Pelicon was a joined EU-funded project of
Institute of Environmental Physics, University of Bremen
British Antarctic Survey, Cambridge
Alfred-Wegener Institute for Polar and Marine Research, Bremerhaven
Danish Center of Remote Sensing, Technical University of Denmark
Sea ice is an integral part of the climate system of the high latitude regions. Up to now, control of its occurrence and extent, especially its interannual and long-term variability, has proved difficult to isolate or quantify, especially in the data devoid regions of the Southern Ocean. The consequences of climate change induced, for example, by the continued increase of greenhouse gases, on the occurrence of sea ice, is even less certain, although the theory at least suggests that sea ice extent changes could be expected to have a positive feedback role, i.e. that reduced ice extent could be expected to enhance warming at high latitudes.
The interaction of sea ice with the polar ocean and atmosphere can be summerized as follows:
Satellite passive microwave imagers are measuring the upwelling radiation from the ground. The instruments are especially suited to monitor sea ice in polar regions because:
Satellite passive microwave imagers have provided this information for most of the past 20 years and are presently an integral part of operational monitoring programmes that will be continued into the foreseeable future. The potential of passive microwave data to observe sea ice distributions on a global scale is demonstrated with an animation of the annual cycle of ice concentration in the Antarctica.. The passive microwave sensors considered in this study are:
The extend of the sea ice in the Antarctica has been investigated from SSM/I ice concentrations. The NASA-Team algorithm (Cavalieri et.al., 1984) was used to calculate the ice concentration. A new scheme to correct the impact of the atmosphere on the microwave signal at low ice concentrations and open water has been delveloped (Thomas and Heygster, 1996).
For the trend analysis ice concentrations were averaged monthly. In order to remove the high seasonal variability, the climatological monthly mean was subtracted from the data to get the ice extend and icearea anomalies. The full SMMR period (1978-87) has been analyzed with an inter-sensor corrected NASA Team algorithm, the full SSM/I period (1987-96) with inter-sensor and weather corrected NASA-Team algorithm. Both time series have been combined in order to form a period of 18 years.
The above figure shows the combined and corrected SMMR-SSMI time series of the last 18 years. The icearea anomalies (blue) have been plotted as well as the trend (red).
The above figure shows the trend when only a subset of the time series was considered, beginning in 1978 ("The trend of the trend"). At the beginning of the investigation periode strong oscillations (blue) and a large standard deviation (red) can be observed. Both are decreasing year by year. In 1994 the trend started to be slightly larger than the standard deviation which confirms a positive trend. This means that we have a statistically significant positive trend of 1.55% per decade at the moment. The standard deviation of the estimate of the slope is 35%. The anomalies of the last years are mostly positive which caused the positive trend.
A detailed statistical analysis with an autoregressive model was made to confirm the results coming from linear regression. It was found that the significance of the trend of the ice-area time series without the weather correction scheme exceed the 95% confidence level. But the trend of the corrected ice-area time serie is not statistical significant.
Further observation of the Antarctic sea ice and a full statistical time serie analyse will be necessary to confirm or reject the trend-hypothesis. The uncertainty in the linear trends decreases with the length of the time series, emphasizing the importance of the intersensor calibration (SMMR and different SSM/I sensors). As the time series needs to be prepared as homogenous as possible, it is desirable to derive the weather correction scheme for the SMMR data by replacing the 21 GHz data with those of the adjacent 18 and 37 GHz channels and it is also desirable to derive a nonlinear intersensor calibration.
Sorry, but your browser can't run JAVA applets, so here's just a picture :
This JAVA-applet shows the Antarctic continent (black) and illustrates the dynamic of the surrounding sea ice. When you push the LOAD button the film will be loaded (about 250 KBytes) and the video player will be started. Weekly mean ice concentrations are shown for a whole year. Usually not all images can be captured during the first cycle (sometimes not even one!), but the remaining ones are picked up step by step in the following loops.
Features of the video player:
Links to oher sites
We greatfully acknowledge the National Snow and Ice Data Center (NSIDC), which delivered all the brightness temperature data on CD-ROM.