- Student Dissertation or Thesis
As the impacts between land cover change, future climates and ecosystems are expected to be substantial (e.g., Feddema et al., 2005), there are growing needs for improving the capability of simulating the global vegetation structure and landscape as accurate as possible. In order to serve these needs, Dynamic Global Vegetation Models (DGVMs) are used to describe the current status of vegetation structure and biogeography as well as estimate their future states, either with prescribed climates or coupled to climate models. Yet, current DGVMs assume ubiquitous availability of seeds and do not consider any seed dispersal mechanisms and/or plant migration process, which may influence the assessment of impacts to the ecosystem that rely on the vegetation structure changes (i.e., change in albedo, runoff, and terrestrial carbon sequestration capacity). For the first time, this study incorporates time-varying wind-driven seed dispersion (i.e., the SEED configuration) as a dynamic constraint to the migration process of natural vegetation in the Community Land Model (CLM)-DGVM.
Compared to the satellite-derived tree covers, the result shows significantly improved representation of vegetation in regions such as boreal forests in Western Siberia and temperate forests in Eastern Europe. The prevailing wind pattern, along with the existing vegetation structure in nearby grid cells, alters the competition dynamics of the trees in these regions by filtering unrealistic saplings out and adjusting their establishment rates.
The SEED configuration is then applied to project future vegetation structures under two climate mitigation scenarios (No-policy vs. 450ppm CO2 stabilization) for the 21st century. The results indicate that regional changes of vegetation structure under changing climates are expected to be significant. In the high latitudes, regions such as Alaska and Siberia are expected to experience substantial shifts of forestry structure, characterized by expansion of needle-leaf boreal forest and shrinkage of C3 grass Arctic. In the mid-latitudes, temperate trees are likely to expand in South America, South Africa, and East Asia, showing sensitive responses to changing climates for the latter part of the 21st century. In Tropics, a most notable degree of change in the composition of tropical trees and C4 grass are projected in Amazon and also regions in Africa.
The vulnerability assessment suggested by this study shows that vegetation structures in Alaska, Greenland, Central America, southern part of South America, East Africa and East Asia are susceptible to changing climates, regardless of the two climate mitigation scenarios. Regions such as Greenland, Tibet, South Asia and Northern Australia, however, may substantially alleviate their risks of rapid change in vegetation structure, given a robust greenhouse gas stabilization target.
The impacts of future vegetation change on radiation budget cannot be neglected. The results suggest that depending upon the climate mitigation scenarios, the vegetation change may accelerate or offset the anticipated warming trend of the 21st century. Proliferation of boreal forests in the high latitudes is expected to amplify the warming trend (i.e., a positive feedback to climate), if no mitigation policy is implemented. In contrast, under the 450ppm scenario, vegetation structure may buffer the warming trend, which is a negative feedback to climate. A series of hydrologic processes including interception of rainfall by forest canopy, evapotranspiration, and runoff are to be influenced by the modified vegetation structures. The magnitude of the runoff response by the vegetation change may not exceed the direct response from hydro-climate change; however, the spatial pattern of runoff change due to vegetation indicates that vegetation change may offset or be complementary to increase in runoff due to enhanced precipitation under climate warming. Reduction of terrestrial productivity and a conservative estimate of vegetation carbon storage (-8PgC/yr and 24PgC, respectively under the NP scenario) in the 21st century may be due to ignoring the CO2 fertilization effect and partially applying the new SEED configuration to project future vegetation structures.
The newly developed SEED configuration may serve to attain more comprehensive representations of future vegetation structures and thus assess the impacts of natural vegetation distribution on the ecosystems. The results may also be used as indicators of assessing vulnerabilities in providing ecosystem goods and services.