North-eastern part of Frýdlantský Spur is situated at the northern foothills of the Jizera Mountains. It represents an area, which was at least twice covered by North-European ice sheet during the Middle Pleistocene. At the same time, it is the region whose hills have the summits around the expected altitudinal limit of the maximum ice sheet extent. Therefore, it is a suitable area to study glacial trimline altitude. Schmidt hammer testing was applied to determine different state of rock surface weathering. Obtained R-values were taken as the most relevant data for glacial trimline determination. Nevertheless, other approaches (GIS analysis, outcrops, block fields and erratics mapping) were also incorporated. Interpretation of these approaches gave the altitudinal limit of ice sheets in the north-eastern part of the Frýdlantský Spur at 500 m a. s. l. ± 20 m.

Dans le contexte actuel du réchauffement climatique, l'une des principales sources d'incertitude pour l'élévation du niveau de la mer est la contribution de la calotte Antarctique. L'étendue et les conditions météorologiques extrêmes de ce continent font de la télédétection spatiale un moyen utile pour son suivi sur le long terme. Les observations satellites altimétriques et radiométriques dans la gamme des micro-ondes rendent compte de l'évolution des propriétés du manteau neigeux de la calotte. L'altimétrie radar, par des mesures répétées de l'élévation de la topographie de surface, permet de quantifier les variations de volume sur l'ensemble du continent. Cependant, la pénétration de l'onde radar dans la neige affecte négativement cette quantification. Les méthodes proposées pour minimiser les erreurs de pénétration sont toutes basées sur des relations avec le coefficient de rétrodiffusion radar. La compréhension des variations annuelles et inter-annuelles du coefficient de rétrodiffusion est nécessaire pour améliorer la précision de l'estimation de l'élévation de la surface donc du bilan de volume de la calotte. Cette thèse a pour objectif d'étudier le coefficient de rétrodiffusion mesuré par les altimètres sur l'ensemble du continent, sujet qui jusqu'à aujourd'hui a reçu peu d'attention. Les altimètres radars embarqués à bord d'ENVISAT (bandes S et Ku) et de SARAL/AltiKa (bande Ka) ont des sensibilités différentes aux propriétés de la neige. Nous nous sommes intéressés aux caractéristiques annuelles et inter-annuelles des coefficients de rétrodiffusion dans ces trois bandes. Une étude de sensibilité a été réalisée avec un modèle électromagnétique afin de déterminer les propriétés du manteau neigeux qui dominent le signal saisonnier. On montre que le signal saisonnier est sensible à la densité et la rugosité de surface dans la bande S, à la température de la neige dans la bande Ka et à l'une ou à l'autre de ces variables selon la région dans la bande Ku. Les caractéristiques saisonnières du coefficient de rétrodiffusion sont ensuite comparées à celles des températures de brillance acquises par les radiomètres à bord de SARAL et de SSM/I. Les résultats indiquent une influence significative de la rugosité de surface sur les températures de brillance de la bande Ka, influence souvent considérée négligeable dans la modélisation de la température de brillance. Cette étude apporte une meilleure connaissance de la dynamique saisonnière des propriétés de proche surface de la calotte Antarctique. Elle fournit de nouveaux indices pour développer dans le futur des algorithmes robustes de correction de l'erreur de pénétration. Elle met également en lumière l'importance des missions altimétriques multi-fréquences et les possibilités qu'offrent le signal de la bande S pour l'étude des variabilités saisonnières de la rugosité de surface. En définitive, la rugosité de surface est un paramètre important à prendre en compte pour obtenir de meilleures estimations et modélisations des coefficients de rétrodiffusion et des températures de brillance. / In the context of global climate changes, the Antarctic ice sheet contribution to sea-level rise is one of the main uncertainty sources. The extent and extreme meteorological conditions of this continent render remote sensing a useful tool for long term monitoring. Altimetry and radiometry observations in the microwave range reveal variations of the volume of the ice sheet and surface properties of the snowpack. Radar altimeters, provide repeated observations of the surface topography elevation, which allow the quantification of volume variations of the ice sheet. However, the penetration of radar waves in dry and cold snowpack adversely affects the estimated surface elevation. Approaches to minimize the penetration error are all based on a relationship with the backscattering coefficient. Understanding the annual and interannual variations of the backscattering coefficient is thus a key issue in order to improve the estimation accuracy of the surface elevation and to refine the ice-sheet volume trend. This thesis aims at studying the backscattering coefficients acquired by radar altimeters, which until now have received little attention. Radar altimeters on board ENVISAT (S and Ku bands) and SARAL/AltiKa (Ka band) have different sensitivities to the snowpack properties. The annual and interannual variations of the backscattering coefficient at the three bands is investigated. Sensitivity tests are carried out with an electromagnetic model to determine the prevailing snowpack properties that drive the signal. The seasonal signal is sensitive to surface density and roughness at S band, to snow temperature at Ka band and to either snow surface density and roughness or temperature depending on the location on the continent at Ku band. The seasonal signal of the backscattering coefficient is then compared with that of the brightness temperature measured by radiometers on SARAL and SSM/I. The results show a significant influence of surface roughness on brightness temperatures at Ka band, which has often been neglected in brightness temperature modeling studies. This thesis provides a better understanding of the seasonal dynamics of the near surface properties of the Antarctic ice sheet. It also provides new clues to build a more robust corrections of the penetration errors in the future. It highlights the importance of multi-frequency altimetry missions and the potential of the S band to study the seasonal variability in surface roughness. In summary, surface roughness is an important property which should be taken into account for a better modeling of backscattering coefficient and brightness temperature.

The West Antarctic Ice Sheet is drained primarily by five major ice streams, which together control the volume of ice discharged into the ocean across the grounding line. The grounding line of Kamb Ice Stream (KIS) is unusual because the ice stream upstream of it is stagnant. Here, a set of surface features--shore-parallel, long wavelength, low amplitude undulations--found only at that grounding line are examined and found to be "pinch and swell" features formed by an instability in the viscous deformation of the ice. When a relatively competent layer is surrounded by lower strength materials, particular wavelength features within the layer may be amplified under certain layer thickness and strain rate conditions. The undulations at KIS grounding line are possible due to the relatively large strain rates and particular ice thickness at that location. Several data sets are used to characterize the surface features. High resolution surface profiles are created using kinematic GPS carried on board a sled that was used to tow ice penetrating radar equipment. The radar data are used to examine the relationship between surface shape and basal crevasses. Additional surface profiles are created using ICESat laser altimeter observations. Repeat GPS surveys of a strain grid across the grounding line yields strain rate information. Analysis of repeat observations over tidal cycles and multi-day intervals shows that the features are not standing or traveling waves. Together, these observations are then used to evaluate the contributions of elastic and viscous deformation of the ice in creating the grounding line undulations.

Die neuen Schwerefeld-Satellitenmissionen CHAMP (Challenging Minisatellite Payload), GRACE (Gravity Recovery and Climate Experiment) und GOCE (Gravity Field and Steady-State Ocean Circulation Explorer) können wesentlich zur Erforschung und Beobachtung des Systems Erde beitragen. Die Antarktis als ein Schlüsselglied im globalen Klimasystem bietet dabei besondere Herausforderungen. GRACE hat hier das Potential, zeitliche Massenänderungen (unter anderem der Eismasse) zu beobachten. Methoden zur Auswertung der Missionsdaten befinden sich gegenwärtig in einem intensiven Entwicklungsprozess, zu dem die vorliegende Arbeit beitragen soll. Inhaltlicher Schwerpunkt ist die Nutzung von GRACE zur Ermittlung zeitlicher Massenvariationen in der Antarktis. Die Analysen erfolgen in erster Linie aus der Position eines Nutzers von Standard-Missionsprodukten, betreffen aber grundsätzlich den gesamten Auswerteprozess. Nach einer Einführung werden zunächst die Hintergründe der Arbeit ausgeführt (Kapitel 2), speziell die theoretischen Grundlagen zu Massen- und Schwerefeldvariationen, Phänomene geophysikalischer Massenvariationen und die neuen Schwerefeldmissionen mit ihrem Potential zur Beobachtung solcher Massenvariationen. Ein Hauptteil der Arbeit behandelt die Frage, welche Signale und Fehler in den Schwerefeldlösungen der Missionen enthalten sind (Kapitel 3). Zunächst werden dazu gegenwärtige Prozessierungskonzepte zur Erstellung von CHAMP- und GRACE-Schwerefeldlösungen skizziert und die GRACE-Monatslösungen des GeoForschungsZentrums Potsdam als ein Standard-GRACE-Produkt vorgestellt. Es folgen verschiedene Analysen zur Fehlerstruktur der Schwerefeldlösungen, wobei insbesondere die Fehlerstruktur von GRACE-Monatslösungen anhand ihres Zeitverhaltens empirisch untersucht werden. Als eine Ursache empirisch festgestellter, aber durch Fehlermodelle nicht vollständig beschriebener Fehlerstrukturen werden schließlich Alias-Effekte von unmodellierten zeitlichen Variationen auf die geschätzten räumlichen Variationen qualitativ und quantitativ beschrieben und diskutiert. Ein zweiter Hauptteil untersucht geophysikalische Rückschlüsse aus GRACE-Schwerefeldlösungen mit Anwendung auf die Schätzung antarktischer Eismassensignale (Kapitel 4). Methoden zur Schätzung von Massensignalen aus den Schwerefeldlösungen werden systematisch zusammengestellt und teilweise weiterentwickelt. Die praktische Anwendung dieser Methoden zur Schätzung von Eismassenänderungen des Antarktischen Eisschildes und seiner großen Eiseinzugsgebiete wird erklärt. Ein Schwerpunkt liegt auf der Untersuchung der unterschiedlichen Mechanismen, die zu Fehlern der geschätzten Massensignale führen, sowie auf der Abschätzung dieser Fehler. Im Lichte der gewonnenen Einsichten in die methodischen Unsicherheiten der angewandten Analysetechniken erfolgt schließlich die Präsentation und Diskussion der Ergebnisse, einschließlich eines Vergleichs mit bisher veröffentlichten Massenbilanzresultaten. Möglichkeiten zu methodischen Verbesserungen, die in den vorangegangenen Untersuchungen deutlich werden, aber über den Rahmen der Arbeit hinausgehen, werden in einem eigenen Kapitel (Kapitel 5) diskutiert. Dies betrifft sowohl solche Verbesserungen, die bereits auf der Basis der gegenwärtigen GRACE-Monatslösungen möglich sind, als auch Verbesserungen in der Generierung dieser Monatslösungen oder, allgemeiner, in der GRACE-Prozessierung. Die Kombination der GRACE-Daten mit komplementären Beobachtungen und Modellen spielt in den unterschiedlichen Stadien der GRACE-Datenanalyse eine Schlüsselrolle. In Bezug auf die Trennung antarktischer Massensignale werden Kombinationsstrategien nochmals gesondert diskutiert. Schließlich werden die Hauptergebnisse der Arbeit nochmals zusammengefasst und eingeordnet (Kapitel 6). / The new gravity field satellite missions CHAMP (Challenging Minisatellite Payload), GRACE (Gravity Recovery and Climate Experiment) and GOCE (Gravity Field and Steady-State Ocean Circulation Explorer) can provide essential contributions to the study and the monitoring of the Earth system. Hereby, Antarctica as a key element of the climate system offers particular challenges. GRACE has the potential to observe temporal variations of masses such as ice masses, in particular. Methods to analyse the mission data are currently in a process of intensive development. The present work aims to contribute to this process. The focus is on the use of GRACE to determine temporal mass variations in Antarctica. The analyses are carried out from the viewpoint of a standard product user. Nonetheless, they concern the entire process of GRACE data analysis. After an introduction, the background of the work is explained, in particular the theoretical fundamentals of mass and gravity field variations, the phenomena of geophysical mass variations and the new gravity field missions with their potential to observe these variations (chapter 2). One main part of the work (chapter 3) treats the question which signals and errors are contained in the missions' gravity field solutions (chapter 3). Current CHAMP and GRACE processing approaches are outlined. The GRACE monthly solutions by GeoForschungsZentrum Potsdam are introduced. Subsequently, different analyses about error structures of gravity field mission solutions are presented. In particular, an empirical analysis of time-variations of the GRACE monthly solutions reveals error structures which are not completely described by error models. As one cause of this discrepancy, alias effects of unmodelled temporal variations on the spatial patterns of the solutions are discussed qualitatively and quantitatively. Another main part of the work (chapter 4) investigates geophysical inferences from the GRACE monthly solutions, with Antarctica taken as a case study. Methods to estimate mass signals are systematised and partly enhanced. The practical applications of these methods for the estimation of Antarctic ice mass changes is explained. The different error mechanisms are investigated in detail, and corresponding errors are assessed. The results about Antarctic ice mass changes are then presented, compared to previous results and discussed in the light of the remaining methodological uncertainties. The studies reveal directions for methodological improvements, and so, related ideas are discussed in a separate chapter (chapter 5). They concern both the analysis of current GRACE monthly solutions and the generation of these solutions, or, more generally, the GRACE processing. The combination of GRACE data with complementary observations and models plays a key role in the different levels of GRACE data analysis. Combination strategies are, hence, once more discussed with regard to Antarctic mass signals. Finally, the main results of the work are summarised and discussed in a broader context.

Ten sediment cores (2.0-8.5 m long) from various locations in Lake of the Woods (LOTWs) and Shoal Lake (SL) were recovered in August 2006, using a Kullenberg piston corer. From the study of the macrofossils (primarily ostracodes and thecamoebians) and the sediments in six processed cores, variations in paleoconditions were observed both spatially and temporally, and the timing of these changes were identified in over 10,000 years of postglacial history. Ostracodes disappeared from the LOTWs record from about 9000 to 7600 calendar years before present (BP) (about 5800 in SL), after LOTWs became isolated from glacial Lake Agassiz. Thecamoebians appeared in many cores around 2000 calendar years BP, with the earliest appearance at 9200. Buried paleosols in three cores indicate portions of the lake dried on several occasions during the Hypsithermal, perhaps indicating the region’s future climate response. One core contained a pink clay bed indicative of the Marquette readvance about 11,300 years (BP), and the subsequent input of water from the Superior basin.

Ten sediment cores (2.0-8.5 m long) from various locations in Lake of the Woods (LOTWs) and Shoal Lake (SL) were recovered in August 2006, using a Kullenberg piston corer. From the study of the macrofossils (primarily ostracodes and thecamoebians) and the sediments in six processed cores, variations in paleoconditions were observed both spatially and temporally, and the timing of these changes were identified in over 10,000 years of postglacial history. Ostracodes disappeared from the LOTWs record from about 9000 to 7600 calendar years before present (BP) (about 5800 in SL), after LOTWs became isolated from glacial Lake Agassiz. Thecamoebians appeared in many cores around 2000 calendar years BP, with the earliest appearance at 9200. Buried paleosols in three cores indicate portions of the lake dried on several occasions during the Hypsithermal, perhaps indicating the region’s future climate response. One core contained a pink clay bed indicative of the Marquette readvance about 11,300 years (BP), and the subsequent input of water from the Superior basin.

Glaciers cover ca 10% of the Earth's land and are found in the high altitudes and latitudes. They are important components of environmental systems due to the multiple feedbacks linking them with the atmosphere, hydrosphere and periglacial landscapes. The cold sloping surfaces of glaciers change the patterns of atmospheric circulation at different scales and at the same time glaciers are largely controlled by climate. They are commonly used as climatic archives for reconstruction of the past environmental changes based on evidences from the areas affected by glaciation at the moment and in the past. Glaciers are the largest fresh-water reservoirs on our planet and runoff thereof significantly affects the global sea level and life in glaciated catchments. However, melt- and rain-induced runoff from glaciers greatly depends on the subsurface conditions which thus need to be taken into account, particularly in a changing climate. This thesis focuses on the processes of subsurface mass and energy exchange in the accumulation zones of glaciers, which are largely driven by the climate at the surface. Results are largely based on empirical data from Lomonosovfonna ice cap, Svalbard, collected during field campaigns in 2012-2017. Observations of subsurface density and stratigraphy using shallow cores, video records from boreholes and radar surveys returned detailed descriptions of the snow and firn layering. The subsurface temperature data collected using multiple thermistor strings provided insights into several subsurface processes. The temperature values measured during three summer seasons were used to constrain the suggested parameterization of deep preferential water flow through snow and firn. The part of data recorded during the cold seasons was employed for an inverse modelling exercise resulting in optimized values of effective thermal conductivity of the subsurface profile. These results are then used to compute the subsurface water content by comparing the simulated and measured rates of freezing front propagation after the melt season in 2014. The field observations and quantitative estimates provide further empirical evidences of preferential water flow in snow/firn packs at glaciers. Results presented in the thesis call for implementation of description of the process in layered models simulating the subsurface fluxes of energy and mass at glaciers. This will result in a better understanding of glacier response to the past and future climatic changes and more accurate estimates of glacier runoff. / Stability and Variations of Arctic Land Ice (SVALI)

With continued warming regional surface air temperatures around the Artic in recent decades, there is growing importance in understanding how ice sheet dynamics interact with a shifting global climate system. This research investigates the spatial variability of sub-glacial hydrology under land terminating ice sheets in Western GrIS, when applying varying overburden ice pressures within Shreve’s (1972) hydraulic potential equation. The application of ArcGIS is used with adjusted k-values in the equation to route hydrological network systems under the ice sheet and help identify the processes taking place within the sub-glacial system. With focus on processes such as water piracy, we are able illustrate the effects that water at the base can have on ice sheet behaviors. i.e. velocity and mas balance. The findings conclude that the ice sheet is operating under conditions at its base with a k-value somewhere within the range of 0.9 to 1.1. This assumption is based on comparisons between modelled pro-glacial output with observed data taken from studies by Mikkelsen et al. (2014) and Smith et al. (2015), using input melt data at the surface (Lindback et. al., 2015). At this level of hydraulic potential, water piracy is effectively changing the course pathways of the hydrological network and therefore manipulating the size and shape of the sub-glacial catchments. As a result, discharge may leave the glacier from a different location than what would be assumed. Identifying the location and volume of water under particular ice sheets, compared to neighboring ice sheets, can be used to explain spatial and temporal differences in ice sheet characteristics. Such research is important in understanding both environmental and socio-economic implications at local to global scale. Although the application of GIS methodology is extremely useful is producing such results, it must be recognised that a high level of uncertainty and error exists in the data results. / Temperatur kan ha en särskilt stark inverkan på hur istäcken (glaciärer) beter sig. Även om det är väl känt att temperaturerna ökar både globalt och runt Arktis så har vi bara börjat förstå vikten av effekten av detta på glaciärer. Med ökande temperaturer har vi börjat se att en ökad avsmältning och en ökad mängd smältvatten i glaciala system kan ändra sättet en ismassa beter sig. Ett exempel på en sådan förändring är en ökad hastighet som ismassan rör sig i eftersom vatten fungerar som ett glidmedel mellan glaciären och den underliggande marken. Var vattnet finns under isen påverkar var istäcket rör sig. Man tror att en ökning i hastighet vid vissa delar av istäcket kan ledda till att det tunnas ut. Med tiden lämnar en allt större mängd vatten systemet vilket bidrar till mindre, retirerande glaciärer och en höjning av den globala havsytenivån då vattnet via vattendrag till slut rinner ut i havet. I den här studien användes ett kartläggningsprogram, Geografiskt Informationssystem (GIS), för att förutsäga var flodsystem under isen befinner sig. Detta kartläggningsprogram används eftersom dessa regioner inte är tillgängliga för observationer i fält. Tjockleken på istäcket som ligger ovanpå flodsystemet utövar ett tryck på vattnet och gör att det kan flöda emot gravitationen och i riktningar och med hastigheter som inte är typiska för flodsystem i varmare klimat. Denna process kallas vattenavlänkning. I denna studie används en ekvation i GIS för att variera trycket från det ovanliggande istäcket. Genom att ändra detta ändras hur stor vattenavlänkningen blir och därför även vilken väg vattnet tar under isen. Denna teknik tillåter oss att se vilken väg vattnet tar från att det kommer in i systemet vid toppen av istäcket till var det lämnar systemet genom vattendrag nedströms. Genom att veta var vattnet befinner sig kan man utröna varför isen rör sig annorlunda gentemot omkringliggande istäcken, och därför också dess påverkan på omgivande miljö såväl som sociala konsekvenser. Även om detta är en väldigt användbar metod för att kartlägga kanalsystem så finns det osäkerheter, till exempel i hur resultatet stämmer överens med verkliga scenarion. Så även om detta är användbart för att förstå teorin bakom processerna så är resultatet kanske inte helt tillförlitligt.

Mass changes of polar ice sheets have an important societal impact, because they affect global sea level. Estimating the current mass budget of ice sheets is equivalent to determining the balance between the surface mass gain through precipitation and the outflow across the grounding line. In Antarctica, the latter is mainly governed by oceanic processes and outlet glacier dynamics.<p>In this thesis, we assess the mass balance of a part of eastern DronningMaud Land via an input/output method. Input is given by recent surface accumulation estimations of the whole drainage basin. The outflow at the grounding line is determined from the radar data of a recent airborne survey and satellite-based velocities using a flow model of combined plug flow and simple shear. We estimate the regional mass balance in this area to be between 1.88±8.50 and 3.78±3.32 Gt a−1 depending on the surface mass balance (SMB) dataset used. This study also reveals that the plug flow assumption is acceptable at the grounding line of ice streams.<p>The mass balance of drainage basins is governed by the dynamics of their outlet glaciers and more specifically the flow conditions at the grounding line. Thanks to an airborne radar survey we define the bed properties close to the grounding line of the West Ragnhild Glacier (WRG) in the Sør Rondane Mountains. Geometry and reflectivity analyses reveal that the bed of the last 65 km upstream of the grounding line is sediment covered and saturated with water. This setting promotes the dominance of basal motion leading to a change in the flow regime: in the interior flow is governed by internal deformation while its relative importance decreases to become driven by basal sliding.<p>Subsequently we present the results of the reconstruction of the SMB across an ice rise through radar data and inverse modelling. The analysis demonstrates that atmospheric circulation was stable during the last millennium. Ice rises induce an orographic uplift of the atmospheric flow and therefore influence the pattern of the SMB across them, resulting in an asymmetric SMB distribution. Since the geometry of the internal reflection horizons observed in radar data depends on the SMB pattern, the asymmetry observed in radar layers reveals the trajectories of air masses at the time of deposit. We present an original and robust method to quantify this SMB distribution. Combining shallow and deep radar layers, SMB across Derwael Ice Rise is reconstructed. Two methods are employed as a function of the depth of the layers: i.e. the shallow layer approximation for the surface radar layers and an optimization technique based on an ice flow model for the deeper ones. Both methods produce similar results. We identify a difference in SMB magnitude of 2.5 between the flanks and the ice rise divide, as well as a shift of ≈4 km between the SMB maximum and the crest. Across the ice rise, SMB exhibits a very large variability, ranging from 0.3 to 0.9 mw.e. a−1. This anomaly is robust in time.<p>Finally we draw a comprehensive description of the Sør Rondane Mountains sector. The glacial system is close to the equilibrium and seems stable but evidences suggest that it is a fragile equilibrium. The proximity of the open ocean certainly favours the interaction between warm water and the ice shelf cavity conducting to potential important melting. The thinning associated with this melting can detach the ice shelf from pinning points. This will reduce the buttressing from the ice shelf, outlet glaciers will accelerate and mass transfer toward the ocean will increase. Therefore, the future of Antarctic Ice Sheet directly depends on the changes affecting its boundaries and assessing the sensitivity of the ice sheets is essential to quantify and anticipate the future variation of mass balance. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

Die neuen Schwerefeld-Satellitenmissionen CHAMP (Challenging Minisatellite Payload), GRACE (Gravity Recovery and Climate Experiment) und GOCE (Gravity Field and Steady-State Ocean Circulation Explorer) können wesentlich zur Erforschung und Beobachtung des Systems Erde beitragen. Die Antarktis als ein Schlüsselglied im globalen Klimasystem bietet dabei besondere Herausforderungen. GRACE hat hier das Potential, zeitliche Massenänderungen (unter anderem der Eismasse) zu beobachten. Methoden zur Auswertung der Missionsdaten befinden sich gegenwärtig in einem intensiven Entwicklungsprozess, zu dem die vorliegende Arbeit beitragen soll. Inhaltlicher Schwerpunkt ist die Nutzung von GRACE zur Ermittlung zeitlicher Massenvariationen in der Antarktis. Die Analysen erfolgen in erster Linie aus der Position eines Nutzers von Standard-Missionsprodukten, betreffen aber grundsätzlich den gesamten Auswerteprozess. Nach einer Einführung werden zunächst die Hintergründe der Arbeit ausgeführt (Kapitel 2), speziell die theoretischen Grundlagen zu Massen- und Schwerefeldvariationen, Phänomene geophysikalischer Massenvariationen und die neuen Schwerefeldmissionen mit ihrem Potential zur Beobachtung solcher Massenvariationen. Ein Hauptteil der Arbeit behandelt die Frage, welche Signale und Fehler in den Schwerefeldlösungen der Missionen enthalten sind (Kapitel 3). Zunächst werden dazu gegenwärtige Prozessierungskonzepte zur Erstellung von CHAMP- und GRACE-Schwerefeldlösungen skizziert und die GRACE-Monatslösungen des GeoForschungsZentrums Potsdam als ein Standard-GRACE-Produkt vorgestellt. Es folgen verschiedene Analysen zur Fehlerstruktur der Schwerefeldlösungen, wobei insbesondere die Fehlerstruktur von GRACE-Monatslösungen anhand ihres Zeitverhaltens empirisch untersucht werden. Als eine Ursache empirisch festgestellter, aber durch Fehlermodelle nicht vollständig beschriebener Fehlerstrukturen werden schließlich Alias-Effekte von unmodellierten zeitlichen Variationen auf die geschätzten räumlichen Variationen qualitativ und quantitativ beschrieben und diskutiert. Ein zweiter Hauptteil untersucht geophysikalische Rückschlüsse aus GRACE-Schwerefeldlösungen mit Anwendung auf die Schätzung antarktischer Eismassensignale (Kapitel 4). Methoden zur Schätzung von Massensignalen aus den Schwerefeldlösungen werden systematisch zusammengestellt und teilweise weiterentwickelt. Die praktische Anwendung dieser Methoden zur Schätzung von Eismassenänderungen des Antarktischen Eisschildes und seiner großen Eiseinzugsgebiete wird erklärt. Ein Schwerpunkt liegt auf der Untersuchung der unterschiedlichen Mechanismen, die zu Fehlern der geschätzten Massensignale führen, sowie auf der Abschätzung dieser Fehler. Im Lichte der gewonnenen Einsichten in die methodischen Unsicherheiten der angewandten Analysetechniken erfolgt schließlich die Präsentation und Diskussion der Ergebnisse, einschließlich eines Vergleichs mit bisher veröffentlichten Massenbilanzresultaten. Möglichkeiten zu methodischen Verbesserungen, die in den vorangegangenen Untersuchungen deutlich werden, aber über den Rahmen der Arbeit hinausgehen, werden in einem eigenen Kapitel (Kapitel 5) diskutiert. Dies betrifft sowohl solche Verbesserungen, die bereits auf der Basis der gegenwärtigen GRACE-Monatslösungen möglich sind, als auch Verbesserungen in der Generierung dieser Monatslösungen oder, allgemeiner, in der GRACE-Prozessierung. Die Kombination der GRACE-Daten mit komplementären Beobachtungen und Modellen spielt in den unterschiedlichen Stadien der GRACE-Datenanalyse eine Schlüsselrolle. In Bezug auf die Trennung antarktischer Massensignale werden Kombinationsstrategien nochmals gesondert diskutiert. Schließlich werden die Hauptergebnisse der Arbeit nochmals zusammengefasst und eingeordnet (Kapitel 6). / The new gravity field satellite missions CHAMP (Challenging Minisatellite Payload), GRACE (Gravity Recovery and Climate Experiment) and GOCE (Gravity Field and Steady-State Ocean Circulation Explorer) can provide essential contributions to the study and the monitoring of the Earth system. Hereby, Antarctica as a key element of the climate system offers particular challenges. GRACE has the potential to observe temporal variations of masses such as ice masses, in particular. Methods to analyse the mission data are currently in a process of intensive development. The present work aims to contribute to this process. The focus is on the use of GRACE to determine temporal mass variations in Antarctica. The analyses are carried out from the viewpoint of a standard product user. Nonetheless, they concern the entire process of GRACE data analysis. After an introduction, the background of the work is explained, in particular the theoretical fundamentals of mass and gravity field variations, the phenomena of geophysical mass variations and the new gravity field missions with their potential to observe these variations (chapter 2). One main part of the work (chapter 3) treats the question which signals and errors are contained in the missions' gravity field solutions (chapter 3). Current CHAMP and GRACE processing approaches are outlined. The GRACE monthly solutions by GeoForschungsZentrum Potsdam are introduced. Subsequently, different analyses about error structures of gravity field mission solutions are presented. In particular, an empirical analysis of time-variations of the GRACE monthly solutions reveals error structures which are not completely described by error models. As one cause of this discrepancy, alias effects of unmodelled temporal variations on the spatial patterns of the solutions are discussed qualitatively and quantitatively. Another main part of the work (chapter 4) investigates geophysical inferences from the GRACE monthly solutions, with Antarctica taken as a case study. Methods to estimate mass signals are systematised and partly enhanced. The practical applications of these methods for the estimation of Antarctic ice mass changes is explained. The different error mechanisms are investigated in detail, and corresponding errors are assessed. The results about Antarctic ice mass changes are then presented, compared to previous results and discussed in the light of the remaining methodological uncertainties. The studies reveal directions for methodological improvements, and so, related ideas are discussed in a separate chapter (chapter 5). They concern both the analysis of current GRACE monthly solutions and the generation of these solutions, or, more generally, the GRACE processing. The combination of GRACE data with complementary observations and models plays a key role in the different levels of GRACE data analysis. Combination strategies are, hence, once more discussed with regard to Antarctic mass signals. Finally, the main results of the work are summarised and discussed in a broader context.