Water body dissolved oxygen concentration remote sensing inversion method based on XGBoost-2DCNN model

Abstract

The invention relates to a water body dissolved oxygen concentration remote sensing inversion method based on an XGBoost-2DCNN model. The method comprises the following steps: acquiring or collecting long-time sequence actually measured water quality parameter concentration of a research area; the method comprises the following steps: acquiring a multispectral remote sensing image of a research area based on a Google Earth Engine platform, and extracting space-time matched water surface remote sensing reflectivity; carrying out sine and cosine function coding on the reflectivity spectrum and the water quality parameter concentration periodic characteristics, and estimating chlorophyll a and water temperature through remote sensing reflectivity by utilizing an XGBoost model; and further developing a two-dimensional convolutional neural network (2DCNN) model, and constructing and training a dissolved oxygen concentration inversion model by taking the chlorophyll a, the water temperature and the periodic characteristics of the codes as input. On the basis of water quality parameters actually measured for many years and optical remote sensing images, a precise dissolved oxygen estimation method is developed in combination with a time period mode of chlorophyll a and water temperature, and short-term and long-term space-time dynamic analysis is realized. The method provides a scientific basis for prevention and control of water eutrophication, and has wide popularization and application values.

Description

(I) Technical field

[0001] The present invention relates to a water body dissolved oxygen concentration remote sensing inversion method based on an XGBoost-2DCNN model, which belongs to the field of water color remote sensing and is of great significance in water ecosystem research and water quality monitoring applications. (II) Background technology

[0002] The concentration of dissolved oxygen (DO) in water is one of the important parameters for measuring water quality and is closely related to the health of aquatic ecosystems and environmental monitoring. Abnormal fluctuations in DO concentration may cause ecological problems such as eutrophication of water bodies and fish deaths, so it is of great significance to monitor DO efficiently and accurately. Traditional DO monitoring methods mainly rely on ground sampling and laboratory analysis. Although they have high accuracy, they have the limitations of limited temporal and spatial coverage, high cost, and are not suitable for large-scale dynamic monitoring.

[0003] With the development of remote sensing technology, the inversion of water quality parameters using satellite image data has become an efficient and economical solution. However, the optical properties of water bodies are complex and changeable, and dissolved oxygen concentration is difficult to invert directly through spectral characteristics. Therefore, how to use indirect information such as optically active substances in water bodies (such as chlorophyll-a, suspended matter, etc.) and water temperature to achieve high-precision estimation of dissolved oxygen through remote sensing images combined with advanced mathematical and physical modeling methods has become the focus and difficulty of current research.

[0004] In recent years, machine learning and deep learning technologies have shown significant advantages in the field of remote sensing water quality inversion. Among them, the XGBoost model can efficiently process nonlinear and multidimensional features, while the two-dimensional convolutional neural network (2DCNN) has outstanding capabilities in mining the spatiotemporal characteristics and high-dimensional associations of data. Based on this, a joint model integrating XGBoost and 2DCNN is constructed by combining remote sensing image data and multi-source water quality data. It can achieve high-precision and spatiotemporal dynamic estimation of dissolved oxygen concentration while making full use of the advantages of remote sensing data. The development and promotion of this technology has important scientific value and practical significance for water environment monitoring and eutrophication prevention and control. (III) Summary of the invention

[0005] 1 A remote sensing inversion method for dissolved oxygen concentration in water based on the XGBoost-2DCNN model, characterized by comprising the following steps:

[0006] (1) Collect, process and standardize data for water dissolved oxygen (DO) concentration model training;

[0007] (2) Construct a water chlorophyll a concentration and water temperature prediction model based on the XGBoost machine learning model;

[0008] (3) Develop a two-dimensional convolutional neural network model (2DCNN) to predict dissolved oxygen concentration;

[0009] (4) Verification of the accuracy of the dissolved oxygen concentration prediction model.

[0010] The specific methods for collecting, processing and standardizing data for predictive model training are as follows:

[0011] (1) Collect and collect long-term in situ water quality measurement data from the study area, including various optically active and inactive water quality parameters, as well as metadata such as the geographical coordinates of the water quality monitoring station, data quality, and collection date; extract necessary data from the collected data sets, including water quality control area, monitoring station code, geographical coordinates of the monitoring station, sampling date, chlorophyll a, water temperature, and dissolved oxygen (DO); compile AOI polygon vector files of the area of ​​interest in the study area; establish a spatiotemporal matching data set containing in situ water quality parameter measurements and corresponding satellite multispectral image water surface reflectance values; use Google Earth Engine (GEE) to extract the water surface reflectance of each water quality monitoring station and sample collection date;

[0012] (2) Incorporate the seasonal variation trend of water quality parameters into the modeling of dissolved oxygen concentration prediction; use the time information—month as the input parameter of the modeling, and encode the month into a periodic variable through sine and cosine transformation, as shown below:

[0013]

[0014] where x represents the month; this transformation ensures that all months are correctly positioned on the circular scale and accurately represents their cycle pattern by capturing the cycle angle and amplitude; the converted month feature is denoted as Month sin and Month cos , which will be used as input features for the dissolved oxygen prediction model;

[0015] (3) Extract data records of the research period from the prepared data set to prepare two training data sets; one training data set contains input features: including spectral reflectance data and converted monthly features and output features: including chlorophyll a and water temperature, which are used to train the machine learning algorithm for univariate prediction of chlorophyll a and water temperature; taking the optical remote sensing image Sentinel-2 as an example, the data set can be expressed as:

[0016] Training_Dataset Chla_Temperature = {B1, B2, …, B8, B 8A ,B9,B 11 ,B 12 ,Month sin ,Month cos ,Chla,Temperature}

[0017] Another training dataset contains input features including chlorophyll a, water temperature, and transition month features, as well as annotated dissolved oxygen output features, which are used to train a convolutional neural network model for univariate prediction of dissolved oxygen; this dataset is represented as

[0018] Training_Dataset DO ={Chla,Temperature,Month sin ,Month cos ,DO}

[0019] (4) Extract the data records of the research period from the prepared data set to prepare the test data set. The data set contains input features including reflectance data and converted monthly features. The first stage is used to predict chlorophyll a and water temperature, and then the prediction is used to predict dissolved oxygen; the data set is represented as

[0020] Test_dataset={B1,B2,…,B8,B 8A ,B9,B 11 ,B 12 ,Month sin ,Month cos}

[0021] The specific implementation method of building a water body chlorophyll a concentration and water temperature prediction model based on the XGBoost machine learning model is as follows:

[0022] Based on the machine learning algorithm eXtreme Gradient Boosting (XGBoost), the complex relationship between water quality parameters and satellite optical characteristics is captured for water quality monitoring; using Training_Dataset Chla_Temperature The dataset was used to train the XGBoost model using the optimal hyperparameters to estimate chlorophyll a and water temperature; for each target variable, the training dataset was split into input features and target output as follows:

[0023] X Chla_Temperature =Training_Dataset Chla_Temperature \{Chla,Temperature}

[0024] Y Chla =Training_Dtaset Chla_Temperature ['Chla']

[0025] Y Temperature =Training_Dataset Chla_Temperature ['Temperature']

[0026] For each parameter model, the XGBoost algorithm is used for training, and its objective function is to minimize the prediction error, which is specifically expressed as follows:

[0027] f XGBOOST,Chla (X Chla_Temperature )=arg min∑ i (Y Chla,i -f(X Chla_Temperature,i )) 2

[0028]

[0029] Among them, i represents the i-th sample value, f(X) represents the function of the model to predict chlorophyll a or water temperature based on the input features, and the goal is to minimize the sum of the squared errors between the predicted value and the actual value.

[0030] The specific implementation method of developing a two-dimensional convolutional neural network model 2DCNN to predict dissolved oxygen concentration is as follows:

[0031] A two-dimensional convolutional neural network (2DCNN) model was developed and trained using input parameters such as chlorophyll a, water temperature, and transition month features to estimate dissolved oxygen concentration; the 2DCNN model was trained using optimal hyperparameters to minimize training and validation losses; the steps for training and developing the 2DCNN model were as follows:

[0032] (1) Train the training data set Training_Dataset DO Divided into input features and target output:

[0033] X DO =Training_Dataset DO \{DO}

[0034] Y DO =Training_Dataset DO ['DO']

[0035] (2) Input feature vector X DO ∈R 4×1 It is represented by 4 features: Chla, Temperature, Month sin ,Month cos One-dimensional data consisting of one value for each feature:

[0036] X′ DO ∈R 4×1×1

[0037] (3) 13 layers are defined with different hyperparameters to perform operations such as convolution, pooling, and batch normalization on the input features to generate the desired output. The definitions of these layers are as follows:

[0038] Layer 1 => Conv2D: 32 convolutional filters W1 of size 4×1 with “same” padding and hyperbolic tangent activation function applied to the input features defined as follows:

[0039] Z1=Conv2D(X′ DO ,W1)+b1

[0040] A1=tanh(Z1)

[0041] Layer 2 => Batch Normalization: Normalize the activation function output A1 to improve the stability of training. It is defined as follows

[0042]

[0043] Among them, μ1 and Respectively represent the mean and variance of A1, γ1 and β1 represent the learnable scaling and shifting parameters, ∈ is a very small number to avoid the denominator of the formula being 0, b1 is the model bias value, and the meanings of similar parameters in other subsequent layers are similar, and so on;

[0044] Layer 3 => MaxPooling2D: applies max pooling to the batch normalized output with a pooling window of 4×1 and a stride of 1, denoted as

[0045]

[0046] Layer 4 => Conv2D: 64 convolutional filters W2 of size 3×1, with “same” padding and hyperbolic tangent activation function as follows:

[0047] Z2=Conv2D(P1,W2)+b2

[0048] A2=tanh(Z2)

[0049] Layer 5 => Batch Normalization: Normalize the activation function output A2 to improve the stability of training. It is defined as follows

[0050]

[0051] Layer 6 => MaxPooling2D: applies max pooling to batch normalized output with a pooling window of 2×1 and a stride of 1, denoted as

[0052]

[0053] Layer 7 => Conv2D: 128 convolutional filters W3 of size 2×1, with “same” padding, and hyperbolic tangent activation function as follows:

[0054] Z3=Conv2D(P2,W3)+b3

[0055] A3=tanh(Z3)

[0056] Layer 8 => Batch Normalization: Normalize the activation function output A3 to improve the stability of training. It is defined as follows

[0057]

[0058] Layer 9 => MaxPooling2D: applies max pooling to batch normalized output with a pooling window of 2×1 and a stride of 1, denoted as

[0059]

[0060] Layer 10 => Conv2D: 256 convolutional filters W4 of size 1×1, with “same” padding, and hyperbolic tangent activation function as follows:

[0061] Z4=Conv2D(P3,W4)+b4

[0062] A4=tanh(Z4)

[0063] Layer 11 => Batch Normalization: Normalize the activation function output A4 to improve the stability of training. It is defined as follows

[0064]

[0065] Layer 12 => Dropout: A dropout rate of 0.3 is used to drop out to reduce overfitting;

[0066]

[0067] Layer 13 => Flatten: Flattens the 4D tensor D4 received after the filtering operation into a 1D vector, which is represented as

[0068] F=Flatten(D4)

[0069] (4) The output layer is defined as a fully connected layer with one unit, and the linear activation of the regression output is expressed as

[0070]

[0071] Among them, W o is the weight vector, bo is the bias term, · represents the dot product operation;

[0072] (5) The model integrated with the above layers consists of a loss function and an Adam optimizer. The loss function is defined by the mean square error (MSE) loss and is used to calculate the predicted output. The difference between the actual value y; the Adam optimizer uses the gradient of the loss function to update the model parameters (W, b), and the initial learning rate (lr) is 0.0001; the loss function and optimization are defined as follows;

[0073]

[0074] Among them, W (t) ,b (t) Respectively represent the weight and bias term in the current iteration state, W (t+1) ,b (t+1) They represent the weight and bias terms updated in the next iterative calculation process, respectively. The gradient of the loss function; lr represents the learning rate; t represents the sequence number of iterative calculations.

[0075] (6) Train the model by setting the training configuration, including defining the minimum batch size (B = 32), EPOCHS (E = 500) for 500 iterations over the entire dataset, the validation dataset, and the early stopping criterion based on restoring the best weights when the validation performance stops improving. The model is saved and available for use with the function f 2DCNN express.

[0076] The specific implementation method of the dissolved oxygen concentration prediction model accuracy verification is as follows:

[0077] The dissolved oxygen concentration prediction model is verified by the test data set. First, the input features (B1, B2, …, B8, B 8A ,B9,B 11 ,B 12 ,Month sin ,Month cos ) predicts chlorophyll a and water temperature values, defined as follows:

[0078]

[0079]

[0080] in, represents the predicted chlorophyll a concentration value, Represents the predicted water temperature value; these prediction parameters are combined with the time feature Month sin ,Month cosConnect them together and prepare the input dataset for the 2DCNN model to predict dissolved oxygen as follows:

[0081]

[0082] This dataset is used as f 2DCNN Inputs to the model, used to predict dissolved oxygen values;

[0083]

[0084] in, represents the predicted dissolved oxygen value;

[0085] 2Finally, the accuracy of the model was evaluated using regression accuracy indicators, including Pearson correlation coefficient r, root mean square error RMSE, mean absolute error MAE, mean absolute percentage error MAPE and unbiased mean absolute percentage error UMAPE to predict dissolved oxygen values. and the actual dissolved oxygen value y DO Make a comparison.

[0086] The advantages of the present invention compared with the prior art are:

[0087] (1) High-precision and multi-dimensional feature fusion

[0088] The present invention combines the nonlinear feature processing capability of the XGBoost model with the spatiotemporal feature mining capability of the 2DCNN model, and indirectly estimates the dissolved oxygen concentration of water bodies through remote sensing reflectivity, which can effectively improve the estimation accuracy of the model and overcome the error accumulation problem caused by single variable inversion in traditional methods.

[0089] (2) Widespread applicability in time and space

[0090] The present invention utilizes satellite multispectral image data, combined with chlorophyll-a, temperature and periodic characteristics, to realize short-term and long-term spatiotemporal dynamic analysis of dissolved oxygen concentration, breaking through the limitations of existing technologies in time continuity and spatial coverage, and providing a scientific basis for large-scale dynamic monitoring of water environment. (IV) Description of the drawings

[0091] Figure 1 This is the technical process of the present invention.

[0092] Figure 2 Time series characteristics of chlorophyll a, water temperature and dissolved oxygen.

[0093] Figure 3 This is the spatiotemporal distribution map of dissolved oxygen concentration in water bodies in Hong Kong waters predicted by the method of the present invention. (V) Specific implementation methods

[0094] In order to better illustrate the remote sensing inversion method of dissolved oxygen concentration in water based on the XGBoost-2DCNN model, the present invention was used to test and simulate the selected research area, and good results were achieved. The specific implementation method is as follows:

[0095] The present invention involves the collection of five years (2019-2023) of water quality data from 94 monitoring stations along the coast of Hong Kong, and the extraction of the corresponding Sentinel-2 satellite surface reflectivity. A machine learning model based on the XGBoost algorithm was developed and trained using surface reflectivity and periodic monthly features encoded by sine and cosine functions to predict Chl-a and temperature. A two-dimensional convolutional neural network (2DCNN) model was then developed and trained using a training dataset (2019-2022) consisting of Chl-a, temperature, and encoded periodic monthly features (2019-2022) to estimate dissolved oxygen. The present invention introduces the XGBoost-2DCNN inversion method for estimating dissolved oxygen from satellite remote sensing reflectivity, which can be used for short-term and long-term spatiotemporal analysis.

[0096] Step 1: Extract water surface reflectance and prepare training and test datasets

[0097] (1) Taking the ESA Sentinel-2 satellite as an example, the 2A level images of the COPERNICUS / S2_SR dataset are obtained from the GEE platform. This dataset contains the surface reflectance value and auxiliary information of each pixel of the preprocessed and atmospherically corrected Sentinel-2 satellite image. In order to extract Level-2A images based on the area of ​​interest (AOI) and a specific date, it is necessary to define a spatial polygon / bounding box and a date range to filter the image set. The definition of the area of ​​interest is as follows:

[0098] AOI={(x,y)∈R 2 ∣x∈[x min ,x max ],y∈[y min ,y max ]}

[0099] Where (x, y) is the coordinate in the defined spatial reference system, x min ,x max ,y min ,y max Represents the AOI bounding box limits in minimum and maximum longitude and latitude, respectively.

[0100] For each target date D, create a date range D range , centered on D, spanning from D-5 days to D+5 days:

[0101] D range={D-5,D-4,D-3,…,D,…,D+3,D+4,D+5}

[0102] The collected images are based on AOI and D range Perform screening to obtain a set of Sentinel-2 images I within a specified time and space range as shown below:

[0103]

[0104] Among them, I j Represents each image in the filter set; Sentinel-2SR represents the Sentinel-2Level-2A surface reflectance image dataset; D i Indicates the date of the target image, in D range within the scope; Represents image I j Intersects with the specified AOI.

[0105] (2) Extract the surface reflectance value and auxiliary information from each image in the image dataset I and compare them with each monitoring station. The specific definitions are as follows.

[0106] P={P i |P i ∈(x i ,y i )}

[0107] Among them, P i is the i-th point in the monitoring station vector file, (x i ,y i ) represents P i The latitude and longitude coordinates of the point.

[0108] Each P i is a sampling point where each image I can be extracted j B1 to B 12 Band (excluding B 10 band) and the scene classification layer (SCL) band (which can classify pixels into clear land, water, cloud, etc.). The pixel coordinate (x i ,y i ) at band B k The reflectivity can be used k (I j ) represents, where k∈{1,2,...,8,8A,9,11,12}. For I j Each pixel position (x i ,y i ), extract the reflectance data of each band and divide it by a scale factor of 10,000. The extracted reflectance data includes the following information;

[0109]

[0110] Likewise, I j Each pixel position (x i ,y i )’s SCL classification is annotated using the following category types and formula (6);

[0111]

[0112] is_cloud: When When is the cloud-covered pixel;

[0113] is_cloud_shadow:When When , it is the pixel in the cloud shadow;

[0114] is_cirrus: When When , it is a cirrus cloud pixel;

[0115] is_snow_ice: When When , it is a pixel covered by snow or ice;

[0116] is_clear: When At this time, it is a clear pixel.

[0117] I j Medium i Each data record is accompanied by a Date(I j ) represents the image acquisition date and reference target date. The final output of each point contains the reflectance value of each band in each image, the category type obtained by SCL classification, the image acquisition date and the target reference date. The mathematical expression is as follows;

[0118]

[0119] Records with scene classification type “is_clear” and duplicate records were removed, while records with a smaller difference in days between the satellite image acquisition date and the target reference date (sample collection date) were retained. In addition, the extracted in situ water quality data were linked to these records based on the monitoring station code and the target reference date. Necessary fields were retained in the dataset, including water body control area, monitoring station code, monitoring station geographic coordinates, band reflectance value, sample collection date, Chl-a, temperature, and dissolved oxygen.

[0120] (3) Extract the records from 2019 to 2022 from the final dataset to prepare two training datasets. One training dataset contains input features (including reflectance data and converted month features) and output features (including Chl-a and temperature) and is used to train a machine learning algorithm for univariate prediction of Chl-a and temperature. The dataset is represented as

[0121] Training_Dataset Chla_Temperature =

[0122] {B1,B2,…,B8,B 8A ,B9,B 11 ,B 12 ,Month sin ,Month cos ,Chl-a,Temperature}

[0123] The other training datasets contain input features, including Chl-a, temperature, and transition month features, as well as annotated dissolved oxygen output features, which are used to train the convolutional neural network model for univariate prediction of dissolved oxygen. This dataset is represented as

[0124] Training_Dataset DO ={Chl-a,Temperature,Month sin ,Month cos ,DO}

[0125] (4) The records of 2023 are extracted from the final dataset to prepare the test dataset, which contains input features, including reflectance data and converted month features. The first stage is used to predict Chl-a and temperature, and then the prediction parameters are used to predict dissolved oxygen. The dataset is represented as

[0126] Test_Dataset={B1,B2,…,B8,B8A,B9,B11,B12,Month sin ,Month cos}

[0127] Step 2: Chl-a and temperature modeling

[0128] Based on the machine learning algorithm eXtreme Gradient Boosting (XGBoost), the complex relationship between water quality parameters and satellite optical characteristics is captured for water quality monitoring; using Training_Dataset Chla_TemperatureThe XGBoost model was trained with the optimal hyperparameters to estimate chlorophyll a and water temperature. For each target variable, the training dataset was divided into input features and target output. For each parameter model, the XGBoost algorithm was used for training, and its objective function was to minimize the prediction error.

[0129] Step 3: Dissolved oxygen concentration prediction modeling

[0130] A two-dimensional convolutional neural network (2DCNN) model was developed and trained using input parameters such as chlorophyll a, water temperature, and transition month features to estimate dissolved oxygen; the 2DCNN model was trained using optimal hyperparameters to minimize training and validation losses; the steps for training and developing the 2DCNN model were as follows:

[0131] (1) Train the training data set Training_Dataset DO Divided into input features and target output:

[0132] X DO =Training_Dataset DO \{DO}

[0133] Y DO =Training_Dataset DO ['DO']

[0134] (2) Input feature vector X DO ∈R 4×1 It is represented by 4 features: Chla, Temperature, Month sin ,Month cos ), each feature has one value:

[0135] X′ DO ∈R 4×1×1

[0136] (3) 13 layers are defined with different hyperparameters to perform operations such as convolution, pooling, and batch normalization on the input features to generate the desired output.

[0137] Step 4: Dissolved oxygen concentration prediction model accuracy verification

[0138] The dissolved oxygen concentration prediction model is verified by the test data set. First, the input features (B1, B2, …, B8, B 8A ,B9,B 11 ,B 12 ,Month sin ,Month cos) predicts chlorophyll a and water temperature values, defined as follows:

[0139]

[0140] in, represents the predicted chlorophyll a concentration value, Represents the predicted water temperature value; these prediction parameters are combined with the time feature Month sin ,Month cos Connect them together and prepare the input dataset for the 2DCNN model to predict dissolved oxygen as follows:

[0141]

[0142] This dataset is used as f 2DCNN Inputs to the model, used to predict dissolved oxygen values;

[0143]

[0144] in, represents the predicted dissolved oxygen value;

[0145] Finally, the accuracy of the model was evaluated using regression accuracy indicators, including r, RMSE, MAE, MAPE, and UMAPE. and the actual dissolved oxygen value y DO Make a comparison.

[0146] Finally, it should be noted that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A remote sensing inversion method for dissolved oxygen concentration in water based on the XGBoost-2DCNN model, characterized in that The following steps are involved: (1) Collect, process and standardize data for training the prediction model of dissolved oxygen (DO) concentration in water bodies; (2) Construct a water chlorophyll a concentration and water temperature prediction model based on the XGBoost machine learning model; (3) Develop a two-dimensional convolutional neural network model (2DCNN) to predict dissolved oxygen concentration; (4) Verification of the accuracy of the dissolved oxygen concentration prediction model.

2. According to claim 1, a remote sensing inversion method for dissolved oxygen concentration in water based on the XGBoost-2DCNN model is characterized in that: The specific method of "collecting, processing and standardizing data for training the water body dissolved oxygen DO concentration prediction model" described in step (1) is as follows: (1) Collect and collect long-term in situ water quality measurement data from the study area, including various optically active and inactive water quality parameters, as well as metadata such as the geographic coordinates of the water quality monitoring stations, data quality, and collection date; Extract necessary data from the collected data sets, including water quality control area, monitoring station code, monitoring station geographic coordinates, sampling date, chlorophyll a, water temperature, dissolved oxygen DO; compile the AOI polygon vector file of the area of ​​interest in the study area; establish a spatiotemporal matching data set containing in-situ water quality parameter measurements and corresponding satellite multispectral image water surface reflectance values; Google Earth Engine GEE was used to extract water surface reflectance spectra for each water quality monitoring station and sample collection date; (2) The seasonal variation trend of water quality parameter concentrations is incorporated into the modeling of dissolved oxygen concentration prediction; the time information—month is used as the input parameter of the modeling, and the month is encoded into a periodic variable through sine and cosine transformations, as shown below: where x represents the month; this transformation ensures that all months are correctly positioned on the circular scale and accurately represents their cycle pattern by capturing the cycle angle and amplitude; the converted month feature is denoted as Month sin and Month cos , which will be used as input features for the dissolved oxygen prediction model; (3) Extract data records of the research period from the prepared data set to prepare two training data sets; one training data set contains input features: including spectral reflectance data and converted monthly features; output features include chlorophyll a and water temperature, which are used to train the machine learning algorithm for univariate prediction of chlorophyll a and water temperature; taking the optical remote sensing image Sentinel-2 as an example, the data set is represented as: Training_Dataset Chla_Temperature ={B1,B2,…,B8,B 8A ,B9,B 11 ,B 12 ,Month sin ,Month cos ,Chla,Temperature} Another training dataset contains input features including chlorophyll a, water temperature, and transition month features, as well as annotated dissolved oxygen output features, which are used to train a convolutional neural network model for univariate prediction of dissolved oxygen; this dataset is represented as Training_Dataset DO ={Chla,Temperature,Month sin ,Month cos ,DO} (4) Extract the data records of the research period from the prepared data set to prepare the test data set. The data set contains input features including reflectance data and converted monthly features. The first stage is used to predict chlorophyll a and water temperature, and then the prediction is used to predict dissolved oxygen; the data set is represented as Test_dataset={B1,B2,…,B8,B 8A ,B9,B 11 ,B 12 ,Month sin ,Month cos } 3. The method for remote sensing inversion of dissolved oxygen concentration in water based on the XGBoost-2DCNN model according to claim 1, characterized in that: The specific method of "building a water body chlorophyll a concentration and water temperature prediction model based on XGBoost machine learning model" described in step (1) is as follows: Based on the machine learning algorithm eXtreme Gradient Boosting (XGBoost), the complex relationship between water quality parameters and satellite optical characteristics is captured to predict water quality concentration; using Training_Dataset Chla_Temperature The dataset was used to train the XGBoost model using the optimal hyperparameters to estimate chlorophyll a and water temperature; for each target variable, the training dataset was split into input features and target output as follows: X Chla_Temperature =Training_Dataset Chla_Temperature \{Chla,Temperature} Y Chla =Training_Dataset Chla_Temperature ["Chla"] Y Temperature =Training_Dataset Chla_Temperature [′Temperature′] For each parameter model, the XGBoost algorithm is used for training, and its objective function is to minimize the prediction error, which is specifically expressed as follows: f XGBOOST,Chla (X Chla_Temperature )=argmin∑ i (Y Chla,i -f(X Chla_Temperature,i )) 2 Among them, i represents the i-th sample value, f(X) represents the function of the model to predict chlorophyll a or water temperature based on the input features, and the goal is to minimize the sum of the squared errors between the predicted value and the actual value.

4. The method for remote sensing inversion of dissolved oxygen concentration in water based on the XGBoost-2DCNN model according to claim 1, characterized in that: The specific method of "developing a two-dimensional convolutional neural network model 2DCNN to predict dissolved oxygen concentration" described in step (1) is as follows: A two-dimensional convolutional neural network (2DCNN) model was developed and trained using input parameters such as chlorophyll a, water temperature, and transition month features to predict dissolved oxygen concentration; the 2DCNN model was trained using optimal hyperparameters to minimize training and validation losses; the steps for training and developing the 2DCNN model were as follows: (1) Train the training data set Training_Dataset DO Divided into input features and target output: X DO =Training_Dataset DO \{DO} Y DO =Training_Dataset DO [′DO′] (2) Input feature vector X DO ∈R 4×1 It is represented by 4 features: Chla, Temperature, Month sin ,Month cos ), each feature has one value: X′ DO ∈R 4×1×1 (3) 13 layers are defined with different hyperparameters to perform operations such as convolution, pooling, and batch normalization on the input features to generate the desired output. The definitions of these layers are as follows: Layer 1 => Conv2D: 32 convolutional filters W1 of size 4×1 with “same” padding and hyperbolic tangent activation function applied to the input features defined as follows: Z1=Conv2D(X′ DO ,W1)+b1 A1=tanh(Z1) Layer 2 => Batch Normalization: Normalize the activation function output A1 to improve the stability of model training. It is defined as follows Among them, μ1 and σ1 2 Respectively represent the mean and variance of A1, γ1 and β1 represent the learnable scaling and shifting parameters, respectively; ∈ is a very small number to avoid the denominator of the formula being 0; b1 is the model bias value, and the meanings of similar parameters in subsequent layers are similar, and so on; Layer 3 => MaxPooling2D: applies max pooling to the batch normalized output with a pooling window of 4×1 and a stride of 1, denoted as Layer 4 => Conv2D: 64 convolutional filters W2 of size 3×1, with “same” padding, and hyperbolic tangent activation function as follows: Z2=Conv2D(P1,W2)+b2 A2=tanh(Z2) Layer 5 => Batch Normalization: Normalize the activation function output A2 to improve the stability of training. It is defined as follows Layer 6 => MaxPooling2D: applies max pooling to batch normalized output with a pooling window of 2×1 and a stride of 1, denoted as Layer 7 => Conv2D: 128 convolutional filters W3 of size 2×1, with “same” padding, and hyperbolic tangent activation function as follows: Z3=Conv2D(P2,W3)+b3 A3=tanh(Z3) Layer 8 => Batch Normalization: Normalize the activation function output A3 to improve the stability of training. It is defined as follows Layer 9 => MaxPooling2D: applies max pooling to batch normalized output with a pooling window of 2×1 and a stride of 1, denoted as Layer 10 => Conv2D: 256 convolutional filters of size 1×1 W4 with “same” padding and hyperbolic tangent activation function as follows: Z4=Conv2D(P3,W4)+b4 A4=tanh(Z4) Layer 11 => Batch Normalization: Normalize the activation function output A4 to improve the stability of model training. It is defined as follows Layer 12 => Dropout: A dropout rate of 0.3 is used to drop out to reduce model overfitting; Layer 13 => Flatten: Flatten the 4D tensor D4 after the Dropout operation into a 1D vector, which is represented by F = Flatten (D4) (4) The output layer is defined as a fully connected layer with one unit, and the linear activation of the regression output is expressed as Among them, W o is the weight vector, b o is the bias term, · represents the dot product operation; (5) The model integrated with the above layers consists of a loss function and an Adam optimizer. The loss function is defined by the mean square error (MSE) loss and is used to calculate the predicted output. The difference between the actual value y; the Adam optimizer uses the gradient of the loss function to update the model parameters (W, b), and the initial learning rate (lr) is 0.0001; the loss function and optimization are defined as follows; Among them, W (t) ,b (t) Respectively represent the weight and bias term in the current iteration state, W (t+1) ,b (t+1) They represent the weight and bias terms updated in the next iterative calculation process, respectively. The gradient of the loss function; lr represents the learning rate; t represents the sequence number of iterative calculations. (6) Train the model by setting the training configuration, including defining the minimum batch size (B = 32), EPOCHS (E = 500) for 500 iterations over the entire dataset, the validation dataset, and the early stopping criterion based on restoring the best weights when the validation performance stops improving. The model is saved and available for use with the function f 2DCNN express.

5. The method for remote sensing inversion of dissolved oxygen concentration in water based on the XGBoost-2DCNN model according to claim 1, characterized in that: The specific method of "dissolved oxygen concentration prediction model accuracy verification" described in step (1) is as follows: The dissolved oxygen concentration prediction model is verified by the test data set. First, the input features (B1, B2, …, B8, B 8A ,B9,B 11 ,B 12 ,Month sin ,Month cos ) predicts chlorophyll a and water temperature values, defined as follows: in, represents the predicted chlorophyll a concentration value, Represents the predicted water temperature value; these prediction parameters are combined with the time feature Month sin ,Month cos Connect them together and prepare the input dataset for the 2DCNN model to predict dissolved oxygen as follows: This dataset is used as f 2DCNN Inputs to the model, used to predict dissolved oxygen values; in, represents the predicted dissolved oxygen value; Finally, the accuracy of the model was evaluated using regression accuracy indicators, including Pearson correlation coefficient r, root mean square error RMSE, mean absolute error MAE, mean absolute percentage error MAPE and unbiased mean absolute percentage error UMAPE to predict dissolved oxygen values. and the actual dissolved oxygen value y DO Make a comparison.

Images