Introduction to special section on Large-Scale Characteristics of the Sea Ice Cover from AMSR-E and Other Satellite Sensors

Authors


1. Introduction

[1] The sea ice cover is one of the key components of the global climate system and is expected to provide early signals of a climate change because of ice-albedo (and other) feedbacks associated with the high reflectivity of snow and ice. In particular, a declining snow and ice cover causes the exposure of larger areas of low-albedo regions that absorbs more heat and causes the surface to get warmer. This would cause the delay of ice freezeup in autumn and the occurrence of early melt in spring that in turn would cause even less ice cover. Sea ice is also a good insulator and therefore serves to keep the warmth of the ocean from leaking to the atmosphere in winter. It redistributes salt in that in areas where it forms, brine is rejected and high-salinity dense water is formed while in areas where it melts, low-salinity water is introduced. Through the accumulation of brine in high ice production coastal polynya regions along the continental shelves, the water mass also gets transformed to the high-salinity bottom water that drives the thermohaline circulation of the World's oceans. In deep oceans, the formation of sea ice can also initiate convection which in turn enables the ventilation of the ocean and redistribution of nutrients, oxygen, carbon and other chemicals.

[2] We have learned a lot about the distribution and large-scale characteristics of the sea ice cover from satellite data. Through passive microwave data, for example, we now know that the coverage of sea ice is vast, representing about 6 to 8% of the ocean surface, and is very seasonal with the average extent changing from 7 × 106 km2 in the summer to 16 × 106 km2 in the winter in the Northern Hemisphere and from 2 × 106 km2 in the summer to 19 × 106 km2 in the winter in the Southern Hemisphere. With almost three decades of continuous passive microwave satellite data already available, we also now know that the ice cover goes through very large interannual variability and is providing mixed signals from the Arctic and the Antarctic with the former showing overall negative trends of about −3% per decade [Bjorgo et al., 1997; Parkinson et al., 1999; Comiso, 2006] while the Antarctic shows a slight positive but significant trend of about 1% per decade [Zwally et al., 2002; Comiso, 2003; Cavalieri et al., 2003]. The most remarkable change, however, has been the Arctic perennial sea ice cover which has been observed to be declining at a rapid rate of about −10%/decade [Comiso, 2002, 2006]. Such rate is about 3 times faster than what has been projected by numerical models [Stroeve et al., 2007]. The unusually large variability and trends observed (for perennial ice) require independent verification, especially since passive microwave data have relatively coarse resolution and algorithms have to be developed and used to extract geophysical ice parameters from the satellite data.

[3] This special section is a collection of papers that discusses the current state of the global sea ice cover including its extent, area, thickness, and snow cover as observed by satellite and aircraft sensors and also by ship based observations. Papers on sea ice studies that make use of the Advanced Microwave Scanning Radiometer on board the EOS-Aqua satellite (AMSR-E) data were especially solicited because the instrument is currently the state of the art with much better resolution than predecessors and likely the best tool currently available for hemispheric sea ice cover studies throughout the year. The set of papers includes results from three aircraft missions primarily meant to validate sea ice parameters from AMSR-E, to assess accuracy of current parameters, and to evaluate relevance to scientific needs. One of the missions was conducted at the Sea of Okhotsk in February 2003 and the other two in the Antarctic (Weddell and Bellingshausen Seas) in August 2003 and October 2004. In August 2003, the aircraft mission was done in conjunction with a dedicated ship program. In this overview paper, we start by introducing papers dealing with sea ice concentration which is the key geophysical parameter that has been derived from AMSR-E and similar data sets and is in turn used to estimate of ice extent and area. We then proceed with the discussion of papers about ice thickness and snow thickness which are equally if not more important for climate studies but their algorithms are still relatively new and need validation. Papers on mesoscale characteristics and on process and modeling studies that make use of satellite data to gain insights into some of the unique processes in the polar regions follow.

2. Ice Concentration, Ice Extent, and Ice Area

[4] The first set of papers discusses the distributions and coverage of the sea ice cover as inferred from satellite data and in situ measurements. The paper by J. C. Comiso et al. (Multi-sensor Characterization of the Antarctic Sea ice Cover, submitted to Journal of Geophysical Research, 2008) provides a general overview of current capabilities for monitoring the sea ice cover using sensors from the visible, infrared and microwave channels, passive and active. The multisensor capabilities are quite comprehensive with enormous potential for both mesoscale and large-scale studies but most were launched only in the recent decade and are not suitable for long-term variability studies. The data used for such long-term studies have been provided by passive microwave sensors, the most advanced of which is the AMSR-E, which is a Japanese instrument. The articles by Parkinson and Comiso [2008] and by Comiso and Parkinson [2008] provide results from analysis of 5-years of AMSR-E data using two techniques and demonstrate that although average concentrations derived are slightly different (1 to 5%), the sea ice extents and ice areas provide very similar characterizations and trends of the sea ice cover. The paper by Comiso and Nishio [2008] provides results from analysis of the longer-term data set consisting of the records from Scanning Multichannel Microwave Radiometer (SMMR) (1978 to 1987) and Special Sensor Microwave/Imager (SSM/I) (1987 to 2007) that have been reprocessed using the same technique as that used in the processing of AMSR-E data. Improvements in accuracies in trend analysis when AMSR-E data are used are also assessed. The paper by Stammerjohn et al. [2008] provide results from analysis of long-term variability in the ice cover in the Southern Hemisphere and at the same time show significant correlation of such variability with ENSO and the Southern Annular Mode. In the Massom et al. [2008] paper, similar studies were made but more specifically in the region west of the Antarctic Peninsula where a climate anomaly has been reported previously. Another regional study is presented in a paper by F. Nishio et al. (Physical and radiative characteristics and long-term variability of the Okhotsk Sea ice cover, submitted to Journal of Geophysical Research, 2008). The paper by Spreen and Hegster [2008] makes use of AMSR-E's highest-resolution data (i.e., 5 km using the 89-GHz channels) and demonstrates the unique capabilities of the sensor to characterize the spatial distribution of the sea ice cover.

3. Ice Thickness and Snow Cover

[5] One of the biggest unknowns about the sea ice cover is its thickness or its overall thickness distribution. Drafts of the Arctic sea ice cover have been recorded by upward looking sonars on board nuclear submarines [Rothrock et al., 1999; Wadhams and Davis, 2000] and although such measurements provide accurate ice thickness data, the spatial and temporal coverages have been very limited. Recent advances with the advent of new space technologies have enabled more comprehensive coverage but interpretation of results is difficult. There are several papers in this special section that discusses various topics associated with sea ice thickness measurements and interpretation of the space based data. In particular, there are two papers presenting studies of large-scale variability of the freeboard of sea ice as inferred from ICESat/Geoscience Laser Altimeter System (GLAS), one for the Arctic [Kwok et al., 2007a] and another one for the Antarctic [Zwally et al., 2008]. The results of an analysis of thickness distribution of sea ice in the Antarctic from a vast collection of ships measurement are also presented in a separate paper by Worby et al. [2008a]. Data used in this paper could serve as the baseline for evaluating aircraft and satellite measurements. Aircraft data from concurrent measurements using the Airborne Themapic Mapper (ATM) which is a Lidar ranging system that measures the top of the snow cover and the Delay-Doppler Phase Monopulse radar (D2P) which is an advanced microwave altimeter system that measures the freeboard of the ice (or ice at the snow ice interface) have also been analyzed and results are presented in a paper by Leuschen et al. [2008]. Regional studies of ice thickness, primarily those of new and young ice, using AMSR-E data in conjunction with ship measurements have also been done, one in the Okhotsk Sea [Naokii et al., 2008] and the other in the Cosmonaut Sea in the Southern Ocean (K. Tateyama et al., Ship-based sea-ice thickness measurement in the Antarctic Ocean using the electro-magnetic inductive device and microwave radiometer, manuscript in preparation, 2008). Snow cover over sea ice is also an important parameter since it is such a good insulator and its albedo is very high. A few AMSR-E snow depth pixels over East Antarctic sea ice were validated with in-situ and aircraft measurements, during the Antarctic Remote Sensing Experiment (ARIS) onboard an icebreaker [Worby et al., 2008b]. Snow-ice formation for sea ice regions around Antarctica was derived using passive microwave snow depth, ice concentration, and ice motion in combination with snow fall data from the European Center for Medium Range Weather Forecasting (ECMWF) reanalysis [Maksym and Markus, 2008]. Surprisingly, snow-ice formation is largely independent of snow depth.

4. Mesoscale Characteristics and Special Studies

[6] There are three papers on specialized sea ice related topics. One paper present results from analysis of thin sea ice cover, using AMSR-E and ENVISAT/ASAR data, at the Ross Sea polynya [Kwok et al., 2007b] which is one of the primary sites of bottom water formation in the Southern Ocean. The paper also provides an evaluation of how the formation process has been impacted by recent calving of large icebergs. Another paper provides results from the first quantification of the areal extent of fast ice in the Antarctic [Giles et al., 2008] and also new insights into the scientific importance of this ice type and how well it can be monitored by satellite data. The third is an article on primary productivity in the Southern Ocean [Smith and Comiso, 2008] that assesses the role of the region as a carbon sink, quantifies the spatial distribution of planktons, which is at the bottom of the food web, and evaluates how plankton blooms are influenced by the melt.

5. Modeling and Process Studies

[7] New insights about the large-scale variability of the Antarctic sea ice cover are provided in a phenomenological study of the climate modes in the southern high latitudes by X. Yuan and C. Li (Climate modes in southern high latitudes and their impacts on Antarctic sea ice, submitted to Journal of Geophysical Research, 2008). The time variability is shown to be correlated with a mode-2 system as derived from pressure data. The use of sea ice concentration and ice motion satellite data as a means to improve sea ice models is also discussed in a paper by Stark et al. [2008]. Currently available satellite instruments for measuring snow thickness and ice thickness have a lot to be desired and in a modeling study by Varnai and Cahalan [2007], a new laser system that makes use of an innovative technique (different from a ranging system) is discussed and its effectiveness and accuracy are quantified.

6. Summary

[8] The sea ice cover protects the ocean from loosing heat in winter and cools the polar regions owing to its high reflectivity in the summer. The recent variability and trends observed in both polar regions manifest a response to climate forcings. Hence, sea ice extent and thickness are sensitive climate proxy values and need to be monitored accurately over decades during all weather conditions to shed light on changes in these remote regions. It is therefore important to evaluate the information that are being provided by the key sensors used for monitoring them. Quantification of errors is also needed to be able to assess the significance of the trends.

[9] The set of papers represent new techniques and results from studies of ice concentration, ice extent, ice area, and ice thickness and its snow cover, and the large-scale synoptic forcing of the sea ice cover. Analysis of errors are provided and procedures needed to be followed to ensure consistency and accuracy of the time series of data derived from satellite sensors are discussed. These results are useful and informative, and they provide incentives for further research that will lead to a greater understanding of the role of sea ice in the climate system.