Comparison of vegetation patch dynamics after the eruptions of the volcano Mount Usu, northern Japan, in 1977-1978 and 2000, detected by imagery chronosequence
Comparison of vegetation patch dynamics after the eruptions of the volcano Mount Usu, northern Japan, in 1977-1978 and 2000, detected by imagery chronosequence
复制标题
1977-1978年和2000年日本北部有珠火山喷发后植被斑块动态的比较,通过影像时序检测
DOI:
10.1111/1440-1703.12199
复制
发表时间:
2021
影响因子:
2
通讯作者:
Shiro Tsuyuzaki
中科院分区:
文献类型:
--
作者:
Lea Vegh;Shiro Tsuyuzaki
Vegetation patch dynamics were analyzed to detect vegetation development patterns after eruptions on two sites (summit destroyed in 1977–1978, and a foothill, Konpira destroyed in 2000) on the volcano Mount Usu, in northern Japan. Aerial photos and satellite images taken in 2000, 2006, and 2014 were used to develop an imagery chronosequence of vegetation patch dynamics. Vegetation patches were identified by the Normalized Difference Vegetation Index (NDVI) for satellite images, and by the Normalized Green‐Red Difference Index (NGRDI) for aerial photos. We categorized the vegetation patch types based on whether the patches overlapped (touching) or not (isolated) with the future vegetation patches and whether their area increased (growing) or decreased (shrinking). Afterwards, patch dynamics were compared between the two sites through changes in patch types, dense vegetation, and patch growth with slope degree, elevation, and time. Isolated patches were established more at the summit and showed high mortality, while at Konpira most isolated patches survived until 2006 and merged into touching patches by 2014. Moreover, the vegetation density of patches was higher at Konpira than at the summit. Patch growth was associated with patch types at both sites. However, the time was more important for the patch dynamics at the summit, and the vegetation density affected the dynamics more at Konpira. Therefore, the two sites had different vegetation patch dynamics, which were related to the characteristics of topography and eruptions. In conclusion, the imagery chronosequence proposed in this study monitored patch dynamics well, and patches developed faster at Konpira.