Correlations between spectral and biophysical data obtained in canola canopy cultivated in the subtropical region of Brazil

Abstract: The objective of this work was to identify the spectral bands, vegetation indices, and periods of the canola crop season in which the correlation between spectral data and biophysical indicators (total shoot dry matter and grain yield) is most significant. The experiment was carried out during the 2013 and 2014 crop seasons at Embrapa Trigo, in the state of Rio Grande do Sul, Brazil. A randomized complete block design was used, with four replicates, and the treatments consisted of five doses of nitrogen topdressing. Plant dry matter, grain yield, and phenology were measured. The canola spectral response was evaluated by measuring the canola canopy reflectance using a spectroradiometer, and, with this data, the SR, NDVI, EVI, SAVI, and GNDVI vegetation indices were determined. Pearson’s correlations between the spectral and biophysical variables of canola showed that the red (620 to 670 nm) and near-infrared (841 to 876 nm) bands were the best to estimate the dry matter. The vegetative period is the most indicated to obtain the most significant correlations for canola. All the used vegetation indices are adequate for estimating the dry matter and grain yield of canola.

摘要：本研究旨在明确油菜生育季内，光谱数据与生物物理指标（地上部总干物质及籽粒产量）相关性最显著的光谱波段、植被指数及生育时期。试验于2013和2014年油菜生育季在巴西南里奥格兰德州的巴西农牧业研究院小麦研究所（Embrapa Trigo）开展。试验采用完全随机区组设计，设置4次重复，处理组为5个氮肥追施剂量。研究测定了植株干物质、籽粒产量及物候期。通过光谱辐射仪（spectroradiometer）测定油菜冠层反射率，并以此计算得到比值植被指数（SR）、归一化差异植被指数（NDVI）、增强型植被指数（EVI）、土壤调节植被指数（SAVI）及绿色归一化差异植被指数（GNDVI）。对油菜光谱变量与生物物理变量开展皮尔逊相关分析结果显示：红光波段（620~670 nm）与近红外波段（841~876 nm）为估算干物质的最优波段；营养生长期是获取油菜相关性最显著结果的最适宜生育时期。本研究所用全部植被指数均可有效估算油菜地上部干物质及籽粒产量。