WO2020207070A1

WO2020207070A1 - Method and system for evaluating shenzhen sea water quality

Info

Publication number: WO2020207070A1
Application number: PCT/CN2019/130578
Authority: WO
Inventors: 段广拓; 韩宇; 陈劲松
Original assignee: 中国科学院深圳先进技术研究院
Priority date: 2019-04-09
Filing date: 2019-12-31
Publication date: 2020-10-15
Also published as: CN109917098A

Abstract

A method and system (10) for evaluating Shenzhen sea water quality. The method comprises: pre-processing Landsat 8 data (S1), wherein the Landsat 8 data is image data from the Landsat 8 satellite; developing a water quality parameter inversion model according to the pre-processed Landsat 8 data (S2); and developing a Shenzhen sea water quality evaluation model on the basis of the developed water quality parameter inversion model (S3). The water quality evaluation model employs multiple existing models to perform water quality evaluation and compares produced results, thereby providing an accurate, meaningful and useful evaluation result.

Description

Water Quality Evaluation Method and System of Shenzhen Sea Area

Technical field

This application belongs to the technical field of sea water quality evaluation, and particularly relates to a method and system for Shenzhen sea water quality evaluation.

Background technique

For a coastal city like Shenzhen, waters are the environmental foundation that carries all aspects of its economic development, transportation, humanities and social sciences. Therefore, periodic high-quality monitoring of waters is of far-reaching significance.

Remote sensing technology has the advantages of large-scale, fast, periodic and low-cost. The use of remote sensing technology to monitor the water area can meet the needs of monitoring for the breadth of space and time continuity, whether it is used as a separate monitoring method or compared with traditional Complementary methods can produce significant benefits.

The existing water quality evaluation system based on remote sensing technology often uses the existing evaluation model in the water quality evaluation model part, and lacks the investigation of water quality parameters and the modification of the model. A mature water quality evaluation system should be based on measured data. Statistical analysis and water area surveys should be conducted on the measured data, and the model should be improved so that the water quality evaluation results of the model are convincing.

Water quality parameters are the basis of water quality evaluation models. The water quality parameters of existing water quality evaluation systems based on remote sensing technology are generally obtained through general inversion algorithms. However, on the one hand, in oceanography, coastal water bodies are called second-class water bodies. The second-class water bodies are strongly affected by human activities, and the composition of the water bodies is complex and often has significant regional characteristics; on the other hand, the existing water quality Most of the research and development of parameter inversion models are based on the measured data of the study area, and they often have regional characteristics. In summary, the existing water quality parameter model has poor portability in the second-class water body and is not universal. If the existing inversion model is directly applied to Shenzhen waters, the accuracy will be very poor, and the results will be inaccurate and have no reference value.

Summary of the invention

This application provides a water quality evaluation method and system in Shenzhen sea area, which aims to solve one of the above-mentioned technical problems in the prior art at least to a certain extent.

In order to solve the above-mentioned problems, this application provides the following technical solutions:

A water quality evaluation method for Shenzhen sea area, the method includes the following steps: a. Preprocessing Landsat 8 data, where Landsat 8 data is Landsat 8 satellite image data; b. Performing water quality parameters based on the preprocessed Landsat 8 data Development of the inversion model; c. Based on the above-mentioned water quality parameter inversion model developed, develop a water quality evaluation model for the Shenzhen sea area.

The technical solution adopted in the embodiment of the application further includes: the step a specifically includes: radiometric correction of Landsat 8 data; atmospheric correction of Landsat 8 data after radiation correction; cloud removal of Landsat 8 data after atmospheric correction Processing: Water extraction of Landsat 8 data after cloud removal processing.

The technical solution adopted in the embodiment of the present application further includes: the water quality parameter inversion model includes: a chlorophyll a concentration inversion model, a suspended matter concentration inversion model, and a sea surface temperature inversion model.

The technical solution adopted in the embodiment of the application further includes: the development of the chlorophyll a concentration inversion model and the suspended solids concentration inversion model specifically include: screening and matching actual measured data and Landsat 8 data through data statistics and analysis; The normalized spectral analysis of the buoy point water area determines the sensitive band; the correlation analysis of the sensitive band combination and the measured data is carried out to find the band combination with the highest correlation; the inversion model is established using the sensitive band with the best correlation, and the best is selected through accuracy verification The best model is the inversion model of chlorophyll a concentration in Shenzhen sea area and the inversion model of suspended solids concentration.

The technical solution adopted in the embodiment of the application also includes: the development of the sea surface temperature inversion model specifically includes: screening and matching the measured data and Landsat 8 data through data statistics and analysis; and improving the universal single-channel algorithm: Use the MODIS near-infrared water vapor secondary product MOD05 to obtain more accurate water vapor content estimation, and fine-tune the correlation coefficients of the model in combination with Shenzhen environmental parameters; establish the sea surface temperature inversion model of Shenzhen sea area based on the above improved model and the fine-tuned model correlation coefficients .

The technical solution adopted in the embodiment of this application further includes: the step c specifically includes: selecting the concentration of chlorophyll a, the concentration of suspended solids, and the sea surface temperature through a water quality parameter inversion model, a statistical analysis of measured data, and a Shenzhen water quality survey Analyze the water environment evaluation index of Shenzhen sea area; establish the water quality evaluation model of Shenzhen sea area according to the water environment evaluation index of Shenzhen sea area: The water quality evaluation model of Shenzhen sea area is based on the comprehensive index method, and selects chlorophyll based on the design idea of time scale anomaly index a Concentration, suspended solids concentration and sea surface temperature are used as evaluation factors to establish a water quality evaluation model for Shenzhen sea area.

Another technical solution adopted by the embodiments of this application is: a Shenzhen sea area water quality evaluation system, which includes a preprocessing module, an inversion model development module, and a water quality evaluation model development module, wherein: the preprocessing module is used for the evaluation of Landsat 8 data is preprocessed, where the Landsat 8 data is Landsat 8 satellite image data; the inversion model development module is used to develop a water quality parameter inversion model based on the preprocessed Landsat 8 data; the water quality evaluation model The development module is used to develop the Shenzhen sea water quality evaluation model based on the above-mentioned water quality parameter inversion model developed.

The technical solution adopted in the embodiment of this application also includes: the pre-processing module is specifically used to: perform radiation correction on Landsat 8 data; perform atmospheric correction on Landsat 8 data after radiation correction; perform atmospheric correction on Landsat 8 data after atmospheric correction Cloud processing: Water extraction of Landsat 8 data after cloud removal processing.

The technical solution adopted in the embodiment of this application also includes: the water quality evaluation model development module is specifically used to select the concentration of chlorophyll a, the concentration of suspended solids, and the concentration of chlorophyll a through water quality parameter inversion model, statistical analysis of measured data, and Shenzhen water quality survey Analysis of the evaluation index of sea surface temperature on the water environment of Shenzhen sea area;

Establish a water quality evaluation model for the Shenzhen sea area based on the Shenzhen sea water environment evaluation index: The water quality evaluation model for the Shenzhen sea area is based on the comprehensive index method, drawing on the design idea of the time scale anomaly index to select the concentration of chlorophyll a, the concentration of suspended matter and the sea surface Temperature is used as the evaluation factor to establish the water quality evaluation model of Shenzhen sea area.

Compared with the prior art, the beneficial effects of the embodiments of the present application are that: the present application develops an inversion model of water quality parameters based on the measured data of the Shenzhen sea area, which results in higher accuracy. Regarding the water quality evaluation model, the evaluation result of the model of this application is more accurate and more indicative. Use multiple existing models to evaluate water quality and compare the results. This application selects appropriate parameters on the basis of actual measured data and water quality investigations, and combines the comprehensive index method and the time anomaly design model. The model has more application value.

Description of the drawings

Fig. 1 is a flowchart of the Shenzhen sea water quality evaluation method according to an embodiment of the application;

Figure 2 is a hardware architecture diagram of the Shenzhen sea water quality evaluation system according to an embodiment of the application;

FIG. 3 is a schematic diagram of the accuracy verification result of the chlorophyll a inversion model in the embodiment of the application;

4 is a schematic diagram of the accuracy verification result of the suspended solids concentration inversion model in the embodiment of the application;

5 is a schematic diagram of the accuracy verification result of the sea surface temperature inversion model according to the embodiment of the application;

Fig. 6 is a schematic diagram of water quality changes in Shenzhen Bay in recent years according to an embodiment of the application.

detailed description

In order to make the purpose, technical solutions, and advantages of this application clearer, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the application, and not used to limit the application.

Refer to Fig. 1, which is a flowchart of a preferred embodiment of the Shenzhen sea water quality evaluation method according to the present application.

Step S1, preprocessing the Landsat 8 data. Among them, the Landsat 8 data is Landsat 8 satellite image data. in particular:

In this embodiment, the preprocessing of Landsat 8 data includes:

a. Radiation correction for Landsat 8 data: Radiation correction is the basic process of remote sensing image processing. Its purpose is to convert the DN value of the original image into radiance, that is, the process of obtaining the reflectivity of the outer surface of the atmosphere. The relevant formula is as follows:

L _sensor ＝K×DN+T formula 1

In the above formula, L _sensor represents the apparent radiance of Landsat 8, and K and T are the two parameters of the gain and offset of the image header file: 0.0003342 and 0.1.

b. Atmospheric correction for Landsat 8 data after radiation correction: Atmospheric correction is the process of calculating the reflectance of the earth's surface from the reflectance of the atmospheric surface from the radiation correction. Atmospheric correction can eliminate the influence of atmospheric components such as carbon dioxide, particulate matter, aerosols and other substances on the radiation transmission process, thereby eliminating the errors caused by electromagnetic waves in the atmospheric transmission process. The atmospheric correction model used is the Flaash model of MODTRAN.

In the above formula, K and M are the correlation coefficients of the model, which are determined by the instantaneous observation environment of the sensor. λ and λ _e are the reflectivity and average reflectivity of the pixel point, and L _sensor represents the apparent radiance of Landsat 8.

c. De-cloud processing of Landsat 8 data after atmospheric correction: De-cloud processing can only be achieved through the cirrus band of Landsat 8 data.

location=B ₉ ＞K ₀ formula 3

In the above formula, Location represents the cloud area, B ₉ is the cirrus band pixel reflectivity value of the image after atmospheric correction, K ₀ is the cloud threshold, and the area where B _{9 is} greater than the threshold is the cloud area. At the same time, in addition to the conventional spectral bands, Landsat 8 images also have a new QA band, that is, the quality control band, which uses numerical values to indicate how the pixels are affected by clouds. In this paper, the QA band and the above formula are combined to perform cloud area detection and visual inspection of the results to ensure the reliability of the inversion model.

d. Perform water area extraction on the Landsat 8 data after cloud removal: the water area extraction work in this embodiment is mainly achieved by normalizing the water body index:

In the formula, NDWI is the water index, B _Green represents the GREEN band of the image data, and B _Nir represents the NIR band. This application selects an appropriate threshold to extract the water area according to the calculation result of the water index.

Step S2, based on the pre-processed Landsat 8 data, develop a water quality parameter inversion model. Wherein, the water quality parameters include: chlorophyll a concentration, suspended solids concentration and sea surface temperature. in particular:

In the first step, the development of the inversion model of chlorophyll a concentration and suspended solids concentration includes:

a. Filter and match the measured data and Landsat 8 data through data statistics and analysis, including data cleaning of the measured data, remove extreme value data and blank data, and then select the matching Landsat 8 data according to the time relationship; b. Determine the sensitive band through normalized spectral analysis of the buoy point water area; c. Perform correlation analysis on the sensitive band combination and the measured data to find the band combination with the highest correlation; d. Establish the inversion model: divide the measured data into thirds The second is used for modeling, and one third is used for accuracy verification. The inversion model is established using the most relevant sensitive band. After accuracy verification, the best model is screened, and the inversion model of chlorophyll a concentration in Shenzhen sea area is obtained as:

In the formula, C _chla represents the concentration of chlorophyll a, and X is the image band of Landsat 8

Band combination, TM5 stands for NIR band, TM4 stands for RED band, this band has a good chlorophyll a concentration correlation after experimental analysis.

The inversion model of suspended solids concentration in Shenzhen sea area is:

In the formula, C _TSM represents the concentration of suspended solids, and X is the image band of Landsat 8

Band combination (TM3 stands for GREEN band, TM4 stands for RED band), which has a good correlation with suspended solids concentration after experimental analysis.

The second step, the development of the sea surface temperature inversion model includes:

a. Same as step a of the first step, screen and match the measured data and Landsat 8 data through data statistics and analysis; b. Model improvement: This embodiment improves the universal single-channel algorithm, and there are some improvements ：Using the MODIS near-infrared water vapor secondary product MOD05 to obtain a more accurate water vapor content estimation, and fine-tune the correlation coefficient of the model with the environmental parameters of Shenzhen; c. Establish the sea surface temperature of Shenzhen sea area based on the above-mentioned improved model and the fine-tuned model correlation coefficient Inversion model.

The sea surface temperature inversion model of Shenzhen sea area is as follows:

Where

In the formula, λ=10.9μm, c ₁ .19104×108W·μm4·m-2·sr-1, c ₂ =1.43877×104μm·K. T ₀ and

In the data preprocessing link, γ and δ are linear approximations of Planck’s rule, and the parameter ε=0.98,

with

Is the atmospheric influence factor, and ω is the estimated water vapor content.

Step S3, based on the developed inversion model for the above water quality parameters, develop a water quality evaluation model for the Shenzhen sea area. in particular:

First, analyze the water environment evaluation indicators of Shenzhen sea area: through the statistical analysis of the measured data and the Shenzhen water quality investigation, select the degree of abnormal sea surface water temperature (ΔT), suspended solids concentration (SS) and chlorophyll content (Chla) in Shenzhen. The parameters with strong indicating significance in the water area are used as indicators and basis for water quality monitoring, evaluation and classification.

The degree of abnormal sea surface temperature change (ΔT): The sea surface temperature changes throughout the year, but for a fixed area, its interannual change should conform to a certain law, and the water temperature distribution in four seasons should have a certain scientific range. When the water temperature changes beyond the conventional interval, the phenomenon is often accompanied by the occurrence of water pollution such as warm drainage, red tide, etc. Therefore, the water temperature change can indicate the health of the water environment. In this embodiment, statistics and analysis are performed on the sea surface temperature measurement results of 13 buoy points in Shenzhen sea area from 2014 to 2016, and the average temperature of each season in the main sea area of Shenzhen is obtained as the standard measurement data T_m of temperature change. The temperature change at the position (x, y) at time t is expressed as: ΔT _{(x, y, t)} = |T _{(x, y, t)} -T _m |, and the statistical results of the four seasons average temperature of each sea area are shown in the following table.

Four seasons average temperature in Shenzhen sea area/℃(T _m )

Chlorophyll a concentration: Chlorophyll a concentration is directly related to water quality. Areas with dense algae and phytoplankton tend to have higher chlorophyll a concentration. Analyzing the measured data of chlorophyll a at Shenzhen Buoy Station, it is found that the range of chlorophyll a concentration in Shenzhen waters is mainly 0-15mg/m3. Therefore, it is reasonable to set the evaluation factor C _t ^{* of} chlorophyll a concentration in Shenzhen waters as 5mg/m3. s Choice.

Suspended matter concentration: The concentration of suspended matter has a direct indicator effect on the value of sediment and particulate matter in the water body, and is a good water environment evaluation index. Statistics of the actual measurement results of the suspended solids concentration at the buoy points in the Shenzhen waters revealed that the main distribution range of the suspended solids concentration in the Shenzhen waters is 0-40g/m3, with an average value of 19.8/m30, so the evaluation factor S ^{* of the} suspended solids is set as 10mg/ m3.

Then, according to the Shenzhen sea water environment evaluation index, the Shenzhen sea water quality evaluation model is established: the evaluation model of this embodiment is based on the comprehensive index method, and the design idea of the time scale anomaly index is used for reference. This embodiment selects the sea surface water temperature change ( ΔT), suspended solids concentration (SS) and chlorophyll content (Chla) are used as evaluation factors. The higher the model score, the more serious the pollution:

POINT _{(x, y)} = l ₁ *ΔT _{(x, y, t)} /T ^* _{(x, y, t))} +l ₂ *SS _{(x, y, t)} /S ^* +l ₃ *CHLA _{( x, y, t)} /C ^*

Formula 12

ΔT _(x,y,t) =|T _(x,y,t) -T _m |Formula 13

T ^* _{(x, y, t)} = | T _{Max (x, y, t0)} -T _{Min (x, y, t0)} | Formula 14

In the formula, POINT _{(x, y)} represents the scoring result at (x, y), ΔT _{(x, y, t)} is the abnormal sea surface temperature change at time t, and T ^* _{(x, y, t)} is the history The difference between the highest and lowest sea temperature during the same period, so the first term in Equation 8 is the anomaly between the current water temperature and the historical sea surface temperature during the same period. SS _{(x, y, t)} is the concentration of suspended particulate matter in the sea at the current time t, S ^* is the evaluation factor of suspended matter CHLA _{(x, y, t)} is the chlorophyll concentration in the sea at the current time t, and C ^* is the historical period of Shenzhen sea Chlorophyll evaluation factors, of which l ₁ , l ₂ , and l ₃ are the weights of the three factors. According to the observation of buoy sites in Shenzhen waters and the national scoring standard, it is believed that the concentration of chlorophyll and suspended solids in coastal waters has a serious impact on ocean health. Therefore, The weight distribution of the three environmental factors is as follows: l ₁ =0.2, l ₂ =0.5, and l ₃ =0.3.

Refer to FIG. 2, which is a hardware architecture diagram of the Shenzhen sea area water quality evaluation system 10 of the present application. The system includes: a preprocessing module 101, an inversion model development module 102, and a water quality evaluation model development module 103.

The preprocessing module 101 is used to preprocess Landsat 8 data. Among them, the Landsat 8 data is Landsat 8 satellite image data. in particular:

In this embodiment, the preprocessing of Landsat 8 data includes:

a. Radiation correction of Landsat 8 data: Radiation correction is the basic process of remote sensing image processing. Its purpose is to convert the DN value of the original image into radiance, which is the process of obtaining the reflectivity of the outer surface of the atmosphere. The relevant formula is as follows:

L _sensor ＝K×DN+T formula 1

b. Atmospheric correction for Landsat 8 data after radiation correction: Atmospheric correction is the process of calculating the reflectance of the earth's surface from the reflectance of the atmospheric surface from the radiation correction. Atmospheric correction can eliminate the influence of atmospheric components such as carbon dioxide, particulate matter, aerosols and other substances on the radiation transmission process, thereby eliminating the errors caused by electromagnetic waves in the atmospheric transmission process. The atmospheric correction model used is the Flaash model of MODTRAN

Location=B ₉ ＞K ₀ Formula 3

The inversion model development module 102 is used to develop a water quality parameter inversion model based on the pre-processed Landsat 8 data. Wherein, the water quality parameters include: chlorophyll a concentration, suspended solids concentration and sea surface temperature. in particular:

The band combination TM5 represents the NIR band, and TM4 represents the RED band, which has a good chlorophyll a concentration correlation after experimental analysis.

The inversion model of suspended solids concentration in Shenzhen sea area is:

The sea surface temperature inversion model of Shenzhen sea area is as follows:

Where

In the formula, λ=10.9μm, c ₁ .19104×108W·μm4·m-2·sr-1, c ₂ =1.43877×104μm·K.

T ₀ and

with

Step S3, based on the developed above-mentioned water quality parameter inversion model, develop a water quality evaluation model for the Shenzhen sea area. in particular:

The degree of abnormal sea surface temperature change (ΔT): The sea surface temperature changes throughout the year, but for a fixed area, its interannual change should conform to a certain law, and the water temperature distribution in four seasons should have a certain scientific range. When the water temperature changes beyond the conventional interval, the phenomenon is often accompanied by the occurrence of water pollution such as warm drainage, red tide, etc. Therefore, the water temperature change can indicate the health of the water environment. In this embodiment, statistics and analysis are performed on the sea surface temperature measurement results of 13 buoy points in Shenzhen sea area from 2014 to 2016, and the average temperature of each season in the main sea area of Shenzhen is obtained as the standard measurement data T_m of temperature change. The temperature change at the location (x, y) at time t is expressed as: ΔT _{(x, y, t)} = |T _{(x, y, t)} -T _m |

Four seasons average temperature in Shenzhen sea area/℃(T _m )

Formula 12

ΔT _(x,y,t) =|T _(x,y,t) -T _m |Formula 13

T ^* _{(x, y, t)} = | T _{Max (x, y, t0)} -T _{Min (x, y, t0)} | Formula 14

Experiment and verification :

This application has been verified by the accuracy of actual measured data, and after experimental testing and simulation, the results have good accuracy and feasibility. The accuracy verification of the inversion model part is as follows:

1. Verification of the accuracy of the chlorophyll a concentration inversion model: 1/3 of the measured data is used for accuracy verification, so a total of 25 data are used for accuracy verification this time. Considering the proximity effect and quality control factors, when performing spatio-temporal matching and comparison, a 3*3 grid centered on the pixel at the measured point is selected and the average value after eliminating anomalies is used as the obtained image value and the measured value for comparison and verification. Calculate the inversion results, the linear correlation between MODIS SST products and the measured data, and calculate the absolute value of the error, please refer to Figure 3.

The correlation coefficient between the predicted inversion value and the measured concentration is 0.7837, which has a good correlation. The maximum absolute error is 2.17mg/m ³ , the minimum is 0.15mg/m ³ , the average error is 0.65mg/m ³ , and the standard deviation is 0.55mg/m ^{3. The} result is accurate and feasible.

2. Verification of the inversion accuracy of suspended solids concentration: 1/3 of the measured data is used for accuracy verification, so a total of 18 data are used for accuracy verification this time. Considering the proximity effect and quality control factors, when performing spatio-temporal matching and comparison, a 3*3 grid centered on the pixel at the measured point is selected and the average value after eliminating anomalies is used as the obtained image value and the measured value for comparison and verification. Calculate the inversion results, the linear correlation between MODIS SST products and the measured data, and calculate the absolute value of the error, please refer to Figure 4.

3. Verification of the sea surface temperature inversion accuracy: In order to further verify the reliability of the inversion results, the inversion results and the MODIS SST product MOD28 are cross-validated using the measured data. In the total of 12 LANDSAT 8 images, a total of 6 images have measured data at the buoy point during the corresponding satellite transit time. The situation of MODIS SST products is the same. Among the 24 measured data corresponding to the 6 images, 8 data corresponding to the points are affected. In order to ensure the accuracy of the verification results, the cloud-covered images were excluded, so a total of 16 measured data were used for accuracy verification.

The accuracy verification results show that compared with MODIS SST products, the stability and accuracy of the experimental algorithm inversion are better. It can be seen from Figure 5 that the sample point distribution of the experimental algorithm is relatively concentrated, the fitted correlation coefficient is 0.9605, the correlation is good, and the algorithm inversion result is stable. However, the distribution of sample points of MODIS SST products is relatively scattered, and the fitted correlation coefficient is 0.6632, which is weaker than the inversion result and has poor stability. In terms of absolute error, the error of the inversion result is smaller than that of the SST product. The average value of the absolute value of the sample point error of the inversion result is 2.0806℃, the standard deviation is 1.4605℃, the maximum error is 4.4836℃, and the minimum error is 0.3495℃. The samples with absolute error within 1℃ account for 37.5%, and the absolute error is 2 The samples within ℃ accounted for 62.5%. The average value of the absolute value of the error of the MODIS SST product sample point is 2.6795℃, the standard deviation is 1.9745℃, the maximum error is 7.105℃, the minimum error is 0.3535℃, the samples with absolute error within 1℃ account for 25%, and the absolute error is 2℃ The samples within 50%. The verification results show that the inversion model proposed in this issue is feasible.

4. Application note of Shenzhen water quality evaluation system based on remote sensing data: In order to explore the feasibility of the evaluation system, Shenzhen Bay was selected as the experimental area for experimentation. According to the available data, the Shenzhen Bay area had large-scale land reclamation from 2000 to 2010. During the reclamation process, a large amount of sediment was discharged into the sea, which had a significant impact on the water quality of Shenzhen Bay. 2003-01-18 , 2005-01-23, 2007-01-29 and 2009-02-03 Shenzhen Bay TM images, extracted the water area, according to the method of this application to invert the area's chlorophyll a, suspended solids and sea surface temperature and other parameters, combined with this application The proposed water environment assessment model is scored. The higher the model score, the more serious the pollution. The result is shown in Figure 6.

Selecting the observational data of the Shenzhen Bay waters for many years in winter, it can be seen that the water quality in Shenzhen has significantly deteriorated with the reclamation. The water quality changes most drastically in the Shenzhen Bay and the sea area located in the northeast of Lingdingyang reclamation. The model evaluation results are consistent with the existing data, which proves that the model has good feasibility.

This application addresses the problem of insufficient universality of the existing inversion models and directly uses the inversion results in Shenzhen that the accuracy of the inversion results are not up to standard. Combined with the actual measurement data of the water quality monitoring buoy site in Shenzhen, the inversion of chlorophyll a, sea surface temperature and suspended solids concentration respectively Improve the performance model. The specific method is to analyze the correlation of each spectral band and its combination of Landsat 8 data with the measured values in the Shenzhen waters, select the most sensitive band combination to build a high-precision regression model. In terms of sea surface temperature, the atmospheric columnar water vapor content in the target area will have a significant impact on the inversion results, but it is often difficult to obtain high-quality data of this type, and most of the data in existing studies are of poor quality. This application is optimized to use the near-infrared water vapor secondary product of MODIS. The data uses the near-infrared band to obtain water vapor estimation. The data is of good quality and the spatial resolution is 1km, which is more suitable for research applications in offshore waters.

In this application, the mathematical analysis of 13 buoy sites in Shenzhen waters, statistics of the value range distribution of various water quality parameters in Shenzhen, to understand the water quality of Shenzhen waters in recent years and important water quality factors, and then borrow the comprehensive index method based on The characteristics of Shenzhen waters can be better explained by choosing evaluation indicators and factor weights. At the same time, we can learn from the design idea of time scale anomaly index to improve the model, and develop a water quality evaluation system based on Landsat 8 data suitable for Shenzhen waters.

Although this application has been described with reference to the current preferred embodiments, those skilled in the art should understand that the above preferred embodiments are only used to illustrate the application and are not used to limit the scope of protection of the application. Any modification, equivalent replacement, improvement, etc., made within the spirit and principle scope of this application shall be included in the protection scope of this application.

The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use this application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined in this document can be implemented in other embodiments without departing from the spirit or scope of the application. Therefore, this application will not be limited to the embodiments shown in this document, but should conform to the widest scope consistent with the principles and novel features disclosed in this document.

Claims

A method for evaluating water quality in Shenzhen sea area, characterized in that the method includes the following steps:

a. Preprocess Landsat 8 data, where Landsat 8 data is Landsat 8 satellite image data;

b. According to the pre-processed Landsat 8 data, develop the water quality parameter inversion model;

c. Based on the above-mentioned water quality parameter inversion model developed, develop the Shenzhen sea water quality evaluation model.
The method according to claim 1, wherein said step a specifically comprises:

Perform radiation correction on Landsat 8 data; perform atmospheric correction on Landsat 8 data after radiation correction; perform cloud removal processing on Landsat 8 data after atmospheric correction; perform water extraction on Landsat 8 data after cloud removal processing.
The method according to claim 2, wherein the water quality parameter inversion model comprises: a chlorophyll a concentration inversion model, a suspended matter concentration inversion model, and a sea surface temperature inversion model.
The method according to claim 3, wherein the development of the chlorophyll a concentration inversion model and the suspended matter concentration inversion model specifically comprises:

Filter and match the measured data and Landsat 8 data through data statistics and analysis; determine the sensitive band through normalized spectral analysis of the buoy point water area; perform correlation analysis on the sensitive band combination and the measured data to find the most relevant band combination; The inversion model was established by using the most relevant sensitive band, and the best model was screened through accuracy verification, and the inversion model of chlorophyll a concentration in Shenzhen sea area and the inversion model of suspended matter concentration were obtained.
The method according to claim 4, wherein the development of the sea surface temperature inversion model specifically comprises:

Screen and match the measured data and Landsat 8 data through data statistics and analysis; improve the universal single-channel algorithm: use the MODIS near-infrared water vapor secondary product MOD05 to obtain a more accurate water vapor content estimation, and combine the Shenzhen environmental parameters to model the model The correlation coefficient is fine-tuned; based on the above-mentioned improved model and the fine-tuned model correlation coefficient, the sea surface temperature inversion model of Shenzhen sea area is established.
The method according to claim 5, wherein the step c specifically comprises:

Through the inversion model of water quality parameters, the statistical analysis of the measured data and the Shenzhen water quality survey, select the concentration of chlorophyll a, the concentration of suspended solids and the sea surface temperature to analyze the water environment evaluation index of Shenzhen sea area;

Establish a water quality evaluation model for the Shenzhen sea area based on the Shenzhen sea water environment evaluation index: The water quality evaluation model for the Shenzhen sea area is based on the comprehensive index method, drawing on the design idea of the time scale anomaly index to select the concentration of chlorophyll a, the concentration of suspended matter and the sea surface Temperature is used as the evaluation factor to establish the water quality evaluation model of Shenzhen sea area.
A Shenzhen sea area water quality evaluation system, characterized in that the system includes a preprocessing module, an inversion model development module, and a water quality evaluation model development module, wherein:

The preprocessing module is used to preprocess the Landsat 8 data, where the Landsat 8 data is Landsat 8 satellite image data;

The inversion model development module is used to develop a water quality parameter inversion model based on the pre-processed Landsat 8 data;

The water quality evaluation model development module is used to develop a water quality evaluation model for Shenzhen sea area based on the developed water quality parameter inversion model.
The system according to claim 7, wherein the preprocessing module is specifically configured to:

Perform radiation correction on Landsat 8 data; perform atmospheric correction on Landsat 8 data after radiation correction; perform cloud removal processing on Landsat 8 data after atmospheric correction; perform water extraction on Landsat 8 data after cloud removal processing.
The system according to claim 8, wherein the water quality parameter inversion model comprises: a chlorophyll a concentration inversion model, a suspended matter concentration inversion model, and a sea surface temperature inversion model.
The system according to claim 9, wherein the development of the chlorophyll a concentration inversion model and the suspended matter concentration inversion model specifically comprises:

Filter and match the measured data and Landsat 8 data through data statistics and analysis; determine the sensitive band through normalized spectral analysis of the buoy point water area; perform correlation analysis on the sensitive band combination and the measured data to find the most relevant band combination; The inversion model was established by using the most relevant sensitive band, and the best model was screened through accuracy verification, and the inversion model of chlorophyll a concentration in Shenzhen sea area and the inversion model of suspended matter concentration were obtained.
The system according to claim 10, wherein the development of the sea surface temperature inversion model specifically comprises:

Screen and match the measured data and Landsat 8 data through data statistics and analysis; improve the universal single-channel algorithm: use the MODIS near-infrared water vapor secondary product MOD05 to obtain a more accurate water vapor content estimation, and combine the Shenzhen environmental parameters to model the model The correlation coefficient is fine-tuned; based on the above-mentioned improved model and the fine-tuned model correlation coefficient, the sea surface temperature inversion model of Shenzhen sea area is established.
The system according to claim 11, wherein the water quality evaluation model development module is specifically used for:

Through the inversion model of water quality parameters, the statistical analysis of the measured data and the Shenzhen water quality survey, select the concentration of chlorophyll a, the concentration of suspended solids and the sea surface temperature to analyze the water environment evaluation index of Shenzhen sea area;

Establish a water quality evaluation model for the Shenzhen sea area based on the Shenzhen sea water environment evaluation index: The water quality evaluation model for the Shenzhen sea area is based on the comprehensive index method, drawing on the design idea of the time scale anomaly index to select the concentration of chlorophyll a, the concentration of suspended matter and the sea surface Temperature is used as the evaluation factor to establish the water quality evaluation model of Shenzhen sea area.