible for differences in vertical smoothing and sampling. Four harmonisation methods are presented to achieve this goal: two for ozone profiles obtained from nadir sounders (UV-visible, IR, and combined UV-IR), and two for tropospheric ozone column products derived by one of the residual methods (Convective Cloud Differential or Limb–Nadir Matching). We discuss to what extent harmonisation may affect assessments of the spatial distribution, seasonal cycle, and long-term changes in free tropospheric ozone, and we anchor the harmonised profile data to ozonesonde measurements recently homogenised as part of TOAR-II. We find that approaches that use global ozone fields as a transfer standard (here the Copernicus Atmosphere Monitoring Service ReAnalysis, CAMSRA) to constrain the harmonisation generally lead to the largest reduction of the inter-product dispersion (IPD) between satellite datasets. These harmonisation efforts, however, only partially account for the observed discrepancies between the satellite datasets, with a reduction of about 10 %–40 % of the IPD upon harmonisation, depending on the products involved and with strong spatiotemporal dependences. This work therefore provides evidence that it is not only the differences in spatiotemporal smoothing and sampling, but rather the differences in measurement uncertainty that pose the main challenge to the assessment of the spatial distribution and temporal evolution of free tropospheric ozone from satellite observations.