the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 May 2023
Investigating the development of clouds within marine cold air outbreaks
Abstract. Marine cold air outbreaks are important parts of the high-latitude climate system, and are characterised by strong surface fluxes generated by the air-sea temperature gradient. These fluxes promote cloud formation, which can be identified in satellite imagery by the distinct transformation of stratiform cloud 'streets' into a broken field of cumuliform clouds downwind of the outbreak. This evolution in cloud morphology changes the radiative properties of the cloud, and therefore is of importance to the surface energy budget. While the drivers of stratocumulus-to-cumulus transitions, such as aerosols or the sea surface temperature gradient, have been extensively studied for subtropical clouds, the factors influencing transitions at higher latitudes are relatively poorly understood. This work uses reanalysis data to create a set of composite trajectories of cold air outbreaks moving off the Arctic ice edge and co-locates these trajectories with satellite data to generate a unique view of cloud development within cold air outbreaks.
The results of this analysis show that clouds embedded in cold-air outbreaks have distinctive properties relative to clouds following other, more stable, trajectories in the region. The initial strength of the outbreak shows a lasting effect on cloud properties, with differences between clouds in strong and weak events visible over 30 hours after the air has left the ice edge. However, while the strength (measured by the magnitude of the marine cold-air outbreak index) of the outbreak affects the magnitude of cloud properties, it does not affect the timing of the transition to cumuliform clouds nor the top-of-atmosphere albedo. In contrast, the initial aerosol concentration does not strongly affect the magnitude of the cloud properties, but aerosol concentration is correlated to cloud break-up, leading to an enhanced cooling effect in clouds moving through high aerosol conditions due to delayed break-up. This evidence of precipitation suppression enhancing cloud lifetime highlights the need for information about aerosol sources at the ice edge to correctly model cloud development. Both the aerosol environment and the strength and frequency of marine cold air outbreaks are expected to change in the future Arctic, and these results provide insight into how these changes will affect the radiative properties of the clouds.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(8331 KB)
-
Supplement
(164 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(8331 KB) - Metadata XML
-
Supplement
(164 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-734', Anonymous Referee #1, 12 May 2023
This manuscript describes the analysis of marine cold-air outbreak (MCAO) events spreading from the North Atlantic to the Kara Sea. By examining subsets of satellite and reanalysis data, the authors examine the influence of parameters such as aerosol loading and stability conditions on cloud-field variables, with a unique temporally-evolving perspective. The manuscript is reasonably written and is generally well-referenced. Â I did not find any major issues with the science and methodology and think that this manuscript could become a nice contribution following minor revisions.
Comments:
l. 11 - recommend rephrasing as trajectories are not stable.
l. 18-21 - I understand what you're trying to communicate in those two sentences, but they seem somewhat detached, so I recommend rearranging/rephrasing.
l. 29-31 - because studies on the effect of clouds on sea ice cover find indications of additional feedbacks and effects, I recommend adding "for example" before "a decrease in cloud reflectivity ..."
l. 51-52 - this sentence is not clear. Stratification of the lower BL creates convective conditions? Do you mean that the cloud effectively forms a secondary BL (in addition to the now-decoupled mixed layer), which can support the formation of convective clouds because the coupled BL is now shallower?
In any case, a reference is required here.l. 51 - define "This"
Fig. 1 - coordinates are pixelated.
l. 59 - add "generally" after "aerosol concentration"
l. 66-67 - This sentence is somewhat repetitive of l. 55-56.
l. 68 - higher --> high
l. 69 - remove 'liquid'
l. 70-75 - sentence too complex. Recommend breaking it in two.
l. 71 - recommend adding Shupe, 2011.
l. 71 - recommend concern --> interest
l. 96 - above --> north of
l. 108 – to my knowledge, eq. 1 originates in Wood and Hartmann (2006). Please also provide a reference for eq. 2.
l. 133 - rho --> rho_w
l. 114 - Q here is not the scattering efficiency but the extinction efficiency
l. 119-124 - I doubt that 263 K and MODIS ice fraction retrievals really remove ice (or even most of it). You can say that ice occurrence is less likely and that you are focusing on liquid-dominated clouds, which is still somewhat vague. As you already noted, it is challenging to detect phase with a passive instrument.
l. 123 - the results of de Boer et al., 2009 refer to mass and don't claim that these clouds are ice-free, as you also demonstrate in Fig. 2. Recommend rephrasing this sentence accordingly.
l. 125 - provide a reference for CALIOP
Fig. 2 caption - 'Liquid' should be 'Warm' per the legend.
l. 150-151 - this sentence is somewhat repetitive - recommend combining it with the final sentence of the previous paragraph.
l. 150 - see 2.2 --> see Section 2.2
SLP can often be below 1000 hPa around cyclones over the North Atlantic, for example. In these cases, the 1000 hPa ERA5 data is wonky. How common are these cases in your dataset and what is the impact (if any) on the results?
l. 154-155 - what are the statistics of cloud top pressure justifying the use of the 750 hPa pressure level? I know that you present CTP later on in the manuscript but as a reader, this can be confusing as it is non-chronological.
l. 158 – The MCAO acronym now refers to both the index and the event itself, so one of these definitions requires a different acronym.
l. 160 - Note thatÂ
l. 170 - I don't understand Fig. 4 and what it represents. If the white areas are mainly regions with no data (a value of 0) as they exceed the range, then color it differently than white. Also, I recommend introducing Fig. 4 right after the following paragraph in which you explain the trajectory calculation.
l. 181-182 - So basically you are limiting the analysis to the region from the North Atlantic to the Kara Sea. This should be specified explicitly, and the boundaries should be highlighted in Fig. 5.
Fig. 6 - what do the shaded regions represent? Standard deviation? Quartiles?
l. 217 - Recommend adding Shupe, 2011
l. 226-229 - This was already discussed extensively in previous MCAO work such as Tornow et al., 2021; 2022.
l. 233-235 - probably higher than previous MODIS-based studies because of the different data filtering methodologies (excluding highly supercooled clouds).
l. 241-243 - these suggestions regarding airmass origin were recently demonstrated to some extent by Silber and Shupe, 2022.Â
l. 254 - Because the 340 W m-2 value is derived from the solar constant incident onto a disc instead of the full sphere recommend changing:
incident solar --> global average incident solarl. 258 - are --> is
l. 267-270 - That is a nice method. Is resampling performed with or without repetition?
Also, please provide the number of samples in each subset here and elsewhere (e.g., in figures).
Also, by "dividing the AOD distribution" do you refer to the AOD time series? This is not clear.l. 294 - reaching 15 um --> reaching the commonly-used 15 um threshold.
l. 300 - more unstable --> less stable
l. 302 - Figure 6 --> Figure 7
Also note that cloud depth is implicit here, so you should note that this is indicated in the figure.l. 303-305 - what is then the source for the steepest Nd decline before 15 h? Just drying entrainment? Nd is assumed to be constant with height here, so this is not just a cloud-top effect. Perhaps this is the ice effect you discuss towards the end, which is more likely at the lower CCT closer to the ice edge.
Section 3.3 - this is a very nice analysis
l. 330 - define INP and CCNÂ
l. 331-332 - right, but please provide a reference for this sentence.
l. 332 - simply --> qualitatively
l. 404 - since this is the future we are dealing with, I would use a weaker verb here such as insinuate, for example.
l. 412 - emit as black bodies --> have an emissivity near unity.
l. 420 - Recommend adding a reference to the MOSAiC overview paper (Shupe et al., 2022, https://doi.org/10.1525/elementa.2021.00060)
l. 423 – following my comment on l. 119-124, Fig. 2 clearly shows that a significant, though not a big, fraction of clouds here contain ice, so I wouldn't treat these clouds as strictly liquid-phase.
You could say liquid-dominated for example.l. 427-428 - right, but alternatively, there could also be higher ice concentrations around relatively higher temperatures around -5 C due to SIP.
l. 429-433 - because of the typical differences in ice and liquid number concentrations, I suspect that this effect is negligible here. That said, misclassification is always possible due to various reasons.
Also, the lower r_e in stronger events is counterintuitive, because these clouds typically deepen with distance from the ice edge (suggesting greater differences from weaker events in cloud top height and temperature), but you could refer to the cloud top temperature panel to support your argument.Â
Citation: https://doi.org/10.5194/egusphere-2023-734-RC1 -
RC2: 'Comment on egusphere-2023-734', Anonymous Referee #2, 23 May 2023
General
This paper describes the temporal development of CAO clouds from a composite of trajectories with collocated satellite and reanalysis data. Although the study period is during sunlit and only liquid clouds are selected to perform the analysis, I found the methodologies rigorous and results promising. The observational data analysis could be a benchmark to guide model development and evaluation. I only have several minor comments as listed below.
Specific
Figure 2 and other similar styled figures: can you add the description of the shadings to the figure caption?
Please provide full name of acronyms, such as DARDAR, SMMR DMSP SSM, ERA5
Line 158: MCAO is used to represent the boundary layer stability here, whereas it is short for marine cold-air outbreak before this point. Please use a different terminology for one of them (you used M to represent MCAO index in line 197).
Figure 3: are the MCAO index and FRO shown here for all March – October 2008 – 2014 regardless of the selection criteria you described in this section? If so, what does the figure look like from the samples after applying various data filters?
Following the previous question, can you show a map of MODIS sample numbers used in the analysis after all the filtering process? You can show that in the response and does not necessarily in the main text.
Line 222: should it be ‘decrease at a rate of 1 g m-2 hr-1’ or ‘change at a rate of -1 g m-2 hr-1’? similar for line 232.
Line 335-336: I am curious how do you resample in the MCAO index space and in the wind speed space at the same time? Do you form the 2-d bins with the MCAO-index and wind-speed and then resample in each 2-d bin?
Citation: https://doi.org/10.5194/egusphere-2023-734-RC2 - AC1: 'Comment on egusphere-2023-734', Rebecca Murray-Watson, 24 Jul 2023
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-734', Anonymous Referee #1, 12 May 2023
This manuscript describes the analysis of marine cold-air outbreak (MCAO) events spreading from the North Atlantic to the Kara Sea. By examining subsets of satellite and reanalysis data, the authors examine the influence of parameters such as aerosol loading and stability conditions on cloud-field variables, with a unique temporally-evolving perspective. The manuscript is reasonably written and is generally well-referenced. Â I did not find any major issues with the science and methodology and think that this manuscript could become a nice contribution following minor revisions.
Comments:
l. 11 - recommend rephrasing as trajectories are not stable.
l. 18-21 - I understand what you're trying to communicate in those two sentences, but they seem somewhat detached, so I recommend rearranging/rephrasing.
l. 29-31 - because studies on the effect of clouds on sea ice cover find indications of additional feedbacks and effects, I recommend adding "for example" before "a decrease in cloud reflectivity ..."
l. 51-52 - this sentence is not clear. Stratification of the lower BL creates convective conditions? Do you mean that the cloud effectively forms a secondary BL (in addition to the now-decoupled mixed layer), which can support the formation of convective clouds because the coupled BL is now shallower?
In any case, a reference is required here.l. 51 - define "This"
Fig. 1 - coordinates are pixelated.
l. 59 - add "generally" after "aerosol concentration"
l. 66-67 - This sentence is somewhat repetitive of l. 55-56.
l. 68 - higher --> high
l. 69 - remove 'liquid'
l. 70-75 - sentence too complex. Recommend breaking it in two.
l. 71 - recommend adding Shupe, 2011.
l. 71 - recommend concern --> interest
l. 96 - above --> north of
l. 108 – to my knowledge, eq. 1 originates in Wood and Hartmann (2006). Please also provide a reference for eq. 2.
l. 133 - rho --> rho_w
l. 114 - Q here is not the scattering efficiency but the extinction efficiency
l. 119-124 - I doubt that 263 K and MODIS ice fraction retrievals really remove ice (or even most of it). You can say that ice occurrence is less likely and that you are focusing on liquid-dominated clouds, which is still somewhat vague. As you already noted, it is challenging to detect phase with a passive instrument.
l. 123 - the results of de Boer et al., 2009 refer to mass and don't claim that these clouds are ice-free, as you also demonstrate in Fig. 2. Recommend rephrasing this sentence accordingly.
l. 125 - provide a reference for CALIOP
Fig. 2 caption - 'Liquid' should be 'Warm' per the legend.
l. 150-151 - this sentence is somewhat repetitive - recommend combining it with the final sentence of the previous paragraph.
l. 150 - see 2.2 --> see Section 2.2
SLP can often be below 1000 hPa around cyclones over the North Atlantic, for example. In these cases, the 1000 hPa ERA5 data is wonky. How common are these cases in your dataset and what is the impact (if any) on the results?
l. 154-155 - what are the statistics of cloud top pressure justifying the use of the 750 hPa pressure level? I know that you present CTP later on in the manuscript but as a reader, this can be confusing as it is non-chronological.
l. 158 – The MCAO acronym now refers to both the index and the event itself, so one of these definitions requires a different acronym.
l. 160 - Note thatÂ
l. 170 - I don't understand Fig. 4 and what it represents. If the white areas are mainly regions with no data (a value of 0) as they exceed the range, then color it differently than white. Also, I recommend introducing Fig. 4 right after the following paragraph in which you explain the trajectory calculation.
l. 181-182 - So basically you are limiting the analysis to the region from the North Atlantic to the Kara Sea. This should be specified explicitly, and the boundaries should be highlighted in Fig. 5.
Fig. 6 - what do the shaded regions represent? Standard deviation? Quartiles?
l. 217 - Recommend adding Shupe, 2011
l. 226-229 - This was already discussed extensively in previous MCAO work such as Tornow et al., 2021; 2022.
l. 233-235 - probably higher than previous MODIS-based studies because of the different data filtering methodologies (excluding highly supercooled clouds).
l. 241-243 - these suggestions regarding airmass origin were recently demonstrated to some extent by Silber and Shupe, 2022.Â
l. 254 - Because the 340 W m-2 value is derived from the solar constant incident onto a disc instead of the full sphere recommend changing:
incident solar --> global average incident solarl. 258 - are --> is
l. 267-270 - That is a nice method. Is resampling performed with or without repetition?
Also, please provide the number of samples in each subset here and elsewhere (e.g., in figures).
Also, by "dividing the AOD distribution" do you refer to the AOD time series? This is not clear.l. 294 - reaching 15 um --> reaching the commonly-used 15 um threshold.
l. 300 - more unstable --> less stable
l. 302 - Figure 6 --> Figure 7
Also note that cloud depth is implicit here, so you should note that this is indicated in the figure.l. 303-305 - what is then the source for the steepest Nd decline before 15 h? Just drying entrainment? Nd is assumed to be constant with height here, so this is not just a cloud-top effect. Perhaps this is the ice effect you discuss towards the end, which is more likely at the lower CCT closer to the ice edge.
Section 3.3 - this is a very nice analysis
l. 330 - define INP and CCNÂ
l. 331-332 - right, but please provide a reference for this sentence.
l. 332 - simply --> qualitatively
l. 404 - since this is the future we are dealing with, I would use a weaker verb here such as insinuate, for example.
l. 412 - emit as black bodies --> have an emissivity near unity.
l. 420 - Recommend adding a reference to the MOSAiC overview paper (Shupe et al., 2022, https://doi.org/10.1525/elementa.2021.00060)
l. 423 – following my comment on l. 119-124, Fig. 2 clearly shows that a significant, though not a big, fraction of clouds here contain ice, so I wouldn't treat these clouds as strictly liquid-phase.
You could say liquid-dominated for example.l. 427-428 - right, but alternatively, there could also be higher ice concentrations around relatively higher temperatures around -5 C due to SIP.
l. 429-433 - because of the typical differences in ice and liquid number concentrations, I suspect that this effect is negligible here. That said, misclassification is always possible due to various reasons.
Also, the lower r_e in stronger events is counterintuitive, because these clouds typically deepen with distance from the ice edge (suggesting greater differences from weaker events in cloud top height and temperature), but you could refer to the cloud top temperature panel to support your argument.Â
Citation: https://doi.org/10.5194/egusphere-2023-734-RC1 -
RC2: 'Comment on egusphere-2023-734', Anonymous Referee #2, 23 May 2023
General
This paper describes the temporal development of CAO clouds from a composite of trajectories with collocated satellite and reanalysis data. Although the study period is during sunlit and only liquid clouds are selected to perform the analysis, I found the methodologies rigorous and results promising. The observational data analysis could be a benchmark to guide model development and evaluation. I only have several minor comments as listed below.
Specific
Figure 2 and other similar styled figures: can you add the description of the shadings to the figure caption?
Please provide full name of acronyms, such as DARDAR, SMMR DMSP SSM, ERA5
Line 158: MCAO is used to represent the boundary layer stability here, whereas it is short for marine cold-air outbreak before this point. Please use a different terminology for one of them (you used M to represent MCAO index in line 197).
Figure 3: are the MCAO index and FRO shown here for all March – October 2008 – 2014 regardless of the selection criteria you described in this section? If so, what does the figure look like from the samples after applying various data filters?
Following the previous question, can you show a map of MODIS sample numbers used in the analysis after all the filtering process? You can show that in the response and does not necessarily in the main text.
Line 222: should it be ‘decrease at a rate of 1 g m-2 hr-1’ or ‘change at a rate of -1 g m-2 hr-1’? similar for line 232.
Line 335-336: I am curious how do you resample in the MCAO index space and in the wind speed space at the same time? Do you form the 2-d bins with the MCAO-index and wind-speed and then resample in each 2-d bin?
Citation: https://doi.org/10.5194/egusphere-2023-734-RC2 - AC1: 'Comment on egusphere-2023-734', Rebecca Murray-Watson, 24 Jul 2023
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 245 | 116 | 13 | 374 | 23 | 4 | 4 |
- HTML: 245
- PDF: 116
- XML: 13
- Total: 374
- Supplement: 23
- BibTeX: 4
- EndNote: 4
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Rebecca J. Murray-Watson
Edward Gryspeerdt
Tom Goren
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(8331 KB) - Metadata XML
-
Supplement
(164 KB) - BibTeX
- EndNote
- Final revised paper