GF-1影像遥感监测指标与冬小麦长势参数的关系

Relationship between remote sensing monitoring indices and growth parameters in winter wheat based on GF-1 images

为了分析高分一号卫星(GF-1)影像在冬小麦长势监测中的有效性和适宜性,以建湖县冬小麦为研究对象,选取12个植被指数作为遥感监测指标,运用回归分析法探讨遥感监测指标与地面实测冬小麦长势参数的关系,并以回归模型的决定系数(R 2)作为反演精度的评价指标。研究发现,叶面积指数(LAI)、密度和生物量的反演精度较高,其中LAI的反演精度在拔节期最高[监测指标:红蓝色归一化植被指数(RBNDVI),R 2:0.6894],密度的反演精度在拔节期最高[监测指标:优化的土壤调节植被指数(OSAVI),R 2:0.5438],生物量的反演精度在孕穗期最高[监测指标:归一化植被指数(NDVI),R 2:0.4486],说明GF-1影像适合在拔节期进行冬小麦LAI、密度的监测,在孕穗期进行生物量监测。土壤含水量、株高和叶绿素含量(SPAD值)的反演精度较差,最佳回归模型的R 2皆低于0.3600,说明所选的12个遥感监测指标不适合反演这3个长势参数。除乳熟期外,其他4个生育期中都是LAI的反演精度最高,可见GF-1影像的遥感监测指标与LAI的相关性最好,反演精度最高。本研究结果说明,在进行冬小麦长势监测时,不同的生育期需要采用不同的监测指标,同时GF-1影像则更适合在拔节期和孕穗期进行冬小麦的长势监测。本研究结果在一定程度上为GF-1影像在农情遥感监测中的应用提供了科学依据。

In order to analyze the effectiveness and validity of GF-1 images in winter wheat growth monitoring,twelve vegetation indices were selected as remote sensing monitoring indices,and the differences between monitoring indices and growth parameters,leaf area index(LAI),aboveground biomass,leaf chlorophyll content(SPAD value),density,plant height and soil water capacity(0-10 cm)during five critical growing stages were analyzed with regression analysis.It took the determination coefficient(R 2)of regression model as retrieval accuracy assessment indicator.The results showed that the highest R 2 between monitoring indices and LAI was 0.6894 at jointing stage using red blue normalized difference vegetation index(RBNDVI).The highest R 2 between monitoring indices and density was 0.5438 at jointing stage using optimal soil adjusted vegetation index(OSAVI).The highest R 2 between monitoring indices and biomass was 0.4486 at booting stage using normalized difference vegetation index(NDVI).It was concluded that GF-1 image was more suitable for monitoring the growth of winter wheat at jointing and booting stages.The inversion accuracy of soil water capacity,plant height and SPAD was poor,and R 2 of the best regression model was lower than 0.3600.These results indicated that the 12 remote sensing monitoring indices were not suitable for inversion of the three growth parameters.The LAI had the highest inversion accuracy at other four stages except milky stage.So,different monitoring indicators should be used to monitor the growth of winter wheat in different growth periods.GF-1 image is more suitable for growth monitoring of winter wheat at jointing and booting stages.These results from this study provide scientific basis for the application of GF-1 image in agricultural monitoring.