干旱区内陆流域净初级生产力时空分异模拟及自然因素分析——以泾河流域为例

Estimation of Vegetation Net Primary Productivity in the Jinghe Basin

生物固碳是土壤-植被-大气-人类活动共同作用下的复杂生态过程，净初级生产力（NPP）是地球系统各圈层相互作用的产物。本文针对干旱区内陆流域的自然环境特征，利用光能利用率模型（CASA）对泾河流域1998年-2010年净初级生产力（NPP）进行时空分异模拟，分析其与土壤、气候、植被、地貌因素的耦合关系，深入理解NPP响应机理。结果表明，1998年-2010年间，泾河流域NPP总量浮动于（3．53—4．55）Tg．C／a之间，单位面积NPP产量约为136．77g．C／（m2·a），低于全国平均水平；NPP季节特征明显，春、夏、秋、冬的NPP产量分别占NPP年总量的22．91％、55．85％、19．99％、1．25％；流域NPP空间格局主要受自然因素控制，分析NPP与18个自然因子的相关性知，NPP与植被生理生态特征及气候因素的相关性最强，其中与NDVI、叶面积指数的相关性系数达0．53、0．78，与风速、温度、太阳辐射相关性分别为0．61、O．61、0．43；地形和土壤因素并非NPP响应的主导因素，但对植被有机物质的合成至关重要。

Net primary productivity （NPP） of terrestrial ecosystems is the material basis for human activity. The process of carbon fixation is an interaction between the biosphere, pedosphere, atmosphere and anthroposphere. Based on SPOT/VEGETATION data, land use/cover change, meteorology and soils this paper applies the Carnegie Ames Stanford Approach （CASA） to simulate yearly and seasonal variation in NPP for the Jinghe Basin from 1998 to 2010. The estimated NPP was validated using observed NPP and previous research. Our results show that the estimated annual average NPP for the Jinghe Basin was 3.53-4.55Tg.C/a. NPP accumulation for summer was 2.25Tg.C, 55.85% of the yearly total; NPP in spring was 0.92Tg.C; NPP in autumn was 0.80Tg.C; and NPP in winter was 0.05Tg.C. The spatial distribution of net primary production declined from the southeast to northwest. Among the 23 sub-watersheds of the basin, the tnmk stream of Jinghe NPP was highest （0.64Tg.C） and Jinghong stream was lowest（0.03Tg.C）. Simulate results compared to observed values and previous research showed that the measured NPP is less than the national average. This result is reasonable considering the natural conditions of the Jinghe Basin and its arid environment. The spatial pattern of NPP in the Jinghe Basin was controlled by natural environment and humanistic factors. Natural factors were the dominant factors influencing the distribution of NPP. Among the influencing factors, climate and physiological features are the leading factors affecting NPP in the arid ecosystem. For example, the correlation coefficient of NPP and NDVI was 0.53; the correlation coefficient of NPP and leaf area index was 0.78; and the correlation coefficient of NPP and wind speed was 0.78. Our data does not completely accord with the actual situation interpolated from limited meteorological stations and soil profiles and data precision should be improved in future studies.