基于NDVI的黄土丘陵沟壑区草地生物量模型研究——以宁夏彭阳县为例

Research of grassland biomass model based on NDVI in loess hilly and gully region——A case of Pengyang County,Ningxia

基于多年TM遥感影像数据NDVI(归一化差异植被指数),与同期彭阳县草地实测地上生物量做相关性分析,建立NDVI与地上生物量的线性和3种非线性(对数、二次多项式、三次多项式)回归模型,并对这些模型的模拟结果进行了比较。结果表明:(1)在黄土丘陵沟壑区典型草地中,地上生物量与NDVI呈现显著相关;利用NDVI监测草地植被生物量的复相关系数均大于0.6,充分说明利用植被指数检测典型草原生物量是一种简单可行的方法。(2)以NDVI建立的生物量回归模型,其模拟地表生物量的效果较好,地上生物量线性与3种非线性(三次多项式、二次多项式、对数)回归模型都表现出较好的模拟效果,但三次多项式生物量回归模型为最优。(3)用NDVI-生物量三次多项式回归模型模拟彭阳县草地的生物量,显示其呈逐年增加趋势,但2010—2015年增速减缓。

Based on the NDVI （normalized difference vegetation index） data of TM remote sensing image data for many years, this paper analyzed the aboveground biomass of groundwater in Pengyang County, and established linear and three nonlinear （logarithmic, quadratic polynomial, cubic polynomial） regression model. The results showed that aboveground biomass had a significant correlation with NDVI in typical grassland in the loess hilly and gully region. The complex correlation coefficients of vegetation biomass were all greater than 0.6. The biomass regression model were also established by NDVI and it is very suitable for simulating the effect of surface biomass in regression linear model or the three non-linear models including third-order polynomial biomass model, quadratic polynomial model and logarithmic model. Especially, the third-order polynomial regression model is best and it was used to simulate the biomass of grassland in Pengyang County, which was increasing year by year, but slowed down in 2010-2015.