the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 12 Dec 2024
Advances in CALIPSO (IIR) cirrus cloud property retrievals – Part 1: Methods and testing
Abstract. In this study, we describe an improved CALIPSO (Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation) satellite retrieval which uses the Imaging Infrared Radiometer (IIR) and the CALIPSO lidar for retrievals of ice particle number concentration Ni, effective diameter De, and ice water content IWC. By exploiting two IIR channels, this approach is fundamentally different than another satellite retrieval based on cloud radar and lidar that retrieves all three properties. A global retrieval scheme was developed using in situ observations from several field campaigns. The Ni retrieval is formulated in terms of Ni/APSD ratios, where APSD is the directly measured area concentration of the ice particle size distribution (PSD), along with the absorption optical depth in two IIR channels and the equivalent cloud thickness seen by IIR. It is sensitive to the shape of the PSD, which is accounted for, and uses a more accurate mass-dimension relationship relative to earlier work. The new retrieval is tested against corresponding cloud properties from the field campaigns used to develop this retrieval, as well as a recent cirrus cloud property climatology based on numerous field campaigns from around the world. In all cases, favorable agreement was found. This analysis indicated that Ni varies as a function of cloud optical depth. By providing near closure to the ice PSD, the natural atmosphere may be used more like a laboratory for studying key processes responsible for the evolution and life cycle of cirrus clouds and their impact on climate.
- Preprint
(4123 KB) - Metadata XML
-
Supplement
(736 KB) - BibTeX
- EndNote
Status: open (extended)
-
RC1: 'Comment on egusphere-2024-3790', Anonymous Referee #1, 18 Mar 2025
reply
The manuscript by Mitchell et al. improves upon the development of an earlier methodology published in 2018 to retrieve the layer properties of optically thin cirrus clouds using a combination of the IIR and CALIOP instruments on the CALIPSO satellite. The synergistic combination of lidar and infrared measurements is a unique property of the CALIPSO satellite and were initially conceived for the very purpose of retrieving cirrus microphysics. The proposed algorithm developed in this manuscript seeks to use the ratio of the 12 micron to 10 micron radiances to derive a relationships between absorption optical depth and microphysics and the vertically integrated attenuated lidar backscatter that is related to the visible optical depth at 523nm to derive the layer-mean ice water content, the effective particle size, and ice crystal number concentration, Ni. The retrieval is formulated using a closed set of simplified equations for the radiative properties. The closure is achieved using analysis of in situ cirrus particle size distribution data from multiple cirrus measurement campaigns along with assumptions regarding the absorption properties of the PSD. The authors compare statistical results of the retrieval algorithm (i.e. regionally derived cirrus properties) to the statistics of the cirrus properties in the campaigns. Overall, I find the methodology unique and innovative builds on a long history of cirrus data analysis by the authors. However, I have several major concerns that I think the authors must address in a major revision of the present manuscript.
- The algorithm is dependent upon in situ cirrus data particle size distributions for relationships that allow for closure of the algorithm equations.While the general approach seems sound, there are several critical issues that should be addressed in more detail. The first of these is the fact that the algorithm retrieves the properties of a layer that may be several kilometers deep where vertical variations in cirrus properties are often considerable over the depth of the layer. However, aircraft data are typically collected along level flight lines within cirrus layers and they do not typically sample the vertical structure. Assuming that the relationships developed from the in situ data are relevant in small volumes, why would the relationships apply to entire cirrus layers particularly when the infrared absorption properties are likely weighted to a different region of the layer than the visible optical depth?
- Continuing with the previous comment, the PSDs measured in cirrus by probes are well known to contain shattering artifacts.While all modern probes now contain tips that reduce shattering and data are analyzed to remove clusters of small particles that likely resulted in shattering, this effect is still present and likely dominates (or at least biases) the number concentration of ice crystals, Ni, in aircraft measurements. I note that Kramer et al. (2020) in their cirrus climatology spend a large section of their analysis on this topic and still feel that it biases their results. The PSDs shown in Figure 8 in this manuscript are representative where there is a main mode of larger particles that is in the several hundred-micron size range and then tails off smoothly toward smaller sizes (like a modified gamma distribution would do) but this mode is interrupted in the 10's of microns range by a smaller mode that occupies the smallest several size bins. Can the authors explain the microphysical mechanism that would result in this behavior? In my experience, this mode of small particles is omnipresent in cirrus measurements, and, in my opinion, it likely represents shattering artifacts that have not been mitigated by tips or removed through software. I'd be happy if the authors could explain to me why I am wrong about this. Can the authors point to in situ data that does not show this small mode? However, it remains the case that this residual small mode biases the Ni in their data so the presence of the small mode should be addressed either via explaining the microphysical processes that produce it or by finding some way of removing it that is more sophisticated than simply dividing it by 10.4 as described in the paper that they apply to the first bin of the 2DS data.
- The authors address algorithm uncertainty in an appendix by deriving a number of equations that propagate the error in the measurements through to the retrieved quantities.While this seems thorough enough, they do not present actual results of the uncertainty analysis – just the equations. They do mention that pixel to pixel uncertainty in Ni can be on the order of a factor of 2, they do not, however, present an actual analysis of the error or show how it depends upon the uncertainties in the input variables or assumptions. Such an analysis is a requirement in my opinion.
- The authors move from algorithm development directly to statistical comparisons with other campaigns. They do not show actual data collected by the sensors and the retrieval results with error bars that would result from application of their algorithm to actual cirrus layers. Furthermore, it seems necessary to demonstrate an actual case or two in circumstances when in situ aircraft data were being collected data underneath the satellites. The SPartICus data that is used by the authors in their analysis has approximately two dozen flights where the Lear Jet flew along the paths of the CloudSat and CALIPSO satellites. These are well documented in Deng et al. (2013; DOI: 10.1175/JAMC-D-12-054.1). As a matter of fact, since the authors use SPartICus data, it seems reasonable for them to replicate the results in Deng et al. This would be quite straightforward. The authors also use data collected in TC4. Mace et al. (2010; doi:10.1029/2009JD012517) illustrate a case when the NASA DC8 flew along the CloudSat and CALIPSO tracks in tropical cirrus. Replicating these direct comparisons between algorithm results and in situ aircraft data is a necessary step toward establishing confidence in the algorithm results and would go a long way toward addressing many of the earlier concerns I have raised.
Citation: https://doi.org/10.5194/egusphere-2024-3790-RC1
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 148 | 41 | 16 | 205 | 22 | 13 | 11 |
- HTML: 148
- PDF: 41
- XML: 16
- Total: 205
- Supplement: 22
- BibTeX: 13
- EndNote: 11
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 97 | 50 |
| Germany | 2 | 13 | 6 |
| China | 3 | 10 | 5 |
| France | 4 | 10 | 5 |
| United Kingdom | 5 | 9 | 4 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 97