Water-gas interface greenhouse gas flux automatic monitoring system and distribution prediction method

Abstract

The embodiment of the invention discloses a water-gas interface greenhouse gas flux automatic monitoring system and a distribution prediction method. The system comprises a wind speed sensor and an air pressure and temperature sensor which are respectively used for collecting the wind speed of air at a monitoring point, the atmospheric pressure and the Kelvin air temperature; the infrared sensor is used for collecting the content of CO2, CH4 and N2O in the air of the monitoring point; the acidity sensor and the temperature sensor are respectively used for collecting the pH value and the centigrade water temperature of monitoring point water; the carbon dioxide sensor is used for collecting the concentration of bicarbonate in water at a monitoring point; the methane sensor and the nitrous oxide sensor are respectively used for collecting the contents of greenhouse gases CH4 and N2O dissolved in the monitoring point water; and the calculation module is used for calculating the flux of the greenhouse gas on the water-gas interface. According to the embodiment of the invention, the water-gas interface flux of various greenhouse gases under actual conditions can be accurately monitored.

Description

Technical Field

[0001] The embodiments of the present invention relate to the technical field of gas flux monitoring, and in particular, to an automatic monitoring system and a distribution prediction method for greenhouse gas fluxes at the water-air interface. Background Art

[0002] The long-term monitoring and estimation of water temperature greenhouse gas fluxes not only have great significance for studying the biogeochemical cycles of carbon and nitrogen and global climate change, but also can provide effective data support for predicting the carbon sequestration capacity of water bodies on a longer time scale and promoting the ecological governance of water environments.

[0003] Currently, the observation equipment for water-air greenhouse gas fluxes in water bodies mostly focuses on carbon fluxes. For example, patent application CN101852723A provides an on-line automatic monitoring system and a monitoring method for carbon fluxes at the water-air interface of inland water bodies. This method lacks the monitoring of the fluxes of two important greenhouse gases in water bodies - methane and nitrous oxide, and these two gases have a greater global warming potential, which are 25 times and 298 times that of carbon dioxide respectively. At the same time, in this method, the gas flux is calculated using the gas partial pressure. Since the gas partial pressure refers to the pressure formed by a certain component in a gas mixture when it occupies the same volume of the gas mixture at the same temperature, the gas flux calculated by the gas partial pressure is assumed to be the gas flux under standard atmospheric pressure, rather than the gas flux at the actual air pressure and temperature at the water-air interface, and there is still a difference from the actual situation.

[0004] In addition, in the prior art, the observation equipment for the fluxes of greenhouse gases carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O) in water bodies mostly adopts the method of collecting water samples or extracting greenhouse gases in water bodies, and then using a gas analyzer for testing. For example, patent application CN221350771U provides a greenhouse gas emission sampling device, which changes the original water environment condition, resulting in a change in the Henry equilibrium of gas-liquid exchange and destroying the inherent water-air interface equilibrium mode in the water body, and will cause a huge error between the measured data and the actual situation. Summary of the Invention

[0005] The embodiments of the present invention provide an automatic monitoring system and a distribution prediction method for greenhouse gas fluxes at the water-air interface to accurately monitor and predict the water-air interface fluxes of various greenhouse gases under actual conditions.

[0006] In a first aspect, the embodiments of the present invention provide an automatic monitoring system for greenhouse gas fluxes at the water-air interface, which is applied to inland water bodies. The system includes: A wind speed sensor and a barometric pressure and air temperature sensor, which are respectively used to collect the wind speed of the air at the monitoring point , and the atmospheric pressure , and the Kelvin air temperature ; An infrared sensor for collecting the contents of CO2, CH4, and N2O in the air at the monitoring point ; An acidity sensor and a temperature sensor for collecting the pH value of the water at the monitoring point and the Celsius water temperature ; A carbon dioxide sensor for collecting the concentration of bicarbonate in the water at the monitoring point ; A methane sensor and a nitrous oxide sensor for collecting the contents of dissolved greenhouse gases CH4 and N2O in the water at the monitoring point ; A calculation module for calculating the flux of greenhouse gases at the water-air interface in the following manner: According to , calculate the solubility of CO2, CH4, and N2O in the surface water , the water-air exchange rate , and the Henry's equilibrium constant of the bicarbonate concentration in the water body ; According to the following formula, calculate the concentrations of dissolved CO2, CH4, and N2O in the water body at the monitoring point :

[0007] According to the following formula, calculate the concentrations of CO2, CH4, and N2O in the air at the water surface of the water body at the monitoring point :

[0008] wherein, is the correction coefficient of the ideal gas state equation for correcting the ideal gas equation to the actual gas equation; According to the following formula, calculate the fluxes of CO2, CH4, and N2O at the water-air interface per unit area and per unit time in the water area at the monitoring point : .

[0009] In a second aspect, an embodiment of the present invention provides a method for predicting the distribution of greenhouse gas fluxes at the water-air interface, including: S210. Divide the water area plane to be monitored into multiple grids; S220. Take each grid as a monitoring point, respectively use the above-mentioned automatic monitoring system for greenhouse gas fluxes at the water-air interface to monitor the flux of any one greenhouse gas at the water-air interface of each grid in real time; and take the fluxes of each grid at the same moment as samples to construct a finite sample set; S230. Average the flux distributions monitored over a period of time, and take multiple extreme value grids and large gradient grids in the averaged flux distribution as multiple initial grids; and construct a neural network model with the fluxes of the multiple initial grids at the same moment as the input and the flux distribution of all grids at the same moment as the output. S240. Use the finite sample set to train and test the neural network model. If the model accuracy does not meet the requirements, cluster the grids with the difference between the predicted flux value and the monitored value exceeding the set threshold in the test results according to the geographical location, and incorporate the cluster centers into the model input. S250. Return to S240 according to the new neural network model, and repeat this process until the model accuracy meets the requirements. S260. Use the genetic algorithm to select the optimal input grid combination that can meet the model accuracy requirements and has the least number of grids from the input grid set of the final model, and use the optimal model corresponding to the optimal input grid combination to predict the water-air interface flux distribution.

[0010] In summary, the embodiments of the present invention disclose an automatic monitoring system and a distribution prediction method for greenhouse gas fluxes at the water-air interface. The system can achieve the following beneficial effects: 1. Through advanced infrared sensors, methane sensors, and nitrous oxide sensors, the contents of two important greenhouse gases, methane and nitrous oxide, are monitored in the air and water bodies, providing comprehensive data support for the monitoring of greenhouse gas fluxes in inland water bodies. 2. Aiming at the problem that the gas fluxes calculated using gas partial pressures are not the gas fluxes at the actual air pressure and temperature at the water-air interface, in this embodiment, the fluxes of each greenhouse gas are calculated through gas concentrations. When calculating the gas concentrations, the correction coefficient of the ideal gas state equation is introduced, considering the influence of the actual air pressure and temperature at the water body interface on the gas fluxes, so as to obtain more accurate gas flux data. 3. The entire system does not collect water or gas samples, does not change the original water environment conditions, and maintains the gas-liquid exchange equilibrium mode of the water body itself, further ensuring that the gas flux data is consistent with the actual situation. 4. Based on the flux distribution prediction method of the automatic monitoring system for greenhouse gas fluxes at the water-air interface, through the combination of partial point monitoring and overall water area prediction, using as few automatic monitoring devices for greenhouse gas fluxes at the water-air interface as possible, the greenhouse gas flux distribution of the entire water area is obtained, reducing the number of devices used for long-term monitoring, improving the device utilization rate, and reducing the monitoring cost of greenhouse gas fluxes. Description of the Drawings

[0011] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0012] Figure 1 It is a schematic diagram of an automatic monitoring system for greenhouse gas fluxes at the water-air interface provided by an embodiment of the present invention; Figure 2 It is an installation schematic diagram of an on-line automatic monitoring device provided by an embodiment of the present invention; Figure 3 It is a flowchart of a method for predicting the distribution of greenhouse gas fluxes at the water-air interface provided by an embodiment of the present invention; Figure 4 It is a schematic diagram of grid division of a water area to be monitored provided by an embodiment of the present invention; Reference numerals: 1. Power supply device; 2. Warning lamp; 3. Industrial control computer; 4. Signal transmitter; 5. Wind speed sensor; 6. Air pressure and temperature sensor; 7. Infrared sensor; 8. Acidity sensor; 9. Temperature sensor; 10. Carbon dioxide sensor; 11. Methane sensor; 12. Nitrous oxide sensor. Specific embodiments

[0013] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0014] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, the terms "first", "second", "third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0015] In the description of the present invention, it should also be noted that unless otherwise clearly specified and limited, the terms "installed", "connected", "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0016] Figure 1 is a schematic diagram of an automatic monitoring system for greenhouse gas fluxes at the water-air interface provided by an embodiment of the present invention. As Figure 1 shown, the system includes an on-line automatic monitoring device and a calculation module.

[0017] Figure 2 is an installation schematic diagram of an on-line automatic monitoring device provided by an embodiment of the present invention. As Figure 2 shown, the on-line automatic monitoring device mainly includes a floating box, a power supply device 1, a series of sensors and an industrial control computer 3. The power supply device 1 mainly provides electrical energy for the on-line automatic monitoring device through the upper photovoltaic panel and the battery under the panel. The warning light 2 is used to prevent collisions with water vehicles.

[0018] During use, the floating box can be fixed at the position of the water body to be monitored in the field by using reinforced concrete anchor points. Through the float type automatic regulator of the floating box, the probes of the acidity sensor 8, the temperature sensor 9, the carbon dioxide sensor 10, the methane sensor 11 and the nitrous oxide sensor 12 are located at a depth of 5-10 cm below the water surface. The on-line automatic monitoring device is powered by photovoltaic solar panels. The monitoring principle of greenhouse gas fluxes is the water chemical ion balance relationship and Henry's law in the water body. The data acquisition module of the industrial control computer 3 collects and calculates the sensor signals, and then uses the signal transmitter 4 of the remote communication module to output the result data (wired RS485 communication interface and wireless GPRS transmission).

[0019] Before installing the on-line automatic monitoring device at the field site, first use pH buffer solutions (2.00, 4.01, 7.00, 9.21 and 10.00) to perform pH electrode standard calibration on the acidity sensor 8 to accurately measure the pH value (unit: dimensionless) of the water environment parameters of the surface water body; then use NaHCO3 standard solutions (0.1 mM, 0.5 mM, 1.0 mM, 5.0 mM and 10.0 mM, mM represents millimole per liter) to perform ion electrode standard calibration on the carbon dioxide sensor 10 to accurately measure the content of bicarbonate in the surface water body (unit: mM).

[0020] The on-line automatic monitoring device uses the accurate signal (±0.01°) transmitted by the temperature sensor 9 to automatically perform temperature compensation and correction on the signals of the acidity sensor 8 and the carbon dioxide sensor 10. The signal transmitted by the wind speed sensor 5 (model Pro10) is the wind speed of the air at the monitoring point (unit: m / s, meter per second), and the signals transmitted by the barometric pressure and air temperature sensor 6 (model BME280) are respectively the atmospheric pressure (unit: atm) and the Kelvin temperature (unit: ° K ) of the air at the monitoring point. The signals transmitted by the infrared sensor 7 (model SGA-900) are respectively the contents of CO2, CH4 and N2O in the air at the monitoring point (unit: ppm). The signal transmitted by the acidity sensor 8 (model Orion 9815JP pH) is the pH value of the water at the monitoring point (unit: dimensionless). The signal transmitted by the temperature sensor 9 (model PT100) is the Celsius water temperature at the monitoring point (unit: °C). The signal transmitted by the carbon dioxide sensor 10 (model Orion 9502BNWP) is the concentration of bicarbonate in the water at the monitoring point (unit: mM). The signal transmitted by the methane sensor 11 (model Franatech METS) is the content of dissolved greenhouse gas CH4 in the water at the monitoring point (unit: ppm). The signal transmitted by the nitrous oxide sensor 12 (model Unisense full electrode) is the content of dissolved greenhouse gas N2O in the water at the monitoring point (unit: ppm).

[0021] The calculation module can be set on the user's computer. After receiving the real-time on-line automatic monitoring data sent by the communication base station, it gradually calculates through the following method to determine the greenhouse gas flux per unit time and per unit area of the water-air interface of the water body: S110. Calculate the solubility of CO2, CH4 and N2O in the surface water (unit: mole·L -1 ·atm -1), water-air exchange rate (unit, cm·h -1 , centimeters per hour) and the Henry's law equilibrium constant of bicarbonate concentration in the water body (unit, atm‧L‧mol -1 ).

[0022] Specifically, the signals of the wind speed sensor 5 and the temperature sensor 9 collected are calculated to obtain the changes in the fluxes of CO2, CH4, and N2O at the water-air interface caused by different physicochemical actions. The calculation formula is as follows:

[0023] In the formula, is the Kelvin temperature (unit, ° K ), which is calculated through . .

[0024] S120. Calculate the concentrations of dissolved CO2, CH4, and N2O in the water body at the monitoring point (unit, μM, micromoles):

[0025] S130. Calculate the concentrations of CO2, CH4, and N2O in the air above the water surface at the monitoring point (unit, μM):

[0026] Among them, , is the correction coefficient of the ideal gas state equation, which is used to correct the ideal gas equation to the actual gas equation; S140. Calculate the greenhouse gas flux per unit area and per unit time at the water-air interface of the monitoring point (unit, mg·m -2 ·h -1 ):

[0027] In summary, this embodiment provides an automatic monitoring system for greenhouse gas fluxes at the water-air interface, which can achieve the following beneficial effects: 1. Through advanced infrared sensors, methane sensors, and nitrous oxide sensors, the content monitoring of two important greenhouse gases - methane and nitrous oxide - is realized in the air and water body, providing comprehensive data support for the monitoring of greenhouse gas fluxes in inland water bodies; 2. Regarding the problem that the gas flux calculated using gas partial pressure in the patent application CN101852723A, "An Inland Water Body Water-Air Interface Carbon Flux On-line Automatic Monitoring System and Its Monitoring Method", is not the gas flux at the actual air pressure and temperature at the water-air interface, in this embodiment, the flux of each greenhouse gas is calculated through gas concentration. When calculating the gas concentration, a correction coefficient of the ideal gas state equation is introduced, considering the influence of the actual air pressure and temperature at the water body interface on the gas flux, so as to obtain more accurate gas flux data; 3. The entire system does not collect samples of water bodies or gases, does not change the original water environment conditions, maintains the gas-liquid exchange equilibrium mode of the water body itself, and further ensures that the gas flux data is consistent with the actual situation.

[0028] Based on the above water-air interface greenhouse gas flux automatic monitoring system, Figure 3 is a flowchart of a method for predicting the distribution of greenhouse gas fluxes at the water-air interface. This method can predict the distribution of greenhouse gas fluxes in the entire water area through as few monitoring points as possible, reducing the number of in-service devices in long-term monitoring. As Figure 3 shown, this method specifically includes: S210. Divide the plane of the water area to be monitored into multiple grids.

[0029] Figure 4 Taking a regular rectangle as an example, a top view of the plane of a water area to be monitored is shown. As shown in the figure, this plane can be divided into multiple grids for zonal monitoring. For an irregularly shaped water area plane, the division method is similar.

[0030] S220. Take each grid as a monitoring point, and install the above-mentioned water-air interface greenhouse gas flux automatic monitoring system respectively to monitor the flux of any greenhouse gas at the water-air interface of each grid in real time; and take the fluxes of each grid at the same moment as samples to construct a finite sample set.

[0031] Combined with Figure 2 , in each grid, in the manner shown in Figure 2 , install on-line automatic monitoring devices respectively to transmit various sensor data collected in real time. Each on-line automatic monitoring device can share a calculation module to calculate the greenhouse gas flux of each grid at each moment in real time.

[0032] After monitoring for a period of time, for any greenhouse gas, the fluxes of the greenhouse gas in each grid at the same moment in the monitoring data are taken as a sample. Multiple samples can be obtained from the data of multiple historical moments, forming a finite sample set. The finiteness of the sample set is emphasized here because for a large water area, the cost of installing an online automatic monitoring device for each grid is very high. Therefore, in this embodiment, only a limited duration of sample collection can be achieved while ensuring that all grids have equipment. In the subsequent steps, a prediction model of the greenhouse gas flux distribution will be trained using the limited samples, and the flux distribution of the entire water area will be predicted through the monitored fluxes of some grids, so as to reduce the number of long-term in-service devices and lower the flux monitoring cost of the entire water area.

[0033] S230. Average the flux distribution monitored for a period of time, and take multiple extreme value grids and large gradient grids in the averaged flux distribution as multiple initial grids; and construct a neural network model with the fluxes of the multiple initial grids at the same moment as the input and the flux distribution of all grids at the same moment as the output.

[0034] This period of time can be selected as a period when the wind speed, temperature, etc. are relatively stable. The flux distribution at each moment during this period is regarded as a two-dimensional matrix arranged according to the geographical location of the grids (in an irregular water area, the grid elements corresponding to the land positions are permanently set to 0). By adding the two-dimensional matrices of multiple moments and dividing by the number of moments in this period, the average flux distribution in this period can be obtained. From the average flux distribution, select the grids where the extreme points are located and the grids where the change gradient exceeds the set threshold, and jointly form the initial set of the input grids of the neural network model. For the convenience of description, the grids in these initial sets are all referred to as initial grids in this embodiment. The reason for selecting these grids is that these grids often represent the key positions where the flux distribution of the entire water area turns or mutates, and can provide key information for predicting the flux distribution of the entire water area later.

[0035] Based on this initial set, a neural network model is constructed. In this embodiment, the model will be trained so that after inputting the greenhouse gas fluxes of these initial grids at the same moment, it can output the greenhouse gas flux distribution of all grids at this moment. Optionally, the model can adopt a multi-layer convolutional neural network, and a fully connected layer is connected after the convolutional neural network to generate results of a specific dimension; the dimensions of the input layer and the output layer can be adjusted according to the quantity of the input data and the output data; the number of convolutional layers is greater than 7 layers, and the convolutional kernel can adopt 3×3 to expand or extract data features layer by layer; the fully connected layer is used to connect the dimensions of the last convolutional layer and the output layer.

[0036] S240. Use the finite sample set to train and test the neural network model. If the model accuracy does not meet the requirements, cluster the grids where the difference between the flux prediction value and the monitored value in the test result exceeds the set threshold according to the geographical location, and incorporate the cluster centers into the model input.

[0037] As described above, the model needs to learn the data relationship between the fluxes of some grids and all grids. Due to the large amount of missing grid data, the learning difficulty of the model is very high. In the case of limited sample size, the accuracy of the trained model is likely to fall short of the ideal requirements. Therefore, in this embodiment, when the model accuracy is poor, by increasing the number of input grids, more data information is provided for the model, and the learning difficulty is gradually reduced.

[0038] In a specific embodiment, to make full use of the limited sample data, the model training can be completed through the following steps: Step 1. Divide the finite sample set into a first training set, a second training set, and a test set. Among them, the first training set and the second training set are used to correspond to and distinguish the subsequent two training stages. The samples in the two training sets can be the same, different, or partially the same, and this embodiment does not make specific restrictions.

[0039] Step 2. Use the first training set to perform a first training on the neural network model. In the first training, update the model parameters by minimizing the flux change between adjacent grids. This step corresponds to the first training stage of the model. In this stage, the model parameters are updated using the law followed by the input data itself, so that the model output has a certain law. As can be seen from the above system embodiment, parameters such as gas concentration and temperature environment used to calculate the flux are continuously changing in space. Therefore, the change of greenhouse gas flux along the water-air interface should also be relatively uniform. Therefore, in this embodiment, the model parameters are updated by constraining the minimum flux change between adjacent grids to minimize the random changes without rules as much as possible. Optionally, the following loss function can be constructed:

[0040] Where represents the loss function value, and respectively represent the grid indices along the horizontal and vertical directions, represents the th column grid flux prediction value of the th row, represents the sum of all grids on the water-air interface (excluding land grids with a permanent flux of 0),

[0041] Step 3: Use the second training set to perform secondary training on the model after the first training. During the secondary training, update the model parameters by minimizing the difference between the predicted flux values and the monitored values of each grid. This step corresponds to the second training stage of the model. In this stage, the actually monitored flux data is used as the true value, and the model output is constrained to continuously approach the true value to perform supervised learning on the model after the first training. Optionally, the following loss function can be constructed:

[0042] where represents the loss function value, represents the th column grid flux true value.

[0043] Furthermore, in this embodiment, it is also considered that due to equipment failures or communication failures in some grids, the true value data is empty. In this case, although the sample data lacks the data of some grids, it can still be used for the secondary training of the model. It only needs to simultaneously constrain the minimum change in the flux between the partial grids and the adjacent grids, and the minimum difference between the predicted flux values and the monitored values of other grids in the loss function. Optionally, the following loss function can be constructed for such samples:

[0044] where represents the loss function value, i 1 and j 1 represent the two-dimensional indices of the grids with complete data, i 2 and j 2 represent the two-dimensional indices of the grids with missing data, represent the weight coefficients respectively. In this way, for samples with complete data, can be used to update the network parameters. For samples with missing data, can be used to update the network parameters, further avoiding the waste of sample data.

[0045] In the above two training stages, the model first learns the laws of the input data itself, and then performs one-to-one supervised training through the true values of each grid. Compared with directly using the true value data for supervised training, the training method of this embodiment can accelerate the model convergence speed and achieve as good a training effect as possible using a limited sample set.

[0046] Step 4: After the training is completed, use the test set to test the model after the secondary training. If the model accuracy still does not meet the requirements, cluster the grids in the test results where the difference between the flux prediction value and the monitored value exceeds the set threshold according to the geographical location, and incorporate the cluster centers into the model input. This step is aimed at the situation where the model accuracy has not reached a good level after two training stages. Select the grids with large prediction errors from the prediction results of each test sample as the input grids of the model.

[0047] Optionally, if large errors occur in some grids in multiple test samples, cluster these grids according to the geographical location to obtain multiple spatial clusters. Each spatial cluster represents an area with relatively close distances. The grids in this area have large measurement errors multiple times, indicating that the model cannot effectively learn the data law of this area from the existing input grids. Therefore, it is necessary to select one or more grids from this area as the model input to provide effective information of this area for the model. Optionally, the DBSCAN clustering method can be used, and finally select the cluster center of each cluster as the model input.

[0048] Furthermore, when adding new model inputs, on the basis of the original model, convolutional kernels connected to the new network and subsequent calculation layers can be added in the first convolutional layer; or a fully connected layer can be added between the input layer and the first convolutional layer to transform the dimension of the new input layer to the dimension of the original hidden layer; the remaining structures in the model remain unchanged and inherit the original parameters as the initial parameters.

[0049] S250: Return to S240 according to the new neural network model, and repeat this process until the model accuracy meets the requirements.

[0050] Specifically, use the finite sample set again to train and test the new neural network model; if the accuracy of the new model still does not meet the requirements, cluster the grids in the test results where the difference between the flux prediction value and the monitored value exceeds the set threshold according to the geographical location again, and incorporate the cluster centers into the model input to obtain an updated neural network, and then return to S240. Repeat this process until the accuracy of the final model reaches the required index. At this time, the neural network model has sufficient effective grid data. According to these data and the variation law of the flux itself, the flux distribution of the entire water surface can be predicted more accurately. For the convenience of distinction and description, the neural network model obtained at this time is called the full-scale model.

[0051] S260: Use the genetic algorithm to select the optimal input grid combination from the input grid set of the full-scale model that can meet the model accuracy requirements and has the smallest number of grids, and use the optimal model corresponding to the optimal input grid combination to predict the water-air interface flux distribution.

[0052] As described above, in multiple cycles of S240 - S250, the model accuracy is improved by continuously increasing the number of input grids. To prevent redundancy in the final input grids, this step uses a genetic algorithm to screen the input grids of the full - scale model to remove redundant grids, and predicts the flux distribution of the entire water plane with the least model inputs and the greenhouse gas flux automatic monitoring system.

[0053] In a specific implementation, the above - mentioned screening process can be completed in the following way: Step 1: According to the number of grids in the full - scale model input grid set, construct a chromosome sequence structure of equal length. Exemplarily, if the number of input grids of the full - scale model is 15, a vector of length 15 can be constructed as the chromosome sequence structure in the genetic algorithm. Among them, each bit in the vector corresponds to a grid in the input grid set; when the numerical bit takes 1, it represents retaining the corresponding grid as the input of the neural network model, and taking 0 represents deleting the corresponding grid from the input grids of the neural network model.

[0054] Step 2: Generate the current population of the genetic algorithm according to the chromosome sequence structure. Exemplarily, a random initialization method can be used to generate multiple chromosomes to form the current population.

[0055] Step 3: Prune the full - scale model according to each chromosome in the current population, and use the finite sample set to train the pruned model three times. Specifically, for a certain chromosome, delete the grids corresponding to the bits with a value of 0 in the chromosome from the input grids of the full - scale model, and delete all branches connecting the first - layer hidden layer of the model to the input grids to obtain a new model; then, use the finite sample set to train this new model. In this embodiment, this training is called three - time training.

[0056] Step 4: Determine whether the model after three - time training meets the optimal goal. If not, perform selection, crossover, and mutation on the current population according to the genetic algorithm, select excellent chromosomes from the mutated current population to form a new current population, and return to Step 3. Prune the full - scale model according to each chromosome in the new current population, and use the finite sample set to train the pruned model three times again. Repeat this cycle until the model after the final three - time training meets the optimal goal. Among them, the optimal goal is that the accuracy of the model after three - time training meets the requirements, and the model with the least number of input grids has been obtained in the three - time training carried out so far.

[0057] Optionally, after each chromosome that meets the model accuracy requirement is obtained, one with the least number of input grids among all the chromosomes that meet the model accuracy requirement accumulated currently can be selected as the optimal chromosome; if the optimal chromosome remains unchanged after three consecutive trainings, it is considered that the model with the least number of input grids has been obtained. Of course, it is also possible to directly iterate to the maximum number of iterations, and after the iteration terminates, select the model with the model accuracy meeting the requirement and the least number of input grids as the optimal model.

[0058] Furthermore, when selecting excellent chromosomes, a fitness function can be constructed according to the change patterns of the predicted air flux values between the input grids in the chromosome; according to the fitness function, excellent chromosomes are selected from the current population after mutation to construct a new current population. Optionally, the following fitness function can be constructed:

[0059] where is the fitness value, represents the number of extreme value grids in the flux distribution predicted by the model after three trainings for the input grids with a value of 1 in the current chromosome, is the two-dimensional index of the input grids with a value of 1 in the current chromosome, represents the number of grids with a value of 0 in the current chromosome; respectively represent the weight coefficients. The larger the fitness value, the more excellent the chromosome.

[0060] Through the above fitness function, it is possible to preferentially select chromosomes with a relatively small number of input grids and the input grids covering more extreme value grids and large gradient grids (these grids represent the key morphological changes of the flux distribution) to evolve the new current population, which helps to obtain the optimal chromosome and the optimal model as soon as possible.

[0061] Finally, after obtaining the optimal input grid combination, only the automatic monitoring systems for the water-air interface greenhouse gas fluxes of the grids in this combination are retained, and the automatic monitoring systems for the water-air interface greenhouse gas fluxes of the remaining grids are removed; the air fluxes of each target grid are monitored in real time by using the remaining systems and input into the optimal model, and then the real-time flux distribution of all grids can be obtained. The removed equipment systems can be used for the flux monitoring collective prediction of other water areas, improving the equipment utilization rate and reducing the monitoring cost of greenhouse gas fluxes.

[0062] Further, taking a certain greenhouse gas as an example, the above method illustrates a specific method for predicting the greenhouse gas flux distribution of the entire water area using as few monitoring points as possible. In practical applications, operations of S220 - S260 can be performed for each greenhouse gas. Among them, when predicting the water - air interface flux distribution using the optimal model corresponding to the optimal input grid combination in S260, the union of the optimal input grid combinations corresponding to various greenhouse gases is taken, and the automatic monitoring systems for the water - air interface greenhouse gas fluxes of each target grid in the union are retained, while the automatic monitoring systems for the water - air interface greenhouse gas fluxes of the remaining grids are removed. The greenhouse gas fluxes of each target grid are monitored in real time using the remaining systems and are respectively input into the optimal models corresponding to various greenhouse gases to obtain the real - time greenhouse gas flux distributions of all grids. In this way, the flux distributions of three greenhouse gases in the entire water area can be predicted using as few monitoring points as possible.

[0063] It should be noted that in the above - mentioned method for predicting the greenhouse gas flux distribution at the water - air interface, except for the steps of installing and removing the in - situ monitoring system for the greenhouse gas flux at the water - air interface, the remaining steps can be automatically executed by an electronic device.

[0064] In summary, this embodiment also provides a method for predicting the greenhouse gas flux distribution at the water - air interface, which can achieve the following beneficial effects: 1. By combining partial - point monitoring and overall - water - area prediction, using as few automatic monitoring devices for the greenhouse gas flux at the water - air interface as possible, the greenhouse gas flux distribution of the entire water area is obtained, reducing the number of devices used for long - term monitoring, improving the device utilization rate, and reducing the monitoring cost of the greenhouse gas flux. 2. Since the neural network model needs to learn the complex flux distribution law based on partial grids, the learning difficulty is large and the sample size is limited, and it is difficult for the trained model accuracy to reach the ideal requirement at one time. To address this difficulty, in this embodiment, when the model accuracy is not good, representative grids are selected from the regions with large prediction errors as model inputs. By increasing the number of input grids, more effective information is provided for the model, gradually reducing the learning difficulty. 3. For the case where the full - grid data samples are limited, this embodiment enables the model to first learn the law of the input data itself through multi - stage training, and then perform one - to - one supervised training through the monitoring values of each grid. Compared with directly using the monitoring data for supervised training, this method can accelerate the model convergence speed and achieve as good a training effect as possible using a limited sample set. 4. For the data - missing samples caused by equipment or communication failures in some grids, this embodiment combines the uniform change of the flux and the monitoring values, adopts different loss - value calculation methods for different grids, fully utilizes the remaining data information in the samples, and further avoids sample waste. 5. After obtaining the full-scale model including sufficient valid input grids, in order to prevent redundancy in the input grids of the full-scale model, in this embodiment, a genetic algorithm is used to screen the input grids and remove the redundant grids therein, so as to predict the flux distribution of the entire water plane with the least model input and the greenhouse gas flux automatic monitoring system. This can greatly reduce the number of final devices and the monitoring cost in the case of a very large water area and predicting the flux distribution of each greenhouse gas.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present invention.

Claims

1. An automatic monitoring system for greenhouse gas fluxes at the water-air interface, characterized in that, Applied to inland waters, the system includes: A wind speed sensor and a barometric pressure and air temperature sensor are respectively used to collect the wind speed of the air at the monitoring point , as well as the atmospheric pressure and the Kelvin air temperature ; An infrared sensor for collecting the contents of CO2, CH4, and N2O in the air at the monitoring point ; An acidity sensor and a temperature sensor, which are respectively used to collect the pH value of the water at the monitoring point and the water temperature in degrees Celsius ; A carbon dioxide sensor for collecting the concentration of bicarbonate in water at a monitoring point ; Methane sensors and nitrous oxide sensors are respectively used to collect the contents of dissolved greenhouse gases CH4 and N2O in the water at the monitoring points ; A calculation module for calculating the flux of greenhouse gases at the water-air interface in the following manner: According to , calculate the solubility of CO2, CH4, and N2O in surface water and the Henry's equilibrium constant of bicarbonate concentration in water ; Calculate the concentrations of dissolved CO2, CH4, and N2O in the water body at the monitoring point according to the following formula : ； Calculate the concentrations of CO2, CH4, and N2O in the air above the water surface at the monitoring points according to the following formula : ； Among them, is the correction coefficient of the ideal gas state equation, which is used to correct the ideal gas equation into the actual gas equation; Calculate the fluxes of CO2, CH4, and N2O at the water-air interface per unit area and unit time at the monitoring point according to the following formula : 。 2. A method for predicting the distribution of greenhouse gas fluxes at the water-air interface, characterized in that, Including: S210. Divide the water area plane to be monitored into multiple grids; S220. Take each grid as a monitoring point and install the automatic monitoring system for the flux of greenhouse gases at the water-air interface as described in claim 1 to monitor the flux of any greenhouse gas at the water-air interface of each grid in real time; and take the fluxes of each grid at the same moment as samples to construct a finite sample set; S230. Average the flux distributions monitored over a period of time, and take multiple extreme value grids and large gradient grids in the averaged flux distribution as multiple initial grids; And construct a neural network model with the fluxes of the multiple initial grids at the same moment as the input and the flux distribution of all grids at the same moment as the output; S240. Use the finite sample set to train and test the neural network model. If the model accuracy does not meet the requirements, cluster the grids with the difference between the predicted flux value and the monitored value exceeding the set threshold in the test results according to the geographical location, and include the cluster centers in the model input; S250. Return to S240 according to the new neural network model, and repeat this process until the model accuracy meets the requirements; S260. Use the genetic algorithm to select the optimal input grid combination that can meet the model accuracy requirements and has the least number of grids from the input grid set of the final model, and use the optimal model corresponding to the optimal input grid combination to predict the flux distribution at the water-air interface.

3. The method according to claim 2, wherein The training and testing of the neural network model using the finite sample set includes: Divide the finite sample set into a first training set, a second training set, and a test set; Use the first training set to train the neural network model once. In one training, update the model parameters by minimizing the flux change between adjacent grids; Use the second training set to perform a second training on the model after one training. In the second training, update the model parameters by minimizing the difference between the predicted flux value and the monitored value of each grid; Use the test set to test the model after the second training.

4. The method according to claim 3, characterized in that, The use of the second training set to perform a second training on the model after one training, and in the second training, update the model parameters by minimizing the difference between the predicted flux value and the monitored value of each grid, includes: If there is missing monitoring data for some grids in the samples in the second training set, in the second training, update the model parameters by minimizing the flux change between the part of the grids and adjacent grids, and minimizing the difference between the predicted flux value and the monitored value of other grids.

5. The method according to claim 4, characterized in that, The update of the model parameters by minimizing the flux change between the part of the grids and adjacent grids, and minimizing the difference between the predicted flux value and the monitored value of other grids, includes: Construct the following loss function: ； Among them, represents the loss function value, i 1 and j 1 represent the two-dimensional indices of the complete data grid, i 2 and j 2 represent the two-dimensional indices of the missing data grid, represents the flux prediction value of the corresponding grid, represents the flux monitoring value of the corresponding grid, represents taking the absolute value, respectively represent the weight coefficients.

6. The method according to claim 2, wherein The use of the genetic algorithm to select the optimal input grid combination that can meet the model accuracy requirements and has the least number of grids from the input grid set of the final model, includes: Construct a chromosome sequence structure of equal length according to the size of the input grid set of the final model, where each numerical bit in the sequence structure corresponds one by one to each grid in the set, and the 1 / 0 value of the numerical bit represents whether the corresponding grid is used as the input grid of the neural network model; Generate the current population of the genetic algorithm according to the chromosome sequence structure; Prune the final model according to each chromosome in the current population, and use the finite sample set to train the pruned model three times; Judge whether the model after three trainings meets the optimal goal. If not, generate a new current population according to the genetic algorithm, and return to the operation of model pruning until the model after the final three trainings meets the optimal goal; Among them, the optimal goal is that the accuracy of the model after three trainings meets the requirements, and the model with the least number of input grids has been obtained in the three trainings carried out so far.

7. The method according to claim 6, characterized in that, The generating a new current population according to the genetic algorithm includes: Construct a fitness function according to the change form of the flux prediction values between the input grids in the chromosome; Select excellent chromosomes from the mutated current population according to the fitness function to construct a new current population.

8. The method according to claim 7, characterized in that, The constructing a fitness function according to the change form of the flux prediction values between the input grids in the chromosome includes: Construct the following fitness function: ； Among them, is the fitness value, which represents the number of extreme grids in the flux distribution predicted by the model after three trainings for the input grids with a value of 1 in the chromosome; is the two-dimensional index of the input grids with a value of 1 in the chromosome, which represents the flux prediction value of the corresponding grid, represents taking the absolute value; represents the number of grids with a value of 0 in the chromosome; respectively represent the weight coefficients; the larger the fitness value, the better the chromosome.

9. The method according to claim 2, characterized in that The predicting the water-air interface flux distribution by using the optimal model corresponding to the optimal input grid combination includes: Retain the water-air interface greenhouse gas flux automatic monitoring system of each target grid in the optimal input grid combination, and remove the water-air interface greenhouse gas flux automatic monitoring systems of the remaining grids; Use the remaining system to monitor the gas flux of each target grid in real time, and input it into the optimal model corresponding to the optimal input grid combination to obtain the real-time flux distribution of all grids.

10. The method according to claim 2, characterized in that Include: For each greenhouse gas, perform the operations of S220-S260; among them, the predicting the water-air interface flux distribution by using the optimal model corresponding to the optimal input grid combination includes: Take the union of the optimal input grid combinations corresponding to various greenhouse gases, Retain the water-air interface greenhouse gas flux automatic monitoring system of each target grid in the union, and remove the water-air interface greenhouse gas flux automatic monitoring systems of the remaining grids; Use the remaining system to monitor the greenhouse gas flux of each target grid in real time, and input it into the optimal models corresponding to the most optimal input grid combinations of various greenhouse gases respectively to obtain the real-time greenhouse gas flux distribution of all grids.

Images