1-3 Dynamic vegetation model

a. Summary

The vegetation dynamic model used as the constituent factor is developed towards construction of a terrestrial integrated model. In order to acquire a fixed result under limited human resources and the limited period, this is taken as the vegetation dynamic model which specialized in vegetation change of the north region considered that the influence of warming arises most notably. In the high latitude area which is the extreme environment for a higher plant, disturbance by a very small climatic change and a very small forest fire has specified the vegetation dynamic state most strongly rather than the gap dynamic state assumed by the existing DGVM. Then, the polar ecosystem shift model ALFRESCO treating a specific dynamic state is used for such a high latitude area as a base. Although ALFRESCO predicts shift of a polar ecosystem based on experiential data, it does not treat the arbor standing crop or size structure which are indispensable information on the occasion of construction of an integrated model. Then, extension of including vegetable growth and diffusion model in ALFRESCO is tried so that these can be treated. Moreover, the next extension is tried in order to obtain the prediction which can place reliance more. (1) Treat appropriately the hetero target scene of (2) polar which treat seed diffusion clearly.

b. Research purpose

Although distribution of vegetation is greatly prescribed by climatic environment, climatic environment is also influenced of vegetation distribution and a feedback-interaction commits it among these both (Foley et al.2003). Therefore, in order to predict the earth environment of the future correctly, it is indispensable to clarify relation between a climate change and vegetation distribution change. It is thought that a large-scale climatic change produces especially a high latitude area for a short period of time (Katteberg et al.1996), and vegetation distribution prediction of this area is the subject that a priority is high.

Construction of a terrestrial integrated model is advanced in the 2nd project of symbiosis. This is a trial which builds an integrative terrestrial carbon cycle model unique to Japan by combining "vegetation physics process model MATSIRO", "vegetation material-recycling model Sim-CYCLE" (Ito and Oikawa 2002), and a "vegetation dynamic model." The target of this research is completing the "vegetation dynamic model" which has not been uniquely built among these constituent factors. Although the model which can predict vegetation distribution change with all ball scales and dynamic state change is finally needed, if it confronts, the model which specialized in the high latitude area is built, and it aims at joining together.

c. A research program, a method, a schedule

As an integrative model treating material recycling of the terrestrial ecosystem which aims at construction in this project, some models are already built as ball vegetation models (Dynamic Global Vegetation Models) dynamic [ all ] (Reviewed by Cramer et al.2001). DGVMs is combining a "material-recycling model", a "biological-processes model", and a "vegetation dynamic model", and it is the model which tries to predict the direction and speed of vegetation composition change when an environmental condition changes. However, no existing DGVMs(es) treat seed diffusion clearly, and cannot predict the vegetation distribution change under the rapid climate change that a seed rate of feed serves as rate controlling. Although the gap dynamic model (e. g. and Prentice and Leemans 1990) that moreover, next-generation sapling grows up from the bright forest floor side which "advanced age tree broke down from DGCM which is existing as a factor which specifies a vegetation dynamic state, and was produced" is assumed, disturbance with a forest fire larger-scale than such a gap dynamic state etc. which has many thin woods has specified change and changes of vegetation strongly primarily, and this assumption is unsuitable in the high latitude area which is the extreme environment for a higher plant, when treating the vegetation dynamic state of a high latitude area.

From these reasons, the vegetation zone change model which incorporated clearly "semination" and "a vegetation dynamic state peculiar to a high latitude area" is developed by this research. Among these, about "semination", the model which paid its attention to a physical process already exists, and it is shown that it can explain survey data well (Tanaka, Shibata, and and Nakashizuka 1998). However, when then how much many seeds are sprinkled, the data which performed measurement direct about weather conditions participating in that what number of sapling(s) can be established and there how is restricted. Then, in the area which recovery of the forest in the forest fire site and the large-scale felling site contiguous to a forest zone is producing as the indirect presumed method, the relation between the distance from a forest, and the number of sapling and an age distribution is measured, and the technique of presuming such a relation is also collectively developed by combining with the weather data of the past in the area. Since quality and quantity sufficient only from reference are not obtained, the field research data to be used performs a field survey in Alaska under cooperation of IARC.

In this research, the vegetation moving projection model ALFRESCO in a high latitude area (Starfield and Chapin 1996) is used as one of the bases. This ALFRESCO is the ecosystem shift model of the frame base which describes the change during an ecosystem type, it was expressed by sub model called a "frame" in each ecosystem type, and growth, disturbance, etc. have determined the conditions which change from a certain vegetation type to another vegetation type for every frame of this. ALFRESCO will calculate the change during four kinds of ecosystem types in a high latitude area at a step for ten years as a function of climate, disturbance (the outbreak of a forest fire and a vegetable feeding habit insect, deforestation), and the seed amount of supply (Fig. 1). For example, in the high-ground tundra, after the climate dried warmly if the degree of arbor covering grew to 50% or more, after it will become a spruce wood and a forest fire will arise continues during a fixed period, it becomes a frigid prairie. Moreover, in the spruce wood which is the climax, unless disturbance, such as a forest fire and felling, arise, it is stability, but after such disturbance arises, it shifts to a deciduous forest, and when disturbance and the climatic conditions unsuitable for growth of an arbor overlap, it shifts to a frigid prairie or the tundra.

Thus, ALFRESCO has described the change pattern during the ecosystem type based on the law learned by experience. However, ALFRESCO does not treat the arbor standing crop or size structure which are indispensable information on the occasion of construction of an integrated model. Then, extension of including vegetable growth and diffusion model in ALFRESCO is tried so that these can be treated. Moreover, although each grid is represented with a single scene at almost all the vegetation dynamic model including ALFRESCO, such a technique is not suitable, in order for the polar area to have the scene with very strong hetero nature with which a lakeside side, a flood plain, a permafrost zone, and heights geographical feature were mixed and to simulate more appropriately the carbon cycle and vegetation dynamic state in a grid (Kittel and Steffen and Chapin 2000). Although such hetero nature in a grid can be treated with parameterization or a statistical method, in order to use the surveyed vegetable population dynamics data and to make the interpretation of a result easy, it needs to calculate every geographical feature type. So, in this research, every two or more picking and scene are made to calculate the sampling calculation section of high resolution in a grid, and the technique of presuming scene change of the whole grid and vegetable biomass change later is developed.

Fig. 1 Outline of structure of ALFRESCO.
Each ecosystem type is expressed with a "frame" and this frame changes on each condition. It changes from Starfield and Chapin (1996).

d. The research program in FY 2002

e. Reports in FY 2002

f. Consideration

g. Bibliography

h. The announcement of a result