基于高分卫星遥感的天然草地牧草营养含量季节动态反演的研究

Seasonal Dynamic Inversion of Natural Grassland Forage Content Based on High-resolution Satellite Remote Sensing

本试验选取了青海省海北地区的海晏县作为研究区域,按照返青期(4-6月)、盛草期(7-9月)、枯黄期(10-12月)和枯草期(1-3月)4个阶段,试图通过实地测量、实验室分析与高分遥感影像相结合,分别筛选与草地各营养指标含量相关性最优的植被指数,建立草地营养含量月际动态估测模型,旨在将高分卫星影像与地面高光谱数据相结合,初步探寻天然草地各营养指标含量的月际动态规律,对指导冷季补饲及实现草地资源最优配置具有重要意义。结果表明:结合高分遥感数据,针对不同月份分别筛选最佳反演模型,对各营养指标含量进行反演,以此分析草场各营养含量月际动态变化规律是可行的;牧草干物质、粗蛋白、钙和磷含量随着返青期-盛草期-枯黄期-枯草期的变化趋势,基本呈现先增后减型变化。4个阶段中,草场干物质和磷含量最大相差约6倍,粗蛋白含量最大相差约7倍,钙含量最大相差约17倍。

Haiyan in Haibei area of Qinghai Province is selected as the research area.According to the four stages of the regreening period(April-June),the grass-bearing period(July-September),the yellowing period(October-December),and the dry grass period(January-March),it attempted to combine the field measurement,laboratory analysis and high-resolution remote sensing images.Based on the remote sensing vegetation index with the best correlation between nutrient output and the dynamic estimation model of forage grass nutrient content.It is intended to combine high-resolution satellite imagery with terrestrial hyperspectral data to initially explore the monthly dynamics of the nutrient content of natural grassland pastures,and play a guiding role in formulating a reasonable supplementary feeding plan and realizing the optimal allocation of grassland resources.The results showed that it was feasible to screen the best inversion models for different months and to invert the content of each nutrient index in combination with high-scoring remote sensing data,so as to analyze the monthly dynamics of nutrient content in grassland.The changing trend of the content of dry matter(DM),crude protein(CP),calcium(Ca)and phosphorus(P)in pastures at the period-to-grass stage-leaf stage-wild stage increased first and then decreased,the difference between dry matter and phosphorus content was about 6 times,the maximum difference of crude protein content was about 7 times,and the maximum difference of calcium content was about 17 times.