the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Pollution slightly enhances atmospheric cooling by low-level clouds in tropical West Africa
Valerian Hahn
Ralf Meerkötter
Christiane Voigt
Sonja Gisinger
Daniel Sauer
Valéry Catoire
Volker Dreiling
Cyrille Flamant
Stefan Kaufmann
Jonas Kleine
Peter Knippertz
Manuel Moser
Philip Rosenberg
Hans Schlager
Alfons Schwarzenboeck
Jonathan Taylor
Abstract. Reflection of solar radiation by tropical low-level clouds has an important cooling effect on climate and leads to decreases in surface temperatures. Still, the effect of pollution on ubiquitous tropical continental low-level clouds and the investigation of the related impact on atmospheric cooling rates are poorly constrained by in-situ observations and modelling, in particular during the West African summer monsoon season. Here, we present comprehensive in-situ measurements of microphysical properties of low-level clouds over tropical West Africa, measured with the DLR aircraft Falcon 20 during the DACCIWA (Dynamics–Aerosol–Chemistry–Cloud Interactions in West Africa) campaign in June and July 2016. Clouds below 1800 meter altitude, identified as boundary layer clouds, were classified according to their carbon monoxide (CO) pollution level into pristine and less polluted clouds (CO < 135 ppbv) and polluted low-level clouds (CO > 155 ppbv) as confirmed by the linear CO to accumulation aerosol correlation. Whereas slightly enhanced aerosol background levels from biomass burning were measured across the entire area, clouds with substantially enhanced aerosol levels were measured in the outflow of major coastal cities, as well as over rural conurbations in the hinterlands. Here we investigate the impact of pollution on cloud droplet number concentration and size during the West African Monsoon season. Our results show that the cloud droplet number concentration (CDNC) measured in the size range from 3 µm to 50 µm around noon increases by 35 % in the elevated aerosol outflow of coastal cities and conurbations with elevated aerosol loadings from median CDNC of 240 cm-3 (52 cm-3 to 501 cm-3 interquartile range to 324 cm-3 (60 cm-3 to 740 cm-3 interquartile range). Higher CDNC resulted in a 17 % decrease in effective cloud droplet diameter from a median deff of 14.8 µm to a deff of 12.4 µm in polluted clouds.
Radiative transfer simulations show a non-negligible influence of droplet number concentrations and particle sizes on the net radiative forcing at the top of atmosphere of -16.3 W m-2 of the polluted with respect to the less polluted clouds and lead to a change in instantaneous heating rates of -18 K day-1 at top of the clouds at noon. It was found that the net radiative forcing at top of atmosphere accounts for only 2.6 % of the net forcing of the cloud-free reference case. Thus, polluted low-level clouds add only a relatively small contribution on top of the already exerted cooling by low-level clouds in view of a background atmosphere with elevated aerosol loading. Additionally, the occurrence of mid- and high-level cloud layers atop buffer this effect further, so that the net radiative forcing and instantaneous heating rate of low-level clouds turn out to be less sensitive towards projected future increases in anthropogenic pollution in West Africa.
Valerian Hahn et al.
Status: final response (author comments only)
-
RC1: 'Comment on acp-2022-795', Anonymous Referee #1, 23 Dec 2022
The authors separate observations of cloud microphysical properties by pollution levels (as indicated by CO) in the DACCIWA campaign and use the median differences to calculate instantaneous radiative forcing values due to the Twomey effect when low clouds occur alone or under midlevel and high cloud layers. The methods appear sound for the most part but the interpretation needs more nuance, especially in terms of the neglect of cloud adjustments. My major concern is that I am left unsure of the primary motivation for the study and how the results have advanced our knowledge of aerosol-cloud interactions in the region. I recommend “minor revisions” because I don’t think any new analyses need to be undertaken, but the motivation and context need to be clarified before publication is warranted.
General comments:
A. Introduction: I don’t understand the primary motivation for this study. There are lots of nice details about prior work but no clear story about what it all adds up to and where important knowledge gaps remain. Why is this new work needed? How does it relate to the previous work?
B. Cloud adjustments and radiative forcing: It would be helpful to be more precise in the discussion here. What you are calculating is the instantaneous radiative forcing due to the Twomey effect alone (effect of greater aerosol leading to greater cloud droplet number concentration and smaller effective radius for the same amount of cloud water). It neglects both cloud adjustments to the Twomey effect and other atmospheric adjustments to the changes in heating profiles and thus is not the effective radiative forcing, which is a distinction worth pointing out. It also may be worth thinking about the effect of aerosols on the clear-sky fluxes, as the “clean” and “polluted” cases would presumably also have different AOD values.
C. Cloud specification in RT model: In general, does the ~1 km cloud make sense? Is that how thick the clouds typically are? It would be helpful to perhaps discuss distributions of cloud top and base heights for the polluted and clean cases. This would also be relevant for thinking about how to interpret the vertical profiles in Figure 6 (see comment below).
Specific comments:
Line 71: What is meant by “large mode”? A large fraction of the accumulation mode?
Lines 72-73: There are many more studies than just Painemal et al. (2014) that study the effect of smoke CCN on low level clouds! Is there a particular point you want to make here about that paper?
Line 72: CCN has not yet been defined.
Line 108: Instantaneous observations are unable to directly measure the “influence of aerosol loading on microphysical properties”.
Line 152: Wouldn’t CO also trace biomass burning plumes?
Line 201: Why not give the details of when these clouds were observed here?
Lines 218-219: Where is the AOD assumed to reside vertically?
Line 292: Is the CLARIFY value referring to median CO in the boundary layer? CO in the free tropospheric biomass burning plumes observed during CLARIFY and ORACLES were substantially higher than this value. Also, the median value between clean air and heavily polluted plumes might not be a particularly meaningful metric.
Line 303: Are these measurements only around noon? This should be clarified in the methods section.
Lines 323-326: Why is this not shown?
Figure 6: The droplet size results are potentially convolving cloud vertical structure and microphysical differences. An adiabatic cloud should have increasing droplet size with height. Are the distributions of cloud tops and bases similar between the more and less polluted cases? If not, that would influence the comparison here.
Lines 344-347: This is a somewhat confusing and oversimplified discussion of indirect aerosol effects. Cloud adjustments to the Twomey effect (holding LWC constant, greater aerosol leads to greater CDNC/lower ED) can be large in magnitude and substantially enhance or counteract the radiative forcing from the Twomey effect alone. Your study only addresses the Twomey effect, but the neglect of adjustments should be mentioned as a source of uncertainty.
Line 361: Twomey effect only, not “pollution effect,” which could encompass direct, semi-direct, and indirect effects.
Lines 379-381: I don’t follow where this discussion is coming from.
Lines 398-399: The SW rate isn’t converging to the LW values, the total is converging to LW, right? Wouldn’t it just be easier to say the SW rate approaches zero?
Line 413: Lower than? Not “smaller.”
Line 427: “For an entire day” is ambiguous here, as the net forcing is positive at night. Integrated over an entire day?
Line 446: I’m not sure how you’re using “climate sensitivity” in this sentence.
Forcing values aren’t accounting for any change in AOD
Citation: https://doi.org/10.5194/acp-2022-795-RC1 -
AC1: 'Reply on RC1', Valerian Hahn, 17 Feb 2023
We thank the Referee for the constructive and valuable feedback.
Please find our reply in the attached pdf-file.
- AC3: 'Reply on AC1', Valerian Hahn, 15 Mar 2023
-
AC1: 'Reply on RC1', Valerian Hahn, 17 Feb 2023
-
RC2: 'Comment on acp-2022-795', Anonymous Referee #2, 06 Jan 2023
The manuscript presents a report of cloud measurements conducted during DACCIWA, which are classified as either clean or polluted clouds, and then used as inputs to radiative transfer calculations to estimate TOA radiative forcing across the cases. The results are consistent with the first indirect effect -- there is more TOA net radiative cooling for the polluted cloud cases than for the clean cloud cases (as estimated by CO). While this is a nice summary of some of the campaign measurement results with extension to radiative forcing via simple calculations, the scientific depth of analysis is rather shallow. What are we to conclude from this campaign with regard to aerosol-cloud interactions, and would we expect that the results of this interesting set of cases from June-July, 2016, would be regionally representative or consistent with other seasons or years? A good first step would be to add a conclusions section to the manuscript that sums up the impact and implications of the campaign results. Overall, the mansuscript is well written and the methods and results sections are appropriately detailed. It is appropriate for ACP. I'd recommend the manuscript be revised to bolster the conclusions and to address the minor comments below:
Line 25: is accumulation aerosol on a number concentration basis?Line 31: Add close parenthesis
Line 36: I don't understand what is being said in the sentence: "Thus, polluted low-level clouds add only a relatively small contribution on top of the already exerted cooling by low-level clouds in view of a background atmosphere with elevated aerosol loading". Are the authors making the case that the indirect cooling from polluted clouds is similar to the direct cooling from pollution aerosols in the absence of clouds? Please clarify.
Line 234: Should this be αΔλ?
Line 266: Were the OPC size distributions fitted (say to a lognormal function) in order to account for the accumulation mode contribution below 250 nm?
Lines 270-272: Were aerosol measurements made within the vicinity of the cloud? Is the accumulation mode aerosol just below cloud well correlated with the CO-correlation-based estimate?
Figure 5: what is the lowest droplet diameter shown on the x-axis?
Citation: https://doi.org/10.5194/acp-2022-795-RC2 -
AC2: 'Reply on RC2', Valerian Hahn, 17 Feb 2023
We appreciate your constructive feedback. Please find our reply in the attached pdf-file.
- AC4: 'Reply on AC2', Valerian Hahn, 15 Mar 2023
-
AC2: 'Reply on RC2', Valerian Hahn, 17 Feb 2023
Valerian Hahn et al.
Valerian Hahn et al.
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
284 | 86 | 13 | 383 | 4 | 6 |
- HTML: 284
- PDF: 86
- XML: 13
- Total: 383
- BibTeX: 4
- EndNote: 6
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1