Paleoclimatic inferences from glacial studies have depended on the theoretical framework of glacial geology as well as field observation and dating, and this may be also true for current global-warming studies. The framework has changed with time due to improved dating competing with integrative theories. In the 1920s Matthes and Blackwelder debated the correlation of advances across a single mountain range; by the 1980s advances were supposedly synchronized globally. Now synchronism appears to be a matter of scale. At the local geographic and short time scales, even adjacent glaciers may advance at different times; at coarser and longer scales advances appear to be synchronized regionally. Viewed globally, such regional advances may or may not be synchronous. The scale of the granularity may be controlled by the scale of storm tracks, ~750 km, and modulated topographically. This is well illustrated in Central Asia. From the Pamir to the Himalaya the LGM (~18 ka) saw smaller glaciers than earlier in the last glacial cycle, in distinction to the familiar pattern of the Alps. Larger advances also occurred early in the Holocene, attributed to the enhanced Monsoon then. In the western Tien Shan the Pleistocene chronology was similar, but the early Holocene advances were absent, and glacierization is thought to be driven by westerly cyclonic storms and Siberian thermal highs. Yet farther east the climate controls are apparently similar but the ELA was strongly depressed during the LGM, and the timing was similar to the “global” pattern.

Paleoclimatic conditions are inferred from ELA depression via simplified relationships between ELA and precipitation and/or temperature. A recent study of ELA sensitivity to Holocene climatic conditions (Rupper, 2007) suggests that this inference may be less certain than assumed, because the ELAs respond strongly to different factors such as sublimation under some climatic regimes. Thus, the most useful information may be the spatial/temporal patterns themselves, which may indicate regions within which disparate types of climatic evidence from palynology and hydrology may be grouped with glacial evidence for interpretation. It follows that “global warming” glacier studies may best treat retreat as a regional, not global, phenomenon.