Early enteromorpha biomass monitoring method

Abstract

The invention discloses an early-stage enteromorpha biomass monitoring method which comprises the following steps: collecting an early-stage enteromorpha frond sample, and obtaining actually measured reflectivity hyperspectral data under different biomass levels through a water tank control experiment; in combination with a spectral response function of a high-resolution satellite sensor, acquiring the equivalent reflectivity of enteromorpha at different biomass levels; calculating a plurality of spectral indexes based on the equivalent reflectivity; establishing an empirical inversion model between various spectral indexes and actually measured biomass, and screening out an optimal inversion model through precision comparison; applying the optimal model to a high-resolution satellite image to generate an early enteromorpha biomass remote sensing monitoring product; according to the invention, refined and quantitative monitoring of early enteromorpha is realized, and effective technical support is provided for scientific early warning and prevention and control of green tide disasters.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ocean environment remote sensing, and particularly relates to an early Enteromorpha biomass monitoring method. BACKGROUND

[0002] In the aspect of Enteromorpha biomass remote sensing monitoring, existing researches on middle-late Enteromorpha have formed two typical methods: (1) Coverage area method: the total biomass is obtained by multiplying the unit area biomass (BPA, kg / m²) and the satellite-inverted Enteromorpha coverage area. However, this method usually assumes that the BPA is uniform in space, and fails to fully consider regional differences, resulting in large uncertainty in the estimated results. (2) Pixel-level BPA empirical relationship method: the empirical relationship between the spectral index and the measured biomass is established to realize the Enteromorpha biomass inversion at the pixel scale. For example, the prior art establishes the relationship between FAI and BPA of middle-late Enteromorpha under laboratory conditions, and applies it to MODIS images to evaluate the interannual variation of Enteromorpha in the Yellow Sea from 2008 to 2015. The prior art constructs an enhanced vegetation index (EVI) BPA inversion model of middle-late Enteromorpha based on RTOA, which is used for total biomass estimation of Enteromorpha disaster. However, the above researches all establish the biomass remote sensing monitoring model based on the peak of Enteromorpha outbreak (i.e. middle-late Enteromorpha). At this stage, the pigment concentration of Enteromorpha is high, and its spectral characteristics are significantly different from those in the early growth stage. Therefore, the applicability of the existing model in early Enteromorpha biomass monitoring is uncertain. At the same time, the existing researches mostly rely on medium-low resolution remote sensing data (such as MODIS 1000 m), which is difficult to identify and invert early small patch Enteromorpha. So far, there is still a lack of early Enteromorpha biomass monitoring method suitable for high-resolution satellite images, which limits the fine and large-scale monitoring of the early disaster. SUMMARY

[0003] The present application aims to provide an early Enteromorpha biomass monitoring method, which solves the problem of the lack of early Enteromorpha biomass monitoring method suitable for high-resolution satellite images.

[0004] The technical scheme of the present application is as follows: (1) Collect early Enteromorpha algal samples, and obtain measured reflectance hyperspectral data at different biomass levels through a water tank control experiment; (2) Obtain the equivalent reflectance of Enteromorpha at different biomass levels in combination with the spectral response function of a high-resolution satellite sensor; (3) Calculate a plurality of spectral indices based on the equivalent reflectance; (4) Establish an empirical inversion model between the plurality of spectral indices and the measured biomass, and select the optimal inversion model through accuracy comparison; (5) Apply the optimal model to high-resolution satellite images to generate early Enteromorpha biomass remote sensing monitoring products.

[0005] Further, in step (2), the calculation formula is as follows: ; Wherein, Ref(λ) is the measured hyperspectral reflectance; SRF i (λ) is the spectral response function of the first i band; Ref i is the sensor equivalent reflectance of the first i band.

[0006] Further, in step (3), the spectral index includes: ratio vegetation index RVI, normalized difference vegetation index NDVI, enhanced vegetation index EVI, greenness index TCG and virtual baseline green algae height index VBFAH.

[0007] Further, the calculation formula is as follows: ; ; ; ; ; Wherein, RVI is the ratio index index, NDVI is the normalized difference vegetation index, EVI is the enhanced vegetation index, TCG is the greenness index, and VBFAH is the virtual baseline green algae height index; ρ i Indicates the reflectivity value Ref of a specific band i , λ i Indicates the center wavelength of the band i , respectively equal to NIR near-infrared band, red light band, green light band, blue light band. i

[0008] Further, step (4) is as follows: the optimal spectral index is screened out by comparing the correlation coefficient R, the root mean square error RMSE and the mean absolute percentage error MAPE; the formula is as follows: ; ; Wherein, BPA is the biomass per unit area kg / m², and VBFAH is the virtual baseline green algae height index.

[0009] The early Enteromorpha biomass monitoring system provided by the application comprises: ​The collection module is used for collecting early Enteromorpha samples and obtaining measured reflectance hyperspectral data under different biomass levels through water tank control experiments; The equivalent reflectance module is used for obtaining Enteromorpha equivalent reflectance under different biomass levels in combination with the spectral response function of a high-resolution satellite sensor; The spectral index module is used for calculating a plurality of spectral indices based on the equivalent reflectance; The screening module is used for establishing an empirical inversion model between the plurality of spectral indices and the measured biomass and screening out an optimal inversion model through accuracy comparison; The monitoring module is used for applying the optimal model to high-resolution satellite images to generate early Enteromorpha biomass remote sensing monitoring products.

[0010] Further, in the equivalent reflectance module, the calculation formula is as follows: ; Wherein, Ref(λ) is the measured hyperspectral reflectance; SRF i (λ) is the spectral response function of the first i waveband; Ref i is the sensor equivalent reflectance of the first i waveband.

[0011] Further, in the spectral index module, the spectral indices include: the ratio vegetation index RVI, the normalized difference vegetation index NDVI, the enhanced vegetation index EVI, the greenness index TCG and the virtual baseline green algae height index VBFAH.

[0012] Further, in the spectral index module, the calculation formula is as follows: ; ; ; ; ; Wherein, RVI is the ratio index, NDVI is the normalized difference vegetation index, EVI is the enhanced vegetation index, TCG is the greenness index, and VBFAH is the virtual baseline green algae height index; ρ i represents the reflectance value Ref of a specific waveband i , λ i represents the central wavelength of the waveband i , and i respectively equal to NIR, red, green and blue.

[0013] Further, in the screening module, the optimal spectral index is screened out by comparing the correlation coefficient R, the root mean square error RMSE and the mean absolute percentage error MAPE, and the formula is as follows: ; ; Wherein, BPA is the biomass per unit area kg / m2, and VBFAH is the virtual baseline green algae height index.

[0014] Beneficial effects: compared with the prior art, the present application has the following remarkable advantages: (1) the present application establishes a special remote sensing monitoring method for early Enteromorpha biomass, which makes up for the deficiency of the existing method mainly for middle and late Enteromorpha, and perfects the dynamic biomass monitoring of the whole process of Enteromorpha disaster. (2) Based on high spatial resolution satellite images, the present method can accurately obtain early Enteromorpha biomass remote sensing products, significantly improve the early prevention and early warning ability of green tide disaster, and enhance the scientific decision-making ability. (3) The constructed biomass inversion model framework is stable and applicable under multi-source high resolution remote sensing data, and has strong portability and application potential. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 is a flowchart of the present application; Figure 2 is a water tank experiment schematic diagram of the present application; wherein Figure 2 (a) in (a) is used for measuring the spectral reflectance under different Enteromorpha BPA conditions; Figure 2 (b) in (b) is a water tank experiment real scene photo, which shows the Enteromorpha under different BPA values; Figure 3 is an empirical model of early Enteromorpha biomass and spectral index of high resolution image of the present application; wherein Figure 3 (a) in (a) GF1-WFV; Figure 3 (b) in (b) GF2-MSS; Figure 4 is an early Enteromorpha biomass remote sensing monitoring product based on GF1-WFV and GF2-MSS images of the present application; wherein Figure 4 (a) in (a) is a GF1-WFV color image (16m); Figure 4 (b) in (b) is an early Enteromorpha biomass monitoring product; Figure 4 (c) in (c) is a GF2-MSS color image (4m); Figure 4 (d) in (d) is an early Enteromorpha biomass monitoring product. DETAILED DESCRIPTION

[0016] The technical solutions of the present application will be further described below in combination with the drawings.

[0017] AsFigure 1 As shown, the embodiment of the present application provides an early Enteromorpha biomass monitoring method, comprising the following steps: (1) Collecting early Enteromorpha algal samples, and obtaining measured reflectance hyperspectral data under different biomass levels through a water tank control experiment; specifically as follows: collecting early Enteromorpha algal samples through sea investigation in the coastal waters of Rudong, Nantong, Jiangsu Province. Using a black water tank experiment, removing water from the Enteromorpha samples, grouping them according to different weight gradients, and using a balance to weigh them. Place the samples in the water tank, calculate the Enteromorpha biomass parameter BPA according to the sample weight and water tank area. Then, use an ASD spectrometer to obtain hyperspectral reflectance data under different BPA conditions at a spectral resolution of 1 nm, as shown in Figure 2 .

[0018] (2) Combined with the spectral response function of a high-resolution satellite sensor, obtaining the equivalent reflectance of Enteromorpha under different biomass levels; specifically as follows: using the spectral response function (SRF) of a high spatial resolution satellite sensor (GF1-WFV, GF2-MSS), converting the measured hyperspectral reflectance data into sensor equivalent reflectance (Ref i ). The calculation formula is: (1); Wherein, Ref(λ) is the measured hyperspectral reflectance; SRF i (λ) is the spectral response function of the first i band; Ref i is the sensor equivalent reflectance of the first i band. This step calculates the equivalent reflectance of the four sensor bands (blue, green, red and near-infrared bands) of GF1-WFV and GF2-MSS respectively.

[0019] (3) Based on the equivalent reflectance, calculate a plurality of spectral indices (such as RVI, NDVI, EVI, TCG, VBFAH); specifically as follows: based on the obtained equivalent reflectance of Enteromorpha, calculate a plurality of typical spectral indices, including the ratio vegetation index (RVI), the normalized difference vegetation index (NDVI), the enhanced vegetation index (EVI), the greenness index (TCG) and the virtual baseline green algae height index (VBFAH), the formula is as shown in formula (2)-(6).

[0020] (2); (3); (4); (5); (6); where, p represents the reflectance value Ref of a specific band, and l represents the center wavelength of the band. For GF1-WFV and GF2-MSS sensors, l green , l red , and l NIR are 555 nm, 660 nm, and 830 nm, respectively.

[0021] (4) An empirical inversion model between multiple spectral indices and the measured biomass is established, and the optimal inversion model is selected by accuracy comparison. Specifically, according to different Enteromorpha biomass (BPA) and the corresponding spectral index values, a "spectral index-BPA" relationship model is established (such as Figure 3 shown in Table 1). By comparing the correlation coefficient (R), root mean square error (RMSE), and mean absolute percentage error (MAPE), the optimal spectral index is selected. Through comparative analysis, for GF1-WFV and GF2-MSS sensors, the VBFAH index performs best in early Enteromorpha biomass inversion. Therefore, the early Enteromorpha biomass inversion models of GF1-WFV and GF2-MSS are constructed, as shown in equations (7) and (8): (7); (8); Table 1 Early Enteromorpha biomass inversion algorithm and accuracy index based on different spectral indices ; (5) The optimal model is applied to high-resolution satellite images to generate early Enteromorpha biomass remote sensing monitoring products, realizing large-scale and quantitative monitoring. Specifically, GF1-WFV and GF2-MSS remote sensing images covering the early Enteromorpha outbreak period in the Yellow Sea are obtained, and the FLAASH model is used for atmospheric correction to obtain the ground reflectance (Ref). Then, the VBFAH index image is calculated, and the Enteromorpha distribution pixels are extracted by threshold method. For non-Enteromorpha pixels, the biomass (BPA) is assigned as 0; for Enteromorpha pixels, the inversion model (equations (7) and (8)) is applied to calculate the BPA remote sensing estimated value, thereby generating early Enteromorpha biomass remote sensing products (such as Figure 4 shown in Table 1).

[0022] Specific experiments: 1. Early Enteromorpha samples are collected through sea investigation, and the water is removed and weighed. Different weights of Enteromorpha samples are placed in a black tank, and the corresponding Enteromorpha biomass BPA is calculated. The ASD hyperspectral instrument is used to measure the hyperspectral reflectance data (spectral resolution 1 nm) under different BPA conditions, and the spectral characteristics of Enteromorpha under different biomass are obtained.

[0023] 2、Combined with the spectral response function of high spatial resolution satellite sensors (GF1-WFV, GF2-MSS), the measured hyperspectral reflectance is converted into sensor equivalent reflectance. For GF1-WFV and GF2-MSS sensors, the equivalent reflectance of four bands of blue, green, red and near infrared is obtained, which provides input data for subsequent index calculation and modeling.

[0024] 3、Based on the sensor equivalent reflectance, a variety of spectral indices (such as RVI, NDVI, EVI, TCG, VBFAH) are calculated. Combined with the spectral index data under different BPA conditions, the empirical relationship model between spectral index and BPA is established. By comparing the correlation coefficient, RMSE and MAPE value of each model, the optimal early Enteromorpha biomass inversion model of spectral index is selected.

[0025] 4、Collect GF1-WFV and GF2-MSS remote sensing images covering early Enteromorpha in the Yellow Sea, and use FLAASH method for atmospheric correction to obtain satellite reflectance data and calculate VBFAH index. Threshold method is used to identify Enteromorpha pixels, and seawater pixels are assigned a value of 0; for Enteromorpha pixels, the established early Enteromorpha biomass inversion model is used to carry out pixel-level BPA inversion, and finally the early Enteromorpha biomass remote sensing product is generated.

Claims

1. A method for monitoring the biomass of Enteromorpha in its early stage, characterized in that: The following steps are involved: (1) Collect early Enteromorpha algae samples and obtain measured reflectance hyperspectral data at different biomass levels through flume control experiments; (2) Combine the spectral response functions of the high-resolution satellite sensors GF1-WFV and GF2-MSS to obtain the equivalent reflectance of Enteromorpha at different biomass levels; (3) Calculate various spectral indices based on equivalent reflectance; (4) Establish empirical inversion models between multiple spectral indices and measured biomass, and select the optimal inversion model through accuracy comparison; (5) Apply the optimal model to high-resolution satellite images to generate early-stage Enteromorpha biomass remote sensing monitoring products.

2. The method for monitoring early Enteromorpha biomass according to claim 1, wherein: In step (2), the spectral response function (SRF) of the high spatial resolution satellite sensors GF1-WFV and GF2-MSS is used to convert the measured hyperspectral reflectance data into sensor equivalent reflectance. The calculation formula is as follows: ； Where Ref(λ) is the measured hyperspectral reflectance; SRF i (λ) is the i Spectral response function of the band; Ref i For the i Sensor equivalent reflectance of the band.

3. The method for monitoring early Enteromorpha biomass according to claim 1, wherein: In step (3), the spectral indices include: ratio vegetation index (RVI), normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), greenness index (TCG), and virtual baseline green algae height index (VBFAH).

4. The method for monitoring early Enteromorpha biomass according to claim 3, wherein: The calculation formula is as follows: ； ； ； ； ； Wherein, RVI is the ratio index, NDVI is the normalized difference vegetation index, EVI is the enhanced vegetation index, TCG is the greenness index, and VBFAH is the virtual baseline green algae height index; ρ i Indicates a specific band i Reflectance value Ref,λ i Indicates band i The central wavelength of i They are respectively equal to the NIR near-infrared band, the red light band, the green light band, and the blue light band.

5. The method for monitoring early Enteromorpha biomass according to claim 1, wherein: Step (4) is as follows: the optimal spectral index is selected by comparing the correlation coefficient R, the root mean square error RMSE and the mean absolute percentage error MAPE; the formula is as follows: ； ； Wherein, BPA is the biomass per unit area kg / m², and VBFAH is the virtual baseline green algae height index.

6. An early-stage Enteromorpha biomass monitoring system, characterized in that: include: Acquisition module: used to collect early Enteromorpha algae samples and obtain measured reflectance hyperspectral data at different biomass levels through flume control experiments; Equivalent reflectance module: used to combine the spectral response function of high-resolution satellite sensors to obtain the equivalent reflectance of Enteromorpha at different biomass levels; Spectral index module: used to calculate various spectral indices based on equivalent reflectance; Screening module: used to establish empirical inversion models between multiple spectral indices and measured biomass, and screen out the optimal inversion model through accuracy comparison; Monitoring module: used to apply the optimal model to high-resolution satellite imagery to generate remote sensing monitoring products for early Enteromorpha biomass.

7. The early enteromorpha biomass monitoring system according to claim 6, characterized in that: In the equivalent reflectivity module, the calculation formula is as follows: ； Where Ref(λ) is the measured hyperspectral reflectance; SRF i (λ) is the i Spectral response function of the band; Ref i For the i Sensor equivalent reflectance of the band.

8. The early enteromorpha biomass monitoring system according to claim 6, characterized in that: In the spectral index module, spectral indices include: ratio vegetation index RVI, normalized difference vegetation index NDVI, enhanced vegetation index EVI, greenness index TCG and virtual baseline green algae height index VBFAH.

9. The early enteromorpha biomass monitoring system according to claim 8, characterized in that: In the spectral index module, the calculation formula is as follows: ； ； ； ； ； Wherein, RVI is the ratio index, NDVI is the normalized difference vegetation index, EVI is the enhanced vegetation index, TCG is the greenness index, and VBFAH is the virtual baseline green algae height index; ρ i Indicates a specific band i Reflectance value Ref,λ i Indicates band i The central wavelength of i They are respectively equal to the NIR near-infrared band, the red light band, the green light band, and the blue light band.

10. The early enteromorpha biomass monitoring system according to claim 6, characterized in that: In the screening module, the specific steps are as follows: the optimal spectral index is screened out by comparing the correlation coefficient R, root mean square error RMSE, and mean absolute percentage error MAPE; the formula is as follows: ； ； Wherein, BPA is the biomass per unit area kg / m², and VBFAH is the virtual baseline green algae height index.

Images