中文摘要：

植被的净初级生产力（net primary productivity，NPP）是全球变化和碳循环研究中的一个非常重要的环节。为了更精确地模拟植被的净初级生产力，选取中国长白山自然保护区作为研究区域，针对基于过程的北部森林生态系统生产力模拟模型（boreal ecosystem productivity simulator，BEPS）仅考虑平坦立地条件的特点，通过对该模型中地面接收到的太阳辐射部分进行地形修正，首先估算长自山自然保护区森林植被的NPP；然后采用高分辨率（30m）的ETM＋遥感数据，利用修正前后的BEPS模型来模拟得到长白山自然保护区森林植被的净初级生产力；最后对模拟得到的NPP结果，再利用地面实测数据进行验证，其相关系数R分别为0．91659和0．92957，算术平均偏差分别为62．8gC／（m^2·a）和44．2gC／（m^2·a）。结果表明，通过对模型的进一步完善，BEPS对NPP的模拟精度有了一定程度的提高。

英文摘要：

Net primary productivity（NPP） is one of important components in global change and carbon cycle research, In order to estimate NPP in Changbaishan Natural Reserve in China accurately, we corrected BEPS （boreal ecosystem productivity simulator） in accidented terrain. In original BEPS, only fiat terrain is considered, which is suitable to Canada. While in China, it＇ s hilly. The simulated accuracy using BEPS will not be very high. According to the differences of accepting solar radiance in different pixels, BEPS is revised theoretically. For the validation of the topographic corrected BEPS, ETM ＋ image（30m resolution） is selected in Changbaishan Natural Reserve. Run the original and corrected BEPS, two NPP results are got. Contrast them with field data, the result of topographic corrected BEPS is better than that of original BEPS, and the correlation coefficients are 0.929 57 and 0. 91659, the average arithmetical errors are 44.2gC/（ m^2· a） and 62.8gC/（m^2· a） respectively. It is an improvement for BEPS in theory.