This paper describes the third edition of the third edition product (CLARA-A3), including surface albedo, surface radiation, and TOA radiation budget products ranging from 1979 to 2020. Various validations for cloud and radiation products were done, and show a good agreement with other or ground-based measurements. Overall, this manuscript is clear. However, there are several issues that need to be taken care of before this paper becomes acceptable for publication.

Specific comments:

1. L285, Figure 2. The cloud fraction from CLARA-A3 and previous version CLARA-A3 show a large difference, especially overestimation of cloud fraction, why?
2. How about the cloud fractions over polar regions? I am curious about these products’ performance over polar regions between snow/ice and cloud detections.
3. Why are there many invalid values of surface albedo in Eastern China?
4. L610, Figure 10. For global mean flux, the CLARA-A3 RSF is close to CERES-SYN, both underestimation of CERES-EBAF, why? And the ISCCP-FH show totally overestimation.
5. The algorithm of radiation updated from CLARA-A2 to CLARA-A3 is not clear, authors should present more details in this paper.
6. The algorithm of instantaneous radiation converting to daily mean is very important (Eq. 1), did the authors try another interpolation algorithms? Such sinusoidal fit, or Wang & Pinker method?
7. How did the authors get daily mean longwave radiation at the surface from instantaneous data?
8. t is recommended to introduce the current status and characteristics of relevant cloud and radiation products in the introduction of the paper, and highlight the advantages and characteristics of the new product (such as long time series characteristics and fine cloud and radiation parameters). Related products including JAXA's GCOM-C product (https://suzaku.eorc.jaxa.jp/GCOM_C/) (Nakajima et al., 2019), CARE (http://www.slrss.cn/care/) cloud characteristics and radiation products (Letu et al., 2020, 2022; Xu et al., 2022). This product is based on the latest geostationary satellites spliced together to form a high spatiotemporal resolution and high-precision remote sensing product. 

Nakajima, T. Y., Ishida, H., Nagao, T. M., Hori, M., Letu, H., Higuchi, R., ... & Yamazaki, A. (2019). Theoretical basis of the algorithms and early phase results of the GCOM-C (Shikisai) SGLI cloud products. Progress in Earth and Planetary Science, 6(1), 1-25.

Letu, H., Yang, K., Nakajima, T. Y., Ishimoto, H., Nagao, T. M., Riedi, J., ... & Shi, J. (2020). High-resolution retrieval of cloud microphysical properties and surface solar radiation using Himawari-8/AHI next-generation geostationary satellite. Remote Sensing of Environment, 239, 111583.

Letu, H., Nakajima, T. Y., Wang, T., Shang, H., Ma, R., Yang, K., ... & Shi, J. (2022). A new benchmark for surface radiation products over the East Asia–Pacific region retrieved from the Himawari-8/AHI next-generation geostationary satellite. Bulletin of the American Meteorological Society, 103(3), E873-E888.

Ri, X., Tana, G., Shi, C., Nakajima, T. Y., Shi, J., Zhao, J., ... & Letu, H. (2022). Cloud, Atmospheric Radiation and Renewal Energy Application (CARE) Version 1.0 Cloud Top Property Product From Himawari-8/AHI: Algorithm Development and Preliminary Validation. IEEE Transactions on Geoscience and Remote Sensing, 60, 1-11.

The authors describe an updated AVHRR-based cloud, radiation, and albedo record CLARA-A3. The AVHRR instruments have made observations from low Earth orbit since the late 1970s and provide a valuable cloud-climate record for the scientific community. However, there are many problems with this instrument: poor calibration, changing instrument characteristics over time (SRFs, channel choices, etc.), challenges with orbital drift, and sparse coverage of the solar and thermal plank functions, to name a few. The authors have been working on this observational record for a long time and should be commended for getting the most out of it. CLARA-A3 has several additional variables compared to the previous version, and the variables common across earlier CLARA versions have shown improvement with respect to other passive and active satellite records. My comments and suggestions regarding the manuscript are fairly minor but should be considered in a revision by the authors.

Throughout the manuscript: At times the use of English can be fairly colloquial and could be tightened up.

L44: extra parenthesis

L46: probably should distinguish between individual VIIRS and MODIS records, and the VIIRS+MODIS continuity product, there should be more recent references for the latter

L79-83: The authors do point out that the reader could go to a validation report for more information. But it would be helpful to explicitly describe the comparisons against CERES fluxes. Since CERES has broadband channels, there are very few photons missed, whereas AVHRR has discrete SRFs and does not measure all frequencies. Also, AVHRR shows really nice spatial structure. It would be worth detailing why one would want to do this with AVHRR data in the first place.

L91: period at start of sentence

L91: ‘has been measuring’ should be e.g. ‘has been making measurements’

L131: period and comma next to each other

L160: it should be made clearer that CLARA-A3 is Level 3, but what is Level 2b? Or is CLARA-A3 also Level 2b? This needs clarification.

L173-174: prepared separately from the rest of the globe? It would be better to simply say “This latitude band is prepared with gridding strategy A, while that latitude band is prepared with another gridding strategy B.” or similar.

L175: why does the sentence end with ‘here’?

L188: also, clause after comma starts with ‘here’, not sure what is meant

L189: this begs the question as to why are some variables considered “products” while other variables are considered “extra data layers”? What is the difference? Why not report everything as “products”?

L191: by implication, a CGT implies a cloud base height (CBH). When researchers use this product, they may very well use it to infer a CBH. Have you examined CGH/CBH?

L194: probably should clarify with ‘…separately for both liquid and ice clouds’. Is this correct?

Same paragraph: May want to cite documentation or papers for MODIS L3 gridded products that provide similar histograms

L220: what is the definition of ‘moisture content’? Vapor only? In the column? Or does it include the sum of vapor, cloud water content, and precipitation?

L232: ‘instead’ rather than ‘in its stead’

L238: Why 231K (-42C) instead of 233K (-40C)? Why 265K (-7C) rather than some other threshold? Can you add a reference(s) to justify these choices?

L253: I did not see any presentation or discussion of uncertainty estimates in this paper. Are we talking about uncertainties in the ancillary data that are somehow utilized to produce the products in Table 2? Or are these uncertainties on the products themselves reported in Table 2? In either case, these should be described, cited, shown in a figure, their methodological approach described and cited, and summarized in some way that demonstrates their use.

L264: do not need ‘in addition’ at the end of the sentence

L279: the phrasing is a bit awkward. Is this supposed to mean ‘more thin cloud with a COT < 0.3 is detected’ or ‘more thin cloud with a COT > 0.3 is detected’?

L292: ‘significant trends’ implies that a formal significance test in the trends was performed. Do the authors mean that there are no trends with 95% significance? How did you do this test or conclude this result?

Figure 4: Why does MODIS Aqua show a decrease in liquid cloud? Is this a calibration issue or a physical change in cloud properties?

L330: Why specifically should we care to read the report? It would help to explain what is in it at a very high level. It is not useful to tell the reader that they need to read another report or series of papers to get the gist of the current paper. The information should be self-contained in some manner.

Same part of manuscript: What about trends in IWP, LWP, and effective radius for liquid and ice clouds? Have these been examined?

L340: spell out Table

L361: look-up table should be spelled out on first use. Also, need to add space between ‘LUT’ and ‘that’

L363: add Hersbach reference for ERA5. Also, define ERA5T. Is this surface temperature? Or something else?

Paragraph starting on line 380: It is unclear what we should take home from the use of a monthly climatology for aerosol averaged over many years. How does this impact the interpretation of the solar products in CLARA-A3 on a daily, instantaneous basis? Relative to the total magnitude of the value of the products, how large is the contribution from aerosol? Less than a few %? Greater than that by a little, or a lot?

L391: end of sentence, also should be noted that it does not contain day-to-day variability, which conceivably can be significant

L406: ‘The grid boxes…’

L407: end of sentence repetitive

L419 and L422: Is ERA5T the same as ERA5(T)?

L530: ‘was very good’ is not entirely useful or quantitative

L562: how many ADMs are used?

Figure 10: The fact that ERA5 and CLARA-A3 track so closely at short and long temporal scales could be interpreted in a way that suggests CLARA-A3 depends heavily on ERA5 ancillary inputs, and that you are basically making another version of ERA5. Is this fair to say or is there a clear advantage to using AVHRR data for the solar and thermal flux and irradiance products? This gets back to the earlier comment about why are these products being generated in the first place? What is the added value over other climate records?

L619: did not see Canty et al in reference list

Figure 11: Why show MAB instead of actual values of OLR? It would be useful to consider showing OLR in the same way as you did with fluxes of RSF in Figure 10.

L707: ‘It is worth…’

L711: rogue letter ‘t’

L733: If the effective radius is not shown in the paper, at least point to this result in the validation report or elsewhere. This is interesting but the microphysics haven’t been shown in the paper.

The authors present some results from the CLARA-A3 validation against other similar cloud and radiation products. Production of this dataset is a significant accomplishment and an important addition to global climate studies of cloud and radiation. It is also clear from the link in Section 5 that there is a great deal of documentation and validation supporting this paper, which therefore necessitates selection of a few ‘highlights’ to be included here. That said I believe there are areas that require more detail in order for this to be considered useful as a stand-alone article.

Minor:

L114: ‘Full’ shouldn’t be capitalized.

L264: ‘in addition’ is repetitive.

L361: There should be a space between ‘LUT’ and ‘that’

L566: ‘then’ should be capitalized.

L712: There is an extraneous ‘t’ after CLARA-A3.

Specific comments:

L180: My understanding is there is no attempt at multilayer detection, and the cloud top phase is assumed to extend throughout the cloud column when deciding whether to calculate IWP or LWP. In that case I assume there is additional uncertainty in IWP estimates where it is unknown when ice overlaps water? It would be nice to know how uncertainty is estimated for these products and whether multi-layer clouds translate to higher uncertainty.

L252: The description of CPH, COT, and CRE are very brief, and we don’t see any figures showing results from these products. Have the updates listed here resulted in lower uncertainty, or simply higher confidence in the reported uncertainty? Some numbers describing the changes in uncertainty for these products would be helpful.

Figure 2: There is a great deal of overlap making it difficult to differentiate among the records. Perhaps an anomaly plot would help? If it were deaseasonalized we could also determine whether the records showed different trends over time.

L292: Was a statistical test applied to the CTP records in Figure 3 to deduce there are no significant trends? If so could that information be included here?

Figure 4: Is the liquid cloud fraction a percentage of total cloudiness phase (i.e. liquid / liquid + ice)?

L380-391: I find this paragraph confusing. When are the instantaneous versus daily irradiance calculations used? Is one for Level 2 and one for Level 3? Also an aerosol record that extends from 1979-2025 is used, but only to create a 12-month climatology? Is this also the case for atmospheric gases? For example is CO2 rise accounted for?

L592: How is the instantaneous OLR estimated for the AVHRR/1 where channels 4 and 5 are the same?