Debris-covered glaciers are widespread in the mountain chains of Asia (e.g. the Karakoram, the Himalaya (Reference Moribayashi and HiguchiMor-ibayashi and Higuchi, 1977) and the Tien Shan). They are also particularly common in New Zealand (Reference Kirkbride and WarrenKirkbride and Warren, 1999), in the Andes and in Alaska. Despite their relatively common occurrence, debris-covered glaciers have not been well studied (Reference Nakawo, Raymond and FountainNakawo and others, 2000). A debris cover that partially or completely masks the glacier ablation zone significantly influences the surface energy flux, the ablation rate and the discharge of meltwater streams. By reducing ablation, a debris cover produces an alteration of the response of the glacier to climate change. It is important, therefore, with the aim of understanding how a persistent debris cover influences terminus fluctuations, to collect quantitative information on length and thickness variations.

In the European Alps, there are only a few examples of such glaciers. Miage and Brenva glaciers, which drain the south slope of Mont Blanc in Valle d’Aosta, western Italian Alps, are two of the best known (Fig. 1). Brenva glacier (8km² wide and 7km long) originates on the east side of Mont Blanc (Monte Bianco; 4810 ma.s.l.) and descends steeply to the floor of Val Veny where it terminates at 1415ma.s.l. This is the lowest-altitude glacier front in the Italian Alps and the closest to a permanently inhabited location (the village of Entreves and the Mont Blanc tunnel entrance are 1 km distant; Fig. 2). The terminal zone is extensively mantled with granitic debris, mainly deriving from large rock falls that were deposited in November 1920. Until that event, its terminus fluctuations were comparable with other Mont Blanc clean glaciers, such as Glacier d’Argentiere or Mer de Glace (Reference Orombelli and PorterOrombelli and Porter, 1982); after the rock falls, Brenva glacier diverged from the general trend. While most alpine glaciers retreated, it continued to advance, in the 1940s almost reaching the 1818 limit. From then, it began to recede. A renewed advance commenced between 1965 and 1967, reaching a new maximum in 1991 (Reference CeruttiCerutti, 1993). The latest measurements of terminal variation show a new reduction phase (Reference DelineDeline, 2002). Despite the often detected short-term terminus fluctuations of Brenva glacier, its tongue thickness and volume variations are poorly known.

Fig. 1.

Brenva glacier location.

Fig. 2.

A view of the east side of Mont Blanc with Brenva glacier.

Current work deals with processing and comparison of large-scale maps and aerial imagery of the glacier, supported by Geographical Information System (GlS)-based methodology. The maps used were: (i) EIRA (1972) (tongue only); (ii) Alifoto (1972) (tongue only); (iii) Regione Valle d’Aosta (1988);and (iv) Regione Valle d’Aosta (1999). For the same purpose, the following aerial photographs have been used: (i) CGR (Italian General Company Aerial Surveys), flight RAVDA (Regione Autonoma Valle d’Aosta), scale 1:17000 image, 1991;and (ii) CGR, flight RAVDA, scale 1: 17000 image, 1997. The 1991 and 1997 stereo pairs were scanned using a Wheerli flatbed photogrammetric scanner at a resolution of 2116 dpi, which results in a pixel size of 12 pm, corresponding to approximately 18 cm on the ground.

A Digital Photogrammetric Workstation (Menci Software) was adopted for the generation of the digital terrain models using semi-automatic and automatic mode. The residuals from external parameter estimation were always <15 cm, demonstrating the overall good quality of the photogrammetric surveys. Ground-control points used to calculate the external orientation parameters were defined in a Universal Transverse Mercator grid (European datum 1950), and elevations measured above sea level. The adopted reference system allows comparison between maps and terrain models to detect multitemporal glacier tongue changes.

The maps and orthophotos were managed after digitization using GIS software. Digital elevation models were produced, and their comparison allowed us: (i) to quantify surface, volume and thickness variations of the Brenva glacier tongue in the second half of the 20th century; (ii) to elaborate ice-thickness variation thematic maps; and (iii) to elaborate longitudinal and transverse profiles to underline thickness variations (Reference Fox and NuttallFox and Nuttall, 1997; Reference KaabKaab, 2000; Reference KaabKaab and Vollmer, 2000).

Preliminary results obtained through differential analysis show that:

It should be noted that the 1997 aerial photograph was acquired in October. In January 1997, a large landslide from the upper basin of Brenva glacier covered the whole tongue. The resulting granitic deposit thickness was about 1 m (Reference DelineDeline, 2002), which was superimposed upon the former debris cover. The above volume and thickness calculations may therefore be a slight underestimate.

Summarizing, for Brenva glacier during the second half of the 20th century (1959-97) a net volume increase of 15.6 × 10⁶m³, i.e. a thickness change of 23.3 m, or 0.41 m a^-1 (0.37 m w.e. a^-1), occurred (Fig. 3). This pattern of thickness change contrasts with recorded fluctuations of non-debris-covered Italian glaciers. For instance, Lys glacier (Monte Rosa group) and Forni glacier (Ortles group), two of the widest glaciers on the south side of the Alpine chain, experienced strong thickness reduction of their tongues: from 1 953 to 1994, Lys glacier lost 15.6 mw.e. (0.38 m w.e. a^-1) (Reference Rota, Bonacina, De, Rossi and SmiragliaRota and others, 2001), and from 1953 to 1988 Forni glacier lost 21.8mw.e. (0.48mw.e.a^-1) (Reference Merli, Pavan, Rossi, Smiraglia, Tamburini and UbialiMerli and others, 2001).

Fig. 3.

Changes in surface elevation on the tongue of Brenva glacier, 1959_97.

The Brenva glacier tongue exhibited two main phases during the second half of the 20th century: a strong positive pattern of volume and thickness change between 1959 and 1983, and a less strong negative pattern between 1983 and 1999. The first phase may be explained by a drop in summer temperature (between 1951 and 1960 the temperature was 0.65°C lower at Courmayeur, close to the glacier, than the 1936-83 mean), accompanied by increased annual precipitation. This climatic pattern caused a reduction in the ablation of Monte Bianco glaciers, resulting in their expansion during the 1962-89 period (Reference CeruttiCerutti, 1993). Furthermore the trend was widespread throughout the Alps (Reference WoodWood, 1988). The second phase was triggered by climate warming after 1985, accompanied by reduced precipitation, causing negative glacier surface mass balance.

Earlier Brenva glacier results are comparable with the general trend of the ‘clean’ Italian glaciers (they also registered a phase of positive mass balance), but the recent behaviour of the glacier is very different. In fact, at the end of the 20th century, Brenva glacier and other debris-covered glaciers such as Miage and Belvedere (Reference Smiraglia, Diolaiuti, Casati and KirkbrideSmiraglia and others, 2000; Reference Diolaiuti and D’AgataDiolaiuti and others, 2003) showed a net positive balance because of the debris insulation effect, allowing more ice at the lower tongue of the glacier to be conserved.