Phenological Analysis of Sub-Alpine Forest on Jeju Island, South Korea, Using Data Fusion of Landsat and MODIS Products

Sang-Jin Park,Dong-Kun Lee,Kyung-Min Park,Seung-Gyu Jeong,Yong Park,Dong-Seok Jang,Yong-Won Mo,Sang-Hyuk Kim

doi:10.3390/f12030286

Abstract

Climate change poses a disproportionate risk to alpine ecosystems. Effective monitoring of forest phenological responses to climate change is critical for predicting and managing threats to alpine populations. Remote sensing can be used to monitor forest communities in dynamic landscapes for responses to climate change at the species level. Spatiotemporal fusion technology using remote sensing images is an effective way of detecting gradual phenological changes over time and seasonal responses to climate change. The spatial and temporal adaptive reflectance fusion model (STARFM) is a widely used data fusion algorithm for Landsat and MODIS imagery. This study aims to identify forest phenological characteristics and changes at the species–community level by fusing spatiotemporal data from Landsat and MODIS imagery. We fused 18 images from March to November for 2000, 2010, and 2019. (The resulting STARFM-fused images exhibited accuracies of RMSE = 0.0402 and R2 = 0.795. We found that the normalized difference vegetation index (NDVI) value increased with time, which suggests that increasing temperature due to climate change has affected the start of the growth season in the study region. From this study, we found that increasing temperature affects the phenology of these regions, and forest management strategies like monitoring phenology using remote sensing technique should evaluate the effects of climate change.

Highlights

Introduction published maps and institutional affilRising global temperatures and atmospheric carbon dioxide levels, changes in precipitation frequency, and increasing severity of extreme climatic events are some of the impacts of climate change
In accordance with previous studies [27], the pilot prediction was conducted by setting the size of the search window to 5, 10, 25, 50, 100, 150, 200, 250, and 300 [27]
As a result of the pilot prediction, when the window size was set to 50, the R2 (0.795) and root mean square error (RMSE) (0.0402) values indicated a sufficient accuracy of the model (Figure 3a)

Summary

Introduction

Introduction published maps and institutional affilRising global temperatures and atmospheric carbon dioxide levels, changes in precipitation frequency, and increasing severity of extreme climatic events are some of the impacts of climate change. The resultant effects on the ecosystems due to these changes will intensify in a negative direction. These effects include changes in the distribution of forest organisms and populations as well as ecosystem function and composition [1,2,3,4]. Forest biodiversity adjusting to new environmental conditions will lead to species migration. Due to their vertical dimensions, mountains have unique climatic and biogeographical features and create gradients of temperature, precipitation, and insolation. Species compositions in forests are highly susceptible to climate change, and as a result vulnerable species and populations may become extinct

Objectives

Methods

Results

Conclusion