Daimaru, H., Ohtani, Y., Ikeda, S., Okamoto, T. and Kajimoto, T. 2002. Paleoclimatic implication of buried peat layers in a subalpine snowpatch grassland on Mt. Zarumori, northern Japan. Catena 48: 53-65.
As background for their study, the authors write that "in the subalpine zone of northeastern Japan, sites where the snow cover disappears after July are usually occupied by 'snowpatch bare grounds' with extremely poor vegetation cover," and they say that this area is "encircled by snowpatch grassland." As a result, they say that "litter fall and the organic content in topsoil decrease toward the center of a snowpatch," so that "in snowpatch grasslands, peaty topsoil is restricted to sites where snowmelt comes early," i.e., the outer reaches of the snowpatch and the surrounding area. Therefore, if there is a buried layer of peat within the confines of the central portion of a snowpatch that lacks peat near the surface today, it follows that at the time the buried peat layer was laid down, the climate was warmer than it has been in more recent times.
Working in a snowpatch grassland within a shallow depression of landslide origin on the southeastern slope of Japan's Mt. Zarumori (~39.8°N, 140.8°E), Daimaru et al. dug 27 soil pits at various locations within and outside the central area of the snowpatch, carefully examining what they found and determining its age based on 14C dating and tephrochronology. In doing so, they encountered peaty topsoils in seven soil pits in the dense grassland (basically outside the area covered by the seasonally-occurring snowpatch), whereas sparse grassland (within the area seasonally covered by the snowpatch) lacked peaty topsoil. However, they report finding buried peat layers in soil pits within both dense and sparse grasslands; and they say that most of the buried peat layers contained a white pumice layer that fell in AD 915. This observation, plus their 14C dating, led them to conclude that the buried peat layers had to have been laid down during a period of time that was warmer than it has been recently, which they estimate to have been between AD 900 and 1000, which is close to the central portion of the global Medieval Warm Period delineated in our Interactive Map and Time Domain Plot.