Spartina alterniflora is the main invasive vegetation in wetland ecosystems, and information on its chlorophyll content is important data for quantitative research on the key ecological functions of wetland ecosystems. The Dongtan wetland of the Yangtze River estuary was used as the experimental research area, and the artificial cultivation field of Eupatorium was taken as the research object. To prevent uncertainty in the calculation of the chlorophyll content inversion factor for Spartina alterniflora leaves, harmonic analysis theory was adopted in this study. First, the original measured spectral data in the range of 400 nm ~ 1000 nm were decomposed and reconstructed by empirical mode decomposition (EMD), and the harmonic characteristic parameters of the EMD reconstruction spectral data were obtained by harmonic analysis (HA). Meanwhile, using the PROSPECT-D radiation transfer model, the simulated data was used to validate the inversion model. Based on these harmonic characteristic parameters, harmonic analysis-back propagation (HA-BP) and stepwise multiple linear regression (SMLR) models were established. Finally, the model was validated with data simulated by the PROSPECT-D model and compared the measured chlorophyll contents using the inversion values of the two models. The results show that the inversion accuracy of the HA-BP model was highest for the measured data. The determination coefficient (R²) and root mean square error (RMSE) of the model were 0.8528 and 6.8968 (μg/cm²), respectively, and the inversion accuracy of the simulated data was slightly lower than that of the measured data. The results show that the original spectral noise could be effectively suppressed by EMD and reconstruction, while harmonic analysis could compress the signal and prevent uncertainty in the spectral parameter calculation. Thus, the harmonic decomposition method could be applied to the inversion of Spartina alterniflora chlorophyll contents. The methods in this research provide a theoretical basis and technical support for the use of frequency domain parameters for the inversion of vegetation pigment contents.

中文翻译：

---

基于谐波分析的叶片叶绿素提取新方法：以互花米草为例

互花米草（Spartina alterniflora）是湿地生态系统中的主要入侵植物，其叶绿素含量的信息对于定量研究湿地生态系统的关键生态功能是重要的数据。以长江口东滩湿地为实验研究区，以紫茎泽兰的人工栽培田为研究对象。为防止互花米草叶绿素含量转化因子计算中的不确定性叶，谐波分析理论在本研究中被采用。首先，通过经验模态分解（EMD）分解和重建原始测量的400 nm〜1000 nm范围内的光谱数据，并通过谐波分析（HA）获得EMD重建光谱数据的谐波特征参数。同时，利用PROSPECT-D辐射传输模型，利用模拟数据对反演模型进行了验证。基于这些谐波特征参数，建立了谐波分析-反向传播（HA-BP）模型和逐步多元线性回归（SMLR）模型。最后，使用PROSPECT-D模型模拟的数据对模型进行验证，并使用两个模型的反演值比较测得的叶绿素含量。结果表明，对于测量数据，HA-BP模型的反演精度最高。测定系数（该模型的R ²）和均方根误差（RMSE）分别为0.8528和6.8968（μg/ cm ²），并且模拟数据的反演精度略低于实测数据。结果表明，通过EMD和重构可以有效地抑制原始频谱噪声，而谐波分析可以压缩信号并防止频谱参数计算的不确定性。因此，谐波分解法可应用于互花米草叶绿素含量的反演。本研究方法为利用频域参数反演植物色素含量提供了理论基础和技术支持。