摘要： Crop leaf chlorophyll content (LCC) is a valuable indicator for agronomists to make fertilization recommendation and can be estimated from canopy reflectance spectra. However, the estimation accuracy of LCC is often influenced by soil background. To alleviate the adverse effect of soil background, this study proposed to collect spectral measurements at non-noon (such as 14:00-16:00 local time) hours and evaluated the performance of these spectral measurements with experimental data and radiative transfer model. The results from the wheat experiment conducted at Rugao demonstrated that the canopy spectra measured at non-noon were less sensitive to the soil background compared with those collected at midday (such as 12:00), which improved the estimation accuracy (R2) for LCC from 0.71 to 0.77. A canopy radiative transfer model called 4SAIL-RowCrop was also used to validate the performance and feasibility of the non-noon measurement scheme. One thousand spectra with different combinations of LCC, soil reflectance, and canopy structure were simulated at three observation times (12:00, 14:00 and 16:00). The CIred-edge calculated from the canopy reflectance spectra simulated for 16:00 exhibited a higher correlation to LCC (R2 = 0.76) than that for 12:00 (R2 = 0.43). These consistent findings from experimental and modeled datasets suggested that the effect of soil background can be alleviated and the estimation accuracy of LCC can be improved by determining a proper timing of spectral observation.