2008 report
Progress: 45%
Partner 25 (Uds)
In the last year, we conducted a seismic experiment, using the Tara drifting base. A network of 5 seismic stations was deployed in April 2007, and continuously monitored the vibrations of the sea-ice cover (in the frequency range 60 s – 100 Hz) until October 2007. The processing of this dataset is to start in November 2007, in collaboration with CNRS (Jérôme Weiss). The detected fracturing events will be characterized(time, location, rupture length, faulting regime). The relation between the fracturing at the local scale (~ 1 km) and deformation at meso and small scales will be investigated.
The objectives of this in-situ measurement campaign were to collect seismological data to monitor the fracturing taking place in the sea-ice cover caused by its deformation. The occurrence of ice quakes is related to the formation of leads cutting through the cover, that accommodates the strain building up in the ice. By locating these ice quakes, by characterizing their sources (direction of slip, size, duration of rupture), by correlating these ice quakes to changes in environmental forcing, this study aims at better understanding the deformation process at the small scales (1m – 1 km). DAMOCLES offered the opportunity to perform such a campaign over a long period of time (months), so allowing for the collection of a unique dataset in size (duration). Also, the vibration characteristics of the ice cover can give an estimate of the ice thickness (between an ice quake and a seismometer, or between two seismometers), and its evolution through time.
Two scientists (Jean-Philippe Métaxian and Jacques Grangeon) were part of the DAMOCLES group that went to the Tara ice camp in April 2007. As detailed in Deliverable D1.3-1 (Field experiment plan), seismic stations were deployed with a ~ 1 km aperture. Due to restrictions in the logistics, only 5 stations instead of 6 were installed. 4 of those stations were equipped with 4 short-period vertical seismometers (L4C) and a three-component, long-period (60s) seismometer (Guralp CMG3). The L4C sensors were deployed at the nodes of a ~ 50 m – long square centred on the CMG3, hence forming a small-aperture seismic antenna for each of those 4 stations. The 5th station was only equipped with a single CMG3 seismometer. The acquisition was initiated successfully, using continuous recording with a 100Hz sampling.
Partner 25 (LGGE-CNRS)
The work performed by LGGE-CNRS during the second year of the DAMOCLES project can be divided in three parts.
(i) In collaboration with UdS (partner #26), who was the task-leader on this part of the work, we participated to the preparation of the field seismic and GPS experiment which has been installed in April 2007 near the Tara ship. A network of 5 seismic stations and GPS was installed around the Tara over a region of about 2 km², and ran until late September. At the end of the summer season, most of the network was removed for logistics reasons, whereas one station remained in the field in the vicinity of Tara until November. More details on this field experiment are given in the UdS report.
(ii) Study of sea ice deformation from the analysis of the dispersion of buoys trajectories. This work was also done in collaboration with UdS. We related first the dispersion of lagrangian trajectories to sea ice deformation. We used the IABP buoy trajectories dataset to analyse the evolution of the sea ice strain-rate with space and time scales. The results show that the strain-rate S is a time and spatial scale-dependent variable, following scaling laws such as S ~ t -a for the time domain, or S ~ L -g for the spatial domain. This expresses the strong heterogeneity and intermittency of sea ice deformation at all scales. This complexity is also characterized by a strong space/time coupling, illustrated by a dependence of the exponent a (respectively b) on the spatial (respectively time) scale. The revealed exponent values show that sea ice deformation cannot be considered as viscous-like (i.e. a=b=0), even at large time and/or spatial scales, instead is accommodated by a multiscale fracturing process. This work has been published recently in JGR Oceans.
(iii) To explain the scaling laws characterizing sea ice dynamics and deformation (point (ii)), we developed a simple conceptual model based on (a) a fractal distribution of faulting events, (b) a fractal clustering of events in the time domain and (c) a power law distribution of event sizes. This model, which explicitly assumes that all the sea ice deformation is accommodated by fracturing/faulting, allows to explain the space and time scaling laws observed, as well as the associated space/time coupling. This argues for an essential role of fracturing processes on sea ice deformation and dynamics.
References:
Rampal, P., Weiss, J., Marsan, D., Lindsay, R., Stern, H., Scaling properties of sea ice deformation from buoy dispersion analysis, J. Geophys. Res., in press
Weiss, J., Marsan, D. and Rampal, P., Space and time scaling laws induced by the multiscale fracturing of the Arctic sea ice cover, IUTAM Symposium on Scaling in Solids Mechanics, Cardiff, UK, 25-29 June, 2007
Partner 5 (FIMR)
FIMR have examined the variability and change of the sea-ice thickness distribution based on the unclassified submarine sonar measurements. The goal is to analyze how much the observed changes are related on the changes of the dynamics (circulation, redistribution of sea-ice) or thermodynamics (radiation balance). We used both analog and digitally recorded data, since the difference between these two forms is very small compared to the draft measurements of typically several meters. The area covered with submarine draft data was divided in six regions: North Pole, Canada Basin, Beaufort Sea, Chukchi Cap, Eastern Arctic and Nansen Basin. Most of the analysis were made for spring (April-May) and autumn (September-October) because those are the periods of most extensive observations. It is also the periods when annual maximum and minimum thicknesses are observed.
For the whole 26-year period the seasonal variability of mean ice draft (difference between spring and autumn mean drafts), is of the same magnitude, 1.2-1.7 meters, in all regions except in Nansen Basin where it is smaller, 0,9 meters. Interannual variations in spring mean drafts are largest in Eastern Arctic and smallest in North Pole and Canada Basin. Both spring and autumn mean draft decreased in almost every region between former and later half of the 26-year study period, and only exception were Nansen Basin in spring (0,47 meters increase) and Chukchi Cap in autumn (no change). In general, decrease in mean draft was larger in spring than in autumn. Also the variability of ice draft is larger in spring than in autumn in every region, and is largest in North Pole and Canada Basin. Ice draft variability decreased in every region and in both seasons between years 1975-1987 and 1988-2000, except in autumn in Nansen Basin .
Draft distributions were calculated for both spring and autumn and for periods of 1975-1987 and 1988-2000 in each region, so that changes in distributions could be seen. It was observed that regional variability in draft distributions is large and also the seasonal variability between spring and autumn varies regionally. There were several considerable changes in draft distributions between former and later half of the 26-year study period. For example, in the Beaufort Sea, the seasonal variability decreased considerably, which was mostly due to remarkable change in spring draft distribution. In Beaufort Sea the modal draft in spring decreased from 2,6 meters to 0,8 meters, indicating the increase in the amount of first year ice and the decrease in multi year ice. In Eastern Arctic changes in draft distribution were larger in autumn, and the shape of autumn draft distribution changed from one to bi-modal due to large increase in the fraction of open water and very thin ice (d<0,2 meters). Also in the Canada Basin the fraction of open water roughly tripled in autumn, while remaining constant in spring. Most uniform of observed changes in all regions was the evolution of the amount of deformed ice . The fraction of deformed ice (defined as draft > 5 meters) decreased both in spring and autumn in every region between years 1975-1987 and 1988-2000, and the decline was strongest in Eastern Arctic (-35%/decade in spring and -40%/decade in autumn) and Beaufort Sea (-36%/decade in spring and -24%/decade in autumn). The decline was most modest in Nansen Basin (-19%/decade in spring and -5%/decade in autumn).
We used NCEP and ERA-40 reanalyzed meteorological data to find thermodynamical or dynamical forcing possible causing observed changes in Arctic ice cover. The mean surface air temperature and wind fields were calculated for winter and summer for former and later 13-year periods. Both in temperature and wind fields changes were larger in winter than in summer. In mean winter air temperature the increase was strongest in Eastern Arctic and in Canadian Archipelago (+2 C) and in winter wind field changes were largest in Beaufort Sea and in Eastern Arctic. These were also the regions with strongest changes in ice cover. Especially the remarkable decrease in the fraction of deformed ice in Eastern Arctic can be explained by the change in wind field. During the former 13-year period wind was piling up the ice in the Siberian coast, while during the later period wind was more oriented offshore.
Figure XX. One hour of cover vibrations recorded on the 27th of April 2007 near Tara by 4 broad-band CMG3 Guralp seismometers. The vertical velocity is shown, in arbitrary units. Several icequakes are clearly seen on those seismograms. The processing of >5 months of such data will imply (1) detecting icequakes, (2) locating their sources using seismic array techniques, (3) computing their magnitudes (sizes), and (4) estimating their orientations and direction of fracturing, whenever the data permit to do so.
