天山北麓山地-绿洲-荒漠生态系统净初级生产力空间分布格局及其季节变化

以天山北麓总面积达93 936 km~2的山地-绿洲-荒漠生态样带为例,利用生态-遥感光能利用率模型NPP-PEM,使用1 km分辨率SPOT/VEGETATION遥感等数据资料,估算了生态样带净初级生产力(NPP)空间分布及其季节变化.结果表明山地-绿洲-荒漠生态样带平均NPP为161.06 g C·m~(-2)·a~(-1),样带陆地生态系统年总碳吸收量或年总NPP累积量为15.081 Tg C(1Tg=10~(12)g),其中绿洲农田、山地草甸草原、平原荒漠草原和山地森林对的碳吸收贡献率分别为32.67%、28.16%、12.41%和9.15%.夏季是各类生态系统NPP增加量最大的季节,而沙漠由于早春短命植被覆盖而具有生长双峰现象.样带NPP空间分布及其季相变化特征是自然环境、地貌、气候以及人类生产活动长期共同作用和影响的结果,其中水热条件和基质是控制干旱区陆地生态系统NPP空间格局的决定因子.结果检验表明模拟效果较为合理,证明NPP-PEM模型在干旱生态系统的应用是可行的.研究为干旱区陆地生态系统碳循环研究开辟了途径,可为干旱区生态系统评估、监测和管理提供研究方法和参考依据.

Spatial pattern and seasonal dynamics of net primary productivity in mountain -oasis -desert ecosystem on the north piedmont of Tianshan Mountains in arid north-west China

Modeling and monitoring of net primary productivity (NPP) is being given high priority for'understand-ing the impacts of land-use/cover change, climate change and anthropogenic perturbations on the terrestrial carbon cycle, food security, and sustainability in arid ecosystem of northwest China. The terrestrial NPP of mountain - oa-sis - desert ecosystem areas, its quantum, spatial and temporal variability as well as distribution across seasons and land cover types are not well understood. Remote sensing of NPP is a critical tool for assessing spatial and temporal patterns of carbon exchange between the atmosphere and biosphere in arid northwest China. In this study, the mod-ified production efficiency model (NPP - PEM) was applied to estimate the NPP of the mountain - oasis - desert e-cosystem areas on the north piedmont of the Tienshan Mountains, using 1 km SPOT/VEGETATION 10 -day com-posites and requiring meteorological data set covering the period from January 2002 to December 2002. Using the NPP- PEM model driven with remote sensing data, this paper aims at estimating spatial and temporal variations in terrestrial NPP of the ecosystem areas. The simulation results show that their annual carbon uptake in net primary production was 15.081 Tg C (1 Tg = 10~(12) g) in 2002, 32.67 % being attributed to oasis, 28. 16 % to alpine meadows, 12.41% to desert plain grassland and 9.15 % to forests. The annual NPP varied greatly among differ-ent kinds of land cover and their annual mean NPP was 150.29 g C · m~(-2) · a~(-1) in 2002. The alpine forest had the highest mean NPP with a value of 422.57 g C · m~(-2)·a~(-1), followed by oasis cropland with a mean value of 375.27 g C · m~(-2) · a~(-1), and the sand dune or gobi had the lowest mean NPP of 28.33 g C · m~(-2) · a~(-1). The sea-sonality of monthly NPP is characterized by strong summer photosynthesis capacities, and the desert ecosystem had not only a short early spring growing season in May but also another growing peak in early August with the maximum monthly NPP less than 5.0 g C · m~(-2) For spatial and temporal heterogeneity of water and thermal resources and combinations, the month of peak net carbon fixation varies with different land covers along elevation gradients on the north piedmont of the Tienshan Mountains. The simulated results are in good agreement with field measure-ments. The gap between stand scale measurements and regional scale modeling can be bridged by combining a car-bon exchange model with satellite observations. The results have vital implications for carbon cycle research, and e-cosystem assessment, monitoring and management in arid northwest China.