Just weeks after that report, there was shock and, for some, a temptation to despair when the startling news was released in May by scientists at both NASA and the University of Washington that the long-feared "collapse" of a portion of the West Antarctic ice sheet is not only under way but is also now "irreversible." Even as some labored to understand what the word "collapse" implied about the suddenness with which this catastrophe will ultimately unfold, it was the word "irreversible" that had a deeper impact on the collective psyche.

Just as scientists 200 years ago could not comprehend the idea that species had once lived on Earth and had subsequently become extinct, and just as some people still find it hard to accept the fact that human beings have become a sufficiently powerful force of nature to reshape the ecological system of our planet, many – including some who had long since accepted the truth about global warming – had difficulty coming to grips with the stark new reality that one of the long-feared "tipping points" had been crossed. And that, as a result, no matter what we do, sea levels will rise by at least an additional three feet.

The uncertainty about how long the process will take (some of the best ice scientists warn that a rise of 10 feet in this century cannot be ruled out) did not change the irreversibility of the forces that we have set in motion. But as Eric Rignot, the lead author of the NASA study, pointed out in The Guardian, it's still imperative that we take action: "Controlling climate warming may ultimately make a difference not only about how fast West Antarctic ice will melt to sea, but also whether other parts of Antarctica will take their turn."

The news about the irreversible collapse in West Antarctica caused some to almost forget that only two months earlier, a similar startling announcement had been made about the Greenland ice sheet. Scientists found that the northeastern part of Greenland – long thought to be resistant to melting – has in fact been losing more than 10 billion tons of ice per year for the past decade, making 100 percent of Greenland unstable and likely, as with West Antarctica, to contribute to significantly more sea-level rise than scientists had previously thought.

[12] The main parameter derived from passive microwave data is sea ice concentration, which is generated using an algorithm that takes advantage of the contrast in the emissivity of sea ice and open water at different frequencies. The two algorithms used most frequently are the Bootstrap Algorithm discussed by Comiso [2009] and the NT2 Algorithm discussed by Markus and Cavalieri [2009]. Parkinson and Comiso [2008] compare the concentration data from the two algorithms and show that they yield similar results. These algorithms have also been validated against field and high-resolution satellite data and the results show accuracies ranging generally from 5% to 15% depending on location and season. The errors are larger near the ice edge and in coastal areas because of large variability of the emissivity of sea ice in these regions. For improved interpretation, we supplement the passive microwave data with other satellite data sets including high-resolution visible, infrared and synthetic aperture radar (SAR) data. In this study, we derive the ice concentration data from November 1978 to December 2009 using the Bootstrap Algorithm. Unfortunately, the NT2 algorithm requires the 85 GHz data that are not available with SMMR (November 1978 to August 1987) and parts of the SSM/I data set (February 1989 to December 1991). This ice concentration data is used to estimate the fraction of open water within the ice pack and to estimate the ice extent, which is the integrated sum of the area of all grid cells containing at least 15% ice. It is also used to estimate the ice area, which represents the total area actually covered by ice and is the integral sum of the products of the area of each grid cell with the corresponding ice concentration. The ice extent and ice area are generally used to quantify the variability and long-term trends of the ice cover.

[13] Because the signatures of sea ice and glacial ice are generally similar, it is difficult to develop an algorithm that discriminates one from the other. As a result, we normally used a land ice mask as described by Comiso [2009]. The use of a land ice mask however, is problematic especially in ice shelf and glacier regions where iceberg calving and ice surges are unpredictable. To illustrate this, concurrent high-resolution visible (250 m) images from Aqua/MODIS and microwave brightness temperature data at 37 GHz (vertical polarization) from Aqua/AMSR-E are presented in Figure 2a and 2b, respectively. These images were acquired over the Ross Sea region on 12 October 2007 after the calving of large icebergs in 2000 and 2002 [Martin et al., 2007]. In both images, the solid black line shows the shelf/ocean boundary as published previously [Defense Mapping Agency, 1992]; it is apparent that because of the loss of shelf ice, the actual ice boundary at this time is south of the published boundary. Because of the importance of the shelf position, we used a mix of SAR and visible imagery to determine the location of the ice shelf front.