中国东北地区植被NPP模拟与时空变化分析

Simulation and spatio-temporal variation analysis of net primary productivity in Northeast China

利用北方生态过程模型(BEPS)模拟我国东北地区2003-2012年间植被生产力,以森林、草地通量站点和区域森林样地生产力调查数据两个尺度验证了模型精度,并对其植被净初级生产力(NPP)的时空变化特征及影响因素进行了分析.结果表明:在站点上,森林的总初级生产力(GPP)与站点观测值的决定系数(R^2)为0.94,均方根误差(RMSE)为1.53 g·(m^2·d)^(-1),草地植被的R^2与RMSE分别为0.79和0.59 g·(m^2·d)^(-1);区域森林生产力调查数据处理结果与模拟的森林NPP的决定系数R^2=0.84,RMSE=42.73 g·(m^2·a)^(-1);研究区域NPP值呈现东北部高西南部低,由东北向西南方向逐级递减的分布格局;10年来,林地NPP增长最为明显,农田增长缓慢,灌木增减不明显,而草地呈弱下降趋势.温度和太阳辐射是影响东北NPP变化的主要因素,森林干扰是引发局部剧烈变化的重要驱动因素.

The boreal ecosystem productivity simulator(BEPS)model was applied to simulate the vegetation productivity over the northeast China from 2003 to 2012.The flux data at the forest and grass sites and the dendrochronological measurements collected from the local forest plots were employed to validate the simulated results.Afterwards,the spatio-temporal variations characteristics of the net primary productivity(NPP)and influencing factors were analyzed.The results showed that:the determination coefficient(R2)of the forest gross primary productivity(GPP)values between the flux measurements and the simulates was 0.94,the root mean square error(RMSE)was 1.53 g·(m2·d)1,and the R2 and RMSE of grassland vegetation were 0.79 and 0.59 g·(m2·d)1,respectively.The R2 of the simulated forest NPPs against the dendrochronological measurements was 0.84,and the RMSE was 42.73 g·(m2·a)1.The local NPPs decreased from high values in the northeast part to the low values in the southwest part.During these 10 years,the NPPs of forest showed the most notable increase among all vegetation types,followed by the crop with a slight increase.There was no obvious increasing or decreasing trends for the shrub,and the grass had a slight decreasing trend.Temperature and solar radiation over the growing seasons were the main factors that affect the NPPs over the northeast China,meanwhile,disturbance was the main cause for the regional dramatic NPP changes.