巴音布鲁克草原高寒草原植被覆盖度反演模型构建与评价

摘要

构建巴音布鲁克高寒草原植被覆盖度反演模型,筛选最佳模型并检验精度,为今后巴音布鲁克高寒草原开展草地监测和科学管理提供科学依据.利用经验回归模型法和像元二分法,通过计算分析植被覆盖度及植被指数,构建反演模型.结果表明,除修正土壤调整植被指数(Modified Soil Adj usted Vegetation Index MSAVI)外,其余植被指数与植被覆盖度均呈显著正相关关系(P<0.01).巴音布鲁克高寒草原植被覆盖度最佳反演模型为:NDVI 二次多项式回归模型y=—0.894x2+1.467x+0.099(R2=0.723),反演精度较高(R2=0.837),可用于估算巴音布鲁克高寒草原的植被覆盖度.%We constructed an inversion model of the vegetation coverage of the alpine prairie in Bayanbulak, screening for the best model and checking its accuracy,which will provide a scientific basis for carrying out grassland monitoring and scientific management in the alpine prairie of Bayanbulak in future.An empirical re-gression model and pixel dichotomy were used to calculate the vegetation coverage and vegetation indices and to construct the inversion model.The results showed that there were significant positive correlations between the vegetation indices,except for the Modified Soil Adj usted Vegetation Index(MSAVI),and vegetation coverage (P<0.01).The best inversion model of alpine steppe vegetation coverage in Bayanbulak is the quadratic poly-nomial regression model of NDVI:y=—0.894x2+1.467x+0.099(R2=0.723);its inversion accuracy is higher (R2=0.837),and can be used to estimate the vegetation coverage of Bayanbulak alpine steppes.