Publications

Feng, MC; Guo, XL; Wang, C; Yang, WD; Shi, CC; Ding, GW; Zhang, XR; Xiao, LJ; Zhang, MJ; Song, XY (2018). Monitoring and evaluation in freeze stress of winter wheat (Triticum aestivum L.) through canopy hyperspectrum reflectance and multiple statistical analysis. ECOLOGICAL INDICATORS, 84, 290-297.

Abstract
In order to evaluate the freeze stress of the winter wheat, field and pot experiments were conducted. The crop variables including plant height, above ground dry biomass (AGDB), chlorophyll content (CHLC), peroxidase (POD), superoxide dismutase (SOD), malondialdehyde (MDA), photosynthetic indexes, and canopy spectra were selected and assessed. Statistical methods of correlation analysis, principal component analysis (PCA), partial least squares (PLS), importance of variables analysis, and multiple linear regression (MLR) were implemented to establish freeze injury comprehensive evaluation index (FICEI), extract sensitive bands, and build the freeze injury predictive model. The results demonstrated that FICEI can be used to evaluate the damage degree. The major spectral regions were selected with the B-coefficient, variable importance in projection (VIP), and the frequencies of selection (FS) parameters derived from the PLS analysis. The combination of all three approaches can be utilized to reduce invalid spectral regions. According to the importance of variables analysis, the seven sensitive bands were extracted, which were 574 nm, 702 nm, 1093 nm, 1111 nm, 1210 nm, 1302 nm, and 1333 nm, respectively. The calibrated MLR model based on the selected bands demonstrated an excellent performance as the R-2 and RMSE were 0.924, 0.279 for FICEI. The model also illustrated an accurate and robust performance by using the pot validation data set since the higher value of coefficient of the determination and lower root mean squared error in validation model were confirmed. Our findings show that the estimation model for FICEI of winter wheat is reliable.

DOI:
10.1016/j.ecolind.2017.08.059

ISSN:
1470-160X