the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 16 Jan 2026
Three decades of sea level multi-mission satellite data reprocessed to improve mesoscale quality while ensuring climate scale consistency
Abstract. Since the launch of TOPEX/Poseidon in 1992, more than 15 satellite altimetry missions have gathered measurements of ocean surface topography. These observations contributed to significant advancements in our understanding of ocean dynamics in the open ocean, coastal and polar areas and at scales ranging from 10 km and a few days to global averages over decades. Heterogeneity across missions and long update cycles of altimeter instrument processing facilities remains a challenge to assemble a multi-mission, consistently processed, state-of-the-art dataset serving the needs of various user types from data assimilation into ocean circulation models to climate science.
In this context, the Delayed Time DT-2024 satellite altimetry reprocessing is a massive endeavor spanning over more than a 100 years' worth of data, from 3 decades, 15 satellites, and 5 climate reference altimeters. In our effort to enhance the user-oriented "reliability" of sea level measurements, we focus on the refinement of altimetry satellite standards (radar processing algorithms and geophysical models) and on the cross-mission consistency. Reliability is here treated as a multi-dimensional spectrum encompassing coverage, precision, accuracy, and stability. These four pillars are essential, not only to capture short-term ocean variability (large and small eddies) for the open, coastal and polar oceans, but also to detect seasonal and long-term climate signals such as global mean sea level rise.
The DT-2024 standards introduce new radar processing algorithms and geophysical corrections. In coastal areas, the error is reduced by 5,6 cm² (or 17%), enhancing monitoring of applications as storm surges and upwelling. In polar regions, error reduction of variance at crossovers exceed 7,7 cm² (or 27%) in the Arctic and 5,9 cm² (or 18%) in the Antarctic, potentially enhancing observation of freshwater fluxes and circulation around ice-covered zones. In open ocean, the error of sea surface height variance at crossovers is reduced by 1,2 cm² (or 6%). Although the contributions of each standard are relatively balanced in the open ocean, the significant improvements observed in coastal and polar regions are largely attributable to the FES-22B tide model, which alone contributes approximately 70% of the gains in these areas. Additional gains come from the TUGO atmospheric correction (forced with ERA5), the CLS/DTU/SIO Hybrid-23 mean sea surface and updated instrumental corrections. These refinements could potentially aid in the detection of mesoscale features, contribute to the assimilation of data into ocean models, and offer insights into dynamic processes such as fronts and internal tides. These improvements stack up with similar gains from previous reprocessing campaigns (e.g. DT-2021), highlighting a continuous progress made in satellite altimetry with every reprocessing cycle since the nineties.
Ensuring the stability of sea level measurements is also crucial for accurate climate monitoring and analysis (Cazenave et al., 2019, Meyssignac et al. 2023). To support climate applications, the DT-2024 definition was rigorous because new algorithms could affect the sea level trends of the multi-mission dataset. To ensure the accuracy and consistency of sea level measurements over time, we align all coverage and precision missions on a unified reference frame based on the reference climate altimeter series. The reference altimeters are extremely consistent with one-another thanks to their so-called tandem phases (formation flight) although we compute the static offsets between subsequent reference altimeters to reduce residual offsets (and to estimate the uncertainty of the transition between reference altimeters). Comparisons with independent in-situ tide gauges yield an agreement within 0.01 mm/year for the regions where tide gauges are located.
The DT-2024 Level-2P dataset is available to users for all the altimeter missions on AVISO+ (https://doi.org/10.24400/527896/a01-2025.004). The Level-3 and Level-4 counterparts are available on the Copernicus Marine Service catalogue (https://marine.copernicus.eu) and C3S (Climate Data Store).
- Preprint
(3648 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on essd-2025-604', Anonymous Referee #1, 06 Mar 2026
-
RC2: 'Comment on essd-2025-604', David Griffin, 15 Jul 2026
This paper documents a huge and complicated data-reprocessing effort. The result of this work is a data set that a large number of people will use (until the next reprocessing is finished). I commend the team of authors for doing a thorough piece of work. For conciseness, many details are left out. I think this is acceptable, and consistent with the authors' choice to not dwell on why the 'chosen' corrections were adopted in favour of others. Some people will disagree with some of these choices, in which case they can simply use a rival correction for their own work. In this context, I am happy with the paper essentially as it is, and only offer a relatively small number of fairly minor comments, or technical errors, as follows.
line Comment/question
78 what are JA2 and JA3 (cf j2 and J3 line 70)79 lower (Climate)
96 "20211" (a typo)
116 were -> was
119-120 "few h or few d" which? For tides, IB, SSB, need h
129-130 "to minimise" How does this minimise?
136 Huh? Geoid is not used for SSHA variance (see footnote p3)
145 & 216 consision -> conciseness
158 "mea sea surface"
176 & 936 "Salvatore et al. 2024". His name is Salvatore DINARDO. (I have not checked all references)
239-244. Here, I get confused what you are comparing to. Are you distinguishing the 'ERA5 meteorological model' (used by DT2021) from the 'reanalysis' (used by DT2024?)
p13 Footnote 6. This is a very good point. I was wondering how you would raise it. Its like a feedback loop, with potential for error growth but hopefully the series converges (as we all expect). I don't think important points belong in footnotes. Can it have its own paragraph? "Some may fear that the MSS, for example, is a product that is both an input to the process and an output. To reassure ourselves that the various product update cycles take us ever closer to the 'truth' rather than away from it, we ….."
Fig 5 Caption. 50 to 66 degrees is no man's land. Or typo?
372-377 I think this depends on how far away from exact crossover points you count as a match.
447 Typo
458 If you have defined 'coverage mission', I missed it. I assume you mean non-reference. But geodetic phase as well?
464 'climate-oriented altimeters' Too many names! I vote to use 'reference' everywhere.
Figure 10 "Note that for J3 vs. S6A, the dynamic atmospheric correction is not applied"
Explain (briefly) why not. Someone skipping through the paper will be surprised to read this.672 "0 to 5days" disagrees with line 129 and 443 (10 days). I think 10 days is too long but I guess its too late to do 5days. Or is there no disagreement? Ie, if a matchup is within a 10-day window: -5 to +5. In which case, this should be more clearly stated since it is a key choice.
Citation: https://doi.org/10.5194/essd-2025-604-RC2 -
AC1: 'Reply on RC1', Cécile Kocha, 10 Sep 2026
REVIEWER 1
Overview
This paper details the extensive efforts by space agencies to enhance the sea level data recording terms of precision, accuracy, continuity, and stability. Specifically, it highlights the advancements integrated into the DT2024 release, focusing on the updated altimeter standards used to calculate sea level anomalies. Furthermore, the study describes the cross-calibration methodology employed to harmonize data across 15 missions, ensuring a seamless 30-year record. While this review of the DT2024 improvements is a vital resource for the user community, there are several areas where the manuscript could be strengthened, as outlined in the major and specific comments bellow.
Response
We thank the reviewer for this constructive, thorough and encouraging evaluation. We have carefully addressed all comments and have revised the manuscript accordingly to improve clarity, completeness, and the scientific value of the presented DT2024 assessment.
Major comment 1
The general writing related to metrology aspects needs to be significantly more rigorous. The authors frequently and incorrectly refer to the "reduction of errors" when they should be referring to the "reduction of the variance of the errors" (see specific comments). Furthermore, the adapted word for describing the variance of errors, "uncertainty," is never used whereas it will provide a better clarity to analyse the improvement of the DT2024 standards. Additionally, the description of the different temporal scales of the errors could be improved (e.g. short-term time-correlated versus long-term time correlated errors). Improving the rigorous wording to describe how the uncertainty (variance of the errors) is reduced will promote a fluent reading of the paper and yield a more rigorous scientific document.
Response
Throughout the revised manuscript, we carefully reviewed the wording concerning errors depending on the context. In addition, Section 2.1 now explicitly explains what crossover statistics measure and clarifies that crossover analyses primarily provide an estimate of changes in the variance of measurement errors on temporal scales shorter than approximately 10 days.
We also expanded the discussion concerning short-term and long-term time-correlated effects to improve consistency with current uncertainty-budget frameworks used in sea-level climate studies.
Major comment 2
The current text frequently contrasts "long-term stability" with "mesoscale" improvements (referring to the enhancement of new DT2024 ). However, as demonstrated in Meyssignac et al. 2023 ("How accuracy is accurate"), all correlated effects across different time scales contribute to the stability uncertainty. Therefore, for the sake of clarity and exactitude, it would be more appropriate to differentiate the DT2024 standard improvement based on the: 1) short-term time-correlated effects (shorter than a few months): these effects primarily contribute to improving mesoscale features; 2) larger time-correlated effects (beyond 1 year, for instance): both short-term and larger time-correlated effects contribute to improving the stability for climate studies. This separation would clearly articulate how different temporal scale improvements impact both mesoscale features and long-term stability claims.
Response
The manuscript has been revised to better distinguish between short-term time-correlated effects (days to months), which mainly affect mesoscale observability and crossover variance, and long-term time-correlated effects (annual to decadal scales), which are more directly related to climate-oriented stability metrics. We also clarify that improvements at short temporal scales also contribute to the overall sea-level uncertainty budget and therefore indirectly contribute to long-term stability.
The methodology section has been revised accordingly.
Major comment 3
In the paper, the authors emphasize the need for homogeneous corrections across all altimeter periods for several corrections (e.g. DAC). However, in the section concerning the wet troposphere correction (WTC), they do not explicitly address the choice of using different WTC corrections, derived from MVR or other approach (GPD+). It would be highly beneficial to explain and justify the decision to use inhomogeneous WTC corrections. Are there any other alternatives ? This explanation should adress how continuity is ensured when different WTC corrections are applied, and so assess the potential impact on mean sea level stability.
Response
We thank the reviewer for highlighting this important point. Our objective is not to apply identical corrections to all missions, but rather to use the most appropriate correction for each mission while preserving the consistency of the overall climate record.
Radiometer-derived wet tropospheric corrections remain the preferred solution whenever available, owing to their direct collocation with the altimeter measurements and the dedicated calibration and validation procedures developed for climate applications. This approach is applied consistently across the climate reference missions (TOPEX/Poseidon, Jason-1, Jason-2, Jason-3, and Sentinel-6), thereby ensuring homogeneity within the climate data record.
When available, alternative radiometer-based solutions can further improve the wet tropospheric correction, particularly when radiometer measurements are affected by coastal contamination, rain effects, sea-ice contamination, or instrumental limitations. In particular, the GPD+ and Neural Network corrections provide improved performance for both mesoscale applications and long-term stability assessments. Importantly, these corrections remain fundamentally based on radiometer observations, allowing us to preserve the homogeneity of the correction approach across missions while benefiting from enhanced accuracy.
Only when radiometer measurements are unavailable or of insufficient quality, and when no satisfactory radiometer-based retrieval can be obtained, is the model-based HRES correction used as the best available alternative.
To clarify this rationale, Section 2.4.3 has been expanded to provide a more detailed description of the correction strategies adopted for the different missions and of the methodological choices and associated trade-offs considered to maintain a consistent climate record.
Major comment 4
While the method used to assess the reduction of variance for short-term errors (less than 10 days) is clearly described and applied systematically to the new DT2024 standards, the approach used for longer timescales (months to a decade) is poorly documented. Additionally, there is no quantification of the improvements in continuity and stability compared to previous versions, despite the authors emphasising the importance of their work with regard to continuity and stability throughout the paper. A more detailed description of the methodology and quantification of these improvements would give users much greater confidence in using DT2024 for climate change studies.
Response
We agree that the methodology used to assess the impact of the different candidate solutions should be described more explicitly. We have therefore expanded Section 2.1 (Methodology to select new DT-2024 standards) to include a dedicated description of the approach used to quantify changes in regional and global mean sea level trends.
Furthermore, the preservation of the long-term stability of the sea level record is assessed through comparisons with independent tide-gauge observations. Although tide gauges do not represent an absolute truth and provide an independent observing system with their own uncertainties, they offer a valuable external reference for assessing the consistency of the reconstructed sea-level trends and detecting potential residual drifts in the altimetry record.
Abstract
Line 35: the current wording, "the error is reduced by 5,6 cm² (or 17%)", is imprecise. It is not the error itself that is reduced, but rather its variance. Therefore, the statement should be corrected to: "The variance of the error is reduced by [...]". Alternatively, given that uncertainty reflects the statistical behavior of the error, an acceptable phrasing is: "The uncertainty is reduced by [...]".
Response
The sentence has been revised to be more precise: "The DT-2024 standards introduce new radar processing algorithms and geophysical corrections. In coastal areas, the uncertainty is reduced by 5,6 cm² (median of variance at sea surface height crossover reduced by 11,2 cm² or 17%),"
line 41 : Is “gain” means “uncertainty reduction” here ?
Response
The term "gain" was ambiguous. We replaced gain by uncertainty reduction : "approximately 70% of the uncertainty reduction in these areas."
lien 55: The statement, "Comparisons with independent in-situ tide gauges yield an agreement within 0.01 mm/year for the regions where tide gauges are located," raises a question: is 0.01 mm/yr the correct value, as it seems exceptionally low? Furthermore, what is the associated confidence interval?
Response
The value corresponds to the residual trend difference observed between the altimetry and tide-gauge estimates over the regions sampled by the tide-gauge network shown on fig 14. To avoid ambiguity, we provide the associated confidence interval in the revised manuscript.
Introduction
line 80 : The statement that "80 precision [is] considered a secondary benefit" is, in my view, inaccurate. The uncertainty budget, as detailed in Guerou et al. (2023) and previously Ablain et al. (2019), clearly indicates that short-term time-correlated effects, those lasting less than one year, also contribute to the uncertainty in sea level stability. This is a point highlighted in Meyssignac et al.'s 2024 work, "How accuracy is accurate?".
Response
We agree that short-term time-correlated errors also contribute to the uncertainty budget of long-term sea-level stability, as discussed by Ablain et al. (2019), Guerou et al. (2023), and Meyssignac et al. (2024).
Our intention in this paragraph was not to imply that precision is irrelevant for climate-oriented missions, but rather to emphasize that these missions are primarily designed to achieve long-term stability and overall measurement accuracy. In our view, the contribution of short-term time-correlated errors to stability uncertainty is already encompassed within the broader concept of measurement accuracy.
To avoid any misunderstanding, we have revised the sentence to better reflect the complementarity between precision, accuracy, and stability. The revised text now reads:
"Climate-oriented missions are primarily designed to optimize long-term stability and measurement accuracy, while also maintaining a high level of precision, which contributes to the overall uncertainty budget of sea-level estimates."
We believe this revised wording better captures the concepts discussed in the cited literature while preserving the intended distinction between reference climate missions and missions primarily optimized for spatial and temporal sampling.
line 106 : To clarify “which now serves as the new regional reference?”
Response
To avoid introducing a concept that is not defined at this stage of the manuscript, we have simplified the wording and removed the term "regional" from the Introduction.
The sentence has been revised accordingly for clarity. The specific role of Sentinel-6 MF as the reference mission used in the regional cross-calibration framework and bias estimation is described in detail later in Section 4 (Cross-calibration methodology), where the corresponding methodology and rationale are introduced and discussed.
Methodology to select new DT24 standards
line 118 : The reduction of SSHA variance at crossover points should be explicitly mentioned, as this is a useful detail.
Response
The revised text now explicitly states that the metric used is the reduction of the variance of SSHA differences observed at crossover points.
line 123 : ”The sentence "changes in the SSHA difference at crossover points is a partial measurement of the altimeter errors" needs clarification. Specifically, explain what is meant by "partial measurement" and "altimeter errors."
Response
By “altimeter errors”, we refer to the residual measurement errors affecting the SSHA estimates after application of all geophysical and instrumental corrections, including instrumental noise, residual orbit errors, imperfect geophysical corrections, and other uncorrected or mismodelled error sources. By “partial measurement”, we mean that crossover differences do not provide a direct estimate of the total altimeter error budget. Instead, they are sensitive only to the error components that contribute to discrepancies between the two SSHA measurements acquired at the crossover point. In particular, error components that are common to both measurements may cancel out, while the observed crossover difference also contains a contribution from true ocean variability occurring between the two acquisition times. Under the assumption that measurement errors are independent from the geophysical signal of interest, changes in crossover variance primarily reflect changes in the residual altimeter error variance. We have revised the text to make this point clearer.
"The difference in the two SSHA measurements on the crossover point is the sum of the natural ocean variability (e.g. displacement of ocean eddies) and altimetry measurement errors (e.g. biases from imperfect geophysical corrections and instrumental noise). Assuming that altimeter errors are independent from the signal of interest 2, changes in the SSHA difference at crossover points is a partial measurement of the altimeter errors. This estimate is only partial because some error components common to both observations cancel out and while the crossover differences also contain a contribution from the natural ocean variability occurring between the two acquisition times."
Line 124 : The sentence, "which is a good indicator of the gain or loss in altimeter errors," needs rephrasing for meteorological accuracy. The word "indicator" is likely inappropriate; "estimate" might be a better choice. Furthermore, it would be more precise to refer to the "gain or loss in the variance of altimeter measurement errors."
Response
The sentence have been replaced to reflect the suggestions : " which is a good estimate of the gain or loss in the variance of altimeter measurement errors".
line 126 : “crossover points evaluate the coherence between ascending and descending ground tracks” => “coherence” should be likely replaced by “consistency”
Response
"Coherence" has been replaced by "consistency".
line “127” : The phrase "or the temporal variability of the errors" should be revised. It is essential to specify that this refers to the error differences, as correlated errors between ascending and descending tracks are effectively cancelled out.
Response
The sentence now explicitly indicates that crossover statistics characterize the temporal variability of measurement-error differences, while error components common to both measurements are cancelled : "the temporal variability of the errors differences, as correlated errors between ascending and descending tracks are cancelled out."
line 129 : I do not understand the sentence "minimize the impact of residual systematic errors." Are the authors intending to say, "to cancel the effect of time-correlated errors larger than 10 days"??
Response
We have revised the sentence to clarify that crossover pairs are restricted to less than 10 days because the sea level signal is expected to remain relatively stable over such a short time interval (except during extreme events), allowing crossover statistics to better characterize short-term measurement uncertainty and to compare different satellite altimeter data whereas longer-term correlated effects are assessed separately through trend and continuity analyses.
line 140-142 : Please provide more detailed information on the methodology developed to assess the stability uncertainty of reference altimetry missions using the new altimeter standards. This description should be similar in depth to the one provided for the SSHA crossover approach, as the current description is insufficient. For instance, how is the continuity of the sea level data record assessed when transitioning between two altimeter satellites?
Response
The comprehensive estimation of stability uncertainties for the climate altimeter record is an important topic that involves a complete uncertainty-budget analysis across a wide range of temporal scales. Such an assessment is beyond the scope of the present manuscript and is the subject of ongoing work that will be presented in a dedicated paper (Quet et al. 2026).
The objective of the present paper is more specifically to assess whether each candidate standard introduces significant changes in regional or global mean sea level estimates and whether such changes are compatible with the current understanding of the observing system.To clarify this point, Section 2.1 has been expanded to describe the methodology used to monitor the impact of each candidate standard. For each processing change, we systematically analyse: (i) differences in GMSL and RMSL trends, (ii) the presence of geographically coherent patterns, offsets, or drifts, and (iii) the consistency of the observed changes with the expected behaviour of the correction and the scientific literature. The selection strategy adopted for DT-2024 is conservative: standards are introduced only when the observed changes are physically understood, documented in the literature.
Regarding continuity between successive reference missions, continuity is primarily assessed using the tandem phases between consecutive reference missions. During these phases, both satellites observe nearly the same ocean surface under very similar conditions, allowing the estimation of inter-mission mean sea level offsets and the evaluation of potential residual error pattern. These inter-mission adjustments constitute the basis of the reference-altimeter alignment described in detail in Section 4.2.
Overview of the DT-2024 standards
lien 160: The term "stability" (in orange) in Figure 1 is inappropriate, as stability refers to both short-term and larger, time-correlated effects (see the general comment). The orange color range in Figure 1 specifically represents the reduction in uncertainty resulting from "large-term and linear time-correlated effects," which naturally enhances stability. Nevertheless, stability is also improved by almost all the other effects at shorter time salce detailed in the green boxes.
Response
We agree that stability may be influenced by error sources acting over a wide range of temporal scales. The original color coding was intended to highlight corrections whose most visible impact is on low-frequency signals, long-term trends, and inter-mission consistency.
To avoid ambiguity, the caption of Figure 1 has been revised. The wording now explicitly refers to corrections that primarily affect "large-scale stability", while acknowledging that improvements at shorter temporal scales also contribute to the overall sea-level uncertainty budget.
Altimeter input dataset reprocessing
line 175 : The authors highlight the improved stability of global mean sea level measurements achieved by implementing the newly developed ocean numerical retracker into the S6A operational processing chain, compared to the historical MLE-4 retracker. While this is a highly relevant finding, it introduces an undocumented source of uncertainty in previous altimeter measurements that relied on the classical MLE-4 (the most recent publication on this is Guerou et al., 2023). Provide a comment on this specific point.
Response
The improvements brought by the Numerical Retracker provide new insight into some limitations of the historical MLE4 approach, particularly regarding the representation of the instrument Point Target Response (PTR) and the resulting sensitivity to sea-state conditions.
However, we would not interpret these results as evidence of a previously undocumented source of error affecting the entire historical altimeter record. Rather, the Numerical Retracker reduces a source of uncertainty that was already known and partially mitigated through calibration activities and correction tables. Its main benefit is to provide a more physically consistent representation of the instrument response by directly accounting for the in-flight PTR evolution, thereby reducing the dependence of sea surface height estimates on sea-state conditions and improving the long-term consistency between Jason-3 and Sentinel-6.
line 183 : The authors claim a 60% reduction in the sea surface height anomaly bias between Sentinel-6 MF and Jason-3 when using a numerical retracker. Clarify what "bias" refers to here—is it the Global Mean Sea Level (GMSL) offset between Jason-3 and Sentinel-6 during a tandem phase? Also, is the 60% reduction applied to the value of the bias itself, or to the variance of the bias?
Response
The term bias does not refer to the GMSL offset between Jason-3 and Sentinel-6, nor to the variance of the bias. It refers to the dependence of SSH differences on Significant Wave Height. The reported 60% reduction therefore corresponds to a reduction of the amplitude of this sea-state-related bias.
lien 202 : The statement regarding the potential drift of the TOPEX-A GMSL with the new TOPX GDR-F is overly cautious. Ongoing studies clearly demonstrate the existence of a drift with independent method (e.g; see Bouih et al, OSTS 2025). It would be more accurate to state that a drift has very likely been detected with the GDR-F, but the corresponding correction has not yet been implemented.
Response
We agree that recent studies provide increasing evidence supporting the existence of a residual drift in the reprocessed TOPEX-A record. The wording has been revised accordingly.
Ocean Tide Correction
line 221 : “spatial resolution improved by a factor of 8” => what is now the new spatial resolution?
Response
The factor-of-8 improvement refers to the resolution of the underlying finite-element model mesh in coastal and high-latitude regions rather than to the resolution of the distributed gridded product. To avoid ambiguity, the manuscript has been revised to provide the corresponding spatial scales : "with a refined finite-element mesh of 4 km resolution in coastal regions reaching locally 500 m and until 30km offshore"
line 229 :The sentence "A decrease in variance indicates better signal consistency, i.e. a reduction in bias or noise errors from tides residuals" is inaccurately formulated. A decrease in variance specifically indicates a reduction in the variance of uncorrelated errors (with periods less than 10 days) observed at SSHA crossovers. Stating that the "error is reduced" is not strictly correct, as variance is a measure related to probability.
Response
The sentence has been revised to specify that crossover analyses quantify reductions in uncertaintes from tides residuals.
line 230 : Similar to the previous point, reword "The error at crossovers is reduced" to "The variance of the errors (or uncertainties) is [...]."
Response
The sentence has been revised accordingly.
line 234 : In the sentence : ”Furthermore, ocean tide impacts the global trends by less than 0.06 mm/year “ , what mean by “global trends” ? is it the GMSL trend ? How is obtain the number of 0.06 mm/yr ? Is the GMLS trend diffrences between 2 ocean tide models over each altimeter period ?
Response
The manuscript now specifies that the quantity corresponds to differences in Global Mean Sea Level (GMSL) trends obtained when applying different ocean tide models over the reference mission periods.
The methodology used to derive the reported value has also been added.
Line 234 : In teh sentence “By design of the correction, ocean tides should barely impact long-term trends” , “long-term trends” could / should be replaced by “stability”
Response
The sentence has been revised accordingly.
Line 235-237 : Ray and Schindelegger (2025) suggest that ocean tide amplitude may change over time due to variations in ocean mass. While I understand this, the expected impact on regional or global Mean Sea Level (MSL) stability and trends seems potentially negligeable. Please, expand on this section to better explain the potential influence of evolving ocean tides on regional or global MSL?
Response
The direct impact of evolving ocean tides on present-day global mean sea level trends is likely negligible. Our intention was rather to highlight that the estimation of long-term tidal trends can be affected by systematic effects, including leakage between tidal and non-tidal variability. We have clarified in the manuscript that the impact on present-day GMSL trends is expected to be negligible.
Dynamical atmospheric and dry tropospheric corrections
Line 246: The authors mentions that Using the dry tropospheric correction and the dynamic atmospheric correction based on ERA5, instead of the HRES operational model, removes a sea level anomaly regional trends up to 0.25 mm/year, and regional trends up to 1 mm/year. Please clarify the following regarding the 0.25 mm/yr value: Does this figure represent the mean or the median? What is the length of the time period over which this value was calculated? Has the impact of the GMSL trend been considered?
Response
The reported values correspond to the typical amplitude of the regional trend differences observed when comparing ERA5-based and HRES-based corrections over the Jason-2 mission period. The impact on the GMSL trend was also evaluated and found to be very small, remaining below 0.02 mm yr⁻¹.
We have clarified the text accordingly : “Using corrections based on ERA5 instead of the HRES operational model removes artificial sea level anomaly trends. For the Jason-2 mission, while the impact on global mean sea level remains below the uncertainty level, regional trend differences reach up to 0.25 mm/year for the dry tropospheric correction and up to 1 mm/year for the DAC. These artificial trends actually originated in various offsets caused by updates of the operational ECMWF model.”
Line 247: Cclarify the intended meaning of "quality" in this sentence: "Moreover, with a more recent model version, the atmospheric reanalysis significantly improved the quality of the atmospheric corrections for older missions"? Would "accuracy" be a more appropriate term?
Response
The wording has been revised accordingly.
Line 265 : same comment as in line 247 for “quality”
Response
The wording has been revised accordingly.
Wet tropospheric correction
line 274 : Clarify the statement: "The radiometer WTC is better by approximately 1 cm² for the global ocean"? Is the SSHA variance at crossovers reduced on average by 1cm² over the global ocean when using the MWR-derived Wet Tropospheric Correction (WTC) instead of the model-derived one?
Response
Yes. The statement refers to the reduction of SSHA variance at crossover points when using a radiometer-derived WTC instead of a model-derived WTC. To avoid ambiguity, the text has been revised as follows: “Compared to model solutions based on ERA5 or HRES, the radiometer WTC reduces the SSHA variance at crossover points by more than 1 cm² for the global ocean and much more in active troposphere regions. “
line 283 : In Guerou et al. (2023), it was noted that the uncertainty in the MWR-derived WTC for Jason-3 was higher due to a detected drift (also discussed in Barnoud et al., 2023). This drift has since been corrected by JPL. This raises two questions: 1) Has this correction been applied for the Jason-3 MWR to the DT2024 dataset? What is the impact of this correction on the GMSL stability?
Response
The Jason-3 MWR drift was known when the DT-2024 reprocessing was finalized. However, the corresponding correction was not yet available for operational implementation in the processing chain and could therefore not be included in DT-2024. The impact of this drift on GMSL is discussed in Barnoud et al. (2023) and Guérou et al. (2023), and future reprocessing activities will benefit from the corrected radiometer calibration as evocated in the section 6.2.
Mean Sea Surface
line 319 : The sentence "it allows us to interpret the reduction of variance as a reduction of the MSS errors" would be more accurate if the term "errors" was replaced with "uncertainty”.
Response
The sentence has been revised accordingly.
Benefits from DT-2024 upgrades for ocean mesoscale
line 353 : to explicitly mention ”SSHA error variance reduction” in the sentence “The Sentinel-3 crossover error reduction in Fig. 5”
Response
The sentence has been revised accordingly.
line 361 : to explicitly mention ”error variance reduction” in the sentence “Fraction of the sea surface height error reduction at crossover points “
Response
The sentence has been revised accordingly.
line 379 : to explicty describe what is expected by “gain”.
Response
The term "gain" is now explicitly defined as the reduction in SSHA error variance (or the equivalent uncertainty reduction) relative to DT-2021.
line 384 : what does “consistency” mean in the sentence “This indicates an improvement in data accuracy and consistency of 6% for the global ocean” ? Do the authors mean that improving the precision and accuracy of SSH data by an average of 6% in the open ocean leads to greater consistency between altimeter missions during the same periods? To clarify in the manuscript.
Response
The manuscript now clarifies that "consistency" refers to the reduction of measurement discrepancies between observations sampling the same ocean signal, as reflected by crossover statistics.
line 391 : To mention the “Temporal evolution of the difference of variance [...] ” in the legend of figure 6.
Response
The sentence has been revised accordingly.
line 401 : to improve “10% reduction in SSH crossover error“ to “10% variance reduction in SSH crossovers”
Response
The sentence has been revised accordingly.
line 421 : ”Uncertainty” is more adapted than “errors” in the this sentence “These results confirm the benefits of the new standards to reduce some error sources in coastal regions”
Response
The sentence has been revised accordingly.
line 431 : to modify “median error reduction” by “ median error variance reduction”
Response
The sentence has been revised accordingly.
line 433 : same comment as in line 431
Response
The sentence has been revised accordingly.
Cross-calibration methodology
line 443 : to improve the wording of this sentence, 'By design, crossovers are limited to errors with periods shorter than 10 days', make it more rigorous. For example, 'By design, SSHA crossovers enable the characterisation of errors with time-correlated scales shorter than 10 days.'
Response
The sentence has been rewritten following the reviewer's suggestion and is now more explicit.
line 446 : Clarify the meaning of 'unadulterated' in the sentence? Does it refer to stability?
Response
Yes, in this context unadulterated refers to the preservation of the long-term stability and integrity of the climate sea-level record. To avoid ambiguity, the wording has been revised accordingly.
line 448 : What is the difference between “reference” and “coverage” missions? I assume it is all the missions that are not "reference" missions. However, is “coverage” the right term for those missions? Perhaps 'complementary' missions is a more appropriate term?
Response
We agree that complementary missions better captures the broader scientific contribution of these missions. However, the term coverage missions was used to emphasize their main role in improving the spatial and temporal sampling of the ocean, particularly for mesoscale applications, in contrast to the reference missions that provide the stability and continuity framework of the climate record and onto which the coverage missions are aligned, particularly in the Level-3 processing.
To avoid ambiguity, we have added the following definition in the Introduction:
“To distinguish missions contributing to the reference climate record from the other missions, we hereafter refer to the former as reference altimeters and to the latter as coverage and precision missions.”
Geodetic phases are therefore included in the latter category when they are not contributing to the reference climate record.
line 452 : To better define what the author means by using the word 'align' (or 'alignment') in this section
Response
In this context, alignment refers to the estimation and application of inter-mission sea level biases in order to place all reference altimeters on a common reference framework anchored to Sentinel-6 MF. We replaced alignment in the text by “estimation and application of inter-mission sea level biases” to better define the alignment.
lien 453 : The author mentions that 'this alignment is small and straightforward'. Is it possible to provide a range of values? Is there also a reference to the method applied? Does it apply only to the global mean or to regional scales?
Response
The purpose of this section is to provide an overview of all processing steps involved in the cross-calibration strategy rather than a detailed description of each individual component. The alignment methodology, the associated values, and their application at global and regional scales are described in detail later in Section 4.2. We have added a reference to this section to guide the reader. The following subsections then revisit each step individually and provide the corresponding methodology, results, and observed values.
line 454 : The author only mentions that ocean variability is cancelled out in cross-mission differences during a tandem phase. However, all effects due to environmental corrections (e.g. ocean tides and dynamic l corrections) are also cancelled out.
Response
We agree with the reviewer. Ocean variability is not the only contribution that is largely cancelled out during tandem-phase comparisons. Since both satellites observe nearly the same ocean surface at nearly the same time, the impact of most environmental corrections, including ocean tides, DAC, and other geophysical signals, is also largely removed from the cross-mission differences. As a result, the residual differences primarily reflect measurement-related errors and calibration differences between the two missions. We have clarified this point in the manuscript. “The tandem phase cancels out ocean variability and most geophysical correction signals in the cross-mission differences,”
line 462 : What does the author mean by 'high-frequency errors'? Are they short-term time-correlated effects less than a few hours or days?
Response
The correction primarily targets geographically correlated biases between neighboring tracks, including long-wavelength residual errors from precise orbit determination and residual barotropic corrections (e.g. tides and DAC). The text has been revised accordingly by "this step mitigates long-wavelength biais from residual error in precise orbit determination, barotropic”
line 467 : What is it the “climate MSL” ? It would be more accurate to refer to the MSL time series.
Response
The sentence has been revised accordingly.
lien “469” : The author mentions 'the extreme stability of reference climate altimeters', whereas the current MSL data record does not meet the scientific stability goals defined in Meyssignac et al., 2024 ('How accurate is accurate?'). This sentence should probably be moderated.
Response
We acknowledge that current stability goals remain an active area of research. The term “extreme” may overstate the current level of stability achieved by the climate altimeter record. The wording has been revised accordingly and now refers to the “stability of reference climate altimeters”.
line 473 : What are the climate indicators mention by authors ?
Response
The manuscript has been revised to explicitly refer to sea-level climate indicators, including Global Mean Sea Level (GMSL), Regional Mean Sea Level (RMSL), and their associated long-term trends and accelerations.
lien 475: The authors mention that '... a core principle of this methodology is to preserve the independence of satellite observations from exogenous data sources, such as in situ measurements or numerical model outputs ...' whereas some of the geophysical corrections applied in the SSH calcualtion are based on models or reanalyses (e.g. ERA-5), which can assimilate exogenous data (e.g. from other sensors or in situ measurements). Could the authors clarify this assertion?
Response
Indeed, several geophysical corrections used in the SSH computation, such as ERA5-based atmospheric corrections, rely on models that assimilate external observations. Our intention was not to imply that the altimeter products are fully independent from any exogenous data source. Rather, the core principle of the methodology is that the estimation of inter-mission adjustments and biases is performed using satellite observations themselves, without relying on external datasets such as tide gauges or model outputs to determine these calibration parameters. We have revised the text accordingly to better reflect this distinction. “, even though some geophysical corrections applied to the altimeter measurements may themselves be derived from exogenous data sources when no suitable alternative is available.”
Precalibration
line 486 : Is the bias mention here is is the GMSL offset between TOPEX-A and TOPEX-B ? To be clarify in the text.
Response
Clarified in the manuscript. The bias corresponds to the estimated offset between the TOPEX-A and TOPEX-B portions of the record.
line 486 : Quet et al. (2023), referenced in the “reference section”, seems to be an oral presentation that is not available online.
Response
The reference is an oral presentation which is available online with the corresponding URL provided in the reference section.
lien 488 : The author mentions that the approach relies on the temporal stability of ERS-2, whereas the stability of ERS-2 has not been demonstrated. The author should address this assertion.
Response
The temporal stability of ERS-2 cannot be considered as fully demonstrated. The method does not rely solely on the assumption of a perfectly stable ERS-2 record. Rather, as shown by Quet et al. (2023), the resulting TOPEX-A/B bias estimate was compared with several independent approaches and found to be consistent with most of them. We have clarified this point in the manuscript and moderated the wording regarding the role of ERS-2 stability.
line 488 : To provide the confidence interval of the TOPEX-A/B GMSL offset ucneratinty (± 0.24) cm
Response
The confidence interval associated with the estimated offset has now been explicitly reported.
line 488 : The author should explain in more detail why this method preserves geophysical signals more effectively than previous constant or linear approaches. For example, Ablain et al. (2009, 2029) and Guerou et al. (2023) found that the uncertainty of the TOPEX-A and TOPEX-B GMSL offset was 2 mm at 1 sigma. Is the impact of the updated GMSL offset value (2.4 mm) really significant?
Response
The constant approach assumes no sea-level variation across the TOPEX-A/B switch, while the linear approach assumes a linear sea-level evolution during the transition period. Both assumptions are restrictive. In contrast, the proposed method uses independent altimeter observations to follow the observed sea-level variability during the transition period, thereby preserving the geophysical signal more realistically. We have clarified this point in the manuscript.
We agree that the difference between the updated offset estimate and previously published values remains small and is of the same order of magnitude as the associated uncertainty estimates. The main benefit of the proposed approach is therefore not the magnitude of the correction itself, but the fact that it relies on a more physically based representation of the observed sea-level variability rather than on prescribed constant or linear assumptions.
Alignment of reference altimeters
line 515 : Could the author confirm that the sub-millimetric accuracy level (better than 0.5 mm) of MSL offset only applies to the global mean? What is the uncertainty of the MSL offset at regional scales ?
Response
Yes, the reported sub-millimetric accuracy (<0.5 mm) applies to the estimation of the global mean MSL offset. The uncertainty of the alignment at regional scales is an important topic but is beyond the scope of the present paper and is currently being investigated in dedicated studies, including within the ESA SL_cci project. The manuscript has been clarified accordingly.
line 605: Quet et al. (2025) not available in the he “reference section”.
Response
The paper is in preparation, the reference of the technical note has been added.
line 600-605 : Given the importance of the GMSL offset in linking GMSL data records of reference missions together accurately, the author should provide new estimates of the GMSL offset and their associated uncertainty with new DT2024 standards (perhaps in the form of a table in the appendix).
Unlike previous approaches based primarily on a global MSL offset, DT-2024 relies first on a regional alignment of the missions. Therefore, the inter-mission differences are largely accounted for within the regional correction maps, and a single offset value is no longer fully representative of the adjustment applied between missions. Deriving an equivalent GMSL offset require an additional analysis. A table has been added in the appendix C with an equivalent GMSL offset and their associated uncertainty.
Alignment of coverage missions
Line 609: What exactly does the author mean by 'climatic reference'?
In this context, it refers to the continuous and stable sea-level record established by the reference altimeter missions (TOPEX, Jason-1, Jason-2, Jason-3 and Sentinel-6), which provides the backbone of the climate sea-level data record. Coverage missions are calibrated against this reference framework in order to ensure consistency with the long-term record. The manuscript has been clarified accordingly.
line 620 : I In the title of Figure 12, it would be better to write 'GMSL evolution' than 'Sea Level Anomaly'.
Response
The sentence has been revised accordingly.
Dynamical calibration of coverage missions on the reference altimeters
line 625 : The author should provide the number of large regional scales. This would help us to better understand why applying the regional MSL offset to link MSL data records from non-reference missions with those from reference missions does not completely solve all the SLA discrepancies.
Response
Here, the term "large regional scales" refers to spatial scales ranging from several hundred to several thousand kilometres, over which geographically correlated errors may persist. We have clarified this point in the manuscript.
line 625 : “Consistent” more adapted than “Coherent”
Response
The sentence has been revised accordingly.
Validation against in-situ data
line 646 : The author should comment in what extent the method is able to detect a GMSL drift for instance providing the ucenartinty of the 0.01 mm/yr trend detected over all teja ltimetry era.
Response
The method is not able to detect a drift of 0.01 mm yr⁻¹. Although 0.01 mm yr⁻¹ is the estimated trend, the associated uncertainty is ±0.32 mm yr⁻¹ at the 90% confidence level. Therefore, this trend cannot be considered significantly different from zero, and the detection threshold for a relative trend between altimetry and tide-gauge datasets is approximately 0.32 mm yr⁻¹.
line 653 : The author mentions the drift of the Jason radiometer, which was already detected by Barnoud et al. (2023). Why was this drift on the Jason (3) MWR not implemented in DT-2024, and why was it not described in more detail in the previous section?
Response
The Jason-3 MWR drift was known when the DT-2024 reprocessing was finalized. However, the corresponding correction was not yet available for operational implementation in the processing chain and could therefore not be included in DT-2024. The impact of this drift on GMSL is discussed in Barnoud et al. (2023) and Guérou et al. (2023), and future reprocessing activities will benefit from the corrected radiometer calibration as evocated in the section 6.2.
Past and future evolutions
Historical context: how does DT-2024 compare with previous upgrades?
line 678 : Improve the Y-axis title. The quantity represented in the diagram is not the SSH crossover error, but rather 'the variance reduction of SSH errors at crossovers' or 'the SSH uncertainty reduction at crossovers'.
Response
The Y-axis title has been modified as suggested.
line 663 : The author mentions that 'the impact of the DT-2024 reprocessing has been quantified in Section 2. 3', whereas Section 3 mainly addresses the impact on SSH error variance reduction at crossovers. Section 3 does not quantify the effect of the new DT-2024 standards on continuity and long-term stability. Nor is it compared with the previous version in sections 4 and 5. It would be relevant in this study to quantify the effects of the new standards on continuity (e.g.,reduction of the MSL offset uncertainty compared to the previous version) and also in terms of stability (e.g. how the MSL trend uncertainty and acceleration have been reduced with the new DT-2024).
Response
We agree that continuity and stability are essential metrics for a climate data record. The impact of each major DT-2024 standard update on GMSL stability was assessed and quantified, and, when relevant, these results are reported in Section 3. We agree that the estimation of GMSL uncertainties deserves a detailed discussion. Because of the scientific importance of this topic and the extensive analyses required, it is the subject of a dedicated publication and will be adressed in Quet et al. 2025.
Toward future reprocessing
line 696 : “Sea level stability” instead of “climate stability” (or “sea level stability stability for climate change studies”).
Response
Here , climate stability does not refer to sea-level stability itself, but rather to the suitability of the correction for climate applications and its ability to preserve a stable long-term climate record. To clarify this point, we have revised the text as follows: . However, it opens up a series of questions about the long-term stability of wave model parameters, or the sea-state bias itself and the impact of their use on the stability of sea level records for climate applications.
line 700 : Is there a peer-reviewed paper that describes the adaptive retracking algorithm?
Response
Additional references describing the adaptive retracker have been added like Tourain et al. 2021 and Thibaut et al. 2021.
line 712 : “altimetric sea level data record” instead of “altimetric climate data record”.
Response
The sentence has been revised accordingly.
line 715 : The authors should clarify the meaning of the phrase 'integrity and homogeneity of long-term sea level observations'. Are they referring to the need to maintain the continuity and stability of the sea level data record?
Response
By integrity and homogeneity, we refer to the continuity, consistency, and stability of the sea-level record. The text has been revised accordingly.
Conclusions
line 747 : The sentence '[...] that reduce the error of sea surface height variance' is incorrect. It should be either 'that reduce the variance of sea surface height errors' or 'that reduce the uncertainty of sea surface height'.
Response
The sentence has been revised accordingly.
line 748 : The author mentions that 'the reference altimeters are extremely consistent with one another', whereas different choices were made for some standards (e.g. WTC) without justification. Could the author elaborate more on this topic? ?
Response
Most processing standards are applied consistently across the reference missions, including the use of onboard radiometer wet tropospheric corrections, similar sea state bias methodologies, ionospheric corrections, and precise orbit solutions whenever available. While the radiometers themselves have evolved from one mission to the next, each reference mission relies on an onboard radiometer specifically designed to provide precise wet tropospheric corrections. This approach ensures greater consistency with the altimeter measurements than the use of external model-based corrections that are independent of the observations. We note that further improvements may still be achieved through enhanced cross-calibration and homogenization of the wet tropospheric corrections, and these aspects are the subject of ongoing studies. Therefore, the statement refers to the consistency of the final calibrated sea-level measurements across the reference mission series rather than to the use of strictly identical instruments or standards. To mitigate this statement, the word extremely was removed.
line 756 / 757 : same comment as in line 600-605, the author should provide new estimates of the GMSL offset and their associated uncertainty with new L2P2024 standards in comparison with previous version. This will demonstrate quantitatively how the continuity has been improved, as mentioned by the author.
Response
To better document the impact of DT-2024 on record continuity, a summary table will be added in Appendix C, providing an overview of the offset between reference missions and their associated uncertainties.
line 762 : The author should demonstrate how the rise in GMSL has changed with the new DT-2024 version compared to the previous one, and explain how uncertainty has been reduced. This will demonstrate quantitatively how the stability has been improved.
Response
We agree that quantifying the impact of DT-2024 on GMSL trends and their uncertainty is an important objective. In this manuscript, trend differences are only illustrated over the most recent decade to highlight the importance of the trend in the context of climate change. A comprehensive DT-2024 versus DT-2021 comparison of GMSL trends over the full record is not straightforward, as the early part of the record is still affected by known TOPEX calibration issues, while the most recent years are not included in DT-2021. Such an assessment requires a dedicated end-to-end analysis. These aspects are currently being addressed in a dedicated uncertainty-budget study in preparation. In the meantime, preliminary results have been reported in the AVISO+ DT-2024 release note, which will be referenced in the revised manuscript.
Citation: https://doi.org/10.5194/essd-2025-604-AC1 -
AC2: 'Reply on RC2', Cécile Kocha, 10 Sep 2026
REVIEWER 2
Overview
This paper documents a huge and complicated data-reprocessing effort. The result of this work is a data set that a large number of people will use (until the next reprocessing is finished). I commend the team of authors for doing a thorough piece of work. For conciseness, many details are left out. I think this is acceptable, and consistent with the authors' choice to not dwell on why the 'chosen' corrections were adopted in favour of others. Some people will disagree with some of these choices, in which case they can simply use a rival correction for their own work. In this context, I am happy with the paper essentially as it is, and only offer a relatively small number of fairly minor comments, or technical errors, as follows.
Response
We sincerely thank the reviewer for this positive and encouraging assessment of the manuscript. We greatly appreciate the recognition of the considerable effort involved in the DT-2024 reprocessing and of the importance of making these data available to the user community.
We also thank the reviewer for acknowledging the necessary balance between conciseness and completeness in a manuscript describing such a large and complex reprocessing effort. We have carefully considered all the comments and technical corrections provided and have revised the manuscript accordingly. These changes have improved the clarity and overall quality of the paper.
line Comment/question
78 what are JA2 and JA3 (cf j2 and J3 line 70)Response
JA2 and JA3 refer to Jason-2 and Jason-3 altimeters, which we should indicate as J2 and J3 as refered line 70. The text has been revised accordingly.
79 lower (Climate)
Response
The text has been revised accordingly.
96 "20211" (a typo)
Response
The text has been revised accordingly.
116 were -> was
Response
The text has been revised accordingly.
119-120 "few h or few d" which? For tides, IB, SSB, need h
We agree with the reviewer that the relevant temporal separation depends on the processes being investigated. Very short time lags (a few hours) are preferable to isolate the impact of tides, IB, SSB, and other rapidly varying signals. However, a sufficient number of crossover points is required to obtain statistically robust estimates and meaningful geographical coverage.
In practice, we therefore consider crossover points separated by up to 10 days. Over such a period, sea level variability is expected to remain limited compared with longer seasonal or interannual timescales, while the applied geophysical corrections already account for most tidal, IB, and SSB signals. Any residual errors in these corrections are precisely what the crossover analysis aims to reveal. We have clarified this point in the manuscript.
129-130 "to minimise" How does this minimise?
Response
We have revised the sentence to clarify that crossover pairs are restricted to less than 10 days because the sea level signal is expected to remain relatively stable over such a short time interval (except during extreme events), allowing crossover statistics to better characterize short-term measurement uncertainty and to compare different satellite altimeter data whereas longer-term correlated effects are assessed separately through trend and continuity analyses.
136 Huh? Geoid is not used for SSHA variance (see footnote p3)
Response
We thank the reviewer for pointing out this mistake. The geoid is not involved in the computation of SSHA, and the example was therefore inappropriate. Our intention was simply to illustrate that different corrections affect the SSHA variance in distinct and predictable geographical patterns. The example has been revised accordingly using mean sea surface instead of geoid.
145 & 216 consision -> conciseness
Response
The text has been revised accordingly.
158 "mea sea surface"
Response
The text has been revised accordingly.
176 & 936 "Salvatore et al. 2024". His name is Salvatore DINARDO. (I have not checked all references)
Response
We thank the reviewer for pointing out this mistake. The text and the reference has been revised accordingly.
239-244. Here, I get confused what you are comparing to. Are you distinguishing the 'ERA5 meteorological model' (used by DT2021) from the 'reanalysis' (used by DT2024?)
Response
Here, we compare the contribution of the ERA5 reanalysis, which is based on a fixed version of the HRES model from 2016, with the operational HRES model whose configuration evolves over time. Both DT-2021 and DT-2024 use ERA5-based corrections for the dry tropospheric correction. For the Dynamic Atmospheric Correction (DAC), ERA5-based forcing had already been adopted in DT-2021 up to the beginning of the Jason-3 mission and was extended in DT-2024 up to the beginning of the Sentinel-6A mission.
We then discuss the benefits of using a stable reanalysis framework rather than an evolving operational model, for both DT-2021 and DT-2024, particularly in terms of continuity and long-term consistency of the sea-level record. The text has been clarified accordingly.
p13 Footnote 6. This is a very good point. I was wondering how you would raise it. Its like a feedback loop, with potential for error growth but hopefully the series converges (as we all expect). I don't think important points belong in footnotes. Can it have its own paragraph? "Some may fear that the MSS, for example, is a product that is both an input to the process and an output. To reassure ourselves that the various product update cycles take us ever closer to the 'truth' rather than away from it, we ….."
Response
We agree that this is an important methodological point that deserves more visibility than a footnote. We have therefore incorporated the underlying idea into the main text.
Fig 5 Caption. 50 to 66 degrees is no man's land. Or typo?
Response
There is no typo. The selected latitude bands are representative regions. The 50°-66° range is influenced by both open-ocean and polar processes and was therefore not treated into this plot.
372-377 I think this depends on how far away from exact crossover points you count as a match.
Response
The objective of this paragraph is to illustrate that crossover statistics under 10 days do not capture all sources of SSHA variance reduction equally well. In particular, improvements associated with static or slowly varying components, such as the MSS, are not fully reflected in crossover metrics, even though they can significantly contribute to the overall SSHA variance reduction. We have clarified the text accordingly.
447 Typo
Response
The text has been revised accordingly.
458 If you have defined 'coverage mission', I missed it. I assume you mean non-reference. But geodetic phase as well?
Response
To avoid ambiguity, we have added the following definition in the Introduction:
“To distinguish missions contributing to the reference climate record from the other missions, we hereafter refer to the former as reference altimeters and to the latter as coverage and precision missions.”
Geodetic phases are therefore included in the latter category when they are not contributing to the reference climate record.
464 'climate-oriented altimeters' Too many names! I vote to use 'reference' everywhere.
Response
The term climate-oriented altimeters was originally introduced to avoid repeatedly using the term reference throughout the sentence. However, we agree that this terminology may create unnecessary confusion for the reader. The manuscript has been revised accordingly.
Figure 10 "Note that for J3 vs. S6A, the dynamic atmospheric correction is not applied"
Explain (briefly) why not. Someone skipping through the paper will be surprised to read this.Response
The DAC is not considered in the regional bias assessment between J3 and S6A because different DAC solutions are used for the two reference missions. J3 as reference mission relies on a DAC forced by the ERA5 reanalysis, whereas S6A uses a DAC forced by the HRES operational atmospheric model.
Although the transition between these two DAC solutions introduces small regional differences, these patterns cannot be directly interpreted as biases that should be propagated to the earlier missions. Indeed, the ERA5-based DAC is expected to provide a more realistic representation of the atmosphere-ocean response, owing to the nature of a reanalysis product. Therefore, the observed differences primarily reflect the change of atmospheric forcing rather than a true inter-mission bias requiring correction. We have clarified this point in the manuscript.
672 "0 to 5days" disagrees with line 129 and 443 (10 days). I think 10 days is too long but I guess its too late to do 5days. Or is there no disagreement? Ie, if a matchup is within a 10-day window: -5 to +5. In which case, this should be more clearly stated since it is a key choice.
Response
The crossover criterion corresponds to a maximum temporal separation of 10 days, i.e. approximately ±5 days around the reference observation. We acknowledge that this was not stated clearly enough and have revised the manuscript accordingly to avoid any ambiguity.
Citation: https://doi.org/10.5194/essd-2025-604-AC2
Data sets
Level-2P dataset Cécile Kocha et al. https://doi.org/10.24400/527896/a01-2025.004
Level-3 dataset Marie-Isabelle Pujol et al. https://doi.org/10.48670/moi-00146
Level-4 dataset Maxime Ballarotta and Marie-Isabelle Pujol https://doi.org/10.48670/moi-00148
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 745 | 454 | 145 | 1,344 | 181 | 162 |
- HTML: 745
- PDF: 454
- XML: 145
- Total: 1,344
- BibTeX: 181
- EndNote: 162
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Overview
This paper details the extensive efforts by space agencies to enhance the sea level data recordin terms of precision, accuracy, continuity, and stability. Specifically, it highlights the advancements integrated into the DT2024 release, focusing on the updated altimeter standards used to calculate sea level anomalies. Furthermore, the study describes the cross-calibration methodology employed to harmonize data across 15 missions, ensuring a seamless 30-year record. While this review of the DT2024 improvements is a vital resource for the user community, there are several areas where the manuscript could be strengthened, as outlined in the major and specific comments below.
Majors comments
Specific comments
Abstract
Introduction
Altimetry standards impact on sea level consistency
Methodology to select new DT24 standards
Overview of the DT-2024 standards
Altimeter input dataset reprocessing
Ocean tide correction
Dynamical atmospheric and dry tropospheric corrections
Wet tropospheric correction
Mean sea surface
Benefits from DT-2024 upgrades for ocean mesoscale
Cross-calibration methodology
Precalibration
Alignment of reference altimeters
Alignment of coverage missions
Dynamical calibration of coverage missions on the reference altimeters
Validation against in-situ data
Past and future evolutions
Historical context: how does DT-2024 compare with previous upgrades?
Toward future reprocessing
Conclusions