Stacking Machine Learning Models Empowered High Time-Height-Resolved Ozone Profiling from the Ground to the Stratopause Based on MAX-DOAS Observation

Ozone (O3) profiles are crucial for comprehending the intricate interplay among O3 sources, sinks, and transport. However, conventional O3 monitoring approaches often suffer from limitations such as low spatiotemporal resolution, high cost, and cumbersome procedures. Here, we propose a novel approach that combines multiaxis differential optical absorption spectroscopy (MAX-DOAS) and machine learning (ML) technology. This approach allows the retrieval of O3 profiles with exceptionally high temporal resolution at the minute level and vertical resolution reaching the hundred-meter scale. The ML models are trained using parameters obtained from radiative transfer modeling, MAX-DOAS observations, and a reanalysis data set. To enhance the accuracy of retrieving the aqueous phosphorus from O3, we employ a stacking approach in constructing ML models. The retrieved MAX-DOAS O3 profiles are compared to data from an in situ instrument, lidar, and satellite observation, demonstrating a high level of consistency. The total error of this approach is estimated to be within 25%. On balance, this study is the first ground-based passive remote sensing of high time-height-resolved O3 distribution from ground to the stratopause (0-60 km). It opens up new avenues for enhancing our understanding of the dynamics of O3 in atmospheric environments. Moreover, the cost-effective and portable MAX-DOAS combined with this versatile profiling approach enables the potential for stereoscopic observations of various trace gases across multiple platforms.

Comparison of Major Sudden Stratospheric Warming Impacts on the Mid-Latitude Mesosphere Based on Local Microwave Radiometer CO Observations in 2018 and 2019

In this paper, a comparison of the impact of major sudden stratospheric warmings (SSWs) in the Arctic in February 2018 (SSW1) and January 2019 (SSW2) on the mid-latitude mesosphere is given. The mesospheric carbon monoxide (CO) and zonal wind in these two major SSW events were observed at altitudes of 70–85 km using a microwave radiometer (MWR) at Kharkiv, Ukraine (50.0°N, 36.3°E). Data from ERA-Interim and MERRA-2 reanalyses and Aura Microwave Limb Sounder measurements were also used. It is shown that: (i) The differences between SSW1 and SSW2, in terms of local variability in zonal wind, temperature, and CO in the stratosphere and mesosphere, were clearly defined by the polar vortex (westerly in cyclonic circulation) and mid-latitude anticyclone (easterly) migrating over the MWR station, therefore; (ii) mesospheric intrusions of CO-rich air into the stratosphere over the Kharkiv region occurred only occasionally, (iii) the larger zonal wave 1–3 amplitudes before SSW1 were followed by weaker polar vortex recovery than that after SSW2, (iv) the strong vortex recovery after SSW2 was supported by earlier event timing (midwinter) favoring vortex cooling due to low solar irradiance and enhanced zonal circulation, and (v) vortex strengthening after SSW2 was accompanied by wave 1–3 amplification in March 2019, which was absent after SSW1. Finally, the influence of the large-scale circulation structures formed in individual major SSW events on the locally recorded characteristics of the atmosphere is discussed.

Impact of intercontinental pollution transport on North American ozone air pollution: an HTAP phase 2 multi-model study

The recent update on the US National Ambient Air Quality Standards (NAAQS) of the ground-level ozone (O₃/ can benefit from a better understanding of its source contributions in different US regions during recent years. In the Hemispheric Transport of Air Pollution experiment phase 1 (HTAP1), various global models were used to determine the O₃ source-receptor (SR) relationships among three continents in the Northern Hemisphere in 2001. In support of the HTAP phase 2 (HTAP2) experiment that studies more recent years and involves higher-resolution global models and regional models' participation, we conduct a number of regional-scale Sulfur Transport and dEposition Model (STEM) air quality base and sensitivity simulations over North America during May-June 2010. STEM's top and lateral chemical boundary conditions were downscaled from three global chemical transport models' (i.e., GEOS-Chem, RAQMS, and ECMWF C-IFS) base and sensitivity simulations in which the East Asian (EAS) anthropogenic emissions were reduced by 20 %. The mean differences between STEM surface O₃ sensitivities to the emission changes and its corresponding boundary condition model's are smaller than those among its boundary condition models, in terms of the regional/period-mean (<10 %) and the spatial distributions. An additional STEM simulation was performed in which the boundary conditions were downscaled from a RAQMS (Realtime Air Quality Modeling System) simulation without EAS anthropogenic emissions. The scalability of O₃ sensitivities to the size of the emission perturbation is spatially varying, and the full (i.e., based on a 100% emission reduction) source contribution obtained from linearly scaling the North American mean O₃ sensitivities to a 20% reduction in the EAS anthropogenic emissions may be underestimated by at least 10 %. The three boundary condition models' mean O₃ sensitivities to the 20% EAS emission perturbations are ~8% (May-June 2010)/~11% (2010 annual) lower than those estimated by eight global models, and the multi-model ensemble estimates are higher than the HTAP1 reported 2001 conditions. GEOS-Chem sensitivities indicate that the EAS anthropogenic NO _x emissions matter more than the other EAS O₃ precursors to the North American O₃, qualitatively consistent with previous adjoint sensitivity calculations. In addition to the analyses on large spatial-temporal scales relative to the HTAP1, we also show results on subcontinental and event scales that are more relevant to the US air quality management. The EAS pollution impacts are weaker during observed O₃ exceedances than on all days in most US regions except over some high-terrain western US rural/remote areas. Satellite O₃ (TES, JPL-IASI, and AIRS) and carbon monoxide (TES and AIRS) products, along with surface measurements and model calculations, show that during certain episodes stratospheric O₃ intrusions and the transported EAS pollution influenced O₃ in the western and the eastern US differently. Free-running (i.e., without chemical data assimilation) global models underpredicted the transported background O₃ during these episodes, posing difficulties for STEM to accurately simulate the surface O₃ and its source contribution. Although we effectively improved the modeled O₃ by incorporating satellite O₃ (OMI and MLS) and evaluated the quality of the HTAP2 emission inventory with the Royal Netherlands Meteorological Institute-Ozone Monitoring Instrument (KNMI-OMI) nitrogen dioxide, using observations to evaluate and improve O₃ source attribution still remains to be further explored.

Analysis of the latitudinal variability of tropospheric ozone in the Arctic using the large number of aircraft and ozonesonde observations in early summer 2008

The goal of the paper are to: (1) present tropospheric ozone (O3) climatologies in summer 2008 based on a large amount of measurements, during the International Polar Year when the Polar Study using Aircraft, Remote Sensing, Surface Measurements, and Models of Climate Chemistry, Aerosols, and Transport (POLARCAT) campaigns were conducted (2) investigate the processes that determine O3 concentrations in two different regions (Canada and Greenland) that were thoroughly studied using measurements from 3 aircraft and 7 ozonesonde stations. This paper provides an integrated analysis of these observations and the discussion of the latitudinal and vertical variability of tropospheric ozone north of 55°N during this period is performed using a regional model (WFR-Chem). Ozone, CO and potential vorticity (PV) distributions are extracted from the simulation at the measurement locations. The model is able to reproduce the O3 latitudinal and vertical variability but a negative O3 bias of 6-15 ppbv is found in the free troposphere over 4 km, especially over Canada. Ozone average concentrations are of the order of 65 ppbv at altitudes above 4 km both over Canada and Greenland, while they are less than 50 ppbv in the lower troposphere. The relative influence of stratosphere-troposphere exchange (STE) and of ozone production related to the local biomass burning (BB) emissions is discussed using differences between average values of O3, CO and PV for Southern and Northern Canada or Greenland and two vertical ranges in the troposphere: 0-4 km and 4-8 km. For Canada, the model CO distribution and the weak correlation (< 30%) of O3 and PV suggests that stratosphere-troposphere exchange (STE) is not the major contribution to average tropospheric ozone at latitudes less than 70°N, due to the fact that local biomass burning (BB) emissions were significant during the 2008 summer period. Conversely over Greenland, significant STE is found according to the better O3 versus PV correlation (> 40%) and the higher 75th PV percentile. A weak negative latitudinal summer ozone gradient -6 to -8 ppbv is found over Canada in the mid troposphere between 4 and 8 km. This is attributed to an efficient O3 photochemical production due to the BB emissions at latitudes less than 65°N, while STE contribution is more homogeneous in the latitude range 55°N to 70°N. A positive ozone latitudinal gradient of 12 ppbv is observed in the same altitude range over Greenland not because of an increasing latitudinal influence of STE, but because of different long range transport from multiple mid-latitude sources (North America, Europe and even Asia for latitudes higher than 77°N).

The Stratospheric Water and Ozone Satellite Homogenized (SWOOSH) database: a long-term database for climate studies

In this paper, we describe the construction of the Stratospheric Water and Ozone Satellite Homogenized (SWOOSH) database, which includes vertically resolved ozone and water vapor data from a subset of the limb profiling satellite instruments operating since the 1980s. The primary SWOOSH products are zonal-mean monthly-mean time series of water vapor and ozone mixing ratio on pressure levels (12 levels per decade from 316 to 1 hPa). The SWOOSH pressure level products are provided on several independent zonal-mean grids (2.5, 5, and 10°), and additional products include two coarse 3-D griddings (30° long × 10° lat, 20° × 5°) as well as a zonal-mean isentropic product. SWOOSH includes both individual satellite source data as well as a merged data product. A key aspect of the merged product is that the source records are homogenized to account for inter-satellite biases and to minimize artificial jumps in the record. We describe the SWOOSH homogenization process, which involves adjusting the satellite data records to a "reference" satellite using coincident observations during time periods of instrument overlap. The reference satellite is chosen based on the best agreement with independent balloon-based sounding measurements, with the goal of producing a long-term data record that is both homogeneous (i.e., with minimal artificial jumps in time) and accurate (i.e., unbiased). This paper details the choice of reference measurements, homogenization, and gridding process involved in the construction of the combined SWOOSH product and also presents the ancillary information stored in SWOOSH that can be used in future studies of water vapor and ozone variability. Furthermore, a discussion of uncertainties in the combined SWOOSH record is presented, and examples of the SWOOSH record are provided to illustrate its use for studies of ozone and water vapor variability on interannual to decadal timescales. The version 2.5 SWOOSH data are publicly available at doi:10.7289/V5TD9VBX.