Context

Estimating the fraction of photosynthetically active radiation (PAR) interception (FIPAR) precisely is critical for monitoring crop eco-physiological status. However, few studies have concentrated on estimating FIPAR using hyperspectral technology, especially for quantifying the FIPAR of different vertical layers within the wheat canopy.

Objective

The objective of this study was to explore an effective method for determining the FIPAR of the wheat canopy and different vertical layers based on hyperspectral data.

Methods

A two-year field experiment including various plant densities and nitrogen rates was conducted to investigate the vertical light distribution (VLD) information characteristics in the wheat canopy, the relationship between the FIPAR of the entire canopy (FIPAR_canopy) and that of vertical layers (FIPAR_Li, L1-L3, vertical layer 1–3 from the top to the bottom of the canopy), and the relationship between different vegetation indices (VIs) and FIPAR_{Li, canopy}.

Results

Results indicated that PAR decreased exponentially with the decrease in relative height (R² > 0.90), and FIPAR_L1 was more susceptible to nitrogen rates than FIPAR_{L2, L3}. Stable exponential relationships existed between FIPAR_{L1, L2, u2 (u2=L1+L2)} and FIPAR_canopy (R² = 0.75–0.95), while a significant quadratic relationship was found between FIPAR_L3 and FIPAR_canopy (R² = 0.35). Vegetation indices (VIs) could directly estimate the FIPAR_canopy using a quadratic model. FIPAR_canopy estimation model performed well based on the red edge chlorophyll index, green normalized difference vegetation index, renormalized vegetation index-2, and soil adjusted vegetation index (R² = 0.72–0.75 and RRMSE = 15.50–16.07%). Additionally, the FIPAR_L3 estimation model gained a higher prediction accuracy based on the integration of VIs and VLD with reduced RRMSE of 0.93–2.53% than the usage of individual VIs.

Conclusion

Combining VLD and hyperspectral data could improve the estimation accuracy for FIPAR of the bottom layer. While the individual use of VIs is the priority choice for estimating the FIPAR of top-middle layers.

Significance

These results support the rapid, accurate, and non-destructive prediction of FIPAR of different vertical layers in the wheat canopy.

中文翻译：

---

大田小麦不同垂直层FIPAR的遥测

语境

精确估算光合有效辐射 (PAR) 拦截 (FIPAR) 的比例对于监测作物生态生理状况至关重要。然而，很少有研究集中于使用高光谱技术估计 FIPAR，特别是用于量化小麦冠层内不同垂直层的 FIPAR。

客观的

本研究的目的是探索一种基于高光谱数据确定小麦冠层和不同垂直层FIPAR的有效方法。

方法

通过为期两年的不同种植密度和施氮量田间试验，研究了小麦冠层的垂直光分布（VLD）信息特征、整个冠层FIPAR（ FIPAR canopy ）与垂直层（ FIPAR ）之间的_关系（FIPAR _Li , L1-L3, 冠层从上到下的垂直层1-3），以及不同植被指数（VIs）与FIPAR _{Li, canopy 的}关系。

结果

结果表明，PAR随着相对高度的降低呈指数下降（R ² > 0.90），FIPAR _L1比FIPAR _L2、L3对施氮量更敏感。FIPAR _{L1、L2、u2 (u2=L1+L2)}与FIPAR_冠层(R ² = 0.75–0.95)之间存在稳定的指数关系，而FIPAR _L3与FIPAR_冠层(R ² = 0.35)之间存在显着的二次关系。植被指数 (VI) 可以使用二次模型直接估计FIPAR_冠层。菲帕基于红边叶绿素指数、绿色归一化差异植被指数、重归一化植被指数-2 和土壤调整植被指数（R 2 ⁼ 0.72–0.75 和 RRMSE = 15.50–16.07%）的_{冠层估计模型表现良好。}此外， FIPAR _L3估计模型基于 VI 和 VLD 的集成获得了更高的预测精度，与使用单个 VI 相比，RRMSE 降低了 0.93–2.53%。

结论

结合VLD和高光谱数据可以提高底层FIPAR的估计精度。而单独使用 VI 是估计中顶层 FIPAR 的优先选择。

意义

这些结果支持对小麦冠层不同垂直层的FIPAR进行快速、准确、无损的预测。