Louis-Marie Gauer

The Antarctic Peninsula (AP), encompassing the ice sheet and its peripheral glaciers, is a highly dynamic component of the cryosphere that disproportionately contributes to sea-level rise relative to its size. However, a large spread remains between the mass changes estimated using gravimetry, altimetry and the mass budget method. Among these techniques, the satellite (radar or laser) altimetry method has a resolution of, at best, 1 km, which is too coarse to resolve the mountainous landscape of the Peninsula. Therefore, we use digital elevation models (DEMs; 30 × 30 m) to provide an estimate of mass changes across the AP between 2007 and 2021. We combine 476 DEMs derived from SPOT5-HRS satellite images (2006–2008) and 2525 DEMs of the Reference Elevation Model of Antarctica (2020–2022) to map elevation changes for the entire Peninsula. We vertically adjusted each DEM using near-coincident ICESat/-2 laser altimetry measurements. Our observations cover 70 % of the AP Ice Sheet and 60 % of its peripheral glaciers, including regions of the Peninsula poorly studied to date, and reveal a spatially complex pattern of elevation changes. After correction with different models of firn air content and solid-earth response, we find that between 2007 and 2021, the AP Ice Sheet lost 28 ± 6 Gt a⁻¹ while its peripheral glaciers lost 14 ± 3 Gt a⁻¹. For the AP Ice Sheet, our new estimate is 4 to 5 times more negative than the one obtained in IMBIE using altimetry data (−5.2 ± 1.8 Gt a⁻¹ from 2007 to 2019) and in better agreement with gravimetry and the mass budget method. Our study highlights the importance of resolving fine scale elevation changes of glaciers and ice sheets in complex, dynamic coastal areas.

The Antarctic Peninsula (AP), encompassing the ice sheet and its peripheral glaciers, is a highly dynamic component of the cryosphere that disproportionately contributes to sea level rise relative to its size. However, a large spread remains between the mass changes estimated using gravimetry, altimetry and the mass budget method. Among these techniques, the satellite (radar or laser) altimetry method has a resolution of, at best, 1 km, which is coarse to resolve the mountainous landscape of the Peninsula. Therefore, we use digital elevation models (DEMs; 30 x 30 m) to provide an estimate of mass changes across the AP between 2007 and 2021. We combine 476 DEMs derived from SPOT5-HRS satellite images (2006–2008) and 2 525 DEMs of the Reference Elevation Model of Antarctica (2020–2022) to map elevation changes for the entire Peninsula. We vertically adjusted each DEM using near-coincident ICESat/-2 laser altimetry measurements. Our observations cover 70 % of the AP Ice Sheet and 60 % of its peripheral glaciers, including regions of the Peninsula poorly studied to date, and reveal a spatially complex pattern of elevation changes. After correction with different models of firn air content and solid-earth response, we find that between 2007 and 2021, the AP Ice Sheet lost -27 ± 9 Gt/a while its peripheral glaciers lost -14 ± 2 Gt/a. For the AP Ice Sheet, our new estimate is four to five times more negative than the one obtained in IMBIE using altimetry data (-5.2 ± 2.3 Gt/a from 2007 to 2019) and in better agreement with gravimetry and the mass budget method. Our study highlights the importance of resolving fine scale elevation changes of glaciers and ice sheets in complex, dynamic coastal areas.

The Antarctic Peninsula (AP), encompassing the ice sheet and its peripheral glaciers, is a highly dynamic component of the cryosphere that disproportionately contributes to sea level rise relative to its size. However, a large spread remains between the mass changes estimated using gravimetry, altimetry and the mass budget method. Among these techniques, the satellite (radar or laser) altimetry method has a resolution of, at best, 1 km, which is coarse to resolve the mountainous landscape of the Peninsula. Therefore, we use digital elevation models (DEMs; 30 x 30 m) to provide an estimate of mass changes across the AP between 2007 and 2021. We combine 476 DEMs derived from SPOT5-HRS satellite images (2006–2008) and 2 525 DEMs of the Reference Elevation Model of Antarctica (2020–2022) to map elevation changes for the entire Peninsula. We vertically adjusted each DEM using near-coincident ICESat/-2 laser altimetry measurements. Our observations cover 70 % of the AP Ice Sheet and 60 % of its peripheral glaciers, including regions of the Peninsula poorly studied to date, and reveal a spatially complex pattern of elevation changes. After correction with different models of firn air content and solid-earth response, we find that between 2007 and 2021, the AP Ice Sheet lost -27 ± 9 Gt/a while its peripheral glaciers lost -14 ± 2 Gt/a. For the AP Ice Sheet, our new estimate is four to five times more negative than the one obtained in IMBIE using altimetry data (-5.2 ± 2.3 Gt/a from 2007 to 2019) and in better agreement with gravimetry and the mass budget method. Our study highlights the importance of resolving fine scale elevation changes of glaciers and ice sheets in complex, dynamic coastal areas.

The Antarctic Peninsula (AP), encompassing the ice sheet and its peripheral glaciers, is a highly dynamic component of the cryosphere that disproportionately contributes to sea level rise relative to its size. However, a large spread remains between the mass changes estimated using gravimetry, altimetry and the mass budget method. Among these techniques, the satellite (radar or laser) altimetry method has a resolution of, at best, 1 km, which is coarse to resolve the mountainous landscape of the Peninsula. Therefore, we use digital elevation models (DEMs; 30 x 30 m) to provide an estimate of mass changes across the AP between 2007 and 2021. We combine 476 DEMs derived from SPOT5-HRS satellite images (2006–2008) and 2 525 DEMs of the Reference Elevation Model of Antarctica (2020–2022) to map elevation changes for the entire Peninsula. We vertically adjusted each DEM using near-coincident ICESat/-2 laser altimetry measurements. Our observations cover 70 % of the AP Ice Sheet and 60 % of its peripheral glaciers, including regions of the Peninsula poorly studied to date, and reveal a spatially complex pattern of elevation changes. After correction with different models of firn air content and solid-earth response, we find that between 2007 and 2021, the AP Ice Sheet lost -27 ± 9 Gt/a while its peripheral glaciers lost -14 ± 2 Gt/a. For the AP Ice Sheet, our new estimate is four to five times more negative than the one obtained in IMBIE using altimetry data (-5.2 ± 2.3 Gt/a from 2007 to 2019) and in better agreement with gravimetry and the mass budget method. Our study highlights the importance of resolving fine scale elevation changes of glaciers and ice sheets in complex, dynamic coastal areas.

The Antarctic Peninsula (AP), encompassing the ice sheet and its peripheral glaciers, is a highly dynamic component of the cryosphere that disproportionately contributes to sea level rise relative to its size. However, a large spread remains between the mass changes estimated using gravimetry, altimetry and the mass budget method. Among these techniques, the satellite (radar or laser) altimetry method has a resolution of, at best, 1 km, which is coarse to resolve the mountainous landscape of the Peninsula. Therefore, we use digital elevation models (DEMs; 30 x 30 m) to provide an estimate of mass changes across the AP between 2007 and 2021. We combine 476 DEMs derived from SPOT5-HRS satellite images (2006–2008) and 2 525 DEMs of the Reference Elevation Model of Antarctica (2020–2022) to map elevation changes for the entire Peninsula. We vertically adjusted each DEM using near-coincident ICESat/-2 laser altimetry measurements. Our observations cover 70 % of the AP Ice Sheet and 60 % of its peripheral glaciers, including regions of the Peninsula poorly studied to date, and reveal a spatially complex pattern of elevation changes. After correction with different models of firn air content and solid-earth response, we find that between 2007 and 2021, the AP Ice Sheet lost -27 ± 9 Gt/a while its peripheral glaciers lost -14 ± 2 Gt/a. For the AP Ice Sheet, our new estimate is four to five times more negative than the one obtained in IMBIE using altimetry data (-5.2 ± 2.3 Gt/a from 2007 to 2019) and in better agreement with gravimetry and the mass budget method. Our study highlights the importance of resolving fine scale elevation changes of glaciers and ice sheets in complex, dynamic coastal areas.

The Antarctic Peninsula (AP), encompassing the ice sheet and its peripheral glaciers, is a highly dynamic component of the cryosphere that disproportionately contributes to sea level rise relative to its size. However, a large spread remains between the mass changes estimated using gravimetry, altimetry and the mass budget method. Among these techniques, the satellite (radar or laser) altimetry method has a resolution of, at best, 1 km, which is coarse to resolve the mountainous landscape of the Peninsula. Therefore, we use digital elevation models (DEMs; 30 x 30 m) to provide an estimate of mass changes across the AP between 2007 and 2021. We combine 476 DEMs derived from SPOT5-HRS satellite images (2006–2008) and 2 525 DEMs of the Reference Elevation Model of Antarctica (2020–2022) to map elevation changes for the entire Peninsula. We vertically adjusted each DEM using near-coincident ICESat/-2 laser altimetry measurements. Our observations cover 70 % of the AP Ice Sheet and 60 % of its peripheral glaciers, including regions of the Peninsula poorly studied to date, and reveal a spatially complex pattern of elevation changes. After correction with different models of firn air content and solid-earth response, we find that between 2007 and 2021, the AP Ice Sheet lost -27 ± 9 Gt/a while its peripheral glaciers lost -14 ± 2 Gt/a. For the AP Ice Sheet, our new estimate is four to five times more negative than the one obtained in IMBIE using altimetry data (-5.2 ± 2.3 Gt/a from 2007 to 2019) and in better agreement with gravimetry and the mass budget method. Our study highlights the importance of resolving fine scale elevation changes of glaciers and ice sheets in complex, dynamic coastal areas.

Le changement climatique aggrave la pénurie d'eau douce, menaçant les populations et les écosystèmes. La fonte des glaces, les précipitations irrégulières et l'évaporation plus rapide exercent une pression importante sur les ressources agricoles et hydriques. De plus, la redistribution de l'eau des continents vers l'océan contribue à l'élévation du niveau de la mer, menaçant les zones côtières. Une approche multidisciplinaire est essentielle pour surveiller l'évolution des ressources en eau douce. La Géodésie, l'étude de l'évolution de la forme de notre planète, est un outil clé dans ce contexte. Les missions satellites Gravity Recovery and Climate Experiment (GRACE) et Follow-On (GRACE-FO) fournissent des mesures mensuelles du champ de gravité terrestre depuis 2002, révélant les redistributions de masse hydrologique à grande échelle. Cependant, ces observations souffrent d'une résolution spatiale limitée, d'un bruit corrélé important et d'observations manquantes, ce qui entraîne de grandes incertitudes dans les estimations de bilans de masse hydrologiques aux échelles régionales et globales. Dans cette thèse, je développe des méthodes innovantes pour réduire le bruit et estimer avec précision les bilans de masse hydrologiques à partir des observations GRACE/-FO. Le chapitre 1 fournit une introduction détaillée aux missions GRACE/-FO, aux techniques de traitement des données et à leurs applications variées en géophysique. L'amélioration de la résolution spatiale et temporelle de ces données est désormais essentielle pour permettre encore plus d'applications. Dans le chapitre 2, je propose une méthode de filtrage, s'appuyant sur les travaux de Prevost et al., 2019, qui améliore la résolution spatiale et comble objectivement les observations manquantes des données GRACE/-FO. Ceci est accompli à l'aide d'une méthode statistique : la "Multichannel Singular Spectrum Analysis" (M-SSA) qui exploite les corrélations spatio-temporelles. Cette méthode filtre efficacement les données, révélant ainsi des signaux à petite échelle tels que certains lacs de barrage ou glaciers. Le chapitre 3 présente une méthode pour estimer le bilan de masse hydrologique de sources spécifiques à partir des données GRACE/-FO et corrige de l'erreur de filtrage. Nous isolons ainsi des sources hydrologiques à partir de mesures gravimétriques intégrées. Nous appliquons la méthode pour déterminer les variations de masse de la mer Caspienne entre 2003 et 2021 et cherchons à valider nos observations par comparaison avec des données marégraphiques et altimétriques. L'excellent accord entre toutes ces estimations démontre l'efficacité de la méthode. Dans le Chapitre 4, nous étendons cette méthode à la calotte Cook, dans l'archipel des Kerguelen. Une comparaison avec des estimations de la masse de glace obtenues à partir de la différenciation de modèles numériques de terrain suggère une contribution viscoélastique au bilan de masse issu de GRACE/-FO, en lien avec la fonte récente de la calotte. Pour confirmer cette hypothèse, nous modélisons la déformation du sol causée par les changements d'élévation de la glace. La comparaison de ces résultats avec les mesures de RADAR interférométrique à synthèse d'ouverture (InSAR) suggère aussi une déformation viscoélastique transitoire en réponse à la fonte décennale des glaces.

The Antarctic Peninsula (AP), encompassing the ice sheet and its peripheral glaciers, is a highly dynamic component of the cryosphere that disproportionately contributes to sea level rise relative to its size. However, a large spread remains between the mass changes estimated using gravimetry, altimetry and the mass budget method. Among these techniques, the satellite (radar or laser) altimetry method has a resolution of, at best, 1 km, which is coarse to resolve the mountainous landscape of the Peninsula. Therefore, we use digital elevation models (DEMs; 30 x 30 m) to provide an estimate of mass changes across the AP between 2007 and 2021. We combine 476 DEMs derived from SPOT5-HRS satellite images (2006–2008) and 2 525 DEMs of the Reference Elevation Model of Antarctica (2020–2022) to map elevation changes for the entire Peninsula. We vertically adjusted each DEM using near-coincident ICESat/-2 laser altimetry measurements. Our observations cover 70 % of the AP Ice Sheet and 60 % of its peripheral glaciers, including regions of the Peninsula poorly studied to date, and reveal a spatially complex pattern of elevation changes. After correction with different models of firn air content and solid-earth response, we find that between 2007 and 2021, the AP Ice Sheet lost -27 ± 9 Gt/a while its peripheral glaciers lost -14 ± 2 Gt/a. For the AP Ice Sheet, our new estimate is four to five times more negative than the one obtained in IMBIE using altimetry data (-5.2 ± 2.3 Gt/a from 2007 to 2019) and in better agreement with gravimetry and the mass budget method. Our study highlights the importance of resolving fine scale elevation changes of glaciers and ice sheets in complex, dynamic coastal areas.

In many mountain regions, glaciers constitute an essential freshwater reservoir and represent a critical water supply for downstream populations during droughts. Moreover, glacier mass changes provide a direct and valuable indicator of ongoing climate change. While long-term glacier mass loss is well documented, the drivers of temporal variability in regional glacier mass balance remain poorly constrained, despite their importance for understanding regional glacier–climate interactions.Satellite gravimetry missions Gravity Recovery and Climate Experiment (GRACE) and GRACE Follow-On (GRACE-FO) provide direct observations of mass variations at monthly timescales with global coverage over the past two decades. Here, we use GRACE/-FO Level-3 products to estimate glacier mass changes over some regions of the Randolph Glacier Inventory (RGI v6.0) and to investigate their sensitivity to climate variability. Two GRACE/-FO solutions are considered: a solution based on Multichannel Singular Spectrum Analysis (M-SSA), and the SAGSA ensemble solution combining products from multiple processing centers. Signal leakage and separation effects related to the coarse spatial resolution of gravimetry are mitigated using geometry-based approaches, enabling regional mass balance estimates.Monthly region-wide glacier mass balance are then used to explore relationships with key climate-related drivers, with an initial focus on near-surface air temperature and precipitation derived from the ERA5 reanalysis. This study aims to explore the links between glacier mass variability and climate-related drivers using GRACE/GRACE-FO observations.

The Antarctic Peninsula (AP), encompassing the ice sheet and its peripheral glaciers, is a highly dynamic component of the cryosphere that disproportionately contributes to sea level rise relative to its size. However, a large spread remains between the mass changes estimated using gravimetry, altimetry and the mass budget method. Among these techniques, the satellite (radar or laser) altimetry method has a resolution of, at best, 1 km, which is coarse to resolve the mountainous landscape of the Peninsula. Therefore, we use digital elevation models (DEMs; 30 x 30 m) to provide an estimate of mass changes across the AP between 2007 and 2021. We combine 476 DEMs derived from SPOT5-HRS satellite images (2006–2008) and 2 525 DEMs of the Reference Elevation Model of Antarctica (2020–2022) to map elevation changes for the entire Peninsula. We vertically adjusted each DEM using near-coincident ICESat/-2 laser altimetry measurements. Our observations cover 70 % of the AP Ice Sheet and 60 % of its peripheral glaciers, including regions of the Peninsula poorly studied to date, and reveal a spatially complex pattern of elevation changes. After correction with different models of firn air content and solid-earth response, we find that between 2007 and 2021, the AP Ice Sheet lost -27 ± 9 Gt/a while its peripheral glaciers lost -14 ± 2 Gt/a. For the AP Ice Sheet, our new estimate is four to five times more negative than the one obtained in IMBIE using altimetry data (-5.2 ± 2.3 Gt/a from 2007 to 2019) and in better agreement with gravimetry and the mass budget method. Our study highlights the importance of resolving fine scale elevation changes of glaciers and ice sheets in complex, dynamic coastal areas.