中文摘要：

以天山北坡森林生态系统为对象,对Biome-BGC模型的生理生态参数进行了校正,模拟了1959—2009年气候变化及不同气候变化情景下,研究区森林生态系统净初级生产力（NPP）、碳收支（NEP）和土壤呼吸的变化趋势。结果表明：1959—2009年该区森林生态系统年均NPP为547.97 g·m-2·a-1,年均NEP为61.24 g·m-2·a-1,表现出适度的碳汇效应,但碳汇效应整体呈递减趋势;在设定的温度、降水情境下,NPP和NEP对降雨变化的敏感性,较其对气温升高的敏感性强;在平均气温升高、降雨量不变的情景下,森林生态系统NPP呈微弱增加态势,但NEP呈减少趋势,主要原因为气温升高,导致土壤呼吸增强。通过分析气温在干旱区的正负效应,表明气温升高促进了天山北坡森林生态系统的NPP;干旱胁迫在该区域不显著。

英文摘要：

Forest ecosystems attract a continuous attention because of their importance for mitigating environmental change. Policy making in the fields of sustainable forest management and agendas related to climate change drives the research on forest carbon budget and its interpretation. The recent adoption of the Kyoto Protocol further spurs the need of a sound understanding of carbon-related processes. There is a big challenge to accurately estimate the contribution of forest ecosystems to the global carbon cycle, as it is difficult to directly measure carbon pools or fluxes over large area. Some models are available to address the issues of carbon budget, and they may be classified into several ways. In terms of their capability to include the causality, the models can be either empirical regression models, or so-called process models including eco-physiological processes describing ecosystem functioning in terms of key processes with their interactions. The adopted simulation system may include all kinds of combinations of ap- proaches to fulfill the modeling goals and create the specific model hybrids. In any case, the models developed based on the generalized approach to simulate ecosystem development must be parameterized for the target ecosys- tems. Regarding the estimation of carbon dynamics, the use of process-based ecosystem models is of particular in- terest because this approach allows not only to estimate the carbon budget under various environmental conditions, but also to interpret and quantify the possible causes of carbon stock change along with the environmental change. The application of ecosystem models, however, encounters some specific difficulties in dryland environment, which is characterized by its climatic and human-induced features. After identifying the proper ecophysiological parame- ters used in Biome-BGC model, in this study the NPP （net primary productivity） and NEP （net ecological productivity） in the Tianshan Mountains were estimated. Furthermore, the responses of NPP/NEP to int