NS11A-0152
Development and First Results of a new Airplane Based Fixed Wing Electromagnetic Induction Sea Ice Thickness Sounder
Regular observation of Arctic and Antarctic sea ice thickness is of high importance for a better understanding of processes of climate change in polar regions. For regular and accurate observations of polar sea ice thickness a long range airborne device is necessary. Airborne electromagnetic induction (AEM) sounding was found to be an ideal method for accurate and wide area sea ice thickness measurements. As a consequence of five years of successful helicopter electromagnetic (HEM) sea ice thickness measurements and to overcome helicopter range restrictions, the Alfred Wegener Institute (AWI) constructed a new airplane based fixed wing EM system. The first test flights were carried out in 2006 over the North Sea and in April 2007 in Svalbard, where the system's performance was proven under arctic conditions. The system operates in frequency domain with 1990 Hz and a vertical coplanar coil configuration. Thus the system produces a horizontal dipole. The coils are mounted beneath the wings with a separation of 11.6 meters. The airplane, a Dornier 228, is also equipped with a laser altimeter to determine the altitude of the instrument with an accuracy of 2cm. The compensation of the transmitter signal at the receiver coil is done electronically. Flights over open sea are used for the calibration of the system, because the ocean functions as a homogeneous half space with well known conductivity. A data acquisition computer records four voltages with a sample rate of 10 Hz. These are the reference voltage of the transmitter, the compensated and raw receiver voltages and the compensation signal. The laser height is recorded with a sample rate of 100 Hz to account for surface roughness. EM instruments for sea ice thickness sounding should have a vertical resolution of 10cm but due to the electrical noise caused by the airplane engines this was not easy to achieve. To account for the noise a time average filter is used. Alternatively, in order to keep the original horizontal resolution a correction with the reference voltage, which includes the noise but not the ocean’s EM signal, is applied.
NS11A-0153
Exploring the Possibilities of Passive Seismic Interferometry to Image Temperate Glaciers
To aid the general scope of a larger project studying hydrological processes occurring within and beneath glaciers, we installed a small passive seismic network on the Bench glacier, near Valdez, AK. In conjunction with active-source seismic and GPR surveys, passive data is recorded over a period of 10 days along two separate receiver arrays. Four receivers are placed on a line parallel to the glacier's movement, while six receivers form a line perpendicular to the first, traversing the glacier. Seismic interferometry is used to extract Green functions between receiver pairs. In this work we explore the use of rock falls from the surrounding mountains, water traveling within the glacier and at the glacier bed, and fracturing of the glacier ice as possible seismic sources to image the englacial and subglacial systems. Aided by the low intrinsic seismic energy loss of ice and a favorable distribution of natural sources, it may be possible to use passive seismic imaging to help us understand key glacial processes such as fracture formation and water movement from the surface to the bed of the glacier. http://pal.boisestate.edu/mediawiki/index.php?title=Bench07_interferometry&action=edit
NS11A-0154
Passive seismic signatures of a fast-flowing Alpine glacier
A multi-disciplinary shallow geophysics experiment was carried out near a prominent break in surface slope of the Grubengletscher, Valais, Switzerland, during the summer melt season of 2007. High-resolution reflection and refraction seismics, differential GPS, passive seismology, seismoelectric sounding and GPR profiling were undertaken. Here we present the initial results of the passive seismic experiment. An array of 6 high-frequency SAQS data loggers with 4.5Hz 3-component geophones was deployed on the surface of the glacier, around the break-of-slope, near the terminus of the glacier. An inter-station distance of around 50m was used, comparable to the ice thickness, with one instrument deployed at the centre of a ring of 5. Data were recorded continuously at 3000sps with GPS timing. Due to the rapid ablation rate the array required daily maintenance involving the re-seating of recording equipment, batteries and solar panels, as well as redeployment of the geophones to maintain correct orientation. Geophones were buried in small pits with a flat base chipped out of the ice with an ice-axe. Holes were then drilled for the three spikes of the sensor casing to achieve maximum coupling. The sensors were then covered with ice and overlain by rocks to inhibit surface melt. With no snow cover at the surface, deployment of the geophones became the most critical aspect of the deployment. Spurious resonances were observed in the data as a result of significant tilt of the geophones. With surface-melt in the cm/day range this kind of deployment would not be possible without daily attention. Dozens of events per hour are observed on all stations at all times of day and night. These signals encompass surface events such as rockfalls, and also crevassing and basal events. To allow the characterisation and identification of signal waveforms, the timing of known events such as rockfalls, footfalls, sledgehammer blows and "felt" glacial events were recorded. Initial results indicate a range of event types from simple impulsive crevasse-like events to highly complex large-amplitude events associated with rock-falls. When combined with the other geophysical data, the extraction of basal events from the plethora of data will allow the subglacial mechanism sustaining the flow instability of the glacier to be isolated. The findings are significant because the flow velocity of the Grubengletscher can reach some 45 m / year, which is fast for an Alpine glacier; thus, the glacier is anticipated to serve as a small-scale analogue for fast-flowing outlet glaciers from larger ice caps or ice sheets.
NS11A-0155
Analysis of Regionally Detected Icequakes Using the STEEP Network, South-Central AK
Glaciers produce seismic energy that is detectable from local to teleseismic distances. Glaciolgical processes including calving, surface crevassing, basal sliding and other, yet unresolved source processes are capable of producing recordable seismicity. Twenty-two broadband sensors deployed in south-central Alaska during the SainT Elias TEctonics and Erosion Project (STEEP) provide an excellent means to study glacier-generated seismicity at regional distances. These instruments surround over 7500 km2 of glacier area including the Bering Glacier, Bagley Icefield and the tidewater calving glaciers of Icy Bay (Yahtse, Guyot, Tyndal). Our analysis shows that icequakes nominally occur several times hourly, and can be separated from tectonic seismicity using their unique spectral characteristics and hypocenter locations. The events typically propagate over 50-75 km distances, but occasionally are recorded at stations over 150 km away from the energy source. Hypocenters for more than 1000 events were manually calculated through a 26-day interval during October 2006, and suggest that a majority of the icequakes are associated with calving at tidewater glaciers that terminate in Icy Bay. Events with similar time and frequency domain characteristics also occur at locations away from calving fronts, but less often, and their mechanical origin remains undetermined. Automated detections from a frequency domain event detector exhibit strong correlation with the handpicked time series, and extend our analysis to all available data collected during 2006. We present the time distribution of several categories of icequakes and compare these distributions to environmental variables such as precipitation, temperature and tides to explore potential forcing for observed variability in icequake occurrence.
NS11A-0156
Radar, electromagnetic, and active seismic investigation of a propagating ice shelf rift tip
A rift system has been growing inward from the seaward edge of the Amery Ice Shelf in East Antarctica for over twenty years. Currently active, the ‘Loose Tooth' rift system is in the process of calving off an iceberg that will be at least 30 x 30 km in size. Over the past five field seasons passive seismic and geodetic measurements have been made at one of the propagating tips of the rift system. During the 2006-07 field season pilot Ground Penetrating Radar and Transient Electromagnetic (TEM) data were collected over and near that propagating rift tip, to provide complimentary constraints regarding the internal structure of the ice shelf and to look for evidence of penetration of electrically conductive seawater into the rift from below. In another auxiliary experiment, small charges were set off near the rift tip to provide calibration and velocity structure information for the primary passive seismic data set. Preliminary results are presented in comparison with the passive seismic and geodetic results, as well as a discussion of practical constraints and challenges in collecting these types of data at an ice shelf rift. http://loose-tooth.ucsd.edu
NS11A-0157
Tracking the Evolution of Sea Ice Properties With In-Situ Dielectric Probes and Cross- borehole Resistivity Tomography
The physical properties of sea ice depend on its temperature, which is easily measured in situ, and salinity, which is typically determined on melted core samples. The ice matrix and the brine inclusions in sea ice have a very large electrical conductivity contrast which makes electrical methods attractive for in-situ salinity measurements. However, the connection between bulk electrical properties and salinity are confounded by the complex, anisotropic and multi-scale microstructure for which only highly simplified models exist. We report on recent measurements of the complex dielectric permittivity at 50 MHz and the DC resistivity of sea ice. Our aims are to develop in-situ salinity measurements and by deriving brine volume fractions (vb) from concurrent temperature measurements to gain insight into microstructural effects. Specifically we seek to identify evidence of critical transitions in transport properties, such as a sudden onset of brine percolation attributable to non-linear increases in the connectivity of the brine inclusions. Cross-borehole resistivity tomography measurements in sea ice were made over the growth and melt season in landfast first-year sea ice near Barrow, Alaska. 2-D and 3-D inversions resolved the seasonal evolution of the horizontal resistivity (ρH) structure with a better resolution than previous, surface-based approaches. When vb is low, ρH obeys Archie's Law with m ≈ 2.9 but a departure to a stronger increase in conductivity is seen for vb > 7-10 % indicating enhanced horizontal brine inclusion connectivity. Laboratory measurements with 50 MHz Vitel Hydraprobes established conditions for accurate measurements of sea-ice dielectric permittivity. Results from single-crystals show anisotropy with respect to the vertically-oriented, sub-parallel intra-crystalline brine layers. In natural sea ice, the real part of the permittivity (`dielectric constant') can be used to estimate salinity and the imaginary part (`loss factor') shows signatures of brine motion during spring warming once vb > 5-6 %.
NS11A-0158
Comparison and Calibration of Sea-Swell with Microseism Observations on Ice-Shelf and Land Based Seismometer Stations
Two broadband seismometer records, one from a location near the seaward iceberg-calving front of the Ross Ice Shelf, and the other at Scott Base, on Ross Island are compared. Despite the fact that the two stations are about 300 km apart, and that one is based on floating ice and the other based on an oceanic island, both records clearly display dispersed sea-swell arrival events from storms in the Indian, Pacific and Southern oceans. The ice shelf record displays the 'direct' sea-swell motion, whereas the land record displays primarily the frequency- doubled microseism motion from near-coastally generated Rayleigh waves arising from incoming and reflected swell interference. Two years of austral summer data are compared to calibrate the microseism observations of the land-based seismometer as a proxy for sea swell signal incident on the ice shelf. Preliminary results are used to estimate sea-swell influences on the Ross Ice Shelf at times prior to the deployment of the near-ice-front seismometer, e.g., to hindcast the sea-swell state at the time when B15 and other large icebergs originally calved from the Ross Ice Shelf.
NS11A-0159
DC Electrical Resistivity Imaging at a High-Arctic Continuous Permafrost Site in Svalbard, Norway
Direct-Current (DC) electrical resistivity imaging has proven to be a suitable technique for a number of permafrost related questions. We present measurements from a high-arctic continuous, maritime permafrost site near Ny Alesund, Svalbard (Norway). The area under investigation features a great diversity of soil types and soil water contents. Sparse vegetation alternating with rock fields and exposed soil characterize the surface. A total of 25 different transects each of 47.5m length were investigated using a DC-Resistivity and Electrode Control System (RESECS) with 96 electrodes at a spacing of 0.5m in Wenner-alpha configuration. At three transects, fixed electrode arrays were installed and measured on a weekly basis in order to capture temporal changes. The study was conducted from August until mid-September, thus covering the period of maximum active layer thickness and the beginning of freeze-up. The specific resistivities at the surface ranged from less than 50Ømegam in areas with damp clay to more than 1000Ømegam in rock fields and on dry hill crests. In most cases, areas with such high resistivities only extended to depths of less than 1m. From depths between 1m and 1.5m onwards, specific resistivities increased continuously, indicating the position of the freeze-thaw interface. This agrees well with thaw depths that were determined by point measurements along individual transects using a drill. The repeated measurements of the fixed electrode arrays displayed the most pronounced changes in the beginning of August, where decreases in specific resistivities of up to 40% over one week period were detected at depths between 1m and 2m. Afterwards, only insignificant changes were observed at these depths. This is interpreted to be the seasonal thawing of the active layer, which stagnates during in the second half of August. At depths less than 1m, both decreases and increases in specific resistivities were detected, most likely due to changes in the water content of the soil.
NS11A-0160
Time-lapse Electrical Resistivity Measurements for Investigating Temperature Changes in Extrazonal Ventilated Permafrost
Field experiments have been carried out to evaluate the ability for time-lapse electrical resistivity tomography and/or vertical electrical sounding for estimating seasonal temperature changes at depth in a low-altitude permafrost terrain in the Swiss Jura mountains. The prospected site is a limestone blocky talus slope located at 1200 m a.s.l. (mean annual air temperature +5.5°C) far below the regional lower limit of discontinuous mountain permafrost. Due to a seasonally reversible air circulation mechanism throughout the whole porous medium, driven by the thermal contrast between the outside and inside air, a strong negative thermal anomaly is observed and permafrost in the lower and deeper parts of the ventilated terrain locally occurs. The predominance of advective heat fluxes causes significant seasonal temperature shifts until the base of debris accumulation at about 10-20 m depth. Indeed, the ascent of relatively warm light air during wintertime (the so-called "chimney effect") provokes the sucking of cold outside air deep inside the ground in the lower and intermediate parts of the slope, building up a "cold" reservoir. Consequently, a permanent gravity discharge of cold dense air occurs in summertime and prevents the ground temperature to increase above about +5°C in the lower section of the slope. Permafrost is thus likely to occur in both the lower and deeper parts of the talus slope. Evaporation during the winter phase of ventilation makes that the ice content of permafrost is probably low. Among other experiments, a 33-m tomography profile was repeated nine times between May 2001 and May 2002 along the lower part of the talus slope using permanently-installed electrodes (1 m electrode spacing) and a Wenner array. Only the uppermost talus layer beneath the organic soil was thus investigated. According to ground temperature measurements performed at 30 cm depth, results show that changes in electrical apparent resistivity collected with our tomography system are not identical for temperature shifts above or below freezing point: the temperature increase in the active layer from 0°C (in May 2001) to about +5°C (in October 2001) was accompanied by a decrease of the mean apparent resistivity of 23%, whereas a drop in temperature (from November to January 2001) down to about -5°C caused a multiplication of the mean apparent resistivity at least by a factor of 15. Theses observations are confirmed on the inversion results. This experiment has shown that a temperature change above freezing point induces a low linear resistivity variation, whereas a temperature shift below freezing point seems to provoke an strong (exponential ?) resistivity change. Vertical soundings were carried out repeatedly at the same place over the permafrost in the lower part of the slope. Close to the surface, the calculated resisitity and its seasonal changes were similar to those observed on the tomography profile. Moreover, interpretation of the data indicated a smaller but significant increase of the ground resistivity until the deepest layer (8 - 20 m depth) of the talus slope by mid-winter. Applying the relationships linking temperature and resistivity determined from the tomography data, it could be estimated that this change has corresponded to a temperature decrease from 0 down to about -0.7°C. This demonstrates the efficiency of the air circulation process to cool the deepest layers of a talus slope in wintertime.
NS11A-0161
Ice penetrating radar surveys along the EGIG line in the percolation zone of Western Greenland
We conducted a variety of ice penetrating radar surveys along the EGIG line on the western side of the Greenland ice sheet. The purpose of these surveys is to1) gain an understanding of the hydrologic pathways of surface generated meltwater, and 2) document the rate of firn densification in the upper 80 m of the percolation zone. Our various data collection techniques included: 1) common midpoint (CMP) gatherers using 100 MHz antennae and a maximum separation of 79 m at six locations spread out along on the EGIG line; 2) single offset 1 GHz transect imaging the approximately the top 2 m of snow. This transect spans a distance of ~40 km from T1 to Crawford point; 3) 1 GHz constant offset imaging of a 100 m x 100 m grid on 10 m grid spacing; 4) a multi-offset transect within the 100 m x 100 m grid utilizing 1 GHz and 500 MHz antennae with a maximum offset of 7.5 m; 5) 20 m x 20 m 3D grid with 20 cm line spacing. This grid images the top 10 m of the snowpack with 500 MHz antennae; and 6) 200 MHz CMP survey of the 3D grid at 5 locations. The radar data is part of a larger data set that includes 21 cores to 10 m depth, logged for detailed density and stratigraphy and 10 m temperature strings at all CMP locations. These additional data provide control for interpretation and modeling of the GPR response
NS11A-0162
Oil Detection In and Under Sea Ice Using Ground-Penetrating Radar
Marine oil spills can occur in the Arctic due to pipeline breaks or leaks and spills from storage or production facilities. Depending on the time of year and scenario, a portion or all of the spill may become trapped under and/or encapsulated within the sea ice sheet. The current methods for locating spilled oil include visually inspecting drilled ice cores or sending divers under the ice. Speed is a key issue in oil clean-up. A non-invasive method of detecting oil quickly and reliably would greatly facilitate the clean-up and lessen the impact on the environment. First-year ice thicknesses of 0.5-2.0 m, typical of the Arctic region, can be well-resolved using radar. Oil film thicknesses can range from a few mm to 20 cm depending on the ice-water interface topography. For typical conditions a frequency of 500 MHz gives a 1/4 wavelength limit of 7 cm; therefore a typical spill scenario is a thin-bed problem for ground-penetrating radar (GPR). Interference due to thin-beds may cause amplitude, phase and frequency anomalies in the reflected wavelet. In April 2006, SINTEF conducted a contained oil-spill under natural Arctic sea ice conditions in a fjord on Svalbard. Using data collected during that experiment from a 500 MHz antenna and complex trace analysis we computed the instantaneous frequency, instantaneous phase and the envelope function and found significant differences in the data before and after the oil was inserted. These results demonstrated the potential of GPR to be a practical system for oil in ice detection under certain conditions. As a follow-on to the 2006 project, we have undertaken a detailed modeling effort to estimate GPR response to specific variables, such as ice and oil thicknesses, ice salinity and temperature.
NS11A-0163
Monitoring Shallow Freeze and Thaw Processes Using High Frequency Surface Ground Penetrating Radar
High frequency GPR (225-900 MHz) was used for the high resolution, non-invasive monitoring of freeze and thaw processes in the shallow subsurface (i.e., within 1.5 meters of the surface) at three sites with different soil types (sand, sandy loam and silt loam). Both common midpoint (CMP) gathers and constant offset reflection profiles were repeatedly done over 2 m transects at each site. CMP gathers were used to monitor direct ground wave propagation and determine vertical velocity profiles; reflection profiles were used to image reflectivity variations and observe near vertical reflection event traveltimes. Direct ground waves were significantly attenuated during the initial freezing process due to interference with short period reflections; modelling results indicate that this phenomenon occurs until the frozen surface layer exceeds 15 cm in thickness for 450 MHz data. Differences in shallow surface freezing rates between sites were inferred from the direct wave data. Direct wave velocities indicate that surface thawing occurred rapidly in late March even though snowpack melting commenced in early March. CMP velocity profiles show the progressive development of the frozen zone and indicate the presence of liquid water in this zone one week prior to the surface thaw. Reflection profiling monitored the development of a continuous reflection event interpreted to be the interface at the base of frozen zone. Significantly diminished reflectivity of stratigraphic interfaces was observed above this interface. Measured reflection event traveltimes indicate the deeper section underwent drainage after the shallow frozen zone development terminated the seasonal autumn recharge. In addition, reflection traveltime data also found the thaw recharge started one week prior to the surface thaw.
NS11A-0164
Object-oriented philosophy in designing adaptive finite-element package for 3D elliptic deferential equations
Although adaptive finite-element (AFE) analysis is becoming more and more focused in scientific and engineering fields, its efficient implementations are remain to be a discussed problem as its more complex procedures. In this paper, we propose a clear C++ framework implementation to show the powerful properties of Object-oriented philosophy (OOP) in designing such complex adaptive procedure. In terms of the modal functions of OOP language, the whole adaptive system is divided into several separate parts such as the mesh generation or refinement, a-posterior error estimator, adaptive strategy and the final post processing. After proper designs are locally performed on these separate modals, a connected framework of adaptive procedure is formed finally. Based on the general elliptic deferential equation, little efforts should be added in the adaptive framework to do practical simulations. To show the preferable properties of OOP adaptive designing, two numerical examples are tested. The first one is the 3D direct current resistivity problem in which the powerful framework is efficiently shown as only little divisions are added. And then, in the second induced polarization£¨IP£©exploration case, new adaptive procedure is easily added which adequately shows the strong extendibility and re-usage of OOP language. Finally we believe based on the modal framework adaptive implementation by OOP methodology, more advanced adaptive analysis system will be available in future.
NS11A-0165
Combining Radar Altimetry and Lidar to Study Snow Accumulation
Calibration and validation is essential for the success of a remote sensing satellite missions. In spring 2006 a major CryoSat-2 campaign was carried out in the Arctic to obtain a calibration and validation dataset using lidar, the ASIRAS radar and in-situ measurements. The ASIRAS is an airborne radar which mimics the CryoSat-2 SIRAL instrument. The radar was mounted on an aircraft together with lidar and support instruments throughout the five week long campaign. Studies show that ASIRAS has the capability to detect density variations several meters into the snow pack on ice sheets. When combining the radar measurements with lidar and in-situ measurement it is possible to estimate the yearly snow accumulation on inland ice. Based on the yearly variation in density detected by the radar it is possible to detect several seasons and thus obtain an accumulation history for a specific area. The combination of radar and lidar on a single platform also makes it possible to estimate the snow cover on sea ice. Studies have shown that a good cross calibration between the lidar and the radar can be used to estimate the snow thickness on first year ice, and ongoing studies are examining the possibility to apply the same method on multi year ice. The multi year ice poses difficulties with a high noise level and false reflectors due to the inhomogeneous ice field consisting of old porous ice, refrozen melt water, ridges and new ice.
NS11A-0166
Englacial Characterization of a Temperate Glacier Using 3D Multioffset Multi-channel GPR
In this study, we used ground penetrating radar to image the geometry and spatial distribution of englacial voids. It is well established that radar velocity is correlated with water content in ice. Accurate laterally continuous velocity measurements are most efficiently acquired with continuous multi-offset data. Previous acquisition of such data was prohibitive due to technological limitations. Recent advances in GPR technology, including faster electronics and multi-channel capability, have made multi-fold acquisition feasible. In spring 2006, we acquired a 100 m x 100 m multi-azimuthal, multi-offset 3D grid. We used 25 MHz antennas with offsets ranging from 5 m to 150 m with 15 different offsets. These data were acquired at the transition from low water input during winter, to relatively high water input and rapid glacier motion during the spring and early summer. During late summer, 2007, we also collected a second dataset consisting of a multi-azimuthal common midpoint survey. Both surveys generated data along multiple azimuths, allowing us to examine velocity anisotropy. Variations in the velocity with respect to direction indicate preferential void orientation. Determining the distribution, direction and size of the voids is vital to understanding englacial flow and its affect on overall glacial hydrology. Preliminary velocity analysis shows anisotropy with the preferred fracture direction.