Effect of a simulated typhoon disturbance on forest and carbon dynamics in a cool-temperate forest in Hokkaido by SEIB-DGVM

Large storm is a major natural disturbance to forest ecosystem in coastal area of the world. Given the changing climate, large storm is expected to become stronger and to appear more frequently in the future, especially in the Northwest pacific by some climate projections. Though destructive typhoon gives strong impact on the carbon and forest dynamics, leading to feedback to climate change, the long-term effect by its frequency and intensity on forest and carbon dynamics has not been estimated under the changing future climate.

To elucidate them, we adjust the parameters of a dynamic vegetation model at a deciduous-coniferous mixed forest in Tomakomai, northern Japan, to represent the development of larch tree plantation from 1954 after the destructive typhoon Toyamaru visit, and the natural regeneration from 2004 after Songda visit for validation with eddy flux and biomass data. Next, we conducted simulation during 2017 to 2100 with forecasted climatic environment along with a series of scenarios for intensity and frequency of typhoon under RCPs 2.6 and 8.5. Frequency and intensity are defined as the number of typhoons during the 84 years and as the tree mortality that is directly caused by typhoon, respectively. We categorize the typhoon frequency into six ranks: 1, 2, 4, 8, 16, and 20 times for 84 years and divided typhoon intensity into 10 levels with interval of 10% from 10-100%. Simulations were conducted for all possible combinations (60 pairs) of the frequency and intensity.

Result showed that there is a significant effect of typhoon frequency and intensity on forest structure and carbon dynamics for both RCPs 2.6 and 8.5. As both typhoon frequency and intensity increase, AGB, tree height and tree LAI decrease, but grass LAI increase. The interannual changes in GPP, NPP and NEE showed the significant positive relationship with increasing typhoon intensity and frequency. We obtained the second order polynomial functions between AGB, NPP and frequency, intensity of typhoon, respectively. The residuals are small and r² are relatively high (r²>0.9) for both RCPs 2.6 and 8.5. Therefore, these two equations can predict the AGB and NPP loss with increasing the frequency and intensity of typhoon under RCPs 2.6 and 8.5.

Keywords: typhoon intensity and frequency, forest dynamics, material cycle, typhoon scenario simulation