融雪期积雪参数对实测光谱反射率的影响

Influence of Snow Features on the Measured Spectral Reflectance during Snowmelt Period

每种地物都有其特殊的光谱特征曲线,积雪作为水体的一种特殊存在形式,其光谱也存在特殊性。在我国西北地区的积雪测试过程中,采集融雪期的积雪光谱数据以及同时期的积雪参数实测数据,对这些数据进行综合分析,并利用积雪参数数据和神经网络模型对光谱反射率进行模拟分析。结果表明:融雪期积雪参数对光谱影响的敏感波段在可见光到短波红外光都有分布;光谱对积雪各参数几乎都有所对应的敏感波段;可见光、近红外光和短波红外波段的反射率可以由不同的积雪参数组合,特别是通过积雪含水率、雪密度和总孔隙率组合来进行模拟反演,而且模拟精度高,在可见光和短波红外波段可达95%以上,拟合残差均小于0.001。

Snow has significant influence on climate, natural environment and human activities. The research on snow melt process not only consists of snow,but also the influence on runoff,soil water moisture change. Especially the seasonal snow plays an irreplaceable role in changing climate and environment in the arid region of Northwest China. Every ground object has its special feature curve. As a special existence form of water,snow also has its own specialization on spectral feature. During the snow experiment in Northwest China,we collected the snow spectroscopic data and the measured snow features data in snowmelt period.Model analysis of spectral reflectance can be carried out through the above data comprehensive analysis,as well as the application of the snow feature data and neural network model. The result showed that the sensitive spectral band distributed in visible light near-infrared as well as short wave infrared. Every snow feature has the certain sensitive band; the combination of different snow features data can simulate the sensitive band spectral reflectance very well,the accuracy of simulation is much higher when the combination consists of snow liquid water content,snow density and total porosity especially,and in the visible light and short wave infrared,the simulation accuracy can reach 95%,meanwhile the Fitting Residue is less than 0. 001.