OMPS LP Version 2.0 Multi-wavelength Aerosol Extinction Coefficient Retrieval Algorithm

The OMPS Limb Profiler (LP) instrument is designed to provide high vertical resolution ozone and aerosol profiles from measurements of the scattered solar radiation in the 290-1000 nm spectral range. It collected its first Earth limb measurement on January 10, 2012 and continues to provide daily global measurements of ozone and aerosol profiles from the cloud top up to 60 15 km and 40 km respectively. The relatively high vertical and spatial sampling allow detection and tracking of sporadic events when aerosol particles are injected into the stratosphere, such as volcanic eruptions or pyrocumulonimbus (PyroCb) events. In this paper we discuss the newly released Version 2.0 OMPS multi-wavelength aerosol extinction coefficient retrieval algorithm. The algorithm now produces aerosol extinction profiles at 510, 600, 674, 745, 869, and 997 nm wavelengths. The OMPS LP Version 2.0 data products are compared to the SAGE III/ISS, OSIRIS and CALIPSO missions and shown to be of good quality and suitable 20 for scientific studies. The comparison shows significant improvements in the OMPS LP retrieval performance in the Southern Hemisphere and at lower altitudes. These improvements arise from use of the longer wavelengths, in contrast with the V1.0 and V1.5 OMPS aerosol retrieval algorithms, which used radiances only at 675 nm and therefore had limited sensitivity in those regions. In particular, the extinction coefficients at 745, 869 and 997 nm are shown to be the most accurate, with relative accuracies and precisions close to 10% and 15% respectively, while the 675 nm relative accuracy and precision are on the order of 20%. The 25 510 nm extinction coefficient is shown to have limited accuracy in SH and is only recommended for use between 20 24 km, and only in the Northern Hemisphere. The V2.0 retrieval algorithm has been applied to the complete set of OMPS LP measurements and the new data set is publicly available.

description given in Sect. 1.2.1 for ground-based lidars generally applies to the space-borne lidars as well. The main difference is that an orbiting lidar provides the additional advantage of mobility (near global coverage in the case of CALIPSO). The spaceborne CALIPSO lidar instrument is used in this study, and is described further in Sect. 3.3. 115

Summary of available measurements
In 2017, accurate solar occultation measurements of stratospheric aerosols resumed after the deployment of SAGE III on International Space Station (ISS) (Cisewski et al., 2014). Combined with the ongoing OSIRIS and CALIPSO missions, we now have coincident stratospheric aerosol measurements from several space-based platforms. The structure of this paper is as follows: In Section 2 we provide a brief description of OMPS LP instrument and V2.0 algorithm changes. The correlative satellite aerosol 120 measurements are described further in Section 3. Section 4 describes the validation methodology. The comparison results are shown in Section 5, followed by conclusions in Section 6.

Instrument review
The OMPS LP sensor images the Earth limb by pointing aft along the spacecraft flight path to measure the sunlit portion of the 125 globe without directly observing the sun. The sensor employs 3 vertical slits separated horizontally to provide near global coverage in 3 -4 days, and more than 7000 profiles a day. The instrument measures limb scattering radiance at the 290 -1000 nm wavelength range and the 0 -80 km altitude range. The instrument is installed in a fixed orientation relative to the spacecraft, which flies in a sun-synchronous ascending orbit with 1:30 PM equator crossing time. As a result, the observed single scattering angle (SSA) varies along the orbit, where the Northern Hemisphere (NH) observations correspond to forward-scattered solar radiation and the 130 Southern Hemisphere (SH) observations correspond to back-scattered radiation. Therefore, the aerosol scattering signal is much larger in NH than in SH, resulting in a sampling of the aerosol phase function magnitude varying by a factor of 50 over the course of OMPS orbit (Loughman et al., 2018). OMPS LP is scheduled to fly on the NOAA JPSS-2, 3 and 4 satellites, to extend the stratospheric aerosol measurements into the next couple of decades. (These satellite launches are currently targeted for 2022, 2026, and 2031, respectively.) 135

OMPS LP V2.0 algorithm improvement
The Version 2.0 (V2.0) OMPS LP aerosol extinction retrieval algorithm builds upon the Version 1.0 (V1.0) and Version 1. 5 (V1.5) algorithms, which were described in Sect. 4 of Loughman et al. (2018) and Sects. 2-3 of Chen et al. (2018), respectively. We therefore begin by briefly reviewing the V1.0 and V1.5 algorithms in Sect. 2.2.1 and defining the key variables used. This is followed by Sect. 2.2.2, which details the algorithm updates made to produce V2.0. 140 measurements by comparison to two analogous sets of calculated radiances, Ic(λ,h) and Ic0(λ,h). These calculated radiance profiles are generated by the GSLS RTMGauss-Seidel Limb Scattering (GSLS) radiative transfer model (RTM) (Loughman et al., 2004) for the same viewing and solar illumination conditions that existed when Im (λ,h) was measured. The model atmospheres used 150 (described further below) are identical for these two calculations, with one exception: In the case of Ic0 (λ,h), the model atmosphere contains no aerosols, while the Ic(λ,h) model atmosphere contains the first-guess aerosol profile.
The model atmosphere consists of static atmospheric temperature and pressure profiles derived from the operational geopotential height product provided by the NASA Global Modeling and Assimilation Office (GMAO). The algorithm uses the (McPeters and Labow, (2012) ozone climatology and the PRATMO photochemical box model NO2 climatology (McLinden et al., 2000) NO2 155 climatologies to define the model atmosphere. The first-guess aerosol extinction profile, x0, is defined based on a single SAGE climatological profile. Aerosols are assumed to consist of spherical liquid sulfate particles (75% H2SO4) with index of refraction m = 1.448 + 0i (Yue and Deepak, 1983;Wang et al., 1996). In the V1.0 algorithm, the aerosol size distribution (ASD) is assumed to be a bi-modal log-normal distribution (Loughman et al., 2018); this was updated to a gamma distribution in V1.5 (Chen et al., 2018). Mie scattering theory is used to calculate the aerosol phase function based on the assumed ASD and optical properties. 160 The Earth's surface is modeled as a Lambertian surface (for which a fraction, R, of the incident downward radiation is reflected as isotropic, unpolarized upward radiance field at each point). The value of R is determined by requiring that Ic0(λ,h) = Im (λ,h) at h = 40.5 km (Loughman et al., 2018). An approximate ozone correction is also applied to the model radiances to correct for possible ozone error, as described in Sect. 4.3 of Loughman et al. (2018). To reduce the sensitivity of the algorithm to a variety of interfering factors, the radiances are normalized with respect to tangent height h. The measured altitude-normalized radiance (ANR) is defined 165 as ρm(λ,h) = Im(λ,h) / Im (λ,hn), with hn =, namely the normalization tangent height =, set to a value of 40.5 km in the V1.0 and V1.5 algorithms. The value of hn is generally selected as a compromise between two competing interests:. It should be as high as possible (to minimize the atmospheric aerosol extinction at hn),, but not so high that the radiance at hn is poorly characterized (due to residual stray light contamination, low signal-to-noise ratio, etc.). SimilarAnalogous expressions define ρc(λ,h) and ρc0(λ,h), respectively, based on the calculated radiance profiles. 170 As a final step, the ANR values are combined to produce the aerosol scattering index (ASI), which serves as the measurement vector y in the retrieval. The measured ASI is defined as ym (λ,h) = [ρm(λ,h) -ρc0(λ,h)] / ρm (λ,h), with similar definitions for yc (λ,h) and yc0 (λ,h). Since the V1.0 and V1.5 algorithms use a single wavelength, this notation can be abbreviated to ym(λ,hi) = ! " , with a similar abbreviation yc(λ,hi) = ! # used to represent the ASI calculated based on the model atmosphere after n iterations of the retrieval algorithm. 175 The V1.0 and V1.5 algorithms use the Chahine nonlinear relaxation method (Chahine, 1970) to derive the aerosol extinction coefficient (which represents the state vector, x) based on the measurement vector y defined above. The state vector is updated iteratively as shown in Eq. (1): algorithm, when the stability of the retrieval was relatively untested. The V1.5 algorithm relaxed these constraints somewhat, using N = 4 and allowing ! # values between 1/30.2 and 3.0. 185

Updates made for the V2.0 OMPS LP algorithm
Since the limb scattering radiances at visible and near-infrared wavelengths are very sensitive to aerosol properties, the V2.0 OMPS LP aerosol algorithm is modified to include multiple wavelengths in this spectral region, similar to the SAGE III aerosol channels (Thomason and Taha, 2003). The V2.0 algorithm uses OMPS LP measurements at wavelengths 510, 600, 675, 745, 869, and 997 nm, selected to minimize the effect of gaseous absorption, with the exception of 600 nm, which will be used primarily for 190 diagnostics. Each wavelength is retrieved independently, as described in the preceding section leading to Eq. 1. Taha et al., (2010) showed that, because of its strong weighting function or Jacobian matrix, retrieving aerosol profiles at longer wavelengths can improve the quality of the profile in the southern hemisphere, where OMPS LP observes backscattered radiation, and extend the retrieval further down in altitude. The Jacobian matrix quantifies the changes in the radiance with respect to the aerosol extinction.
Multiple wavelength aerosol measurements can also provide limited information about aerosol particle size and can be used to 195 identify cloud presence. Notice that the 997 nm radiance measurements are only available after 26 November 2013.
The assumed ASD is the same in V2.0 as in V1.5, but the single first-guess aerosol extinction profile has been replaced by a firstguess climatology that varies with wavelength, latitude, and season, again based on the SAGE aerosol data record. The V2.0 algorithm further relaxes the constraints that were previously applied to the Chahine iteration results: N = 5 and ! # has an upper bound of 10.0 and no lower bound. The V2.0 algorithm also checks for convergence after each iteration, rather than always 200 performing the stated number of iterations: Iterations end when the retrieved aerosol extinction changes by < 2% at 20 km. or when it reaches maximum number of iterations. The planned V2.1 release next year will use modified convergence criteria that checks for multiple altitudes.
Limb-scatter instruments such as OSIRIS, SCIAMACHY and OMPS LP suffer from increased stray light at increasing wavelength and altitude due in part to diminishing scattered signal (Jaross et al., 2014;. To reduce the stray light effect on 205 the retrieval at longer wavelengths, hn was lowered to 38.5 km in V2.0 (from the 40.5 km value used in previous versions). The GSLS radiative transfer model used in the V2.0 algorithm was also updated as described by Loughman et al. (2015). The main improvement associated with this change involves use of several zeniths to calculate the multiple scattering source function along the limb line of sight, which improves the radiance calculations near the terminator. Unlike the V1.0 and V1.5 algorithms, the V2.0 GSLS model also includes refraction in the line -of -sight calculation. The V2.0 algorithm also excludes polarization (which had 210 been included in the V1.5 radiance calculations). The exclusion of polarization is primarily done for speed purposes: Scalar (unpolarized) radiance calculations are considerably faster than their vector (polarized) counterparts, and the resulting change in ρc(λ,h) is very small. Recent calculations performed for a RTM comparison project (Zawada et al., 2020) allow the ρc(λ,h) values computed by the scalar and vector versions of GSLS to be compared, for a variety of atmospheres and illumination conditions. For the relevant wavelengths (500 nm and greater), these values agree to within 1% or better at 20 km, and within 2% or better at 215 strong sensitivity of all 6 wavelengths to aerosol when the scattering angle is small. In the NH, the OMPS LP measurement vector for all wavelengths is positive for all altitudes, and the aerosol retrieval quality does not vary significantly with wavelength. Notice that all retrieved aerosol wavelengths can detect the cloud near the tropopauselayer evident as enhanced extinction near 10.5 km.

SAGE III/ISS
The SAGE series of instruments started with Stratospheric Aerosol Monitor (SAM) in 1975, (SAM II) in 1978(McCormick et al., 1982, SAGE I in 1979, SAGE II (Chu et al., 1993 in 1984, and SAGE III Meteor 3M (M3M) (Thomason and Taha, 2003) in 2001, spanning over 26 years. On February 19, 2017, SAGE III was launched to the International Space Station (ISS) to resume the SAGE series of measurements and provide. It provides high-resolution vertical profiles of aerosol extinction at multiple 230 wavelengths, the molecular densities of ozone, nitrogen dioxide, and water vapor, as well as profiles of temperature, pressure, and cloud presence. The aerosol extinction is computed as a residual after accounting for Rayleigh scattering and gaseous absorption, and thus, itthe retrieval makes no prior assumptions of the aerosol size or phase function. However, the technique is limited in coverage and number of profiles, to typically about 30 per day. The SAGE III/ISS retrieval algorithm is essentially the same its predecessor on the Meteor 3M platform. The quality of the SAGE III on Meteor 3M aerosol data was evaluated by Thomason and 235 Taha, (2003); Thomason et al. (2007b); and Thomason et al. (2010). These studies found that the aerosol extinction measurements accuracy and precision are on the order of 10% between 15 to 25 km, with the exception of 601 and 675 nm above 20 km, which exhibit substantial bias that was caused by the ozone clearing. A recent study by Wang et al. (2020) about SAGE III/ISS ozone validation also stated that an error in ozone correction caused an under estimation of the aerosol retrievals at wavelengths near the Chappuis band at altitudes where the aerosol loading is minimal. Thomason et al., (2020) also reported a defect in these 240 wavelengths below 20 km due to an error caused by the oxygen dimer (O4) cross section used in V5.1.

OSIRIS
OSIRIS (Optical Spectrograph and InfraRed Imager System)OSIRIS is an instrument that measures vertical profiles of limb scattered sunlight from the upper troposphere into the lower mesosphere. It was launched on February 2001 onboard the Odin satellite and continues to take measurements to the present. The instrument measures ozone, aerosol and NO2 profiles. Initial 245 (V5.07) aerosol retrievals were obtained by combining measurements at 470 and 750 nm, and were reported as aerosol extinction profiles at 750 nm. Rieger et al., (2014) compared coincident aerosol extinction observations by interpolating the SAGE II 525nm and 1020nm channels to the OSIRIS extinction wavelength of 750 nm. They found mean differences of less than 10% in the tropics to mid-latitudes, with larger biases at higher latitudes and at altitudes outside the main aerosol layer.
More recently, the V7 OSIRIS retrieval was introduced, which combines information from measurements at 470, 675, 750 and 250 805 nm to produce multi-wavelength aerosol extinction retrievals. The expanded wavelength usage reduces biases caused by measurement geometry, and improves the retrieval coverage and quality in the upper troposphere and lower stratosphere (UT/LS) region. The V7 algorithm also uses a modified version of the Chen et al. (2016) cloud detection algorithm (for PSC detection and general cloud screening).  report agreement at the 10% level between SAGE II and the Version 7 OSIRIS retrieval, with exceptions at high altitudes, which exhibit low bias due to sensitivity to stray light and nonzero aerosol in OSIRIS 255 normalization altitudes. Overall, the V7 product agreement with coincident SAGE data is comparable to the V5.07 performance, while the agreement with the CALIPSO-GOCCP product (Chepfer et al., 2010) is improved relative to V5.07. However, Kovilakam et al. (2020) noted that OSIRIS extinction is consistently higher than SAGE II in the lower stratosphere with difference exceeding 30% near the tropopause when comparing monthly means.

CALIPSO 260
The spaceborne lidar on CALIPSO which was launched in April 2006, provides global measurements of vertically resolved aerosol and cloud attenuated backscatter coefficients at 532 and 1064 nm . Significant improvement in calibration in V4 of CALIPSO data products makes it possible to retrieveobtain extinction coefficients in the stratosphere even with limited signal-to-noise ratio. The V1 Level 3 CALIPSO stratospheric aerosol profile product was produced using only the nighttime measurements and substantial spatial (vertical averaging to 900 m, 5 o latitude bins, 20 o longitude bins) and temporal (monthly) 265 averaging were applied. A constant lidar ratio (extinction to backscatter ratio) of 50 sr was used to retrieveobtain the extinction profiles., which is a typical value used for stratospheric aerosol background conditions (Kremser et al., 2016). The extinction profiles were retrieved using two different methods. In the "background" mode, all detected cloud and aerosol layers were removed and thin cirrus clouds within a few kms above the tropopause were filtered using a threshold on the volume depolarization ratio. (Kar et al., 2019). In the "all aerosol mode", all layers detected as aerosols in the stratosphere were retained and thin cirrus were 270 filtered using a threshold on the attenuated color ratio. In this work, we use the gridded extinction profiles from the "all aerosol" mode for consistently comparing with OMPS. It should be noted that in this mode, the cirrus cloud removal is not as efficient as in the "background" mode.
Initial validation of V1.0 CALIPSO L3 532 nm stratospheric aerosol profiles is described by Kar et al. (2019). This study concluded that CALIPSO agrees well with SAGE III/ISS aerosol, with CALIPSO about 25% higher between 20-30km30 km in tropics, and 275 larger differences at. However, the difference with SAGE III/ISS at the middle to high latitudes and low altitudes. was substantially larger, often exceeding 100%.

Data Comparison Methodology
In order to evaluate the accuracy of OMPS LP aerosol V2.0 retrievals, we have used a variety of methods. This includes comparison with the space-based instruments SAGE III/ISS, OSIRIS and CALIPSO, as well as performing internal consistency tests, which 280 can quantify the uncertainty of the aerosol model assumptions and the diffuse upwelling radiance effect. To provide detailed assessment of OMPS performance at different altitudes, latitudes, and time, we use two different approaches; coincident observations comparison, and zonal mean climatology comparison. While the first approach is used to eliminate any geographical and time biases, the latter is proved to be useful for monitoring the health and stability of the instrument and retrieval algorithm under different conditions and periods. However, zonally averaged comparisons can produce large biases following large volcanic 285 eruptions, where the aerosol load is high and spatially inhomogeneous, and therefore coincident comparison is preferred under these conditions . The percent difference is defined as Wherewhere reference is the correlative measurement of aerosol extinction. All correlative aerosol profiles were interpolated to 1 km vertical intervals, matching OMPS LP reported altitudes. Zonal mean climatologies were constructed using monthly mean profiles within 5 o or 10 o latitude bins. 290 For all comparisons shown in this paper, the center slit aerosol retrieval is used, since it has the most accurate radiometric calibration and stray light corrections (Jaross et al., 2014). The OMPS LP algorithm identifies cloud top height using the cloud detection method described in Chen et al. (2016). However, this algorithm also flags aerosols from fresh volcanic eruptions or PyroCbs. OMPS LP V2.0 data files now contain both cloud filtered and unfiltered data, as well as separate fields containing cloud height and type. Cloud type classifies the identified cloud as cloud, enhanced aerosol, or PSC. The "enhanced aerosol" definition 295 requires the cloud altitude to be at least 1.5 km above the tropopause. The "PSC" definition requires the cloud altitude to be at least 4 km above the tropopause, and the ancillary temperature at the cloud altitude to be less than 200 K. Users may wish to use both cloud height and cloud type flags to filter the data based on their own needs. To avoid removing aerosols from fresh volcanic or PyroCb plumes, we filtered the data by removing the extinction coefficient at and below cloud top height only if the reported cloud top height is in the troposphere. SAGE III is filtered for cloud contamination by using wavelength coloronly data with extinction 300 ratio at 510 nm / 1022 nm greater than 2 , while OSIRIS and CALIPSO provide cloud screened data.

Algorithm internal consistency
So as to estimate the uncertainty of the assumed aerosol size model (ASD)distribution and phase function, we compare 305 measurements taken at similar location but with different viewing geometry. Such measurements take place at high latitudes during the summer of both hemispheres, when the OMPS orbit allows observations of a given latitude in both the ascending and descending nodes. The ascending and descending nodes provide two daily observations of the same latitude, but with different scattering angles. The main assumption is that, if the retrieved aerosol values are different when the instrument is measuring the same air mass but with different scattering angle, then there is an error in the assumed phase function and ASD model. As shown 310 by , the ASD errors can introduce seasonal variations that correlate well with the SSA. Herein, we compare the daily zonal mean aerosol climatology between ascending and descending nodes in the Northern Hemisphere, where the aerosol signal is stronger. little, if any sensitivity to the aerosol model errors. At 16.5 km, the dependency of the aerosol retrieval on the scattering angle shows a linear trend of ~0.25% per degree for the 745 and 869 nm, and 0.5% per degree for the 675 nm. The trend is almost doubled to negative -1% per degree at 25.5 km, although it's distorted by sensor noise and inhomogeneity of the aerosol loading above the Junge layer in theat northern hemisphere high latitudelatitudes, especially when events occur inside the polar vortex 320 where the aerosol extinction is very low (Thomason and Poole, 1993). Nevertheless, the increase of the difference per unit of difference in SSA suggests that the aerosol model used in the retrieval is less representative of the aerosol measured at this altitude.  have shown that the OSIRS V7.0 aerosol extinction SSA dependence is 0.5% per degree.

The Diffuse Upwelling Radiance (DUR) uncertainties 325
As described in Sect. 2.2.1, the aerosol retrieval algorithm uses a simple Lambertian model of the reflecting surface to estimate an effective scene reflectivity (R). It doesn't mean the Earth's surface reflectivity, since the scene can contain clouds or aerosols.
Although the sensitivity of the aerosol retrieval to Diffuse Upwelling Radiance (DUR) uncertainties is reduced significantly by using normalized radiances (Flittner et al., 2000;Loughman et al., 2018), the error associated with assuming the Lambertian surface is difficult to estimate, and possibly not negligible. In order to quantify DUR uncertainties, we compare OMPS LP daily zonal 10 mean climatology for R values less than 0.3 (cloud free) and greater than 0.3 (bright or cloudy). Figure 4Figure 4 is a plot of the percent difference between the two aerosol climatologies at three different wavelengths. The three figures show a very similar picture; very large differences below the tropopause in the tropics, where the cirrus clouds are more frequent, and 5% positive bias above 20 km, just over that cloudy region. The 5% bias may be caused by scattered light originating from cirrus clouds near the tropopause, which wasn't properly accounted for in the radiative transfer model (which simply used a bright Lambertian surface 335 at sea level). The bias is negligible away from the cloudy regions, except for the SH lower altitudes (745 and 869 nm) and the NH (997 nm), which may be caused by a variety of reasons not related to cloud presence, since the large reflectivity outside the tropics is not necessarily an indication of cloud presence. (Larger R values are generally inferred from OMPS LP data at higher latitudes in both hemispheres.)Outside the tropics, the mean value of R is generally greater than 0.3, with strong seasonal dependence that peaks in the winter. Therefore, any observed differences outside the tropics are uncorrelated with cloud presence. 340

Coincidence comparison
To evaluate the quality of OMPS LP aerosol retrieval with SAGE III, we use a coincidence criteria of same calendar day measurements, Dlat. = ± 3 o and Dlon. = ± 10 o , which is selected to minimize the effect of spatial and temporal differences between the two instruments.  Figure 6Figure 6 is a summary plot of the mean difference between OMPS and SAGE III coincidences for wavelengths 510, 600, 675, 745, 869, and 997 nm. In general, wavelength 869 nm is the best OMPS retrieved wavelength comparedrelative to SAGE III 355 with differences of 5% or less 5% for most altitudes and latitudes, while the other. Other wavelengths agree with SAGE III to within 10%. Exceptions to this occur at high altitudes (above ~28 km) where the aerosol loading is minimal, and near the tropopause, causedwhich is affected by cloud contamination. The 510 and 600 nm OMPS extinction values have a slightly larger bias of 20% in the tropics. In contrastThis is due to other wavelengths observed differences, the ozone interference in both OMPS and SAGE III 600 nm aerosol retrievals. The 997 nm OMPS extinction values have a systematic bias of -10% between 60 o S and 360 20 o N, which might be affectedcaused by stray light contamination in the OMPS measurements. Unlike the other wavelengths, the 997 nm laboratory characterization is poor, and its stray light correction, therefore, has lower quality (Jaross et al., 2014). In the SH, wavelengths 510 nm shows largera large positive bias relative to SAGE III below 18km, caused by its lack of sensitivity toward aerosol at large scattering angle.18 km. This is an artifact in the OMPS retrieval algorithm, which often results in noisy and large extinction values when the measurement vector is too small (see Figure 12). 365 It is worth noting that the best agreement between OMPS and SAGE are found in the NH, where OMPS is observing in forward scattering and the weighting function is strong for all wavelengths. In that region, the agreement is mostly within 5% for an altitude between 14 and 22 km. Above 24 km, the observed biases for 510, 600, 675, and to some extent, 745, 869, and 997 nm, gradually increase with altitude, mainly caused by instrument noise and errors under low aerosol conditions, although OMPS assumed aerosol size model uncertainty also contributes to the larger differences. 370 Figure 7Figure 7 summarizes the quality of OMPS LP aerosol extinction at 6 retrieved wavelengths, showing the zonally averaged mean differences between OMPS LP and SAGE III aerosol (in percent) at 510, 600,675,745,869, and 997 nm. The comparison shows that below). Below 25 km, the differences between OMPS LP and SAGE III are largely driven by OMPS weighing functions or Jacobians. The weaker Jacobians for short wavelengths under backscatter conditions in the SH and below 20 km leads to limited accuracy, while stronger Jacobians at the longer wavelengths improve its accuracy significantly (Taha et al., 2011;Rieger et al., 375 2019). Overall, the shorter wavelengths (510, 600, and 675 nm) are biased low against SAGE III with a difference greater than 25% below 20 km in the SH. In addition, these short wavelengths exhibit pronounced large aerosol in the tropics below 20 km caused by the algorithm's reduced accuracy when the measurement vector is very small. The agreement is well within 25% at altitude range 20 -25 km and better in the NH. Above 25 km, the comparison between the two instruments is poor, caused by either SAGE III ozone correction errors and/or OMPS reduced sensitivity ofto aerosol at these short wavelengths. The best 380 agreement between OMPS and SAGE III can be seen at 869 and 997 nm, where they are mostly within 10% of each other for all altitudes and latitudes. The 745 nm OMPS extinction agrees with SAGE to within 15% everywhere except for the SH tropics below 18km.
The standard deviation shown in Figure 8Figure 8 is influenced by several factors: OMPS LP uncertainties such as measurement noise, forward model errors, and retrieval algorithm sensitivities, in addition to SAGE III/ISS own uncertainty and atmospheric 385 variability. In general, the standard deviation is 15% -20% for altitudes that show good agreement with SAGE III (Figure 7Figure 7). The large standard deviation of ~50% at high altitude is due to instrument noise and low aerosol loading. Below 20 km, the standard deviations for the shorter wavelengths increase to 50%, caused by the OMPS LP reduced accuracy. In the UT/LS, the standard deviation is >50% due to larger dynamical variability, especially during periods when dispersal of plumes due to volcanic eruptions and other events cause longitudinal variations, as well as cloud interference. 390 Based on SAGE IIIthe comparison with SAGE III, we can estimate the OMPS aerosol retrieval relative accuracy to be ~10% for 745, 869, and 997 nm in the stratosphere, and 20% for the 675 nm above 20 km and in the NH. The 510 and 600 nm retrievals have limited accuracy in the SH and 25% relative accuracy at altitudes between 20 -26 km and in the NH. Furthermore, the standard deviation can be used to determine the retrieval relative precision, which can be estimated to be better than 15% for the longer wavelengths, and close to 20% for wavelengths less than 745 nm. The real precision is probably better than the quoted 395 values, since the calculated standard deviation includes atmospheric variability and both instrumentsinstruments' biases, none of which was removed (Rault and Taha, 2007;Wang et al., 2020).

Zonal mean comparison
In order to investigate the OMPS LP retrieval performance under different seasonal or geographical conditions, we compare the OMPS LP monthly zonal mean time series with the SAGE III/ISS monthly zonal mean time series for 4 wavelengths at 3 different 400 altitudes. The comparison is also divided into 3 different regions, SH (Figure 9Figure 9), tropics (Figure 10Figure 10), and NH (Figure 11Figure 11). In general, the agreement between the two instruments in the SH is mostly within 10-20%. The 675 nm extinction at 20.5 km is a notable exception, as the OMPS LP aerosol extinction values drop significantly when the SSA is greater than 145 o and the attenuation of Rayleigh scattering below 20 km becomes significant. This behavior appears as an apparent seasonal pattern, in which the OMPS LP / SAGE III difference becomes much larger during SH winter months. It is therefore 405 recommended that OMPS aerosol measurements at λ ≤ 675 nm should be excluded when SSA is greater than 145 o below 21 km.
SimilarA similar agreement is found in the tropics, at or above 20.5 km, with the exception of the first few months following Aoba volcanic eruption in July 2018, where OMPS LP initially measured morereported higher aerosol extinction than SAGE III. This might be caused by the different coverage and frequency of measurements for each instrument, where the monthly zonal mean is heavily skewed by few daily measurements in the case of SAGE III.. Still, the difference between the two measurements was 410 mostly within 20% in the aftermath of this volcanic eruption, due in part to OMPS use of fixed background aerosol size distribution model.. At 18.5 km, the difference is often greater than 20%, reaching more than 60% following the subsidence of the volcanic plume. The reason for such large differences is unclear as OMPS LP still shows elevated aerosol levels when SAGE III measurements indicate that the aerosol values are back to pre-eruption levels, although spatial variability and spatial resolutions can contribute to such large differences. SAGE III aerosol extinction profiles are produced on a 0.5 km grid with an estimated 415 vertical resolution is 0.7 km (Thomason et al., 2010;SAGE III ATBD, 2002)  They showed that both instruments have very similar layered vertical structure and magnitude. However, they noted that some differences in layer height and magnitude can be expected from differing vertical resolutions. Another possible explanation 420 is that OMPS LP cloud clearing can be incomplete and residual cloud contamination can contribute to the large differences near the tropopause. The best agreement between the two instruments can be found at 25.5 km, well within 10%.
In the NH, all OMPS LP wavelengths show similar robust agreement to SAGE III, mostly within 10%, since OMPS LP observes in the forward scattering and all wavelengths are strongly sensitive to aerosol. (Figure 11). A notable exception is the first couple of months of the August 2017 Canadian PyroCb period and the June 2019 Raikoke eruption, when the aerosol loading was very 425 high and spatially inhomogeneous. Spatial inhomogeneity also caused thea large bias after 2018 following the sharp drop in aerosol extinction at 25.5 km. While OMPS assumed ASD model may contribute to the larger differences at 25.5 km, instruments noise and calibration errors are also more significant under low aerosol conditions. OMPS LP 997 nm is affected by stray light contamination at the normalization altitudes in the NH high latitudes, which might explain the negative bias during 2019. On the other hand, SAGE III ozone correctionscorrection uncertainty near the Chappuis band can cause a dip in SAGE III aerosol 430 extinction measurements at 676 nm. In particular, the SAGE III 676 nm values at 25.5 km are either zero or negative during 2019 when the measured aerosol is at its lowest levels in the NH during the short lifetime of ISS SAGE III.

Comparison OMPS LP with OSIRIS and CALIPSO
In this section, we compare OMPS LP aerosol at 510 and 745 nm with V7 OSIRIS at 750 nm, and V1 L3 CALIPSO at both 532 435 and 745 nm. CALIPSO 532 nm extinction ("all aerosol mode") is converted to 745 nm using an Angstrom exponent of 1.9, similar to the Angstrom exponent for the OMPS assumed aerosol model. We also included SAGE III/ISS measurements at 755 nm as an independent reference, since SAGE measurements are widely considered as the most accurate stratospheric aerosol dataset (SPARC, 2006;von Savigny et al., 2015;Kremser et al., 2016;Thomason et al., 2018;Kar et al., 2019). Although coverage and sampling differences can make such comparisons difficult, it provides a chance to evaluate the entire OMPS LP data record relative 440 to these two datasets. As both OSIRIS and CALIPSO approach the ends of their lives, it is now more critical than ever to extend the stratospheric aerosol record that has been developed from SAGE/OSIRIS/CALIPSO into OMPS LP/SAGE III/ISS records.  (Vernier et al., 2011b). Notice that agreement between 450 OMPS LP and the 3 instruments is generally within 20% for all shown altitudes, except for the 25.5 km in SH, where CALIPSO is somewhat biased high, and in the tropics at or below 20.5 km, where the aerosol loading is greatly enhanced by several moderate volcanic eruptions. Part of this large difference can be due to aerosol model uncertainties, as both OMPS LP and OSIRIS assume fixed background aerosol model, while CALIPSO uses fixed lidar ratio. In addition, differences at 18.5 km in the tropics can be affected by residual cloud contamination, as all three instruments use different criteria for screening cloudy events. Kar et al., 455 (2019) reported that CALIPSO havehas larger biases 2-3 km above the tropopause that might be due to cloud contamination. The OMPS LP 510 nm comparison shows good agreement with CALIPSO at 20.5 km in the SH, with periods of larger difference when the SSA is greater than 120 o in the SH. At 18.5 km, the difference is also 20% with OMPS exhibiting periodic jumps in the aerosol extinction values at 510 nm, caused by the algorithm's reduced accuracy when the magnitude of the measurement vector is very small (see Figure 1Figure 1). At 25.5 km, both the 510 and 745 nm extinction values show similar variability to CALIPSO, well 460 within 25%, except for the SH. In the NH, the accuracy of the 510 nm aerosol retrieval is comparable to the 745 nm accuracy.
OSIRIS monthly means in the NH are slightly noisier because of the limited number of profiles used. Figure 15Figure 15(a, b). The differences between OMPS LP and CALIPSO are, in general, within 25% between 50 o S and 50 o N, except for the tropics, which isare closer to 25% at some altitudes and greater in case of 510 nm. The difference is substantially large in mid to high latitudes of both hemispheres, 465 with CALIPSO showing a rather large bias of 50%, and exceeding 100% at altitudes above 25 km., This is also consistent with an earlier study that shows the agreement of 25% between CALIPSO and SAGE III/ISS between 30 o S and 30 o N, and larger biases of 100 -200% at the middle to high latitudes (Kar et al., 2019). They also noted that the primary parameter affecting the comparison is likely the fixed lidar ratio of 50 sr used in the CALIPSO retrieval, which is dependent on the aerosol optical and physical properties. It is plausible that the comparison between OMPS and CALIPSO can be further improved by using a different lidar 470 ratio.

A summary of the comparison between OMPS LP and CALIPSO is shown in
Agreement between OMPS LP and OSIRIS (Figure 15Figure 15c) is generally very good, with differences less than 20% for most latitudes in the stratosphere below 30 km, except for the tropics, where OSIRIS is 30% low compared to OMPS. The reason for the increased differences in the tropics is unclear, however, a similar negative bias of 15% was also noted for comparisons in the tropics between OSIRIS and SAGE III/ISS , while OMPS 745 nm has 10% positive bias in the tropics relative 475 to SAGE III/ISS (Figure 5Figure 5). Combining both biases might explain the difference seen in the tropics. The large bias seen at NH high latitude above 25 km is consistent with the comparison with SAGE III and CALIPSO and highlighthighlights the difficulties of retrieving very low aerosol for both instruments. Rieger et al. (2018) also reported a large negative bias at high latitude above 25 km when comparing OSIRIS to SAGE II that is caused by non-zero aerosol in the normalization altitudes.
Similar to the previous comparison with SAGE III (Figure 7Figure 7), the standard deviation for the three comparisons is generally 480 less than 15% for altitudes that shownshow good agreement with either CALIPSO or OSIRIS (Figure 15Figure 15d, e, and f).
There is a large standard deviation of ~50% above 22 km at the SH high latitude due to OMPS reduced accuracy under low aerosol loading and large scattering angle. In the UT/LS, the standard deviation is greater than 40% due to larger dynamical variability, especially during volcanic eruptions, and cloud interference.

Recommendations for use of OMPS LP aerosol extinction data 485
OMPS LP provides good quality multi-wavelength aerosol extinction retrievals that can be useful for scientific studies. In particular, we find the relative accuracy at 745, 869, and 997 nm is on the order of 10% and relative precision better than 15% in the primary aerosol layer in the stratosphere. These retrievals are suitable for continuing the long-term record of stratospheric aerosol that was started by SAGE II in 1984, although the OMPS 997 nm retrieval is affected by stray light contamination and shows slight negative bias in the SH. Since the 869 and 997 nm retrievals have the strongest sensitivity to aerosol in almost all 490 altitudes and regions, these are most suitable for scientific studies like detection and tracking periodic events when aerosol particles are injected into the stratosphere, such as volcanic eruptions or PyroCbs. The 675 nm relative accuracy is on the order of 20% above 20 km, and its relative precision is 15%. In the SH, its accuracy is reduced, with measurements deemed unusable when the SSA exceeds 145 o . We find the 600 nm of comparable accuracy to 675 nm at altitudes 20-25km25 km which suggests that the applied ozone corrections at this wavelength are reasonable. Still, we recommend the user avoid using this channel since it is only 495 meant for diagnostic purposes. The 510 nm relative accuracy is 25% at a limited altitude ragerange of 20-24 km in the SH and tropics, with similar accuracy in the NH for all altitudes below 25 km. Because of its weak sensitivity to aerosol in the backscatter, we recommend cautious use of the 510 nm retrieval, only in the NH. We also recommend that the user be cautious when attempting to derive aerosol size information from quantities such as Angstrom exponent, since the accuracy of each wavelength retrieval is affected by its weighing function at some altitudes and latitudes, and the 997 nm retrieval is affected by stray light contamination, 500 which may bias the result.

Conclusions
The new V2.0 OMPS LP aerosol extinction products at wavelengths 510, 600,675,745,869, and 997 nm have been processed.
Comparisons with coincident measurements by the SAGE III/ISS, OSIRIS, and CALIPSO instruments indicate that the OMPS LP retrievals are suitable for scientific studies. By comparing OMPS measurements at different scattering angles, we demonstrate that 505 the retrieval's dependency on viewing geometry is negligible at 20.5 km and reduced significantly for the longer wavelengths at lower altitudes relative to the short wavelengths. In addition, we estimate the uncertainty in the aerosol retrieval caused by diffuse upwelling radiance (DUR) to be in the order of 5%. The 745, 869, and 997 nm extinction profiles in the stratosphere are shown to be the most accurate and most suitable for continuing the long-term record of stratospheric aerosol, with relative accuracies and precisions close to 10% and 15% respectively, while the relative accuracy and precision of 675 nm extinction profiles are on the 510 order of 20%. Differences can be larger for individual profiles or zonal mean comparison, which can be affected by differences in instrument's coverage and inhomogeneity along the line of sight for fresh volcanic eruptions. The 510 nm extinction profile was shown to have limited accuracy in SH and is only recommended for use between 20-24 km and in the NH only. The 600 nm extinction profile is mainly retrieved for diagnostic purposes and is not recommended for scientific use. Future versions of the OMPS LP retrieval algorithm may improve on the assumed aerosol size model to account for the different types of aerosol at 515 different altitudes. Additionally, a better cloud clearing whichthat can utilize OMPS multiple wavelength dependence may further improve the aerosol products in the UT/LS region.

Author contributions. 525
GT and RL were responsible for the development of the OMPS LP V2.0 multi-wavelength algorithm, which is described in this paper. TZ was responsible for code improvements and testing. GT wrote the initial draft of the paper with help from RL. GT carried out the analysis shown here. LT participated in the scientific discussion about SAGE III data. JK participated in the scientific discussion about CALIPSO data. LR and AB participated in the scientific discussion in regard to OSIRIS. ALL authors reviewed the manuscript and provided with advice on the text and figures. 530

Competing interests.
The authors declare that they have no conflict of interest.         The 1/3 value is an error, and has been corrected to 0.2 (same as the value used in the V1 algorithm). 770 L201: This is my first major criticism: stopping the retrieval based on convergence at one height is a mistake. This will mean poor low accuracy particularly below this tangent height but even at altitudes at/above this tangent height.
We understand the reviewer's concern and we are already planning on modifying the algorithm to check for 775 convergence for all altitudes of the main aerosol layer, which will be released next year as Version 2.1. However, our analysis indicates that this change will have a limited impact on the retrieved aerosol profiles, mostly affecting the shorter wavelengths (675 nm or less), and only in the tropics in the aftermath of large aerosol enhancements such as volcanic eruptions, when it converges after 3 iterations. We have revised the sentence and it reads as "Iterations end when the retrieved aerosol extinction changes by < 2% at 20 km or when it reaches maximum number of iterations. 780 The planned V2.1 release next year will use modified convergence criteria that checks for multiple altitudes."

L227: "it" is not defined
We replaced "it" with "the retrieval" 785 L235: "by" -> "caused by" The cloud screening of all instruments is addressed in section 4. We have modified the text in L299 to add more details to the cloud clearing process. The text now reads "SAGE III is filtered for cloud contamination by using only data with 795 extinction ratio at 510 nm / 1022 nm greater than 2 " L258: Regarding "retrieve extinction", CALIPSO does not really retrieve extinction.
We replaced "retrieve" with "obtain" Done L288: What is a "cloud type flag" and how is this determined?

815
We have added the following text to explain the cloud types and how it is determined: "Cloud type classifies the identified cloud as cloud, enhanced aerosol, or PSC. The "enhanced aerosol" definition requires the cloud altitude to be at least 1.5 km above the tropopause. The "PSC" definition requires the cloud altitude to be at least 4 km above the tropopause, and the ancillary temperature at the cloud altitude to be less than 200 K." Wavelength dependence of this error is generally expected since the different wavelengths have different sensitivity to 840 aerosol particle size (Reiger et al., 2014).
As we explained in the text (L304 -L306), because of the spacecraft orbit, these measurements only take place at high latitudes during the summer of both hemispheres. OMPS LP observations at high latitudes during the winter are mostly in the dark.
We are confident that the convergence criteria have no effect on the analysis shown in Figure 3, since these analyses 845 were made in the NH for a period not affected by any volcanic perturbations (see our comment above). The aerosol differences seen at 16 and 25 km are largely driven the uncertainty in the assumed aerosol model. Lack of SSA dependence at 20 km means that the a-priori aerosol model used in the retrieval is more representative of the measured aerosol at 20 km. Similar pattern, albeit with larger difference, was found in V1.0 retrieval algorithm that used bimodal aerosol size model and different convergence criteria (see section 2.2.1). The retrieved aerosol extinction 850 dependencies on the SSA were subsequently reduced in V1.5, which indicates that the gamma distribution ASD used in V1.5 is more accurate than the bi-modal ASD used in V1.0 (see figure 1 below). Similar pattern was also seen by .
We have also added the following text "The main assumption is that, if the retrieved aerosol values are different when the instrument is measuring the same air mass but with different scattering angle, then there is an error in the assumed 855 phase function and ASD model. As shown by , the ASD errors can introduce seasonal variations that correlate well with the SSA." We also added "Similar analyses made by  have shown that the OSIRS V7.0 aerosol extinction SSA dependence is 0.5% per degree."

865
L316: Since at L290, the authors inform us that clouds are being removed, any differences in aerosol extinction between low and high R are therefore not expected differences in extinction due to cirrus, but rather point to cirrus being missed by the cloud flagging. I suppose this is difficult to avoid in the tropical tropopause region.
That is correct. It is either cloud being missed or incomplete cloud clearing. 870 L322: Do the authors believe that the larger R (effective scene reflectivity) at higher latitudes is real or an artifact of the retrieval? If it is real, R should have a seasonal dependence, being higher in winter when there is snow covering the land at northern high latitudes.

875
That is correct, we do see seasonal dependence of R being higher in the winter, and lower during the summer, although it is nowhere as low in the tropics. We have revised the sentence and it now reads "Outside the tropics, R mean value is generally greater than 0.3, with strong seasonal dependence that peaks in the winter. Therefore, any observed differences outside the tropics are uncorrelated with cloud presence."

895
We agree with the reviewer that this sentence is not accurate. The accuracy of the 510 nm is discussed in more details in section 5.2, figure 12, which shows OMPS 510 nm exhibiting periodic jumps in the aerosol extinction value. This is caused by the algorithm's reduced accuracy when the measurement vector is very small. We have now replaced this sentence with "This is an artifact in OMPS retrieval algorithm, which often results in noisy and large extinction values when the measurement vector is too small (see Figure 12)." 900 Figure 6: It seems a bit odd that the aerosol extinction bias relative to SAGE is higher at 600 nm than 510 nm for low latitudes at/below 18 km?
This is most likely caused by the ozone contamination for both OMPS and SAGE retrievals at this wavelength. We 905 have added the following text "This is due to the ozone interreference in both OMPS and SAGE III 600 nm aerosol retrievals." We disagree with the reviewer. While this is true for the shorter wavelengths (675 or less), the longer wavelengths have strong sensitivity to aerosol, even in the SH (see section 2.2.2). Figures 5,7,and 9, also show the longer wavelengths agreement with SAGE III is mostly within 10% for most altitudes.

930
We deleted the sentence. We are glad that the reviewer shares our assessment of the high quality of the V2.0 OMPS LP aerosol. Figure 9 clearly 935 support the argument made above, that the V2.0 longer wavelengths are of good quality in the SH.

L405: Regarding "corrections", does this need to be plural?
We changed it to correction.

24)line 300: No need to redefine SSA
We deleted the SSA definition. 1125 Fig. 3 demonstrates that the algorithm is insensitive to errors in the assumed ASD. You need to justify this statement with additional detail.

25)lines 302-303: It is not clear how the results in
We never made the claim that the algorithm is insensitive to the errors of the assumed aerosol model. The main 1130 assumption here is that, if the retrieved aerosol values are different when the instrument is measuring the same air mass but with different scattering angle, then there is an error in the assumed phase function and ASD model. As shown by , the ASD errors can introduce seasonal variations that correlates well with the SSA. Figure 3 shows little dependency on the SSA at 20 km, and somehow larger dependency at 16 and 25 km. Those results are similar or better than the 0.5% per degree reported by . 1135 We have modified the text to include the following sentences "The main assumption is that, if the retrieved aerosol values are different when the instrument is measuring the same air mass but with different scattering angle, then there is an error in the assumed phase function and ASD model. As shown by , the ASD errors can introduce seasonal variations that correlates well with the SSA." And "Similar analyses made by  has shown that the OSIRS V7.0 aerosol extinction SSA dependence is 0.5% per degree." 1140 26)line 310: By reflecting surface, do you mean Earth surface?
Not necessary, it can also be clouds or aerosols. We have now added the following sentence "It doesn't mean the Earth's surface reflectivity, since the scene can contain clouds or aerosols." 1145 27)lines 319-320: this sentence should be clarified.
We have revised the sentence and it now reads "Outside the tropics, R mean value is generally greater than 0.3, with strong seasonal dependence that peaks in the winter. Therefore, any observed differences outside the tropics are 1150 uncorrelated with cloud presence." 28)line 321: Remove the parentheses from this sentence.

Done.
30)line 337and elsewhere): The preferred syntax would be "869 nm wavelength" We deleted "wavelength" as suggested by the reviewer 1.

31)line 336: This paragraph is a bit clumsy overall. 1165
We have revised the paragraph and it now reads as " Figure 6 is a summary plot of the mean difference between OMPS and SAGE III coincidences for wavelengths 510, 600, 675, 745, 869, and 997 nm. In general, 869 nm is the best OMPS retrieved wavelength relative to SAGE III with differences of 5% or less for most altitudes and latitudes. Other wavelengths agree with SAGE III to within 10%. 1170 Exceptions to this occur at high altitudes (above ~28 km) where the aerosol loading is minimal, and near the tropopause, which is affected by cloud contamination. The 510 and 600 nm OMPS extinction values have a slightly larger bias of 20% in the tropics. This is due to the ozone interference in both OMPS and SAGE III 600 nm aerosol retrievals. The 997 nm OMPS extinction values have systematic bias of -10% between 60 o S and 20 o N, caused by stray light contamination in the OMPS measurements. Unlike the other wavelengths, the 997 nm laboratory characterization 1175 is poor, and its stray light correction therefore has lower quality (Jaross et al., 2014). In the SH, 510 nm shows large positive bias relative to SAGE III below 18 km. This is an artifact in OMPS retrieval algorithm, which often results in noisy and large extinction values when the measurement vector is too small, (see Figure 12)." 32)line 351: There is no need to list the wavelengths at the end of this sentence. We deleted "The comparison shows that". Figure 8: Please correct the label on the color bar, which should say 1 -the standard deviation of the difference (or 1 -sigma). 1190

Done.
35)line 369: This is one xample of a poorly formed sentence, which seem to be common in this section. "Based on SAGE III comparison, ..."should be "Based on the comparisons with SAGE III,..." We have changed the sentence to "where OMPS LP initially reported higher aerosol extinction than SAGE III." 37)line 387: What do you mean by "heavily skewed by few daily measurements..."? Please explain this effect.
We agree that this sentence is not clear, so we deleted it. The original text that reads "This might be caused by the 1205 different coverage and frequency of measurements for each instrument." Is sufficient enough to explain the differences.

1215
Vertical resolution differences were previously reported by various studies (Chen et al., 2020;. We have now added the following text "Bourassa et al. (2019)  Done.