Interactive comment on “ Using airborne observations to improve estimates of short-lived halocarbon emissions during summer from Southern Ocean ”

This study of airborne observations of halogenated VOCs (HVOCs) represents a valuable addition to the knowledge of these compounds over the Southern Ocean, where few data exist. The study confirms the current view that the main sources of CHBr3 and CH2Br2 are biological, and that CH3I has both biological and non-biological sources. The authors have put forward a novel concept of using enrichment ratios of HVOCs to O2 to infer the contribution or otherwise of ocean biological sources, and propose a new function to estimate non-biological emission fluxes of CH3I. The dataset has been used

parameterizations of HVOC emissions in a global atmospheric chemistry transport model, assess 114 contributions from previously hypothesized regional sources for the Southern Ocean, and  Atmospheric measurements for this study were collected at high latitudes in the Southern 121 Hemisphere as part of the O 2 /N 2 Ratio and CO 2 Airborne Southern Ocean (ORCAS) study 122 (Stephens et al., 2018), and the second NASA Atmospheric Tomography Mission (ATom-2), 123 near Punta Arenas, Chile (Fig. 1). The ORCAS field campaign took place from Jan. 15 -Feb. During ORCAS and ATom-2 TOGA provided mixing ratios of over 60 organic compounds, 141 including HVOCs, at background levels. The instrument, described in Apel et al. (2015), 142 continuously collects and analyzes samples with a 35-second sampling period and repeats the 143 cycle every two-minutes using online fast gas chromatography and mass spectrometry. HVOCs 144 reported here have an overall ±15% relative accuracy and ±3% relative precision, and detection 145 limits of ≤ 0.2 ppt for CHBr 3 , CH 2 Br 2 , CHClBr 2 , CHBrCl 2 , and CH 3 I. This study also leverages 146 measurements of CH 3 Br with a detection limit of 0.2 ppt from whole air samples from the U.

147
Miami / NCAR Advanced Whole Air Sampler (AWAS; Schauffler et al., 1999) onboard the GV 148 during the ORCAS campaign and the UC Irvine Whole Air Sampler (WAS; Blake et al., 2001) 149 onboard the DC-8 during the ATom-2 campaign. In addition, comparisons between onboard 150 collected whole air samples and in-flight TOGA measurements, when sharing over half of their 151 sampling period with TOGA measurements, showed good correlations for CHBr 3 , CH 2 Br 2 , CH 3 I, 152 and CHClBr 2 , although there were some calibration differences ( Fig. S1 and Fig. S2). In 153 addition to the comparison between co-located atmospheric measurements, we also conducted a 154 lab inter-comparison following the campaign between NOAA's programmable flask package 155 (PFP) and TOGA (Table S1; see supplement for details). The AO2 instrument measures variations in atmospheric O 2 , which are reported as relative 159 deviations in the oxygen to nitrogen ratio (δ(O 2 /N 2 )), following a dilution correction for CO 2 160 (Keeling et al., 1998;Stephens et al. 2018). The instrument's precision is ± 2 per meg units (one 161 in one million relative) for a 5 second measurement (Stephens et al., 2003;Stephens et al., 162 manuscript in preparation, 2019). Anthropogenic, biogenic, and oceanic processes introduce O 2 163 perturbations that are superimposed on the background concentrations of O 2 in air (XO 2 , in dry 164 air = 0.2093). O 2 is consumed when fossil fuels are burned and produced during terrestrial 165 photosynthesis. Seasonal changes in the ocean heat content lead to small changes in atmospheric 166 N 2 . As others have done (Keeling et al., 1998;Garcia and Keeling, 2001;Stephens et al., 2018), 167 we isolated the air-sea O 2 signal by subtracting model estimates of the terrestrial photosynthesis, 168 fossil-fuel combustion, and air-sea N 2 flux influences from the δ(O 2 /N 2 ) measurement (Equation 169 1). The difference of the δ(O 2 /N 2 ) measurement and these modeled values is multiplied by XO 2 170 to convert to ppm equivalents as needed (ppm eq; Keeling et al., 1998;Equation 1).

171
O 2-ppm-equiv = [δ (O 2 205 3a, c), which suggests that these species may be co-emitted. Previous studies have documented 206 co-located source regions of CHBr 3 and CH 2 Br 2 in the Southern Ocean (e.g. Hughes et al., 2009;207 For these comparisons, both O 2 and CO 2 mixing ratios from the upper troposphere (5-7 km) were 220 subtracted from the data to detrend for seasonal and inter-annual variability ( Fig. 4; Fig. S4). In 221 Fig. 4 we present type II major axis regression fits to data between the ocean surface and the 222 lowest 7 km for bromocarbons with photochemical lifetimes of ≥ 1 month and from the lowest 2 223 km for CH 3 I with a photochemical lifetime of ~ 1 week. We used a type II major axis regression 224 model (bivariate) to balance the influence of measurement uncertainty in HVOCs (on the y-axis) 225 and the measurement uncertainty in O 2 and CO 2 (on the x-axis) on the regression slope (Ayers et 226 al. 2001;Glover et al., 2011). As noted by previous studies, simple least squares linear 227 regressions fail to account for uncertainties in predictor variables (e.g. Cantrell et al. 2008).

228
The robust correlations of CHBr 3 and CH 2 Br 2 with δ(O 2 /N 2 ), in both 2016 and 2017 and in 229 Region 1 and Region 2, provides support for a regional biogenic source of these two HVOCs Ocean is driven by net community production (the excess of photosynthesis over respiration) in 232 the surface mixed layer, surface warming, and to a lesser extent ocean advection and mixing (e.g. 233 Stephens et al., 1998;Tortell and Long 2009;Tortell et al., 2014). Note that we adjust for 234 influences on the δ(O 2 /N 2 ) from thermal N 2 fluxes (see Equation 1, Sect. 2.3 for details).

235
Biological O 2 supersaturation in the surface mixed layer develops quickly in the first several 236 days of a phytoplankton bloom and diminishes as community respiration increases and air-sea 237 gas exchange equilibrates the surface layer with the atmosphere on a timescale of ~ 1 week.

238
CHBr 3 (and CH 2 Br 2 ) is emitted from phytoplankton during the exponential growth phase 239 (Hughes et al., 2013), which often coincides with high net community production and the and other HVOCs is also similar to O 2 , although the photochemical loss of HOVCs will alter 242 their ratio over time.

243
Our observations suggest a biological source for CHBr 3 and CH 2 Br 2 in Region 1 ( Fig. 4a and Fig.   244 4b). In contrast to CHBr 3 and CH 2 Br 2 , we observe a weaker relationship between CH 3 I and O 2 in 245 Region 1 (Fig. 4c), consistent with the existence of other, non-biological sources of CH 3 I in this 246 region. Figure 4d-f illustrates strong relationships between all three HVOCs and O 2 in Region 2.

247
This implies that the dominant source of CH 3 I emissions over the Patagonian shelf is biological.

248
The slope of the regression between CHBr 3 and O 2 also changes noticeably between Region 1 249 and Region 2. Molar enrichment ratios are 0.20 ± 0.01, and 0.07 ± 0.004 nmol : mol for CHBr 3 250 and CH 2 Br 2 to O 2 in Region 1, and 0.32 ± 0.02, 0.07 ± 0.004 pmol : mol in Region 2. In Region 251 2, we also report enrichment ratios of CH 3 I to O 2 of 0.38 ± 0.03 pmol : mol, based on the 252 correlation in Figure 4f.

253
In contrast to O 2 , air-sea fluxes of CO 2 over the Southern Ocean during summer reflect the 254 balance of opposing thermal and biological drivers (e.g. Stephens et al., 1998;. Ocean  Ocean CO 2 fluxes are much less certain than for O 2 (Anav et al., 2015;Nevison et al., 2016). As

324
Region 2 also has much higher chl a (Fig. S4), supporting biogenic sources for these gases.

357
For this study, remotely sensed and reanalysis data were used with STILT influence functions in 358 linear and multi-linear regressions to explain observed mixing ratios of CHBr 3 , CH 2 Br 2 , CH 3 Br 359 and CH 3 I. These data included a combination of chl a, sea ice concentration, absorption due to 360 ocean detrital material, and downward shortwave radiation at the ocean surface.

361
We used daily sea ice concentration data (https://nsidc.org/data/nsidc-0081) at a 25 km x 25 km 362 spatial resolution between 39.23º S and 90º S, 180º W -180º E from the NASA National Snow 363 and Ice Data Center Distributed Active Archive Center (NSIDC; Maslanik et al., 1999). This observed at all visible wavelengths . This reanalysis data is available at a 388 higher temporal resolution and better spatial coverage than satellite retrievals of PAR or 389 temperature. 392 We used STILT to explore the relationships between observed mixing ratios and the upstream

Relationships between predicted influences and observations
We found statistically significant negative correlations between the upstream sea ice influence 422 and both CHBr 3 and CH 2 Br 2 mixing ratios, and no positive relationships between upstream sea-423 ice influence and any measured HVOC, such as CH 3 I in Region 1 (Fig. 7). Note, sea ice did not 424 include land ice; however, we also found a negative correlation between upstream land ice 425 influence and mixing ratios of HVOCs. We interpret this result to mean that increased campaigns are needed to further study the seasonality and regional strength sea ice related 442 HVOC emissions. 443 We observed a statistically significant positive correlation between the footprints of 8-day 444 satellite composites of the chl a concentration, which is widely used as a proxy for near-surface 445 phytoplankton biomass, and mixing ratios of CHBr 3 and CH 2 Br 2 in Region 1 (Fig. 8a and Fig.   446 8b). This finding corroborates previous findings from ship-borne field campaigns and laboratory 447 studies that have suggested a biogenic source for these two bromocarbons (e.g., Moore et al.,448 1996; Hughes et al., 2013), and further substantiates the current CAM-Chem parameterization of 449 regional bromocarbon emissions using satellite retrievals of chl a in polar regions. CH 3 Br 450 mixing ratios were not significantly correlated with chl a footprints (Fig. 8c). Although 451 potentially suggesting that marine phytoplankton and microalgae were not a strong regional 452 source of CH 3 Br during ORCAS, it is also possible that the relatively long lifetime of CH 3 Br 453 precludes a definitive analysis of its origin based on chl a using 7-day back-trajectories. Neither

454
CHCl Br 2 nor CHBrCl 2 were significantly correlated with chl a composite footprints (data not 455 shown); however, more observations of these short-lived species in the remote MBL are needed 456 to substantiate this result.

457
Similar to Lai et al. (2011), we observed a significant correlation between mixing ratios of CH 3 I 458 and total weekly upstream influence functions of 8-day chl a composites (Fig. 8d) correlations were observed with upstream influence functions on shorter timescales than seven 460 days. We found that CH 3 I, particularly in Region 1, was better explained by a multi-linear 461 regression with two predictors: 1) the influence function of downward shortwave radiation at the 462 surface (Fig. 9a) and 2) the absorption of light due to detrital material (Fig. 9b), yielding 463 improved agreement between predicted and observed CH 3 I (Fig. 9c). and PAR to the solar radiation necessary for the photo-production of CH 3 I in surface waters (e.g. 477 Happell et al., 1996;Yokouchi et al., 2001). We note that chl a, which is a proxy for living algal 478 biomass, was correlated with CDOM in Region 1 and Region 2, (r 2 = 0.24; data not shown).

479
Finally, we note that photochemical loss during transport is not accounted for in this analysis.

480
Low OH mixing ratios, cold temperatures, and lower photolysis rates due to angled sunlight at  CESM model nudged to reanalysis temperatures and winds as described in Stephens et al. (2018) to facilitate comparisons across regions and atmospheric models (Fig. 9). An earlier free running 499 version of CESM was one of the best evaluated for reproducing the seasonal cycle of O 2 /N 2 over 500 the Southern Ocean (Nevinson et al., 2015;. To date, the north-south gradient in 501 atmospheric O 2 has not been well reproduced by any models (Resplandy et al., 2016). Vertical  Figure 10 shows the mean emissions for Jan. and Feb. of CHBr 3 , CH 2 Br 2 , and CHClBr 2 in 520 Region 1 and Region 2. Mean regional emissions of CHBr 3 and CH 2 Br 2 and CHClBr 2 are 91 ± 8, 521 31 ± 17, and 11 ± 4 pmol m -2 hr -1 in Region 1 and 329 ± 23, 69 ± 5, and 24 ± 5 pmol m -2 hr -1 in 522 Region 2 (Table 1). The mean flux of CH 3 I in Region 2 is 392 ± 32 (Table 1). Table 1 (Table 1). We note that in Region 2, CAM-Chem fluxes 532 of CHBr 3 and CH 2 Br 2 , although still significantly different, are more similar to our estimated 533 fluxes. radiation and detrital material influence function coefficients and an interaction term from a 539 multi-linear regression (Fig. 9) were used to estimate an average non-biological flux of CH 3 I 540 ( Fig. 11; Table 1). This method could be used in place of the current Bell et al. (2002) 541 climatology to update near weekly (~8 day) emissions of CH 3 I in future versions of CAM-Chem.

542
Our estimated regional mean flux in Region 1 (35 ± 29 pmol m -2 hr -1 ) is significantly lower than 543 the current CAM-Chem estimated emissions (Table 1). As noted in Sect. 3, our observations of 544 CH 3 I are also much lower than the modeled mixing ratios. As discussed above, the strong 545 correlations between CH 3 I and O 2 in Region 2 also suggest a dominant biological source for this 546 compound. As a result, we have not used this relationship to parameterize a flux for CH 3 I in 547 Region 2 (see Sect. 2.5 and 5.1 for details). (Region 2) are moderate regional sources of CHBr 3 , CH 2 Br 2 , and CH 3 I, and weak sources of 559 CHClBr 2 and CHBrCl 2 in January and February. CAM-Chem provided a good foundation for 560 parameterizing HVOC emissions, particularly for CHBr 3 and CH 2 Br 2 in Region 1 and Region 2.

561
Conversely, CHClBr 2 and CHBrCl 2 emissions were underestimated by a factor of two or three in 562 the model, while CH 3 I emissions were overestimated by a factor of more than three, and airborne 563 observations indicated that the CAM-Chem CH 3 Br surface boundary condition may be too low 564 by ~25%.

565
Our results suggested that summertime biological HVOC fluxes may be parameterized with 566 some success based on airborne observations of enrichment ratios, as well the influence of 567 remotely sensed parameters. CHBr 3 and CH 2 Br 2 exhibited strong and robust correlations with O 2 568 as well as weaker correlations with the influence of chl a, which is a proxy for phytoplankton 569 biomass. CHClBr 2 and CHBr 3 were well correlated with one another. Together, these 570 correlations suggested a biological source for these gases over the Southern Ocean. We found 571 that CH 3 I mixing ratios in Region 1 were best correlated with a non-biological geophysical 572 influence function, although biogenic CH 3 I emissions appear important in Region 2.

573
Our flux estimates based on the relationship of HVOC mixing ratios to other airborne 574 observations and remotely sensed parameters compared relatively well with those derived from 575 global models and ship-based studies (Table 1). Our emission estimates of CHBr 3 , CH 2 Br 2 , 576 CH 3 I, and CHClBr 2 were lower than most prior estimates from the Antarctic polar region in 577 summer, although they were significantly higher than CAM-Chem's prescribed emissions in 578 Region 1, where HVOC mixing ratios are under predicted (Table 1; Fig. 5). In the case of CH 3 I,

579
our estimated emissions suggest that the prescribed emissions in CAM-Chem may be too high.  Region 1 (Fig.3a,b) and in Region 2 (Fig.3c,d). Type II major axis regression model (bivariate 937 least squares regressions) are based on ORCAS data below 2 km illustrates a regional 938 enhancement ratio. Error bars represent the uncertainty in HVOC measurements.