SNOWLINE RECONSTRUCTIONS DURING THE LAST GLACIAL MAXIMUM: A PALAEOENVIRONMENTAL DATABASE AND SYNTHESIS OF TROPICAL CLIMATE

A new compilation of snowline reconstructions from the tropics and sub-tropics at the last glacial maximum (LGM=21±ka) provides information on glacier equilibrium line altitudes (ELAs) for over 400 glacier-valley specific localities. The data base contains metadata allowing the reliability of the reconstructions and the site chronology to be evaluated, as well as environmental information (e.g. aspect, elevation, glacier type) that could help explain intra-regional variability in the reconstructed ELA changes.

The reconstructed mean changes, weighted by aspect, in the ELA are smallest in the Himalaya region (500 m), relatively small in the southern central Andes (600 m) and East Africa (950 m), larger in the northern Andes (1200 m) and Papua New Guinea (1100 m) and largest in Central America/Mexico and Japan (1250 m). Similar general patterns are apparent in the minimum and maximum estimates of the changes in ELA. There are, however, large differences in the minimum and maximum estimates of change in ELA in some regions, and these differences are clearly related to the orientation of the glacial valley.

We have computed the implied change in temperature, assuming no change in precipitation, using modern lapse rates (computed from the CRU 10' modern terrestrial climate database) and assuming that the ELA position corresponds to the mean 0° C isotherm in summer (i.e. during the ablation season). By comparing the independent estimates of temperature changes with elevation, based on vegetation data (the 21ka TROPICS data set: Farrera et al., 1999, Clim Dyn 15:823-856), we are able to quantify the role of precipitation change in producing the changes in ELA at the LGM. Our database should serve as a useful benchmark dataset for the Palaeoclimate Modelling Intercomparison Project (PMIP).