Abstract
In Japanese high mountains, glaciers advanced during the Last Glacial period to form cirques and mo-raines. The chronology of these glacial landforms is important for examining landforms development in the high mountains and for reconstructing the Last Glacial paleoenvironment. In high mountains, however, it is difficult to find the key tephra layers and/or radiocarbon samples for dating. There-fore, weathering-rind thickness (WRT) of gravels (Fig. 4) has frequently been used for dating mo-raines and debris slopes. Recent studies in Japan have also determined the age of glacial and pen-glacial landforms using the WRT, yet little attention has been paid to the statistical significance of the dating. This study discusses the glacial fluctuation history of the northern Central Japan Alps on the basis of the statistical interpretation of the WRT values. Eleven moraines in five cirques were se-lected for the analysis (Fig. 3-a-c).
The measured WRT values are large for lower moraines, but small for higher moraines. In addi-tion, the WRT values have no correlation with the depth of the granular exfoliation pits developed on gravels, which reflect the effect of denudation on gravel surfaces (Fig. 7). Thus, the WRT values depend not on the difference in the denudation process but on the age of the moraines.
In the study area, two moraines were dated at 35ka and 25ka, respectively, using tephrochronolog-ical methods (Yanagimachi, 1983). The relationships between the ages and the WRT values for the two moraines can be expressed as:
W=4.33×10-2 t0.50
or
W=10.26log (1+1.36×10-4t)
The ages of the other undated moraines can be estimated by the substitution of their WRT values into these equations (Fig. 6; Table 2). The moraines can be divided into three groups according to their estimated ages: ca. 35ka, 1.9-2.6ka, and 11-12ka. This grouping is confirmed by the Kruskal-Wallis test, which is a nonparametric statistical method (Table 3).
Based on the WRT values and the mathematical errors involved in the WRT method, the ages of the glacial advance stages are determined as follows (Fig. 8) 1) Inagawa Stage, the early stage of the Last Glacial Age; 2) Nishisenjojiki Stage, around the Last Glacial Maximum; and 3) San'nosawa Stage, the Late Glacial. The age of the third stage has not been determined in previous research in the Central Japan Alps. The stage can be correlated to the Younger Dryas stage, because glacial ad-vances in Japan are known to have occurred during the colder substages of the Last Glacial Age.
