Quantitative inversion method and system for concentration of chlorophyll in water body

Abstract

The invention relates to the technical field of water body remote sensing, in particular to a quantitative inversion method and system for the concentration of chlorophyll a in a water body, and the method comprises the steps: determining a target photon set, and carrying out the data optimization and enhancement processing based on offset correction and block photon accumulation, and obtaining optimized photon data; constructing an impulse response function reflecting the overall scattering characteristics of different height layers in the earth surface relative to the laser pulses, and performing deconvolution correction based on multiple iterations on the optimized photon data based on the impulse response function to obtain corrected photon data; constructing a water column signal model covering a water body diffusion attenuation coefficient and a backscattering coefficient on the basis of the corrected photon data, carrying out simultaneous operation on the water column signal model and a semi-analysis model which is determined on the basis of the remote sensing data and corresponds to the working wavelength of 532nm, and resolving to obtain the water body diffusion attenuation coefficient; and realizing quantitative inversion of the chlorophyll a concentration based on an empirical conversion relationship between the water diffusion attenuation coefficient and the chlorophyll a concentration.

Description

Technical Field

[0001] The present invention relates to the technical field of water body remote sensing, and in particular to a quantitative inversion method and system for chlorophyll a concentration in water body. Background Art

[0002] With global climate change and intensified human activities, aquatic ecosystems face numerous challenges. Accurate monitoring of chlorophyll a concentration is crucial for assessing eutrophication, ecosystem health, and the ocean carbon cycle. Traditional methods rely primarily on field sampling and laboratory analysis, but this approach is time-consuming and labor-intensive, making it difficult to monitor large areas with high temporal and spatial resolution. Recent advances in remote sensing technology have provided a new approach for monitoring chlorophyll a concentration in water. However, existing remote sensing technology, primarily based on optical satellite data, has limited spatial and vertical resolution, making it difficult to accurately monitor the vertical distribution of chlorophyll a concentration in water. To improve monitoring accuracy, researchers have proposed inverting chlorophyll a concentration in water by combining photon-counting lidar data with an inversion model linking water optical properties and chlorophyll a concentration. However, practical applications still face numerous technical challenges, such as noise interference from photon signals, signal offset caused by refraction, and complex water optical properties, all of which affect the accuracy and efficiency of chlorophyll a monitoring. Therefore, to ensure the reliable application of photon counting lidar in monitoring chlorophyll a concentration in water bodies, it is necessary to conduct in-depth research and optimize the signal processing algorithm to reduce noise interference, correct the refraction effect, and fully consider the complexity of the optical properties of water bodies, so as to improve monitoring accuracy and efficiency. Summary of the Invention

[0003] In order to improve the accuracy and efficiency of chlorophyll a concentration monitoring in water bodies, the present invention aims to provide a quantitative inversion method and system for chlorophyll a concentration in water bodies. The technical solutions adopted are as follows:

[0004] In a first aspect, the present application discloses a method for quantitatively inverting chlorophyll a concentration in water, the method comprising:

[0005] S1. Determine the target photon set that has distinguished the water surface and water column photons, and perform data optimization and enhancement processing based on offset correction and block photon accumulation to obtain optimized photon data;

[0006] S2. constructing an impulse response function that reflects the overall scattering characteristics of different altitude layers on the surface relative to the laser pulse, and performing deconvolution correction based on multiple iterations on the optimized photon data based on the impulse response function to obtain corrected photon data;

[0007] S3. Constructing a water column signal model that includes a water diffusion attenuation coefficient and a backscattering coefficient based on the corrected photon data, and combining the water column signal model with a semi-analytical model corresponding to a working wavelength of 532 nm determined based on remote sensing data to calculate the water diffusion attenuation coefficient;

[0008] S4. Based on the empirical conversion relationship between the water body diffusion attenuation coefficient and the chlorophyll a concentration, quantitative inversion of the chlorophyll a concentration is achieved.

[0009] Furthermore, in step S1, determining the target photon set that has distinguished the water surface and water column photons, and performing data optimization and enhancement processing based on offset correction and block photon accumulation includes:

[0010] S11, determining a target photon set with a confidence level higher than a preset threshold, and distinguishing water surface photons and water column photons from the target photon set based on density threshold segmentation;

[0011] S12. Based on the refraction effect, correct the axis offset and elevation offset of each photon to eliminate the photon path offset;

[0012] S13. Use 4km along-track partitioning for block photon accumulation to enhance the signal quality of water column photons.

[0013] Furthermore, in step S11, determining a target photon set with a confidence level higher than a preset threshold, and distinguishing water surface photons and water column photons from the target photon set based on density threshold segmentation, includes:

[0014] S111, obtaining the initial photon set, and determining the confidence level of each photon in the set;

[0015] S112. Comparing and screening the confidence levels to obtain a target photon set having a confidence level higher than a preset threshold;

[0016] S113. Calculate the number of photons in a preset neighborhood radius corresponding to the neighborhood density of each photon in the target photon set based on the Euclidean distance of the position within the area;

[0017] S114. Obtaining a density threshold of the target photon set by averaging the obtained values of the number of photons;

[0018] S115. Based on the density threshold, the target photon set having a neighborhood density greater than or equal to the density threshold is regarded as a water surface photon, and the target photon set having a neighborhood density less than the density threshold is regarded as a water column photon.

[0019] Furthermore, in step S2, the impulse response function is constructed based on the following steps:

[0020] S21, acquiring target observation data, and screening out surface reflection photon data having a higher confidence level than a preset confidence level from the target observation data;

[0021] S22, preprocessing the surface reflected photon data, and extracting a plurality of profiles within 20 meters from the preprocessed data;

[0022] S23, constructing an impulse response curve for each profile based on the contribution of the backscatter value of each altitude layer in each profile to the set of backscatter values in the corresponding altitude interval determined based on the peak position of the ground echo signal;

[0023] S24. Average each pulse response curve along the vertical height direction to obtain a pulse response function reflecting the overall scattering characteristics of different height layers on the surface of the earth to the laser pulse.

[0024] Furthermore, in step S2, deconvolution correction is performed in each iteration based on the following formula:

[0025]

[0026] in, represents the correction result obtained after the k+1th iteration, represents the correction result after the kth iteration, F(z) represents the pre-built impulse response function, * represents the convolution operation, N u (z) represents the average underwater photon number obtained at different depths z.

[0027] Furthermore, in step S3, the water column signal model is constructed by the following steps:

[0028] S31. estimating environmental and system parameters based on the corrected water surface photon data;

[0029] S32. Based on the estimated environmental and system parameters, the water diffusion attenuation coefficient and the backscattering coefficient are taken as parameters to be solved to construct a water column signal model of the following form

[0030]

[0031] Where A represents the estimated environment and system parameters, b b (z) represents the backscattering coefficient, K d represents the water diffusion attenuation coefficient, Δz represents half of the depth layer interval, z represents the depth, and θ represents the refraction angle of the laser pulse when it enters the water column.

[0032] Furthermore, in step S31, the environment and system parameters are estimated using the following formula:

[0033]

[0034] Among them, s 2 represents the mean square slope of the water surface, T w is the water surface transmittance, n w is the refractive index of water, ρ s is the sea surface reflectivity.

[0035] Furthermore, in step S3, the semi-analytical model is constructed by the following steps:

[0036] S31. Based on the OC-CCI remote sensing dataset, obtain the initial absorption coefficient at a working wavelength of 490 nm.

[0037] S32. Based on a corresponding wavelength conversion formula, the initial absorption coefficient is converted into a target absorption coefficient having an operating wavelength of 532 nm;

[0038] S33. Based on the target absorption coefficient, a semi-analytical model for calculating the water diffusion attenuation coefficient is constructed using the following formula:

[0039] K d =(1+0.005θ a )a(532)+3.47b b ;

[0040] Among them, K d represents the water diffusion attenuation coefficient, a(532) represents the target absorption coefficient at a working wavelength of 532 nm, and b b represents the backscattering coefficient, θ a Represents the solar zenith angle at the Earth's surface.

[0041] Furthermore, in step S4, the empirical conversion relationship between the water body diffusion attenuation coefficient and the chlorophyll a concentration is determined by the following formula:

[0042] K d (λ,z)=K w (λ,z)+K bio (λ,z);

[0043]

[0044] Among them, K d (λ,z) is the water diffusion attenuation coefficient, K w (λ,z) is the diffusion attenuation coefficient of pure seawater, K bio (λ,z) is the attenuation coefficient of water particles and chlorophyll, χ(λ) and e(λ) are empirical parameters, C chl is the chlorophyll a concentration.

[0045] In a second aspect, the present application discloses a quantitative inversion system for chlorophyll a concentration in water bodies, the system comprising a data optimization module, a data correction module, a water body diffusion attenuation coefficient calculation module, and a chlorophyll a concentration quantitative inversion module, wherein:

[0046] The data optimization module is used to determine the target photon set that has distinguished the water surface and water column photons, and perform data optimization and enhancement processing based on offset correction and block photon accumulation to obtain optimized photon data;

[0047] The data correction module is used to construct an impulse response function that reflects the overall scattering characteristics of different altitude layers on the surface relative to the laser pulse, and perform deconvolution correction based on multiple iterations on the optimized photon data based on the impulse response function to obtain corrected photon data;

[0048] The water diffusion attenuation coefficient calculation module is used to construct a water column signal model that includes the water diffusion attenuation coefficient and the backscattering coefficient based on the corrected photon data, and to combine the water column signal model with a semi-analytical model corresponding to a working wavelength of 532 nm determined based on remote sensing data to calculate the water diffusion attenuation coefficient;

[0049] The chlorophyll a concentration quantitative inversion module is used to achieve quantitative inversion of chlorophyll a concentration based on the empirical conversion relationship between the water body diffusion attenuation coefficient and the chlorophyll a concentration.

[0050] The present invention has the following beneficial effects:

[0051] 1) Through offset correction and block photon accumulation, noise and systematic errors are effectively eliminated, and the signal-to-noise ratio and spatial resolution of photon data are improved;

[0052] 2) Using the impulse response function to perform multiple iterative deconvolutions can separate the scattered signals at different altitudes and eliminate the blurring effect of the laser pulse;

[0053] 3) Combining the water diffusion attenuation coefficient and backscattering coefficient, the constructed water column signal model can quantitatively describe the propagation and scattering process of laser in water. This model is then solved simultaneously with the semi-analytical model to achieve high-precision inversion of the water diffusion attenuation coefficient.

[0054] 4) Based on the empirical conversion relationship, the water diffusion attenuation coefficient is converted into chlorophyll a concentration, thereby realizing rapid and non-contact monitoring of water ecological parameters and improving the accuracy and efficiency of chlorophyll a concentration monitoring in water. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0056] Figure 1 A flow chart of a method for quantitatively inverting chlorophyll a concentration in water provided by one embodiment of the present invention;

[0057] Figure 2 This is a flowchart of the overall implementation of a quantitative inversion method for chlorophyll a concentration in water provided by one embodiment of the present invention;

[0058] Figure 3 This is a system structure diagram of a quantitative inversion system for chlorophyll a concentration in water provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0059] To further illustrate the technical means and effectiveness of the present invention in achieving its intended objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail a method and system for quantitatively inverting chlorophyll a concentration in water, including its specific implementation, structure, features, and effectiveness. In the following description, references to different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0060] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0061] The specific scheme of the quantitative inversion method and system for chlorophyll a concentration in water provided by the present invention is described in detail below with reference to the accompanying drawings.

[0062] See also Figure 1 and Figure 2 , which shows a method flow chart of a quantitative inversion method for chlorophyll a concentration in water provided by one embodiment of the present invention, the method comprising:

[0063] Step S1, determining the target photon set that has distinguished the water surface and water column photons, and performing data optimization and enhancement processing based on offset correction and block photon accumulation to obtain optimized photon data.

[0064] Step S2: constructing a pulse response function that reflects the overall scattering characteristics of different altitude layers on the surface relative to the laser pulse, and performing deconvolution correction based on multiple iterations on the optimized photon data based on the pulse response function to obtain corrected photon data.

[0065] Step S3, constructing a water column signal model covering the water body diffusion attenuation coefficient and the backscattering coefficient based on the corrected photon data, and combining the water column signal model with a semi-analytical model corresponding to a working wavelength of 532 nm determined based on remote sensing data to solve the water body diffusion attenuation coefficient.

[0066] Step S4: based on the empirical conversion relationship between the water body diffusion attenuation coefficient and the chlorophyll a concentration, quantitative inversion of the chlorophyll a concentration is achieved.

[0067] From the above, it can be seen that the quantitative inversion method of chlorophyll a concentration in water bodies disclosed in the present application effectively eliminates noise and systematic errors through offset correction and block photon accumulation, thereby improving the signal-to-noise ratio and spatial resolution of photon data; multiple iterative deconvolutions using the pulse response function can separate the scattered signals of different height layers and eliminate the blurring effect of the laser pulse; the constructed water column signal model is capable of quantitatively describing the propagation and scattering process of the laser in the water body by combining the water body diffusion attenuation coefficient and the backscattering coefficient, and the model is jointly solved with the semi-analytical model to achieve high-precision inversion of the water body diffusion attenuation coefficient; based on the empirical conversion relationship, the water body diffusion attenuation coefficient is converted into chlorophyll a concentration, thereby achieving rapid and non-contact monitoring of water body ecological parameters and improving the accuracy and efficiency of water body chlorophyll a concentration monitoring.

[0068] In one embodiment, in step S1, determining a target photon set that has distinguished water surface and water column photons, and performing data optimization and enhancement processing based on offset correction and block photon accumulation includes:

[0069] Step S11 : determining a target photon set whose confidence level is higher than a preset threshold, and distinguishing water surface photons and water column photons from the target photon set based on density threshold segmentation.

[0070] Specifically, the present application calculates the distribution density of each photon in space and sets a suitable density threshold, treating photons above the threshold as water surface photons and photons below the threshold as water column photons.

[0071] Step S12: Based on the refraction effect, correct the axis offset and elevation offset of each photon to eliminate the photon path offset.

[0072] Specifically, the axial offset Δx and the elevation offset Δz of each photon are calculated based on the following formula:

[0073]

[0074] Where z represents the uncorrected water depth, α represents the incident angle of the laser transmitter, and β = arcsin(sinα / n w ) represents the refraction angle of the laser emitter, n w Represents the refractive index of water.

[0075] In step S13, block photon accumulation is performed using 4 km along-track partitioning to enhance the signal quality of water column photons.

[0076] Specifically, since the number of water column photons emitted by a single laser is small, this application uses 4km along-track partitioning for photon accumulation to enhance signal quality. Specifically, in the vertical direction, the water column photons will be divided into 1m depth layers and 0.15m step size will be used for sliding smoothing to reduce data fluctuations. Among them, within each 4km×1m partition, this application will count the number of water column photons at different depths, and normalize it according to the number of laser emissions of ICESat-2 to obtain the average number of photons in each depth layer, and then store these average photon numbers according to the partitions, thereby completing the process of block photon accumulation.

[0077] In the above embodiment, first, by determining the target photon set with a confidence level higher than a preset threshold and segmenting based on the density threshold, the water surface photons and water column photons are effectively distinguished, providing a clear data basis for subsequent analysis; secondly, the axial offset and elevation offset of the photons are corrected based on the refraction effect, which effectively eliminates the photon path offset and improves the positioning accuracy; finally, the 4km along-track partitioning is used for block photon accumulation, which significantly enhances the signal quality of the water column photons and provides strong support for subsequent data processing.

[0078] In one embodiment, in step S11, determining a target photon set having a confidence level higher than a preset threshold, and distinguishing water surface photons and water column photons from the target photon set based on density threshold segmentation, includes:

[0079] Step S111 : obtaining the initial photon set and determining the confidence level of each photon in the set.

[0080] Step S112 , obtaining a target photon set having a confidence level higher than a preset threshold through comparison and screening of the confidence levels.

[0081] Specifically, this application sets the preset threshold to 4, and uniformly filters out photons with a confidence level greater than 4 and integrates them to obtain the target photon set. For details, please refer to the formula: P s ={(x i ,h i )∣Ci ≥C th ,i∈[1,N p ]} to understand, where P s represents the target photon set formed by integration, x i represents the position of the i-th photon in the orbital direction, h i represents the height corresponding to the i-th photon in the orbital direction, C i represents the confidence of the i-th photon, C th Indicates the preset threshold, N p Represents the total number of photons within the orbital window.

[0082] Step S113, based on the Euclidean distance of the position in the area, calculate the number of photons in the neighborhood density corresponding to each photon in the target photon set within a preset neighborhood radius.

[0083] Specifically, the Euclidean distance calculation formula is defined as:

[0084]

[0085] Among them, d ij represents the Euclidean distance between two photons i and j, x i represents the horizontal coordinate of photon i, x j represents the horizontal coordinate of photon j, z i represents the vertical coordinate of photon i, z j represents the vertical coordinate of photon j.

[0086] Specifically, the present application calculates the number of photons within a preset neighborhood radius corresponding to the neighborhood density of each photon in the target photon set using the following formula:

[0087]

[0088] Among them, p i Indicates the number of photons in the neighborhood density corresponding to the i-th photon in the set within the preset neighborhood radius, N represents the total number of photons adjacent to the i-th photon in the field, and K represents the schematic The kernel function of ∈ represents the preset neighborhood radius (set in the range of 0.2 to 0.5 m).

[0089] Step S114 : Based on the obtained values of the number of photons, a density threshold of the target photon set is obtained by average calculation.

[0090] Specifically, the density threshold p of the target photon set is calculated by the following formula: initial :

[0091]

[0092] Among them, n sp Represents the target photon set P s The total number of photons in i Represents the target photon set P s The neighborhood density corresponding to the i-th photon is the number of photons within the preset neighborhood radius.

[0093] Step S115 , based on the density threshold, the target photon set having a neighborhood density greater than or equal to the density threshold is regarded as a water surface photon, and the target photon set having a neighborhood density less than the density threshold is regarded as a water column photon.

[0094] In the above embodiment, by obtaining the initial photon set and determining the confidence, the high-quality target photon set is effectively screened out, and the interference of low-confidence photons is eliminated; the neighborhood density is calculated based on the Euclidean distance, and the number of photons within the preset neighborhood radius is used to quantify the degree of photon aggregation, and a reasonable density threshold is obtained by mean calculation. This process comprehensively considers the distribution characteristics of the target photon set, making the density threshold more representative and reliable.

[0095] In one embodiment, in step S2, the impulse response function is constructed based on the following steps:

[0096] Step S21 : acquiring target observation data, and screening out surface reflection photon data having a higher confidence level than a preset level from the target observation data.

[0097] Specifically, considering that when ICESat-2 laser pulses propagate at night, the interference from solar radiation is greatly reduced and the energy attenuation is relatively stable, this application screens out surface reflection photon data that is higher than a preset confidence level based on the nighttime observation data of ICESat-2 passing through desert areas.

[0098] Step S22 , preprocessing the surface reflected photon data, and extracting a plurality of profiles within 20 meters from the preprocessed data.

[0099] Specifically, the preprocessing operations here include denoising and time synchronization correction of surface reflected photon data to eliminate outliers and errors in the data and ensure the consistency and accuracy of the data.

[0100] Step S23 : constructing an impulse response curve for each profile based on the contribution of the backscatter value of each altitude layer in each profile to the set of backscatter values in the corresponding altitude interval determined based on the peak position of the ground echo signal.

[0101] Specifically, this application constructs the impulse response curve of each profile based on the following formula:

[0102]

[0103] Among them, β(z j ) represents the magnitude of the j-th height backscatter value, p represents the peak position of the ground echo signal, β(z i ) represents the magnitude of the backscatter value corresponding to the i-th peak position.

[0104] Step S24 , averaging the pulse response curves along the vertical height direction to obtain a pulse response function reflecting the overall scattering characteristics of different height layers on the ground to the laser pulse.

[0105] Specifically, for the pulse response curves of all profiles, this application will accumulate them at the same altitude, that is, sum the curve values at each altitude to obtain the total scattering intensity at that altitude. Then, for averaging calculations, this application will divide the accumulated total scattering intensity by the total number of profiles to obtain the average scattering intensity at each altitude. In this way, a pulse response function is obtained that reflects the overall scattering characteristics of the laser pulse at different altitudes on the surface. It can describe the average scattering intensity of the target area to the laser pulse at different altitudes.

[0106] In the above embodiment, high-confidence surface reflected photon data is screened out from the target observation data, and multiple short-distance profiles are extracted through preprocessing. Then, a pulse response curve is constructed based on the contribution of the backscattering value of each height layer of the profile to the set of scattering values in a specific height interval. Finally, the pulse response function is obtained by averaging each curve along the vertical height direction. This process can accurately characterize the overall scattering characteristics of different height layers on the surface to the laser pulse, providing more accurate data support for subsequent data processing.

[0107] In one embodiment, in step S2, in each iteration, deconvolution correction is performed based on the following formula:

[0108]

[0109] in, represents the correction result obtained after the k+1th iteration, represents the correction result after the kth iteration, F(z) represents the pre-built impulse response function, * represents the convolution operation, N u (z) represents the average underwater photon number obtained at different depths z.

[0110] It should be noted that the iteration termination conditions include when the residual When the value of is less than the preset threshold, it indicates that the correction result of the current iteration is approaching convergence. Specifically, in each iteration, the correction result calculated by the above formula will gradually approach the average underwater photon number obtained at different depths. When the residual between the correction result and the true value after the iteration (i.e., the absolute value of the difference between the two) is less than the preset threshold, the iteration process is considered to have converged and the iteration will be stopped.

[0111] In one embodiment, in step S3, the water column signal model is constructed by the following steps:

[0112] Step S31 : estimating environment and system parameters based on the corrected water surface photon data.

[0113] In step S32, based on the estimated environmental and system parameters, the water diffusion attenuation coefficient and the backscattering coefficient are used as parameters to be solved to construct a water column signal model of the following form:

[0114]

[0115] Where A represents the estimated environment and system parameters, b b (z) represents the backscattering coefficient, K d represents the water diffusion attenuation coefficient, Δz represents half of the depth layer interval, z represents the depth, and θ represents the refraction angle of the laser pulse when it enters the water column.

[0116] It should be noted that the backscattering coefficient b b (z) is expressed as b by using the isotropic backscattering formula b (z) = 2πβ(π,z), water diffusion attenuation coefficient K d The relationship with the lidar attenuation coefficient α mainly depends on the size of the field of view (FOV). In ICESat-2, these two parameters are almost the same, that is, α(z)≈K d (z). Therefore, the water column signal model can be changed from the original form: Transformed into the above formula form.

[0117] In one embodiment, in step S31, the environment and system parameters are estimated using the following formula:

[0118]

[0119] Among them, s 2 represents the mean square slope of the water surface, T w is the water surface transmittance, n w is the refractive index of water, ρ s is the sea surface reflectivity.

[0120] Specifically, the mean square slope s of the water surface 2 , water surface transmittance T w , the refractive index of water n w , and sea surface reflectivity ρ s These are all known quantities, and their specific values depend on the actual application scenario and measurement data, which are not limited in this application.

[0121] In one embodiment, in step S3, the semi-analytical model is constructed by the following steps:

[0122] Step S31 : obtaining an initial absorption coefficient at a working wavelength of 490 nm based on the OC-CCI remote sensing data set.

[0123] Specifically, this application obtains the absorption coefficient a corresponding to phytoplankton at 490nm based on the OC-CCI remote sensing data set. phyto (490), corresponding absorption coefficient a of dissolved organic matter cDOM (490), and the absorption coefficient a corresponding to non-algae particles NAP (490), the above absorption coefficients are combined into a phyto (490)+a CDOM (490)+a NAP (490), and the initial absorption coefficient a(490) with a working wavelength of 490 nm is obtained.

[0124] Step S32: Based on a corresponding wavelength conversion formula, the initial absorption coefficient is converted into a target absorption coefficient with an operating wavelength of 532 nm.

[0125] Specifically, the present application converts the initial absorption coefficient a(490) into a target absorption coefficient a(532) having an operating wavelength of 532 nm through a linear transformation operation based on the following wavelength conversion formula:

[0126] a(532)=0.68×(a(490)-0.022)+0.054.

[0127] In step S33, based on the target absorption coefficient, a semi-analytical model for calculating the water diffusion attenuation coefficient is constructed using the following formula:

[0128] K d =(1+0.005θ a )a(532)+3.47b b ;

[0129] Among them, K d represents the water diffusion attenuation coefficient, a(532) represents the target absorption coefficient at a working wavelength of 532 nm, and b brepresents the backscattering coefficient, θ a Represents the solar zenith angle at the Earth's surface.

[0130] Specifically, after the target absorption coefficient a (532) is obtained based on the solution in step S32, due to the water body diffusion attenuation coefficient K d , and the backscatter coefficient b b Also appears in the water column signal model Therefore, the semi-analytical model is combined with the water column signal model By combining them, we can get the water diffusion attenuation coefficient K d The solution.

[0131] In one embodiment, in step S4, the empirical conversion relationship between the water diffusion attenuation coefficient and the chlorophyll a concentration is determined by the following formula:

[0132] K d (λ,z)=K w (λ,z)+K bio (λ,z);

[0133]

[0134] Among them, K d (λ,z) is the water diffusion attenuation coefficient, K w (λ,z) is the diffusion attenuation coefficient of pure seawater, K bio (λ,z) is the attenuation coefficient of water particles and chlorophyll, χ(λ) and e(λ) are empirical parameters, C chl is the chlorophyll a concentration.

[0135] It should be noted that when the water body diffusion attenuation coefficient K is obtained by solving the water column signal model and the semi-analytical model, d (λ, z), and the predetermined diffusion attenuation coefficient K of pure seawater w (λ,z), water particles and chlorophyll attenuation coefficient K bio By bringing (λ, z) and related empirical parameters into the above formula for mathematical calculation, the chlorophyll a concentration can be solved, thereby realizing the quantitative inversion of chlorophyll a concentration.

[0136] Please refer to Figure 3 The present application discloses a quantitative inversion system for chlorophyll a concentration in water bodies, the system comprising a data optimization module, a data correction module, a water body diffusion attenuation coefficient calculation module, and a chlorophyll a concentration quantitative inversion module, wherein:

[0137] The data optimization module is used to determine the target photon set that has distinguished the water surface and water column photons, and perform data optimization and enhancement processing based on offset correction and block photon accumulation to obtain optimized photon data.

[0138] The data correction module is used to construct a pulse response function that reflects the overall scattering characteristics of different altitude layers on the surface relative to the laser pulse, and based on the pulse response function, perform deconvolution correction based on multiple iterations on the optimized photon data to obtain corrected photon data.

[0139] The water body diffusion attenuation coefficient calculation module is used to construct a water column signal model covering the water body diffusion attenuation coefficient and the backscattering coefficient based on the corrected photon data, and to combine the water column signal model with a semi-analytical model corresponding to a working wavelength of 532nm determined based on remote sensing data to calculate the water body diffusion attenuation coefficient.

[0140] The chlorophyll a concentration quantitative inversion module is used to achieve quantitative inversion of chlorophyll a concentration based on the empirical conversion relationship between the water body diffusion attenuation coefficient and the chlorophyll a concentration.

[0141] In one embodiment, the above modules are also used to implement a quantitative inversion method for chlorophyll a concentration in water as described in any one of the above method embodiments, which is not limited in this application.

[0142] From the above, it can be seen that the quantitative inversion system of chlorophyll a concentration in water bodies disclosed in the present application effectively eliminates noise and systematic errors through offset correction and block photon accumulation, thereby improving the signal-to-noise ratio and spatial resolution of photon data; multiple iterative deconvolutions using the pulse response function can separate the scattered signals of different height layers and eliminate the blurring effect of the laser pulse; the water column signal model constructed by combining the water body diffusion attenuation coefficient and the backscattering coefficient can quantitatively describe the propagation and scattering process of the laser in the water body, and the model is solved jointly with the semi-analytical model to achieve high-precision inversion of the water body diffusion attenuation coefficient; based on the empirical conversion relationship, the water body diffusion attenuation coefficient is converted into chlorophyll a concentration, thereby achieving rapid and non-contact monitoring of water body ecological parameters and improving the accuracy and efficiency of water body chlorophyll a concentration monitoring.

[0143] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0144] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A quantitative inversion method for chlorophyll a concentration in water, characterized in that: The method comprises: S1. Determine the target photon set that has distinguished the water surface and water column photons, and perform data optimization and enhancement processing based on offset correction and block photon accumulation to obtain optimized photon data; S2. constructing an impulse response function that reflects the overall scattering characteristics of different altitude layers on the surface relative to the laser pulse, and performing deconvolution correction based on multiple iterations on the optimized photon data based on the impulse response function to obtain corrected photon data; S3. Constructing a water column signal model that includes a water diffusion attenuation coefficient and a backscattering coefficient based on the corrected photon data, and combining the water column signal model with a semi-analytical model corresponding to a working wavelength of 532 nm determined based on remote sensing data to calculate the water diffusion attenuation coefficient; S4. Based on the empirical conversion relationship between the water body diffusion attenuation coefficient and the chlorophyll a concentration, quantitative inversion of the chlorophyll a concentration is achieved.

2. The method according to claim 1, characterized in that In step S1, the target photon set that has been distinguished between water surface and water column photons is determined, and data optimization and enhancement processing is performed based on offset correction and block photon accumulation, including: S11, determining a target photon set with a confidence level higher than a preset threshold, and distinguishing water surface photons and water column photons from the target photon set based on density threshold segmentation; S12. Based on the refraction effect, correct the axis offset and elevation offset of each photon to eliminate the photon path offset; S13. Use 4km along-track partitioning for block photon accumulation to enhance the signal quality of water column photons.

3. The method according to claim 2, characterized in that In step S11, determining a target photon set with a confidence level higher than a preset threshold, and distinguishing water surface photons and water column photons from the target photon set based on density threshold segmentation, includes: S111, obtaining the initial photon set, and determining the confidence level of each photon in the set; S112. Comparing and screening the confidence levels to obtain a target photon set having a confidence level higher than a preset threshold; S113. Calculate the number of photons in a preset neighborhood radius corresponding to the neighborhood density of each photon in the target photon set based on the Euclidean distance of the position within the area; S114. Obtaining a density threshold of the target photon set by averaging the obtained values of the number of photons; S115. Based on the density threshold, the target photon set having a neighborhood density greater than or equal to the density threshold is regarded as a water surface photon, and the target photon set having a neighborhood density less than the density threshold is regarded as a water column photon.

4. The method according to claim 1, wherein In step S2, the impulse response function is constructed based on the following steps: S21, acquiring target observation data, and screening out surface reflection photon data having a higher confidence level than a preset confidence level from the target observation data; S22, preprocessing the surface reflected photon data, and extracting a plurality of profiles within 20 meters from the preprocessed data; S23, constructing an impulse response curve for each profile based on the contribution of the backscatter value of each altitude layer in each profile to the set of backscatter values in the corresponding altitude interval determined based on the peak position of the ground echo signal; S24. Average each pulse response curve along the vertical height direction to obtain a pulse response function reflecting the overall scattering characteristics of different height layers on the surface of the earth to the laser pulse.

5. The method according to claim 1, wherein In step S2, in each iteration, deconvolution correction is performed based on the following formula: in, represents the correction result obtained after the k+1th iteration, represents the correction result after the kth iteration, F(z) represents the pre-built impulse response function, * represents the convolution operation, N u (z) represents the average underwater photon number obtained at different depths z.

6. The method according to claim 1, characterized in that In step S3, the water column signal model is constructed by the following steps: S31. estimating environmental and system parameters based on the corrected water surface photon data; S32. Based on the estimated environmental and system parameters, the water diffusion attenuation coefficient and the backscattering coefficient are taken as parameters to be solved to construct a water column signal model of the following form Where A represents the estimated environment and system parameters, b b (z) represents the backscattering coefficient, K d represents the water diffusion attenuation coefficient, Δz represents half of the depth layer interval, z represents the depth, and θ represents the refraction angle of the laser pulse when it enters the water column.

7. The method according to claim 6, characterized in that In step S31, the environment and system parameters are estimated using the following formula: Among them, s 2 represents the mean square slope of the water surface, T w is the water surface transmittance, n w is the refractive index of water, ρ s is the sea surface reflectivity.

8. The method according to claim 1, characterized in that In step S3, the semi-analytical model is constructed by the following steps: S31. Based on the OC-CCI remote sensing dataset, obtain the initial absorption coefficient at a working wavelength of 490 nm. S32. Based on a corresponding wavelength conversion formula, the initial absorption coefficient is converted into a target absorption coefficient having an operating wavelength of 532 nm; S33. Based on the target absorption coefficient, a semi-analytical model for calculating the water diffusion attenuation coefficient is constructed using the following formula: K d =(1+0.005θ a )a(532)+3.47b b ; Among them, K d represents the water diffusion attenuation coefficient, a(532) represents the target absorption coefficient at a working wavelength of 532 nm, and b b represents the backscattering coefficient, θ a Represents the solar zenith angle at the Earth's surface.

9. The method according to claim 1, characterized in that In step S4, the empirical conversion relationship between the water body diffusion attenuation coefficient and the chlorophyll a concentration is determined by the following formula: K d (λ,z)=K w (λ,z)+K bio (λ,z); Among them, K d (λ,z) is the water diffusion attenuation coefficient, K w (λ,z) is the diffusion attenuation coefficient of pure seawater, K bio (λ,z) is the attenuation coefficient of water particles and chlorophyll, χ(λ) and e(λ) are empirical parameters, C chl is the chlorophyll a concentration.

10. A quantitative inversion system for chlorophyll a concentration in water, characterized in that: The system includes a data optimization module, a data correction module, a water body diffusion attenuation coefficient calculation module, and a chlorophyll a concentration quantitative inversion module, wherein: The data optimization module is used to determine the target photon set that has distinguished the water surface and water column photons, and perform data optimization and enhancement processing based on offset correction and block photon accumulation to obtain optimized photon data; The data correction module is used to construct an impulse response function that reflects the overall scattering characteristics of different altitude layers on the surface relative to the laser pulse, and perform deconvolution correction based on multiple iterations on the optimized photon data based on the impulse response function to obtain corrected photon data; The water diffusion attenuation coefficient calculation module is used to construct a water column signal model that includes the water diffusion attenuation coefficient and the backscattering coefficient based on the corrected photon data, and to combine the water column signal model with a semi-analytical model corresponding to a working wavelength of 532 nm determined based on remote sensing data to calculate the water diffusion attenuation coefficient; The chlorophyll a concentration quantitative inversion module is used to achieve quantitative inversion of chlorophyll a concentration based on the empirical conversion relationship between the water body diffusion attenuation coefficient and the chlorophyll a concentration.

Images