the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
What CloudSat can't see: Liquid water content profiles inferred from MODIS and CALIOP observations
Richard M. Schulte
Matthew D. Lebsock
John M. Haynes
Abstract. Single layer nonprecipitating warm clouds are integral to Earth’s climate, and accurate estimates of cloud liquid water content for these clouds are critical for constraining cloud models and understanding climate feedbacks. As the only cloud-sensitive radar currently in space, CloudSat provides very important cloud profiling capabilities. However, a significant fraction of clouds are missed by CloudSat, because they are either too thin or too close to the earth’s surface. We find that the CloudSat 2B-CWC-RVOD product misses about 73 % of nonprecipitating liquid cloudy pixels, and about 63 % of total nonprecipitating liquid cloud water content, compared to coincident MODIS observations. Those percentages increase to 84 % and 69 %, respectively, if MODIS “partly cloudy” pixels are included. We develop a method, based on adiabatic parcel theory but modified to account for the fact that observed clouds are often subadiabatic, to estimate profiles of cloud liquid water content based on MODIS observations of cloud top effective radius and cloud optical depth combined with CALIPSO observations of cloud top height. We find that, for cloudy pixels that are detected by CloudSat, the resulting subadiabatic profiles of cloud water are similar to what is retrieved from CloudSat. For cloudy pixels that are not detected by CloudSat, the subadiabatic profiles can be used to supplement the CloudSat profiles, recovering much of the missing cloud water and generating realistic-looking merged profiles of cloud water. Adding this missing cloud water to the CWC-RVOD product increases the mean cloud liquid water path by 228 % for single layer nonprecipitating warm clouds. This method will be included in a subsequent reprocessing of the 2B-CWC-RVOD algorithm.
Richard M. Schulte et al.
Status: final response (author comments only)
-
RC1: 'Comment on amt-2023-49', Anonymous Referee #1, 18 Apr 2023
-
AC1: 'Reply on RC1', Rick Schulte, 26 May 2023
We thank the reviewer for their constructive comments. Reviewer comments are in italics, and our responses in regular font.
- Eq. (11) on page 7: The cloud depth H (i.e., geometric thickness) is derived from cloud optical thickness and effective radius from the MODIS cloud product. I wonder if this formulation guarantees that the cloud bottom does not reach the ground level or does not provide such situations in practice.
In rare cases, the method for retrieving H described in the appendix results in a cloud depth that is greater than the cloud top height – that is, physically impossible. In these cases, we iteratively increase the value of the condensation rate (c) by 1% and re-compute H, repeating this process until we arrive at a value of H that is less than the cloud top height. We have added this explanation to the revised manuscript (Lines 416-418).
- Lines 191-192 on page 8: As the MODIS product assumes the single-layer homogeneous cloud in the retrieval process, applying this assumption to vertically inhomogeneous clouds leads to a systematic bias in the retrieval products (i.e., t and re) due to vertically inhomogeneous microphysical properties (Platnick, 2000). Although these biases are small for adiabatic clouds (Saito et al., 2019), it would be good to mention this here.
This is a good point. We have mentioned this in Lines 193-196 and now cite both Platnick 2000 and Saito et al. 2019.
- Line 239 on page 8: “estimates of estimates of” should be “estimates of.”
Fixed as suggested.
- Line 281 on page 10: Figure 9a shows SW reflectance (true color), but the corresponding description indicates IR brightness temperature. Please clarify this.
The figure description should say true color reflectance. This has been corrected in the revised manuscript (Line 271).
- Line 290-291 on page 10: “are some are some” should be “are some.”
Fixed as suggested.
Citation: https://doi.org/10.5194/amt-2023-49-AC1
-
AC1: 'Reply on RC1', Rick Schulte, 26 May 2023
-
RC2: 'Comment on amt-2023-49', Anonymous Referee #2, 29 Apr 2023
The article discusses the importance of accurately estimating cloud liquid water content in single-layer nonprecipitating warm clouds. It notes that the CloudSat radar misses a significant fraction of these clouds due to below-threshold radar signals or overlapping with the surface clutter zone. To address this issue, the authors develop a method to estimate cloud liquid water content using MODIS and CALIPSO observations, which can supplement the CloudSat profiles for missing cloud water. Adding this missing cloud water increases the mean cloud liquid water path for single-layer nonprecipitating warm clouds by 228%. This method will be included in a reprocessing of the CloudSat algorithm.
Manuscripts of this kind of technique often become very technical. But this article is excellently written and very appealing to read. I only have a few minor comments.
- Line 210: The SLNPW cloud fraction shown in Figure 3 represents the SLNPW cloud fraction of CloudSat pixels, not CALIOP or MODIS cloud fraction. The title of Figure 3 and the text in the manuscript could be modified to explicitly state that the cloud fractions of MODIS or CALIOP are CloudSat pixels-based cloud fractions. Additionally, including the actual MODIS and CALIOP SLNPW cloud fractions (or the total cloud fraction), would provide the reader with a comparison of MODIS and CALIOP cloud fractions to the CloudSat pixels-based cloud fraction.
- Line 258: The MODIS-based sub-adiabatic LWP estimations show a positive bias in Fig. 7 panel (d), which could be attributed to the COT biases in the MODIS observations. Perhaps the authors could elaborate on the criteria used to choose the best COT observations from MODIS (e.g., flags) and consider the solar zenith angle when choosing MODIS COT observations.
- Line 267: The title of Figure 8 panel (a) says it is MODIS True color, but the text states it is the 11um brightness temperature.
- Line 65: Is -> It
Citation: https://doi.org/10.5194/amt-2023-49-RC2 -
AC2: 'Reply on RC2', Rick Schulte, 26 May 2023
We thank the reviewer for their constructive comments. Reviewer comments are in italics, and our responses in regular font.
Line 210: The SLNPW cloud fraction shown in Figure 3 represents the SLNPW cloud fraction of CloudSat pixels, not CALIOP or MODIS cloud fraction. The title of Figure 3 and the text in the manuscript could be modified to explicitly state that the cloud fractions of MODIS or CALIOP are CloudSat pixels-based cloud fractions. Additionally, including the actual MODIS and CALIOP SLNPW cloud fractions (or the total cloud fraction), would provide the reader with a comparison of MODIS and CALIOP cloud fractions to the CloudSat pixels-based cloud fraction.
This is a good clarification, and we have adjusted the text (Lines 210-211; 214-215) and the title of Figure 3 to clearly state that what is shown is the fraction of CloudSat pixels with clouds, not cloud fraction at CALIOP or MODIS resolution. Because the SLNPW classification as we have defined it is only possible at the CloudSat pixel level, we have refrained from adding any additional plots of MODIS or CALIOP cloud fraction.
Line 258: The MODIS-based sub-adiabatic LWP estimations show a positive bias in Fig. 7 panel (d), which could be attributed to the COT biases in the MODIS observations. Perhaps the authors could elaborate on the criteria used to choose the best COT observations from MODIS (e.g., flags) and consider the solar zenith angle when choosing MODIS COT observations.
The LWP estimates (in panel b) are actually almost total unbiased with respect to the RVOD estimates. What panel (d) shows is that the MODIS-based maximum LWC (i.e., the LWC in the radar bin at or near the top of the cloud) is positively biased. This has to do with the tendency of the subadiabatic model to make clouds that are thinner (with a steeper growth rate in LWC) than those retrieved by RVOD, as discussed in Lines 259-264. Nevertheless, it is true that MODIS COT observations have been shown to be biased at high solar zenith angles. We plan to implement a correction for this (as suggested in Lebsock and Su 2014, doi: 10.1002/2014JD021568) in the next reprocessing of the algorithm. However, as the focus in this paper is on warm clouds, most of the pixels considered are at low latitudes and so any solar zenith angle effects are small.
Line 267: The title of Figure 8 panel (a) says it is MODIS True color, but the text states it is the 11um brightness temperature.
MODIS true color is correct. The figure description in the text (Line 271) has been updated.
Line 65: Is -> It
Fixed as suggested.
Citation: https://doi.org/10.5194/amt-2023-49-AC2
Richard M. Schulte et al.
Richard M. Schulte et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 281 | 77 | 14 | 372 | 3 | 4 |
- HTML: 281
- PDF: 77
- XML: 14
- Total: 372
- BibTeX: 3
- EndNote: 4
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1