荒漠区人工林NPP估算模型比较研究

Comparative Study on Net Primary Productivity Estimation Model of Artificial Forest in Desert Area

利用LANDSAT8遥感影像数据、结合气象数据和野外调查数据,采用CASA模型、C-FIX模型和GLO-PEM模型,分别估算和分析了2014年墨玉县森林植被净初生产力时空分布特征,并进行验证和比较。结果表明:植被净初生产力空间分布差异比较明显,主要集中于研究区中部和东南部;CASA模型的NPP平均值为70gC·m^(-2)、C-FIX模型的为56C·m^(-2),GLO-PEM模型的为28C·m^(-2),其中CASA模型模拟结果最接近于前人研究成果;通过实测值验证,各模型实测值和模拟值相当吻合,其中CASA模型的平均值最接近,相关系数R2=0.994 15,大部分样本点的误差小于5%。

By Using LANDSAT8 remote sensing image data,combined with the meteorological data&field survey data,using CASA model,C-FIX model&using GLO-PEM model,forest vegetation net primary productivity spatial&temporal distribution characteristics in Moyu County of 2014 were estimated,and were validated&compared.Result shows that three NPP model calculation result shows that vegetation net primary productivity spatial distribution differences,mainly concentrated in the central&southeast of the study area.The average NPP of each model is 70 g C·m-2,C-FIX model respectively.The model is 56 C·m-2,a GLO-PEM model 28 c·m-2 which CASA model simulation result is most close to the former research results;measured value of verification,the model test value and the simulated values are in optimal agreement,which CASA model average closest,correlation coefficient R 2=0.994 15.Most of the samples of the error is less than 5%.