Overview and update of the SPARC Data Initiative: Comparison of stratospheric composition measurements from satellite limb sounders

Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 © Author(s) 2021. This work is distributed under the Creative Commons Attribution 4.0 License. Overview and update of the SPARC Data Initiative: comparison of stratospheric composition measurements from satellite limb sounders Michaela I. Hegglin1, Susann Tegtmeier2, John Anderson3, Adam E. Bourassa2, Samuel Brohede4, Doug Degenstein2, Lucien Froidevaux5, Bernd Funke6, John Gille7, Yasuko Kasai8, Erkki T. Kyrölä9, Jerry Lumpe10, Donal Murtagh11, Jessica L. Neu5, Kristell Pérot11, Ellis E. Remsberg12, Alexei Rozanov13, Matthew Toohey2, Joachim Urban11,, Thomas von Clarmann14, Kaley A. Walker15, Hsiang-Jui Wang16, Carlo Arosio13, Robert Damadeo12, Ryan A. Fuller5, Gretchen Lingenfelser12,, Christopher McLinden17, Diane Pendlebury17, Chris Roth2, Niall J. Ryan15, Christopher Sioris17, Lesley Smith18, and Katja Weigel13 1Department of Meteorology, University of Reading, Reading, UK 2Institute of Space and Atmospheric Studies, University of Saskatchewan, Saskatoon, Canada 3Atmospheric Science, Hampton University, Hampton, VA, USA 4FluxSense AB, Gothenburg, Sweden 5Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA 6Instituto de Astrofísica de Andalucía, CSIC, Granada, Spain 7National Center for Atmospheric Research, Boulder, CO, USA 8National Institute of Information and Communications Technology, Tokyo, Japan 9Earth observation, Finnish Meteorological Institute, Helsinki, Finland 10Computational Physics, Inc., Boulder, CO, USA 11Department of Space, Earth, and Environment, Chalmers University of Technology, Gothenburg, Sweden 12NASA Langley Research Center, Hampton, VA, USA 13Institute of Environmental Physics (IUP), University of Bremen, Bremen, Germany 14Karlsruhe Institute of Technology, IMK, Karlsruhe, Germany 15Department of Physics, University of Toronto, Toronto, Canada 16School of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta, GA, USA 17Environment and Climate Change Canada, Toronto, Canada 18Cooperative Institute for Research in Environmental Sciences, University of Colorado, Boulder, CO, USA deceased retired Correspondence: Michaela I. Hegglin (m.i.he[email protected]) Received: 10 November 2020 – Discussion started: 21 November 2020 Revised: 15 March 2021 – Accepted: 24 March 2021 – Published: 5 May 2021 Abstract. The Stratosphere-troposphere Processes and their Role in Climate (SPARC) Data Initiative (SPARC, 2017) performed the first comprehensive assessment of currently available stratospheric composition measurements obtained from an international suite of space-based limb sounders. The initiative’s main objectives were (1) to assess the state of data availability, (2) to compile time series of vertically resolved, zonal monthly mean trace gas and aerosol fields, and (3) to perform a detailed intercomparison of these time series, summarizing useful information and highlighting differences among datasets. The datasets extend over the region from the upper troposphere to the lower mesosphere (300–0.1 hPa) and are provided on a common latitude–pressure grid. Published by Copernicus Publications. 1856 M. I. Hegglin et al.: SPARC Data Initiative overview They cover 26 different atmospheric constituents including the stratospheric trace gases of primary interest, ozone (O3) and water vapor (H2O), major long-lived trace gases (SF6, N2O, HF, CCl3F, CCl2F2, NOy), trace gases with intermediate lifetimes (HCl, CH4, CO, HNO3), and shorter-lived trace gases important to stratospheric chemistry including nitrogen-containing species (NO, NO2, NOx, N2O5, HNO4), halogens (BrO, ClO, ClONO2, HOCl), and other minor species (OH, HO2, CH2O, CH3CN), and aerosol. This overview of the SPARC Data Initiative introduces the updated versions of the SPARC Data Initiative time series for the extended time period 1979–2018 and provides information on the satellite instruments included in the assessment: LIMS, SAGE I/II/III, HALOE, UARS-MLS, POAM II/III, OSIRIS, SMR, MIPAS, GOMOS, SCIAMACHY, ACE-FTS, ACEMAESTRO, Aura-MLS, HIRDLS, SMILES, and OMPS-LP. It describes the Data Initiative’s top-down climatological validation approach to compare stratospheric composition measurements based on zonal monthly mean fields, which provides upper bounds to relative inter-instrument biases and an assessment of how well the instruments are able to capture geophysical features of the stratosphere. An update to previously published evaluations of O3and H2O monthly mean time series is provided. In addition, example trace gas evaluations of methane (CH4), carbon monoxide (CO), a set of nitrogen species (NO, NO2, and HNO3), the reactive nitrogen family (NOy), and hydroperoxyl (HO2) are presented. The results highlight the quality, strengths and weaknesses, and representativeness of the different datasets. As a summary, the current state of our knowledge of stratospheric composition and variability is provided based on the overall consistency between the datasets. As such, the SPARC Data Initiative datasets and evaluations can serve as an atlas or reference of stratospheric composition and variability during the “golden age” of atmospheric limb sounding. The updated SPARC Data Initiative zonal monthly mean time series for each instrument are publicly available and accessible via the Zenodo data archive (Hegglin et al., 2020). 1 Introduction The past four decades starting in the late 1970s represent a “golden age” of stratospheric composition measurements from satellite limb sounders, which capture the spatiotemporal structure of stratospheric composition with a vertical resolution of approximately 1 to 5km. These limb observations have been used extensively to monitor the state of the stratospheric ozone layer that protects human and ecosystem health (e.g., Randel et al., 1999; Harris et al., 2015; WMO, 2011, 2014, 2018) and to study the processes leading to anthropogenic ozone depletion (e.g., Manney et al., 1994; Dessler et al., 1995; Santee et al., 2008). Such research provided the crucial science basis that underpinned actions taken under the Montreal Protocol and its amendments for the protection of the ozone layer, which is considered to be the most successful international treaty on an environmental issue to date. Limb observations, and merged products thereof, are also becoming increasingly important for the detection and attribution of climate change and potential feedback mechanisms, including the role of stratospheric water vapor and aerosol trends and variability in radiative forcing of climate (e.g., Solomon et al., 2010, 2011; Gilford et al., 2016;Schmidt et al., 2018). More generally, limb observations are used for the study of stratospheric dynamics and transport (e.g., Gray and Pyle, 1986; Solomon et al., 1986; Holton and Choi, 1988; Funke et al., 2005a; Manney et al., 2009), empirical studies of stratospheric climate and variability (e.g., Randel et al., 2006, 2010; Randel and Thompson, 2011; Manney et al., 2008; Hegglin et al., 2009; Bourassa et al., 2010; Stiller et al., 2012; Gille et al., 2014), data merging, and trend evaluation activities (e.g., Randel and Wu, 1999; Hegglin et al., 2014; Shepherd et al., 2014; Froidevaux et al., 2015; Harris et al., 2015; Davis et al., 2016; Arosio et al., 2019; SPARC, 2019), with merged datasets also being used as forcing databases in climate models (e.g., Cionni et al., 2011, for ozone; Thomason et al., 2018, for aerosol) and for the validation of the representation of transport and chemistry in numerical models (e.g., Eyring et al., 2006; Gettelman et al., 2010; Hegglin et al., 2010; Strahan et al., 2011; Kolonjari et al., 2018; Froidevaux et al., 2019). The validity of any data and trend analysis, however, strongly depends on the understanding of the observational uncertainty and overall quality of the datasets used, which hitherto was deemed unsatisfactory (SPARC, 2010). Uncertainty and bias estimates are particularly important to inform chemical data assimilation systems (Inness et al., 2013; Errera et al., 2016) and to develop observational metrics for the evaluation of model performance (Douglass et al., 1999; Waugh and Eyring, 2008). In response to this need, the Stratosphere-troposphere Processes and their Role in Climate (SPARC) core project of the World Climate Research Programme (WCRP) initiated the SPARC Data Initiative with the aim to coordinate a comprehensive assessment of available vertically resolved chemical trace gas and aerosol observations obtained from an international suite of satellite limb sounders. The SPARC Data Initiative’s main objectives were (1) to assess the availability of datasets, (2) to compile time series of vertically resolved, zonal monthly mean trace gas and aerosol fields, and (3) to perform a detailed Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1857 intercomparison of these time series, summarizing useful information and highlighting differences among datasets. The SPARC Data Initiative thereby complements other SPARC activities that have focused on the assessment of stratospheric ozone (e.g., Harris et al., 2015; SPARC, 2019), water vapor (SPARC, 2000; Khosrawi et al., 2018; Lossow et al., 2019), and aerosol (SPARC, 2006; Kremser et al., 2016). The provision of error estimates for atmospheric temperature and composition measurements from space following a unified methodological approach, which was highlighted by the SPARC Data Initiative (SPARC, 2017) to be a missing component of its analysis, is now the focus of the SPARC Towards Unified Error Reporting (TUNER) Initiative (von Clarmann et al., 2020). A first application of the SPARC Data Initiative zonal monthly mean time series is the evaluation of stratospheric ozone and water vapor in global reanalyses (Davis et al., 2017) as part of the SPARC Reanalysis Intercomparison Project (S-RIP) (Fujiwara et al., 2017). SPARC Data Initiative gridded datasets have also been contributed to the annual State of the Climate reports in the Bulletin of the American Meteorological Society (Blunden and Arndt, 2019, 2020). Here, we present an update of the SPARC Data Initiative (SPARC, 2017), which focused on composition measurements from 1979–2010, extending its evaluation of the gridded data time series up to the end of 2018 (see Fig. 1). The update features gridded datasets based on more recent retrieval versions and adds the observations of OMPS-LP (on Suomi-NPP) and SAGE III/ISS to the original list of satellite limb sounders presented in SPARC (2017) (LIMS, SAGE I/II/III, HALOE, UARS-MLS, POAM II/III, OSIRIS, SMR, MIPAS, GOMOS, SCIAMACHY, ACE-FTS, ACEMAESTRO, Aura-MLS, HIRDLS, and SMILES; see Sect. 2 for the full definitions of these acronyms). The gridded datasets include the stratospheric trace gases of primary interest (O3and H2O), major long-lived trace gases (SF6, N2O, HF, CCl3F, CCl2F2, NOy), trace gases with intermediate lifetimes (HCl, CH4, CO, HNO3), and shorterlived trace gases important to stratospheric chemistry including nitrogen-containing species (NO, NO2, NOx, N2O5, HNO4), halogens (BrO, ClO, ClONO2, HOCl) and other minor species (OH, HO2, CH2O, CH3CN), and aerosol. The observations considered have been compiled on a common latitude–pressure grid, covering the region from the upper troposphere to the lower mesosphere (300–0.1 hPa) with a latitudinal resolution of 5◦. A summary of the available trace gas and aerosol gridded datasets from each instrument is given in Fig. 2. Almost half of these are based on newer data versions than those used in SPARC (2017) (highlighted in Fig. 2 and with details provided in Tables 1 and 2). The data are published via Zenodo (https://doi.org/10.5281/zenodo.4265393, Hegglin et al., 2020). Note that early data versions of chemical trace gases (i.e., research products) are not included (except for the SAGE III/ISS H2O product) and many more species could be made available. Also, there are a handful of early satellite limb sounders such as the Stratospheric and Mesospheric Sounder (SAMS) on Nimbus 7 (Jones et al., 1986), the Improved SAMS (ISAMS) (Taylor et al., 1993) and the Cryogenic Limb Array Etalon Spectrometer (CLAES) (Roche et al., 1993) on UARS, the Atmospheric Trace Molecule Spectroscopy (ATMOS) (Gunson et al., 1996), and the Millimeter-Wave Atmospheric Sounder (MAS) (Hartmann et al., 1996) on the Atlas Space Shuttle missions, and the Improved Limb Atmospheric Spectrometer (ILAS) on the Advanced Earth Observing Satellite (ADEOS) (Sasano et al., 1999) that could not be evaluated in this assessment due to a lack of resources and generally shorter time series than those from other datasets. The paper is organized as follows. Section 2 provides information on the participating satellite instruments, which vary in terms of measurement method, geographical coverage, spatial and temporal sampling and resolution, time period, and retrieval algorithm. The methodology used to create and compare the trace gas and aerosol time series is described in Sect. 3. The SPARC Data Initiative introduced a top-down climatological validation approach to the evaluation of stratospheric composition measurements (Hegglin et al., 2013; Tegtmeier et al., 2013, 2016; SPARC, 2017), based on the comparison of gridded trace gas and aerosol datasets. This top-down approach complements (but does not replace) the more traditional validation approach that uses coincident profile measurements and sometimes focuses on bottom-up error budgets to characterize measurement uncertainty. The top-down climatological validation approach has the advantages that it is consistent between all instruments, avoids sensitivity to arbitrary coincidence criteria, and generally produces larger sample sizes, which minimizes the random part of the measurement error (or in other words, cancels any kind of random fluctuations). The information gained from the SPARC Data Initiative approach thereby allows us to obtain upper bounds of systematic biases between instruments by reducing the noise from single measurements through averaging. Importantly, it enables assessing the latitude dependence of these systematic biases. This work also provides unique information on how well the different instruments are capable of capturing distinct chemical and geophysical features in stratospheric composition, with the consistency among the instruments constraining our current knowledge of the state of the stratosphere. Section 4 includes example trace gas evaluations of the longer-lived trace gases ozone (Sect. 4.1), water vapor (Sect. 4.2), and methane (CH4; Sect. 4.3); and the mediumto shorter-lived trace gases carbon monoxide (CO; Sect. 4.4), nitrogen-containing species (NO, NO2, NOx, HNO3, and NOy; Sect. 4.5), and also hydroperoxyl (HO2; Sect. 4.6). These evaluations all use updated versions of the datasets used in SPARC (2017), with differences to the old versions highlighted. A summary and conclusions of the updated and evaluated SPARC Data Initiative data, including an overview https://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1858 M. I. Hegglin et al.: SPARC Data Initiative overview Figure 1. Mission lifetime of limb satellite instruments (left-hand side of bars) evaluated within the SPARC Data Initiative. Also indicated are the mission platforms (right-hand side of bars). The colors classify the instruments according to their observation geometry. Note that the SPARC Data Initiative Report (SPARC, 2017) only evaluated zonal monthly mean datasets up to 2010. Here, we evaluate the datasets up to 2018. Seven satellite limb sounders currently remain in space: Aura-MLS, ACE-FTS, ACE-MAESTRO, Odin/SMR, OSIRIS, OMPS, and SAGE III/ISS, of which the first five long passed their expected lifetimes. Figure 2. Colored boxes indicate SPARC Data Initiative zonal monthly mean time series of atmospheric constituents, listed by instruments and available from Zenodo (https://doi.org/10.5281/zenodo.4265393, Hegglin et al., 2020). Dark blue are time series originally submitted and evaluated in SPARC (2017), light blue updated time series based on new data versions, and orange newly added instruments and/or time series. Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1859 Table 1. Data versions used for the construction of the gridded zonal monthly mean datasets submitted to the SPARC Data Initiative assessment and as deposited in the Zenodo data archive (except for aerosol). Italics indicate data versions that have been updated since SPARC (2017). Bold italics indicate datasets that have been added recently, i.e., those which were not part of SPARC (2017). Instrument O3H2O CH4N2O CCl3F CCl2F2CO HF SF6NO NO2NOxHNO3 ACE-FTS v3.6 v3.6 v3.6 v3.6 v3.6 v3.6 v3.6 v3.6 v3.6 v3.6 v3.6 v3.6 v3.6 Aura-MLS v4.2 v4.2 v4.2 v4.2 v4.2 GOMOS v6.01 v6.01 HALOE v19 v19 v19 v19 v19 v19 v19 HIRDLS v7.0 v7.0 v7.0 v7.0 v7.0 v7.0 v7.0 LIMS v6.0 v6.0 v6.0 v6.0 ACE-MAESTRO v3.13 v31 MIPAS(1) v21 v20 v21 v21 v20 v20 v20 v20 v20 v20 v20 v22 MIPAS(2) v224 v220 v224 v224 v220 v220 v220 v222 v220 v220 v220 v224 OSIRIS v5.10 v3.0 v3.0 POAM II v6.0 v6.0 POAM III v4.0 v4.0 v4.0 SAGE I v5.9 SAGE II v7.0 v7.0 v7.0 SAGE III v4.0 v4.0 v4.0 SCIAMACHY v3-5 v4-2 v4-0 v4-0 SMILES v2.1.5 v2.0.1 Odin/SMR(1) v3.1 v2.1 v2.1 v2.1 v2.1 v2.0 Odin/SMR(2) v2.0 UARS-MLS v5 v6 v6 IUP-OMPS v2-6 USask-OMPS v1.1.0 SAGE III/ISS v5.1 v5.1 of our knowledge of the mean state of atmospheric trace gas distributions, are given in Sects. 5 and 6. Note that, due to the complicating factor that aerosol extinction measurements are wavelength dependent, the aerosol evaluations are based on a modified comparison approach, which will be presented in a follow-on publication. In addition to this paper, a special issue in the Journal of Geophysical Research (JGR) – Atmospheres on the SPARC Data Initiative has presented the evaluations of water vapor (Hegglin et al., 2013), ozone (Tegtmeier et al., 2013), the comparison of ozone from limb sounders with the nadir-viewing Aura Tropospheric Emission Spectrometer (Aura-TES) instrument (Neu et al., 2014), an assessment of the impact of instrument-specific sampling patterns on measurement bias (Toohey et al., 2013), and a single instrument study on SMILES observations (Kreyling et al., 2013). A comparison featuring SPARC Data Initiative datasets of long-lived species CFC-11, CFC-12, HF, and SF6 can be found in Tegtmeier et al. (2016) and the dependence of the standard error of the mean on the sample size for profiles obtained with a non-random sampling pattern in Toohey and von Clarmann (2013). The reader is also referred to the WCRP SPARC Data Initiative Report (SPARC, 2017) which offers the complete assessment of all the different original atmospheric trace gas observations and aerosol, and is accessible online. 2 Satellite instruments The SPARC Data Initiative (SPARC, 2017) originally evaluated observations from 18 different satellite limb sounders and additionally, the nadir sounder Aura-TES (Beer, 2006; Beer et al., 2001). The latter instrument was used for https://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1860 M. I. Hegglin et al.: SPARC Data Initiative overview Table 2. Table 1 continued. Note that Odin NOyv3.0 is based on both OSIRIS and SMR data; hence, it has a double entry. Instrument HNO4N2O5ClONO2NOyHCl ClO HOCl BrO OH HO2CH2O CH3CN Aerosol ACE-FTS v3.6 v3.6 v3.6 v3.6 v3.6 v3.6 Aura-MLS v4.2 v3.3 v3.3 v3.3 v3.3 GOMOS v6.01 AERGOM v1 HALOE v19 v19 HIRDLS v7.0 v7.0 MIPAS(1) v20 v21 v21 v22 v20 v20 v20 MIPAS(2) v220 v222 v222 v224 v220 OSIRIS v3.0 v5 v5.7 POAM II v6.0 POAM III v4.0 SAGE II v7.0 SAGE III/M3M v4.0 SCIAMACHY v4.1 v1.4 SMILES v2.1.5 v2.0.1 v2.1.5 v2.0.1 v2.1.5 v2.0.1 Odin/SMR v3.0 v2.1 v2.1 comparisons in the upper troposphere and lower stratosphere (UTLS) only, focusing on the comparability between limb (with high-vertical-resolution measurements) and nadir sounders (with high-horizontal-resolution measurements) applying observation operators (Neu et al., 2014). In this update, TES is no longer included, but the instruments SAGE III on the ISS (hereafter SAGE III/ISS) and OMPS-LP on Suomi-NPP are added for evaluations including trace gas datasets between 2011 and 2018. The instruments considered here all use passive remote sensing techniques, which are based on the detection of natural radiation emitted from the Sun or stars, or from the atmosphere itself (unlike active sounders such as lidars). The different instruments can be classified according to their observation geometry (limb emission, solar or stellar occultation, limb scattering, or nadir emission) and the wavelengths they are measuring at, as compiled in Table 3. In the following, we provide a short description of each instrument, with the most important instrument characteristics summarized in Tables 4 and 5, and the representative sampling patterns provided in Fig. 3. Note that the vertical range observed can depend on the retrieved species. Further information on the instrument and retrieval algorithms can be found in the SPARC Data Initiative Report (SPARC, 2017). 2.1 LIMS on Nimbus 7 The Limb Infrared Monitor of the Stratosphere (LIMS) instrument was launched aboard the Nimbus 7 satellite in October 1978 (Gille and Russell, 1984). The spacecraft occupied a Sun-synchronous orbit, crossing the ascending node at ∼13:00 local time (LT) and the descending node at ∼23:00 LT, taking observations from 64◦S to 84◦N latitude. LIMS used broadband radiometry to observe infrared limb emission, with two radiometer channels for sensing temperature (atmospheric CO2) centered near 15µm, and further four channels for sensing trace gases: 6–7µm for H2O and NO2, 9–10 µm for O3, and 11–12µm for HNO3(Remsberg et al., 2004). LIMS obtained radiance profiles at every ∼0.8◦latitude along its orbital, tangent-point tracks, yielding ∼260 profiles per orbit with ∼14 orbits per day. LIMS operated successfully from launch through its end date in May 1979, when there was final depletion of the cryogen gas supply for cooling its detectors. 2.2 SAGE I on AEM-2, SAGE II on ERBS, SAGE III on Meteor-3M, and SAGE III on the ISS The Stratospheric Aerosol and Gas Experiment (SAGE) series of instruments consists of five instruments including the Stratospheric Aerosol Measurement (SAM II) on Nimbus 7 that span the period from 1978 through 2005 (McCormick et al., 1989) and after a pause continuing from 2017 to present. Note that SAM II has not been included in the SPARC Data Initiative evaluations. The Stratospheric Aerosol and Gas Experiment I (SAGE I) was launched aboard the Applications Explorer MissionB (AEM-B) satellite in February 1979 (McCormick et al., 1979). The spacecraft was in a ∼600km orbit with an inclination of 560◦that allowed for solar occultation measureEarth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1861 Figure 3. Representative sampling patterns for the instruments are shown in time–latitude space for solar occultation sounders to reflect annual sampling patterns (upper two rows) and in longitude–latitude space for emission/scattering and stellar occultation sounders to reflect daily sampling patterns (lower three rows). Different years or days are chosen to give a sense of change in the observed sampling patterns over time. See also Fig. 1 in Toohey et al. (2013) for the resulting measurement density in latitude–time space for the original SPARC Data Initiative instruments. Note that the sampling patterns of ACE-MAESTRO and POAM III are the same for ACE-FTS and POAM II, respectively, and thus are not shown here. The sampling pattern of SAGE I is very similar to that of SAGE II and HALOE. The gap in the sampling seen in OMPS and SCIAMACHY over South America is the result of the South Atlantic Anomaly, a dip in Earth’s magnetic field that allows charged particles to penetrate lower into the atmosphere and as a consequence causes irregularities in the recorded spectral signals by these instruments. https://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1862 M. I. Hegglin et al.: SPARC Data Initiative overview Table 3. Instruments classified according to their observation geometry and wavelength categories. Only instruments that participated in the SPARC Data Initiative are listed. Microwave/sub-mm Mid-IR Near-IR Vis–UV 100 µm–10 cm 2.5–20 µm 1–2.5 µm <1 µm Limb emission UARS-MLS LIMS Aura-MLS MIPAS Odin/SMR HIRDLS SMILES Solar occultation HALOE POAM II/III POAM II/III ACE-FTS SAGE I/II/III SAGE I/II/III ACE-MAESTRO SAGE III/ISS SAGE III/ISS Stellar or lunar GOMOS occultation SAGE III/ISS SAGE III/ISS Limb scattering SCIAMACHY SCIAMACHY OSIRIS OMPS-LP ments from 79◦S to 79◦N. The SAGE I instrument had four spectral channels centered at wavelengths of 1000, 600, 450, and 385 nm for measurements of aerosol extinction, O3, and NO2concentration profiles. SAGE I made 15 sunrise and 15 sunset measurements per day that each covered a narrow latitude band and were separated by ∼24◦in longitude. It took ∼1.5 months for the SAGE I sampling location to shift from one latitude extreme to the other. While there were sunset measurements during the entire 34-month lifetime of SAGE I, there were only 6 months of sunrise observations due to a spacecraft power problem early in the mission. SAGE I ceased operation in November 1981 due to a power system failure. The Stratospheric Aerosol and Gas Experiment II (SAGE II) was launched aboard the Earth Radiation Budget Satellite (ERBS) in October 1984 (Mauldin III et al., 1985; McCormick et al., 1989). The spacecraft occupied a 57◦inclined orbit at an altitude of ∼610 km that allowed for observations from 80◦S to 80◦N. The SAGE II instrument was a broadband spectrometer that operated in the spectral range of ∼375–1030 nm for aerosol and trace gas observations (Mauldin et al., 1985). SAGE II measured 15 sunrise and 15 sunset measurements each day that covered a narrow latitude band and are separated by ∼24◦in longitude. After late 2000, an azimuthal pointing problem resulted in the instrument operating at half-duty cycle. The ERBS mission was decommissioned in October 2005. The Stratospheric Aerosol and Gas Experiment III (SAGE III/M3M) was launched aboard the Russian Meteor-3M (M3M) spacecraft in December 2001 (Mauldin et al., 1998). The spacecraft was placed on a Sun-synchronous orbit, with an altitude of ∼1020 km, inclination of 99.50◦, and equatorial crossing time (ascending node) at 09:15 local time (LT). The SAGE III/M3M provided both solar and lunar measurements, with satellite sunrise events at 60 to 30◦S and satellite sunset events at 45 to 80◦N. Lunar events varied from pole to pole. The SAGE III instrument used a grating spectrometer that operated in the spectral range of ∼295–1025 nm and a single photodiode near 1550 nm for aerosol and trace gas observations (Mauldin et al., 1998). The M3M spacecraft ceased functioning in January 2006. The Stratospheric Aerosol and Gas Experiment III on the International Space Station (SAGE III/ISS) is the second instrument from the SAGE III project (Mauldin et al., 1998). It was launched on the SpaceX Falcon 9 spacecraft in February 2017. Unlike the first SAGE III instrument on the Meteor-3M spacecraft (SAGE III/M3M), SAGE III/ISS is in a mid-inclination orbit (51.6◦). The solar observations can provide near-global (70◦S–70◦N) measurements on a monthly basis with sampling similar to that of the SAGE II measurements. The SAGE III/ISS uses a grating spectrometer operating between ∼280 and ∼1035 nm as well as a single photodiode covering 1542nm ±15nm to retrieve aerosol and other trace gases (SAGE III ATBD, 2002). It can provide vertical profiles of O3, H2O, NO2, and aerosol extinctions at multiple wavelengths through the solar occultation technique. The lunar occultation measurements can augment the sampling of solar observations with measurements of O3 and NO2, as well as NO3and ClO2. The sampling pattern and resulting monthly and annual sampling density of SAGE III/ISS are shown in Fig. 4, equivalent to what is shown for the other instruments in chap. 2 of the SPARC Data Initiative Report (SPARC, 2017). Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1863 Table 4. Satellite instrument characteristics including satellite platform, observation period, spatial coverage, vertical range, vertical resolution, data density, local time (LT) at the Equator, LT of measurement, inclination, and instrument references. Note that the vertical range is generally species dependent and the vertical resolution is often species and altitude dependent; see Tables 9–16 for details. For instruments providing data on a pressure grid, the vertical range is also given in kilometers. n/a – not applicable Instrument, Obs. Spatial Vertical Vertical Data LT at LT of Incl. References platform period coverage range res. density Equator meas. (prof. per day) LIMS 11/1978 to 64◦S–84◦N 250–0.01 hPa 3.7km 3000 a: 11:51 a: 13:00 99.3◦Gille and Russell (1984) Nimbus 7 05/1979 (daily) 10–80 km d: 23:51 d: 23:00 SAGE I 2/1979 to 75◦S–75◦N surface/cloud top 1km 30 n/a sunrise 56◦McCormick et al. (1979) AEM-2 11/1981 (∼1 month) to 55km sunset SAGE II 10/1984 to 75◦S–75◦N surface/cloud top 1km 30 n/a sunrise 57◦Mauldin III et al. (1985) ERBS 08/2005 (∼1 month) to 70km sunset McCormick et al. (1989) UARS-MLS 10/1991 to 80◦S–80◦N 100–0.01 hPa 3.5–5 km 1318 n/a n/a 57◦Waters et al. (1993) UARS 10/1999 (∼2 months) 17–80km Livesey et al. (2003) HALOE 10/1991 to 75◦S–75◦N 250–0.002 hPa 2.5 km 30 n/a sunrise 57◦Russell et al. (1993) UARS 11/2005 (∼1 month) 10–90km sunset POAM II 10/1993 to 88–63◦S 15–50 km 1 km 28 a: 22:30 n/a 98.7◦Glaccum et al. (1996) SPOT-3 11/1996 55–71◦N d: 10:30 (1 year) POAM III 04/1998 to 88–63◦S 5–60 km 1 km 28 a: 22:30 n/a 98.7◦Lucke et al. (1999) SPOT-4 12/2005 55–71◦N d: 10:30 (over 1 year) SMR 07/2001 to 82◦S–82◦N 100–0.001 hPa 2–6km 600–975 a: 18:30 a: 18:30 97.8◦Murtagh et al. (2002) Odin present (daily) 15–120km d: 06:30 d: 06:30 OSIRIS 10/2001 to 82◦S–82◦N 10–60 km 2 km 300–975 a: 18:30 a: 18:30 97.8◦Murtagh et al. (2002) Odin present (daily, no winter) d: 06:30 d: 06:30 SAGE III 02/2002 to 60–30◦S surface/cloud top 1km 30 a: 09:30 sunrise 99.6◦Mauldin et al. (1998) Meteor-3M 12/2005 40–80◦N to 100 km d: 21:30 sunset Thomason et al. (2010) (∼1 month) https://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1870 M. I. Hegglin et al.: SPARC Data Initiative overview Figure 5. Variables in a typical data file that follows SPARC Data Initiative standards are N2O volume mixing ratio, N2O standard deviation (N2O_SD), N2O number, average day of month, average latitude, and minimum, mean, and maximum local solar time (LST_MIN, LST_MAX, and LST_MEAN). This example shows April data from the 2008 MIPAS zonal monthly mean file. Table 6. Description of the content included in each of the SPARC Data Initiative data files, here using N2O as an example variable. Each file includes the time series of zonal monthly mean data for 1 year. Not-a-number values are filled in with “−999.0”. See Fig. 4 for an example. Note that while much effort has been put into applying a consistent file format across the different instruments, some files may still differ from the description here. Variable name Long name Variable type time time 1-D plev pressure 1-D lat latitude 1-D N2O volume mixing ratio of N2O in air Geo3D N2O_NR number of N2O measurements Geo3D N2O_SD volume mixing ratio of N2O in air standard deviation Geo3D AVE_DOM average day of month Geo2D AVE_LAT average latitude Geo2D LST_MIN minimum local solar time Geo2D LST_MAX maximum local solar time Geo2D LST_MEAN mean of local solar time Geo2D 2019), or sudden stratospheric warmings (SSWs) (e.g., Manney et al., 2009) and which may not be resolved by some instruments, will most likely average out in the zonal monthly mean fields. Nonetheless, it is best to compare zonal mean fields averaged over the exact same years and for the maximum time period for which all instruments overlap (ideally for more than 4–5 years). When this is not possible, as many years as possible should be included, keeping in mind a potential tradeoff with underlying trends in a given trace gas over the time period considered. For most species, Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1871 Table 7. Terminology used to define agreement between instruments with respect to the multi-instrument mean (MIM). Description Deviation from MIM Excellent agreement ±2.5 % Very good agreement ±5 % Good agreement ±10 % Reasonably good agreement ±20 % Considerable disagreement ±50 % Large disagreement ±100 % SPARC (2017) concluded that expected trends are generally smaller than inter-instrument differences. Where the instruments’ temporal coverage allowed for it, inter-instrument differences should be tested for different time periods to get a sense of the influence of temporal inconsistencies in the comparison. Again, SPARC (2017) concluded that the general structure in the different instruments’ biases relative to another did not significantly change. However, there are some examples where the previous conclusion was not applicable. SAGE II versus HALOE differences in particular show inter-instrument differences changed over time, which was indicative of a drift in one of the instruments or an influence of volcanic aerosol that could not be fully accounted for in the retrieval. Note that, in this case, the change in the biases was not attributable to sampling, since the instruments were compared over the same time periods (see SPARC, 2017). Finally, within the SPARC Data Initiative, agreement between instruments is defined using the terminology specified in Table 7. All these numbers indicating a certain level of agreement are with respect to the multi-instrument mean (MIM; see Sect. 3.2.3), so that where two instruments show excellent agreement of ±2.5%, the two instruments could show a maximum difference of 5 % between them. 3.2.2 Evaluation diagnostics A set of standard diagnostics is used to investigate the differences between the time series obtained from the different instruments. The diagnostics include comparisons of annual or zonal monthly mean trace gas fields, vertical and meridional mean profiles, seasonal cycles for a single year or averaged over multiple years, and multi-annual averages of latitude– month evolution. Additional evaluations of interannual variability and known tracer-specific features (such as the taperecorder signal in water vapor or the QBO signal in ozone) which test the physical consistency of the datasets, were also carried out and those not presented here can be found in SPARC (2017). The evaluation methods for the trace gas species time series and more examples are more thoroughly described in Hegglin et al. (2013), Tegtmeier et al. (2013), and SPARC (2017). 3.2.3 Multi-instrument mean reference The SPARC Data Initiative’s approach is to use the MIM as a reference to which all instruments are compared. The MIM is calculated by taking the annual or monthly mean of all available instrument datasets within a given time period of interest, aiming at maximum spatial and temporal data coverage for each instrument in order to limit the impact of sampling bias. Note that the MIM does not represent the best estimate of the atmospheric state but rather is motivated by the need that it does not favor a certain instrument. Most datasets are included in its calculation regardless of their quality and without any weighting applied to them. In particular, the datasets from instruments with sparse sampling have the same weight as datasets from instruments with much higher sampling in the calculation of the MIM. Only if measurements from a particular instrument are deemed unrealistic (i.e., outside the ±3σrange), or if another version of a specific trace gas data product is available from the same instrument, are they not included. The relative percentage differences between the trace gas mixing ratios of an instrument (χi) and the MIM (χMIM) are then given by 100 ·(χi−χMIM)/χMIM.(2) One always has to keep in mind when interpreting relative differences with respect to the MIM that the composition of instruments from which the MIM was calculated may have changed between time periods. Hence, changes in derived differences are not to be interpreted as changes in the performance (or drifts) of an individual instrument. Also, if there is an unphysical behavior in one instrument, the MIM and thus the differences with respect to the MIM of the other instruments will most certainly reflect this unphysical behavior as well, although we have tried to eliminate the largest outliers. Finally, if one instrument does not have global coverage for every month, some sampling biases may be introduced into the MIM (see discussion in Sect. 3.2.1). Due to its changing nature, the MIM is thus not made available via the Zenodo data archive. 3.2.4 Summary evaluation Finally, a summary evaluation (seen in Figs. 15–17 and discussed in Sect. 5) is presented, which provides an estimate of the uncertainty in our knowledge of the atmospheric mean state of a given trace gas. This uncertainty is expressed as the relative standard deviation (i.e., calculated relative to the MIM) over all instrument values at a given latitude–pressure grid point or, in other words, the spread between the datasets around the MIM. https://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1872 M. I. Hegglin et al.: SPARC Data Initiative overview Table 8. Definitions and abbreviations of different atmospheric regions as used in this study. The full height range corresponds to about 9–65 km. The tropopause is latitude dependent (approx. 200–300 hPa in the extratropics and 80–100 hPa in the tropics depending on season), while the transition between the stratosphere and the mesosphere (i.e., the stratopause) is here defined uniformly across all latitudes as the 1 hPa pressure level. Note that the abbreviations are often used in combination (e.g., UTLS for upper troposphere and lower stratosphere and USLM for upper stratosphere and lower mesosphere). Region Abbreviation Lower boundary Upper boundary Upper troposphere UT 300 hPa tropopause Lower stratosphere LS tropopause 30 hPa Middle stratosphere MS 30 hPa 5 hPa Upper stratosphere US 5 hPa 1 hPa Lower mesosphere LM 1 0.1 hPa 4 Examples of SPARC Data Initiative trace gas evaluations The approach of the SPARC Data Initiative for evaluating chemical trace gas datasets from stratospheric limb sounders is illustrated in the following providing updates to the ozone (Tegtmeier et al., 2013) and water vapor evaluations (Hegglin et al., 2013), and presenting additional examples based on CH4, CO, different nitrogen-containing species like NO, NO2, HNO3, and NOy, and HO2measurements. These species were chosen to highlight particular differences in the evaluation approach that were necessary to account for the wide range of average lifetimes valid for the lower stratosphere among the species considered (e.g., 8 years for CH4, 3 months for CO, seconds for HO2). Note that the definitions and abbreviations of different altitude regions in the atmosphere as used throughout this study are given in Table 8. 4.1 Ozone (O3) Ozone is one of the most important trace species in the stratosphere due to its absorption of biologically harmful ultraviolet radiation and its role in determining the temperature structure of the atmosphere. A systematic comparison of the SPARC Data Initiative ozone datasets has been provided in Tegtmeier et al. (2013) and SPARC (2017), revealing that the uncertainty in our knowledge of the O3mean state is smallest in the tropical MS and midlatitude LS and MS (see Table 8 for abbreviations). Notable differences between the datasets, on the other hand, exist in the tropical LS and at high latitudes. Here, the multi-instrument spread increases to ±30 % at the tropical tropopause (hence indicating considerable disagreement between the instruments) and ±15% at polar latitudes (reasonably good agreement), which is partially related to inter-instrumental differences in vertical resolution and geographical sampling. It should be noted that diurnal ozone variations are of ∼10 % below 1hPa and grow with increasing altitude up to more than 100% for upper mesospheric levels (e.g., Wang et al., 1996; Schneider et al., 2005). In addition, the impact of temperature uncertainties on the conversion from altitude to pressure during the gridded dataset production may cause additional errors that are particularly pronounced in the LM. Therefore, the mesospheric ozone observations were not corrected (as was done for the nitrogen-containing species; see Sect. 4.5). Instead, we present the ozone evaluations up to 1 hPa only. An update of Fig. 2 from Tegtmeier et al. (2013) is given in Fig. 6 including new versions of SAGE II, SMR, OSIRIS, MIPAS, GOMOS, SCIAMACHY, ACE-FTS, ACEMAESTRO, Aura-MLS, and HIRDLS ozone datasets. Note that MIPAS measured in a high-spectral-resolution measurement mode between 2002 and 2004 (hereafter called MIPAS(1)), which switched to a low-spectral-resolution measurement mode after 2004 (hereafter called MIPAS(2)). The latter led to the opportunity to measure at a higher vertical resolution. In addition, new datasets obtained from OMPSLP and SAGE III/ISS have been added. Tables 1 and 9 provide detailed information on time period, vertical range, vertical resolution, and other information on the different data versions evaluated here. Overall, the updated datasets agree better with notably smaller differences found for SMR, SCIAMACHY, ACE-FTS, GOMOS, and MIPAS. For SAGE II, the updated data version (v7.0) shows very similar structures in the relative differences to the MIM as version v6.2 used in Tegtmeier et al. (2013), albeit tending to more negative values throughout the atmosphere. Some of the rapid transitions between positive and negative values are a result of the combination of seasonal and diurnal sampling biases during the last few years of the mission (as evaluated here), when sampling became more sparse. For SMR, a new data product (v3.1) is evaluated here, based on frequency mode 2 that monitors the band 544.102– 544.902 GHz. This product has been improved from earlier versions (not included in SPARC, 2017) by adjusting the line broadening constant and removing the pointing offset (Murtagh et al., 2020). Compared to SMR frequency mode 1, version 2.1 ozone product (included in SPARC, 2017), the negative bias of 10%–20 % in the upper stratosphere has been reduced to values of 2.5 %–10% (Fig. 6), thus showing very good to good agreement with the other instruments. The updated MIPAS(2) ozone (v224) benefits from better temEarth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1873 Table 9. Time period, vertical range, vertical resolution, references, and other comments for O3measurements. Note that tp refers to tropopause and c.t. to cloud top in this table. Instrument (version) Time Vertical Vertical References Additional period range resolution comments LIMS (v6.0) 11/1978–05/1979 10–50km 3.7 km Remsberg et al. (2007, 2021) SAGE I (v5.9) 10/1984–08/2005 surface/c.t. to 50km 1–2.5km McCormick et al. (1989) data above 3hPa excluded SAGE II (v7.0) 10/1984–08/2005 surface/c.t. to 70km 1km Wang et al. (2002) Damadeo et al. (2013) UARS-MLS (v5) 10/1991–06/1997 18–45 km 3–4 km Livesey et al. (2003) 45–80 km 5–8 km HALOE (v19) 10/1991–11/2005 tp to 80 km 3.5km Grooß and Russell (2005) data below tp excluded SAGE III/M3M (v4.0) 05/2002–12/2005 surface/c.t. to 70 km 1 km Wang et al. (2006) only solar products POAM II (v6.0) 10/1993–11/1996 15–50km 1 km Lumpe et al. (1997) Rusch et al. (1997) POAM III (v4.0) 04/1998–12/2005 5–60km 1km Lumpe et al. (2002) Randall et al. (2003) SMR (v3.1) 08/2001–present 170–0.3hPa 3–4 km Murtagh et al. (2020) OSIRIS (v5.10) 11/2001–present tp to 59.5 km 2.2–3.5km Bourassa et al. (2018) Degenstein et al. (2009) GOMOS (ALGOM2s) 04/2002–04/2012 15–100km 2–3km Sofieva et al. (2017) MIPAS MIPAS(1) (v21) 03/2002–03/2004 6–68km 3.5–5.0km Steck et al. (2007) change in spectral MIPAS(2) (v224) 01/2005–04/2012 6–70km 2.7–3.5km Laeng et al. (2014) resolution in 2004 SCIAMACHY (v3.5) 09/2002–04/2012 11–25km 3–5 km Jia et al. (2015) ACE-FTS (v3.6) 03/2004–present 5–95km 3–4 km Sheese et al. (2017) ACE-MAESTRO (v3.13) 03/2004–present 5–60km 1–2 km Bognar et al. (2019) Aura-MLS (v4.2) 08/2004–present 261–0.02 hPa 2.5–5km Livesey et al. (2018) Hubert et al. (2016) HIRDLS (v7.0) 02/2005–03/2008 422–0.1hPa 1 km Gille and Gray (2013) data degrade after 12/2007 IUP-OMPS (v2.6) 02/2012 8–60 km 2–5km Arosio et al. (2018) USask-OMPS (v1.1.0) 02/2012–present tp to 58 km 1.5–2km Zawada et al. (2018) SAGE III/ISS (v5.1) 06/2017–present surface/c.t. to 70 km 0.75 km Wang et al. (2020) perature data in the mesosphere and optimization of spectroscopic data for some spectral regions. In comparison to the old MIPAS(2) ozone (v220), differences in the upper stratosphere are now reduced to 2.5 %–5%, which is about half of their original amount. SCIAMACHY provides an updated data version (V3-5) based on a new retrieval algorithm (Jia et al., 2015), which improved the retrievals considerably compared to the previously evaluated version (V2.5; Tegtmeier et al., 2013), with a positive bias in the MS and US now reduced from 10 %–20 % to 2.5 %–10 %. Updated ACE-FTS ozone (v3.6) in the MS and US shows considerably smaller differences to the MIM (mostly up to 5 %, Fig. 6) than the old dataset (v2.2), which had a low bias in the MS of up to 10 % and a high bias in the US of up to 10 %–20 % (Tegtmeier et al., 2013). Interpolation of mixing ratios to the SPARC Data Initiative grid in log pressure, data filtering based on quality flag information (Sheese et al., 2015), and reduced non-physical oscillations in the updated pressure and temperature retrievals all contribute to the improved performance (Koo et al., 2017; Waymark et al., 2014). The ACE-MAESTRO dataset (v3.13), on the other hand, has larger biases than the previously evaluated version (v2.1; Tegtmeier et al., 2013). In particular, the low bias in the LS and the high bias in the US increased from 2.5%–5 % to 10 %–20 % (Fig. 6; see also Bognar et al., 2019). Both MAESTRO versions use ACE-FTS temperature profiles in the retrieval, which requires information on the relative time difference between the measurements. For v3.13, this time difference is determined from MAESTRO O2slant column and ACE-FTS air mass slant column instead of using a conhttps://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1874 M. I. Hegglin et al.: SPARC Data Initiative overview Figure 6. Cross sections of the MIM annual zonal mean ozone for 2003–2018 and differences between the individual instruments and the MIM are shown (update from Fig. 2 in Tegtmeier et al., 2013). The MIM includes SAGE II, HALOE, SMR, OSIRIS, MIPAS(1) and MIPAS(2), GOMOS, SCIAMACHY, ACE-FTS, ACE-MAESTRO, Aura-MLS, HIRDLS, IUP-OMPS, USask-OMPS-LP, and SAGE III/ISS. Note that while none of the instruments cover the full time period, detailed evaluations of shorter time periods (e.g., 2012–2018, 2005–2010) give very similar results. Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1875 stant value based on the best match between the ozone profiles. However, it is not clear if these changes cause the larger biases or if they are related to other issues of the v3.13 processing. This is under investigation. The GOMOS O3dataset (v5.0) used previously has shown a substantial positive bias in the LS (30%) and UT (80 %) (Tegtmeier et al., 2013) due to the high sensitivity of the retrieval algorithms to the aerosol extinction model. The new GOMOS datasets (ALGOM2s; Sofieva et al., 2017) are based on a new O3profile inversion algorithm, which is optimized by enhancing the spectral inversion at visible wavelengths for the UTLS, thus decreasing the impact of the aerosol model. As a result, GOMOS performs much better with excellent agreement in the LS (Fig. 6). In the UT, GOMOS retrieves lower ozone values than the other instruments, with differences to the MIM of 20 % to 50%. New O3data products from IUP-OMPS (Arosio et al., 2018) and USask-OMPS (based on a 2-D retrieval) agree very well with the other datasets in the middle and upper stratosphere (Fig. 6). The two data products are based on different retrieval algorithms but show very similar structure with positive differences of 2.5 %–5 % in the MS and US increasing up to 10%–20 % at the SH high latitudes and higher deviations of up to 50 % in the tropical UTLS. The new O3 data product from SAGE III/ISS (v5.1) agrees also well with the other datasets with a reasonably good agreement up to the US. For this work, the “AO3” product was used because it has reduced noise compared to the “MLR” product, particularly in the UT and US (see Wang et al., 2020 for details). In summary, the updated O3datasets show improved agreement in most regions of the atmosphere (Fig. 15). In particular, the 1σmulti-instrument spread in the UT decreased significantly at all latitudes from ±45 % on average to ±25 %, among other things due to improved GOMOS performance. The region of very good agreement (1σof ±5 %), previously restricted to below 3 hPa, extends now further up into the US reaching the level of 1hPa. In the LM, agreement also improves, with maximum deviations of ±30 % due to POAM III not being included in the updated evaluations. At polar latitudes, however, deviations are still large with maximum values of ±30 % found in the Antarctic LS, indicating considerable disagreement between the datasets. 4.2 Water vapor (H2O) H2O is the single most important natural greenhouse gas and provides a positive feedback to climate change driven by anthropogenic emissions of carbon dioxide and other greenhouse gases. H2O is also a key constituent in atmospheric chemistry as source gas of the hydroxyl (OH) radical, which controls the lifetime of atmospheric pollutants, ozone, and greenhouse gases. A comprehensive assessment of the SPARC Data Initiative H2O gridded datasets has been provided by Hegglin et al. (2013) and SPARC (2017). These evaluations revealed that the uncertainty in our knowledge of the H2O mean state is best in the LS and MS, with a relative uncertainty of only ±2 %–6 %. However, substantial biases were found between the datasets in the LM (±15 %), the polar regions (±10 %– 15 %), and the UTLS below 100hPa (±30 %–50 %), where sampling issues add uncertainty due to large gradients and high natural variability. However, once these biases are removed, the instruments showed very good agreement in the magnitude and structure of interannual variability. Figure 7 shows an update of Fig. 5 from Hegglin et al. (2013) including new data versions for ACE-FTS, AuraMLS, MIPAS(1), and MIPAS(2), SAGE II and SCIAMACHY, and adding new datasets obtained from HIRDLS, ACE-MAESTRO, and SAGE III/ISS. Tables 1 and 10 provide detailed information on data versions, time period, vertical range, vertical resolution, and other information on the different data versions evaluated here. LIMS and UARSMLS (although having been measured during an earlier period) are also added for comparison. All other datasets remain the same. Notable changes in the difference patterns arising from the updated data versions are identified in the following. SAGE II (v7.0) shows large changes when compared to SAGE II (v6.2) used in Hegglin et al. (2013) and SPARC (2017), with positive differences replacing negative differences over large parts of the stratosphere. In the MS, the differences to the MIM have decreased from between −5 % and −10 % (v6.2) to values mostly within ±2.5 % (v7.0) (Fig. 7), now indicating excellent agreement with the other datasets. Much smaller differences compared to the MIM (±5 %), indicating very good agreement, are also found in the UTLS, where large negative biases (>10%–20 %) existed in the previous version (v6.2) (Hegglin et al., 2013). This overall improvement is a consequence of modifying a spectral filter channel correction in the SAGE II retrieval (Thomason et al., 2004) using SAGE III/M3M as the basis for comparison in v7.0 instead of HALOE in v6.2 (Damadeo et al., 2013; see also Hegglin et al., 2014). In the US, on the other hand, differences from the MIM have increased from near zero to 5% and higher. The new MIPAS(1) (V3o_H2O_21) and MIPAS(2) (V5r_H2O_224) data versions show generally very similar features in the differences to the MIM compared to the earlier data versions (V3o_H2O_13 and V5r_H2O_220) used in Hegglin et al. (2013), respectively. MIPAS(2) exhibits some improvements in the tropical US, where differences to the MIM decreased from around 10 % to 5 % in the newer version. MIPAS(1) improved in the LM, where differences to the MIM decreased from >10% in V3o_H2O_13, which was evaluated in Hegglin et al. (2013), to smaller or even slightly negative values (between 2.5 % to −5 %). As a consequence, the new data versions of MIPAS(1) and MIPAS(2) seem more similar in character throughout the stratosphere and LM, except in the UTLS, where MIPAS(2) generally shows positive differences compared to the MIM (>10 %), https://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1876 M. I. Hegglin et al.: SPARC Data Initiative overview Figure 7. Cross sections of the MIM annual zonal mean water vapor for 1998–2008 and differences between the individual instruments and the MIM are shown (update from Fig. 5 in Hegglin et al., 2013). Note that LIMS, UARS-MLS, ACE-MAESTRO, SMR(2), HIRDLS, and SAGE III/ISS are not included in the calculation of the MIM to allow for a more direct comparison with Hegglin et al. (2013). Note that while none of the instruments cover the full time period, detailed evaluations of shorter time periods (e.g., 2005–2010) give very similar results. Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1877 Table 10. Time period, vertical range, vertical resolution, references, and other comments for H2O measurements. Note that tp refers to tropopause, p to pressure, and c.t. to cloud top in this table. Instrument (version) Time Vertical Vertical References Additional period range resolution comments LIMS (v6.0) 11/1978–05/1979 10–50km 3.7 km Remsberg et al. (2009, 2021) SAGE II (v7.0) 10/1984–08/2005 surface/c.t. to 50km 1–2.5km Thomason et al. (2004) data above Damadeo et al. (2013) 3hPa excluded UARS-MLS (v6) 10/1991–03/1993 18–50 km 3–4 km Pumphrey (1999) H2O stops early, 50–80 km 5–7 km radiometer failure HALOE (v19) 10/1991–11/2005 10–80km 3.5 km Grooß and Russell (2005) data below tp are excluded SAGE III (v4.0) 05/2002–12/2005 surface/c.t. to 50km 1.5km Thomason et al. (2010) only solar products used here POAM III (v4.0) 04/1998–12/2005 5–45km 1–2 km Lumpe et al. (2006) Lucke et al. (1999) SMR SMR(2) (v2-0) 07/2001–present 16–20 km 3–4 km Urban (2008) 544 GHz-band SMR(1) (v2-1) 07/2001–present 20–75 km 3 km Urban et al. (2007) 489 GHz-band MIPAS MIPAS(1) (V3o_H2O_21) 03/2002–03/2004 6 km/c.t. to 70 km 4–5 km Milz et al. (2005) change in spectral MIPAS(2) (V5r_H2O_224) 01/2005–04/2012 6 km/c.t. to 70km 2–3.7 km Milz et al. (2009) resolution in 2004 von Clarmann et al. (2009) SCIAMACHY (v4.2) 09/2002–04/2012 11–25km 3–5 km Weigel et al. (2016) Weaver et al. (2019) ACE-FTS (v3.6) 03/2004–present 5 km/c.t. to 101km 3–4 km Sheese et al. (2017) Lossow et al. (2019) ACE-MAESTRO (v31) 03/2004–present 5 km/c.t. to 20km 1–2km Sioris et al. (2010, 2016) Lossow et al. (2019) Aura-MLS (v4.2) 08/2004–present 316–100 hPa 2–3 km Read et al. (2007) 100–0.2 hPa 3–4 km Lambert et al. (2007) <0.1 hPa 6–11 km Livesey et al. (2018) HIRDLS (v7.0) 02/2005–03/2008 100–10hPa 1 km Gille and Gray (2013) values high at p > 100 hPa Lossow et al. (2019) values low at p < 40 hPa SAGE III/ISS (v5.1) 06/2017–present 5 km/c.t.–100km 1.5 km Davis et al. (2021) while MIPAS(1) shows both positive (>5%) and negative differences compared to the MIM (>−5 %) depending on the region. Note that it is expected that the application of averaging kernels would likely improve the comparison (which should be tested in future work). The new ACE-FTS (v3.6) and Aura-MLS (v4.2) data versions both show slight improvements in the UTLS, and AuraMLS also has slightly smaller positive differences to the MIM in the US. The negative bias seen in Aura-MLS around 200 hPa in the evaluation of Hegglin et al. (2013), which extended the findings by Vömel et al. (2007) based on balloon soundings to all latitudes, is, however, still apparent. ACE-MAESTRO (v31), a new instrument in the comparison, shows rather large positive differences to the MIM (mostly >10 %–20 %) across its measurement range in the UTLS (except between approximately 300 and 205hPa in the extratropics, where negative values of a similar size are found). The wet bias in the tropical LS is a known issue for this version of ACE-MAESTRO (Lossow et al., 2019). SCIAMACHY’s negative bias to the MIM of around 10 % found for data version v3.0 by Hegglin et al. (2013) in the NH LS slightly improved in the version evaluated here (v4.0) to 5 %, as well as the positive bias when compared to the MIM in the tropical UTLS (from 20 % to 10 %). HIRDLS (v7.0) exhibits a negative bias of >10% with respect to the MIM extending across the MS, and SMR (v2.0) shows an even larger negative bias of >20 %. While LIMS (v6.0) and UARS-MLS (v6) are not directly comparable to the other instruments due to the time period they measured in, the very different character in the differences still highlights that trends in H2O are of minor importance when compared to inter-instrument differences. UARS-MLS shows a very uniform negative bias with respect to the MIM of −10 %, whereas LIMS exhibits a positive bias https://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1878 M. I. Hegglin et al.: SPARC Data Initiative overview Table 11. Time period, vertical range, vertical resolution, references, and other comments for CH4measurements. Note that tp refers to tropopause and c.t. to cloud top in this table. Instrument (version) Time Vertical Vertical References Additional period range resolution comments HALOE (v19) 10/1991–11/2005 tp up to 80 km 3.5 km Grooß and Russell (2005) MIPAS Glatthor et al. (2005) MIPAS-1 (v21) 03/2002–03/2004 6km/c.t. to 70km 4–5km von Clarmann et al. (2009) change in spectral MIPAS-2 (v224) 01/2005–04/2012 6 km/c.t. to 70 km 2–3.7 km Plieninger et al. (2016) resolution in 2004 ACE-FTS (v3.6) 03/2004–present 5km/c.t. to 75 km 3–4 km Plieninger et al. (2016) Olsen et al. (2016) in the extratropical LS and MS, and a more negative bias across the US and in the tropical LS. A part of the negative H2O bias for the US in LIMS may be due to the increases in CH4and its conversion to H2O during the intervening years. The new dataset (v5.1) of the SAGE III/ISS instrument shows excellent agreement with the MIM across the MS, US, and into the LM (with relative differences of ±2.5 % only), although the data can be less trusted at altitudes above 0.5 hPa, where strong positive relative differences from the MIM (>20 %) are found. This feature persists even when comparing datasets from instruments available during the same years (2017–2018) (including ACE-FTS and AuraMLS) (not shown) and is likely due to some reminiscent profiles that “keel over” to very high values in the USLM, potentially biasing the mean field high. These profiles will be filtered out and/or corrected in future versions. In the UT and LS, SAGE III/ISS (v5.1) shows generally negative differences to the MIM, with the values improving from <−20 % at 300 hPa to −5 % around 30hPa. Overall, the update in the H2O datasets has only led to some small improvements and only in some regions of the atmosphere (see Fig. 15). In the NH LS, the 1σmultiinstrument spread decreased from ±10 % to ±5 %, and in the tropical UTLS from ±20 % to ±10 %, among other reasons due to improved performance of SCIAMACHY and SAGE II. In the US, on the other hand, the multi-instrument spread increased slightly from ±10 % to ±12.5%, most likely due to the changes found in the new data version of SAGE II. 4.3 Methane (CH4) CH4is the most abundant hydrocarbon in the atmosphere, and with a lifetime of around 8 years (Lelieveld et al., 1998) it is considered long lived. It is a very effective greenhouse gas and the second-largest contributor to anthropogenic radiative forcing since pre-industrial times after CO2. CH4is a source gas for stratospheric water vapor (resulting in a positive climate feedback), which affects stratospheric ozone chemistry, and in the troposphere it acts to reduce the atmosphere’s oxidizing capacity. The earliest CH4measurements from space were obtained from SAMS on Nimbus 7 between 1979 and 1981 (Taylor, 1987), followed by measurements from ATMOS starting from the mid-1980s (Gunson et al., 1996), and from ISAMS (Taylor et al., 1993) and CLAES (Roche et al., 1993) on UARS (along with HALOE). As mentioned above, these datasets were not considered in the SPARC Data Initiative. The first vertically resolved satellite datasets of CH4available to the SPARC Data Initiative were made by HALOE in 1991. MIPAS started measuring CH4in 2002, providing about 4 years of overlap (although with a major gap in 2004). From 2004 onwards, there are also ACE-FTS measurements available for comparison. Tables 1 and 11 provide information on the availability of CH4measurements, including data version, time period, height range, vertical resolution, and references relevant for the data product. Figure 8 shows meridional profiles of CH4at different pressure levels for August averaged over 1998–2008. These comparisons provide information on the latitudinal distribution of CH4and latitude–height dependency of the differences between the instruments. At 50 hPa, the instruments tend to agree very well with each other mostly within ±5 %. The same is largely true for the 10 hPa level. In both cases, ACE-FTS (v3.6) and HALOE (v19) agree best with each other, while MIPAS(2) (v224) seems to show somewhat higher differences from the MIM than from the other instruments and also exhibits differences that vary more with latitude. At 5 hPa, however, the differences of the instruments with respect to the MIM increase to ±20 %. Here, HALOE is closest to the MIM, ACE-FTS shows largest negative values and MIPAS(1) (v21) largest positive values. The deterioration in the agreement between the instruments with height is qualitatively consistent with the results of SPARC (2017). However, the new data versions used here agree quantitatively much better with each other, particularly at the 50 and 10 hPa levels. We now turn to an example which can be used to test the physical consistency of the available datasets. To this end, the latitude–time evolution of CH4for the different instruments at 2 hPa is shown in Fig. 9. ACE-FTS and HALOE fields Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1879 Figure 8. Meridional profiles of zonal monthly mean CH4at 5, 10, and 50hPa and averaged over 1998–2008 are shown for the different instruments and the MIM (a, b, c). Differences between the individual instruments and the MIM are shown in the lower panels (d, e, f). The grey shading indicates where the relative differences are smaller than ±5 % (thus where the datasets show very good to excellent agreement). Error bars indicate the uncertainty in the relative differences based on the SEM of each instrument. Figure 9. Latitude–time evolution of zonal monthly mean CH4at 2 hPa and averaged over 1998–2008. Shown are absolute values for the MIM (top panel) and the different instruments (middle row), and for the relative differences with respect to the MIM (lower row). Note that HALOE and ACE-FTS show linearly interpolated fields, with hatched regions indicating where no measurements are available. https://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1886 M. I. Hegglin et al.: SPARC Data Initiative overview Table 15. Time period, vertical range, vertical resolution, references, and other comments for HNO3measurements. Instrument Time Vertical Vertical References Additional period range resolution comments LIMS v6.0 11/1978–05/1979 10 km/c.t.–50km 3.7 km Remsberg et al. (2010, 2021) original vertical resolution is 2km but adjusted to make compatible with lower-resolution LIMS products UARS-MLS v6 10/1991–10/1999 100–4.6 hPa 5–10 km Livesey et al. (2003) data with significant (1–3ppbv) low bias at p < 15 hPa and high bias below the VMR peak SMR v2.0 07/2001–present 18–45km 1.5–2 km Urban et al. (2006) empirical Urban et al. (2005) scaling applied MIPAS Mengistu Tsidu et al. (2005) MIPAS V22 03/2002–03/2004 6 km/c.t. 4–6km Wang et al. (2007) change in spectral MIPAS V224 01/2005–04/2012 to 70 km 3–5km von Clarmann et al. (2009) resolution in 2004 ACE-FTS v3.6 03/2004–present 5 km/c.t.–62 km 3–4 km Sheese et al. (2016) Aura-MLS v4.2 08/2004–present 215–1.5 hPa 3–5 km Santee et al. (2007) Livesey et al. (2018) Fiorucci et al. (2013) HIRDLS v7.0 01/2005–01/2008 215–5.1 hPa 1 km Gille and Gray (2013) latitude range 63◦S–80◦N SMILES v2.0.1 10/2009–04/2010 18–45 km 3–4km Kreyling et al. (2013) bias due to problems in spectroscopic parameter and altitude shift can be evaluated without chemical scaling and show mostly a very good agreement of the mean values except for higher ACE-FTS values in the SH midlatitudes during austral winter. The updated HIRDLS dataset (v7.0) shows an improved performance compared to the old data version (v6.0, SPARC, 2017), since the too-low HNO3values during boreal autumn and the resulting semi-annual signal are now removed. For all regions above 30hPa, Aura-MLS and HIRDLS are on the low side, while ACE-FTS, MIPAS, and SMR are on the high side. Below 30 hPa, the situation is reversed. Finally, evaluations of the NOyseasonal cycle (Fig. 12, fifth row) show some severe differences (although not necessarily in the mean value, just the amplitude), most notably in the SH midlatitudes where the seasonal cycle from Odin is completely the opposite of those from ACE-FTS and MIPAS. These deviations can be understood from the OSIRIS NO2 and NOxas well as the SMR HNO3seasonal cycles in the SH, which show a smaller amplitude than the respective MIPAS and ACE-FTS datasets. In general, we expect increasing NOyvalues during the dynamically quiescent springand summertime, and this is observed by ACE-FTS and MIPAS. In the NH, the NOymaximum is observed in boreal autumn by all three instruments. In the SH spring, Odin shows a secondary maximum and an apparently opposite seasonality to the other datasets. For ACE-FTS, the too-low NOxvalues in the SH and NH boreal winter cancel out with the too-high HNO3values, resulting in overall good NOyagreement with MIPAS. The overall annual mean state of NOyis well known, and the three datasets show excellent agreement (Fig. 16) with differences smaller than ±5 %. However, deviations can be larger for individual months (up to ±10%; Fig. 12) and cancel out in the annual mean. Apart from the climatological and seasonal differences between the datasets, it is of interest to evaluate how well the instruments detect signals of interannual variability. Figure 13 shows the time series of NO2mean values (upper panels) and deseasonalized anomalies (lower panels) for the tropical latitude band (20◦S–20◦N) at 10hPa. We focus on the evaluation of the NO2interannual anomalies of the longer time series of SAGE II and HALOE in comparison with interannual variability of ACE-FTS, MIPAS, OSIRIS, SCIAMACHY, GOMOS, and HIRDLS. Anomalies calculated in an additive sense by subtracting monthly multi-year mean values for each month might also display a diurnal cycle and are therefore not suitable evaluation tools for unscaled datasets. However, anomalies calculated in a multiplicative sense as percentage deviations from the monthly multi-year mean values are less affected by the diurnal variations. Since no scaled versions of SAGE II and HALOE data are available, the comparison focuses on multiplicative anomalies of the sunset/nighttime NO2datasets including SAGE II, HALOE, and ACE-FTS local sunset datasets and MIPAS, OSIRIS, SCIAMACHY, GOMOS 22:00LT, and HIRDLS night datasets. The comparison of the mean values (upper panel) shows very good agreement of MIPAS, GOMOS, and scaled SCIAMACHY measurements. Scaled OSIRIS data are somewhat Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1887 Figure 13. Time series of tropical NO2mean values (a) and deseasonalized anomalies (b) between 20◦S–20◦N at 10hPa for 1993–2010. Datasets correspond to local sunset or to 22:00 LST as described in the text. The s10pm denotes zonal monthly mean fields scaled to 22:00LT, the ss zonal monthly mean fields from sunset measurements. lower than those in the other three datasets. Diurnal NO2 variations between 22:00 LT and local sunset at the 10 hPa level are so small that SAGE II, HALOE, and ACE-FTS data taken at local sunset mostly agree with the other datasets for the overlap period (2003–2005). From 2003 onwards, the multiplicative anomalies of all datasets display the expected QBO signal with the best agreement between MIPAS, OSIRIS, GOMOS, and SCIAMACHY. The 3 years of HIRDLS measurements display a larger amplitude of the QBO signal and also larger month-to-month fluctuations, possibly due its higher vertical resolution (which should be tested in future work). Interannual anomalies from ACE-FTS agree for some months with the other datasets but show large deviations for other months. Due to the sparse sampling, it is not possible to diagnose a QBO signal in the ACE-FTS time series. Local sunset evaluations from SAGE II and HALOE show also large month-to-month variations but agree reasonably well on their interannual variability and display the QBO signal over the whole time period. The same is not true, however, for the local sunrise evaluations of the two instruments, where HALOE shows only a weak and SAGE II shows no clear indication of a QBO signal (SPARC, 2017). The overall knowledge on the atmospheric mean state of the different trace gases treated in this section as expressed by the 1σmulti-instrument spread is shown in Fig. 16. In comparison to earlier evaluations (SPARC, 2017), the updated nitrogen datasets show a slightly improved agreement. In particular, the scaled ACE-FTS datasets agree better with the other time series in terms of absolute bias and seasonal cycle. 4.6 Hydroperoxyl (HO2) comparisons Hydroperoxyl (HO2) together with the hydrogen atom (H) and hydroxyl (OH) form the HOxfamily. HO2is formed in the reaction between a hydrogen atom (H) and molecular oxygen (O2), or between ozone (O3) and OH. OH affects stratospheric ozone chemistry through its role in the HOxcatalytic reaction cycle that destroys ozone. The HOxcycle was the first catalytic reaction cycle to be identified (Bates and Nicolet, 1950). HOxchemistry dominates ozone destruction above 40km, while NOxdominates ozone destruction in the MS (Salawitch et al., 2005). In the troposphere, HO2is generated as an intermediate product of the oxidation of many hydrocarbons. Measurements of HO2are available from instruments that measure in the sub-mm/microwave wavelength bands, namely SMILES, SMR, and Aura-MLS. Other available HO2datasets are restricted to balloon campaigns, such as from the Far Infrared Spectrometer (FIRS-2) (Johnson et al., 1995; Jucks et al., 1998). There is no temporal overlap between the three satellite instruments, since SMR currently only provides HO2data as research product during 1 year (October 2003–2004). SMILES, on the other hand, operated between October 2009 to April 2010 only. While SMILES measures the full diurnal cycle, Aura-MLS measures at 01:30 and 13:30 LT, and SMR at 06:30 and 18:30 LT. Since HO2 does not exhibit very strong variations during the day, dayhttps://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1888 M. I. Hegglin et al.: SPARC Data Initiative overview Figure 14. Zonal monthly mean HO2cross sections for Aura-MLS and SMILES daytime (LT) data (left two columns) and Aura-MLS relative differences from the MIM (c, f) are shown for November 2009 (a, b, c) and February 2010 (d, e, f), respectively. Note that the SMILES relative differences from the MIM would look like exact opposites of these figures and thus are not shown. Table 16. Time period, vertical range, vertical resolution, references, and other comments for HO2measurements. Instrument Time Vertical Vertical References Additional period range resolution comments SMR v2 10/2003–10/2004 30–60km 3–4 km Khosravi et al. (2013) (10–0.3 hPa) Aura-MLS v3.3 07/2004–present 22–0.0046hPa 4–10km Pickett et al. (2008) daytime fields Khosravi et al. (2013) with nighttime mean as background correction SMILES v2.0.1 10/2009–04/2010 26–95 km 4–5 km Kreyling et al. (2013) (20–0.001 hPa) Khosravi et al. (2013) Kuribayashi et al. (2013) time datasets are compared only. Tables 2 and 16 compile information on the availability of HOxmeasurements, including data version, time period, height range, vertical resolution, and references relevant for the data product used in this study. Figure 14 shows the zonal monthly mean evaluation between Aura-MLS and SMILES for November 2009 and February 2010. SMR is not shown due to a very limited temporal and spatial coverage (see Fig. 4.23.2 in SPARC, 2017). Mixing ratios are similar in both months in the tropics (where SZAs do not vary much with season), indicating only a weak seasonal cycle in the daytime zonal monthly mean field. Lowest mixing ratios are found in the polar region of the winter hemisphere (during high SZA conditions), indicating a somewhat more pronounced seasonal cycle in these regions of the atmosphere. The differences to the MIM indicate very good (up to ±5%) to excellent (up to ±2.5 %) agreement between SMILES and Aura-MLS, except in the lower part of the measurement range (around 20hPa) where differences compared to the MIM increase to ±10 % and more. The results presented here are comparable to (if not somewhat better than) what was found in SPARC (2017), where multi-year monthly mean HO2fields were used for the comparison. Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1889 Figure 15. Synopsis of the uncertainty in the annual zonal mean state of the longer-lived species evaluated within the SPARC Data Initiative. The relative standard deviation over all instruments’ multi-annual zonal mean datasets is presented for different chemical trace gas species (color contours). The relative standard deviations are calculated by dividing the absolute standard deviations by the MIM. The black contour lines in each panel represent the MIM trace gas distribution for each species. The number of instruments included is given by the right-hand grey bar. Note that the time periods used depend on the availability of the instruments included in the assessment and hence differ from trace gas to trace gas. 5 Summary evaluations The SPARC Data Initiative provides an estimate of the systematic uncertainty in our knowledge of the measured fields’ mean state derived from the inter-instrument spread defined as ±1σ. Figure 15 shows these fields for the long-lived trace gases. Note that we adopt the same vocabulary (see Table 7) for the summary comparisons (based on relative standard deviations) as used earlier for instrument-specific evaluations (based on relative differences). For CH4, the uncertainty is smallest in the tropical and midlatitude MS and LS and larger towards the UTLS, US, and LM. The same has been found for other long-lived trace gases such as O3, H2O, N2O, and HF. In contrast, the trace gases CFC-11 (or CCl3F), CFC-12 (or CCl2F2), and SF6show the best agreement in the UTLS and larger deviations in the MS. Nearly all trace gases show larger deviations in the polar regions than at lower latitudes, which is at least partially due to increased sampling biases found at higher latitudes. Datasets of CO, which is a trace gas with an intermediate lifetime, are characterized by large relative differences throughout most of the measurement range. The large CO differences in the annual zonal mean structure (±30 % in the LS) should be further addressed in forthcoming retrieval revisions. Overall, the ±1σmulti-instrument spread has decreased for all long-lived trace gas species by up to 10 % since SPARC (2017), except possibly for CO, indicating a more consolidated knowledge of the state of the atmosphere resulting from improvements in the retrievals of these species. The agreement of the nitrogen species NO, NO2, and HNO3, as derived from the relative deviations between the datasets, depends strongly on the atmospheric distribution of the respective gas with larger relative differences in regions of smaller mixing ratios (Fig. 16). While NO and NOxagree very well in the tropical and subtropical MS and US, NO2 and HNO3have larger deviations in the US and show the best agreement in the tropical and midlatitude MS and for HNO3 also in the LS. All datasets (except for HNO3and NOyin the https://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1890 M. I. Hegglin et al.: SPARC Data Initiative overview Figure 16. Same as Fig. 14 but for nitrogen-containing species. The assessment of the uncertainty in the annual mean state of NO, NOx, and NO2is based on gridded datasets corresponding to 10:00 and 22:00LT, and for the latter also on datasets corresponding to local sunrise (sr) and local sunset (ss). Note that some of the included datasets have been derived by scaling the individual measurements with a chemical box model to 10:00 and 22:00 local solar time (LST). See SPARC (2017) for more detailed information. Reactive nitrogen (NOx) is here defined as NO+NO2. The odd nitrogen family (NOy) is defined as NOx+HNO3+2×N2O5+ClONO2+HNO4. For N2O5, ClONO2, and HNO4, an assessment of the uncertainty in the annual mean field cannot be provided since no data products at the same local solar time are available. Northern Hemisphere) have considerably larger deviations in the polar regions, at least in part again because of sampling issues and the large atmospheric variability that is less well sampled by the measurements going into the monthly mean datasets (see Toohey et al., 2013). Finally, the NOy datasets show excellent agreement throughout most of the measurement range except for the polar latitude LM. Overall, the ±1σmulti-instrument spread in the nitrogen species has decreased only slightly (by 5%) when compared to SPARC (2017). The agreement between datasets of chlorine compounds (Fig. 17) and shorter-lived species depends strongly on the lifetime of the trace gas considered. HCl, which is longer lived, exhibits very good agreement, and the daytime datasets of the shorter-lived ClO show good to reasonable agreement in the MS and US, where mixing ratios are highest. HOCl, which is short lived, shows mostly reasonable agreement in the US during nighttime. HO2is available from a small number of instruments only and is thus not included in the synopsis plots, although the HO2comparisons show promising results with mostly good agreement throughout the MS, US, and LM. The large deviations between the datasets of shorterlived species stem partially from the difficulty of accounting for the strong diurnal cycles these trace gases exhibit. Scaling of the data to a common daytime or nighttime using a chemical box model helped improve the comparisons in some cases. However, it remains a challenge to estimate how much these deviations are related to errors introduced by the scaling procedures and how many of the deviations correspond to direct measurement differences. Overall, the ±1σ multi-instrument spread in the chlorine-containing species has improved for HCl but has remained very similar for ClO and HOCl when compared to SPARC (2017). Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1891 Figure 17. Same as Fig. 14 but for chlorine-containing species. The assessment of the uncertainty in the annual mean state is based on ClO daytime and HOCl nighttime datasets. Note that for ClO, the dataset from SMR is included which has been derived by scaling the individual measurements with a chemical box model to 13:30LST. See SPARC (2017) for more detailed information. 6 Data availability All SPARC Data Initiative zonal monthly mean datasets can be found in the Zenodo data archive (Hegglin et al., 2020, https://doi.org/10.5281/zenodo.4265393). 7 Conclusions This paper presents an overview and update of the evaluations performed within the WCRP SPARC Data Initiative as published in the SPARC Data Initiative Report (SPARC, 2017). To date, the SPARC Data Initiative represents the most comprehensive assessment of stratospheric composition measurements obtained from an international suite of limb sounders from various space agencies and other national institutions. The SPARC Data Initiative thereby offers the first systematic assessment of the availability of chemical trace gas and aerosol observations from satellite limb sounders, provides these observations in a common and easy-to-handle data format (zonal monthly means), and presents a detailed comparison between these datasets, importantly covering different generations of satellite limb instruments and contrasting the products of different agencies around the world. Here, we extended the SPARC (2017) evaluations, which covered the period 1978–2010, up to the end of 2018 and used the most recent data versions that have become available in the meantime. New observations from OMPS-LP (on Suomi-NPP) and SAGE III/ISS are also added to the original list presented in SPARC (2017), which included LIMS, SAGE I/II, SAGE III/M3M, HALOE, UARS-MLS, POAM II/III, OSIRIS, SMR, MIPAS, GOMOS, SCIAMACHY, ACE-FTS, ACE-MAESTRO, AuraMLS, HIRDLS, and SMILES. (Note that aerosol evaluations and zonal monthly mean time series data will be presented in a follow-on study.) The SPARC Data Initiative comparisons are based on vertically resolved zonal monthly mean datasets of 26 different atmospheric constituents, including the stratospheric trace gases of primary interest (O3and H2O), major long-lived trace gases (SF6, N2O, HF, CCl3F, CCl2F2, NOy), trace gases with intermediate lifetimes (HCl, CH4, CO, HNO3), and shorter-lived trace gases important to stratospheric chemistry including nitrogen-containing species (NO, NO2, NOx, N2O5, HNO4), halogens (BrO, ClO, ClONO2, HOCl) and other minor species (OH, HO2, CH2O, CH3CN), and aerosol. The observations considered have been compiled on a common latitude–pressure grid, covering the region from the upper troposphere to the lower mesosphere (300– 0.1 hPa) with a latitudinal resolution of 5◦. The zonal monthly mean time series are available from the Zenodo data archive (https://doi.org/10.5281/zenodo.4265393, Hegglin et al., 2020). A consistent file format was designed and is being used across the different composition measurements and instruments, so as to allow for easy handling by the user (see Popp et al., 2020, for a discussion of the importance of a consistent data format in the provision of observational datasets). The trace gas time series have then been evaluated by a common approach, comparing multi-year annual or monthly mean fields, allowing for maximum overlap between different instruments. By evaluating zonal monthly mean averages, the SPARC Data Initiative has taken a “climatological” approach to data validation (Hegglin et al., 2008, 2013; Tegtmeier et al., 2013; SPARC, 2017) in contrast to the more common approach of using coincident profile measurements. The climatological comparison method averages over multiple measurements, thereby reducing both instrument noise and geophysical variability from single profile comparisons and offering a top-down instead of a bottom-up assessment of the (systematic) biases between different measurements. Importantly, the climatological validation approach resolves these biases in the full latitude–height space. The climatological validation method has therewith the advantage that it is consistent for all instrument comparisons, avoids sensitivity to chosen limits defining coincident measurements, and produces larger sample sizes, which should in theory minimize the random part of the measurement error. This climatological approach, however, has the disadvantage that climatological means can be biased due to non-uniformity of sampling or potential long-term trends in the trace gases. The extent to which the monthly and annual zonal mean datasets are https://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1892 M. I. Hegglin et al.: SPARC Data Initiative overview representative of the true mean has been evaluated as part of the SPARC Data Initiative for two trace gases (O3and H2O) in a separate paper by Toohey et al. (2013). This study yields information on the potential sampling bias in the zonal monthly mean fields of these tracers and instruments and provides an approximate measure of the sampling bias also for trace gases with similar lifetimes to users who examine variability and trends or perform comparisons with free-running models. The findings of the trace gas datasets comparisons presented here are generally consistent with the results of previous validation efforts based on the classical validation approach using profile coincidences (where available). Instruments with sparser sampling show noisier zonal means. Profiles with wide averaging kernels do not resolve sharp structures such as those found across the tropopause region. However, the climatological approach yields generally more comprehensive information on measurement uncertainty in terms of latitude–pressure range covered. The comparisons of the datasets have in many cases improved our knowledge of the systematic biases between the available data products. Although not shown here, the comparison results generally do not change substantially when changing the number of years going into a averaged field or, in the case of the longerlived species, when calculating instrument differences for a month instead of a year. From this, it follows that the comparisons shown yield relatively robust conclusions about instrument/retrieval performance (see SPARC, 2017, for detailed examples). The conclusions from the SPARC Data Initiative highlight the use (or necessity) of observations from multiple instruments in order to characterize retrieval behavior and overall observation quality as a function of latitude and pressure (or altitude). The small number of stratospheric limb sounders currently remaining in space (with most of them being long past their expected lifetime) and the even smaller number of planned future missions will likely have serious implications. These may impact not only our ability to perform a robust assessment of the quality of stratospheric composition measurements but more importantly to derive stratospheric composition changes from these measurements, which are needed to better understand the state of the ozone layer that protects life on Earth and its response to (as well as feedbacks on) climate change (e.g., Hegglin and Shepherd, 2009). As such, the gridded trace gas datasets from the SPARC Data Initiative may serve as an atlas and reference of stratospheric composition mean state and variability during the “golden age” of limb satellite sounding of the atmosphere well into the future. Author contributions. MIH and ST designed and co-led the SPARC Data Initiative, performed all the evaluations, and wrote the text. The instrument PIs and their research staff compiled the SPARC Data Initiative datasets to their best current knowledge and contributed to the writing and interpretation of the evaluation results. Competing interests. The authors declare that they have no conflict of interest. Acknowledgements. While the SPARC Data Initiative has been driven from a user perspective, the measurement partners have been critical to its success. These partners to whom the SPARC Data Initiative extends its thanks include the relevant instrument teams, the various space agencies (CSA, ESA, NASA, JAXA, SNSA, and other national agencies), and organizations such as CEOS-ACC and IGACO. We thank the World Climate Research Programme (WCRP) for travel funding through the SPARC office to support our activities. The SPARC Data Initiative also thanks the International Space Science Institute in Bern (ISSI) who supported the activity through their ISSI International Team activity program and facilitated two successful team meetings in Bern. Financial support. The research of Michaela I. Hegglin was supported by the CSA SSEP (grant no. 9SCIGRA-29), the ESA STSESPIN (contract no. 4000105291/12/I-NB), and the ESA Water Vapour Climate Change Initiative (contract no. 4000123554). The research of Susann Tegtmeier was funded by the WGL project TransBrom and the EU project SHIVA (grant no. FP7-ENV-20071-226224). Work at the Jet Propulsion Laboratory, California Institute of Technology, was funded by the National Aeronautics and Space Administration (NASA). The Atmospheric Chemistry Experiment is a Canadian-led mission mainly supported by the CSA. Development of the ACE-FTS gridded datasets was supported by grants from the Canadian Foundation for Climate and Atmospheric Sciences and the CSA. MIPAS data analysis and validation was supported by the German Federal Ministry for Economic Affairs and Energy (grant no. 50EE1547) and by the ESA Ozone Climate Change Initiative. Bernd Funke acknowledges support by the Spanish MCINN (grant no. ESP2017-87143-R and PID2019110689RB-I00) and EC FEDER funds. Development of the SCIAMACHY and IUP-OMPS gridded datasets at the University of Bremen was funded in part by the German Research Foundation (DFG) Research Units SHARP (grant no. FOR1095) and VolImpact (grant no. FOR2820), the German Aerospace Agency (DLR) SADOS project, ESA SQWG and Ozone CCI projects, EU/ECMWF C3S project, and the University and State of Bremen. Alexei Rozanov and Carlo Arosio also acknowledge the German HLRN (HighPerformance Computer Center North) and the thread-safe FORTRAN library GALAHAD. In addition, Carlo Arosio acknowledges the support by the PRIME program of the German Academic Exchange Service (DAAD) and ESA’s Living Planet Fellowship SOLVE. Work on HIRDLS was supported in the US by the National Aeronautics and Space Administration (NASA) and in the UK by the National Environmental Research Council (NERC). Development of the Odin/SMR gridded datasets was supported by the Swedish National Space Agency (SNSA). Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1893 Review statement. This paper was edited by David Carlson and reviewed by Sean Davis and two anonymous referees. References Arosio, C., Rozanov, A., Malinina, E., Eichmann, K.-U., von Clarmann, T., and Burrows, J. P.: Retrieval of ozone profiles from OMPS limb scattering observations, Atmos. Meas. Tech., 11, 2135–2149, https://doi.org/10.5194/amt-11-2135-2018, 2018. Arosio, C., Rozanov, A., Malinina, E., Weber, M., and Burrows, J. P.: Merging of ozone profiles from SCIAMACHY, OMPS and SAGE II observations to study stratospheric ozone changes, Atmos. Meas. Tech., 12, 2423–2444, https://doi.org/10.5194/amt12-2423-2019, 2019. Barath, F. T., Chavez, M. C., Cofield, R. E., Flower, D. A., Frerking, M. A., Gram, M. B., Harris, W. M., Holden, J. R., Jarnot, R. F., Kloezeman, W. G., and Klose, G. J.: The upper atmosphere research satellite microwave limb sounder instrument, J. Geophys. Res.-Atmos., 98, 10751–10762, https://doi.org/10.1029/93JD00798, 1993. Bates, D. R. and Nicolet, M.: The photochemistry of atmospheric water vapor, J. Geophys. Res.-Atmos., 55, 301–327, 1950. Bauer, R., Rozanov, A., McLinden, C. A., Gordley, L. L., Lotz, W., Russell III, J. M., Walker, K. A., Zawodny, J. M., LadstätterWeißenmayer, A., Bovensmann, H., and Burrows, J. P.: Validation of SCIAMACHY limb NO2profiles using solar occultation measurements, Atmos. Meas. Tech., 5, 1059–1084, https://doi.org/10.5194/amt-5-1059-2012, 2012. Beer, R.: TES on the Aura mission: scientific objectives, measurements, and analysis overview, IEEE T. Geosci. Remote, 44, 1102–1105, https://doi.org/10.1109/TGRS.2005.863716, 2006. Beer, R., Glavich, T. A., and Rider, D. M.: Tropospheric Emission Spectrometer for the Earth Observing System’s Aura satellite, Appl. Optics, 40, 2356–2367, https://doi.org/10.1364/AO.40.002356, 2001. Belmonte Rivas, M., Veefkind, P., Boersma, F., Levelt, P., Eskes, H., and Gille, J.: Intercomparison of daytime stratospheric NO2 satellite retrievals and model simulations, Atmos. Meas. Tech., 7, 2203–2225, https://doi.org/10.5194/amt-7-2203-2014, 2014. Bernath, P. F.: Atmospheric chemistry experiment (ACE): Analytical chemistry from orbit, TrAC.-Trend. Anal. Chem., 25, 647– 654, 2006. Bernath, P. F., McElroy, C. T., Abrams, M. C., Boone, C. D., Butler, M., Camy-Peyret, C., Carleer, M., Clerbaux, C., Coheur, P.- F., Colin, R., DeCola, P., DeMazière, M., Drunnond, J. R., Dufour, D., Evans, W. F. J., Fast, H., Fussen, D., Gilber, K., Jennings, D. E., Llewellyn, E. J., Lowe, R. P., Mahieu, E., McConnell, J. C., McHugh, M., McLeod, S. D., Michaud, R., Midwinter, C., Nassar, R., Nichitiu, F., Nowlan, C., Rinsland, C. P., Rochon, Y. ., Rolands, N., Sememiuk, K., Simom, P., Skelton, R., Sloan, J. J., Soucy, M.-A., Strong, K., Tremblay, P., Turnbull, D., Walker, K. A., Walkty, I., Wardle, D. A., Wehrle, V., Zander, R., and Zou, J., Atmospheric Chemistry Experiment (ACE): mission overview, Geophys. Res. Lett., 32, L15S01, https://doi.org/10.1029/2005GL022386, 2005. Bertaux, J. L., Kyrölä, E., Fussen, D., Hauchecorne, A., Dalaudier, F., Sofieva, V., Tamminen, J., Vanhellemont, F., Fanton d’Andon, O., Barrot, G., Mangin, A., Blanot, L., Lebrun, J. C., Pérot, K., Fehr, T., Saavedra, L., Leppelmeier, G. W., and Fraisse, R.: Global ozone monitoring by occultation of stars: an overview of GOMOS measurements on ENVISAT, Atmos. Chem. Phys., 10, 12091–12148, https://doi.org/10.5194/acp-10-12091-2010, 2010. Blunden, J. and Arndt, D. S.: State of the Climate in 2018, B. Am. Meteorol. Soc., 100, Si-S306, available at: https://doi.org/ 10.1175/2019BAMSStateoftheClimate.1 (last access: 24 January 2020), 2019. Blunden, J. and Arndt, D. S.: State of the Climate in 2018, B. Am. Meteorol. Soc., 101, Si–S429, https://doi.org/10.1175/2020BAMSStateoftheClimate.1, 2020. Bognar, K., Zhao, X., Strong, K., Boone, C. D., Bourassa, A. E., Degenstein, D. A., Drummond, J. R., Duff, A., Goutail, F., Griffin, D., Jeffery, P. S., Lutsch, E., Manney, G. L., McElroy, C. T., McLinden, C. A., Millán, L. F., Pazmino, A., Sioris, C. E., Walker, K. A., and Zou, J.: Updated validation of ACE and OSIRIS ozone and NO2 measurements in the Arctic using ground-based instruments at Eureka, Canada, J. Quant. Spectrosc. Ra., 238, 106571, https://doi.org/10.1016/j.jqsrt.2019.07.014, 2019. Bourassa, A. E., Degenstein, D. A., Elash, B. J., and Llewellyn, E. J.: Evolution of the stratospheric aerosol enhancement following the eruptions of Okmok and Kasatochi: OdinOSIRIS measurements, J. Geophys. Res.-Atmos., 115, D00L03, https://doi.org/10.1029/2009JD013274, 2010. Bourassa, A. E., Roth, C. Z., Zawada, D. J., Rieger, L. A., McLinden, C. A., and Degenstein, D. A.: Drift-corrected Odin-OSIRIS ozone product: algorithm and updated stratospheric ozone trends, Atmos. Meas. Tech., 11, 489–498, https://doi.org/10.5194/amt11-489-2018, 2018. Bovensmann, H., Burrows, J. P., Buchwitz, M., Frerick, J., Noël, S., Rozanov, V. V., Chance, K. V., and Goede, A. P. H.: SCIAMACHY: Mission objectives and measurement modes, J. Atmos. Sci., 56, 127–150, 1999. Brohede, S., McLinden, C. A., Berthet, G., Haley, C. S., Murtagh, D., and Sioris, C. E.: A stratospheric NO2climatology from Odin/OSIRIS limb-scatter measurements, Can. J. Phys., 85, 1253–1274, https://doi.org/10.1139/P07-141, 2007. Brohede, S., McLinden, C. A., Urban, J., Haley, C. S., Jonsson, A. I., and Murtagh, D.: Odin stratospheric proxy NOymeasurements and climatology, Atmos. Chem. Phys., 8, 5731–5754, https://doi.org/10.5194/acp-8-5731-2008, 2008. Burrows, J. P., Hölzle, E., Goede, A. P. H., Visser, H., and Fricke, W.: SCIAMACHY-Scanning imaging absorption spectrometer for atmospheric chartography, Acta Astronaut., 35, 445–451, https://doi.org/10.1016/0094-5765(94)00278-T, 1995. Choi, W. K. and Holton, J. R.: Transport of N2O in the stratosphere related to the equatorial semiannual oscillation, J. Geophys. Res.- Atmos., 96, 22543–22557, 1991. Cionni, I., Eyring, V., Lamarque, J. F., Randel, W. J., Stevenson, D. S., Wu, F., Bodeker, G. E., Shepherd, T. G., Shindell, D. T., and Waugh, D. W.: Ozone database in support of CMIP5 simulations: results and corresponding radiative forcing, Atmos. Chem. Phys., 11, 11267–11292, https://doi.org/10.5194/acp-11-112672011, 2011. Crutzen, P. J.: The influence of nitrogen oxides on the atmospheric ozone content, Q. J. Roy. Meteor. Soc., 96, 320–325, 1970. https://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1894 M. I. Hegglin et al.: SPARC Data Initiative overview Cunnold, D. M., Zawodny, J. M., Chu, W. P., Pommereau, J. P., Goutail, F., Lenoble, J., McCormick, M. P., Veiga, R. E., Murcray, D., Iwagami, N., Shibasaki, K., Simon, P. C., and Peetermans, W.: Validation of SAGE II NO2 measurements, J. Geophys. Res., 96, 12913–12925, https://doi.org/10.1029/91JD01344, 1991. Damadeo, R. P., Zawodny, J. M., Thomason, L. W., and Iyer, N.: SAGE version 7.0 algorithm: application to SAGE II, Atmos. Meas. Tech., 6, 3539–3561, https://doi.org/10.5194/amt-6-35392013, 2013. Damadeo, R. P., Zawodny, J. M., Remsberg, E. E., and Walker, K. A.: The impact of nonuniform sampling on stratospheric ozone trends derived from occultation instruments, Atmos. Chem. Phys., 18, 535–554, https://doi.org/10.5194/acp-18-535-2018, 2018. Daniel, J. S. and Solomon, S.: On the climate forcing of carbon monoxide, J. Geophys. Res.-Atmos., 103, 13249–13260, 1998. Davis, S. M., Rosenlof, K. H., Hassler, B., Hurst, D. F., Read, W. G., Vömel, H., Selkirk, H., Fujiwara, M., and Damadeo, R.: The Stratospheric Water and Ozone Satellite Homogenized (SWOOSH) database: a long-term database for climate studies, Earth Syst. Sci. Data, 8, 461–490, https://doi.org/10.5194/essd8-461-2016, 2016. Davis, S. M., Hegglin, M. I., Fujiwara, M., Dragani, R., Harada, Y., Kobayashi, C., Long, C., Manney, G. L., Nash, E. R., Potter, G. L., Tegtmeier, S., Wang, T., Wargan, K., and Wright, J. S.: Assessment of upper tropospheric and stratospheric water vapor and ozone in reanalyses as part of S-RIP, Atmos. Chem. Phys., 17, 12743–12778, https://doi.org/10.5194/acp-17-127432017, 2017. Davis, S. M., Damadeo, R., Flittner, D., Rosenlof, K. H., Park, M., Randel, W. J., Hall, E. G., Huber, D., Hurst, D. F., Jordan, A. F., Kizer, S., Millan, L. F., Selkirk, H., Taha, G., Walker, K. A., and Vömel, H.: Validation of SAGE III/ISS solar water vapor data with correlative satellite and balloon-borne measurements, J. Geophys. Res.-Atmos., 126, e2020JD033803, https://doi.org/10.1029/2020JD033803, 2021. Degenstein, D. A., Bourassa, A. E., Roth, C. Z., and Llewellyn, E. J.: Limb scatter ozone retrieval from 10 to 60 km using a multiplicative algebraic reconstruction technique, Atmos. Chem. Phys., 9, 6521–6529, https://doi.org/10.5194/acp-9-6521-2009, 2009. Dessler, A. E., Considine, D. B., Morris, G. A., Schoeberl, M. R., Russell, J. M., Roche, A. E., Kumer, J. B., Mergenthaler, J. L., Waters, J. W., Gille, J. C., and Yue, G. K.: Correlated observations of HCl and ClONO2from UARS and implications for stratospheric chlorine partitioning, Geophys. Res. Lett., 22, 1721–1724, 1995. Diallo, M., Riese, M., Birner, T., Konopka, P., Müller, R., Hegglin, M. I., Santee, M. L., Baldwin, M., Legras, B., and Ploeger, F.: Response of stratospheric water vapor and ozone to the unusual timing of El Niño and the QBO disruption in 2015–2016, Atmos. Chem. Phys., 18, 13055–13073, https://doi.org/10.5194/acp-1813055-2018, 2018. Douglass, A. R., Prather, M. J., Hall, T. M., Strahan, S. E., Rasch, P. J., Sparling, L. C., Coy, L., and Rodriguez, J. M.: Choosing meteorological input for the global modeling initiative assessment of high-speed aircraft, J. Geophys. Res.-Atmos., 104, 27545– 27564, 1999. Dupuy, E., Urban, J., Ricaud, P., Le Flochmoën, É., Lautié, N., Murtagh, D., De La Noë, J., El Amraoui L., Eriksson, P., Forkman, P., Frisk, U., Jégou, F., Jiménez, C., and Olberg, M.: Stratomesospheric measurements of carbonmonoxide with the Odin Sub-Millimetre Radiometer: Retrievaland first results, Geophys. Res. Lett., 31, L20101, https://doi.org/10.1029/2004GL020558, 2004. Errera, Q., Ceccherini, S., Christophe, Y., Chabrillat, S., Hegglin, M. I., Lambert, A., Ménard, R., Raspollini, P., Skachko, S., van Weele, M., and Walker, K. A.: Harmonisation and diagnostics of MIPAS ESA CH4and N2O profiles using data assimilation, Atmos. Meas. Tech., 9, 5895–5909, https://doi.org/10.5194/amt-9-5895-2016, 2016. Eyring, V., Butchart, N., Waugh, D. W., Akiyoshi, H., Austin, J., Bekki, S., Bodeker, G. E., Boville, B. A., Brühl, C., Chipperfield, M. P., Cordero, E., Dameris, M., Deushi, M., Fioletov, V. E., Frith, S. M., Garcia, R. R., Gettelman, A., Giorgetta, M. A., Grewe, V., Jourdain, L., Kinnison, D. E., Mancini, E., Manzini, E., Marchand, M., Marsh, D. R., Nagashima, T., Newman, P. A., Nielsen, J. E., Pawson, S., Pitari, G., Plummer, D. A., Rozanov, E., Schraner, M., Shepherd, T. G., Shibata, K., Stolarski, R. S., Struthers, H., Tian, W., and Yoshiki, M.: Assessment of temperature, trace species and ozone in chemistry-climate model simulations of the recent past, J. Geophys. Res.-Atmos., 111, D22308, https://doi.org/10.1029/2006JD007327, 2006. Fadnavis, S., Schultz, M. G., Semeniuk, K., Mahajan, A. S., Pozzoli, L., Sonbawne, S., Ghude, S. D., Kiefer, M., and Eckert, E.: Trends in peroxyacetyl nitrate (PAN) in the upper troposphere and lower stratosphere over southern Asia during the summer monsoon season: regional impacts, Atmos. Chem. Phys., 14, 12725–12743, https://doi.org/10.5194/acp-14-127252014, 2014. Fahey, D. W., Gao, R. S., Carslaw, K. S., Kettleborough, J., Popp, P. J., Northway, M. J., Holecek, J. C., Ciciora, S. C., McLaughlin, R. J., Thompson, T. L., and Winkler, R. H.: The detection of large HNO3-containing particles in the winter Arctic stratosphere, Science, 291, 1026–1031, 2001. Fiorucci, I., Muscari, G., Froidevaux, L., and Santee, M. L.: Ground-based stratospheric O3and HNO3measurements at Thule, Greenland: an intercomparison with Aura MLS observations, Atmos. Meas. Tech., 6, 2441–2453, https://doi.org/10.5194/amt-6-2441-2013, 2013. Fischer, H., Birk, M., Blom, C., Carli, B., Carlotti, M., von Clarmann, T., Delbouille, L., Dudhia, A., Ehhalt, D., Endemann, M., Flaud, J. M., Gessner, R., Kleinert, A., Koopman, R., Langen, J., López-Puertas, M., Mosner, P., Nett, H., Oelhaf, H., Perron, G., Remedios, J., Ridolfi, M., Stiller, G., and Zander, R.: MIPAS: an instrument for atmospheric and climate research, Atmos. Chem. Phys., 8, 2151–2188, https://doi.org/10.5194/acp-8-21512008, 2008. Flocke, F., Herman, R. L., Salawitch, R. J., Atlas, E., Webster, C. R., Schauffler, S. M., Lueb, R. A., May, R. D., Moyer, E. J., Rosenlof, K. H., Scott, D. C., Blake, D. R., and Bui, T. P., An examination of chemistry and transport processes in the tropical lower stratosphere using observations of long-lived and short-lived compounds obtained during STRAT and POLARIS, J. Geophys. Res., 104, 26625–26642, https://doi.org/10.1029/1999JD900504, 1999. Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1895 Froidevaux, L., Anderson, J., Wang, H.-J., Fuller, R. A., Schwartz, M. J., Santee, M. L., Livesey, N. J., Pumphrey, H. C., Bernath, P. F., Russell III, J. M., and McCormick, M. P.: Global OZone Chemistry And Related trace gas Data records for the Stratosphere (GOZCARDS): methodology and sample results with a focus on HCl, H2O, and O3, Atmos. Chem. Phys., 15, 10471– 10507, https://doi.org/10.5194/acp-15-10471-2015, 2015. Froidevaux, L., Kinnison, D. E., Wang, R., Anderson, J., and Fuller, R. A.: Evaluation of CESM1 (WACCM) free-running and specified dynamics atmospheric composition simulations using global multispecies satellite data records, Atmos. Chem. Phys., 19, 4783–4821, https://doi.org/10.5194/acp-19-4783-2019, 2019. Fujiwara, M., Wright, J. S., Manney, G. L., Gray, L. J., Anstey, J., Birner, T., Davis, S., Gerber, E. P., Harvey, V. L., Hegglin, M. I., Homeyer, C. R., Knox, J. A., Krüger, K., Lambert, A., Long, C. S., Martineau, P., Molod, A., Monge-Sanz, B. M., Santee, M. L., Tegtmeier, S., Chabrillat, S., Tan, D. G. H., Jackson, D. R., Polavarapu, S., Compo, G. P., Dragani, R., Ebisuzaki, W., Harada, Y., Kobayashi, C., McCarty, W., Onogi, K., Pawson, S., Simmons, A., Wargan, K., Whitaker, J. S., and Zou, C.-Z.: Introduction to the SPARC Reanalysis Intercomparison Project (S-RIP) and overview of the reanalysis systems, Atmos. Chem. Phys., 17, 1417–1452, https://doi.org/10.5194/acp17-1417-2017, 2017. Funke, B. and von Clarmann, T.: How to average logarithmic retrievals?, Atmos. Meas. Tech., 5, 831–841, https://doi.org/10.5194/amt-5-831-2012, 2012. Funke, B., López-Puertas, M., Gil-López, S., von Clarmann, T., Stiller, G. P., Fischer, H., and Kellmann, S.: Downward transport of upper atmospheric NOxinto the polar stratosphere and lower mesosphere during the Antarctic 2003 and Arctic 2002/2003 winters, J. Geophys. Res.-Atmos., 110, D24308, https://doi.org/10.1029/2005JD006463, 2005a. Funke, B., López-Puertas, M., Von Clarmann, T., Stiller, G. P., Fischer, H., Glatthor, N., Grabowski, U., Höpfner, M., Kellmann, S., Kiefer, M., and Linden, A.: Retrieval of stratospheric NOxfrom 5.3 and 6.2 mm nonlocal thermodynamic equilibrium emissions measured by Michelson Interferometer for Passive Atmospheric Sounding (MIPAS) on Envisat, J. Geophys. Res.-Atmos., 110, D09302, https://doi.org/10.1029/2004JD005225, 2005b. Funke, B., López-Puertas, M., García-Comas, M., Stiller, G. P., von Clarmann, T., Höpfner, M., Glatthor, N., Grabowski, U., Kellmann, S., and Linden, A.: Carbon monoxide distributions from the upper troposphere to the mesosphere inferred from 4.7 µm non-local thermal equilibrium emissions measured by MIPAS on Envisat, Atmos. Chem. Phys., 9, 2387–2411, https://doi.org/10.5194/acp-9-2387-2009, 2009. Funke, B., López-Puertas, M., Stiller, G. P., and Von Clarmann, T.: Mesospheric and stratospheric NOyproduced by energetic particle precipitation during 2002–2012, J. Geophys. Res.-Atmos., 119, 4429–4446, 2014. Garcia, R. R., Marsh, D. R., Kinnison, D. E., Boville, B. A., and Sassi, F.: Simulation of secular trends in the middle atmosphere, 1950–2003, J. Geophys. Res.-Atmos., 112, D09301, https://doi.org/10.1029/2006JD007485, 2007. Gettelman, A., Hegglin, M. I., Son, S. W., Kim, J., Fujiwara, M., Birner, T., Kremser, S., Rex, M., Añel, J. A., Akiyoshi, H., and Austin, J.: Multi-model assessment of the upper troposphere and lower stratosphere: Tropics and trends, J. Geophys. Res.-Atmos., 115, D00M08, https://doi.org/10.1029/2009JD013638, 2010. Gilford, D. M., Solomon, S., and Portmann, R. W.: Radiative impacts of the 2011 abrupt drops in water vapor and ozone in the tropical tropopause layer, J. Climate, 29, 595–612, 2016. Gille, J.: High Resolution Dynamics Limb Sounder: Experiment overview, recovery, and validation of initial temperature data, J. Geophys. Res., 113, D16S43, https://doi.org/10.1029/2007JD008824, 2008. Gille, J. and Barnett, J.: The High-Resolution Dynamics Limb Sounder (HIRDLS), An instrument for the study of global change, in: The Use of EOS for Studies of Atmospheric Physics, edited by: Gille, J. and Visconti, G., North-Holland, 433–450, 1992. Gille, J. and Gray, L.: High Resolution Dynamics Limb Sounder Earth Observing System (EOS) data description and quality version 7 (v7) (HIRDLS version 7.00.00), Tech. Rep. SC-HIR1511K, Oxford Univ., Univ. of Colo., Nat’l. Cent. for Atmos. Res., available at http://www.eos.ucar.edu/hirdls/data/products/ HIRDLS-DQD_V7-1.pdf (last access: January 2020), 2013. Gille, J. and Russell III, J. M.: The Limb Infrared Monitor of the Stratosphere (LIMS) experiment: Experiment description, performance, and results, J. Geophys. Res.-Atmos., 88, 5125–5140, https://doi.org/10.1029/JD089iD04p05125, 1984. Gille, J., Grey, L., Cavanaugh, C., Coffey, M., Dean, V., Halvorson, C., Karol, S., Khosravi, R., Kinnison, D., Massie, S., Nardi, B., Rivas, M. B., Smith, L., Torpy, B., Waterfall, A., and Wright, C.: HIRDLS data description and quality version 7, available at: http://docserver.gesdisc.eosdis.nasa.gov/repository/Mission/ HIRDLS/3.3_Product_Documentation/3.3.5_Product_Quality/ HIRDLS-DQD_V7.pdf (last access: 7 July 2019), 2012. Gille, J., Karol, S., Kinnison, D., Lamarque, J.-F., and Yudin, V.: The role of midlatitude mixing barriers in creating the annual variation of total ozone in high northern latitudes, J. Geophys. Res.-Atmos., 119, 9578–9595, https://doi.org/10.1002/2013JD021416, 2014. Glaccum, W., Lucke, R. L., Bevilacqua, R. M., Shettle, E. P., Hornstein, J. S., Chen, D. T., Lumpe, J. D., Krigman, S. S., Debrestian, D. J., Fromm, M. D., and Dalaudier, F.: The polar ozone and aerosol measurement instrument, J. Geophys. Res.-Atmos., 101, 14479–14487, 1996. Glatthor, N., von Clarmann, T., Fischer, H., Funke, B., Grabowski, U., Höpfner, M., Kellmann, S., Kiefer, M., Linden, A., Milz, M., Steck, T., Stiller, G. P., Tsidu, G. M., and Wang, D-Y., Mixing Processes during the Antarctic Vortex Split in September– October 2002 as Inferred from Source Gas and Ozone Distributions from ENVISAT–MIPAS, J. Atmos. Sci., 62, 787–800, https://doi.org/10.1175/JAS-3332.1, 2005. Gray, L. J. and Pyle, J. A.: The semi-annual oscillation and equatorial tracer distributions, Q. J. Roy. Meteor. Soc., 112, 387–407, 1986. Grooß, J.-U. and Russell III, J. M.: Technical note: A stratospheric climatology for O3, H2O, CH4, NOx, HCl and HF derived from HALOE measurements, Atmos. Chem. Phys., 5, 2797–2807, https://doi.org/10.5194/acp-5-2797-2005, 2005. Gunson, M. R., Abbas, M. M., Abrams, M. C., Allen, M., Brown, L. R., Brown, T. L., Chang, A. Y., Goldman, A., Irion, F. W., Lowes, L. L., and Mahieu, E.: The Atmospheric Trace Molecule Spectroscopy (ATMOS) experiment: Deployment on the AThttps://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021 1902 M. I. Hegglin et al.: SPARC Data Initiative overview II version 6.2 data set, J. Geophys. Res., 109, D06312, https://doi.org/10.1029/2003JD004465, 2004. Thomason, L. W., Moore, J. R., Pitts, M. C., Zawodny, J. M., and Chiou, E. W.: An evaluation of the SAGE III version 4 aerosol extinction coefficient and water vapor data products, Atmos. Chem. Phys., 10, 2159–2173, https://doi.org/10.5194/acp10-2159-2010, 2010. Thomason, L. W., Ernest, N., Millán, L., Rieger, L., Bourassa, A., Vernier, J.-P., Manney, G., Luo, B., Arfeuille, F., and Peter, T.: A global space-based stratospheric aerosol climatology: 1979–2016, Earth Syst. Sci. Data, 10, 469–492, https://doi.org/10.5194/essd-10-469-2018, 2018. Toohey, M. and von Clarmann, T.: Climatologies from satellite measurements: the impact of orbital sampling on the standard error of the mean, Atmos. Meas. Tech., 6, 937–948, https://doi.org/10.5194/amt-6-937-2013, 2013. Toohey, M., Hegglin, M. I., Tegtmeier, S., Anderson, J., Añel, J. A., Bourassa, A., Brohede, S., Degenstein, D., Froidevaux, L., Fuller, R., and Funke, B.: Characterizing sampling biases in the trace gas climatologies of the SPARC Data Initiative, J. Geophys. Res.-Atmos., 118, 11847–11862, 2013. Tung, K. K.: On the two-dimensional transport of stratospheric trace gases in isentropic coordinates, J. Atmos. Sci., 39, 2330–2355, 1982. Urban, J.: Tropical ascent of lower stratospheric air analysed using measurements of the Odin Sub-Millimetre Radiometer, in: Proc. Reunion Island Int. Symp. Tropical Stratosphere – Upper Troposphere, 5–9 November 2007, edited by: Bencherif, H., Universite de la Reunion, St. Gilles, Reunion Island, France, 29–34, 2008. Urban, J., Lautié, N., Le Flochmoën, E., Jiménez, C., Eriksson, P., de La Noë, J., Dupuy, E., Ekström, M., El Amraoui, L., Frisk, U., Murtagh, D., Olberg, M., and Ricaud, P.: Odin/SMR limb observations of stratospheric trace gases: Level 2 Processing of ClO, N2O, O3, and HNO3, J. Geophys. Res., 110, D14307, https://doi.org/10.1029/2004JD005741, 2005. Urban, J., Lautie, N., Murtagh, D., Eriksson, P., Kasai, Y., Lossow,S., Dupuy, E., de la Noe, J., Frisk, U., Olberg, M., Le Flochmoen, E., and Ricaud, P.: Global observations of middle atmosphericwater vapour by the Odin satellite: An overview, Planet. SpaceSci., 55, 1093–1102, https://doi.org/10.1016/j.pss.2006.11.021, 2007. Urban, J., Pommier, M., Murtagh, D. P., Santee, M. L., and Orsolini, Y. J.: Nitric acid in the stratosphere based on Odin observations from 2001 to 2009 – Part 1, A global climatology, Atmos. Chem. Phys., 9, 7031–7044, https://doi.org/10.5194/acp-9-70312009, 2009. Vömel, H., David, D. E., and Smith, K.: Accuracy of tropospheric and stratospheric water vapor measurements by the cryogenic frost point hygrometer: Instrumental details and observations, J. Geophys. Res.-Atmos., 112, D08305, https://doi.org/10.1029/2006JD007224, 2007. von Clarmann, T., Höpfner, M., Kellmann, S., Linden, A., Chauhan, S., Funke, B., Grabowski, U., Glatthor, N., Kiefer, M., Schieferdecker, T., Stiller, G. P., and Versick, S.: Retrieval of temperature, H2O, O3, HNO3, CH4, N2O, ClONO2and ClO from MIPAS reduced resolution nominal mode limb emission measurements, Atmos. Meas. Tech., 2, 159–175, https://doi.org/10.5194/amt-2159-2009, 2009. von Clarmann, T., Degenstein, D. A., Livesey, N. J., Bender, S., Braverman, A., Butz, A., Compernolle, S., Damadeo, R., Dueck, S., Eriksson, P., Funke, B., Johnson, M. C., Kasai, Y., Keppens, A., Kleinert, A., Kramarova, N. A., Laeng, A., Langerock, B., Payne, V. H., Rozanov, A., Sato, T. O., Schneider, M., Sheese, P., Sofieva, V., Stiller, G. P., von Savigny, C., and Zawada, D.: Overview: Estimating and reporting uncertainties in remotely sensed atmospheric composition and temperature, Atmos. Meas. Tech., 13, 4393–4436, https://doi.org/10.5194/amt13-4393-2020, 2020. Wang, H.-J., Cunnold, D. M., Thomason, L. W., Zawodny, J. M., and Bodeker, G. E.: Assessment of SAGE version 6.1 ozone data quality, J. Geophys. Res.-Atmos., 107, 4691, https://doi.org/10.1029/2002JD002418, 2002. Wang, H.-J., Cunnold, D. M., Trepte, C., Thomason, L. W., and Zawodny J. M.: SAGE III solar ozone measurements: Initial results, Geophys. Res. Lett., 33, L03805, https://doi.org/10.1029/2005GL025099, 2006. Wang, H.-J., Damadeo, R., Flittner, D., Kramarova, N., Taha, G., Davis, S., Thompson, A. M., Strahan, S., Wang, Y., Froidevaux, L., Degenstein, D., Bourassa, A., Steinbrecht, W., and Walker, K. A.: Validation of SAGE III/ISS solar occultation ozone products with correlative satellite and ground based measurements, J. Geophys. Res.-Atmos., 125, e2020JD032430, https://doi.org/10.1029/2020JD032430, 2020. Waters, J. W., Froidevaux, L., Read, W. G., Manney, G. L., Elson, L. E., Flower, D. A., Jarnot, R. F., and Harwood, R. S.: Stratospheric ClO and ozone from the Microwave Limb Sounder on the Upper Atmosphere Research Satellite, Nature, 362, 597–602, https://doi.org/10.1038/362597a0, 1993. Waters, J. W., Read, W. G., Froidevaux, L., Jarnot, R. F., Cofield, R. E., Flower, D. A., Lau, G. K., Pickett, H. M., Santee, M. L., Wu, D. L., Boyles, M. A., Burke, J. R., Lay, R. R., Loo, M. S., Livesey, N. J., Lungu, T. A., Manney, G. L., Nakamura, L., Perun, V. S., Ridenoure, B. P., Shippony, Z., Siegel, P. H., Thurstans, R. P., Harwood, R. S., and Filipiak, M. J.: The UARS and EOS Microwave Limb Sounder Experiments, J. Atmos. Sci., 56, 194–218, https://doi.org/10.1175/15200469(1999)056<0194:TUAEML>2.0.CO;2 , 1999. Waters, J. W., Froidevaux, L., Harwood, R. S., Jarnot, R. F., Pickett, H. M., Read, W. G., Siegel, P. H., Cofield, R. E., Filipiak, M. J., Flower, D. A., Holden, J. R., Lau, G. K., Livesey, N. J., Manney, G. L., Pumphrey, H. C., Santee, M. L., Wu, D. L., Cuddy, D. T., Lay, R. R., Loo, M. S., Perun, V. S., Schwartz, M. J., Stek, P. C., Thurstans, R. P., Boyles, M. A., Chandra, K. M., Chavez, M. C., Gun-Shing Chen and Chudasama, B. V., Dodge, R., Fuller, R. A., Girard, M. A., Jiang, J. H., Yibo Jiang and Knosp, B. W., LaBelle, R. C., Lam, J. C., Lee, K. A., Miller, D., Oswald, J. E., Patel, N. C., Pukala, D. M., Quintero, O., Scaff, D. M., Van Snyder, W., Tope, M. C., Wagner, P. A., and Walch, M. J.: The earth observing system microwave limb sounder (EOS MLS) on the Aura satellite, IEEE T. Geosci. Remote, 44, 1075–1092, 2006. Waugh, D. W. and Eyring, V.: Quantitative performance metrics for stratospheric-resolving chemistry-climate models, Atmos. Chem. Phys., 8, 5699–5713, https://doi.org/10.5194/acp-8-56992008, 2008. Waymark, C., Walker, K. A., Boone, C. D., and Bernath, P. F.: ACEFTS version 3.0 data set: validation and data processing update, Ann. Geophys.-Italy, 56, https://doi.org/10.4401/ag-6339, 2013. Earth Syst. Sci. Data, 13, 1855–1903, 2021 https://doi.org/10.5194/essd-13-1855-2021 M. I. Hegglin et al.: SPARC Data Initiative overview 1903 Weaver, D., Strong, K., Walker, K. A., Sioris, C., Schneider, M., McElroy, C. T., Vömel, H., Sommer, M., Weigel, K., Rozanov, A., Burrows, J. P., Read, W. G., Fishbein, E., and Stiller, G.: Comparison of ground-based and satellite measurements of water vapour vertical profiles over Ellesmere Island, Nunavut, Atmos. Meas. Tech., 12, 4039–4063, https://doi.org/10.5194/amt12-4039-2019, 2019. Weigel, K., Rozanov, A., Azam, F., Bramstedt, K., Damadeo, R., Eichmann, K.-U., Gebhardt, C., Hurst, D., Kraemer, M., Lossow, S., Read, W., Spelten, N., Stiller, G. P., Walker, K. A., Weber, M., Bovensmann, H., and Burrows, J. P.: UTLS water vapour from SCIAMACHY limb measurementsV3.01 (2002–2012), Atmos. Meas. Tech., 9, 133–158, https://doi.org/10.5194/amt-9133-2016, 2016. WMO: Scientific Assessment of Ozone Depletion: 2010, in: Global Ozone Research and Monitoring Project-Report No. 52, World Meteorological Organization, Geneva, Switzerland, 516 pp., 2011. WMO: Scientific Assessment of Ozone Depletion: 2014, in: Global Ozone Research and Monitoring Project-Report No. 55, World Meteorological Organization, Geneva, Switzerland, 416 pp., 2014. WMO: Scientific Assessment of Ozone Depletion: 2018, in: Global Ozone Research and Monitoring Project-Report No. 58, World Meteorological Organization, Geneva, Switzerland, 588 pp., 2018. Zawada, D. J., Rieger, L. A., Bourassa, A. E., and Degenstein, D. A.: Tomographic retrievals of ozone with the OMPS Limb Profiler: algorithm description and preliminary results, Atmos. Meas. Tech., 11, 2375–2393, https://doi.org/10.5194/amt-112375-2018, 2018. https://doi.org/10.5194/essd-13-1855-2021 Earth Syst. Sci. Data, 13, 1855–1903, 2021