Time-series monitoring of the regeneration process of tsunami-affected coastal forests through comparative study of multiple vegetation indexes using near-infrared high-resolution satellite images

The 2011 off the Pacific coast of Tohoku Earthquake occurred, and coastal forests were severely damaged by a huge tsunami. Since the disaster, coastal forest restoration projects have been underway by the Forestry Agency and local governments. Detailed time-series monitoring of the regeneration process of coastal forests is important in order to proceed with regeneration appropriately. The Normalized Difference Vegetation Index (NDVI), which uses near-infrared and visible red images obtained from optical satellite observations, has been widely used to survey trees and vegetation. However, it has been reported that NDVI tends to be saturated depending on the observation period and vegetation type. In addition, there is a tendency for index values to be overestimated on the soil surface. In particular, in the case of coastal forest regeneration, the influence of the soil surface is even greater because the complex mixture of soil surface and afforestation is assessed from observation images. To date, many improvement vegetation indices have been proposed to reduce soil surface effects and more appropriately evaluate vegetation activity. However, the applicability of improvement indexes using higher-resolution satellite images for evaluating the regeneration of tsunami-affected coastal forests has not yet been sufficiently investigated.