Abstract A major international effort has been made to monitor sun-induced chlorophyll fluorescence (SIF) from space as a proxy for the photosynthetic activity of terrestrial vegetation. However, the effect of spatial heterogeneity on the SIF retrievals from canopy radiance derived from images with medium and low spatial resolution remains uncharacterised. In images from forest and agricultural landscapes, the background comprises a mixture of soil and understory and can generate confounding effects that limit the interpretation of the SIF at the canopy level. This paper aims to improve the understanding of SIF from coarse spatial resolutions in heterogeneous canopies by considering the separated contribution of tree crowns, understory and background components, using a modified version of the FluorFLIGHT radiative transfer model (RTM). The new model is compared with others through the RAMI model intercomparison framework and is validated with airborne data. The airborne campaign includes high-resolution data collected over a tree-grass ecosystem with the HyPlant imaging spectrometer within the FLuorescence EXplorer (FLEX) preparatory missions. Field data measurements were collected from plots with a varying fraction of tree and understory vegetation cover. The relationship between airborne SIF calculated from pure tree crowns and aggregated pixels shows the effect of the understory at different resolutions. For a pixel size smaller than the mean crown size, the impact of the background was low (R2 > 0.99; NRMSE 0.2). This study demonstrates that using a 3D RTM model improves the calculation of SIF significantly (R2 = 0.83, RMSE = 0.03 mW m−2 sr−1 nm−1) when the specific contribution of the soil and understory layers are accounted for, in comparison with the SIF calculated from mixed pixels that considers only one layer as background (R2 = 0.4, RMSE = 0.28 mW m−2 sr−1 nm−1). These results demonstrate the need to account for the contribution of SIF emitted by the understory in the quantification of SIF within tree crowns and within the canopy from aggregated pixels in heterogeneous forest canopies.

中文翻译：

---

通过 3-D 辐射转移模型和现场数据评估林下阳光诱导的叶绿素荧光的贡献

摘要 国际上已经做出了一项重大努力，从空间监测太阳诱导的叶绿素荧光 (SIF)，作为陆地植被光合作用活动的代表。然而，空间异质性对来自中低空间分辨率图像的冠层辐射的 SIF 检索的影响仍未得到表征。在来自森林和农业景观的图像中，背景包括土壤和林下植被的混合物，并且会产生混淆效应，从而限制了在冠层水平上对 SIF 的解释。本文旨在通过考虑树冠、林下和背景成分的分离贡献，使用 FluorFLIGHT 辐射传输模型 (RTM) 的修改版本，从异构冠层的粗略空间分辨率中提高对 SIF 的理解。新模型通过RAMI模型比对框架与其他模型进行比较，并通过机载数据进行验证。机载活动包括使用荧光探索者 (FLEX) 准备任务中的 HyPlant 成像光谱仪在树草生态系统上收集的高分辨率数据。现场数据测量是从具有不同比例的树木和林下植被覆盖的地块中收集的。从纯树冠计算出的空中 SIF 与聚合像素之间的关系显示了林下在不同分辨率下的影响。对于小于平均冠尺寸的像素尺寸，背景的影响较低（R2 > 0.99；NRMSE 0.2）。本研究表明，使用 3D RTM 模型显着改善了 SIF 的计算（R2 = 0.83，RMSE = 0。03 mW m−2 sr−1 nm−1)，当考虑土壤和林下层的特定贡献时，与从仅考虑一层作为背景的混合像素计算的 SIF 相比（R2 = 0.4，RMSE = 0.28 mW m-2 sr-1 nm-1)。这些结果表明需要考虑林下排放的 SIF 在量化树冠内和来自异质森林冠层聚合像素的冠层内的 SIF 的贡献。