Effects of polar stratospheric clouds in the Nimbus 7 LIMS Version 6 data set

The historic Limb Infrared Monitor of the Stratosphere (LIMS) measurements of 1978–1979 from the Nimbus 7 satellite were re-processed with Version 6 (V6) algorithms and archived in 2002. The V6 data set employs updated radiance registration methods, improved spectroscopic line parameters, and a common vertical resolution for all retrieved parameters. Retrieved profiles are spaced about every 1.6 of latitude along orbits and include the additional parameter of geopotential height. Profiles of O3 are sensitive to perturbations from emissions of polar stratospheric clouds (PSCs). This work presents results of implementing a firstorder screening for effects of PSCs using simple algorithms based on vertical gradients of the O3 mixing ratio. Their occurrences are compared with the co-located, retrieved temperatures and related to the temperature thresholds needed for saturation of H2O and/or HNO3 vapor onto PSC particles. Observed daily locations where the major PSC screening criteria are satisfied are validated against PSCs observed with the Stratospheric Aerosol Monitor (SAM) II experiment also on Nimbus 7. Remnants of emissions from PSCs are characterized for O3 and HNO3 following the screening. PSCs may also impart a warm bias in the co-located LIMS temperatures, but by no more than 1–2 K at the altitudes of where effects of PSCs are a maximum in the ozone; thus, no PSC screening was applied to the V6 temperatures. Minimum temperatures vary between 187 and 194 K and often occur 1 to 2 km above where PSC effects are first identified in the ozone (most often between about 21 and 28 hPa). Those temperature–pressure values are consistent with conditions for the existence of nitric acid trihydrate (NAT) mixtures and to a lesser extent of super-cooled ternary solution (STS) droplets. A local, temporary uptake of HNO3 vapor of order 1–3 ppbv is indicated during mid-January for the 550 K surface. Seven-month time series of the distributions of LIMS O3 and HNO3 are shown based on their gridded Level 3 data following the PSC screening. Zonal coefficients of both species are essentially free of effects from PSCs on the 550 K surface, based on their average values along PV contours and in terms of equivalent latitude. Remnants of PSCs are still present in O3 on the 450 K surface during midJanuary. It is judged that the LIMS Level 3 data are of good quality for analyzing the larger-scale, stratospheric chemistry and transport processes during the Arctic winter of 1978– 1979.

Retrieved profiles are spaced about every 1.6˚ of latitude along orbits and include the additional following the screening. PSCs may also impart a warm bias in the co-located LIMS 32 temperatures, but by no more than 1-2 K at the altitudes of where effects of PSCs are a 33 maximum in the ozone; thus, no PSC screening was applied to the V6 temperatures. Minimum It is now well known that heterogeneous chemical reactions on surfaces of polar stratospheric 49 clouds (PSCs) lead to an acceleration of the depletion of polar ozone in late winter and early 50 spring (e.g., Solomon et al., 2015). Chemistry/climate models are showing that changes in the 51 formation and persistence of PSC particles and of their related chemical effects are sensitive to 52 changes in polar stratospheric temperatures, as well as to trends in ozone depleting substances 53 (ODS). One current scientific need regarding stratospheric ozone is the "evaluation of the 54 Antarctic ozone hole and Arctic winter/spring ozone depletion and the predicted changes in these 55 phenomena, with a particular focus on temperatures in the polar stratosphere" (WMO, 2014). 56 Therefore, it is important to characterize the occurrence frequency of PSCs, to compare their 57 presence with the local temperature and chemical fields each winter/spring season, and to relate 58 their current frequencies and effects with those of past decades.

60
The present study is an analysis for the effects of PSCs as determined for the Arctic winter of 61 microphysical models for the formation of PSCs. It is shown that significant zonal wave-1 114 forcings bring about rapid exchanges of O3 and HNO3 between middle and polar latitudes in 115 early December and late January (e.g., Leovy et al., 1985). The mapped data and the time series 116 analyses are examined for instances of uptake or loss of HNO3 vapor as well as instances of the 117 advection of low values of O3 and HNO3 to the vortex from lower latitudes. Section 7 provides 118 evidence for a temporary uptake of HNO3 vapor onto PSC particles, when the temperature is less  instantaneous field-of-view (IFOV) vertical width of 3.6 km at the horizon, while the other four 138 channels have half that width or 1.8 km (see Gille and Russell (1984) and Remsberg et al. (1990) 139 for more details about the LIMS instrument and the domain of its atmospheric measurements). radiances about the best that one can do is to consider anomalies in profiles of the ratio of the 146 water vapor radiance to the CO2W radiance. But the IFOVs of the H2O and CO2W channels are 147 not compatible, which means that one cannot properly account for the corresponding radiances 148 from temperature itself. Effects of emissions from clouds are delineated best with the narrow-149 IFOV, O3 and HNO3 channels. Yet, cloud occurrences are more pronounced in the retrieved O3 150 than in HNO3 because the relation between mixing ratio and radiance is non-linear for LIMS O3.

151
Pressure-altitude locations of perturbations from cloud tops are included in a header line for 152 every V6 profile, so that effects of the clouds can be screened from them prior to their processing 153 with the LIMS Level 3 mapping algorithm (e.g., Remsberg et al., 1990).

155
Although the spectral effects of emissions from PSCs were not understood well in the 1970s, 156 there are obvious effects from them in the LIMS retrieved ozone (Fig. 1) and in H2O (not 7 their presence showed up clearly as "bull's-eye-like" features or as localized ozone maxima in 163 preliminary maps of the daily parameters on pressure surfaces. Those occurrences are reported 164 in Remsberg et al. (1986, their Tables 6 and 7 and Fig. 4). Then, the vertical segments that gave 165 rise to them were removed using conservative latitude/longitude templates of the affected regions 166 followed by a final mapping of the V5 profile data.

168
The same PSC templates were applied in the map analyses of the V5 temperatures, and the maps 169 were inspected for evidence of perturbing effects      The LIMS V6 Level 2 screening criteria are described as follows. The primary criterion is 256 denoted by a parameter labeled DIF in the daily files of ozone data points that were screened out.

257
The DIF threshold was evaluated and then finally set as an absolute mixing ratio change of 258 greater than 1.7 ppmv between two adjacent LIMS ozone profile points, spaced 0.88 km apart.

259
As an example, Figure 3 shows five successive V6 ozone profiles from 11 January along the  (1) 273 Specifically, the RTO threshold was met when there was a change in the ozone mixing ratio q 274 with decreasing altitude between two adjacent profile points, n-1 and n, (spaced 0.88 km apart) 275 such that f in Eq. (1) was greater than 0.7 whenever q(n-1) had a somewhat arbitrary value of 276 greater than 0.5 ppmv. Note that the point index n for a profile in the data file increases as 277 altitude decreases, due to the nature of the "top-down" retrieval algorithm. As with DIF, the 278 screening based on the RTO criterion begins four points above where the threshold is met. The

279
RTO threshold is complementary to DIF and accounts for occurrences of anomalous vertical 280 structure in ozone, where the ozone mixing ratio is small (but > 0.5 ppmv). Only two profiles 281 met the RTO threshold on 11 January (near 82˚N, 180˚E), and they occurred at the level of 88 282 hPa. The DIF threshold was also met for one of them at that level. where PSCs are located based on the ozone field, but also to be confident that there is little 334 impact on the co-located temperature fields and no impact on the geopotential height fields for 335 calculations of the associated species transport. It is presumed that the LIMS temperatures are 336 within their estimated accuracies at and above where the effects of PSCs are small in the co-337 located ozone. Tighter criteria can screen out more instances of PSC contamination from the 338 species but may also remove some profile segments having vertical structures due to real, 339 transport-induced effects. In particular, a few instances of an overly tight screening have been 340 found based on the RTO criterion at pressure-altitudes from 52.7 to 100 hPa, as indicated by co-341 located temperatures that are much too warm for saturation and maintenance of PSC particles.  January and perhaps a slight excess in the nitric acid field in the same region. Figure 5 shows the  ozone is required in the tangent layer to achieve a match between the calculated and observed 419 radiances for the retrieval of ozone. The several profiles that were screened out along 118 to 420 126˚E are "false positives" for PSCs; their ozone mixing ratios dropped below 0.2 ppmv in a 421 region of strong temperature gradients.  According to Figure 9 and    . 17a). In addition, Fig.   507 9 shows that the PSCs were found within a 10-km deep layer. The RH (ice) calculations in 508 Table 1 (Table 1). Figure 11 shows that HNO3 has a maximum of 12 ppbv over Siberia (120˚E) 597 on that date and smaller values (~9 ppbv) at the center of the vortex (0 to 60˚E and ~21.5 km) 598 and where the co-located ozone is only 3 ppmv. Although the relative minimum of HNO3 of 9 599 ppbv is equivalent to its values at the middle latitudes, the 3-ppmv ozone is an isolated, absolute  Figure 12 is the 4-panel plot for 17 January, and it characterizes the polar 611 stratosphere during a period of a large number of PSCs (see Fig. 11). In fact, the largest PSC indicates a region of de-nitrification by 2-3 ppbv of the air flowing across the PSC particles.

621
That apparent, temporary uptake of HNO3 is considered further in the next section. The points in Figure 13 are also colored by whether the maximum ozone (of the right Hovmöller  Table 1 indicate that the STS threshold was 661 achieved on 19 and 20 January. Pitts et al. (2013) also found that that an uptake of 2-3 ppbv of 662 HNO3 vapor onto NAT particles can occur, when the temperature remains below 193 K for more 663 than 2 to 3 days. Those conditions were met from 5 to 11 January 1979 (Table 1) only up to 6 zonal waves can be resolved from the data and that the daily character of the actual 709 zonal waves may be aliased slightly (Remsberg et al., 1990). Therefore, a 7-day smoother was 710 applied to the PV time series for Fig. 15 to minimize effects of that sampling bias.  The time series of the MLM for ozone is shown in Figure 17, and its isolines deviate appreciably

783
There was also transport of lower ozone values from φ = 70 to 85˚ for a week or two thereafter.

784
From late February and onward there is good correspondence between ozone and PV at middle 785 and high equivalent latitudes. Finally, while the meridional gradients for ozone are much weaker 786 than for HNO3 near φ = 65˚, their respective seasonal patterns remain similar to those of PV. (RTO) that may also be triggered due to normal, dynamically-induced structures in the profiles.

803
In fact, the RTO criterion was met more frequently at the lower levels of 52 to 100 hPa. During The retrieved V6 temperature profiles were not screened, even though the co-located temperature wide-band, 15-μm CO2 channel that is nearly insensitive to PSC/aerosol emissions, so that co-824 located temperature profiles are also available with little to no bias.

826
The LIMS V6 rather than V5 data are more compatible with stratospheric datasets from recent 827 satellite experiments. It is primarily for this reason that the V6 data have been included as part March it is clear that there was a significant accumulation of ozone at middle and higher 844 equivalent latitudes at 550 K. It is also concluded that the V6 data provide accurate, associated 845 temperature and GPH fields for conducting transport studies of that time period.