the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 13 Dec 2024
New Radar Altimetry Datasets of Greenland and Antarctic Surface Elevation, 1991–2012
Abstract. Over the past three decades, there has been a 4.5-fold increase in the loss of ice from the Greenland and Antarctic Ice Sheets, resulting in an enhanced contribution to global sea level rise. Accurately tracking these changes in ice mass requires comprehensive, long-term measurements, which are only feasible from space. Satellite radar altimetry provides the longest near-continuous record of ice sheet surface elevation and volume change, dating back to the launch of ERS-1 in 1991, and maintained through the successive ERS-2, Envisat, CryoSat-2 and Sentinel-3 missions. To reliably constrain multi-decadal trends in ice sheet imbalance, and to place current observations within a longer-term context, requires continued efforts to optimise the processing of data acquired by the older historical missions, and to evaluate the accuracy of these measurements. Here, we present new ERS-1, ERS-2 and Envisat altimeter datasets, which are derived using consistent and improved retrieval methods, and provide measurements of ice sheet elevation spanning two decades. Through comparison with independent airborne datasets, we provide a comprehensive assessment of the accuracy of these measurements, and the improvements delivered relative to previously available products. These new datasets will be of benefit to a broad range of applications, including the quantification of ice sheet mass imbalance, investigations of the processes driving contemporary ice loss, and the constraint of numerical ice sheet models.
- Preprint
(13909 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 13 Feb 2025)
-
RC1: 'Comment on egusphere-2024-3446', Anonymous Referee #1, 17 Jan 2025
reply
The manuscript reports the generation and assessment of new data products of Greenland Ice Sheet and Antarctic Ice Sheet surface elevation from the ERS-1, ERS-2 and Envisat radar altimetry missions. The underlying reprocessing has been performed within ESA's Fundamental Data Records for Altimetry (FDR4ALT) project. It is to supersede the previous REprocessing Altimeter Products for ERS-2 and ERS-2 (REAPER) and the version 3 products for Envisat.
The data products are on the Level-2 level and contain position and surface elevation along track of satellite measurements. In addition a Thematic Data Product (TDP) is provided (which could be classified as Level-2b). It refers all measurements to reference ground tracks and thereby corrects for the non-exact-repeat nature of the orbits and measurements and for the associated topography-induced variations. I expect that this will largely facilitate higher-level analyses for temporal surface elevation change.The improvements in the processing methodology include up-to-data retracking methods and the Roemer et al. 2007 approach of accounting for local topography by relocation.
The new data products, as well as the previous REAPER and Envisat version 3 products are extensively assessed and validated. Comparison to independent surface elevation data from Operatoin ICEBridge is done for Greenland. For both ice sheets, and for the three missions, waveforms are classified (using 21 different classes) through a Neural Network classification. The geographic distribution and histograms of the waveform classes are shown. For Greenland, this allows to assess surface elevation accuracy per waveform class, with very interesting insights and hints to future work on the waveform classes associated with low accuracy. It is shown convincingly that the new products outperform the previous products in virtually all aspects considered.
I enjoyed reading this manuscript. It is extremely clear and informative, starting with the helpful and succinct summary of the three missions in Section 2.1.-2.3 through the illustration on how the TMP simplifies variability in Section 3.4. Thank you for making the review work so easy.
I only have a few minor-to-moderate comments on how the presentation could be further improved.
From the abstract and even from the introduction it is not clear what kind of dataset is presented. For example, readers of a certain background might expect a gridded data product. So you might specify that it is a Level-2 dataset comprising horizontal coordinates (lat lon, or some projection?) and surface elevation (ellipsoidal heights?) sampled along the satellite track.
The manuscript presents datasets (rather than cryospheric science results). Therefore it would be really appropriate to provide public access to the datasets at the time of publishing the paper, at latest. Currently, the manuscript does not comply with TC data policy (https://www.the-cryosphere.net/policies/data_policy.html)
One might even argue that a data journal like ESSD might be more appropriate for this work. Anyway, I would not mind to see this work published in "The Cryosphere" because the excellent analysis of the data (such as the dependence of accuracy on slope and waveform classes) really facilitates understanding the nature of altimetric surface elevation measurements. And this is the basis for a lot of science.
On a similar line, the reader is referred to technical documents for details of the processing (The FDR4ALT Detailed Processing Model Document; the FDR4ALT Product User Guide), which are currently available at websites. It would be desirable to see some commitment for their persistent availability. (It's also curious that these documents have "confidential" at the bottom of every page.)
I'm not going to insist on this comment, though, as I understand it may not be trivial.The abstract could be a bit more specific as to what kind of improvements have been implemented in the processing and what improvements could be achieved w.r.t. previous datasets.
Concerning the validation by ICEBridge data, the reader might wonder why this exercise was restricted to Greenland. Maybe you can explain or justify this limitation in a sentence or two.
line 84 "had the impact of" --> "had the effect of"
line 704 something wrong about this reference "Team, I"
line 702 same problemIn the disucssion of the relocation you might mention that the geographic position assigned to the measurement is corrected together with the height. This is (of course) not reflected in Eq. 1.
line 187: As fas as I understand, the ocean tide and inverse barometer effect are zero over ice sheets. This could be mentioned for completeness.
289 "mean square error". Maybe "mean square difference" might be a more appropriate term.
line 292ff: The reader learns that this classification was performed for waveforms from not only over the ice sheet.
Make sure in the text that the reader understands which steps are done for global altimetry data and which steps are done specifically for ice sheets.line 345 "less continuous along-track sampling"
I think you could sell this point more positively and avoid the impression (at first reading of this sentence) that this is some disadvantage.line 374 replace "Threshold Centre of Gravity" by TCOG for consistency and brevity.
Fig. 6: This is a great visualisation. Just the histogram insets are too small to be well readable, at least in a printed version. You may try to make them a bit bigger. And convince the journal in the layout process to use the full paper width for the figure. In turn you might cut the headline "Envisat | cycle 78" and put this information into the figure description.
Somehow, Figure 10 is missing completely.
Fig. 12a: It appears that slopes are shown not only for the ice sheet area, which is inconsistent with what is sayed in the figure caption and with the other figures in the manuscript.
line 490f "Comparing performance across the three missions shows an apparent lower accuracy of Envisat at higher slopes, which is likely to
reflect the increased coverage that Envisat achieves in these regions (Figure 5)."
I suggest to replace "Comparing performance" by "Comparing the performance of FDR4ALT" in order to be cristal-clear.
I suggest the reference to Fig. 5 should be replaced by a reference to Table 7, because the increased coverage cannot easily seen from Fig. 5, while it is clearly quantified in Table 7Fig. 12 caption: I suggest to shorten the description by writing something like:
"d and e: same as b and c, but for ERS-2. f and g: same as b and c, but for Envisat"Fig 13 and 14. The caption says "The spatial distribution of the five most common waveform classes". However it seems that not only the five most common classes are depicted. For example, panel a and b show a lot of orange dots, too.
Also, the the color bar is somewhat ambiguous in assigning colors to waveform classes. (Maybe stretch it from -0.5 to 20.5, then the ticks at 0, 1, ..., 20 are in the centre of their color block)
Congratulations for Figure 15. I like your graphical indication of waveform classes at the top of the figure!
Yet I have some minor comments on this figure.
- Choose one name for class 9: either "very complex echo" (as in Tab. 2, 8, 9) or "unclassified" as in the top of Fig. 15.
- There is maybe too much graphical emphasis on the outliers. - certainly a matter of taste.
- Figure 15 caption, last sentence: I guess the order must be changed into "Classes 8, 12, 14, 19-20 and 10, 13-15, 17-20" to match the order "ERS and Envisat"Throughout the paper, be consistent in writing Envisat (rather than ENVISAT)
line 534ff "In contrast, relatively specular (classes 2 and 4) and multipeaked (class 3) waveforms generally produce a negative elevation bias, although the degradation in performance is less severe for Envisat than the preceding missions."
The second half of this sentence is substantiated for class 3 but not for classes 2 and 4. Maybe limit this consideration to class 3, as occurance of class 2 and 4 is rare.line 539 "For Envisat, beyond classes 1, 3 and 7, the most common remaining classes are 9 (very complex) and 11 (stepped leading edge)."
This sentence seems to suggest that classes 9 and 11 are less common than classes 1, 3, 7, but this is not the case.Fig. 16 caption: in line 571, "(row 1)" does not appear to make sense to me.
Citation: https://doi.org/10.5194/egusphere-2024-3446-RC1
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 184 | 64 | 11 | 259 | 5 | 4 |
- HTML: 184
- PDF: 64
- XML: 11
- Total: 259
- BibTeX: 5
- EndNote: 4
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1