5:15 PM - 6:45 PM
[BCG04-P03] Topography and vegetation-based estimates of soil methane flux in a complex forested watershed
Keywords:Methane, Watershed, Vegetation, Topography, Digital Elevation Model, Machine learning
Methane (CH4) is a potent greenhouse gas contributing to global warming which concentration in the atmosphere has dramatically increased in recent years. Forest soils, especially soils from upland forests are thought to be a biological sink of atmospheric CH4 but landscape heterogeneity may influence the sink capacity of the soil due to spatial variations of its characteristics. Soil aeration, which can be determined by soil moisture content, is a key factor of CH4 flux. In temperate forested watershed, topographic variations and vegetation composition controlled the soil hydrology, which may impact on soil moisture content, and therefore soil CH4 flux. Therefore, the role of vegetation and the influence of the topography on CH4 flux in temperate mountainous forests should be accounted for when scaling flux at landscape level.
We measured CH4 flux in the upper watershed of the Yura River in the Ashiu Experimental Forest of Kyoto University (area 62 ha and elevation 600-850 m) covering different topographic positions and vegetation composition (53 sampling points). Fluxes were measured using cavity-enhanced absorption spectroscopy gas analyzer nine times from April to December 2023. Across the mountain landscape we classified the topography into ridge, middle slope, bottom slope, creek, valley and wetland, and vegetation into evergreen, deciduous and mixed types. A five-meter Digital Elevation Model was used to characterize the landscape features (slope, aspect, topographic position index, upslope accumulated area, topographic wetness index) that influence the circulation and accumulation of water, and energy availability. These landscape features were further used as potential spatial drivers of CH4 flux to model and upscale the CH4 flux at the level of a complex forested watershed.
The results showed that the soil CH4 fluxes varied across the landscape, with the soil from ridges, middle slopes, bottom slopes, creek, and valley being net sinks of CH4 whereas wetlands being sources of CH4 throughout the study period. The most negative soil CH4 fluxes were observed on the ridge and middle part of the slope and the least one in the valley. A random forest algorithm is currently used to upscale CH4 flux measured each month to the upper watershed of the Yura River. Preliminary results showed that the topographic wetness index, which is derived from the slope and the upslope accumulated area, seems to have a significant influence on soil CH4 flux across seasons.
The findings of this study indicated that topography, which can be finely characterized over large areas using Digital Elevation Model, offers valuable insights for future estimations of CH4 budgets in mountainous regions.
We measured CH4 flux in the upper watershed of the Yura River in the Ashiu Experimental Forest of Kyoto University (area 62 ha and elevation 600-850 m) covering different topographic positions and vegetation composition (53 sampling points). Fluxes were measured using cavity-enhanced absorption spectroscopy gas analyzer nine times from April to December 2023. Across the mountain landscape we classified the topography into ridge, middle slope, bottom slope, creek, valley and wetland, and vegetation into evergreen, deciduous and mixed types. A five-meter Digital Elevation Model was used to characterize the landscape features (slope, aspect, topographic position index, upslope accumulated area, topographic wetness index) that influence the circulation and accumulation of water, and energy availability. These landscape features were further used as potential spatial drivers of CH4 flux to model and upscale the CH4 flux at the level of a complex forested watershed.
The results showed that the soil CH4 fluxes varied across the landscape, with the soil from ridges, middle slopes, bottom slopes, creek, and valley being net sinks of CH4 whereas wetlands being sources of CH4 throughout the study period. The most negative soil CH4 fluxes were observed on the ridge and middle part of the slope and the least one in the valley. A random forest algorithm is currently used to upscale CH4 flux measured each month to the upper watershed of the Yura River. Preliminary results showed that the topographic wetness index, which is derived from the slope and the upslope accumulated area, seems to have a significant influence on soil CH4 flux across seasons.
The findings of this study indicated that topography, which can be finely characterized over large areas using Digital Elevation Model, offers valuable insights for future estimations of CH4 budgets in mountainous regions.