Estimating leaf area index in spring wheat using alternative remote sensing approaches.

摘要

The reflectance based vegetation indices commonly used to infer leaf area index (LAI) are affected by external factors such as soil background, and view and illumination angles. Simulation of a modified SUITS directional reflectance model, MODSUITS, showed that there are some combinations of near infrared and red reflectance that correlate with LAI while minimizing the effects of these factors. We proposed the MSAVI3 (Modified Soil Adjusted Vegetation Index) that is suited in minimizing soil brightness effect particularly for soils with soil line slope greater than unity. This index decreases the residual soil effect caused by the multiple scattering between the soil background and the plant at intermediate LAI values. The MODSUITS model predicts that soil brightness affects less canopy transmittance than reflectance. Hence in determining LAI, transmittance based indices should be less dependent on soil brightness.; We assumed that the transmission coefficient is linearly related to the logarithm of LAI. Therefore, the extinction coefficient for diffuse visible radiation varied between 0.70 and 0.75 for the 42 genotypes. Similarly, from the ratio of near infrared to red transmittance, the difference in spectral extinction coefficient under total incidence oscillated between 0.40 and 0.52. The consistency in the extinction coefficients among the 42 genotypes and across geographical locations suggest that LAI of spring wheat can only be inferred from the diffuse transmission coefficient. Moreover, we found that the normalized transmittance vegetation index (NTVI), which is conceptually similar to the common normalized difference vegetation index (NDVI), is linearly related to the LAI up to a value of 6.; Leaf area index was also assessed from the analysis of digitized infrared color films. While the difference vegetation index (DVI) is related to growth or LAI, the low spectral sensitivity of the CIR film, particularly in the red band, limits its ability to resolve differences among the 42 spring wheat genotypes. The DVI integrated from emergence up to the onset of senescence showed slightly positive and moderately negative relationship with final grain yield at Pullman and Mayview, respectively.