WO2024098587A1 - 基于ceemdan的农田土壤水分胁迫甄别与监测方法 - Google Patents

基于ceemdan的农田土壤水分胁迫甄别与监测方法 Download PDF

Info

Publication number
WO2024098587A1
WO2024098587A1 PCT/CN2023/078316 CN2023078316W WO2024098587A1 WO 2024098587 A1 WO2024098587 A1 WO 2024098587A1 CN 2023078316 W CN2023078316 W CN 2023078316W WO 2024098587 A1 WO2024098587 A1 WO 2024098587A1
Authority
WO
WIPO (PCT)
Prior art keywords
soil water
water stress
soil
stress
ceemdan
Prior art date
Application number
PCT/CN2023/078316
Other languages
English (en)
French (fr)
Inventor
李旭青
金永涛
王小丹
李国洪
顾行发
刘原萍
朱霞
赵起超
刘玉燕
杨秀峰
王延仓
刘天娇
张文豪
赵辰雨
Original Assignee
北华航天工业学院
中国科学院空天信息创新研究院
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北华航天工业学院, 中国科学院空天信息创新研究院 filed Critical 北华航天工业学院
Priority to US18/143,849 priority Critical patent/US20240167947A1/en
Publication of WO2024098587A1 publication Critical patent/WO2024098587A1/zh

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
    • G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
    • G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
    • G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
    • G01N21/359—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light using near infrared light
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
    • G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
    • G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
    • G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
    • G01N21/3554—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for determining moisture content
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
    • G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
    • G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
    • G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
    • G01N21/3563—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for analysing solids; Preparation of samples therefor
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/0098—Plants or trees
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/24—Earth materials
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
    • G01N2021/1793—Remote sensing
    • G01N2021/1797—Remote sensing in landscape, e.g. crops
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N2021/8466—Investigation of vegetal material, e.g. leaves, plants, fruits
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/24—Earth materials
    • G01N33/245—Earth materials for agricultural purposes

Definitions

  • the invention relates to a method for identifying and monitoring farmland soil water stress, and relates to a method for identifying and monitoring farmland soil water stress based on CEEMDAN, and belongs to the technical field of remote sensing.
  • Traditional soil moisture measurement methods include contact measurement methods and indirect measurement methods.
  • Contact measurement methods include time domain reflectometry and capacitive sensor methods; indirect measurement methods include passive microwave radiometers and scatterometers.
  • the traditional method can measure soil moisture content in real time, but it is only suitable for small-scale situations on the experimental site.
  • the traditional method is time-consuming and labor-intensive when conducting large-scale measurements, and it is difficult to meet the needs of large-area monitoring.
  • remote sensing technology has developed rapidly, with fast remote sensing data acquisition speed and high information update frequency, which fully reflects its significant advantages in large-area dynamic real-time monitoring.
  • the optical remote sensing method has the advantage of high spatial resolution.
  • the optical vegetation index data is highly correlated with soil moisture and is the most commonly used data for soil moisture monitoring.
  • NDVI Normalized Difference Vegetation Index
  • the complexity of the crop growth environment in natural farmland ecosystems causes them to be affected by multi-source composite stress factors such as soil moisture stress, pest and disease stress, heavy metal stress, and soil fertility stress during their growth process.
  • the various stress factors have characteristics such as interactivity and accompaniment.
  • the information represented by the long-term vegetation index sequence extracted from remote sensing image data is affected by multi-source and complex factors, aliasing information of multiple time scales, and has the characteristics of nonlinearity and non-stationarity.
  • Ensemble Empirical Mode Decomposition (EEMD) algorithm is used in Empirical Mode Decomposition (Empirical Mode Decomposition).
  • the EEMD algorithm was improved on the basis of the EEMD algorithm, which can continuously decompose the original long-time series data, obtain different intrinsic mode function (IMF) components, and finally decompose into several single-frequency sequences and a residual component.
  • IMF intrinsic mode function
  • Some scholars have decomposed the monthly precipitation data using the EEMD algorithm and established a statistical forecast model, realizing the monthly forecast of summer precipitation in the middle and upper reaches of the Yellow River; other researchers have proposed a rice heavy metal stress identification model based on spatiotemporal characteristic indicators, realizing the monitoring of rice heavy metal stress; some researchers have identified the characteristics of rice heavy metal stress signals based on the EEMD algorithm, and the extraction results can better reflect the response of stress signals to rice heavy metal stress. Previous methods have proved the feasibility of using long time series data for stress identification and the advantages of the EEMD algorithm in extracting features at different time scales.
  • soil moisture monitoring is mostly based on precipitation or drought and flood disaster data, and the advantages of remote sensing image vegetation index data are not fully utilized; and the decomposition and identification of remote sensing vegetation index data are mostly for long-term stress information with stability such as heavy metals, but the soil moisture stress, which belongs to the short-term stress component, cannot be accurately identified and extracted; in addition, although the EEMD algorithm can extract features at different time scales, there will still be residual noise in the IMF after its decomposition, which affects the subsequent processing and analysis of feature components, resulting in low final monitoring accuracy.
  • CEEMDAN Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) algorithm can decompose long time series and solve the problems of noise transmission and residue, providing a new idea for the identification and monitoring of soil moisture stress in farmland.
  • CEEMDAN Adaptive Noise
  • the purpose of the invention is to provide a method for identifying and monitoring farmland soil water stress based on CEEMDAN.
  • the technical solution adopted by the present invention is:
  • a method for identifying and monitoring farmland soil water stress based on CEEMDAN includes the following steps:
  • Step 1 Remote sensing image preprocessing: Perform radiometric calibration, atmospheric correction and geometric correction preprocessing on N remote sensing image data;
  • Step 2 Construction of NDVI long time series: Calculate the NDVI long time series x(n) based on the preprocessed remote sensing image data, 1 ⁇ n ⁇ N:
  • ⁇ NIR (n) is the reflectance of the near infrared band of the nth remote sensing image
  • ⁇ RED (n) is the reflectance of the red band of the nth remote sensing image.
  • Step 3 CEEMDAN decomposition: Decompose the NDVI long time series x(n) based on the CEEMDAN algorithm.
  • the decomposition process is as follows:
  • Step 5 Identification of soil water stress sequence: The statistical descriptive index is combined with the mechanism analysis characteristics of soil water stress to identify the soil water stress sequence, and the soil water stress subsequence is identified according to the preset conditions; the IMF components that meet the short-term soil water stress characteristics and the identification conditions are selected and accumulated to form a soil water stress sequence;
  • Step 6 Ground data measurement: Use a handheld chlorophyll meter to measure the chlorophyll content of crop plant leaves, weigh the fresh biomass FB, use the drying method to dry the leaves at the preset drying temperature for a predetermined time, then turn the temperature to the preset drying temperature and dry to constant weight, then weigh the dry biomass DB and calculate the water content PWC of the plant:
  • Step 8 Fitting the soil water stress response characteristic curve: Taking soil water stress as the independent variable and the measured chlorophyll content as the dependent variable, construct the soil water stress chlorophyll response index; taking soil water stress as the dependent variable Independent variable, taking the measured plant water content as the dependent variable to construct the soil water stress plant water response index; fitting the functional relationship between soil water stress and chlorophyll content and the functional relationship between soil water stress and plant water content;
  • Step 9 Predict the degree of soil moisture stress: construct a model between the soil moisture stress amount, the chlorophyll response amount to soil moisture stress and the plant moisture response amount to soil moisture stress and soil moisture content to invert the soil moisture content and predict the degree of soil moisture stress.
  • the conditions for identifying the soil water stress sequence in step 5 are that the fluctuation period P r is less than 7, the mean M r is minimum, and the variance V r , variance contribution rate C r and Pearson correlation coefficient PS r are maximum.
  • step 5 identifies the soil water stress sequence and adds the first and second IMF components IMF 1 and IMF 2 to form a soil water stress sequence;
  • x1 represents the chlorophyll response to soil water stress
  • x2 represents the plant water response to soil water stress
  • x3 represents the soil water stress amount
  • the present invention combines the CEEMDAN algorithm with remote sensing technology to achieve accurate monitoring of soil moisture in large areas of farmland and improve inversion accuracy;
  • the CEEMDAN decomposition is adopted, which has good completeness and solves the noise transmission and residual problems of the traditional algorithm; it not only solves the modal aliasing problem and the noise transmission problem, but also improves the adaptability of the algorithm and the reconstruction accuracy of the component data;
  • the present invention constructs a fitting function of multiple indicators and soil moisture content, covering the multivariate indicators on soil moisture content.
  • the multi-dimensional impact of soil moisture can be realized to achieve accurate monitoring of soil moisture content.
  • Fig. 1 is a flow chart of the present invention
  • FIG. 2 is a long time series diagram of NDVI in Example 2 of the present invention.
  • Embodiment 1 is a diagrammatic representation of Embodiment 1:
  • a method for identifying and monitoring farmland soil water stress based on CEEMDAN includes the following steps:
  • Step 1 Remote sensing image preprocessing: Perform radiometric calibration, atmospheric correction and geometric correction preprocessing on N GF-1 image data;
  • Step 2 Construction of NDVI long time series: Calculate the NDVI long time series x(n) based on the preprocessed GF-1 image data, 1 ⁇ n ⁇ N:
  • ⁇ NIR (n) is the reflectance of the near-infrared band of the n-th GF-1 image
  • ⁇ RED (n) is the reflectance of the red band of the n-th GF-1 image.
  • Step 3 CEEMDAN decomposition: Decompose the NDVI long time series x(n) based on the CEEMDAN algorithm.
  • the decomposition process is as follows:
  • adaptive white noise Z (m) (n) to the NDVI long time series x(n), where m represents the number of times the noise is added, which is generally 10 to 50. In this embodiment, it is 50.
  • the NDVI long time series signal after the first addition of the mth adaptive white noise can be expressed as:
  • the IMF components are obtained by EMD decomposition.
  • the EMD decomposition method adopts the method proposed by Guo Tengxiao et al. in "A New Method for Preprocessing Infrared Remote Sensing Spectral Signals Based on EMD” published in “Infrared and Laser Engineering” in 2013.
  • For the new signal to be decomposed Perform EMD decomposition again to obtain the IMF components
  • the second IMF component decomposed by CEEMDAN is:
  • Step 5 Identification of soil water stress sequence: The statistical descriptive index is combined with the mechanism analysis characteristics of soil water stress to identify the soil water stress sequence.
  • the soil water stress sequence is identified under the conditions that the fluctuation period P r is less than 7, the mean M r is minimum, and the variance V r , variance contribution rate C r and Pearson correlation coefficient PS r are maximum. Sequence identification; the first and second IMF components IMF 1 and IMF 2 meet the characteristics and identification conditions of short-term soil moisture stress, and the first and second IMF components IMF 1 and IMF 2 are accumulated to form a soil moisture stress sequence;
  • Step 6 Ground data measurement: Use a handheld chlorophyll meter to measure the chlorophyll content of crop plant leaves, and avoid the leaf veins and nearby positions of the plants during measurement. Use an engineer shovel to dig up complete crop plants at the sampling point, place them in a sealed bag and bring them back to the laboratory. After cleaning the impurities at the roots, use absorbent paper to absorb the moisture, and use a balance with an accuracy of 0.1g to weigh the fresh biomass FB. After using the drying method at 105°C for two hours, turn the temperature to 80°C and dry to constant weight, then weigh the dry biomass DB, and finally calculate the plant water content PWC.
  • Step 8 Fitting the soil water stress response characteristic curve: taking soil water stress as the independent variable and the measured chlorophyll content as the dependent variable to construct the soil water stress chlorophyll response index; taking soil water stress as the independent variable and the measured plant water content as the dependent variable to construct the soil water stress plant water response index; fitting the functional relationship between soil water stress and chlorophyll content and the functional relationship between soil water stress and plant water content;
  • Step 9 Predict the degree of soil water stress: construct a model between the soil water stress amount, the chlorophyll response amount to soil water stress and the plant water response amount to soil water stress and the soil water content to invert the soil water content and predict the degree of soil water stress;
  • x1 represents the chlorophyll response to soil water stress
  • x2 represents the plant water response to soil water stress
  • amount represents the amount of soil water stress.
  • This embodiment uses a quadratic curve to fit the functional relationship between chlorophyll content and soil water stress, and uses a composite curve to fit the functional relationship between plant water content and soil water stress.
  • the fitting results show that the soil water stress chlorophyll response index and the soil water stress plant water response index can effectively reflect the response degree of winter wheat leaf chlorophyll content and plant water content to soil water stress.
  • This embodiment constructs a long NDVI series of winter wheat based on the Gaofen-1 remote sensing image, decomposes the constructed long time series by CEEMDAN to obtain each IMF component, and calculates the fluctuation period, mean, variance, variance contribution rate and Pearson correlation coefficient and other statistical descriptive indicators of each component.
  • Ground data is measured at different growth stages of winter wheat to measure the chlorophyll content of plant leaves, plant water content, and soil water content data to obtain the true values of multiple ground indicators in the natural farmland ecosystem.
  • a comparative analysis is performed to identify and extract the soil water stress subsequence, and then the identified soil water stress subsequence is synthesized to obtain the soil water stress sequence.
  • Combine the soil water stress data with the ground measured data to construct the soil water stress chlorophyll response index and the soil water stress plant water response index, and finally construct a multi-index inversion model to achieve accurate monitoring of soil water.
  • CEEMDAN decomposition reaches 100%, indicating that the properties of the original data can be obtained by adding the decomposed components IMF r .
  • CEEMDAN decomposition only produces a unique IMF remainder, which effectively solves the problem of white noise transfer from high frequency to low frequency, fully demonstrates the completeness of the algorithm decomposition data reconstruction, and reflects the advantages of CEEMDAN decomposition.
  • CEEMDAN reduces the interference of white noise on the original data and retains the detailed information of the original data.
  • the use of the CEEMDAN algorithm can better capture the instantaneous impact of soil moisture stress, which belongs to the short-term stress component, and is conducive to the accurate identification and extraction of soil moisture stress and the subsequent improvement of soil moisture inversion accuracy.
  • Soil moisture stress is a short-term stress with a short fluctuation period; if the stress duration is short, the correlation between the component data characterizing short-term stress and the original data is greater than that of the component characterizing long-term stress.
  • the first and second IMF components IMF1 and IMF2 have a fluctuation period of less than 7 months (winter wheat growth period), the lowest mean, the largest variance and variance contribution rate, and the largest Pearson correlation coefficient between the two and the original data.
  • the first and second IMF components IMF1 and IMF2 are determined to be soil moisture stress sub-series.
  • the first and second IMF components IMF1 and IMF2 are synthesized to characterize the impact of soil moisture stress during the growth period of winter wheat.
  • Table 1 is a statistical description of each IMF in this embodiment.

Landscapes

  • Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Pathology (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Immunology (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Engineering & Computer Science (AREA)
  • Medicinal Chemistry (AREA)
  • Food Science & Technology (AREA)
  • General Life Sciences & Earth Sciences (AREA)
  • Geology (AREA)
  • Remote Sensing (AREA)
  • Environmental & Geological Engineering (AREA)
  • Botany (AREA)
  • Wood Science & Technology (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)

Abstract

一种基于CEEMDAN的农田土壤水分胁迫甄别与监测方法，其特征在于：包括步骤遥感影像预处理、NDVI长时间序列构建、CEEMDAN分解、统计性描述指标计算、土壤水分胁迫序列甄别、地面数据实测、构建土壤水分胁迫特征曲线、拟合土壤水分胁迫响应特征曲线和预测土壤水分胁迫程度。该方法采用CEEMDAN分解，解决了以往方法中存在的噪声残留问题以及重构精度不高等问题，分解分量数据的高重构精度更有利于捕捉土壤水分胁迫的瞬时影响，结合实测数据实现对土壤水分胁迫的甄别提取；综合多指标对土壤含水量的影响拟合多指标与土壤含水量的函数模型，结合CEEMDAN算法与遥感技术手段实现大面积农田土壤水分的精准监测。

Description

基于CEEMDAN的农田土壤水分胁迫甄别与监测方法 技术领域
本发明涉及一种农田土壤水分胁迫甄别与监测方法，涉及一种基于CEEMDAN的农田土壤水分胁迫甄别与监测方法，属于遥感技术领域。
背景技术
传统的土壤含水量测量方法包括接触式测量方法和间接测量法，接触式测量方法如时域反射法、电容感应器法等；间接测量法如被动式微波辐射仪、散射仪等。传统方式可以实时测得土壤水分含量，但其只适于实验场地上小尺度的情况。传统方式在进行大尺度测量时费时费力，难以满足大面积监测的需求。近年来遥感技术发展迅猛，遥感数据采集速度快、信息更新频率高，充分体现出其在大面积动态实时监测中的显著优势。其中，光学遥感方法具备高空间分辨率的优势，光学植被指数数据与土壤水分高度相关，是进行土壤水分监测最常用的一种数据。我国国产高分一号卫星突破了高时间分辨率、多光谱与宽覆盖相结合的光学遥感等关键技术，具有16m分辨率的多光谱波段影像，并且重访周期较短，在分辨率和幅宽的综合指标上达到了目前国内外民用光学遥感卫星的领先水平，在相关研究中得到了广泛应用。归一化植被指数(Normalized Difference Vegetation Index,NDVI)定义为近红波段与红光波段的反射率差值除以近红外波段与红光波段的反射率和值，是反映农作物长势的重要参数之一。NDVI可以消除大部分与仪器定标、太阳角、地形、云阴影和大气条件有关辐照度的变化，增强对植被的响应能力。
自然农田生态系统中农作物生长环境的复杂性导致其在生长过程中受土壤水分胁迫、病虫害胁迫、重金属胁迫、土壤肥力胁迫等多源复合胁迫因素的影响，各胁迫因素间具有交互性、伴生性等特征。由遥感影像数据提取出的长时间植被指数序列所表征信息受多源性以及复杂性因素的影响，混叠了多种时间尺度的信息，具有非线性以及非平稳性的特点。在实现特定特征的提取时需要进行处理，特别是具有自适应非线性的信号处理方法，对植被指数序列信号进行分解，提取出表征特定特征的分量。准确地甄别提取出土壤水分胁迫，剔除农作物生长过程中所受病虫害胁迫、重金属胁迫及土壤肥力胁迫等其他胁迫因素的影响是农田土壤水分精准监测的关键所在。集合经验模态分解(Ensemble Empirical Mode Decomposition，EEMD)算法在经验模态分解(Empirical Mode  Decomposition，EMD)算法的基础上进行了改进，能将原始长时间序列数据进行不断分解，获取不同本征模态函数(Intrinsic Mode Function，IMF)分量，最终分解为若干单一频率的序列以及一个残差组分的形式。有学者对逐月降水资料数据进行EEMD算法分解并建立了统计预报模型，实现了黄河中上游区域夏季降水的月预报；另有学者提出了基于时空特征指标的水稻重金属胁迫识别模型，实现了水稻重金属胁迫的监测；部分学者基于EEMD算法对水稻重金属胁迫信号特征进行了甄别，提取结果能较好地反映胁迫信号对水稻重金属胁迫的响应。以往的方法证明了长时间序列数据用于胁迫甄别的可行性以及EEMD算法提取不同时间尺度特征量的优势，但是仍存以下几个方面的缺陷与不足：首先对土壤水分的监测多是基于降水或旱涝灾害数据，没有充分利用遥感影像植被指数数据的优势；而利用遥感植被指数数据进行分解与甄别提取的大都是针对重金属等具有稳定性的长期胁迫信息，对属于短期胁迫组分的土壤水分胁迫却未能实现精准甄别与提取；此外，虽然EEMD算法可以提取不同时间尺度的特征量，但是其分解后的IMF中仍会残留噪声，影响后续特征分量的处理与分析，导致最终的监测精度不高。
自适应噪声完全集合经验模态分解(Complete Ensemble Empirical Mode Decomposition with Adaptive Noise，CEEMDAN)算法对长时间序列进行分解，能够很好解决噪声传递及残留问题，为农田土壤水分胁迫甄别与监测提供了新的思路。
发明内容
发明的目的是提供一种基于CEEMDAN的农田土壤水分胁迫甄别与监测方法。
为解决上述技术问题，本发明采用的技术方案是：
一种基于CEEMDAN的农田土壤水分胁迫甄别与监测方法，包括以下步骤：
步骤1：遥感影像预处理：对N幅遥感影像数据均进行辐射定标、大气校正以及几何校正预处理；
步骤2：NDVI长时间序列构建：基于预处理后的遥感影像数据计算NDVI长时间序列x(n),1≤n≤N：
其中ρNIR(n)为第n幅遥感影像近红外波段的反射率，ρRED(n)为第n幅遥感影像红光波段的反射率，构建自然农田生态系统中作物的NDVI长时间序列；
步骤3：CEEMDAN分解：基于CEEMDAN算法对NDVI长时间序列x(n)进行分解，其分解流程如下：
首先在NDVI长时间序列x(n)中添加自适应白噪声Z(m)(n)，其中m表示添加噪声的次数，1≤m≤M，得到第一待分解信号：
式中，m＝1,2,...,50，εm为第m次加入白噪声的标准差；CEEMDAN分解得到的第一IMF分量为：
表示第一待分解信号进行EMD分解得到的IMF分量，第一余项为S1(n)＝x(n)-IMF1；
在第r-1余项上叠加白噪声得到第r待分解信号对第r待分解信号再次进行EMD分解，得到IMF分量则第r IMF分量为：
第r余项为Sr(n)＝Sr-1(n)-IMFr r＝1,...,R；
步骤4：统计性描述指标计算：计算第一个至第R IMF分量IMFr,r＝1,...,R的统计性描述指标，所述统计性描述指标包括波动周期Pr、均值Mr、方差Vr、方差贡献率Cr和Pearson相关系数PSr，其中，Kr为第r本征模态分量IMFr，r＝1,...,R的极值点个数，




步骤5：土壤水分胁迫序列甄别：所述统计性描述指标与土壤水分胁迫的机理分析特征相结合甄别出土壤水分胁迫序列，按照预设条件进行土壤水分胁迫子序列甄别；选择符合短周期土壤水分胁迫特征及甄别条件的IMF分量累加合成为土壤水分胁迫序列；
步骤6：地面数据实测：使用手持式叶绿素测定仪对作物植株叶片进行叶绿素含量测定，称取其鲜生物量FB，采用烘干法在预设杀青温度条件下杀青预定时间后，将温度转为预设烘干温度烘干至恒重，再称取其干生物量DB，计算植株含水量PWC：
称取土壤样本湿重FW，采用烘干法在预设烘干温度下烘干至恒重，再称取其干重DW，计算土壤含水量SMC：
步骤7：构建土壤水分胁迫特征曲线：取土壤水分胁迫序列上包含实测时间t在内的两点n与n+1及对应数值y(n)与y(n+1)，拟合实测时间t对应数值，作为实测时间点对应的土壤水分胁迫量Y(t)：
Y(t)＝(t-n)(y(n+1)-y(n))+y(n)      (公式9)
步骤8：拟合土壤水分胁迫响应特征曲线：以土壤水分胁迫量为自变量，以实测叶绿素含量为因变量构建土壤水分胁迫叶绿素响应量指标；以土壤水分胁迫量为 自变量，以实测植株含水量为因变量构建土壤水分胁迫植株水分响应量指标；拟合土壤水分胁迫量与叶绿素含量的函数关系以及土壤水分胁迫量与植株含水量的函数关系；
步骤9：预测土壤水分胁迫程度：构建土壤水分胁迫量、土壤水分胁迫叶绿素响应量和土壤水分胁迫植株水分响应量三者多指标与土壤含水量间的模型反演土壤含水量，预测土壤水分胁迫程度。
进一步，所述步骤5甄别出土壤水分胁迫序列的条件为波动周期Pr小于7，均值Mr最小，方差Vr、方差贡献率Cr与Pearson相关系数PSr最大。
进一步，所述步骤5甄别出土壤水分胁迫序列将第一和第二IMF分量IMF1与IMF2累加合成为土壤水分胁迫序列；
进一步，使用二次方曲线拟合叶绿素含量与土壤水分胁迫量的函数关系，使用复合曲线拟合植株含水量与土壤水分胁迫量的函数关系。
更进一步，叶绿素含量与土壤水分胁迫量的函数关系为y＝58.241-19.917x-393.742x2；植株含水量与土壤水分胁迫量的函数关系为y＝68.121+0.404x。
进一步，所述步骤9中土壤水分胁迫量、土壤水分胁迫叶绿素响应量和土壤水分胁迫植株水分响应量与土壤含水量间的模型为：
SMC＝20.58-0.02x1+10-2x2-1.76x3-2×10-3x1x2-0.065x1x3-0.045x2x3-
1.83×10-4x1 2+7.91×10-5x2 2+3.99x3 2
其中，x1代表土壤水分胁迫叶绿素响应量，x2代表土壤水分胁迫植株水分响应量，x3代表土壤水分胁迫量。
采用上述技术方案，本发明的有益效果是：
1.本发明结合CEEMDAN算法与遥感技术手段实现大面积农田土壤水分的精准监测，提升反演精度；
2.采用CEEMDAN分解，具有较好的完备性，解决了传统算法的噪声传递及残留问题；不仅解决了模态混叠问题以及噪声传递问题，还提高了算法的自适应性以及分量数据的重构精度；
3.本发明构建多指标与土壤含水量的拟合函数，涵盖多变量指标对土壤含水 量的多方位影响，实现土壤含水量的精准监测。
附图说明
下面结合附图和具体实施方式对本发明作进一步详细的说明。
图1是本发明的流程图；
图2是本发明的实施例2中的NDVI长时间序列图。
具体实施方式
实施例1：
一种基于CEEMDAN的农田土壤水分胁迫甄别与监测方法，包括以下步骤：
步骤1：遥感影像预处理：对N幅高分一号影像数据均进行辐射定标、大气校正以及几何校正预处理；
步骤2：NDVI长时间序列构建：基于预处理后的高分一号影像数据计算NDVI长时间序列x(n),1≤n≤N：
其中ρNIR(n)为第n幅高分一号影像近红外波段的反射率，ρRED(n)为第n幅高分一号影像红光波段的反射率，构建自然农田生态系统中作物的NDVI长时间序列；
步骤3：CEEMDAN分解：基于CEEMDAN算法对NDVI长时间序列x(n)进行分解，其分解流程如下，
首先在NDVI长时间序列x(n)中添加自适应白噪声Z(m)(n)，其中m表示添加噪声的次数，一般取10至50，本实施例取50，则第一次添加第m次自适应白噪声后的NDVI长时间序列信号可表示为：
式中，m＝1,2,...,50，εm为第m次加入白噪声的标准差。则CEEMDAN分解得到的第一IMF分量为：
表示进行EMD分解得到的IMF分量，EMD分解方法采用郭腾霄等于2013年发表在《红外与激光工程》中“基于EMD的红外遥测光谱信号预处理新方法”所提的方法。第一余项为S1(n)＝x(n)-IMF1，在第一余项 上叠加白噪声得到新的待分解信号对新的待分解信号再次进行EMD分解，得到IMF分量则CEEMDAN分解得到的第二IMF分量为：
第二余项为S2(n)＝S1(n)-IMF2。重复上述流程，得到第一至第RIMF分量IMFr,r＝1,...,R，最终余项为：
SR＝SR-1(n)-IMFR
步骤4：统计性描述指标计算：计算第一个至第R IMF分量IMFr,r＝1,...,R的统计性描述指标，所述统计性描述指标包括波动周期Pr、均值Mr、方差Vr、方差贡献率Cr和Pearson相关系数PSr，其中，Kr为第r本征模态分量IMFr，r＝1,...,R的极值点个数，




步骤5：土壤水分胁迫序列甄别：所述统计性描述指标与土壤水分胁迫的机理分析特征相结合甄别出土壤水分胁迫序列，以波动周期Pr小于7，均值Mr最小，方差Vr、方差贡献率Cr与Pearson相关系数PSr最大为条件进行土壤水分胁迫子 序列甄别；其中第一和第二IMF分量IMF1与IMF2符合短周期土壤水分胁迫特征及甄别条件，将第一和第二IMF分量IMF1与IMF2累加合成为土壤水分胁迫序列；
步骤6：地面数据实测：使用手持式叶绿素测定仪对作物植株叶片进行叶绿素含量测定，测量时避开植株叶片叶脉及附近位置。使用工兵铲挖取采样点完整作物植株，置于密封袋内带回实验室，将根部杂质处理干净后用吸水纸吸干水分，用0.1g精度的天平称取其鲜生物量FB。采用烘干法在105℃下杀青两小时后，将温度转为80℃烘干至恒重，再称取其干生物量DB，最后计算植株含水量PWC。用工兵铲挖取样本点5至10cm内的土壤，将土壤样本后密封好带回实验室，用0.1g精度的天平称取土壤样本的湿重FW，在105℃的烘箱内将土壤样本烘干至恒重，然后称取其干重DW，最后计算土壤含水量SMC：

步骤7：构建土壤水分胁迫特征曲线：取土壤水分胁迫序列上包含实测时间t在内的两点n与n+1及对应数值y(n)与y(n+1)，拟合实测时间t对应数值，作为实测时间点对应的土壤水分胁迫量Y(t)：
Y(t)＝(t-n)(y(n+1)-y(n))+y(n)     (公式10)
步骤8：拟合土壤水分胁迫响应特征曲线：以土壤水分胁迫量为自变量，以实测叶绿素含量为因变量构建土壤水分胁迫叶绿素响应量指标；以土壤水分胁迫量为自变量，以实测植株含水量为因变量构建土壤水分胁迫植株水分响应量指标；拟合土壤水分胁迫量与叶绿素含量的函数关系以及土壤水分胁迫量与植株含水量的函数关系；
步骤9：预测土壤水分胁迫程度：构建土壤水分胁迫量、土壤水分胁迫叶绿素响应量和土壤水分胁迫植株水分响应量三者多指标与土壤含水量间的模型反演土壤含水量，预测土壤水分胁迫程度；
SMC＝20.58-0.02x1+10-2x2-1.76x3-2×10-3x1x2-0.065x1x3-0.045x2x3-1.83×10-4x1 2+7.91×10-5x2 2+3.99x3 2
其中，x1代表土壤水分胁迫叶绿素响应量，x2代表土壤水分胁迫植株水分响应 量，x3代表土壤水分胁迫量。
本实施例使用二次方曲线拟合叶绿素含量与土壤水分胁迫量的函数关系，使用复合曲线拟合植株含水量与土壤水分胁迫量的函数关系。叶绿素含量与土壤水分胁迫量的的函数关系为y＝58.241-19.917x-393.742x2；植株含水量与土壤水分胁迫量的函数关系为y＝68.121+0.404x。拟合结果表明土壤水分胁迫叶绿素响应量指标以及土壤水分胁迫植株水分响应量指标可以有效反映冬小麦叶片叶绿素含量以及植株含水量对土壤水分胁迫的响应程度。
本实施例基于高分一号遥感影像对冬小麦进行NDVI长时间序列构建，对所构建的长时间序列进行CEEMDAN分解得到各IMF分量，计算各分量的波动周期、均值、方差、方差贡献率以及Pearson相关系数等统计性描述指标。于冬小麦不同生育期进行地面数据实测，测定植株叶片叶绿素含量、植株含水量、土壤含水量数据，获取自然农田生态系统下地面多指标真实值。结合各IMF分量的统计性描述指标进行对比分析，甄别提取出土壤水分胁迫子序列，然后将甄别出的土壤水分胁迫子序列进行合成获得土壤水分胁迫序列。结合土壤水分胁迫数据与地面实测数据构建土壤水分胁迫叶绿素响应量指标以及土壤水分胁迫植株水分响应量指标，最终构建多指标反演模型实现土壤水分的精准监测。
本实施例使用CEEMDAN分解的重构精度达到了100％，表示将分解后的各个分量IMFr相加能够获得原始数据的性质，CEEMDAN分解只产生唯一的IMF余项，有效地解决了白噪声从高频到低频的转移传递问题，充分展示出算法分解数据重构的完备性，体现了CEEMDAN分解的优势。CEEMDAN减少了白噪声对原始数据的干扰，保留原始数据的细节信息，使用CEEMDAN算法能更好的捕捉属于短期胁迫组分的土壤水分胁迫的瞬时影响，有利于土壤水分胁迫的准确甄别提取及后续土壤含水量反演精度的提升。
农作物受到土壤水分胁迫后，存在可见光波段高反射率，近红外波段低反射率现象，导致NDVI值下降明显。土壤水分胁迫属于短期胁迫，波动周期较短；胁迫持续时间短，则表征短期胁迫的分量数据与原始数据的相关性较之表征长期胁迫的分量更大些。第一和第二IMF分量IMF1与IMF2的波动周期小于7个月(冬小麦生长周期)、均值最低、方差与方差贡献率最大且二者与原始数据的Pearson相关系数最大，判定第一至和第二IMF分量IMF1与IMF2为土壤水分胁迫子系列。将第一至和第二IMF分量IMF1与IMF2进行合成，表征冬小麦生长周期内受到的土壤水分胁迫的影响。表1为本实施例的各IMF统计性描述。
表1

Claims (6)

  1. 一种基于CEEMDAN的农田土壤水分胁迫甄别与监测方法，包括以下步骤：
    一种基于CEEMDAN的农田土壤水分胁迫甄别与监测方法，包括以下步骤：
    步骤1：遥感影像预处理：对N幅遥感影像数据均进行辐射定标、大气校正以及几何校正预处理；
    步骤2：NDVI长时间序列构建：基于预处理后的遥感影像数据计算NDVI长时间序列x(n),1≤n≤N：
    其中ρNIR(n)为第n幅遥感影像近红外波段的反射率，ρRED(n)为第n幅遥感影像红光波段的反射率，构建自然农田生态系统中作物的NDVI长时间序列；
    步骤3：CEEMDAN分解：基于CEEMDAN算法对NDVI长时间序列x(n)进行分解，其分解流程如下：
    首先在NDVI长时间序列x(n)中添加自适应白噪声Z(m)(n)，其中m表示添加噪声的次数，1≤m≤M，得到第一待分解信号：
    式中，m＝1,2,...,50，εm为第m次加入白噪声的标准差；CEEMDAN分解得到的第一IMF分量为：
    表示第一待分解信号进行EMD分解得到的IMF分量，第一余项为S1(n)＝x(n)-IMF1；
    在第r-1余项上叠加白噪声得到第r待分解信号对第r待分解信号再次进行EMD分解，得到IMF分量则第r IMF分量为：
    第r余项为Sr(n)＝Sr-1(n)-IMFr r＝1,...,R；
    步骤4：统计性描述指标计算：计算第一个至第R IMF分量IMFr,r＝1,...,R的统计性描述指标，所述统计性描述指标包括波动周期Pr、均值Mr、方差Vr、方差贡献率Cr和Pearson相关系数PSr，其中，Kr为第r本征模态分量IMFr，r＝1,...,R的极值点个数，




    步骤5：土壤水分胁迫序列甄别：所述统计性描述指标与土壤水分胁迫的机理分析特征相结合甄别出土壤水分胁迫序列，按照预设条件进行土壤水分胁迫子序列甄别；选择符合短周期土壤水分胁迫特征及甄别条件的IMF分量累加合成为土壤水分胁迫序列；
    步骤6：地面数据实测：使用手持式叶绿素测定仪对作物植株叶片进行叶绿素含量测定，称取作物植株鲜生物量FB，采用烘干法在预设杀青温度条件下杀青预定时间后，将温度转为预设烘干温度烘干至恒重，再称取其干生物量DB，计算植株含水量PWC：
    称取土壤样本湿重FW，采用烘干法在预设烘干温度下烘干至恒重，再称取其干重DW，计算土壤含水量SMC：
    步骤7：构建土壤水分胁迫特征曲线：取土壤水分胁迫序列上包含实测时间t在内的两点n与n+1及对应数值y(n)与y(n+1)，拟合实测时间t对应数值，作为实测时间点对应的土壤水分胁迫量Y(t)：
    Y(t)＝(t-n)(y(n+1)-y(n))+y(n)  (公式9)
    步骤8：拟合土壤水分胁迫响应特征曲线：以土壤水分胁迫量为自变量，以实测叶绿素含量为因变量构建土壤水分胁迫叶绿素响应量指标；以土壤水分胁迫量为自变量，以实测植株含水量为因变量构建土壤水分胁迫植株水分响应量指标；拟合土壤水分胁迫量与叶绿素含量的函数关系以及土壤水分胁迫量与植株含水量的函数关系；
    步骤9：预测土壤水分胁迫程度：构建土壤水分胁迫量、土壤水分胁迫叶绿素响应量和土壤水分胁迫植株水分响应量三者多指标与土壤含水量间的模型反演土壤含水量，预测土壤水分胁迫程度。
  2. 根据权利要求1所述的基于CEEMDAN的农田土壤水分胁迫甄别与监测方法，其特征在于：所述步骤5甄别出土壤水分胁迫序列的条件为波动周期Pr小于7，均值Mr最小，方差Vr、方差贡献率Cr与Pearson相关系数PSr最大。
  3. 根据权利要求1所述的基于CEEMDAN的农田土壤水分胁迫甄别与监测方法，其特征在于：所述步骤5甄别出土壤水分胁迫序列将第一和第二IMF分量IMF1与IMF2累加合成为土壤水分胁迫序列；
  4. 根据权利要求1所述的基于CEEMDAN的农田土壤水分胁迫甄别与监测方法，其特征在于：所述步骤8中使用二次方曲线拟合叶绿素含量与土壤水分胁迫量的函数关系，使用复合曲线拟合植株含水量与土壤水分胁迫量的函数关系。
  5. 根据权利要求4所述的基于CEEMDAN的农田土壤水分胁迫甄别与监测方法，其特征在于：所述步骤8中叶绿素含量与土壤水分胁迫量的函数关系为y＝58.241-19.917x-393.742x2；植株含水量与土壤水分胁迫量的函数关系为y＝68.121+0.404x。
  6. 根据权利要求2所述的基于CEEMDAN的农田土壤水分胁迫甄别与监测方法，其特征在于：所述步骤9中土壤水分胁迫量、土壤水分胁迫叶绿素响应量和土壤水分胁迫植株水分响应量与土壤含水量间的模型为：
    SMC＝20.58-0.02x1+10-2x2-1.76x3-2×10-3x1x2-0.065x1x3-0.045x2x3-
    1.83×10-4x1 2+7.91×10-5x2 2+3.99x3 2
    其中，x1代表土壤水分胁迫叶绿素响应量，x2代表土壤水分胁迫植株水分响应量，x3代表土壤水分胁迫量。
PCT/CN2023/078316 2022-11-10 2023-02-27 基于ceemdan的农田土壤水分胁迫甄别与监测方法 WO2024098587A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US18/143,849 US20240167947A1 (en) 2022-11-10 2023-05-05 CEEMDAN-based method for screening and monitoring soil moisture stress in agricultural fields

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211402702.3A CN115661601A (zh) 2022-11-10 2022-11-10 基于ceemdan的农田土壤水分胁迫甄别与监测方法
CN202211402702.3 2022-11-10

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/143,849 Continuation US20240167947A1 (en) 2022-11-10 2023-05-05 CEEMDAN-based method for screening and monitoring soil moisture stress in agricultural fields

Publications (1)

Publication Number Publication Date
WO2024098587A1 true WO2024098587A1 (zh) 2024-05-16

Family

ID=85015413

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/078316 WO2024098587A1 (zh) 2022-11-10 2023-02-27 基于ceemdan的农田土壤水分胁迫甄别与监测方法

Country Status (3)

Country Link
US (1) US20240167947A1 (zh)
CN (1) CN115661601A (zh)
WO (1) WO2024098587A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115661601A (zh) * 2022-11-10 2023-01-31 北华航天工业学院 基于ceemdan的农田土壤水分胁迫甄别与监测方法
CN118366059A (zh) * 2024-06-20 2024-07-19 山东锋士信息技术有限公司 一种基于光学和sar数据融合的作物需水量计算方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110322032A (zh) * 2019-04-17 2019-10-11 四川大学 一种基于ceemdan的金融时间序列组合预测方法
CN112506663A (zh) * 2020-12-17 2021-03-16 华南理工大学 基于去噪和误差修正的云服务器cpu负载预测方法、系统和介质
CN113111799A (zh) * 2021-04-19 2021-07-13 北华航天工业学院 一种基于集合经验模态分解的耕地土壤肥力水平监测方法
CN114676636A (zh) * 2022-04-01 2022-06-28 自然资源部国土卫星遥感应用中心 一种综合植被和生境特征的草原区土壤水分快速反演方法
US20220248617A1 (en) * 2017-08-22 2022-08-11 Sentek Pty Ltd. Method of Determination of Water Stress in a One or More Plants in a Crop Located in the region of a Soil Moisture Sensor Array and Knowledge of ETo
CN115661601A (zh) * 2022-11-10 2023-01-31 北华航天工业学院 基于ceemdan的农田土壤水分胁迫甄别与监测方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20220248617A1 (en) * 2017-08-22 2022-08-11 Sentek Pty Ltd. Method of Determination of Water Stress in a One or More Plants in a Crop Located in the region of a Soil Moisture Sensor Array and Knowledge of ETo
CN110322032A (zh) * 2019-04-17 2019-10-11 四川大学 一种基于ceemdan的金融时间序列组合预测方法
CN112506663A (zh) * 2020-12-17 2021-03-16 华南理工大学 基于去噪和误差修正的云服务器cpu负载预测方法、系统和介质
CN113111799A (zh) * 2021-04-19 2021-07-13 北华航天工业学院 一种基于集合经验模态分解的耕地土壤肥力水平监测方法
CN114676636A (zh) * 2022-04-01 2022-06-28 自然资源部国土卫星遥感应用中心 一种综合植被和生境特征的草原区土壤水分快速反演方法
CN115661601A (zh) * 2022-11-10 2023-01-31 北华航天工业学院 基于ceemdan的农田土壤水分胁迫甄别与监测方法

Also Published As

Publication number Publication date
CN115661601A (zh) 2023-01-31
US20240167947A1 (en) 2024-05-23

Similar Documents

Publication Publication Date Title
WO2022166939A1 (zh) 一种基于Sentinel-2卫星影像红边波段改进小麦生长早期叶面积指数估算的方法
WO2024098587A1 (zh) 基于ceemdan的农田土壤水分胁迫甄别与监测方法
CN104266982B (zh) 一种大面积虫害量化监测系统
CN105608293A (zh) 融合光谱与纹理特征的森林地上生物量反演方法及系统
CN116665073A (zh) 一种基于多源数据的玉米产量遥感估测方法
CN106033052B (zh) 一种基于高光谱数据敏感波段的植被信息提取方法
CN112348812A (zh) 林分年龄信息测量方法及装置
CN113252583B (zh) 一种基于枯草植被指数计算高寒枯草覆盖度的方法
CN106485345A (zh) 棉花全生育期冠层spad值遥感估算及估算模型构建方法
CN108876172B (zh) 一种基于改进型modis植被供水指数的地表土壤含水量评估方法
Song et al. Mapping paddy rice agriculture over China using AMSR-E time series data
CN114782834A (zh) 基于比值法的农田土壤水分反演方法及系统
CN113111799A (zh) 一种基于集合经验模态分解的耕地土壤肥力水平监测方法
Wang et al. Improving remote sensing of salinity on topsoil with crop residues using novel indices of optical and microwave bands
Lou et al. An effective method for canopy chlorophyll content estimation of marsh vegetation based on multiscale remote sensing data
Zhang et al. Improved model for estimating the biomass of Populus euphratica forest using the integration of spectral and textural features from the Chinese high-resolution remote sensing satellite GaoFen-1
CN114486783A (zh) 基于无人机多源遥感的冬小麦田土壤水分反演方法
CN116934548A (zh) 植被覆盖与生物措施因子计算方法
Yu et al. Hyperspectral database prediction of ecological characteristics for grass species of alpine grasslands
Matongera et al. Characterizing bracken fern phenological cycle using time series data derived from Sentinel-2 satellite sensor
CN113252592B (zh) 枯草参数率定方法
CN114529838A (zh) 基于卷积神经网络的土壤含氮量反演模型构建方法及系统
Shihua et al. Modeling of maize gross primary production using MODIS imagery and flux tower data
Peng et al. Hybrid inversion of radiative transfer models based on topographically corrected Landsat surface reflectance improves leaf area index and aboveground biomass retrievals of grassland on the hilly Loess Plateau
Ma et al. Using a modified water cloud model to retrive leaf area index (LAI) from RADARSAT-2 SAR data over an agriculture area

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23887297

Country of ref document: EP

Kind code of ref document: A1