Review of “Verification of the AIRS and MLS ozone algorithm based on retrieved daytime and nighttime ozone” by Wang et al.
The authors have improved their paper and especially the clarity of their discussion and conclusions. An evaluation of the diurnal variability in AIRS and MLS ozone retrievals have scientific significance. It is important to understand the strengths and weaknesses of different satellite data products to correctly interpret their depiction of real atmospheric change. For this reason, I gave this paper a careful read. The authors are correct when they say that not much literature exists that adequately characterizes AIRS O3 diurnal variability and their paper could, therefore, add value to the broader scientific community. After careful assessment, however, my decision is that this paper, in its current form, is not ready for publication because the results presented here do not adequately support the authors’ conclusions. The authors present an interesting problem, namely that the AIRS total column ozone (TCO) product sometimes has diurnal variability over some areas that cannot be explained by photochemical processes or O3 transportation. They conclude that these changes must be due to retrieval artefacts and specifically those that are caused by the mischaracterization of surface emissivity and cloud top pressure. Below I attempt to justify why I think this is an oversimplification.
First, I will address specific statements in the Conclusion (Section 4, pages 9-10) followed by a discussion of the figures and technical issues.
(1) AIRS TCO diurnal variation due to surface emissivity (Lines 266–278).
- Figure 1 shows AIRS TCO diurnal variation over the Sahara desert for the months of Dec through May and over the Congo (tropical forest) for the months of Jun through Nov. If dry, hot surfaces were the primary reason for these variational biases then we would not see this seasonal latitudinal shift and would see similar TCO variations over Namibia, Australia and the USA Southwest, all dry hot regions. How do the authors explain the diurnal variability of TCO over the Congo, which is not a dry surface?
- The AIRS a-priori for O3 is a climatology without diurnal variation. Similarly, its a-priori for emissivity is a monthly average without diurnal variation. If the radiance information content for O3 is very low and the retrieval mostly sticks to the a-priori, then there will be no diurnal variability in the AIRS TCO product.
- I do not doubt that retrieval artifacts indeed cause the observed TCO variability, but I argue that they cannot be explained by surface emissivity alone for a number of reasons, (i) not all surface types have diurnal variability in their emissivity and most of the patterns observed globally in Figure 1 cannot be explained by emissivity alone, (ii) using IASI measurements, Masiello et al., (2014) presented a very clear case for why and how the Sahara desert emissivity pose a problem to infrared O3 retrievals that the authors should consider here, (iii) apart from emissivity, AIRS O3 channels are also sensitive to surface temperature as well as boundary layer temperature and water vapor that have large diurnal variabilities that can affect O3 retrievals, and (iv) infrared O3 retrievals depend on an accurate representation of atmospheric temperature and water vapor; any uncertainty in their depiction at a given point in time and space will propagate into the O3 retrieval (Smith and Barnet, 2019).
(2) AIRS TCO diurnal variation due to cloud top height variation (Lines 284–293).
- Line 284: “detection of cloud features in AIRS TCO day-night differences is difficult due to the presence of land surface emissivity related bias”. How can this be true? Clouds present, by far, the largest source of uncertainty to atmospheric retrievals from space-based infrared instruments. AIRS observe the atmosphere from the top-down so clouds dominate the signal, not surface emissivity.
- The conclusion the authors draw here, that cloud top pressure cause diurnal TCO variation is not substantiated by their results; they do not provide quantitative information nor any figures in the main text. The only figure that they cite is Figure S2 in the supplement, which does not make a clear case.
- Again, I do not dispute the fact that diurnal variation in cloud uncertainty can affect the quality of O3 retrievals. I question whether the authors presented enough evidence to substantiate their claim.
Figures
I think the authors will strengthen the value and significance of their results by improving their figure captions. For example, it helps a great deal if acronyms used in the figures are defined in the caption.
Figure 1:
- The text of the colorbars for panels (e) and (f) is too small to read comfortably.
- I don’t understand panel (f) which the authors describe as the “longitude gradient value using absolute difference between two pixels adjacent at the same latitude in (e)”. It is unclear to me how the values were calculated and what they mean in the present context. In the main text the authors merely mention that “Figure 1f shows significant TCO changes at the land-ocean interface.” Looking at Figure 1f, I do not distinguish a clear land-ocean interface everywhere and have no good sense if a value of ~2 (units?) can be deemed “significant”. Can the authors explain how this figure contributes to the discussion and substatiates their conclusions?
Figure 5: Here the authors depict AIRS stratospheric column ozone (SCO), which they define as the vertical pressure range from 250 hPa to 1 hPa.
- While this range makes sense, given that MLS profiles (Figures 2 and 4) are defined within the same pressure bounds, it complicates interpretation of the AIRS SCO results. Here’s why:
(1) Unlike MLS, AIRS channels in the 10 micron spectral region are sensitive to tropospheric and stratospheric O3. In the high latitudes, the tropopause occurs at higher pressure levels (~220 hPa) than in the Tropics (~100 hPa). When the authors define AIRS SCO using a fixed pressure range across all latitudes then their results will include upper tropospheric effects in the low latitudes, but not the high latitudes. Their results of SCO in the mid-latitudes and Tropics (panels a and b), therefore is not strictly stratospheric but includes O3 amounts from the upper troposphere. In contrast, the results depicted for high latitudes (panels a through f) represents information mostly about the stratosphere. A comparison between, say panels b and f, is therefore not as straightforward as the authors make it out to be. AIRS O3 sensitive channels are predominantly, not exclusively, sensitive to O3. In addition to O3, channels in the 10 micron AIRS band are also sensitive to temperature, lapse rate, water vapor gradients, surface temperature, emissivity and clouds, for example.
Olson et al. (2017, page 104 of the “AIRS Version 6 Release Level 2 Product User Guide”, which the authors cite) explain the AIRS O3 product as such: “Errors in temperature profiles and water vapor mixing ratios will adversely affect the ozone retrieval. Significant biases (0 - 100%) may exist in the region between ~300 hPa and ~80 hPa; such biases currently being evaluated. Ozone mixing ratio data may not be reliable at pressures greater than 300 hPa or if the tropospheric mixing ratio is less 100 ppbv, however results may be qualitatively correct under conditions of high upper tropospheric ozone (such as in a tropopause fold). Mixing ratios and columns should not be considered reliable under conditions of very low skin temperatures (< 240 K). V6 is a slight improvement over V5, but there are still problems when the surface skin temperature does not compare well to AMSR-E or the scene has a significant cloud amount.”
(2) When surface temperature is very low and atmospheric temperature lapse rate is close to zero (conditions often occurring in the Polar regions), AIRS channel sensitivity to O3 decreases significantly. This means the AIRS O3 retrieval will have a strong dependence on its climatological a-priori, which has no diurnal variation. This, at least, partly explains why AIRS SCO have low diurnal variability in the Polar regions. The authors oversimplify their interpretation of results by attributing the lack of SCO diurnal variability in the Polar regions to homogeneity in surface emissivity.
Figure 6:
- I would recommend that the authors enlarge panels e and f. This is an interesting result and its value will improve if the error bars can be distinguished and interpreted more readily.
Figure S1:
- I have studied both panels in this figure and their meaning remain unclear to me since they do not depict a correlation between diurnal variability and surface type. It is true that desert emissivity can have diurnal variation due to changes in soil moisture (Li et al., 2012; Masiello et al., 2014) but the same does not hold for forests. What do the authors wish to demonstrate here?
- In short, AIRS O3 verticality does not depend on surface emissivity, but predominantly on O3 amount, temperature, instrument spectral resolution, lapse rate, and so on. It is true that the AIRS O3 channels are sensitive to emissivity in the boundary layer and that errors in the representation of surface emissivity over deserts can introduce a small bias in stratospheric O3 retrievals (Masiello et al., 2014), but surface emissivity is not a dominant source of uncertainty in TCO.
- The definition the authors give here for verticality is not correct. Their definition, instead, applies to averaging kernels. Verticality is derived from the averaging kernels as a total column representation of the kernel functions and can therefore be greater than unity.
Figure S2: In addition to surface emissivity, the authors attribute diurnal variation in AIRS TCO to cloud top pressure. This is the only figure they have included in their paper that attempts to correlate cloud properties with TCO. I do not see a clear statistical correlation in this scatter plot and it is unclear to me how the authors reached their conclusion. The results will be more meaningful if the authors explain their figure more clearly and also include statistical values that quantify the correlations they wish to highlight.
Data sources
The authors switch between AIRS total column ozone (TCO) and stratospheric column ozone (SCO) without adequately explaining why or how. Similarly, their use of AIRS Level 2 and Level 3 products creates confusion that obscures the relevance of their results. I recommend that the authors update Section 2.1 to clarify which products they used, where they sourced them from and why they switched between L2 and L3, SCO and TCO.
- In Section 2.1 the authors state that they employed AIRS V6 Level 3 products and list a weblink as source: https://disc.gsfc/nasa.gov/datasets/AIRS3STD_006/summary”. When I follow this link, I see a tab at the bottom of the page called “Data Citation” that gives an example of how this data product should be cited. I recommend that the authors use the correct citation for AIRS V006 L3 with DOI number included. Here it is:
AIRS Science Team/Joao Teixeira (2013), AIRS/Aqua L3 Daily Standard Physical Retrieval (AIRS-only) 1 degree x 1 degree V006, Greenbelt, MD, USA, Goddard Earth Sciences Data and Information Services Center (GES DISC), Accessed: [Data Access Date], doi:10.5067/Aqua/AIRS/DATA303
- On Page 11 in the “Data Availability” section, the authors list a different source for the AIRS TCO products as https://giovanni.gsfc.nasa.gov/giovanni. Giovanni is a GES DISC data tool, not data archive. Similarly, for MLS the authors list here a different source from the one they reported in Section 2.2. I recommend that the authors list a single source for their AIRS and MLS products to avoid confusion and promote confidence in their work.
- In Supplemental figures S1 and S2, the authors list “AIRS L2” as the product they depict yet they do not list a source for it or discuss why they used the Level 2 as opposed to the Level 3 product. Can the authors clarify this?
Technical Issues
- The following statement on lines 9-11 gave me pause: “Based on knowledge of the chemistry and transport of O3, significant deviations between daytime and nighttime O3 are only expected either in the planetary boundary layer or high in the stratosphere or mesosphere, having little effect on the TCO.” If O3 variation in the stratosphere or PBL is significant, then it would affect TCO, not so? The fact is that O3 concentrations vary with atmospheric pressure. I expect that variation in stratospheric O3, where concentrations are large relative to the free troposphere, will most definitely affect TCO.
- Lines 50-51: “UV absorption spectroscopy with the sun or stars as sources of UV light is the most used method to derive O3” To substantiate this statement the authors cite a publication from 1978 and 2000, both well before the launch of modern-era instruments. What metric did the authors use to define “UV light” as the “most used method”?
- Replace “long record” with “long term record” throughout the text to promote ease of reading (e.g., lines 13, 48).
- Line 21: Awkward sentence – “support for retrieval method origin of AIRS day-night TCO differences”. Rephrase.
- Line 30: “has become as an important topic”, remove “as”.
- Line 73: “Further” should be “Furthermore”.
- Line 152: Error in “over oceans differences”
- Line 158:” In Figure 1e shows” should be “Figure 1e shows”.
- Line 159: Awkward sentence – “with persistent clouds and Arctic regions”. Rephrase.
- Line 175: “cloud covers is another problem” should be “cloud cover”.
- Line 192: “about 500 pixels”. This fact will be more meaningful if the authors report it as a percentage.
- Line 249: “large areas at night besides in tropical western Pacific”. Rephrase, awkward sentence.
- Line 269: “points to the emissivity rather than the averaging kernels dominating these differences”… Averaging kernels quantify the signal-to-noise and is a product of the retrieval system, not a cause for it.
- Line 303: Awkward sentence – “only in the mesosphere the loss timescale for O3”. Rephrase.
- The Reference section needs work since there appears to be many inconsistencies in punctuation, journal abbreviations, doi numbers and so on.
- The font style of figure captions in the main text is different from those in the supplement.
This paper lacks references to current and relevant research in the field of remote sounding and I provide here a list of papers for the authors to consider:
- AIRS instrument (Aumann et al., 2003, 2020; Chahine et al., 2006)
- AIRS retrieval algorithm (Susskind et al., 2003, 2011, 2014)
- IR ozone retrieval quality, uncertainty and evaluation (Li et al., 2012; Maddy et al., 2009; Maddy and Barnet, 2008; Nalli et al., 2018; Smith and Barnet, 2019, 2020; Tian et al., 2007; Wang et al., 2012; Wei et al., 2010)
Bibliography
Aumann, H. H., Chahine, M. T., Gautier, C., Goldberg, M. D., Kalnay, E., McMillin, L. M., Revercomb, H., Rosenkranz, P. W., Smith, W. L., Staelin, D. H., Strow, L. L. and Susskind, J.: AIRS/AMSU/HSB on the aqua mission: design, science objectives, data products, and processing systems, IEEE Trans. Geosci. Remote Sens., 41(2), 253–264, doi:10.1109/TGRS.2002.808356, 2003.
Aumann, H. H., Broberg, S. E., Manning, E. M., Pagano, T. S. and Wilson, R. C.: Evaluating the Absolute Calibration Accuracy and Stability of AIRS Using the CMC SST, Remote Sens., 12(17), 2743, doi:10.3390/rs12172743, 2020.
Chahine, M. T., Pagano, T. S., Aumann, H. H., Atlas, R., Barnet, C., Blaisdell, J., Chen, L., Divakarla, M., Fetzer, E. J., Goldberg, M., Gautier, C., Granger, S., Hannon, S., Irion, F. W., Kakar, R., Kalnay, E., Lambrigtsen, B. H., Lee, S.-Y., Le MARSHALL, J., Mcmillan, W. W., Mcmillin, L., Olsen, E. T., Revercomb, H., Rosenkranz, P., Smith, W. L., Staelin, D., Strow, L. L., Susskind, J., Tobin, D., Wolf, W. and Zhou, L.: AIRS: Improving Weather Forecasting and Providing New Data on Greenhouse Gases, Bull. Am. Meteorol. Soc., 87(7), 911–926, doi:10.1175/BAMS-87-7-911, 2006.
Li, Z., Li, J., Li, Y., Zhang, Y., Schmit, T. J., Zhou, L., Goldberg, M. D. and Menzel, W. P.: Determining diurnal variations of land surface emissivity from geostationary satellites: EMISSIVITY DIURNAL VARIATIONS, J. Geophys. Res. Atmospheres, 117(D23), n/a-n/a, doi:10.1029/2012JD018279, 2012.
Maddy, E. S. and Barnet, C. D.: Vertical Resolution Estimates in Version 5 of AIRS Operational Retrievals, IEEE Trans. Geosci. Remote Sens., 46(8), 2375–2384, doi:10.1109/TGRS.2008.917498, 2008.
Maddy, E. S., Barnet, C. D. and Gambacorta, A.: A Computationally Efficient Retrieval Algorithm for Hyperspectral Sounders Incorporating A Priori Information, IEEE Geosci. Remote Sens. Lett., 6(4), 802–806, doi:10.1109/LGRS.2009.2025780, 2009.
Masiello, G., Serio, C., Venafra, S., DeFeis, I. and Borbas, E. E.: Diurnal variation in Sahara desert sand emissivity during the dry season from IASI observations: DIURNAL EMISSIVITY VARIATION, J. Geophys. Res. Atmospheres, 119(3), 1626–1638, doi:10.1002/jgrd.50863, 2014.
Nalli, N. R., Gambacorta, A., Liu, Q., Tan, C., Iturbide-Sanchez, F., Barnet, C. D., Joseph, E., Morris, V. R., Oyola, M. and Smith, J. W.: Validation of Atmospheric Profile Retrievals from the SNPP NOAA-Unique Combined Atmospheric Processing System. Part 2: Ozone, IEEE Trans. Geosci. Remote Sens., 56(1), 598–607, doi:10.1109/TGRS.2017.2762600, 2018.
Smith, N. and Barnet, C. D.: Uncertainty Characterization and Propagation in the Community Long-Term Infrared Microwave Combined Atmospheric Product System (CLIMCAPS), Remote Sens., 11(10), 1227, doi:10.3390/rs11101227, 2019.
Smith, N. and Barnet, C. D.: CLIMCAPS observing capability for temperature, moisture, and trace gases from AIRS/AMSU and CrIS/ATMS, Atmospheric Meas. Tech., 13(8), 4437–4459, doi:10.5194/amt-13-4437-2020, 2020.
Susskind, J., Barnet, C. D. and Blaisdell, J. M.: Retrieval of atmospheric and surface parameters from AIRS/AMSU/HSB data in the presence of clouds, IEEE Trans. Geosci. Remote Sens., 41, 390–409, 2003.
Susskind, J., Blaisdell, J. M., Iredell, L. and Keita, F.: Improved Temperature Sounding and Quality Control Methodology Using AIRS/AMSU Data: The AIRS Science Team Version 5 Retrieval Algorithm, IEEE Trans. Geosci. Remote Sens., 49(3), 883–907, doi:10.1109/TGRS.2010.2070508, 2011.
Susskind, J., Blaisdell, J. M. and Iredell, L.: Improved methodology for surface and atmospheric soundings, error estimates, and quality control procedures: the atmospheric infrared sounder science team version-6 retrieval algorithm, J. Appl. Remote Sens., 8(1), 084994, doi:10.1117/1.JRS.8.084994, 2014.
Tian, B., Yung, Y. L., Waliser, D. E., Tyranowski, T., Kuai, L., Fetzer, E. J. and Irion, F. W.: Intraseasonal variations of the tropical total ozone and their connection to the Madden-Julian Oscillation: THE MJO IN TROPICAL TOTAL OZONE, Geophys. Res. Lett., 34(8), doi:10.1029/2007GL029451, 2007.
Wang, H., Zou, X. and Li, G.: An Improved Quality Control for AIRS Total Column Ozone Observations within and around Hurricanes, J. Atmospheric Ocean. Technol., 29(3), 417–432, doi:10.1175/JTECH-D-11-00108.1, 2012.
Wei, J. C., Pan, L. L., Maddy, E., Pittman, J. V., Divarkarla, M., Xiong, X. and Barnet, C.: Ozone Profile Retrieval from an Advanced Infrared Sounder: Experiments with Tropopause-Based Climatology and Optimal Estimation Approach, J. Atmospheric Ocean. Technol., 27(7), 1123–1139, doi:10.1175/2010JTECHA1384.1, 2010. |
- Printer-friendly version
- Supplement
