Method, system and electronic device for determining carbon-water flux information of a meteorological station

Abstract

The present invention discloses a method, system and electronic device for determining carbon and water flux information of a meteorological station, and relates to the technical field of electrical digital data processing. Based on limited eddy covariance flux stations, the present invention integrates remote sensing big data and machine learning / deep learning technology, constructs tens of thousands of eddy covariance flux tower carbon and water flux simulation models, and obtains a comprehensive Euclidean distance threshold for migrating the model to the meteorological station. If the flux model migrates to the minimum comprehensive Euclidean distance of the meteorological station and is less than or equal to the distance threshold, the flux model can be determined as a carbon and water flux model applicable to the meteorological station, and the carbon and water flux of the meteorological station is simulated at multiple time scales. Compared with the method of obtaining carbon and water flux based on eddy covariance flux stations, the present invention can mine the carbon and water flux of meteorological stations that meet high-precision requirements in the target area, which significantly increases the amount of high-precision carbon and water flux data obtained, and can meet the data volume and accuracy requirements of ecosystem carbon monitoring and carbon accounting.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical digital data processing, and in particular to a method, system and electronic equipment for determining carbon and water flux information of a meteorological station. Background Art

[0002] The temporal and spatial distribution of terrestrial ecosystem productivity and carbon sources and sinks is still not very clear. First, there is a serious lack of ecosystem carbon observation data and insufficient support for carbon monitoring. Eddy covariance flux towers are the mainstream channel for obtaining ecosystem carbon flux observation data, but the construction, observation, management and data processing costs of flux towers are high, resulting in a scarcity of global flux sites, only 721, and very uneven distribution. The scarcity of carbon flux observation sites has led to the inability to fully and effectively cover complex and diverse ecosystems, completely unable to meet the needs of ecosystem carbon monitoring and carbon accounting, and unable to fully understand the driving mechanism of ecosystem carbon dynamics, which has become one of the main bottlenecks restricting the study of global terrestrial ecosystem carbon cycle. Second, traditional ecosystem carbon accounting technologies and methods face many limitations and challenges. Biogeochemical models / ecosystem models, atmospheric carbon dioxide inversion models, land surface process models and empirical statistical models, either alone or in combination, are the mainstream methods for carbon accounting and assessment. These technical methods or models are basically based on the integration and coupling of known mechanism processes or empirical knowledge, and they do not adequately consider or simplify the mechanisms and processes that are being studied or are not yet mature, and lack sufficient carbon observation data to support verification; the input data of these model technical methods are mostly spatial grid data with large uncertainties, such as reanalysis data, and the spatial resolution is usually 25-100km, which results in high uncertainty in the estimation results and is completely unable to meet the requirements for the accuracy of ecosystem carbon accounting.

[0003] Compared with the very rare eddy covariance flux stations, meteorological stations are more abundant and widely distributed, with nearly 30,000 publicly available worldwide, more than 40 times the number of flux stations; however, the potential of meteorological stations in ecosystem carbon monitoring has not yet been tapped. The support capacity for ecosystem carbon monitoring and carbon accounting is seriously insufficient, and there is a lack of high-precision carbon monitoring and carbon accounting new technologies and methods. Summary of the invention

[0004] The purpose of the present invention is to provide a method, system and electronic equipment for determining carbon-water flux information of a meteorological station, which can use the observation data of the meteorological station to determine the carbon-water flux, increase the data volume of the carbon-water flux, and at the same time improve the accuracy of carbon monitoring and carbon accounting.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A method for determining carbon-water flux information of a meteorological station, comprising:

[0007] Acquire an observation data set in the target area; the observation data set includes observation data of an eddy covariance flux station in the target area, observation data of a meteorological station, and vegetation, soil, topography data and remote sensing data at locations where the eddy covariance flux station and the meteorological station are located;

[0008] The observation data set is quality controlled and standardized, and the flux station data set is obtained by using geographic information system and remote sensing technology; the flux station data set includes: carbon flux, water flux, temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine; the meteorological station data set includes: temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine; the vegetation soil topography data includes soil texture, land cover type and topography; the remote sensing data includes vegetation index, band spectrum, water index, surface temperature and surface humidity;

[0009] Based on the flux station data set and the meteorological station data set, eddy covariance flux stations are selectively combined according to vegetation functional type, climate type and / or region to construct multiple combinations and flux station data sets corresponding to each combination; each flux station data set includes training set and test set flux station data subsets of different partitioning schemes;

[0010] Using a machine learning algorithm or a deep learning algorithm, with training sets of different schemes in different combinations as input, the eddy covariance flux tower carbon-water flux machine learning model or the deep learning simulation model is trained to obtain a carbon flux site carbon-water flux model;

[0011] The test set is used as the input of the carbon-water flux model, and the coefficient of determination is calculated using a spatial cross-validation method;

[0012] Determine the Euclidean distance of each carbon-water flux influencing factor between the training set and the test set to obtain a comprehensive Euclidean distance;

[0013] Determine the average value of all comprehensive Euclidean distances corresponding to the determination coefficient being greater than or equal to 0.5 and less than 0.55, and use the average value as the distance threshold;

[0014] After migrating the carbon-water flux model to the weather station, determining the integrated Euclidean distance corresponding to the weather station;

[0015] Determining a carbon-water flux model applicable to the weather station based on a relationship between the comprehensive Euclidean distance corresponding to the weather station and the distance threshold;

[0016] The carbon-water flux of the weather station is determined using a carbon-water flux model applicable to the weather station.

[0017] Optionally, based on the flux station data set, according to the ecosystem combination type and in accordance with the requirements of the K-fold cross-validation model, flux station data subsets including training sets and test sets with different partitioning schemes are constructed through random sampling.

[0018] Optionally, a machine learning algorithm or a deep learning algorithm is used to construct a machine learning model or a deep learning simulation model of the eddy covariance flux tower carbon-water flux through hyperparameter optimization and K-fold cross validation.

[0019] Optionally, the carbon-water flux model corresponding to the weather station and having the minimum comprehensive Euclidean distance less than or equal to the distance threshold is determined as the carbon-water flux model applicable to the weather station.

[0020] Optionally, the method further comprises:

[0021] The carbon flux dataset of the meteorological station was constructed using the carbon-water flux model applicable to the meteorological station.

[0022] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0023] The method for determining carbon-water flux information of a meteorological station provided by the present invention is based on limited eddy covariance flux stations, integrates remote sensing big data and machine learning / deep learning technology, constructs thousands of eddy covariance flux tower carbon-water flux simulation models, and determines the carbon-water flux model applicable to the meteorological station based on the relationship between the comprehensive Euclidean distance determined by migrating the carbon-water flux model to the meteorological station and the distance threshold, so as to use the carbon-water flux model applicable to the meteorological station to simulate carbon-water flux at multiple time scales. Compared with the method of obtaining carbon-water flux based on eddy covariance flux stations, the present invention can mine high-precision carbon-water flux information of all meteorological stations in the target area, and the amount of carbon-water flux data obtained is significantly increased, and the calculation accuracy is high, while being able to meet the data volume and accuracy requirements of ecosystem carbon monitoring and carbon accounting.

[0024] The present invention also provides the following implementation structure:

[0025] A carbon-water flux information determination system for a meteorological station is applied to the carbon-water flux information determination method for a meteorological station provided above; the system comprises:

[0026] A data set acquisition module is used to acquire an observation data set in a target area; the observation data set includes observation data of an eddy covariance flux station in the target area, observation data of a meteorological station, and vegetation, soil, topography data and remote sensing data at the locations of the eddy covariance flux station and the meteorological station;

[0027] The data set processing module is used to perform quality control and standardization on the observation data set, and obtain flux station data sets respectively by using geographic information system and remote sensing technology; the flux station data set includes: carbon flux, water flux, temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine; the meteorological station data set includes: temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine; vegetation soil topography data includes soil texture, land cover type and topography; remote sensing data includes vegetation index, band spectrum, water index, surface temperature and surface humidity;

[0028] A data set partitioning module is used to selectively combine eddy covariance flux stations according to vegetation functional type, climate type and / or region based on the flux station data set, and construct a plurality of combinations and a flux station data set corresponding to each combination; each flux station data set includes a training set and a test set of flux station data subsets of different partitioning schemes;

[0029] A carbon-water flux model building module is used to use a machine learning algorithm or a deep learning algorithm, with training sets of different schemes in different combinations as input, to train an eddy covariance flux tower carbon-water flux machine learning model or a deep learning simulation model to obtain a carbon-water flux site carbon-flux model;

[0030] A determination coefficient determination module, used to use the test set as the input of the carbon-water flux model and calculate the determination coefficient using a spatial cross-validation method;

[0031] A first distance determination module is used to determine the Euclidean distance of each carbon-water flux influencing factor between the training set and the test set to obtain a comprehensive Euclidean distance;

[0032] A distance threshold determination module, used to determine the average value of all comprehensive Euclidean distances corresponding to the determination coefficient being greater than or equal to 0.5 and less than 0.55, and use the average value as the distance threshold;

[0033] A second distance determination module is used to determine the comprehensive Euclidean distance corresponding to the weather station after migrating the carbon-water flux model to the weather station;

[0034] A station applicable model determination module, used to determine a carbon-water flux model applicable to a meteorological station based on a relationship between a comprehensive Euclidean distance corresponding to the meteorological station and the distance threshold;

[0035] The carbon-water flux determination module is used to determine the carbon-water flux of the weather station by using a carbon-water flux model applicable to the weather station.

[0036] An electronic device, comprising:

[0037] Memory for storing computer programs;

[0038] A processor is connected to the memory and is used to retrieve and execute the computer program to implement the carbon-water flux information determination method for a meteorological station provided above.

[0039] Since the technical effects achieved by the above two implementation structures provided by the present invention are the same as the technical effects achieved by the method for determining carbon-water flux information of a meteorological station provided by the present invention, they will not be described in detail here. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 A flow chart of a method for determining carbon-water flux information of a meteorological station provided by the present invention;

[0042] Figure 2 This is a schematic diagram of the implementation principle of the method for determining carbon-water flux information of a meteorological station provided by the present invention. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] The purpose of the present invention is to provide a method, system and electronic equipment for determining carbon-water flux information of a meteorological station, which can use the observation data of the meteorological station to determine the carbon-water flux, increase the data volume of the carbon-water flux, and at the same time improve the accuracy of carbon monitoring and carbon accounting.

[0045] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0046] like Figure 1 and Figure 2 As shown, the method for determining carbon-water flux information of a meteorological station provided by the present invention includes:

[0047] Step 100: Obtain an observation data set in the target area. The observation data set includes observation data of the eddy covariance flux station in the target area, observation data of the meteorological station, and vegetation soil topography data and remote sensing data at the location of the eddy covariance flux station. Among them, the vegetation soil topography data and remote sensing data at the location of the eddy covariance flux station and the meteorological station can be the soil texture, land cover type, topography and corresponding remote sensing-based surface biophysical parameter data such as vegetation index, slope spectrum, water index, surface temperature, surface humidity, etc. matched by the station.

[0048] In actual application, the main purpose of this step of the present invention is to collect and process data sets for carbon and water flux monitoring and evaluation of flux stations that meet the target area. For example, the target area can be a river basin, a region, a country (such as China), a continent (such as Eurasia), or the entire land (such as the world).

[0049] Step 101: Perform quality control and standardization on the observation data set, and use geographic information system and remote sensing technology to obtain flux station data set and meteorological station data set respectively. Figure 2 The flux station dataset 1 in the data set may include: carbon and water flux datasets with remote sensing information and carbon and water flux datasets without remote sensing in the target area. For example, the flux station dataset includes: carbon flux, water flux, temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine. The meteorological station dataset is Figure 2 The meteorological point data set 2 in the target area may include: a meteorological station observation data set containing remote sensing information and a meteorological station observation data set without remote sensing information. For example, the meteorological station data set includes: temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine, etc.

[0050] Step 102: Based on the flux station data set, the eddy covariance flux stations are selectively combined according to the vegetation function type, climate type and / or region to construct multiple combinations and carbon and water flux data sets corresponding to each combination; each carbon and water flux data set includes a subset of carbon and water flux station data at different time scales of the training set and the test set of different partitioning schemes. Among them, each combination can include overall, vegetation function type (Plant Function Type, PFT), climate type (arid, semi-arid, semi-humid, humid), etc. The flux stations can be combined according to regions (continents, countries), or according to climate zones (tropical, subtropical, warm temperate, temperate, cold temperate), or according to drought index zones (humid areas, semi-humid areas, semi-arid areas, arid areas, etc.), or according to ecosystem types. Within each combination, there can be different partitioning schemes for training sets and test sets.

[0051] In the actual application process, the implementation process of this step can be: using the carbon and water flux data set containing remote sensing information and the carbon and water flux data set without remote sensing in the target area in the above step, according to the overall (overall), vegetation function type (Plant Function Type, PFT), climate type (arid, semi-arid, semi-humid, humid) and other ecosystem combination types, according to the requirements of the K-fold cross-validation model, through random sampling, hundreds of thousands of carbon and water flux data subsets of different time scales including training sets and test sets are constructed to characterize hundreds of thousands of scenarios of the ecosystem. The different time scales here can be scales such as days, months and quarters. In addition, in the process of constructing the training set and the test set, the ratio of the data in the training set and the test set can be determined according to actual needs. For example, the data ratio of the training set and the test set is 9:1, etc.

[0052] Among them, the main PFT types considered are: cropland (CRO), deciduous broadleaf forest (DBF), evergreen broadleaf forest (EBF), evergreen needleleaf forest (ENF), grassland (GRA), mixed forest (MF), savannah (SAV), shrubland (SHR), wetland (WET), etc.

[0053] Step 103: Using a machine learning algorithm or a deep learning algorithm, with training sets of different schemes in different combinations as input, train the eddy covariance flux tower carbon-water flux machine learning model or the deep learning simulation model to obtain a carbon flux site carbon-water flux model.

[0054] In practical applications, machine learning algorithms and deep learning algorithms can be used to build thousands of eddy covariance flux tower carbon and water flux machine learning models or deep learning simulation models through hyperparameter optimization and K-fold cross validation. For example, taking the random forest model algorithm as an example, thousands of eddy covariance flux tower carbon and water flux random forest simulation models (Random Forest Model, RFM) are built to simulate multi-time scale carbon and water fluxes in hundreds of scenarios of ecosystems. Among them, the determination coefficient R 2 Or relative error characterizes the accuracy of RFM on the test set.

[0055] Furthermore, the machine learning algorithm and deep learning algorithm used in the present invention can also be: long short-term memory deep learning model (Long Short-Term Memory deep learning model), random forest algorithm learning (Random Forest, RF), support vector machine (support vector machine, SVM), back-propagation artificial neural network (back-propagation artificial neural network, BP-ANN), modeling animation and rendering system (Modelling Animation And Rendering System, MARS), extreme gradient boosting (eXtreme GradientBoosting, XGB), multi-layer perceptron (Multi-Layer Perceptron, MLP), Stacking model, Boosting method and Bagging algorithm, etc.

[0056] Step 104: Using the test set as the input of the carbon-water flux model, the coefficient of determination is calculated using a spatial cross-validation method.

[0057] In practical applications, the obtained RFM is verified by using spatial cross validation (SCV) to verify the simulation accuracy of each eddy covariance flux station in the test set one by one, so as to generate thousands of determination coefficients (R 2 ). 2 The level of determines the applicability of these RFMs at the flux station. 2 The higher the value, the more similar the training set for building the model is to the flux station dataset (usually R 2 ≥0.5), otherwise it will be lower (usually R 2 <0.5).

[0058] Among them, the R of each inversion model is 2 The value depends on the geographical and ecological differences between the carbon flux influencing factors in the model training set and the corresponding influencing factors in the test set. The Euclidean distance between the influencing factors in the training set and the test set can be represented as d(x,y):

[0059]

[0060] Where x is the carbon flux influencing factor in the model training set, that is, the explanatory variable, such as daily average temperature, y is the corresponding influencing factor in the model test set, and t is the number of samples of the factor.

[0061] Step 105: Determine the Euclidean distance between each carbon and water flux influencing factor in the training set and the test set to obtain a comprehensive Euclidean distance. Carbon and water flux influencing factors may be air temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine, vegetation soil topography data and remote sensing data, etc. Among them, vegetation soil topography data includes soil texture, land cover type and topography, and remote sensing data includes vegetation index, band spectrum, water index, surface temperature and surface humidity.

[0062] In this step, the similarity between data sets consisting of the same model explanatory variables can be characterized by Euclidean distance. The determination coefficient obtained in step 102 and the Euclidean distance between the corresponding explanatory variables and the comprehensive Euclidean distance can form a data set of thousands of samples. The data set is as follows: Figure 2 The data set 3 shown in the figure has a structure as shown in Table 1.

[0063] In practical applications, the comprehensive Euclidean distance d of each model kw The calculation process is: d2,…,d u-1 , d u is the Euclidean distance between different influencing factors (explanatory variables) in the carbon flux model training set and the corresponding factors of a flux station in the model test set. Based on this, d kw =d1×a1+d2×a2+…+d u-1 ×a u-1 +d u ×a u , where a1, a2, …, a u is the importance value of the influencing factor, which can be derived from RFM.

[0064] Table 1 Determination coefficient (R) of flux inversion model 2 ) and its impact factor Euclidean distance d u Dataset Comprehensive Euclidean distance data d kw Set (dataset 3) structure table

[0065]

[0066]

[0067] Step 106: Determine the average value of the comprehensive Euclidean distance corresponding to the determination coefficient satisfying the preset condition, and use the average value as the distance threshold.

[0068] In this step, we can calculate the 2 The average value of the comprehensive Euclidean distance corresponding to the condition of <0.55 is used as the threshold (i.e., distance threshold) for whether RFM can be migrated and applied to the weather station.

[0069] Step 107: After migrating the carbon-water flux model to the weather station, determine the integrated Euclidean distance corresponding to the weather station.

[0070] The dataset of Euclidean distance between the explanatory variables of the carbon and water flux model and the same variables of the meteorological stations (i.e. Figure 2 The structure of dataset 4 is shown in Table 2.

[0071] Table 2 Influence factors of flux inversion model and meteorological station influence factors d u Euclidean distance between two entities and combined Euclidean distance d kw Dataset structure table

[0072] ID <![CDATA[d1]]> <![CDATA[d2]]> ··· <![CDATA[d u-1 ]]> <![CDATA[d u ]]> <![CDATA[d kw ]]> 1 2 ···

[0073] Step 108: Based on the relationship between the comprehensive Euclidean distance corresponding to the meteorological station and the distance threshold, determine the carbon-water flux model applicable to the meteorological station. According to the evaluation criteria between the determined comprehensive Euclidean distance and the threshold, evaluate whether the constructed flux station carbon flux inversion model can be migrated and applied to the meteorological station, or which meteorological stations can be selected to apply the flux station carbon flux inversion model. Among them, if the minimum comprehensive Euclidean distance calculated by RFM migration to a certain meteorological station is less than or equal to the threshold, then the model can be migrated to the meteorological station for carbon-water flux simulation. If the comprehensive Euclidean distance calculated by the model migration to a certain meteorological station is greater than the threshold, it is considered that the migration accuracy of the model is low and cannot be applied to the meteorological station.

[0074] At this point, a framework and method system for evaluating the applicability of the migration and application of the carbon-water flux model to grid points has been established.

[0075] Step 109: Determine the carbon and water flux of the weather station using the carbon and water flux model applicable to the weather station.

[0076] Furthermore, in order to evaluate the accuracy of regional / global scale ecosystem carbon dynamic simulation based on ecological process models / land surface process models, the present invention also uses the selected RFMs of each grid station to construct a carbon and water flux dataset of the meteorological station (i.e. Figure 2 The dataset in 5).

[0077] Finally, the carbon and water flux dataset mined from eddy covariance flux stations and meteorological stations (i.e. Figure 2 The dataset 5) was used to analyze the carbon and water dynamics and characteristics of carbon fluxes in different ecosystems and ecosystem combinations in the region at the interannual, seasonal, and daily scales.

[0078] Based on the above description, the distribution and number of global flux stations and meteorological stations, the spatial heterogeneity of geographical and climatic conditions are fully evaluated, and in-depth research on the theory, methods and technologies of carbon flux construction at meteorological stations is carried out. First, using global eddy covariance flux stations and remote sensing, ecological and geographical big data, based on machine learning, multiple sampling technology, remote sensing and geographic information technology, hundreds of carbon flux machine learning models based on flux stations in ecosystems are developed. On this basis, an adaptive assessment framework and method system for the migration and application of carbon flux inversion models to meteorological stations are developed. Suitable carbon flux intelligent models that meet the accuracy requirements and can be migrated to meteorological stations are selected to construct a carbon flux data set for meteorological stations. This data set has a "quasi-observational" attribute and can be used to monitor and calculate the carbon flux rate, process and spatiotemporal change trend of different terrestrial ecosystems.

[0079] The constructed meteorological station carbon flux dataset can also be used to evaluate the accuracy of regional / global scale ecosystem carbon dynamics simulation based on ecological process models / land surface process models, and can also be used to assimilate regional / global model simulation results to construct a higher accuracy regional / global scale carbon flux dataset.

[0080] Therefore, the present invention can deepen the carbon cycle of terrestrial ecosystems and has the potential to significantly improve the accuracy of assessment of carbon dynamic processes in ecosystems at regional / global scales.

[0081] Furthermore, the present invention also provides the following implementation structure:

[0082] A carbon-water flux information determination system for a meteorological station is applied to the carbon-water flux information determination method for a meteorological station provided above. The system comprises:

[0083] The data set acquisition module is used to acquire the observation data set in the target area; the observation data set includes the observation data of the eddy covariance flux station in the target area, the observation data of the meteorological station, and the vegetation soil topography data and remote sensing data at the locations of the eddy covariance flux station and the meteorological station;

[0084] The data set processing module is used to perform quality control and standardization on the observation data set, and obtain the flux station data set and the meteorological station data set respectively by using the geographic information system and remote sensing technology; the flux station data set includes: carbon flux, water flux, temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine; the meteorological station data set includes: temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine; the vegetation soil topography data includes soil texture, land cover type and topography; the remote sensing data includes vegetation index, band spectrum, water index, surface temperature and surface humidity;

[0085] A data set partitioning module is used to selectively combine eddy covariance flux stations according to vegetation functional type, climate type and / or region based on the flux station data set, and construct multiple combinations and flux station data sets corresponding to each combination; each flux station data set includes training set and test set flux station data subsets of different partitioning schemes;

[0086] A carbon-water flux model building module is used to use a machine learning algorithm or a deep learning algorithm, with training sets of different schemes in different combinations as input, to train an eddy covariance flux tower carbon-water flux machine learning model or a deep learning simulation model to obtain a carbon-water flux site carbon-flux model;

[0087] A determination coefficient determination module is used to use the test set as the input of the carbon-water flux model and calculate the determination coefficient using a spatial cross-validation method;

[0088] A first distance determination module is used to determine the Euclidean distance between each carbon-water flux influencing factor in the training set and the test set to obtain a comprehensive Euclidean distance;

[0089] A distance threshold determination module is used to determine the average value of all comprehensive Euclidean distances corresponding to the determination coefficient being greater than or equal to 0.5 and less than 0.55, and use the average value as the distance threshold;

[0090] The second distance determination module is used to determine the comprehensive Euclidean distance corresponding to the weather station after the carbon-water flux model is migrated to the weather station;

[0091] A station applicable model determination module is used to determine the carbon-water flux model applicable to the meteorological station based on the relationship between the comprehensive Euclidean distance corresponding to the meteorological station and the distance threshold;

[0092] The carbon-water flux determination module is used to determine the carbon-water flux of the weather station by using the carbon-water flux model applicable to the weather station.

[0093] An electronic device, comprising:

[0094] Memory, used to store computer programs.

[0095] The processor is connected to the memory and is used to retrieve and execute a computer program to implement the carbon-water flux information determination method for a meteorological station provided above.

[0096] In addition, when the computer program in the above-mentioned memory is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory, a random access memory, a magnetic disk or an optical disk.

[0097] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0098] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for determining carbon-water flux information of a meteorological station, characterized in that: include: Obtaining observation data sets in the target area; The observation data set includes observation data of eddy covariance flux stations in the target area, observation data of meteorological stations, and vegetation soil topography data and remote sensing data at the locations of the eddy covariance flux stations and meteorological stations; Performing quality control and standardization on the observation data set, and using geographic information system and remote sensing technology to obtain flux station data set and meteorological station data set respectively; The flux station data set includes: carbon flux, water flux, temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine; the meteorological station data set includes: temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine; vegetation soil topography data includes soil texture, land cover type and topography; remote sensing data includes vegetation index, band spectrum, water index, surface temperature and surface humidity; Based on the flux station data set, eddy covariance flux stations are selectively combined according to vegetation functional type, climate type and / or region to construct multiple combinations and flux station data sets corresponding to each combination; each flux station data set includes training set and test set flux station data subsets of different partitioning schemes; Using a machine learning algorithm or a deep learning algorithm, with training sets of different schemes in different combinations as input, the eddy covariance flux tower carbon-water flux machine learning model or the deep learning simulation model is trained to obtain a carbon flux site carbon-water flux model; The test set is used as the input of the carbon-water flux model, and the coefficient of determination is calculated using a spatial cross-validation method; Determine the Euclidean distance of each carbon-water flux influencing factor between the training set and the test set to obtain a comprehensive Euclidean distance; Determine the average value of all comprehensive Euclidean distances corresponding to the determination coefficient being greater than or equal to 0.5 and less than 0.55, and use the average value as the distance threshold; After migrating the carbon-water flux model to the weather station, determining the integrated Euclidean distance corresponding to the weather station; Determining a carbon-water flux model applicable to the weather station based on a relationship between the comprehensive Euclidean distance corresponding to the weather station and the distance threshold; The carbon-water flux of the weather station is determined using a carbon-water flux model applicable to the weather station.

2. The method for determining carbon-water flux information of a meteorological station according to claim 1, characterized in that: Based on the flux station data set, according to the ecosystem combination type and the requirements of the K-fold cross-validation model, carbon and water flux station data subsets of different time scales including training sets and test sets with different partitioning schemes are constructed through random sampling.

3. The method for determining carbon-water flux information of a meteorological station according to claim 1, characterized in that: Using machine learning algorithms or deep learning algorithms, through hyperparameter optimization and K-fold cross validation, a machine learning model or deep learning simulation model of eddy covariance flux tower carbon and water flux is constructed.

4. The method for determining carbon-water flux information of a meteorological station according to claim 1, characterized in that: The carbon-water flux model corresponding to the weather station and having the minimum comprehensive Euclidean distance less than or equal to the distance threshold is determined as the carbon-water flux model applicable to the weather station.

5. The method for determining carbon-water flux information of a meteorological station according to claim 1, characterized in that: The method further comprises: The carbon flux dataset of the meteorological station was constructed using the carbon-water flux model applicable to the meteorological station.

6. A carbon-water flux information determination system for a meteorological station, characterized in that: A method for determining carbon-water flux information of a meteorological station as claimed in any one of claims 1 to 5; the system comprising: A data set acquisition module is used to acquire an observation data set in a target area; the observation data set includes observation data of an eddy covariance flux station in the target area, observation data of a meteorological station, and vegetation, soil, topography data and remote sensing data at the locations of the eddy covariance flux station and the meteorological station; A data set processing module is used to perform quality control and standardization on the observation data set, and obtain a flux station data set and a meteorological station data set by using a geographic information system and remote sensing technology respectively; the flux station data set includes: carbon flux, water flux, temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine; the meteorological station data set includes: temperature, precipitation, air pressure, relative humidity, wind speed, wind direction and sunshine; vegetation soil topography data includes soil texture, land cover type and topography; remote sensing data includes vegetation index, band spectrum, water index, surface temperature and surface humidity; A data set partitioning module is used to selectively combine eddy covariance flux stations according to vegetation functional type, climate type and / or region based on the flux station data set, and construct multiple combinations and flux station data sets corresponding to each combination; each flux station data set includes training set and test set flux station data subsets of different partitioning schemes; A carbon-water flux model building module is used to use a machine learning algorithm or a deep learning algorithm, with training sets of different schemes in different combinations as input, to train an eddy covariance flux tower carbon-water flux machine learning model or a deep learning simulation model to obtain a carbon-water flux site carbon-flux model; A determination coefficient determination module, used to use the test set as the input of the carbon-water flux model and calculate the determination coefficient using a spatial cross-validation method; A first distance determination module is used to determine the Euclidean distance of each carbon-water flux influencing factor between the training set and the test set to obtain a comprehensive Euclidean distance; A distance threshold determination module, used to determine the average value of all comprehensive Euclidean distances corresponding to the determination coefficient being greater than or equal to 0.5 and less than 0.55, and use the average value as the distance threshold; A second distance determination module is used to determine the comprehensive Euclidean distance corresponding to the weather station after migrating the carbon-water flux model to the weather station; A station applicable model determination module, used to determine a carbon-water flux model applicable to a meteorological station based on a relationship between a comprehensive Euclidean distance corresponding to the meteorological station and the distance threshold; The carbon-water flux determination module is used to determine the carbon-water flux of the weather station by using a carbon-water flux model applicable to the weather station.

7. An electronic device, characterized in that: include: Memory for storing computer programs; A processor is connected to the memory and is used to retrieve and execute the computer program to implement the method for determining carbon and water flux information of a meteorological station according to any one of claims 1 to 5.

Images