目前,利用夫琅和费暗线提取叶绿素荧光的3种最常用的算法有标准FLD方法、3FLD和iFLD方法.上述3种夫琅和费暗线算法在叶绿素荧光反演中得到了广泛应用,但各算法的不确定性研究相对薄弱,尚缺乏系统的分析.因此,本文的目标是阐明氧气吸收波段叶绿素荧光反演的不确定性,优化叶绿素荧光遥感探测指标,提高叶绿素荧光反演精度.利用FluorMOD模型,模拟不同植被冠层参数、光谱分辨率SR、信噪比SNR条件下的冠层光谱,并比较分析这3种反演方法在不同参数条件下的不确定性.结果表明:3种方法在O₂-A波段的反演精度均比在O₂-B波段精度高,其中,iFLD和3FLD算法的反演精度相对较高,标准FLD的反演结果较差;随着SR的下降,3种算法的反演精度均有不同程度的下降,标准FLD算法的荧光提取精度受传感器SR的影响最大;随着SNR的增大,3种算法的反演精度有不同程度的升高,iFLD算法的荧光提取精度受信噪比的影响最大.由结果可以得出,3种反演算法在不同的参数条件下有其各自的局限性和优势;利用氧气吸收波段进行叶绿素荧光反演存在诸多不确定性,不确定性主要来源于吸收线内外反射率和荧光比值的真实值与估计值的偏差,叶绿素含量是影响这种偏差的一个主导因素;传感器性能对荧光提取结果也有显著的影响.

Sun-induced chlorophyll Fluorescence signal (Fs) is related to photosynthesis and can serve as a direct indicator to monitor plant photosynthesis status. Fs is retrieved using the three most common Fraunhofer Line Depth (FLD) retrieval methods, namely, original FLD method (sFLD), modified FLD (3FLD), and improved FLD (iFLD). These methods exploit spectrally narrow atmospheric oxygen absorption bands and relate Fs to the difference in absorption feature depth between fluorescensing and non-fluorescensing surfaces. However, owing to the nature of these narrow bands, Fs retrieval results depend not only on vegetation species type or environmental conditions, but also on instrument technology and processing algorithms. Thus, many uncertainties remain in different Fs retrieval algorithms that use the two oxygen absorption Fraunhofer lines at 688 nm and 760 nm. This research clarified the uncertainties in different Fs retrieval algorithms that use the two oxygen absorption Fraunhofer lines to optimize the remote sensing detection index of chlorophyll fluorescence and improve the inversion accuracy of chlorophyll fluorescence.
This study employed the FluorMOD model to simulate canopy spectra under different chlorophyll contents, Spectral Resolutions (SRs), and Signal-to-Noise Ratios (SNRs). sFLD, 3FLD, and iFLD algorithms were also used to retrieve chlorophyll fluorescence. The Fs retrieval accuracies of these three popular algorithms were investigated under different chlorophyll contents, SRs, and SNRs using the simulated spectral data by FluorMOD model.
Results are as below. (1) All the three algorithms have higher precision in the O₂-A band than in the O₂-B band. (2) In general, the sFLDs method strongly overestimates Fs, whereas 3FLD and iFLD provide an accurate estimation of Fs. (3) In the O₂-B band, iFLD method performs best when chlorophyll content is 10—40 μg/cm², 3FLD method performs best when chlorophyll content is 40—70 μg/cm², and the sFLDs method performs Verhoef best when chlorophyll content is 70—80 μg/cm². In the O₂-A band, 3FLD method always performs best in any value of chlorophyll content. (4) SR and SNR specifications would introduce a noticeable error for retrieved Fs. SR is the dominant factor for sFLD method, whereas SNR is the dominant factor for iFLD method.
In conclusion, the three algorithms have their own limitations and advantages under different parameters. Fs retrieval error results from the estimation error of the ratios of reflectance and Fs inside and outside of Fraunhofer lines, in which chlorophyll content is the most important key variable affecting the three Fs retrieval methods. Sensor performance also has a significant effect on fluorescence extraction results. Technical sensor specifications and retrieval methods cause significant variability in retrieved Fs signals. Results are intended to be one relevant component of the total uncertainty budget of Fs retrieval and must be considered in the interpretation of retrieved Fs signals.