Water transparency estimation method based on OLCI data

Abstract

The invention discloses a water transparency estimation method based on OLCI data, and belongs to the field of environmental science and remote sensing monitoring. The method comprises the following steps: S1, data acquisition: setting a plurality of sampling points in a target area, and obtaining a water transparency measured value; s2, data preprocessing: obtaining water-leaving reflectivity data, and eliminating external interference factors; s3, constructing an estimation model: constructing a transparency remote sensing estimation model in a wave band combination form based on the off-water reflectivity characteristics of the Sentinel-3OLCI sensor; and S4, model verification: carrying out linear fitting on an estimated value and a measured value. According to the method, the effect of estimating the transparency of the inland water body is good, the model precision is enough to support subsequent research through verification of an actual measurement sample point model, and the method has extremely high accuracy and applicability.

Description

Technical Field

[0001] The present invention belongs to the fields of environmental science and remote sensing monitoring, and particularly relates to a method for estimating water transparency based on OLCI data. Background Art

[0002] As an important freshwater resource reservoir, flood regulation reservoir, and species gene pool on the earth, lakes play an irreplaceable role in maintaining the ecological balance of the basin, ensuring the water supply for production and living, reducing flood disasters, and providing rich aquatic products. The monitoring of lake water quality is of great significance for the management and restoration of the ecological environment. Water transparency (Secchi disk depth, SDD) is used to characterize the turbidity of water bodies and is an essential and important indicator in water quality assessments such as lake eutrophication level, water quality quality, and primary productivity.

[0003] The size of transparency depends on the concentration of various plankton and suspended solids in the water. Empirical methods use the correlation between remote sensing data and measured data for regression analysis to construct a transparency inversion model. This method is easy to implement and conforms to the optical properties of the water bodies in the target lake, enabling long-term and large-scale monitoring of lake transparency.

[0004] Sentinel-3 is an observation satellite under the Copernicus program of the European Space Agency. It consists of a satellite constellation of two identical satellites, 3A and 3B. The dual satellites flying in the same orbit well expand the coverage area and shorten the revisit time. The Ocean and Land Colour Instrument (OLCI) carried by it can provide high-quality ocean and land surface images, comprehensively covering land and ocean with a spatial resolution of 300 meters, including 21 spectral bands with central wavelengths between 400 - 1020 nm, featuring high accuracy and reliability.

[0005] Existing transparency inversion models are usually constructed based on measured data of specific lakes, lacking universality and being difficult to adapt to the optical complexity of lakes of different regional types. Moreover, there is currently no research on remotely sensing the estimation of water transparency using Sentinel 3-OLCI data. Summary of the Invention

[0006] In view of the problems mentioned in the above background art, the present invention proposes a method for estimating water transparency based on OLCI data, which combines Sentinel-3 and can make up for the problem that it is difficult to monitor the transparency of inland water bodies in the long term in the prior art.

[0007] A method for estimating water transparency based on OLCI data includes the following steps:

[0008] S1: Data collection: Set multiple sampling points in the target area to obtain the measured values of water transparency;

[0009] S2: Data preprocessing: Obtain the water-leaving reflectance data and eliminate interference factors;

[0010] S3: Construct an estimation model: Based on the water-leaving reflectance characteristics of the Sentinel-3 OLCI sensor, construct a remote sensing estimation model of water transparency in the form of a band combination;

[0011] S4: Model verification: Perform linear fitting on the estimated values and the measured values.

[0012] Preferably, in S1, based on multiple sampling points, measure the transparency of inland water bodies in past years.

[0013] Preferably, in S2, eliminate the interference factors of clouds, aerosols, and land in the data.

[0014] Preferably, in S3, use the water-leaving reflectance characteristic combination model of Sentinel-3 OLCI to estimate the water transparency. The water-leaving reflectance characteristic combination model is the ratio of the water-leaving reflectance Oa6 of the 6th band of the OLCI sensor to the water-leaving reflectance Oa11 of the 11th band. Substitute this ratio into the following formula to estimate the water transparency:

[0015] SDD = 0.6374×ln(Oa6 / Oa11) + 0.0051,

[0016] where SDD represents the water transparency.

[0017] Preferably, select the bands with the highest correlation to construct the model according to the Pearson correlation coefficient. The calculation formula of the Pearson correlation coefficient is:

[0018]

[0019] where Γ represents the correlation coefficient, x i and y i respectively represent the remote sensing reflectance and transparency data, and respectively represent the means of the remote sensing reflectance and transparency.

[0020] Preferably, in S4, perform linear fitting on the estimated values and the measured values. The fitting linear equation is:

[0021] y = 0.7804x + 0.0777,

[0022] R 2 = 0.84,

[0023] Use MAPE and RMSE to verify the accuracy of the estimation method, specifically:

[0024]

[0025] Among them, y i represents the measured value of transparency, and y’ i represents the remote sensing estimation result, and n represents the number of samples.

[0026] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0027] (1) The method of the present invention can realize the estimation of the transparency of inland waters. For the first time, a remote sensing estimation model for the transparency of Hongze Lake is constructed based on Sentinel-3 OLCI data. By obtaining the water-leaving reflectance data of the 6th and 11th bands of the satellite sensor, an empirical model is constructed to estimate the transparency value of the target water body, quickly understand the spatio-temporal distribution characteristics of the water transparency in a large-scale range, and realize the long-term high-precision monitoring of the transparency of inland waters. The present invention not only fills the application gap of Sentinel-3 OLCI data in the field of estimating the transparency of Hongze Lake water body, but also helps water quality monitoring and evaluation, and has important significance for the ecological restoration of water environment.

[0028] (2) The effect of estimating the transparency of inland waters by the present invention can meet the research needs in this field. After being verified by the measured sample points model, the model has high accuracy in estimating the transparency of inland waters. Compared with the traditional semi-analytical model and machine learning model, the empirical model constructed by the present invention has higher accuracy and operability. The semi-analytical model usually requires a complex parameterization process and a large amount of measured data support, while although the machine learning model can handle non-linear relationships, its "black box" characteristic leads to poor model interpretability and high requirements for data quality and quantity. In contrast, the empirical model based on Sentinel-3 OLCI data of the present invention is simple and easy to implement, and can make full use of the advantages of satellite data, significantly improving the efficiency and practicality of transparency estimation. Description of the drawings

[0029] Figure 1 is the sampling distribution map of the method for estimating the water transparency based on OLCI data of the present invention;

[0030] Figure 2 is the schematic diagram of the remote sensing estimation model of transparency constructed by 112 sample points in the method for estimating the water transparency based on OLCI data of the present invention;

[0031] Figure 3 is the schematic diagram of the remote sensing estimation model of transparency constructed by verifying 28 sample points in the method for estimating the water transparency based on OLCI data of the present invention;

[0032] Figure 4 is the spatial variation map of the water transparency in spring of Hongze Lake water body inverted according to the model of the present invention;

[0033] Figure 5 The spatial variation map of the summer transparency of the Hongze Lake water body is inversed based on the model of the present invention;

[0034] Figure 6 The spatial variation map of the autumn transparency of the Hongze Lake water body is inversed based on the model of the present invention;

[0035] Figure 7 The spatial variation map of the winter transparency of the Hongze Lake water body is inversed based on the model of the present invention. Specific embodiments

[0036] To further understand the content of the present invention, the technical solution of the present invention will be clearly and completely described below in conjunction with the specific embodiments of the present invention and the corresponding drawings. It should be understood that the embodiments are only for explaining the present invention and not for limiting it.

[0037] The method for estimating the water body transparency based on OLCI data provided in this embodiment uses the water quality spectra in-situ collected by an ASD ground object spectrometer and the measured water body transparency data to construct a remote sensing estimation model for the transparency applicable to the Hongze Lake, realizing high-precision real-time estimation of the transparency. The method of this embodiment is aimed at inland turbid water bodies represented by the Hongze Lake to realize long-time series continuous monitoring of the transparency.

[0038] The specific implementation steps are as follows:

[0039] S1: Data collection: Set multiple sampling points in the target area to obtain the measured values of the water body transparency;

[0040] During the period from 2018 to 2023, the water body transparency of the Hongze Lake was measured, and a total of 169 sampling points were set in the Hongze Lake. The distribution of the sampling points is as Figure 1 shown. The specific sampling time and quantity are shown in Table 1.

[0041] Table 1 Measured data

[0042] Sampling time Number of sampling points 2023.11 32 2021.11 27 2021.03 30 2020.11 29 2019.04 29 2018.09 22

[0043] S2: Data preprocessing: Obtain the water-leaving reflectance data and eliminate interference factors;

[0044] Obtain the water-leaving reflectance data of the Sentinel-3 OLCI sensor. The original image data can be downloaded from the website https: / / scihub.copernicus.eu / (European Space Agency Copernicus Data Center). During the data acquisition process, eliminate interference factors such as clouds, aerosols, and land to ensure the accuracy of the data.

[0045] In this embodiment, cloud and other interferences are mainly excluded through atmospheric correction, and the atmospheric correction is processed using the SNAP (The Sentinel Application Platform) application platform developed by ESA.

[0046] S3: Construct an estimation model: Based on the characteristics of the water-leaving reflectance of the Sentinel-3 OLCI sensor, construct a remote sensing estimation model of transparency in the form of a band combination.

[0047] The water-leaving reflectance characteristic combination model of the Sentinel-3 OLCI sensor is used to estimate the transparency of inland waters. The water-leaving reflectance combination model is the ratio of Oa6 to Oa11. The central wavelength of Oa6 is 560 nm, and the central wavelength of Oa11 is 709 nm.

[0048] After taking the natural logarithm of this ratio, it is linearly fitted with the measured transparency data. The specific fitting formula is:

[0049] SDD = 0.6374×ln(Oa6 / Oa11)+0.0051,

[0050] R 2 = 0.81,

[0051] Among them, Oa6 represents the water-leaving reflectance data of the 6th band of the OLCI sensor, and Oa11 represents the water-leaving reflectance data of the 11th band of the OLCI sensor; SDD represents the water transparency; R 2 represents the coefficient of determination.

[0052] In this embodiment, the fitting curve is as Figure 2 shown.

[0053] In the SDD calculation formula of this embodiment, 0.6374 and 0.0051 are obtained by linearly fitting the measured remote sensing reflectance data and transparency data.

[0054] The band selection is obtained based on the Pearson correlation coefficient. The Pearson correlation coefficient is calculated for the remote sensing reflectance and transparency data of each band, and the bands with the highest correlation, namely Oa6 and Oa11, are selected to participate in the construction of the model. The formula for the Pearson correlation coefficient is as follows:

[0055]

[0056] Among them, Γ represents the correlation coefficient, x i and y i represent the remote sensing reflectance and transparency data respectively, and are the means of the remote sensing reflectance and transparency respectively.

[0057] S4: Model Verification: Perform linear fitting on the estimated values and the measured values.

[0058] The estimated value of the transparency of inland water bodies is obtained. Perform linear fitting on the estimated value and the measured value, as Figure 3 shown. The results show that the measured value data points are evenly distributed on both sides of the 1:1 line. The fitted linear equation is:

[0059] y = 0.7804x + 0.0777,

[0060] R 2 = 0.84,

[0061] where x represents the measured transparency and y represents the estimated transparency.

[0062] Use MAPE (Mean Absolute Percentage Error) and RMSE (Root Mean Square Error) to verify the accuracy of the estimation method. Specifically:

[0063]

[0064] where y i represents the measured value of transparency, y' i represents the remote sensing estimation result, and n represents the number of samples.

[0065] In this embodiment, MAPE and RMSE are 21.33% and 0.07 m respectively.

[0066] The estimated water body transparency according to the present invention is close to the measured transparency, and the accuracy meets the requirements, with extremely high accuracy and applicability. As Figure 4 shown, the spatio-temporal variation of the transparency of Hongze Lake water body in different seasons is inversely calculated based on the model of the present invention. It can be seen that the transparency is relatively high in spring and relatively low in autumn. The estimation result conforms to the actual change trend and can be used as an effective method for estimating the transparency of Hongze Lake water body.

[0067] For the experimental result data, compare the present application with the existing models. Specifically, see Table 2 below.

[0068] Table 2 Model Comparison

[0069] Model MAPE RMSE <![CDATA[SD = 10 -0.01096-19.57*Oa6+0.2214*Oa6 / Oa7 > 95.07％ 0.24m <![CDATA[SD=e 1.638*Oa4 / Oa6-8.241*Oa4-12.876*Oa6+1.26 > 4978.56％ 5.54m SD = 0.036 * (Oa4 / (Oa6 * Oa8)) + 0.79 880.56％ 1.05m SD = 0.6374ln(Oa6 / Oa11) + 0.0051 (This invention) 21.33％ 0.07m

[0070] As shown in Table 2, when applying the existing model method to Hongze Lake, the resulting accuracy is difficult to meet the estimation requirements, and the accuracy of the present invention is much higher than that of other solutions. Therefore, the effect of estimating the transparency of inland water bodies by the present invention can meet the research needs in this field. After model comparison and verification, the accuracy of this model for estimating the transparency of inland water bodies is high.

[0071] An electronic device includes a processor and a memory storing a computer program. When the processor executes the computer program, the steps of the above-mentioned method for estimating water transparency based on Sentinel-3 OLCI are implemented.

[0072] The processor can be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logical blocks, modules, and circuits described in connection with the disclosure of this application. The processor can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0073] The memory can be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or it can also be an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory), or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), magnetic disk storage media, or other magnetic storage devices, or any other medium that can be used to carry or store the desired computer program in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

[0074] A computer-readable storage medium stores a computer program or instructions. When the computer program or instructions are executed by a processor, the steps of the method for estimating water transparency based on Sentinel-3 OLCI are implemented.

[0075] The above-mentioned computer-readable storage medium provided by the present application includes, but is not limited to, any type of disk (including floppy disks, hard disks, optical disks, CD-ROMs, and magneto-optical disks), ROM, RAM, EPROM (Erasable Programmable Read-Only Memory), EEPROM, flash memory, magnetic cards, or optical cards. That is, the readable medium includes any medium that stores or transmits information in a form readable by a device (e.g., a computer).

[0076] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A method for estimating water transparency based on OLCI data, characterized in that: Including the following steps: S1: Data collection: Set multiple sampling points in the target area to obtain the measured values of water transparency; S2: Data preprocessing: Obtain the water-leaving reflectance data and eliminate interference factors; S3: Construct an estimation model: Based on the water-leaving reflectance characteristics of the Sentinel-3 OLCI sensor, construct a remote sensing estimation model of transparency in the form of a band combination; S4: Model verification: Perform linear fitting on the estimated values and the measured values.

2. The method for estimating water transparency based on OLCI data according to claim 1, wherein: In S1, based on multiple sampling points, measure the transparency of inland water bodies in past years.

3. The method for estimating water transparency based on OLCI data according to claim 1, characterized in that: In S2, eliminate the interference factors of clouds, aerosols, and land in the data.

4. The method for estimating water transparency based on OLCI data according to claim 1, characterized in that: In S3, use the water-leaving reflectance characteristic combination model of Sentinel-3 OLCI to estimate the water transparency. The water-leaving reflectance characteristic combination model is the ratio of the water-leaving reflectance Oa6 of the 6th band of the OLCI sensor to the water-leaving reflectance Oa11 of the 11th band. Substitute this ratio into the following formula to estimate the water transparency: SDD = 0.6374 × ln(Oa6 / Oa11) + 0.0051, where SDD represents the water transparency.

5. The method for estimating water transparency based on OLCI data according to claim 4, characterized in that: Select the bands with the highest correlation to construct the model according to the Pearson correlation coefficient. The calculation formula of the Pearson correlation coefficient is: Among them, Γ represents the correlation coefficient, x i and y i respectively represent remote sensing reflectance and transparency data, and respectively represent the mean values of remote sensing reflectance and transparency.

6. The method for estimating water transparency based on OLCI data according to claim 1, wherein: In S4, perform linear fitting on the estimated values and the measured values. The fitting linear equation is: y = 0.7804x + 0.0777, R 2 ＝0.84， Use MAPE and RMSE to verify the accuracy of the estimation method, specifically: Among them, y i represents the measured transparency value, y' i represents the remotely sensed estimation result, and n represents the number of samples.

Images