Water quality parameter inversion model training method, water quality parameter inversion method, terminal and medium
By using remote sensing and meteorological data for time and space registration in the water quality parameter inversion method, training models are constructed, and the deviation problems caused by not considering relevant factors in the existing technology are solved, and the accuracy of water quality parameter inversion is improved. objectivity.
Patent Information
- Application Number
- CN202211537762.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-12-01
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2042-12-01
AI Technical Summary
The existing water quality parameter inversion methods fail to fully consider factors such as temperature, vegetation index, wind speed and aerosol, resulting in a large deviation from the actual situation.
Using the water quality parameter inversion method based on remote sensing and meteorological data, we use water quality observation data, remote sensing reflectivity data and environmental data to register time and space, build sample pairs, and train the pre-constructed water quality inversion model. The trained model is used to invert water quality parameters.
The accuracy of the inversion of water quality parameters is improved, so that the inversion process can fully consider related factors such as temperature, vegetation index, wind speed and aerosol, and enhance the accuracy and objectivity of the results.
Smart Images

Figure CN118133479B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of remote sensing technology, and in particular to a water quality parameter inversion model training method, a water quality parameter inversion method, a terminal and a computer storage medium. Background Art
[0002] Non-optically active water quality parameters are important indicators for water quality monitoring and can be used to characterize water quality status. In the process of water quality monitoring, remote sensing technology is usually used to invert water quality parameters. Currently, in the water quality inversion process based on remote sensing technology, the data source used usually only includes satellite remote sensing data, and does not fully consider relevant factors such as temperature, vegetation index, wind speed and aerosols.
[0003] However, since non-optically active water quality parameters are usually affected by factors such as water body temperature and chlorophyll concentration, and there is often a strong correlation between water body temperature and the surface temperature around the water body; similarly, there is often a strong correlation between the vegetation index of the water body and the surface temperature around the water body. For example, Cândido, Rajib et al. provide that the coverage of vegetation around rivers and lakes will affect and reflect the chlorophyll concentration in the water body in a complex way; in addition, when remote sensing technology is used to observe water bodies, it is usually also affected by the wind speed on the surface of rivers and lakes and aerosols in the air. Among them, the wind speed on the surface of rivers and lakes affects the reflectivity of the water surface by causing changes in the roughness of the water surface, and aerosols affect the scattering and refraction of sunlight and change the transmission path of sunlight and its reflected light in the atmosphere. Therefore, both wind speed and aerosol size will have a serious impact on the accuracy of the observation results. Summary of the invention
[0004] In view of the shortcomings of the prior art mentioned above, the purpose of the present invention is to provide a water quality parameter inversion method, device and computer storage medium based on remote sensing and meteorological data, which can solve the problem that the existing water quality parameter inversion method does not fully consider relevant factors such as temperature and vegetation index, resulting in a large deviation between the obtained water quality parameter inversion results and the actual situation.
[0005] To achieve the above-mentioned purpose and other related purposes, the first aspect of the present invention provides a water quality parameter inversion method based on remote sensing and meteorological data, comprising: obtaining water quality observation data, remote sensing reflectance data and environmental data of a water body; wherein the environmental data includes various environmental sub-data; based on the acquisition time, aligning the water quality observation data, the remote sensing reflectance data and the environmental data to construct each sample pair; based on each of the sample pairs, training a pre-constructed water quality inversion model to obtain a trained water quality inversion model.
[0006] In one embodiment of the present invention, the water quality observation data, the remote sensing reflectance data and the environmental data are aligned based on the acquisition time, including: dividing each environmental sub-data into a first category of environmental sub-data and a second category of environmental sub-data based on the length of the update period of each environmental sub-data; the update period of the first category of environmental sub-data is longer than that of the second category of environmental sub-data; based on the acquisition date, the water quality observation data is aligned with the first category of environmental sub-data to obtain each observation data group; based on the acquisition time, the remote sensing reflectance data is aligned with the second category of environmental sub-data to obtain a remote sensing data group; and, based on the acquisition time, the observation data group and the remote sensing data group are aligned.
[0007] In one embodiment of the present invention, the environmental data includes surface temperature data, aerosol data, vegetation coverage data and wind speed data; then the environmental sub-data are divided into first-category environmental sub-data and second-category environmental sub-data, including: using the surface temperature data and the vegetation coverage data as the first-category environmental sub-data; and using the aerosol data and the wind speed data as the second-category environmental sub-data.
[0008] In one embodiment of the present invention, the method for obtaining the observation data group includes: obtaining the collection date of the water quality observation data; based on the collection date, filtering out data with the same collection date from the surface temperature data as the aligned surface temperature data; based on the collection date of the water quality observation data, filtering out data with a collection date that differs from the collection date of the water quality observation data by less than a first time threshold from the vegetation cover data as the aligned vegetation cover data; constructing the observation data group based on the water quality observation data, the aligned surface temperature data and the aligned vegetation cover data.
[0009] In one embodiment of the present invention, the method for acquiring the remote sensing data group includes: obtaining the collection time of the remote sensing reflectivity data; based on the collection time, filtering out data from the wind speed data whose collection time difference with the collection time of the remote sensing reflectivity data is less than a second time threshold, as the aligned wind speed data; based on the collection time of the remote sensing reflectivity data, filtering out data from the aerosol data whose collection time difference with the collection time of the remote sensing reflectivity data is less than a third time threshold, as the aligned aerosol data; constructing the remote sensing data group based on the remote sensing reflectivity data, the aligned wind speed data and the aligned aerosol data.
[0010] In one embodiment of the present invention, the alignment of the observation data group and the remote sensing data group based on the acquisition time includes: taking the acquisition time of each water quality observation data as the acquisition time of the observation data group; taking the acquisition time of each remote sensing reflectivity data as the acquisition time of the remote sensing data group; based on the acquisition time of the observation data group, screening out from the remote sensing data group a data group whose acquisition time difference with the acquisition time of the observation data group is less than a fourth time threshold as the aligned remote sensing data group.
[0011] The second aspect of the present invention provides a water quality parameter inversion method, comprising: obtaining remote sensing reflectance data and environmental data of a water body to be detected; performing time alignment and spatial alignment on the remote sensing reflectance data and the environmental data to obtain aligned remote sensing reflectance data and environmental data; inputting the aligned remote sensing reflectance data and environmental data into a water body parameter inversion model to obtain water quality parameter information of the water body; wherein the water body parameter inversion model is a model trained using any of the water quality parameter inversion model training methods described above.
[0012] In one embodiment of the present invention, the environmental data includes surface temperature data, aerosol data, vegetation coverage data and wind speed data related to the water body; and the time alignment of the remote sensing reflectivity data and the environmental data includes: obtaining the acquisition date of the remote sensing reflectivity data; based on the acquisition date, filtering out data with the same acquisition date in the surface temperature data as the aligned surface temperature data; according to the acquisition date of the remote sensing reflectivity data, filtering out data with a collection date less than a fifth time threshold from the vegetation coverage data as the aligned vegetation coverage data; obtaining the acquisition time of the remote sensing reflectivity data; based on the acquisition time, filtering out data with a collection time less than a sixth time threshold from the wind speed data as the aligned wind speed data; according to the acquisition time of the remote sensing reflectivity data, filtering out data with a collection time less than a seventh time threshold from the aerosol data as the aligned aerosol data.
[0013] The third aspect of the present invention provides a terminal, comprising: a processor and a memory; the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to execute any of the water quality parameter inversion model training methods described above or any of the water quality parameter inversion methods described above.
[0014] The fourth aspect of the present invention also provides a computer storage medium, which stores a computer program, and the computer program is executed by a processor to perform any of the water quality parameter inversion model training methods described above or any of the water quality parameter inversion methods described above.
[0015] As described above, the water quality parameter inversion model training method, water quality parameter inversion method, terminal and computer storage medium provided by the present invention perform inversion calculations on water quality parameters based on remote sensing reflectivity data and surface temperature, vegetation coverage, aerosol and other data information related to the water body, so that relevant factors such as temperature, vegetation index, wind speed and aerosol can be fully considered in the water quality parameter inversion process, thereby effectively improving the accuracy of water quality parameter inversion. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 Shown is a flow chart of a water quality parameter inversion model training method provided by an embodiment of the present invention;
[0017] Figure 2 Shown is a flow chart of step S200 in the water quality parameter inversion model training method provided by an embodiment of the present invention during execution;
[0018] Figure 3 Shown is a schematic flow chart of a water quality parameter inversion method provided by an embodiment of the present invention;
[0019] Figure 4 Shown is a schematic diagram of the structure of a terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0020] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.
[0021] It should be noted that the illustrations provided in the following embodiments are only for schematically illustrating the basic concept of the present invention.
[0022] In order to solve the technical problems existing in the prior art, the present invention provides a water quality parameter inversion model training method in a first aspect, which is used to obtain a trained water quality parameter inversion model; wherein the water quality parameter inversion model is used to obtain water quality parameter information of a water body.
[0023] In a specific embodiment, the water quality parameters include dissolved oxygen (DO), chemical oxygen demand (COD), ammonia nitrogen (NH3-N) or other existing water quality parameters.
[0024] See also Figure 1 , which is a flow chart of the water quality parameter inversion model training method; Figure 1 As shown, the method includes:
[0025] S100, obtaining water quality observation data, remote sensing reflectance data and environmental data of water bodies;
[0026] The water quality observation data are measured data including water quality parameter values.
[0027] The remote sensing reflectivity data is water body reflectivity data obtained based on remote sensing data;
[0028] In a specific embodiment, the remote sensing reflectance data is remote sensing reflectance data obtained after image processing of Sentinel multispectral satellite remote sensing data.
[0029] The environmental data includes surface temperature data, aerosol data, vegetation coverage data and wind speed data related to the water body; wherein the surface temperature data, the aerosol data and the wind speed data are the surface temperature data, aerosol data and the wind speed data within the water body area; the vegetation coverage data are the vegetation coverage data within a preset range outside the water body.
[0030] Optionally, the surface temperature data, aerosol data and vegetation cover data are all data obtained based on MODIS products, namely, MODIS surface temperature data, MODIS aerosol data and MODIS vegetation cover data; compared with the surface temperature data, aerosol data and vegetation cover data obtained based on other remote sensing data, they can better reflect the spatiotemporal changes of atmospheric parameters, and can achieve dynamic matching with the remote sensing reflectivity of water bodies in time and space, thereby making up for the problems of poor adaptability and high uncertainty in the existing atmospheric correction process.
[0031] Optionally, the vegetation coverage data adopts the MODIS vegetation index.
[0032] Optionally, the wind speed data is data obtained based on ERA5 meteorological data.
[0033] S200, based on the acquisition time, registering the water quality observation data, the remote sensing reflectance data and the environmental data, and constructing each sample pair based on each data after time registration;
[0034] In this embodiment, when this step is executed, it includes the following sub-steps:
[0035] S201, dividing each of the environmental sub-data into first-category environmental sub-data and second-category environmental sub-data based on the length of the update cycle of each of the environmental sub-data;
[0036] The update period of the first type of environment sub-data is longer than that of the second type of environment sub-data.
[0037] Specifically, the update period of each of the environmental sub-data is obtained, and the data with an update period greater than or equal to 24 hours is taken as the first type of environmental sub-data; and the data with an update period less than 24 hours is taken as the second type of environmental sub-data.
[0038] In this embodiment, the environmental data includes surface temperature data, aerosol data, vegetation cover data and meteorological data; wherein the update period of the surface temperature data is 1 day; the update period of the aerosol data is 1 day; the update time of the vegetation cover data is 8 days; and the update time of the meteorological data is 1 hour. The surface temperature data and the vegetation cover data are used as the first type of environmental sub-data; the aerosol data and the wind speed data are used as the second type of environmental sub-data.
[0039] S202, based on the collection date, aligning the water quality observation data with the first type of environmental sub-data to obtain each observation data group;
[0040] Specifically, the acquisition date of the water quality observation data is obtained; based on the acquisition date, the data with the same acquisition date is screened out from the surface temperature data as the surface temperature data after alignment; and based on the acquisition date of the water quality observation data, the data with the acquisition date different from the acquisition date of the water quality observation data by less than a first time threshold is screened out from the vegetation cover data as the vegetation cover data after alignment;
[0041] Optionally, the first time threshold is 2 days.
[0042] Constructing the observation data set based on the water quality observation data, the registered surface temperature data and the registered vegetation cover data;
[0043] Specifically, for a single water quality observation data, it is taken together with the aligned surface temperature data and the aligned vegetation coverage data as a group of the observation data; this operation is performed on each water quality observation data to obtain each group of the observation data.
[0044] S203, based on the acquisition time, registering the remote sensing reflectance data with the second type of environmental sub-data to obtain a remote sensing data group;
[0045] Specifically, the acquisition time of the remote sensing reflectivity data is obtained; based on the acquisition time, data whose acquisition time difference with the acquisition time of the remote sensing reflectivity data is less than a second time threshold is selected from the wind speed data as the registered wind speed data; and based on the acquisition time of the remote sensing reflectivity data, data whose acquisition time difference with the acquisition time of the remote sensing reflectivity data is less than a third time threshold is selected from the aerosol data as the registered aerosol data;
[0046] Optionally, the second time threshold is 1 hour;
[0047] Optionally, the third time threshold is 1 hour.
[0048] constructing the remote sensing data set based on the remote sensing reflectivity data, the registered wind speed data and the registered aerosol data;
[0049] Specifically, for a single remote sensing reflectivity data, it is taken together with the aligned wind speed data and the aligned aerosol data as a group of the remote sensing data sets; this operation is performed on each remote sensing reflectivity data to obtain each group of the remote sensing data sets.
[0050] S204: Based on the acquisition time, align the observation data set and the remote sensing data set to obtain a constructed sample pair.
[0051] Specifically, the acquisition time of each of the water quality observation data is obtained as the acquisition time of the observation data group; and the acquisition time of each of the remote sensing reflectance data is used as the acquisition time of the remote sensing data group; based on the acquisition time of the observation data group, a data group whose acquisition time difference with the acquisition time of the observation data group is less than a fourth time threshold is screened out from the remote sensing data group as the registered remote sensing data group;
[0052] For a single observation data set, it and the registered remote sensing data set are taken as a set of sample pairs; this operation is performed on each observation data set to obtain each set of sample pairs.
[0053] Optionally, the fourth time threshold is 4 hours.
[0054] S300: Based on each of the sample pairs, a pre-constructed water quality inversion model is trained to obtain a trained water quality inversion model.
[0055] In this embodiment, the model sample pairs are randomly divided into a training set and a test set; optionally, the model sample pairs are randomly divided into 80% training set and 20% test set;
[0056] Based on the training set, the preset water quality model is trained, and based on the test set, the trained model is tested, and the process is repeated until the model test result meets the preset training requirements to obtain the final water quality model.
[0057] In one embodiment, the model test results include the coefficient of determination (R 2 ), mean absolute error (MAE), and root mean square error (RMSE).
[0058] Optionally, the pre-constructed water quality model is a model constructed using a support vector machine regression method, that is, the input in a low-dimensional feature space is mapped to a high-dimensional feature space through a nonlinear transformation, and the regression hyperplane is found so that the difference between the output prediction value and the actual observation value is less than a preset tolerance deviation.
[0059] Optionally, the pre-constructed water quality model is a model constructed using a random forest method, that is, a characteristic attribute is selected for each non-leaf node in the decision tree for testing, and each test result is output through a branch, and finally the output value of the random forest is jointly determined by each decision tree.
[0060] Optionally, the pre-constructed water quality model is a model constructed using a multi-layer perceptron method, that is, the connection weights in the multi-layer perceptron network are changed through supervised training so that it can approximate any function on the set with arbitrary precision.
[0061] Optionally, the water quality parameter inversion method based on remote sensing and environmental data, when executing step S100, further includes:
[0062] The remote sensing reflectance data and the environmental data are preprocessed to perform subsequent steps based on the preprocessed remote sensing reflectance data and environmental data.
[0063] Specifically, preprocessing the remote sensing reflectance data and the environmental data includes:
[0064] Based on the spatial resolution of the remote sensing reflectance data, each environmental sub-data in the environmental data is resampled respectively, so that the spatial resolution of each environmental sub-data is the same as the spatial resolution of the remote sensing reflectance data.
[0065] Optionally, a cubic interpolation resampling method is used to resample each of the environmental sub-data separately.
[0066] Further, preprocessing the remote sensing reflectivity data and the environmental data also includes:
[0067] Based on the spatial distribution of the water body range, the remote sensing reflectance data and the environmental data are spatially calibrated respectively to make the spatial position information of the water body on each data the same.
[0068] The present invention provides a water quality parameter inversion method in a second aspect, so as to obtain water quality parameter information of a water body to be detected based on remote sensing data and environmental data; wherein the water quality parameters include dissolved oxygen (DO), chemical oxygen demand (COD), ammonia nitrogen (NH3-N) or other existing water quality parameters.
[0069] See also Figure 3 , which is a schematic diagram of the process of the water quality parameter inversion method; Figure 3 As shown, the method includes:
[0070] S10, obtaining remote sensing reflectance data and environmental data of the water body to be detected;
[0071] Wherein, the remote sensing reflectivity data is water body reflectivity data obtained based on remote sensing data;
[0072] In a specific embodiment, the remote sensing reflectance data is remote sensing reflectance data obtained after image processing of Sentinel multispectral satellite remote sensing data.
[0073] The environmental data includes surface temperature data, aerosol data, vegetation coverage data and wind speed data related to the water body; wherein the surface temperature data, the aerosol data and the wind speed data are the surface temperature data, aerosol data and the wind speed data within the water body area; the vegetation coverage data are the vegetation coverage data within a preset range outside the water body.
[0074] Optionally, the surface temperature data, aerosol data and vegetation cover data are all data acquired based on MODIS products, namely, MODIS surface temperature data, MODIS aerosol data and MODIS vegetation cover data.
[0075] Optionally, the vegetation coverage data adopts the MODIS vegetation index.
[0076] Optionally, the wind speed data is data obtained based on ERA5 meteorological data.
[0077] S20, performing temporal registration and spatial registration on the remote sensing reflectance data and the environmental data to obtain registered remote sensing reflectance data and environmental data;
[0078] Specifically, the acquisition date of the remote sensing reflectivity data is obtained; based on the acquisition date, the data with the same acquisition date is screened out from the surface temperature data as the surface temperature data after registration; and based on the acquisition date of the remote sensing reflectivity data, the data with the acquisition date different from the acquisition date of the remote sensing reflectivity data less than a fifth time threshold is screened out from the vegetation cover data as the vegetation cover data after registration;
[0079] Optionally, the fifth time threshold is 2 days.
[0080] Obtaining the collection time of the remote sensing reflectivity data; based on the collection time, screening out data whose collection time difference with the collection time of the remote sensing reflectivity data is less than a sixth time threshold from the wind speed data as the aligned wind speed data; and based on the collection time of the remote sensing reflectivity data, screening out data whose collection time difference with the collection time of the remote sensing reflectivity data is less than a seventh time threshold from the aerosol data as the aligned aerosol data;
[0081] Optionally, the sixth time threshold is 1 hour.
[0082] Optionally, the seventh time threshold is 1 hour.
[0083] For a single remote sensing reflectivity data, it is combined with the registered surface temperature data, vegetation cover data, wind speed data and aerosol data as a group of model input data; this operation is performed on each remote sensing reflectivity data to obtain each group of model input data.
[0084] S30, inputting the registered remote sensing reflectivity data and environmental data into a water body parameter inversion model to obtain water quality parameter information of the water body.
[0085] Among them, the water body parameter inversion model is based on Figure 1 Model trained with the method shown.
[0086] Specifically, each group of the model input data is input into the water body parameter inversion model to obtain the water quality parameter information at the time of collection.
[0087] The acquisition time is the remote sensing data acquisition time corresponding to the remote sensing reflectivity data.
[0088] In order to solve the problems existing in the prior art, the present invention further provides a terminal in a third aspect, see Figure 4 , showing a schematic diagram of the structure of the terminal of the present invention; Figure 4As shown, the terminal 5 includes a memory 51 and a processor 52 that are interconnected; the memory 51 is used to store computer programs, and the processor 52 is used to execute the computer programs stored in the memory, so that the steps in the water quality parameter inversion model training method or the water quality parameter inversion method can be implemented when the terminal executes.
[0089] Optionally, the number of the memories may be one or more, and the number of the processors may be one or more. Figure 4 Take one as an example.
[0090] Optionally, the processor in the terminal loads one or more instructions corresponding to the application process into the memory in accordance with the steps in the water quality parameter inversion model training method or the water quality parameter inversion method as described above, and the processor runs the application stored in the memory, thereby realizing the various functions in the water quality parameter inversion model training method or the water quality parameter inversion method, which will not be described in detail here.
[0091] It should be noted that the memory includes but is not limited to random access memory (RAM), and may also include non-volatile memory, such as at least one disk storage. Similarly, the processor may also be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gates or transistor logic devices, and discrete hardware components.
[0092] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when called by a processor, implements the water quality parameter inversion model training method or the water quality parameter inversion method.
[0093] Among them, the computer-readable storage medium can be a tangible device that can hold and store instructions used by the instruction execution device. The computer-readable storage medium can be, for example, (but not limited to) an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the above. More specific examples of computer-readable storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, and a mechanical encoding device.
[0094] The computer-readable program described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. A network adapter card or a network interface in each computing / processing device receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing / processing device.
[0095] In summary, the water quality parameter inversion model training method, water quality parameter inversion method, terminal and computer storage medium provided by the present invention perform inversion calculations on water quality parameters based on remote sensing reflectivity data and surface temperature, vegetation coverage and aerosol data information related to water bodies, so that relevant factors such as temperature, vegetation index, wind speed and aerosol can be fully considered in the process of water quality parameter inversion, thereby effectively improving the accuracy of water quality parameter inversion; and, by aligning the water quality observation data with the environmental sub-data with a slower update period by the acquisition date, and by aligning the remote sensing inversion data by the acquisition date, the water quality observation data can be inverted by the remote sensing inversion data. The reflectivity data and the environmental sub-data with a faster update cycle are time-aligned at the time of collection, so as to improve the correlation between different types of data and improve the accuracy of the inversion. In addition, when analyzing the relationship between water parameter concentration and remote sensing reflectivity information, MODIS products are used to obtain aerosol data, temperature data and vegetation cover data. Dynamic matching with remote sensing reflectivity can be performed in time and space to make up for the poor adaptability and uncertainty of many existing atmospheric correction algorithms, thereby further ensuring the objectivity and accuracy of the water quality parameter inversion process.
[0096] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.
Claims
1. A water quality parameter inversion model training method, characterized in that: include: Acquire water quality observation data, remote sensing reflectance data and environmental data of a water body; wherein the environmental data includes various environmental sub-data; Based on the acquisition time, registering the water quality observation data, the remote sensing reflectance data and the environmental data to construct each sample pair; Based on each of the sample pairs, a pre-constructed water quality inversion model is trained to obtain a trained water quality inversion model; wherein the registering of the water quality observation data, the remote sensing reflectance data and the environmental data based on the acquisition time includes: Based on the length of the update period of each of the environmental sub-data, each of the environmental sub-data is divided into a first type of environmental sub-data and a second type of environmental sub-data; the update period of the first type of environmental sub-data is longer than that of the second type of environmental sub-data; Based on the acquisition date, aligning the water quality observation data with the first type of environmental sub-data to obtain each observation data group; based on the acquisition time, aligning the remote sensing reflectance data with the second type of environmental sub-data to obtain a remote sensing data group; and based on the acquisition time, aligning the observation data group and the remote sensing data group; The environmental data includes surface temperature data, aerosol data, vegetation cover data and wind speed data.
2. The water quality parameter inversion model training method according to claim 1, characterized in that: The dividing each of the environmental sub-data into first-category environmental sub-data and second-category environmental sub-data comprises: The surface temperature data and the vegetation coverage data are used as the first type of environmental sub-data; and the aerosol data and the wind speed data are used as the second type of environmental sub-data.
3. The water quality parameter inversion model training method according to claim 2, characterized in that: The method for obtaining the observation data group includes: Obtaining the collection date of the water quality observation data; based on the collection date, selecting data with the same collection date from the surface temperature data as the registered surface temperature data; According to the collection date of the water quality observation data, data whose collection date is less than a first time threshold from the vegetation coverage data are selected as the registered vegetation coverage data; The observation data set is constructed based on the water quality observation data, the aligned surface temperature data and the aligned vegetation cover data.
4. The water quality parameter inversion model training method according to claim 2, characterized in that: The remote sensing data group is obtained by: Obtaining the acquisition time of the remote sensing reflectivity data; based on the acquisition time, selecting data whose acquisition time difference with the acquisition time of the remote sensing reflectivity data is less than a second time threshold from the wind speed data as the registered wind speed data; According to the collection time of the remote sensing reflectivity data, data whose collection time is less than a third time threshold from the collection time of the remote sensing reflectivity data is selected from the aerosol data as the registered aerosol data; The remote sensing data set is constructed based on the remote sensing reflectivity data, the registered wind speed data and the registered aerosol data.
5. The water quality parameter inversion model training method according to claim 2, characterized in that: The registering the observation data set and the remote sensing data set based on the acquisition time includes: The collection time of each of the water quality observation data is used as the collection time of the observation data group; The collection time of each remote sensing reflectivity data is used as the collection time of the remote sensing data group; Based on the acquisition time of the observation data group, a data group whose acquisition time difference with the acquisition time of the observation data group is less than a fourth time threshold is selected from the remote sensing data group as the registered remote sensing data group.
6. A water quality parameter inversion method, characterized in that: include: Obtain remote sensing reflectance data and environmental data of the water body to be detected; Performing temporal registration and spatial registration on the remote sensing reflectance data and the environmental data to obtain registered remote sensing reflectance data and environmental data; Input the registered remote sensing reflectance data and environmental data into the water body parameter inversion model to obtain water quality parameter information of the water body; Wherein, the water body parameter inversion model is a model trained using the water quality parameter inversion model training method as described in any one of claims 1 to 5.
7. The water quality parameter inversion method according to claim 6, characterized in that: The environmental data includes surface temperature data, aerosol data, vegetation cover data and wind speed data related to water bodies; and performing time registration on the remote sensing reflectivity data and the environmental data includes: Obtaining the acquisition date of the remote sensing reflectivity data; based on the acquisition date, filtering out data with the same acquisition date from the land surface temperature data as the registered land surface temperature data; According to the acquisition date of the remote sensing reflectance data, data whose acquisition date is less than a fifth time threshold from the vegetation coverage data are selected as the registered vegetation coverage data; Obtaining the acquisition time of the remote sensing reflectivity data; based on the acquisition time, selecting data whose acquisition time difference with the acquisition time of the remote sensing reflectivity data is less than a sixth time threshold from the wind speed data as the registered wind speed data; According to the collection time of the remote sensing reflectivity data, data whose collection time is less than the collection time of the remote sensing reflectivity data by less than a seventh time threshold is selected from the aerosol data as the registered aerosol data.
8. A terminal, characterized in that: include: Processor and memory; The memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to perform the water quality parameter inversion model training method as described in any one of claims 1 to 5 or the water quality parameter inversion method as described in any one of claims 6 to 7.
9. A computer storage medium storing a computer program, characterized in that: The computer program is executed by a processor to execute the water quality parameter inversion model training method according to any one of claims 1 to 5 or the water quality parameter inversion method according to any one of claims 6 to 7.
