Speckle tracking of synthetic aperture RADAR imagery (Radarsat-1/2, ALOS PALSAR) and feature tracking of optical (Landsat-7 ETM+) imagery is used to determine the entire surface velocity structure of the major ice masses of the Canadian High Arctic in 2000, 2010-2015 and for select tidewater terminating glaciers from 1999-2010. At the termini of tidewater glaciers, surface ice velocities are combined with measured/modelled ice thicknesses to derive an estimate of mass loss via dynamic (iceberg) discharge. The total dynamic discharge for the ice masses of the southern Canadian Arctic Archipelago (SCAA: Baffin and Bylot Islands) is between ~17 and 180 Mt a-1 (0.017 to 0.180 Gt a-1) for the period 2007-2011, compared to a dynamic discharge of ~2.47  ± 0.88 Gt a-1 for the northern Canadian Arctic Archipelago (NCAA: Devon, Ellesmere, Axel Heiberg Islands) for the period 2011-2015. A comparison of these values with rates of mass loss via climatic mass balance (surface melt and runoff) indicates that dynamic discharge accounted for ~3.1% of total ablation for the NCAA in 2012 and ~0.11% of total ablation in the SCAA between 2007 and 2010. This reveals that total ablation in the Canadian Arctic is currently dominated by surface melt and runoff. The glacier velocity dataset provides the most comprehensive record of ice motion and dynamic discharge in the Canadian Arctic to date and reveals a large degree of variability in glacier motion within the region over the last ~15 years. Most of the major glaciers in the NCAA have decelerated and their resultant dynamic discharge has decreased over the observation period, which is largely attributed to cyclical phases attributed to surging and pulsing. On pulse-type glaciers, variation in ice motion is largely confined to regions where the bed is located below sea level. A notable departure from the overall trend of regional velocity slowdown is the widespread acceleration of the Trinity and Wykeham Glaciers of the Prince of Wales Icefield (the largest glacier complex in the Canadian Arctic), which cannot be explained by surge or pulse mechanisms. The increased discharge from these two glaciers nearly compensates (within error) for the decrease in iceberg discharge from other glaciers across the study region and indicates that total dynamic discharge from the Canadian Arctic can be sensitive to the variations of ice flow of just a few glaciers.

Since the 1960s, warming air and sea surface temperatures have led to decreasing sea ice extent and longer periods of open water in the Canadian Arctic Archipelago (CAA). Recent and rapid changes have also been observed in the ice discharge patterns of glaciers in this region. For example, Trinity and Wykeham glaciers on the Prince of Wales Icefield (POW), SE Ellesmere Island, contributed ~62% of total ice discharge to the ocean from the Canadian Arctic Archipelago in 2016, compared to ~22% in 2000. Given these changes, an important question is whether there is a relationship between changing sea ice conditions (e.g., extent, freeze up dates, break up dates) and iceberg production from these glaciers. This study used synthetic aperture radar (Radarsat-1, 2 and ALOS PALSAR) and optical (Landsat-7 and 8) imagery to identify iceberg plume events and sea ice break-up/freeze-up dates between 1997 and 2015 for 40 tidewater glaciers around the POW. Results show a clear relationship between the presence of sea ice and the production of icebergs from glaciers, with most events occurring during the open water season and fewer when sea ice was present. While there have not been clear increasing trends of icebergs produced from all glaciers in the POW, Trinity and Wykeham glaciers show that increases in detected iceberg plumes coincide with increases in previously measured glacier velocity and significant terminus retreat. Comparison to ocean temperature, surface air temperature from NCEP/NCAR reanalysis and tidal data showed no clear relationship with increased calving events, however further research into all factors is recommended. It is likely that there are several factors contributing to the spatial and temporal variability of iceberg production from the POW.

No description available.

No description available.

No description available.

Speckle é um ruído multiplicativo e aleatório, característico de imagens de radar de abertura sintética (SAR). Devido a esse ruído, até mesmo áreas com feições marcantes em superfícies contínuas (e.g., fendas nas superfícies das geleiras) são caracterizadas em imagens SAR por grande variabilidade nos números digitais e pelo efeito "sal e pimenta", comum nesse tipo de imagens. Portanto, o speckle deve ser reduzido para que as imagens SAR possam ser utilizadas em algoritmos de correlação cruzada com o objetivo de extrair informações sobre a velocidade das geleiras. Para solucionar este problema, quatro formas de utilização de um filtro adaptativo (i.e., filtro Lee) foram testadas para o pré-processamento de imagens antes da extração de vetores de velocidade de geleiras. O filtro Lee foi utilizado de duas formas: (i) uma filtragem e (ii) filtragem sucessiva (i.e., dupla filtragem). Além disso, dois parâmetros foram utilizados para informar a variabilidade dos dados: o número de looks da cena e o desvio padrão da cena. A análise dos resultados foi realizada comparando os vetores de velocidade gerados pelas imagens originais e filtradas com dados publicados sobre a dinâmica das geleiras da parte setentrional da Península Antártica. Em termos de supressão do speckle, todos os métodos produziram resultados positivos. No entanto, a dupla filtragem não preservou as bordas das fendas, fundindo as feições. Dessa forma, produtos com dupla filtragem foram descartados da análise final. Em geral, as imagens com uma filtragem apresentam melhores resultados na extração de vetores de velocidade por algoritmos de correlação cruzada que as imagens originais. Assim, a cadeia de pré-processamento incluindo uma só filtragem foi escolhida para a extração de parâmetros dinâmicos de geleiras. Quando comparados com dados já publicados, os vetores de velocidade resultantes da análise mostram um ligeiro aumento na velocidade das geleiras da área de estudo entre 2001 e 2005. / Speckle is a characteristic random noise from coherent imaging systems like synthetic aperture radar (SAR). Due to this noise, even areas with sharp features in continuous surfaces (e.g., crevasses on glaciers) will be characterized in SAR images by grainy texture and high variation in digital numbers. Therefore, the speckle must be reduced before SAR images can be used for measuring glacier velocity by image crosscorrelations algorithms. To solve this problem, four approaches based on adaptive filtering (i.e., Lee filter) were tested for data pre-processing prior to extracting the velocity fields from glaciers. The Lee filter was used in two ways: (i) one-pass and (ii) two-pass filtering. Furthermore, two parameters were used to explain the data variability: number of looks and standard deviation of the scene. Results evaluation was carrying out comparing the velocity vectors resulting from original and filtered images with published data on the dynamics of the glaciers in the northern part of the Antarctic Peninsula. In terms of speckle suppression, all approaches yielded positive results. However, the two-pass filter does not preserve the crevasses edges and the resulting images are not considered for the final result comparison. In general, images processed with one-pass filter showed better results for extraction of velocity vectors with the cross-correlation algorithm than the original ones, and were accepted for an automatic processing chain to derive dynamic parameters of glaciers. Furthermore, resulting velocity vectors agree with published data and show a slight increase in velocity between 2001 and 2005.

Speckle é um ruído multiplicativo e aleatório, característico de imagens de radar de abertura sintética (SAR). Devido a esse ruído, até mesmo áreas com feições marcantes em superfícies contínuas (e.g., fendas nas superfícies das geleiras) são caracterizadas em imagens SAR por grande variabilidade nos números digitais e pelo efeito "sal e pimenta", comum nesse tipo de imagens. Portanto, o speckle deve ser reduzido para que as imagens SAR possam ser utilizadas em algoritmos de correlação cruzada com o objetivo de extrair informações sobre a velocidade das geleiras. Para solucionar este problema, quatro formas de utilização de um filtro adaptativo (i.e., filtro Lee) foram testadas para o pré-processamento de imagens antes da extração de vetores de velocidade de geleiras. O filtro Lee foi utilizado de duas formas: (i) uma filtragem e (ii) filtragem sucessiva (i.e., dupla filtragem). Além disso, dois parâmetros foram utilizados para informar a variabilidade dos dados: o número de looks da cena e o desvio padrão da cena. A análise dos resultados foi realizada comparando os vetores de velocidade gerados pelas imagens originais e filtradas com dados publicados sobre a dinâmica das geleiras da parte setentrional da Península Antártica. Em termos de supressão do speckle, todos os métodos produziram resultados positivos. No entanto, a dupla filtragem não preservou as bordas das fendas, fundindo as feições. Dessa forma, produtos com dupla filtragem foram descartados da análise final. Em geral, as imagens com uma filtragem apresentam melhores resultados na extração de vetores de velocidade por algoritmos de correlação cruzada que as imagens originais. Assim, a cadeia de pré-processamento incluindo uma só filtragem foi escolhida para a extração de parâmetros dinâmicos de geleiras. Quando comparados com dados já publicados, os vetores de velocidade resultantes da análise mostram um ligeiro aumento na velocidade das geleiras da área de estudo entre 2001 e 2005. / Speckle is a characteristic random noise from coherent imaging systems like synthetic aperture radar (SAR). Due to this noise, even areas with sharp features in continuous surfaces (e.g., crevasses on glaciers) will be characterized in SAR images by grainy texture and high variation in digital numbers. Therefore, the speckle must be reduced before SAR images can be used for measuring glacier velocity by image crosscorrelations algorithms. To solve this problem, four approaches based on adaptive filtering (i.e., Lee filter) were tested for data pre-processing prior to extracting the velocity fields from glaciers. The Lee filter was used in two ways: (i) one-pass and (ii) two-pass filtering. Furthermore, two parameters were used to explain the data variability: number of looks and standard deviation of the scene. Results evaluation was carrying out comparing the velocity vectors resulting from original and filtered images with published data on the dynamics of the glaciers in the northern part of the Antarctic Peninsula. In terms of speckle suppression, all approaches yielded positive results. However, the two-pass filter does not preserve the crevasses edges and the resulting images are not considered for the final result comparison. In general, images processed with one-pass filter showed better results for extraction of velocity vectors with the cross-correlation algorithm than the original ones, and were accepted for an automatic processing chain to derive dynamic parameters of glaciers. Furthermore, resulting velocity vectors agree with published data and show a slight increase in velocity between 2001 and 2005.

Speckle é um ruído multiplicativo e aleatório, característico de imagens de radar de abertura sintética (SAR). Devido a esse ruído, até mesmo áreas com feições marcantes em superfícies contínuas (e.g., fendas nas superfícies das geleiras) são caracterizadas em imagens SAR por grande variabilidade nos números digitais e pelo efeito "sal e pimenta", comum nesse tipo de imagens. Portanto, o speckle deve ser reduzido para que as imagens SAR possam ser utilizadas em algoritmos de correlação cruzada com o objetivo de extrair informações sobre a velocidade das geleiras. Para solucionar este problema, quatro formas de utilização de um filtro adaptativo (i.e., filtro Lee) foram testadas para o pré-processamento de imagens antes da extração de vetores de velocidade de geleiras. O filtro Lee foi utilizado de duas formas: (i) uma filtragem e (ii) filtragem sucessiva (i.e., dupla filtragem). Além disso, dois parâmetros foram utilizados para informar a variabilidade dos dados: o número de looks da cena e o desvio padrão da cena. A análise dos resultados foi realizada comparando os vetores de velocidade gerados pelas imagens originais e filtradas com dados publicados sobre a dinâmica das geleiras da parte setentrional da Península Antártica. Em termos de supressão do speckle, todos os métodos produziram resultados positivos. No entanto, a dupla filtragem não preservou as bordas das fendas, fundindo as feições. Dessa forma, produtos com dupla filtragem foram descartados da análise final. Em geral, as imagens com uma filtragem apresentam melhores resultados na extração de vetores de velocidade por algoritmos de correlação cruzada que as imagens originais. Assim, a cadeia de pré-processamento incluindo uma só filtragem foi escolhida para a extração de parâmetros dinâmicos de geleiras. Quando comparados com dados já publicados, os vetores de velocidade resultantes da análise mostram um ligeiro aumento na velocidade das geleiras da área de estudo entre 2001 e 2005. / Speckle is a characteristic random noise from coherent imaging systems like synthetic aperture radar (SAR). Due to this noise, even areas with sharp features in continuous surfaces (e.g., crevasses on glaciers) will be characterized in SAR images by grainy texture and high variation in digital numbers. Therefore, the speckle must be reduced before SAR images can be used for measuring glacier velocity by image crosscorrelations algorithms. To solve this problem, four approaches based on adaptive filtering (i.e., Lee filter) were tested for data pre-processing prior to extracting the velocity fields from glaciers. The Lee filter was used in two ways: (i) one-pass and (ii) two-pass filtering. Furthermore, two parameters were used to explain the data variability: number of looks and standard deviation of the scene. Results evaluation was carrying out comparing the velocity vectors resulting from original and filtered images with published data on the dynamics of the glaciers in the northern part of the Antarctic Peninsula. In terms of speckle suppression, all approaches yielded positive results. However, the two-pass filter does not preserve the crevasses edges and the resulting images are not considered for the final result comparison. In general, images processed with one-pass filter showed better results for extraction of velocity vectors with the cross-correlation algorithm than the original ones, and were accepted for an automatic processing chain to derive dynamic parameters of glaciers. Furthermore, resulting velocity vectors agree with published data and show a slight increase in velocity between 2001 and 2005.

Surge-type glaciers experience cyclical variations between long periods of slow flow, followed by shorter periods of rapid flow. These glaciers have been thoroughly analysed in many regions, but detailed studies of surging in the Canadian Arctic are lacking. This thesis provides the first comprehensive reconstruction of the dynamics of Iceberg and Airdrop glaciers, located on western Axel Heiberg Island, and reveals detailed observations of a surge for the first time in the Canadian Arctic. A variety of remotely sensed data, including historical aerial photographs, declassified intelligence satellite photographs, optical satellite imagery (e.g., ASTER, Landsat), and synthetic aperture radar data (e.g., ERS-1, ERS-2) were used to quantify changes in terminus position, ice velocity, and ice thickness since the 1950s. A surge initiated at the terminus of Iceberg Glacier in 1981 and terminated in 2003, suggesting an active phase length of 22 years. High surface velocities, peaking at ~2300 m a⁻¹ at the terminus in summer 1991, were accompanied by a terminus advance of >7 km over the period 1981-1997 and a large transfer of mass down-glacier, causing significant median surface elevation changes reaching >3 ± 1 m a⁻¹ across the entire trunk width. The ensuing quiescent period has seen a continual decrease in flow rates to an average centreline velocity of 11.5 m a⁻¹ in 2020-2021, a gradual steepening of the glacier surface, and a terminus retreat of >2.5 km. Observations on Airdrop Glacier show a continuous advance totalling ~6 km since 1950 and notably less variability in its surface velocities in comparison to Iceberg Glacier. This advance can be attributed to consistently high flow rates of Airdrop’s entire surface, resulting in significant thickening near its terminus since at least 1977. However, velocities have more than halved within the last 15 years, but without any clear evidence of previous fast flow events, we cannot confirm whether Airdrop’s behaviour is cyclic in nature and therefore characteristic of a surge. Instead, Airdrop Glacier could be experiencing a delayed response to positive mass balance conditions of the Little Ice Age, and its recent slowdown could be indicative of a gradual adjustment to recent climatic conditions.

The aim of this study is to understand the extent, depth, magnitude and significance of subglacial sediment deformation. It will examine the role of this deformation in controlling glacier dynamics and landform generation in glaciers in general, and polythermal glaciers in particular. A detailed multi-dimensional approach is used to study recently exposed glacigenic sediments on the forefields of three polyglaciers in the Tarfala Valley, northern Sweden. Overridden fluted moraines and diamicton plains occur in each forefield. These palimpsest landforms consist of multiple subglacial traction tills. Flutes have quasi-regular geometry and about half of those studied have no initiating boulder. It is suggested here that flute formation by forced-mechanisms was superimposed on flute formation related to a topographically-induced flow instability. In each forefield the depth of the deforming-bed averaged between 0.2m and 0.6m thickness. Detailed clast fabric data suggest the diamicton plain is composed of thin layers of traction tills that accreted over time as the zone of deformation moved upwards. Laboratory shear box tests show that subglacial deformation required elevated pore-water pressures, which suggests deforming-bed conditions and flute formation were restricted to the temperate zones of polythermal glaciers. Magnetic fabrics suggest strain magnitudes were moderate (≤10), rather than the very high strain magnitudes (>10²) required by the deforming-bed model. The application of the micro-structural mapping technique demonstrates that subglacial deformation was multi-phase, heterogeneous, and partitioned into the softer and more easily deformed parts of the matrix. Consequently, deformation is controlled by variations in sediment granulometry and pore-water pressure, and is likely to have been spatially and temporally variable, a finding that supports the ice-bed mosaic model. The strain magnitudes and deforming-bed thickness suggest that soft-bed deformation did not exert a major control on glacier dynamics during the Little Ice Age advance.