Sea ice observations

This page is part of the topic Observations, data accuracy and tools

The pre-satellite era

Since the days of the earliest explorers, ships’ logs have recorded encounters with sea ice. Captain James Cook frequently reported the presence of sea ice as he tried to push south toward the continent, as did Captain Fabian von Bellingshausen during his exploration in 1831. Mackintosh and Herdman (1940) compiled a circumpolar map of the monthly variation of the average sea ice edge, based on data from ships’ logs during the 1920s and 1930s. These were later updated and republished by Mackintosh (1972).

Whaling vessels also made many valuable observations of the postion of the sea ice edge. The region close to the ice edge is rich in food and many whales congregate there, attracting the whaling fleets. Most of these observations are for the summer season. De la Mare (1997) examined the whaling records, which provide the location of every whale caught since 1931. He suggested that there had been a big change in the location of the whaling vessels between the 1940s and the 1970s, with the summer sea ice edge having moved southward by 2.8° of latitude between the mid 1950s and the early 1970s. He inferred this as meaning that there had been a decrease of 25% in the area covered by sea ice. However, there is a great deal of debate over whether or not the locations of whale catches can be translated into ice edge estimates that are comparable to those made from satellite observations. It is particularly unfortunate that there is no overlap between the period covered by the whale catch data and the modern satellite observations of the ice edge. At the moment the de la Mare results are questioned in many quarters and cannot be regarded as proof of a major decrease in sea ice extent between the 1940s and 1970s.

Coastal stations have also made valuable sea ice observations, albeit from a very limited number of locations. Nevertheless, direct observation of the ice edge from land is difficult because the ice edge extends well into the Southern Ocean for much of the year, out of sight from most coastal observatories. Some island stations, such as Signy in the South Orkney Islands, have provided information on sea ice variability over many years, and revealed details of some important modes of climate variability, such as the Antarctic Circumpolar Wave (Nowlin and Klinck, 1986).

Satellite observations of sea ice extent and concentration

From the 1960s it was possible to obtain a broad-scale view of the distribution of sea ice from visible and infra-red satellite imagery. However, the imagery was only of value in cloud-free or partly cloudy conditions, which was a major handicap as the Antarctic sea ice zone is characterised by extensive low cloud cover. With the introduction of reliable satellite passive microwave observations in the early 1970s (Gloersen et al., 1992), the extent (the area bounded by the ice edge, which is often taken as 15% ice concentration) and area (the integrated area of ice within the ice edge) of Antarctic sea ice became confidently measureable.

The US Nimbus-5 Electrically Scanning Microwave Radiometer (ESMR) was launched in December 1972 and allowed the first all-weather mapping of Antarctic sea ice. This instrument only had one channel at 19 GHz, but the large contrast in the emissivity of sea ice and ice-free ocean enabled the development of an ice concentration algorithm, allowing the production of sea ice concentration maps from 1973 to 1976. The ESMR data provided the first observational data on the growth and decay patterns of sea ice for the entire Antarctic region.

Despite hopes for further developments in sea ice monitoring with the launch in 1975 of Nimbus-6/ESMR-2, with its dual polarized 37 GHz radiometer the instrument failed to perform well and no useful data were obtained. Further useful passive microwave data were obtained with the launch of the Scanning Multichannel Microwave Radiometer (SMMR), first on board the SeaSat satellite in July 1978, and then on Nimbus-7 in October 1978. The SMMR was a multifrequency system covering five frequencies from 6 to 37 GHz, all of them dual polarized (horizontal and vertical). The sensor was also conically scanning (i.e., incidence angle constant), and with its multifrequency capability ice concentrations were derived at a much better accuracy than with ESMR data. SMMR lasted for about 9 years, before it failed in August 1987, the DMSP/Special Scanning Microwave Imager (SSM/I) was already in operation and provided overlap data from mid-July to mid-August 1987. The SSM/I sensor has only 7 channels from 19 to 89 GHz, among which is the same set used for generating ice concentration maps from SMMR. The sensor is also conically scanning with similar resolutions, but has a wider swath; it has provided continuous data up to the present. The overlap allowed for comparison of the performance of the two radiometers and a confirmation that data from both sensors provided approximately the same results. In May 2002, the EOS/Advanced Microwave Scanning Radiometer (AMSR-E) was launched, and with 14 channels from 6 to 89 GHz, and much higher resolution, it has provided the baseline for sea ice studies.

The ESMR data set was very valuable and was unique when it first came out, but there were problems using it together with the other sets of data for time series studies. First, since it is a one-channel instrument, the ice concentration data are not as accurate because variations in temperature and emissivity of the ice cover could not be taken into account. Second, it is a horizontally scanning radiometer going from nadir to around 50o with varying resolution and with different incident angles. Third, there were many missing bits in the data stream, causing the elimination of a large fraction of the data and big data gaps in the time series. And fourth, there was no overlap of ESMR and SMMR data to enable assessment of differences of ice edge locations and concentrations derived from the two sensors. For uniformity, and accuracy in the trend analysis in Variability and trends in Antarctic sea ice using satellite data, we use data from the two sets of similar sensors (i.e., SMMR and SSM/I) to evaluate the variability and trends in the ice cover over the last 28 years. We also discuss, how we can utilize ESMR data as well as some ship observations during the pre-satellite era to improve our understanding of the long-term trend.

SARs flown on spacecraft can provide high resolution data on sea ice and reveal mesoscale ice motion and deformation, the development of leads and polynyas, ice type discrimination, sea ice roughness data and iceberg detection. In addition to the scientific applications of these images, ship operators in polar seas have benefited from recent advances in the near real time processing of SAR data, allowing them to be delivered quickly enough to assist ship navigation in sea ice.

Observations of sea ice thickness

While the early explorers made many observations of sea ice location and type, their logs do not contain information on sea ice thickness. Only in recent decades have vessels become more ice capable and spent more time south of the ice edge in support of logistic and scientific activities. Consequently the sea ice logs from these ships have become more comprehensive and often include an estimate of sea ice thickness, or ice type from which thickness can be inferred.

In 1997, SCAR established the Antarctic Sea Ice Processes and Climate (ASPeCt) programme. One of the programe’s first objectives was to collate the many disparate sea ice logs kept from icebreakers operating in the Antarctic sea ice zone. This effort focused primarily on the Australian, German, US and Russian national Antarctic programmes, which were known to have dozens of data sets containing information on the concentration, thickness and snow cover characteristics of the Antarctic sea ice zone. The data constituted a compilation of 23,391 individual ship-based observations collected from 81 voyages to Antarctica over the period 1981 – 2005, plus 1,663 aircraft-based observations. The ship-based observations are typically recorded hourly and include the ship’s position, total ice concentration and an estimate of the areal coverage, thickness, floe size, topography, and snow cover characteristics of the three dominant ice thickness categories within a radius of approximately 1 km around the ship (Worby and Allison, 1999). Not all observations contain this level of information, but at a minimum the partial ice concentrations and thicknesses (or ice types) were necessary for inclusion in the data set. The data are publicly available via the ASPeCt website (http://www.aspect.aq) or from the Australian Antarctic Data Centre (http://data.aad.gov.au/).

Time series of sea ice thickness can be measured by moored upward looking sonars. They consist of echosounders moored about 150 m below the ocean surface which record by travel time changes the variation of the ice draft. Mooring motions and changes of the sound velocities require extended corrections. Significant efforts occurred in the Weddell Sea to quantify the annual cycle of sea ice thickness and horizontal transports in the gyre circulation.

A new method for sea ice thickness measurements is by means of non-destructive electromagnetic inductive (EM) sounding which can be performed by moving an EM instrument above the snow surface either airborne or terrestrially.

It is hoped that in the future, processing of satellite altimeter data will allow the recovery of sea ice freeboard and therefore ice thickness. This may be easier in the Arctic since there sea ice is generally thicker and there is more multi-year ice present. In the Antarctic, the predominance of first year ice may make this a great challenge.

Sea ice extent proxies

ITASE and other Antarctic ice coring programmes are in the process of developing proxies for sea ice, a critical component in the climate system, through studies of sulfur compounds such as sulfate and MSA (methane sulphonic acid) (Welch et al., 1993; Curran et al., 2003; Dixon et al., 2005, Abram et al., 2007a). ENSO-sea ice connections are noted utilizing ice core MSA and sulfate series over the Ross Sea embayment region (Meyerson et al., 2002). The sea ice proxies rest on the premise that biogenic sulfur emissions from within the sea ice zone, when integrated regionally and through time, may be related to the total sea ice extent. This approach has been supported by observed correlations between MSA and sea ice (Curran et al., 2003; Foster et al., 2006, Abram et al., 2007a), however it is likely to be dependent upon the regional sea ice regime. For example, the positive relationship observed by Curran et al. is in the East Antarctic, where there is little residual summer sea ice and very little multi-year ice. Results from East Antarctica show evidence of a 20% decline in mean sea ice extent since the mid-20^th century, consistent with results reconstructed from whaling records (de la Mare, 1997) although the validity of such observations is questioned (Ackley et al., 2003). However this continuous sea ice proxy illustrates large decadal-scale variations superimposed on the trend.

2.20 The cross correlation pattern with R values for June - August MSLP anomalies for the Indian Ocean and Southern Ocean region and MSA variability at GF12, (97° E) in Queen Mary Land. This shows the strong relationship between MSA and the SAM.

Curran et al. (2003) have demonstrated that MSA variability in the coastal DSS ice core can be used as a proxy for regional sea ice extent. Accordingly they reconstructed the sea ice extent between 80° to 140° E back to 1840 AD, and showed the consistent decline in sea ice extent since the 1950s (decreasing MSA in the DSS ice). However, regional sea ice extent is a function of ocean temperature and circulation and atmospheric windfield divergence and convergence patterns. Hence, changes in the longwave atmospheric pattern may cause contrasting regional temperature and windfield patterns. Preliminary analysis of MSA concentrations in the GF12 core (Queen Mary Land, East Antarctica) indicates that MSA concentrations increased from 1950 to 1986 when the core was retrieved. These analyses may indicate a growth in regional sea ice to the west of 90° E. This is consistent with increased southerly wind outflow from the Lambert Basin and cool SST anomalies in this region. Hence, ice core sites located at different longitudes have the potential to provide historical data on regional sea ice changes around Antarctica. Figure 2.20 shows the June – August MSLP pattern associated with June - August MSA variations recorded in the GF12 ice core in Queen Mary Land, near 100° E.

Sea-salt levels have been proposed as an additional proxy of sea ice extent (Wolff et al., 2003) and applied to the long EPICA ice core record (Wolff et al., 2006). The basis of this proxy is the observation from near-coastal sites that ‘frost flowers’ are a significant source of sea-salt species. These ‘flowers’ form on new ice under calm conditions, producing fragile crystalline structures that are vulnerable to destruction and transport by wind. Salt species from frost flowers provide a distinguishing signature resulting from fractionation of sulfur content relative to other species. At coastal sites, where winter snow can be isolated, this signature is evident, however it appears to diminish in relative importance at sites further inland (M. Curran, personal communication) and its utility as a proxy at sites in the interior has not yet been clearly demonstrated.