Pine Island Glacier has thinned and accelerated over recent decades, significantly contributing to global sea level rise. Increased oceanic melting of its ice shelf is thought to have triggered those changes. Observations and numerical modeling reveal large fluctuations in the ocean heat available in the adjacent bay and enhanced sensitivity of ice shelf melting to water temperatures at intermediate depth, as a seabed ridge blocks the deepest and warmest waters from reaching the thickest ice. Oceanic melting decreased by 50% between January 2010 and 2012, with ocean conditions in 2012 partly attributable to atmospheric forcing associated with a strong La Niña event. Both atmospheric variability and local ice shelf and seabed geometry play fundamental roles in determining the response of the Antarctic Ice Sheet to climate.One Sentence Summary: Ocean melting of the Pine Island Glacier ice shelf was halved in two years as an underlying seabed ridge makes it highly sensitive to climatic forcing. Main Text:Austral summer observations in the Amundsen Sea, West Antarctica, show that lightlymodified, warm (0.5-1.2°C) and saline (>34.6) Circumpolar Deep Water (CDW), 2-4°C above the in-situ freezing point, pervades a network of glacially scoured seabed troughs (1, Fig. 1A).The CDW reaches nearby Antarctic glaciers and delivers heat to the base of their 200-1000 mthick ice shelves (2-4). It is overlain by a 200-300 m-thick layer of cold (-1.5°C) and fresh (salinity<34.4) Winter Water (WW, Fig. 2A) that is seasonally replenished by interaction with the atmosphere and sea ice.Pine Island Glacier (PIG), a major outlet glacier feeding one such ice shelf, has shown apparently continuous thinning (5, 6) and intermittent acceleration (7-9) from 1973 to 2009.During this period, its ice shelf has also thinned (6,(10)(11)(12), and the reduction in buttressing driven by oceanic melting is believed to be responsible for the changes inland. Earlier analysis indicated that a higher CDW volume and temperature in Pine Island Bay (PIB) in January 2009caused an increase in ice-shelf melting and in the associated meltwater-driven circulation, relative to 1994 (2). The lack of sub-annual variability in CDW temperature during one yearlong measurement in PIB (1) and the long-term correlation between the oceanic melting and the mass loss required to sustain thinning of the ice shelf gave the impression that the ice-ocean system had shown progressive change over the last two decades. This is consistent with a positive geometrical feedback, with oceanic melt enlarging the cavity under the ice shelf, allowing stronger circulation and further melting.However, such ice-ocean systems are likely to be more complex. The glacier's rapid change over the last few decades was probably triggered by its ungrounding from a the top of a seabed ridge transverse to the ice flow at some time before the 1970s (4). Subsequent migration of the glacier's grounding line (13) down the seabed slope upstream from the ridge crest was probably an inevitable respon...

A brain-computer interface (BCI) is a system that allows its users to control external devices with brain activity. Although the proof-of-concept was given decades ago, the reliable translation of user intent into device control commands is still a major challenge. Success requires the effective interaction of two adaptive controllers: the user's brain, which produces brain activity that encodes intent, and the BCI system, which translates that activity into device control commands. In order to facilitate this interaction, many laboratories are exploring a variety of signal analysis techniques to improve the adaptation of the BCI system to the user. In the literature, many machine learning and pattern classification algorithms have been reported to give impressive results when applied to BCI data in offline analyses. However, it is more difficult to evaluate their relative value for actual online use. BCI data competitions have been organized to provide objective formal evaluations of alternative methods. Prompted by the great interest in the first two BCI Competitions, we organized the third BCI Competition to address several of the most difficult and important analysis problems in BCI research. The paper describes the data sets that were provided to the competitors and gives an overview of the results.

Interest in developing a new method of man-to-machine communication--a brain-computer interface (BCI)--has grown steadily over the past few decades. BCIs create a new communication channel between the brain and an output device by bypassing conventional motor output pathways of nerves and muscles. These systems use signals recorded from the scalp, the surface of the cortex, or from inside the brain to enable users to control a variety of applications including simple word-processing software and orthotics. BCI technology could therefore provide a new communication and control option for individuals who cannot otherwise express their wishes to the outside world. Signal processing and classification methods are essential tools in the development of improved BCI technology. We organized the BCI Competition 2003 to evaluate the current state of the art of these tools. Four laboratories well versed in EEG-based BCI research provided six data sets in a documented format. We made these data sets (i.e., labeled training sets and unlabeled test sets) and their descriptions available on the Internet. The goal in the competition was to maximize the performance measure for the test labels. Researchers worldwide tested their algorithms and competed for the best classification results. This paper describes the six data sets and the results and function of the most successful algorithms.

Abstract. Designing a Brain Computer Interface (BCI) system one can choose from a variety of features that may be useful for classifying brain activity during a mental task. For the special case of classifying EEG signals we propose the usage of the state of the art feature selection algorithms Recursive Feature Elimination [1] and Zero-Norm Optimization [2] which are based on the training of Support Vector Machines (SVM) [3]. These algorithms can provide more accurate solutions than standard filter methods for feature selection [4]. We adapt the methods for the purpose of selecting EEG channels. For a motor imagery paradigm we show that the number of used channels can be reduced significantly without increasing the classification error. The resulting best channels agree well with the expected underlying cortical activity patterns during the mental tasks. Furthermore we show how time dependent task specific information can be visualized.

[1] Cold shelf waters flowing out of the Filchner Depression in the southern Weddell Sea make a significant contribution to the production of Weddell Sea Bottom Water (WSBW), a precursor to Antarctic Bottom Water (AABW). We use all available current meter records from the region to calculate the flux of cold water (<À1.9°C) over the sill at the northern end of the Filchner Depression (1.6 ± 0.5 Sv), and to determine its fate. The estimated fluxes and mixing rates imply a rate of WSBW formation (referenced to À0.8°C) of 4.3 ± 1.4 Sv. We identify three pathways for the cold shelf waters to enter the deep Weddell Sea circulation. One path involves flow constrained to follow the shelf break. The other two paths are down the continental slope, resulting from the cold dense water being steered northward by prominent ridges that cross the continental slope near 36°W and 37°W. Mooring data indicate that the deep plumes can retain their core characteristics to depths greater than 2000 m. Probably aided by thermobaricity, the plume water at this depth can flow at a speed approaching 1 m s À1 , implying that the flow is occasionally supercritical. We postulate that such supercriticality acts to limit mixing between the plume and its environment. The transition from supercritical to slower, more uniform flow is associated with very efficient mixing, probably as a result of hydraulic jumps.