模拟多光谱卫星传感器数据的冬小麦白粉病遥感监测

为了解利用遥感技术快速大范围监测小麦白粉病病害情况的可行性,以Landsat5TM波段响应函数为基础,将地面实测冠层高光谱数据模拟为TM多光谱数据,从而分析卫星传感器多光谱波段对病害的响应情况,并构建多光谱指数(PMSI)估测白粉病严重度。在此基础上,采用2010年星-地配套数据对PMSI估测精度进行验证。结果表明,PMSI能够较准确地反映冬小麦白粉病发生的程度,获得较理想的病情严重度反演精度(r2=0.475,RMSE=0.129)。因此采用多光谱卫星遥感影像在小麦大面积种植区域进行病害监测具有应用潜力。

Remote Sensing Monitoring Powdery Mildew of Winter Wheat Using Simulated Data of Multi spectral Satellite Sensor

In recent years, the rapid development of remote sensing technology provides novel means to monitor crop diseases at regional scale. Taking powdery mildew in winter wheat as an example, the present study assessed the feasibility of using multi spectral satellite images for disease monitoring. Firstly, a simulation of multi spectral reflectance was processed based on canopy hyperspectral measurements and the spectral response function of Landsat5 TM sensor. Then, the sensitivity of TM bands to powdery mildew was examined according to a correlation analysis. Based on those disease sensitive bands, a multi spectral vegetation index (PMSI) was conducted for estimating severity of powdery mildew. To examine the efficiency of the proposed index, a satellite image from 2010 and corresponding field investigation data was used for validation. The results showed that the index was able to estimate the disease severity at a reasonable accuracy, with r=0.475, and RMSE=0.129. The result suggested that the multi spectral satellite images had a potential to monitor powdery mildew in winter wheat at regional scale.