基于HJ-1A高光谱数据的滇池悬浮物浓度估测研究

A Study on the Suspended Sediment Concentration in Dianchi Lake Using HJ-1A Hyperspectral Data

  • 摘要: 以云南滇池悬浮物为研究对象,通过手持GPS记录采样点,实地采集水样,利用过滤烘干法测定水样的悬浮物浓度,基于HJ-1A高光谱遥感影像数据建立悬浮物含量估算模型。结果表明:滇池水体反射光谱在470 nm附近有1个反射峰,与平原地区湖泊II类水体光谱特征有所不同;在不同的悬浮物浓度下,可见光和近红外波段的反射率表现不同;通过光谱变换可以突显原始光谱的局部特征。双波段组合的变量能有效提高相关性;悬浮物浓度最佳估算模型为基于倒数变换的多元逐步回归模型,R2为0.435,RMSE为10.402 mg/L,RE为27.601%。

     

    Abstract: Taking the suspended sediment in Dianchi Lake of Yunnan Province as the research object, the sample sites were recorded by handheld GPS devices, the water samples were collected, and its suspended sediment concentration was measured by filtration and drying method. The estimation model of suspended sediment content was established by hyperspectral data based on HJ-1A HSI images. The results showed that there was a reflecting peak near 470 nm on spectral curves of Dianchi Lake, which is different from case II water in plain area. The different reflectance performance of visible and near infrared bands in different concentration of suspended sediment. The partial features of the original spectrum could be highlighted by spectral transformation. The correlation could be effectively improved by the variables of dual-band combination. The best estimation model of suspended sediment concentration was the multivariate stepwise regression model based on reciprocal transformation, and R2, RMSE and RE were respectively 0.435, 10.402 mg/L and 27.601%.

     

/

返回文章
返回