the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Nov 2024
Smoothed monthly Greenland ice sheet elevation changes during 2003–2023
Abstract. The surface elevation of the Greenland Ice Sheet is constantly changing due to the interplay between surface mass balance processes and ice dynamics, each exhibiting distinct spatiotemporal patterns. Here, we employ satellite and airborne altimetry data with fine spatial (1 km) and temporal (monthly) resolutions to document this spatiotemporal evolution from January 2003 to August 2023. To estimate elevation changes of the Greenland Ice Sheet (GIS), we utilize radar altimetry data from CryoSat-2 and EnviSat, laser altimetry data from the ICESat and ICESat-2, and laser altimetry data from NASA’s Operation IceBridge Airborne Topographic Mapper. We produce continuous monthly ice surface elevation changes from January 2003 to August 2023 on a 1 km grid covering the entire GIS. We estimate cumulative ice loss of 4,352 Gt ± 315 Gt (12.1 ± 0.9 mm sea level equivalent) during this period, excluding peripheral glaciers. Between 2003 and 2023, the ice sheet land-terminating margin underwent a significant cumulative thinning of several meters. Ocean-terminating glaciers exhibited thinning between 20–40 m, with Jakobshavn Isbræ experiencing an exceptional thinning of nearly 70 m. This dataset of fine-resolution altimetry data in both space and time will support studies of ice mass loss and useful for GIS ice sheet modelling. To validate our monthly mass changes of the Greenland ice sheet, we use mass change from satellite gravimetry and mass change from the Input-Output method. On multiannual timescales, there is a strong correlation between the time series, with R values ranging from 0.88 to 0.92.
- Preprint
(3546 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 24 Jan 2025)
-
CC1: 'Comment on essd-2024-348', Ken Mankoff, 22 Nov 2024
reply
The comment was uploaded in the form of a supplement: https://essd.copernicus.org/preprints/essd-2024-348/essd-2024-348-CC1-supplement.pdf
-
RC1: 'Comment on essd-2024-348', Anonymous Referee #1, 05 Jan 2025
reply
This manuscript presents monthly GrIS elevation changes from a long-term series using multi-sources satellite and airborne altimeter data. The authors improved the previous annual elevation change method to detect monthly elevation changes. They also separated the seasonal surface variation from the time series of surface elevation observations. This method seems to be effective; however, I still have some concerns about this paper.
Major comments:
- This paper resembles more a technical report than a scientific paper because it lacks careful organization of original data and a logical description of the methods.
- The authors use the seasonal terms derived from ICESat or ICESat-2 to represent the seasonal surface elevation changes observed in other satellite altimeters. The rationale and the associated uncertainties should be discussed further.
- The validation and cross-comparison with other monthly GrIS elevation change methods should be discussed, such as the method developed by Lai et al. R. Lai and L. Wang, Monthly Surface Elevation Changes of the Greenland Ice Sheet From ICESat-1, CryoSat-2, and ICESat-2 Altimetry Missions, IEEE Geoscience and Remote Sensing Letters, vol. 19, pp. 1-5, 2022, doi: 10.1109/LGRS.2021.3058956
- The accuracy of the time series elevation change detection method depends on the validity of observations in a specific grid. With higher resolution grids, there are fewer observations. Did the authors analyze the distribution of valid observations at 1 km resolution across the whole GrIS on a monthly scale? If so, please add this distribution.
Minor comments:
- In Section 2, please add a table to summarize the data and its characteristics, such as time span and original accuracy.
- Line 155: “For each grid point with the center at (x0, y0), we identify the nearest data point within a 1000 m radius (xi, yi, hi, ti )……”. Since the resolution of the grid is 1000 m, why not use a 500 m radius instead?
- Line 185: “Using the above equation, we only need to estimate two unknowns, A and 𝜑.” In fact, Equation 5 has three parameters.
- Line 250: “For the 2009-2018 period, (when data from both missions is available), we derive the seasonal signal averaged from ICESat and ICESat-2.” The ICESat spans 2003–2009, while ICESat-2 spans 2018–2023. How can the seasonal signal be averaged over these periods?
- Line 380: The ice loss from other methods should be listed here.
- Line 395 and Figure 15(a): “with R values ranging from 0.88 to 0.92”. The R value should be referred to as the coefficient of determination (R²).
- I suggest listing the data in the appendix.
Citation: https://doi.org/10.5194/essd-2024-348-RC1 -
RC2: 'Comment on essd-2024-348', Anonymous Referee #2, 19 Jan 2025
reply
Please refer to the attached file for the peer review comments.
-
RC3: 'Comment on essd-2024-348', Anonymous Referee #3, 19 Jan 2025
reply
This paper presents a valuable new altimetric dataset of the Greenland Ice Sheet (GIS), derived from satellite and airborne altimetry data. The authors describe the processing steps for generating gridded (1 km × 1 km) monthly time series of surface elevation change for the GIS. The dataset was created using altimetry data from Envisat, ICESat, CryoSat-2, ICESat-2, and Operation IceBridge ATM. The authors also validate their monthly GIS elevation products against results from satellite gravimetry and the Input-Output method. However, I have a few suggestions and points of clarification before the manuscript is finalized. Detailed comments are outlined below.
Major Comments:
- Section 3.2: I notice the spatial resolution of ICESat is much lower than 1 km, especially in lower-latitude regions of the GIS. Given this, ICESat data points within 1 km of grid nodes are typically located along repeat tracks from different cycles. How can the authors ensure the stability of the multi-parameter solution (7th-order polynomial, 3rd-order surface topography, seasonal term, and 21 parameters in total) with such sparse data points? If the number of height observations is smaller than the number of parameters to be solved, could the authors clarify how this issue is addressed?
- Lines 240-253: In this step, the radar seasonal signal is removed and replaced with the laser seasonal signal. ICESat/GLAS data covers only 18 discontinuous cycles between 2003 and 2010. How much will this substitution improve the estimate of the seasonal term, especially during the period between 2009 and 2017, when laser altimeter data is missing? Could the authors elaborate on this aspect?
- Please double check the following references:
Nilsson, J. and Gardner, A. S.: Elevation Change of the Greenland Ice Sheet and its Peripheral Glaciers: 1992–2023, Earth Syst. Sci. Data Discuss. [preprint], https://doi.org/10.5194/essd-2024-311, in review, 2024.
Nilsson, J., Gardner, A. S., and Paolo, F. S.: Elevation change of the Antarctic Ice Sheet: 1985 to 2020, Earth Syst. Sci. Data, 14, 3573–3598, https://doi.org/10.5194/essd-14-3573-2022, 2022.
Minor Comments:
- Line 234: "Furthermore, we detect and remove outliers from each time series." How are outliers removed? Please provide more detail on the method used.
- Lines 255-262: The processing steps for Envisat and CryoSat-2 are similar, except for two individual time sub-intervals for CryoSat-2. I suggest the authors separately introduce the separate processing steps for radar and laser altimeter data to make this section clearer.
- Section 3.7: When creating the multi-sensor monthly grid, how are the estimated monthly change rates for the same month and grid cell merged? Did the authors consider the potential inconsistency in reference frames between different altimetry missions? Please specify.
- Line 296 and Figure 11: The merged data also contains many NaN grids. The interpolation method used is crucial in such cases. Could the authors provide the average percentage of effective raw grids used each month?
- Figure 12: The time series of cumulative monthly ice mass change is presented, but it would be helpful to include the average annual rate of mass change in the same period calculated by different methods. This would provide additional context and comparison.
- Line 180: SMB -> Surface Mass Balance (SMB).
- Section 3.8.3: How did the authors account for the impact of SMB in the step of converting volume to mass?
Citation: https://doi.org/10.5194/essd-2024-348-RC3
Data sets
Smoothed monthly Greenland ice sheet elevation changes during 2003-2023 Shfaqat A. Khan, Helene Seroussi, Mathieu Morlighem, William Colgan, Veit Helm, Gong Cheng, Danjal Berg, Valentina R. Barletta, Nicolaj K. Larsen, William Kochtitzky, Michiel van den Broeke, Kurt H. Kjær, Andy Aschwanden, Brice Noël, Jason E. Box, Joseph A. MacGregor, Robert S. Fausto, Kenneth D. Mankoff, Kuba Oniszk, Dominik Fahrner, Anja Løkkegaard, Eigil Y. H. Lippert, and Javed Hassan https://datadryad.org/stash/share/RFbPIGTZRn0Vs0u8hj1PSgjde11IPAL-k8TvEOOEBDA
Model code and software
Matlab code related to paper: Smoothed monthly Greenland ice sheet elevation changes during 2003-2023 by Khan et al, 2024. (in ESSD) Shfaqat Abbas Khan https://doi.org/10.5281/zenodo.13276108
Video supplement
Cumulative surface elevation change of GIS from 2003 to 2023 Shfaqat Abbas Khan https://ftp.space.dtu.dk/pub/abbas/ESSD2024/
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 402 | 73 | 14 | 489 | 11 | 10 |
- HTML: 402
- PDF: 73
- XML: 14
- Total: 489
- BibTeX: 11
- EndNote: 10
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1