According to the AIS map 426a, 315, the morpho-lexical type commented on below is attested in Arabba (municipality of Livinallongo del Col di Lana). This is a designation for the concept GLACIER. The word is also listed in Blad under marmolada with the same meaning.
The Marmolata, at 3343 metres a.s.l., is the highest mountain in the Dolomites. Its northern face is covered by a glacier that used to reach almost as far as the Fedaia Pass (2057 metres a.s.l.), but is nowadays confined to the uppermost part of the mountain.
It can also be found in Tagliavini 1934 (202) as well as in Pult 1947 (41) where the word marmolada, meaning 'glacier', is given for the dialect from Livinallongo del Col di Lana. Consequently, the word would primarily designate the above-mentioned Dolomite peak, while the meaning 'glacier' could be the result of a metonymic process. Here, however, the following remark should be made: the Marmolata is a most impressive mountain, situated between the municipal areas of Livinallongo, Canazei, Rocca Pietore and Falcade. Accordingly, one could expect the same word or variants of it in the neighbouring idioms (at least in Canazei, since Rocca Pietore and Falcade do not represent AIS survey points). However, the type marmolada is only recorded as 'glacier' in Arabba on the AIS map. In Penia near Canazei (p. 313), /ˈʤaʧɐ/ and /ʤaʧˈoŋ/ (i.e. the morpho-lexical type glace (also with the suffix -on) (roa. f.) are attested, while in the slightly eastern Zuel near Cortina d'Ampezzo (p. 316), /ˈʒatso/ (morpho-lexical type ghiaccio roa. m.) appears.
Despite the record with the meaning GLETSCHER in the relevant literature listed above, it appears from our own surveys in Livinallongo del Col di Lana that Marmolada is only known as the name of the mountain in this area. Unfortunately, Crowdsourcing does not at the moment provide any evidence for the municipality of Livinallongo and the concept GLACIER. Nevertheless, the people who were interviewed are all native speakers of the dialect in question. It therefore seems justified to raise the question of the legitimacy of the linguistic record marmolada. With respect to this, it is possible to hypothesise that the word marmolada attested in the AIS is the result of a misunderstanding that arose in the course of data collection: while the researcher wanted to know which word was used for GLACIER in the dialect, the informant may have mistakenly given the name of the mountain. The commentaries in Tagliavini 1934 and Pult 1947, as well as Masarei's dictionary entry (Blad), might also have been written based on this very questionable AIS record (Volume III, 1930). On the other hand, one could also assume that the word was still common for GLACIER at the beginning of the 20th century when the AIS surveys were carried out, but is no longer known today. Among the people we interviewed were speakers of an older age who also claimed never to have heard or used the word marmolada as meaning 'glacier'. The first hypothesis therefore seems more plausible.