引入地形因子的黑土区大豆干生物量遥感反演模型及验证

为了对田块尺度农作物地上干生物量进行估测,提高大豆地上干生物量反演模型的精度和稳定性,该文获取了研究区地块2016年7、8月份的SPOT-6多光谱数据,并测定不同地形坡位的大豆地上干生物量,以归一化植被指数(normalized difference vegetation index,NDVI)和增强型植被指数(enhanced vegetation index,EVI)为输入量,建立田块尺度大豆地上干生物量一元线性回归模型;加入与地上干生物量相关的地形因子,建立逐步多元回归和神经网络多层感知反演模型.结果表明:1)使用传统的单一植被指数模型预测大豆地上干生物量有可行性,但模型精度和稳定性不高.2)加入地形因子(海拔、坡度、坡向)的神经网络多层感知器模型,有较高的精度和可靠性,模型准确度达到90.4％,验证结果显示预估精度为96.2％.反演结果与地块的地形、地貌、气温和降水特征基本吻合,反映了作物长势的空间分布特征,可以为田块尺度大豆地上干生物量动态监测和精准管理,提供借科学依据.

Remote sensing inversion models and validation of aboveground biomass in soybean with introduction of terrain factors in black soil area

Abstract: In order to accurately estimate the crop aboveground biomass at the field scale, improving the accuracy and stability of soybean aboveground biomass inversion model, this paper obtained SPOT-6 6-meter multi-spectral data from July and August of the study area as well as the soybean aboveground biomass of different terrain slope. At the same time, the terrain data of the study area were measured and the topographic factors such as elevation, slope and aspect were extracted. Using the above measured data, intended to build three models, which were the traditional linear regression model, the multiple regression model and the neural network model. At first, the correlation relationships among enhances vegetation index (EVI), normalized difference vegetation index (NDVI) and observed date of soybean aboveground biomass were analyzed by linear regression model. Then added the terrain factors related to the aboveground biomass, establishing multilayer perception stepwise multiple regression and neural network inversion model. Through the model accuracy comparison and estimation accuracy analysis, The results are following 1) In the linear regression model established by the two vegetation indices, the NDVI Model fit degree is high, and the coefficient of determination (R2) reaches 0.712, root mean square error (RMSE) of 0.116. The results can be explained that the use of traditional single vegetation index model to predict soybean aboveground biomass is feasible, but the model accuracy and the stability is not high. 2) After adding the topographic factors such as elevation, slope, aspect and so on, the neural network multilayer sensor model was established. This model has high accuracy and reliability. The results show that R2 reaches 0.904 and RMSE is 0.047. The results of model validation show that the average absolute error and the average relative error of the total aboveground biomass of the whole verification area using the neural network model are the smallest, the values are 0.113 kg/m2 and 0.212, respectively. In the three types of inversion models, the inversion results of the neural network model are closest to the actual situation of crop aboveground biomass distribution. The inversion results of this study are in good agreement with the terrain, topography, temperature and precipitation characteristics of the plot. Accurately reflects the space distribution features of crop condition and the spatial distribution of crop growth. The results of this study provide a scientific basis for the dynamic monitoring and precise management of soybean aboveground biomass at the field scale.