Method for retrieving total suspended sediment concentration in turbid waters using geostationary operational environmental satellite-goci

Abstract

The present application is to solve the technical problem that most of the existing total suspended particulate matter remote sensing inversion algorithm is constructed on the basis of 550 nm remote sensing reflectance, and the remote sensing reflectance appears double peak characteristics with the increase of total suspended particulate matter concentration, which makes the expression of total suspended particulate matter concentration change of high turbidity water body gradually ineffective, and provides a turbid water total suspended particulate matter inversion method suitable for stationary orbit satellite GOCI. The inversion method takes the in-situ observed remote sensing reflectance and the equivalent remote sensing reflectance calculated by the satellite spectral response function as the input, determines the absorption peak and the two shoulder band positions of the non-absorption baseline of the spectral absorption characteristic curve through the correlation coefficient between the equivalent remote sensing reflectance, the first derivative, the second derivative and the total suspended particulate matter concentration, fits the linear relationship between the spectral absorption index and the total suspended particulate matter based on the spectral absorption index, and finally establishes the remote sensing inversion algorithm of the total suspended particulate matter suitable for the stationary orbit water color satellite of the high turbidity water body.

Description

Technical Field

[0001] This invention relates to the field of marine remote sensing technology, specifically to a method for inverting total suspended particulate matter in highly turbid water bodies suitable for geostationary satellite GOCI. Background Technology

[0002] In 2010, South Korea successfully launched the world's first geostationary ocean color satellite sensor, GOCI (Geostationary Ocean Color Imager), which boasts high temporal frequency and spatial resolution. GOCI can remotely sense and invert hourly-level high spatiotemporal dynamic changes in suspended matter, yellow substances, chlorophyll a (Chla), red tides, phytoplankton, macroalgae, sea ice, pollutant diffusion, fishing ground information, and marine oil spills. This unprecedented high spatiotemporal observation capability significantly enhances the ability to monitor high spatiotemporal dynamic changes in the ocean. Total suspended particulate matter (TSP) is a crucial water quality parameter affecting optical properties such as water transparency, turbidity, and color, and is of great significance in nearshore marine biogeochemical cycles and estuarine and coastal engineering research. The distribution, diffusion, sedimentation, and movement of suspended sediment in estuaries and nearshore waters have a profound impact on radiative transfer processes, underwater light field distribution, shoreline and underwater topography changes, and ecological environment changes.

[0003] Hangzhou Bay consistently exhibits high concentrations of suspended solids, sometimes exceeding 5000 mg / L, making it one of the world's most turbid bodies of water. This is due to two main factors: firstly, the inflow of sediment from the three major rivers (Yangtze, Yellow, and Lancang) into Hangzhou Bay; and secondly, the significant sediment transport from the Yangtze River estuary, resulting in a rich sediment source for the bay. Furthermore, Hangzhou Bay's shallow depth, typically between 5 and 20 meters, makes it highly susceptible to the influence of tides and waves, leading to easy resuspension of seabed sediment. The rapid and intense exchange and mixing of substances between the sea surface and seabed further contributes to the extremely high concentration of suspended solids.

[0004] The spectral reflectance characteristics of water bodies form the basis for establishing quantitative remote sensing models of total suspended particulate matter (TSP) concentration. Due to the scattering effect of particulate matter, the visible light reflectance of water bodies with TSP increases compared to natural water bodies dominated by chlorophyll and yellow substances. With increasing TSP concentration, the remote sensing reflectance gradually rises in the wavelength range above 500 nm, with a reflection peak appearing between 545-570 nm. Furthermore, the peak exhibits a "redshift" phenomenon with increasing TSP concentration. In highly turbid water bodies, the remote sensing reflectance exhibits a "double-peak" characteristic with increasing TSP concentration. The first reflection peak in the yellow light band does not change significantly with concentration, but the reflection peak in the near-infrared band (800-820 nm) gradually increases in size, and the difference between the two reflection peaks gradually decreases with continuous concentration changes. The absorption and scattering characteristics of TSP determine the spectral characteristics of water-leaving radiation, and the extraction of inherent optical property parameters of water bodies and the establishment of remote sensing inversion algorithms for suspended particulate matter are also based on the spectral characteristics of TSP water-leaving radiation.

[0005] Furthermore, in highly turbid waters, due to the strong scattering effect of total suspended particulate matter (TSP), the near-infrared remote sensing reflectance often does not approach zero, rendering the "Black Ocean" assumption of the atmospheric correction algorithm used in current operational ocean color satellites ineffective in the near-infrared band. Therefore, the atmospheric correction algorithm fails in highly turbid waters, leading to an overestimation of near-infrared aerosol contribution and an underestimation of shortwave water-leaving radiation. Thus, accurate remote sensing estimation of TSP concentration is fundamental to studying nearshore marine material transport and is crucial for addressing the failure of atmospheric correction algorithms in highly turbid nearshore waters. However, current TSP remote sensing inversion algorithms gradually fail to represent changes in TSP concentration in highly turbid waters. This is primarily because operational inversion algorithms are built based on 550 nm remote sensing reflectance, which gradually saturates in highly turbid waters and cannot characterize higher TSP concentration changes. Therefore, a dedicated remote sensing inversion algorithm for TSP in highly turbid waters, suitable for geostationary orbit ocean color satellites (GOCI), is needed. Summary of the Invention

[0006] The purpose of this invention is to address the technical problem that most existing total suspended particulate matter remote sensing inversion algorithms are built on the basis of 550nm remote sensing reflectance. As the concentration of total suspended particulate matter increases, the remote sensing reflectance exhibits a "double peak" characteristic, which gradually renders the expression of changes in the concentration of total suspended particulate matter in highly turbid water bodies ineffective. Therefore, this invention provides a method for inverting total suspended particulate matter in turbid water bodies that is suitable for geostationary orbit satellite GOCI.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] A method for inverting total suspended particulate matter in turbid water bodies using the geostationary satellite GOCI is characterized by the following steps:

[0009] Step 1: Calculate the equivalent remote sensing reflectance R of each band based on the spectral responsivity of the geostationary ocean color satellite sensor in each band. rs (λ i ):

[0010]

[0011] Where i is the operating band number of the geostationary ocean color satellite sensor, 1≤i≤8, λ i The center wavelength of the band corresponding to the GOCI satellite;

[0012] S(λ i ) represents the spectral responsivity of the geostationary ocean color satellite sensor in each band; λ i1 and λ i2 These represent the left and right wavelengths corresponding to a spectral responsivity of 1% in the i-th band of the GOCI satellite, λ. i1 ≤λ i ≤λ i2 ;

[0013] Step 2: Using the equivalent remote sensing reflectance R for each band rs (λ i Using this as data input, the correlation coefficients between remote sensing reflectance, the first derivative and the second derivative of remote sensing reflectance and changes in suspended particulate matter concentration are calculated for each band, and the optimal sensitive band for spectral absorption characteristics is determined.

[0014] Step 3: Determine the spectral absorption characteristic curve based on the optimal sensitive band, and calculate the spectral absorption index using the remote sensing reflectance at the optimal sensitive band;

[0015] Step 4: Fit the linear relationship between the spectral absorption index (SAI) and the total suspended particulate matter (TSM) using the least squares method, calculate the slope (a) and intercept (b) of the linear relationship, and determine the TSM inversion concentration based on the spectral absorption index. E :

[0016] TSM E =Q a×SAI+b

[0017] A total suspended particulate matter inversion algorithm based on spectral absorption index was obtained;

[0018] Step 5: Verify the accuracy of the total suspended particulate matter inversion algorithm based on spectral absorption index using in-situ observation data, and calculate the relative deviation of the total suspended particulate matter inversion algorithm;

[0019] Step 6: Apply the total suspended particulate matter inversion algorithm based on spectral absorption index to the GOCI satellite, and use in-situ data to calculate the accuracy of the total suspended particulate matter inversion algorithm of the geostationary orbit water color satellite GOCI, obtain the operational remote sensing thematic map of GOCI total suspended particulate matter, and complete the quantitative inversion of total suspended particulate matter in turbid water bodies applicable to geostationary orbit satellite GOCI.

[0020] Furthermore, step 2 specifically involves:

[0021] 2.1. Using the equivalent remote sensing reflectance R for each band rs (λ i Using this as data input, calculate the first and second derivatives of the remote sensing reflectance for each band.

[0022] 2.2 Calculate the correlation coefficients between remote sensing reflectance, the first derivative and the second derivative of remote sensing reflectance and the change in suspended particulate matter concentration for each band;

[0023] The formula for calculating the correlation coefficient is:

[0024]

[0025] Where n is the number of samples of remote sensing reflectance, first derivative of remote sensing reflectance, second derivative of remote sensing reflectance, or concentration of suspended particulate matter observed in situ; j is the sample number observed in situ, 1≤j≤n; n≥30.

[0026] x j (λ i ) represents the remote sensing reflectance, first derivative, or second derivative of the remote sensing reflectance for each band corresponding to the j-th sample. y represents the average of the remote sensing reflectance, the first derivative, or the second derivative of the remote sensing reflectance for each band corresponding to all samples; j Let be the concentration of suspended particulate matter observed in situ for the j-th sample. This represents the average concentration of suspended particulate matter observed in situ for all samples.

[0027] 2.3 Determine the optimal sensitive band for remote sensing reflectance as a function of suspended particulate matter concentration;

[0028] The optimal sensitive band is the band corresponding to the maximum correlation coefficient.

[0029] Furthermore, step 3 specifically involves:

[0030] 3.1 Determining the band λ corresponding to the spectral absorption valley in the spectral absorption characteristic curve based on the optimal sensitive band. m and the corresponding band λ of its two shoulders p and λ q ;

[0031] The wavelength band λ corresponding to the spectral absorption valley point m The first derivative of the remote sensing reflectance is the largest and it is negatively correlated with the change in the concentration of suspended particulate matter.

[0032] The two shoulders correspond to the waveband λ p and λ q The bands corresponding to the spectral absorption and reflection peaks on both sides of the spectral absorption valley, with one shoulder corresponding to the band λ. p The optimal sensitive band determined in step 2.3;

[0033] 3.2 Calculate the spectral absorption index SAI:

[0034]

[0035] Where d is the weighting coefficient.

[0036] Further, in step 5, the relative deviation RE is:

[0037]

[0038] Among them, TSM E Total suspended particulate matter concentration (TSM) is derived from the inversion of the spectral absorption index. in-situ This represents the total suspended particulate matter value observed in situ.

[0039] Furthermore, in step 4, Q = 10.

[0040] Compared with the prior art, the present invention has the following beneficial technical effects:

[0041] 1. The method for inverting total suspended particulate matter in turbid water bodies applicable to the geostationary satellite GOCI provided by this invention fully considers the reason why the algorithm for official operational applications fails due to the saturation of remote sensing reflectance (Rrs) in the yellow and red bands of highly turbid water bodies caused by excessively high concentration. It introduces spectral absorption features to construct a quantitative relationship between spectral absorption index and total suspended particulate matter concentration, and then verifies the accuracy of GOCI satellite remote sensing inversion through in-situ observation data.

[0042] 2. The method for inverting total suspended particulate matter in turbid water bodies provided by the present invention, applicable to the geostationary satellite GOCI, uses the optimal sensitive band that varies with the total suspended particulate matter concentration as input to calculate the spectral absorption index. Therefore, the constructed spectral absorption index has a good correlation with the total suspended particulate matter concentration, and the selected optimal sensitive band will not gradually become saturated due to the strong backscattering effect of highly turbid water bodies. Thus, it is suitable for inverting the total suspended particulate matter concentration in highly turbid water bodies.

[0043] 3. The total suspended particulate matter inversion method for turbid water bodies provided by this invention, applicable to geostationary satellite GOCI, involves all parameters in the inversion calculation process calculated based on remote sensing reflectance data from in-situ observations. Therefore, it has strong regional applicability and can be flexibly adjusted for different sampling areas, with strong algorithm adjustment capabilities.

[0044] 4. The method for retrieving total suspended particulate matter (TSP) in turbid water bodies provided by this invention, applicable to the Geostationary Satellite GOCI, involves band-equivalent remote sensing reflectance based on the GOCI satellite band response function. Therefore, the fitting algorithm based on in-situ observation data can be directly applied to GOCI satellite remote sensing inversion. Compared to existing operational algorithms that saturate and fail in nearshore highly turbid water bodies and cannot represent changes in TSP concentration, this invention, verified by in-situ observation data, demonstrates superior accuracy compared to official operational algorithms, significantly improving the remote sensing application effectiveness of the GOCI satellite in highly turbid water bodies. Attached Figure Description

[0045] Figure 1 This is a flowchart of the method for inverting total suspended particulate matter in turbid water bodies applicable to the geostationary satellite GOCI of the present invention;

[0046] Figure 2 This is a schematic diagram illustrating the principle of calculating the spectral absorption index in an embodiment of the present invention;

[0047] Figure 3 The above are the optimal band sensitivity analysis diagrams in the embodiments of the present invention; wherein, (a) is the remote sensing reflectance data diagram of the in-situ observation of Hangzhou Bay; (b) is the remote sensing reflectance data diagram of the corresponding GOCI band of the geostationary orbit water color satellite; and (c) is the diagram of the first and second derivatives of the remote sensing reflectance of each band of the GOCI band of the geostationary orbit water color satellite, and the correlation coefficient between the remote sensing reflectance and the concentration of suspended particulate matter.

[0048] Figure 4 The following are comparison charts showing the fitting results after applying the embodiments of the present invention; wherein, (a) is a graph showing the variation of total suspended particulate matter concentration with spectral absorption index (SAI); and (b) is a graph showing the inversion results of total suspended particulate matter concentration based on spectral absorption index.

[0049] Figure 5 The following is a comparison of the results obtained by using different inversion methods for total suspended particulate matter in Hangzhou Bay; (a) is the inversion result using the empirical model proposed by Bai Yan; (b) is the inversion result using the empirical model proposed by Shen Fang; (c) is the inversion result using the empirical model proposed by Tang Junwu; and (d) is the inversion result using the official operational application model.

[0050] Figure 6The inversion method of this invention is used to retrieve the total suspended particulate matter thematic map of the high turbidity water body of Hangzhou Bay using the geostationary orbit water color satellite GOCI; (a) to (h) are the total suspended particulate matter inversion results of eight observation times of the geostationary orbit water color satellite GOCI; (i) is the diagram of the total suspended particulate matter concentration variation with tides. Detailed Implementation

[0051] To make the objectives, advantages, and features of this invention clearer, the following detailed description, in conjunction with the accompanying drawings and specific embodiments, provides a method for inverting total suspended particulate matter in turbid water bodies applicable to geostationary satellite GOCI. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of this invention and are not intended to limit the scope of protection of this invention.

[0052] The concept of this invention is as follows: using the equivalent remote sensing reflectance calculated from the spectral response function of each band of the GOCI satellite based on the remote sensing reflectance observed in situ as input, the position of the absorption peak band and the positions of the two shoulder bands of the non-absorption baseline are determined by the correlation coefficient between the equivalent remote sensing reflectance, the first derivative, the second derivative and the total suspended particulate matter concentration. Then, the ratio of the remote sensing reflectance of the absorption peak to the remote sensing reflectance of the non-absorption baseline is calculated as the spectral absorption index parameter. The linear relationship between the spectral absorption index and the total suspended particulate matter is fitted again. Finally, a remote sensing inversion algorithm for total suspended particulate matter from the GOCI satellite of geostationary orbit water color is established, which is suitable for highly turbid water bodies.

[0053] The key step in the specific implementation plan is to automatically calculate the spectral absorption index at the optimal sensitive band based on the equivalent remote sensing reflectance calculated by the spectral response function of each band of the GOCI satellite.

[0054] like Figure 1 As shown in this embodiment, a method for inverting total suspended particulate matter in turbid water bodies applicable to the geostationary satellite GOCI is provided, which specifically includes the following steps:

[0055] Step 1: Calculate the equivalent remote sensing reflectance R of each band based on the spectral responsivity of the geostationary ocean color satellite sensor in each band. rs (λ i ):

[0056]

[0057] Where i is the operating band number of the geostationary ocean color satellite sensor, 1≤i≤8, λ i The center bands corresponding to the GOCI satellites are 412nm, 443nm, 490nm, 555nm, 660nm, 680nm, 745nm, and 865nm.

[0058] S(λ i ) represents the spectral responsivity of the geostationary ocean color satellite sensor in each band; λ i1 and λ i2 These represent the left and right wavelengths corresponding to a spectral responsivity of 1% in the i-th band of the GOCI satellite, λ. i1 ≤λ i ≤λ i2 .

[0059] Step 2: Based on the spectral response functions of each band of the GOCI satellite and the remotely sensed reflectance (R0) observed in situ... rs Calculate the equivalent remote sensing reflectance. Using the equivalent remote sensing reflectance of each band as the data input, calculate the correlation coefficient between the remote sensing reflectance of each band, the first derivative and the second derivative of the remote sensing reflectance and the change in suspended particulate matter concentration. The optimal sensitive band for spectral absorption characteristics is determined by the maximum value of the correlation coefficient.

[0060] Based on the remote sensing reflectance R obtained from in-situ observations of each band in step 1 rs (i) Calculate the correlation coefficients between the equivalent remote sensing reflectance, the first derivative and the second derivative of the equivalent remote sensing reflectance and the change in suspended particulate matter concentration for each band, and determine the optimal sensitive band for the change in remote sensing reflectance with the concentration of suspended particulate matter, which is the location of the spectral absorption characteristic band.

[0061]

[0062] Where n is the number of samples of remote sensing reflectance, first derivative of remote sensing reflectance, second derivative of remote sensing reflectance, or concentration of suspended particulate matter observed in situ; j is the sample number observed in situ, i.e. 1≤j≤n.

[0063] x j (λ i ) represents the remote sensing reflectance, first derivative, or second derivative of the remote sensing reflectance for each band corresponding to the j-th sample; x(λ) i ) represents the remote sensing reflectance, the average of the first or second derivative of the remote sensing reflectance for each band corresponding to all samples; y j y represents the concentration of suspended particulate matter observed in situ for the j-th sample; y represents the average concentration of suspended particulate matter observed in situ for all samples.

[0064] Step 3: Determine the spectral absorption characteristic curve and calculate the spectral absorption index using the remote sensing reflectance at the optimal sensitive band.

[0065] 3.1. Determine the spectral absorption valley and its two shoulders in the spectral absorption characteristic curve based on the optimal sensitive wavelength; a schematic diagram of the spectral absorption index (SAI) calculation process is shown below. Figure 2As shown, the spectral absorption characteristics of a spectral curve can be composed of the spectral absorption valley (reflection peak) point M and the two shoulders S1 and S2 of the spectral absorption valley (reflection peak).

[0066] according to Figure 3 As shown in (a) to (c), the optimal wavelength range for sensitivity to changes in total suspended particulate matter concentration in highly turbid water is 745 nm, with a correlation coefficient reaching a maximum of 0.8. As the suspended particulate matter concentration gradually increases, R... rs The wavelength range (745nm) gradually increases, and even when the total suspended particulate matter concentration reaches 5000 mg / L, this band does not saturate, thus representing extremely high changes in total suspended particulate matter concentration. The first derivative of remote sensing reflectance is largest at 555nm, representing R... rs The correlation coefficient (555nm) shows the most significant change with suspended particulate matter concentration, exhibiting the largest negative correlation. As the wavelength decreases from 490nm to 412nm, the correlation coefficients between remote sensing reflectance and suspended particulate matter concentration gradually decrease across all three bands. Furthermore, the atmospheric correction algorithm for the near-infrared band exhibits significant errors in the shortwave 412nm band when processing highly turbid water bodies. Figure 2 By comparison, the remote sensing reflectance at 490nm, 555nm, and 745nm can be considered to form an inverse spectral absorption characteristic, that is, the remote sensing reflectance peak at 555nm corresponds to... Figure 2 The corresponding spectral absorption valley point M. When the concentration of suspended particulate matter is low, R increases with increasing concentration of suspended particulate matter. rs The concentration (555nm) gradually increases, changing the shape of the spectral absorption characteristics, i.e., the spectral absorption index (SAI) changes with the concentration of suspended particulate matter. However, when the concentration exceeds a certain threshold, R... rs (555nm) gradually saturates, but the second reflection peak R rs The spectral absorption index (SAI) gradually increases (745nm), and continues to change with the concentration of suspended particulate matter. The SAI does not change significantly due to excessively high concentrations of suspended particulate matter, thus solving the problem of inverting the concentration of suspended particulate matter in highly turbid water bodies.

[0067] Therefore, choose λ m =555nm is taken as the characteristic absorption valley point M, and the two bands of the non-absorption baseline are λ. p =490nm, λ q =745nm.

[0068] 3.2 Calculate the spectral absorption index (SAI) using remote sensing reflectance at the optimal sensitive band:

[0069]

[0070]

[0071] Where d is the weighting coefficient.

[0072] Step 4: Fit the linear relationship between the spectral absorption index (SAI) and the total suspended particulate matter (TSM) using the least squares method, calculate the slope (a) and intercept (b) of the linear relationship, and determine the TSM inversion concentration based on the spectral absorption index. E :

[0073] TSM E =Q a×SAI+b

[0074] In this embodiment, Q = 10, and the total suspended particulate matter concentration is represented by a logarithm to the base 10. This can express the change in total suspended particulate matter concentration from 0 to 5000 mg / L, which is three orders of magnitude larger, and has a good applicability range of the algorithm.

[0075] Step 5: Verify the accuracy of the total suspended particulate matter inversion algorithm based on spectral absorption index using in-situ observation data, and calculate the applicable range and relative deviation of the total suspended particulate matter inversion algorithm.

[0076] The applicable range of the total suspended particulate matter (TSP) retrieval algorithm: Based on the fitting formula above, ranging from 0 to infinity, it is suitable for retrieving TSP concentrations in highly turbid waters. Generally, ocean waters have lower TSP concentrations, while nearshore waters have higher TSP concentrations due to terrestrial river inputs and resuspension effects; the TSP concentration in Hangzhou Bay can reach 5000 mg / L. Therefore, the applicability of this algorithm is determined by the range of TSP concentrations at the in-situ observation sample points. Thus, the selection of in-situ sample points should cover the entire study area to ensure the representativeness of the in-situ data.

[0077] The formula for calculating relative deviation is as follows:

[0078]

[0079] Among them, TSM E Total suspended particulate matter concentration (TSM) is derived from the inversion of the spectral absorption index. in-situ This represents the total suspended particulate matter value observed in situ at Hangzhou Bay.

[0080] like Figure 4 As shown in (a) and (b), using the inversion method of this invention in remote sensing of high turbidity water in Hangzhou Bay, the relative error of the remote sensing inversion algorithm for total suspended particulate matter concentration is 24.48%, which is far higher than the industry standard of 35% for operational applications.

[0081] Figure 5As shown, a comparison of the results obtained by using different inversion methods for total suspended particulate matter in Hangzhou Bay is presented. In (a), the method used is the suspended particulate matter inversion algorithm (Bai's model-MERIS) developed by Bai Yan, and the relative deviation obtained is 29.86%.

[0082] log 10 (TSM)=A+B1×log 10 X+B2log 10 X 2 ,X=[Rrs(555)+Rrs(670)]×(Rrs(555) / Rrs(490))

[0083] A=2.355455615, B1=1.24771879, B2=0.138429417

[0084] (b) The method used is the suspended particulate matter inversion algorithm (Shen's model-MERIS) developed by Shen Fang, and the relative deviation obtained is 130.33%.

[0085]

[0086] (c) The method used is the suspended particulate matter inversion algorithm (Tang's mode) developed by Tang Junwu, and the relative deviation obtained is 88.76%.

[0087] TSM = 0.6–1760 mg / L

[0088] log 10 (TSM)=0.638+23.934×(Rrs(555)+Rrs(670))-0.5310×(Rrs(490) / Rrs(555))

[0089] (d) The method used is the official business-oriented algorithm (YOC), and the relative deviation obtained is 88.76%.

[0090] log 10 (TSM)=0.649+25.625×(Rrs(555)+Rrs(660)-0.646×(Rrs(490) / Rrs(555)

[0091] Therefore, comparison Figure 4 (b) and Figure 5 The relative deviation of the inversion method of this invention is 24.48%, and the remote sensing algorithm has the highest accuracy. The relative deviations of the other four methods fail when expressing high concentrations of total suspended particulate matter in highly turbid water bodies, and the relative errors are much greater than the operational application requirements of 35%.

[0092] The comparison results of the algorithm of this invention with commonly used algorithms in Hangzhou Bay, where the YOC algorithm is the official operational algorithm, show that the remote sensing algorithm of this invention has the highest accuracy, while the relative deviation of the operational algorithm is 107.89%. Similarly, the relative errors of the other two algorithms are 130.33% and 88.76%, respectively. The above four inversion algorithms include both empirical inversion algorithms and semi-analytical algorithms. They fail in expressing high concentrations of total suspended particulate matter in highly turbid water bodies, with relative errors far exceeding the 35% requirement for operational applications.

[0093] Step 6: Apply the total suspended particulate matter inversion algorithm based on spectral absorption index to the GOCI satellite, and use in-situ data to calculate the accuracy of the GOCI total suspended particulate matter inversion algorithm on geostationary water color satellite, and obtain the operational remote sensing thematic map of GOCI total suspended particulate matter.

[0094] Effect demonstration Figure 6 The image shows a remote sensing thematic map of total suspended particulate matter (TSP) retrieved from the GOCI ocean color satellite in Hangzhou Bay, a highly turbid water body. The suspended sediment retrieval algorithm based on spectral absorption index, established using measured data from Hangzhou Bay, can be successfully applied to the GOCI ocean color satellite, significantly improving the accuracy of TSP retrieval algorithms in highly turbid water bodies. Thematic maps (a)–(i) clearly depict the high spatiotemporal dynamics of TSP concentration in the highly turbid water body of Hangzhou Bay (08:30–15:30), with the highest TSP concentration reaching 5000 mg / L. High TSP concentrations are mainly distributed in the Yangtze River Estuary and the northern part of Hangzhou Bay, providing a reference for material transport and translocation processes in the nearshore waters of the Yangtze River Estuary.

[0095] Furthermore, data from the Zhapu tide gauge station in Hangzhou Bay shows that the total suspended particulate matter (TSP) concentration in Hangzhou Bay varies significantly with tidal processes, and the hourly variation pattern of TSP in Hangzhou Bay is mainly caused by ocean tidal action. Therefore, the TSP inversion algorithm of this invention, applicable to the geostationary ocean color satellite GOCI, can provide a new inversion approach and approach for quantitative analysis of reactive substance concentrations in nearshore highly turbid waters.

[0096] In summary, the inversion method of this invention calculates the remote sensing reflectance (R0) of in-situ observations based on the spectral response functions of each band of the GOCI satellite. rs Therefore, the constructed total suspended particulate matter fully considers the spectral characteristics of the GOCI satellite and can be directly applied to the GOCI satellite. The total suspended particulate matter inversion algorithm for highly turbid water bodies uses spectral absorption characteristics as fitting parameters, and then uses in-situ observation data and satellite data to verify the accuracy of the fitting model.

[0097] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the present invention.

Claims

1. A method for inverting total suspended particulate matter in turbid water bodies using geostationary satellite GOCI, characterized in that, Includes the following steps: Step 1: Calculate the equivalent remote sensing reflectance R of each band based on the spectral responsivity of the geostationary ocean color satellite sensor in each band. rs (λ i ): Where i is the operating band number of the geostationary ocean color satellite sensor, 1≤i≤8, λ i The wavelength of the center band corresponding to the GOCI satellite; S(λ i ) represents the spectral responsivity of the geostationary ocean color satellite sensor in each band; λ i1 and λ i2 These represent the left and right wavelengths corresponding to a spectral responsivity of 1% in the i-th band of the GOCI satellite, λ. i1 ≤λ i ≤λ i2 ; Step 2: Using the equivalent remote sensing reflectance R for each band rs (λ i Using this as data input, the correlation coefficients between remote sensing reflectance, the first derivative and the second derivative of remote sensing reflectance and changes in suspended particulate matter concentration are calculated for each band, and the optimal sensitive band for spectral absorption characteristics is determined. Step 3: Determine the spectral absorption characteristic curve based on the optimal sensitive band, and calculate the spectral absorption index using the remote sensing reflectance at the optimal sensitive band; Step 4: Fit the linear relationship between the spectral absorption index (SAI) and the total suspended particulate matter (TSM) using the least squares method, calculate the slope (a) and intercept (b) of the linear relationship, and determine the TSM inversion concentration based on the spectral absorption index. E : TSM E =Q a×SAI+b A total suspended particulate matter inversion algorithm based on spectral absorption index was obtained; Step 5: Verify the accuracy of the total suspended particulate matter inversion algorithm based on spectral absorption index using in-situ observation data, and calculate the relative deviation of the total suspended particulate matter inversion algorithm; Step 6: Apply the total suspended particulate matter inversion algorithm based on spectral absorption index to the GOCI satellite, and use in-situ data to calculate the accuracy of the GOCI total suspended particulate matter inversion algorithm, obtain the operational remote sensing thematic map of GOCI total suspended particulate matter, and complete the inversion of total suspended particulate matter in turbid water bodies applicable to the GOCI geostationary satellite.

2. The method for inverting total suspended particulate matter in turbid water bodies applicable to geostationary satellite GOCI as described in claim 1, characterized in that, Step 2 is as follows: 2.

1. Using the equivalent remote sensing reflectance R for each band rs (λ i Using this as data input, calculate the first and second derivatives of the remote sensing reflectance for each band. 2.2 Calculate the correlation coefficients between remote sensing reflectance, the first derivative and the second derivative of remote sensing reflectance and the change in suspended particulate matter concentration for each band; The formula for calculating the correlation coefficient is: Where n is the number of samples of remote sensing reflectance, first derivative of remote sensing reflectance, second derivative of remote sensing reflectance, or concentration of suspended particulate matter observed in situ; j is the sample number observed in situ, 1≤j≤n; n≥30. x j (λ i ) represents the remote sensing reflectance, first derivative, or second derivative of the remote sensing reflectance for each band corresponding to the j-th sample. y represents the average of the remote sensing reflectance, the first derivative, or the second derivative of the remote sensing reflectance for each band corresponding to all samples; j Let be the concentration of suspended particulate matter observed in situ for the j-th sample. This represents the average concentration of suspended particulate matter observed in situ for all samples. 2.3 Determine the optimal sensitive band for remote sensing reflectance as a function of suspended particulate matter concentration; The optimal sensitive band is the band corresponding to the maximum correlation coefficient.

3. The method for inverting total suspended particulate matter in turbid water bodies applicable to geostationary satellite GOCI as described in claim 2, characterized in that, Step 3 specifically involves: 3.1 Determining the band λ corresponding to the spectral absorption valley in the spectral absorption characteristic curve based on the optimal sensitive band. m and the corresponding band λ of its two shoulders p and λ q ; The wavelength band λ corresponding to the spectral absorption valley point m The first derivative of the remote sensing reflectance is the largest and it is negatively correlated with the change in the concentration of suspended particulate matter. The two shoulders correspond to the waveband λ p and λ q The bands corresponding to the spectral absorption and reflection peaks on both sides of the spectral absorption valley, with one shoulder corresponding to the band λ. p The optimal sensitive band determined in step 2.3; 3.2 Calculate the spectral absorption index SAI: Where d is the weighting coefficient.

4. The method for inverting total suspended particulate matter in turbid water bodies applicable to geostationary satellite GOCI as described in claim 3, characterized in that: In step 5, the relative deviation RE is: Among them, TSM E Total suspended particulate matter concentration (TSM) is derived from the inversion of the spectral absorption index. in-situ This represents the total suspended particulate matter value observed in situ.

5. The method for inverting total suspended particulate matter in turbid water bodies applicable to geostationary satellite GOCI according to any one of claims 1-4, characterized in that: In step 4, Q = 10.

Images