青海省植被净初级生产力的模拟研究

THE STUDY ON SIMULATING NET PRIMARY PRODUCTIVITY OF VEGETATION IN QINGHAI PROVINCE

遥感技术根据植被光谱特征可以有效地估算植被的净初级生产力(NPP)。利用遥感数据估算植被NPP,是近10年来NPP的模型研究和估算方法最为突出的特点。文章借鉴了CASA,GLO-PEM,VPM等光能利用率NPP模型的优点,同时充分考虑了研究区域其植被光能利用率和环境因素的典型特点,建立了针对研究区域基于光能利用率原理的植被净初级生产力遥感估算模型。在此基础上,模拟了青海省2006年逐月植被NPP,较深入地分析了NPP的空间分布和季相变化特征。研究结果表明,2006年青海省植被净初级生产力介于0～422gC/m^2/a之间,平均值为151gC/m^2/a;青海省NPP空间分布规律总体上呈由西北向东南地区递增的趋势;NPP在不同植被类型中的差异相当明显,单位面积平均NPP最高的为阔叶林(314gC/m^2/a),最低的为稀疏灌丛(101gC/m^2/a);青海省NPP随季节的推移变化显著,NPP的积累期主要发生在水热搭配较好的4～9月份,7月植被平均NPP达到最大值(43gC/m^2)。

Remote sensing is widely used to effectively simulate net primary productivity （NPP） of vegetation based on spectral characteristics. In recent 10 ycars the method of simulating vegetation NPP by remote sensing is most representative. The study is conducted in Qinghai Province, which is located in Western China. Forest is sparse and grassland is dominant in the study area. Natural grassland covers 50.46% of the total area of the whole Province. The dominant grassland in Qinghai Province belongs to alpine meadow. The paper aims at simulating NPP of vegetation in Qinghai Province and analyzing spatio-temporal trends of simulated NPP for each month in 2006 in Qinghai Province. During making the NPP model, advantages of CASA, GLO-PEM, VPM models based on light use efficiency（LUE） are used for reference, and typical characters of vegetation light use efficiency and environment in the study area are taking into account. Afterwards,the NPP model for vegetation based on LUE and remote sensing is built. Photosynthetic active radiation （PAR） , fractional photosynthetic active radiation （ FPAR）, water stress factor （WS）, temperature stress factor （T） and maximum light use efficiency are five basic parameters in the NPP model. Maximum LUE values of main vegetation in the study area are simulated according to relative literature and field measured NPP. The linear equation between FPAR and simple ratio（SR） is made to calculate FPAR values. Because Evaporation fraction（EF） can indicate water conditions of ecosystem and be derived from vegetation index of remote sensing and land surface temperature products, it is used as water stress factor. The formula from the Terrestrial Ecosystem Model is used to calculate temperature stress factor. The result shows that simulated mean value of NPP for vegetation ranges from 0 to 422gC/m^2/a in 2006 in Qinghai Province, and the average is 151 gC/m^2/a. NPP gradually increases from north-west to south-east, and NPP values in south-east, south and regions around the Qinghai Lake are larger than NPP in other areas in Qinghai. Different kind of vegetation takes on different value of NPP. Among plants the mean NPP of broad-leaved forest is the largest （314gC/m^2/a）, and the mean NPP of sparse shrub is the smallest（ 101 gC/m^2/a）. The annual total NPP of the whole grassland composed of sparse grassland,middle density grassland and dense grassland is 36gC/a,and it accounts for 59.9% of the total NPP of Qinghai Province. The value of NPP in Qinghai Province varies largely with the seasons. The accumulation of NPP occurs from April to September,which accounts for 79. 8% of the annual total NPP. The growth of vegetation reaches the peak,and at this moment the mean NPP of vegetation is the largest（43gC/m^2 ）. Two kind of shrub composed of dense shrub and sparse shrub take on similar variable trend. Simulated NPP values from April to September are larger than NPP in other months. NPP values of sparse shrub are larger than 8gC/m^2, and NPP of dense shrub are larger than 21 gC/m^2. From April to September,the accumulation of NPP for sparse shrub accounts for 76. 3% of its annual total NPP and NPP accumulation for dense shrub accounts for 72.3%.