TL;DR
This paper presents a refined estimation of the Antarctic Ice Sheet's mass balance between 2003 and 2008, leveraging ICESat laser altimetry data. By systematically improving data processing, including inter-campaign bias correction and a novel three-step elevation change model, the authors provide a more accurate picture of ice sheet dynamics, crucial for understanding global sea-level rise.
What Problem Does This Paper Address?
Accurately quantifying the mass balance of the Antarctic Ice Sheet (AIS) is critical for predicting global sea-level rise, as Antarctica holds enough ice to raise sea levels by approximately 60 meters. Previous estimates using ICESat data have shown significant discrepancies, highlighting the need for more robust methodologies. The authors aim to address this by systematically improving the data processing and modeling techniques used to derive mass balance from ICESat altimetry, focusing on the period 2003-2008. This includes tackling challenges like inter-campaign biases in satellite measurements and accurately accounting for non-mass-change related elevation variations.
Methods
The study employs a multi-faceted approach to refine AIS mass balance estimation:
- Data Preprocessing and Correction: ICESat GLA12 data underwent rigorous preprocessing, including filtering, classification, and projection. Crucially, they addressed the Inter-Campaign Bias (ICB) by re-estimating it based on fitting results from previous studies, aiming to reconcile differences between satellite observation periods. Saturation correction was also applied.
- Three-Step Elevation Change Model: A novel three-step model was developed to calculate elevation changes across different regions of Antarctica:
- High-Density Areas: A simplified spatial/temporal polynomial model was used, incorporating surface slope, seasonal variations (modeled using a cosine function), and elevation trends. This model was applied to grids with sufficient data points and observation periods.
- Moderate-Density Areas: A simplified version of the first step was used, focusing on height change and local topography without seasonal signals, for grids with at least 10 points and over 2.5 years of data.
- Low-Latitude/Sparse Areas: A linear solution along the cross-track direction was employed, accounting for slope and elevation change trends, for areas like the Antarctic Peninsula where ICESat footprint distribution is sparse.
- Volume to Mass Conversion: To convert elevation changes to mass changes, the authors corrected for components that do not contribute to mass loss, such as global isostatic adjustment (GIA) and firn compaction. Firn compaction was modeled using surface accumulation and temperature data. A surface snow density model was constructed using field measurements and advanced algorithms. The ice sheet was then divided into zones, with ablation zones multiplied by ice density and accumulation zones by the derived surface density.
Datasets
- ICESat GLA12 Dataset (R634): This is the primary dataset used for elevation measurements, covering 15 campaigns from October 2003 to December 2008.
- Verification Data: The study utilized various datasets for validating their elevation change rate results, including:
- Global Position System (GPS) data (e.g., over Subglacial Lake Vostok, Pine Island Glacier).
- Airborne Terrain Mapper (ATM) data (over the Antarctic Peninsula).
- CHINARE field measurements (from Zhongshan Station to Dome A).
- Other published results from GPS, stakes, ATM, and CHINARE.
- Firn Compaction Modeling Inputs:
- ERA-Interim reanalysis data (for accumulation and temperature).
- AVHRR surface temperature data (for mean surface temperature and simulating densification).
- Surface Snow Density Model Inputs:
- In-situ density data from field measurements around Antarctica.
Key Insights and Ideas
- Systematic Improvements Matter: The paper strongly emphasizes that meticulous data processing, including accurate ICB correction and a nuanced elevation change modeling approach, is crucial for reducing discrepancies in AIS mass balance estimates.
- Regional Modeling is Key: Recognizing that different regions of Antarctica have varying data densities and characteristics, the authors developed a three-step modeling strategy to optimize elevation change estimation across the entire ice sheet.
- Distinguishing Mass Change: A significant contribution is the explicit correction for elevation changes not related to mass loss (firn compaction, GIA), ensuring that the final mass balance reflects actual ice melt or accumulation.
- Validation Reinforces Confidence: The agreement between their ICESat-derived elevation changes and independent datasets (GPS, ATM, field measurements) provides strong validation for their refined methodology.
Key Results
- The study presents a "systematically improved new estimation" of the Antarctic Ice Sheet's mass balance for the period 2003-2008.
- The elevation change rates derived from their improved ICESat analysis showed good agreement with independent verification data from GPS, ATM, and field measurements.
- The paper highlights the importance of correcting for firn compaction and GIA to accurately convert elevation changes to mass changes.
- A refined surface snow density model was developed and used for the conversion, contributing to a more accurate mass balance calculation.
Limitations and Future Work
- Data Coverage: While ICESat provides valuable data, its footprint and orbital spacing can still lead to sparse coverage in certain areas, necessitating different modeling approaches for different regions.
- Uncertainty in Density Models: Despite improvements, uncertainties in density models, especially for firn compaction and accumulation, can still influence the final mass balance estimates.
- Time Period: The study focuses on the 2003-2008 period. Extending this analysis to longer time series and incorporating data from newer missions like CryoSat-2 and Sentinel-3 would provide a more comprehensive understanding of long-term trends.
- Future Work: The authors indicate they will present their improved mass balance results for the 2003-2008 period at a workshop, suggesting that the full quantitative results are forthcoming. Further work could involve integrating their refined methods with other satellite missions and climate models.
Summary updated