Abstract
The glaciers of the Southern Alps of New Zealand play an important role in understanding regional and global patterns of climate change. They constitute the third largest ice mass in the Southern Hemisphere after Antarctica and South America, and present unique glaciological features in the region: high sensitivity, high input-output systems that represent the temperate, maritime end of glacier process-behaviour continuum. Since 1977 equilibrium-line altitudes (ELA) of 48 New Zealand index glaciers have been surveyed using oblique aerial photographs as a low-cost proxy for monitoring glacier mass balance. Although this method yields good results, it is location-specific. The use of synoptic high resolution satellite imagery can potentially maximise opportunities to get a wider picture of glacier response to climate variability throughout the New Zealand Southern Alps. Hence, our objective was to investigate how well ELAs detected from 15 m ASTER satellite imagery compare with reference ELAs obtained from aerial photographs taken approximately at the same time, and thus determine whether this imagery could in the future provide a more comprehensive view of glacier wellbeing and trends throughout the Southern Alps. End-of-summer ASTER images were selected and orthorectified. Glaciers were extracted, and ice and snow patterns were classified, using supervised classification of principal components. ELAs were digitised and projected over a digital elevation model. The comparison of ELAs gave a mean height departure of 61 ± 55 m, or 14.4 ± 13.4% expressed relative to the glacier elevation range. We conclude that ASTER data could extend ELAs estimated by the current snowline program across the whole Southern Alps. However, small and steep glaciers with complex shapes generate large errors (dominant in the New Zealand context), and thus we suggest that only the largest glaciers can be monitored reliably. ASTER imagery cannot easily replace the annual aerial surveys, mainly because of the difficulty in obtaining the appropriate imagery. The short temporal window, high percentage of cloud cover within the area, and the potential for early unseasonable snowfall are significant limitations to the use of ASTER imagery, without mentioning the recent deterioration of the shortwave infrared ASTER instrument.