Abstract:Accurately and efficiently estimating corn LAI data within a region is of crucial importance for field management decisions, predicting land yield, and implementing precision agriculture. In response to the problems of scale effect, low accuracy, and poor universality in multi-scale and large-scale remote sensing inversion, taking the silage corn experimental field in Minle County, Zhangye City as the research area, silage corn was selected as the research object, based on Landsat-8 hyperspectral and Modis multispectral remote sensing images, combined with ground measured data. Through local and global sensitivity analysis of the input parameters of the PROSAIL model，the lookup table of canopy reflectance-LAI of silage corn in multiple growth periods and the inversion strategy of the minimum optimization cost function were constructed, and the optimal LAI inversion model for the study area was determined. The accuracy verification and linear fitting of the inversion results were completed by using the measured values in different growth periods of silage corn. The results showed that the inversion results of LAI were generally good, with high fitting accuracy and strong correlation with the measured values. The optimal determination coefficients R2 for the jointing stage, tasseling stage, and maturity stage were 0.85, 0.91, and 0.90, respectively. The root mean square error (RMSE) were 0.35, 0.58, and 0.51, respectively. Therefore, the inversion strategy based on multi-source hyperspectral remote sensing data combined with the PROSAIL model can provide scientific basis and methods for crop parameter inversion.
