3:45 PM - 4:00 PM
[HTT14-02] Assessing watershed environments from mountainous areas to coastal zones and estimating sediment and large wood debris volumes by considering forest and landscape metrics
Keywords:landscape metrics, land use, ecosystem management, major river watershed, open data
Vegetation, sediment, and water interact as key components within a watershed, with upstream forest conditions playing a crucial role in sediment and large wood debris generation. Forests also contribute to water retention, regulating fluctuations in water volume and runoff characteristics through processes such as precipitation infiltration, storage, and evapotranspiration. These hydrological processes influence sediment and large wood debris transport to downstream areas, affecting erosion and deposition at river mouths and coastal zones. Therefore, assessing the impact of forest environments on water and sediment dynamics within a watershed is essential for effective watershed-wide management.
The National Forest Plan, outlined by the Forestry Agency of Japan, assesses timber harvest and planning volumes at a broad watershed scale, necessitating the development of assessment methods for forest and watershed environments across large spatial scales. Similarly, the River Basin Disaster Resilience and Sustainability Project, promoted by the Ministry of Land, Infrastructure, Transport, and Tourism, aims to implement flood control measures and biodiversity conservation at a broad watershed scale, emphasizing the increasing importance of integrated watershed-scale management. Moreover, the recent availability of nationwide spatial datasets has expanded opportunities for research utilizing open data.
Given this background, this study constructed a database covering 109 major river watersheds in Japan, developed regression models based on sediment and large wood debris data from upstream watersheds, and explored methods for extrapolating these models to broader watershed scales. Previous studies primarily utilized variables related to topography, geology, and land cover, with forest-related indices largely limited to forest cover ratios and vegetation type area proportions. In contrast, this study focused on forest and ecosystem management by incorporating more detailed forest indicators (e.g., vegetation naturalness and aboveground biomass) as well as landscape metrics into the analysis.
First, we developed a generalized linear mixed model (GLMM) using sediment and large wood debris volumes in upstream watersheds as response variables and various explanatory variables, including socio-economic factors (e.g., population density, road density), climatic factors (e.g., precipitation, temperature), topographic factors (e.g., watershed area, watershed gradient), land cover factors (e.g., geology, vegetation), and forest structural factors (e.g., tree height, timber volume), which influence processes across multiple spatial scales. Additionally, we quantified the spatial patterns of land cover using landscape metrics to account for landscape ecological characteristics. The regression model was then extrapolated to a broader watersheds scale to estimate sediment and driftwood volumes across Japan’s major river watersheds. Finally, we validated the estimates by comparing them with changes in coastline morphology.
The National Forest Plan, outlined by the Forestry Agency of Japan, assesses timber harvest and planning volumes at a broad watershed scale, necessitating the development of assessment methods for forest and watershed environments across large spatial scales. Similarly, the River Basin Disaster Resilience and Sustainability Project, promoted by the Ministry of Land, Infrastructure, Transport, and Tourism, aims to implement flood control measures and biodiversity conservation at a broad watershed scale, emphasizing the increasing importance of integrated watershed-scale management. Moreover, the recent availability of nationwide spatial datasets has expanded opportunities for research utilizing open data.
Given this background, this study constructed a database covering 109 major river watersheds in Japan, developed regression models based on sediment and large wood debris data from upstream watersheds, and explored methods for extrapolating these models to broader watershed scales. Previous studies primarily utilized variables related to topography, geology, and land cover, with forest-related indices largely limited to forest cover ratios and vegetation type area proportions. In contrast, this study focused on forest and ecosystem management by incorporating more detailed forest indicators (e.g., vegetation naturalness and aboveground biomass) as well as landscape metrics into the analysis.
First, we developed a generalized linear mixed model (GLMM) using sediment and large wood debris volumes in upstream watersheds as response variables and various explanatory variables, including socio-economic factors (e.g., population density, road density), climatic factors (e.g., precipitation, temperature), topographic factors (e.g., watershed area, watershed gradient), land cover factors (e.g., geology, vegetation), and forest structural factors (e.g., tree height, timber volume), which influence processes across multiple spatial scales. Additionally, we quantified the spatial patterns of land cover using landscape metrics to account for landscape ecological characteristics. The regression model was then extrapolated to a broader watersheds scale to estimate sediment and driftwood volumes across Japan’s major river watersheds. Finally, we validated the estimates by comparing them with changes in coastline morphology.