CN117253344A - Seawater acidification early warning and forecasting method, system and electronic equipment - Google Patents
Seawater acidification early warning and forecasting method, system and electronic equipment Download PDFInfo
- Publication number
- CN117253344A CN117253344A CN202311263362.5A CN202311263362A CN117253344A CN 117253344 A CN117253344 A CN 117253344A CN 202311263362 A CN202311263362 A CN 202311263362A CN 117253344 A CN117253344 A CN 117253344A
- Authority
- CN
- China
- Prior art keywords
- seawater
- early warning
- acidification
- data
- forecasting
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000013535 sea water Substances 0.000 title claims abstract description 220
- 230000020477 pH reduction Effects 0.000 title claims abstract description 145
- 238000013277 forecasting method Methods 0.000 title claims abstract description 9
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 claims abstract description 104
- 229910002092 carbon dioxide Inorganic materials 0.000 claims abstract description 50
- 239000001569 carbon dioxide Substances 0.000 claims abstract description 47
- BPQQTUXANYXVAA-UHFFFAOYSA-N Orthosilicate Chemical compound [O-][Si]([O-])([O-])[O-] BPQQTUXANYXVAA-UHFFFAOYSA-N 0.000 claims abstract description 19
- 229910019142 PO4 Inorganic materials 0.000 claims abstract description 19
- NBIIXXVUZAFLBC-UHFFFAOYSA-K phosphate Chemical compound [O-]P([O-])([O-])=O NBIIXXVUZAFLBC-UHFFFAOYSA-K 0.000 claims abstract description 19
- 239000010452 phosphate Substances 0.000 claims abstract description 19
- 238000000034 method Methods 0.000 claims description 42
- 238000012545 processing Methods 0.000 claims description 29
- 238000003860 storage Methods 0.000 claims description 24
- 238000012937 correction Methods 0.000 claims description 21
- 238000004422 calculation algorithm Methods 0.000 claims description 13
- 150000003839 salts Chemical class 0.000 claims description 13
- 229910002651 NO3 Inorganic materials 0.000 claims description 11
- NHNBFGGVMKEFGY-UHFFFAOYSA-N Nitrate Chemical compound [O-][N+]([O-])=O NHNBFGGVMKEFGY-UHFFFAOYSA-N 0.000 claims description 11
- 150000003863 ammonium salts Chemical class 0.000 claims description 11
- 238000011156 evaluation Methods 0.000 claims description 11
- 238000004519 manufacturing process Methods 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 11
- 230000002308 calcification Effects 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 10
- 235000015097 nutrients Nutrition 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 9
- 238000012546 transfer Methods 0.000 claims description 7
- 230000004927 fusion Effects 0.000 abstract description 8
- 239000010410 layer Substances 0.000 description 14
- 230000006870 function Effects 0.000 description 9
- 238000013135 deep learning Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 8
- 238000004088 simulation Methods 0.000 description 8
- 238000013528 artificial neural network Methods 0.000 description 6
- 238000010801 machine learning Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000005096 rolling process Methods 0.000 description 6
- 230000000007 visual effect Effects 0.000 description 6
- 229910021532 Calcite Inorganic materials 0.000 description 5
- VTYYLEPIZMXCLO-UHFFFAOYSA-L Calcium carbonate Chemical compound [Ca+2].[O-]C([O-])=O VTYYLEPIZMXCLO-UHFFFAOYSA-L 0.000 description 5
- 238000013473 artificial intelligence Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 5
- 230000008878 coupling Effects 0.000 description 5
- 238000010168 coupling process Methods 0.000 description 5
- 238000005859 coupling reaction Methods 0.000 description 5
- 230000007547 defect Effects 0.000 description 5
- 230000007774 longterm Effects 0.000 description 5
- 239000000126 substance Substances 0.000 description 5
- 235000014653 Carica parviflora Nutrition 0.000 description 4
- 241000243321 Cnidaria Species 0.000 description 4
- 238000010494 dissociation reaction Methods 0.000 description 4
- 230000005593 dissociations Effects 0.000 description 4
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 3
- 230000009471 action Effects 0.000 description 3
- 229910052799 carbon Inorganic materials 0.000 description 3
- 229910052729 chemical element Inorganic materials 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000003203 everyday effect Effects 0.000 description 3
- 238000011835 investigation Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 239000002344 surface layer Substances 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 125000004122 cyclic group Chemical group 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- GPRLSGONYQIRFK-UHFFFAOYSA-N hydron Chemical compound [H+] GPRLSGONYQIRFK-UHFFFAOYSA-N 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000003908 quality control method Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- BVKZGUZCCUSVTD-UHFFFAOYSA-L Carbonate Chemical compound [O-]C([O-])=O BVKZGUZCCUSVTD-UHFFFAOYSA-L 0.000 description 1
- 241001465754 Metazoa Species 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000002354 daily effect Effects 0.000 description 1
- 238000004090 dissolution Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 244000005700 microbiome Species 0.000 description 1
- 238000012821 model calculation Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 235000013619 trace mineral Nutrition 0.000 description 1
- 239000011573 trace mineral Substances 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
- 238000010792 warming Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/02—Alarms for ensuring the safety of persons
- G08B21/12—Alarms for ensuring the safety of persons responsive to undesired emission of substances, e.g. pollution alarms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Toxicology (AREA)
- Emergency Management (AREA)
- Business, Economics & Management (AREA)
- Environmental & Geological Engineering (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a seawater acidification early warning and forecasting method, a system and electronic equipment, which relate to the field of seawater acidification early warning and forecasting and comprise the following steps: acquiring early warning and forecasting data, wherein the early warning and forecasting data comprise at least one of wind speed, atmospheric carbon dioxide partial pressure, sea water temperature, salinity, phosphate, silicate, sea water surface carbon dioxide partial pressure and sea water pH; and inputting the early warning and forecasting data to a trained LSTM model based on a TensorFlow framework to finish early warning and forecasting of seawater acidification and obtain a forecasting result. According to the invention, through a plurality of judgment indexes, the pre-warning prediction of the seawater acidification is completed based on the LSTM model, the pre-warning prediction of the seawater acidification in a continuous and future long-time scale aiming at the full seawater space range can be realized, and the requirement of fusion observation of the seawater acidification pre-warning prediction is met.
Description
Technical Field
The invention relates to the field of seawater acidification early warning and forecasting, in particular to a seawater acidification early warning and forecasting method, a seawater acidification early warning and forecasting system and electronic equipment.
Background
At present, the seawater acidification phenomenon is mainly judged through ship investigation and buoy early warning prediction in China or is subjected to historical evaluation based on a numerical model, but a blank exists in the aspect of seawater acidification early warning prediction in large sea areas or key areas, namely in long-term or short-term event scale. The prior art has the following defects:
1. The current pre-warning of sea water acidification is mainly based on-site pre-warning and forecasting to carry out short-term pre-warning prompt on whether sea water at individual points is acidified or not, is limited by space-time limitation of on-site observation data, and lacks full-space real-time continuous sea water acidification pre-warning for the whole south China sea or for a key area;
2. at present, no seawater acidification early warning and forecasting product aiming at a future long time scale is available in China;
3. at present, the judging index for the seawater acidification phenomenon in China has the defect of singleness;
4. at present, the requirements of the domestic fusion observation and numerical model cannot be met.
Disclosure of Invention
The invention aims to solve the technical problems that continuous pre-warning and forecasting of the whole seawater space range cannot be realized, and seawater acidification pre-warning and forecasting and single judgment index cannot be realized in a future long-time scale.
A seawater acidification early warning and forecasting method, a seawater acidification early warning and forecasting system and electronic equipment comprise the following steps:
acquiring early warning forecast data; the early warning forecast data comprise at least one of wind speed, atmospheric carbon dioxide partial pressure, seawater temperature, salinity, phosphate, silicate, seawater surface carbon dioxide partial pressure and seawater pH;
And inputting the early warning and forecasting data to a trained LSTM model based on a TensorFlow framework to finish early warning and forecasting of seawater acidification and obtain a forecasting result.
Further, the method for forming the trained LSTM model comprises the following steps:
constructing an input factor according to a historical data set, wherein the historical data set comprises at least one of wind speed, atmospheric carbon dioxide partial pressure, sea water temperature, salinity, nutrient salt, sea water surface carbon dioxide partial pressure and sea water pH;
constructing a preliminary LSTM model based on a TensorFlow framework;
and training the preliminary LSTM model through a time sequence inverse transfer algorithm according to the input factors to obtain a trained LSTM model.
Further, wherein the historical dataset comprises data from 2000 to 2022 for near 23 south China sea buoy observations, ship surveys, satellite telemetry, numerical simulations.
Further, wherein the input factor of the LSTM model is a subset of the historical dataset of the predicted date and its previous 2 days, and the output of the trained LSTM model is an acidification or calcification indicator of the predicted date.
Further, the method further comprises the following steps:
acquiring a plurality of sea water area positions;
Setting an acidification warning value set according to the positions of all the sea water areas;
and sending seawater acidification proximity warning information according to the acidification warning value and the prediction result.
Further, the method further comprises the following steps:
obtaining standard layer lattice point data according to the prediction result;
and updating the trained LSTM model into a seawater pH value calculation model according to the standard layer lattice point data.
Further, the method further comprises the following steps:
acquiring real-time observation information, wherein the real-time observation information comprises at least one of sea temperature, salinity, nitrate, ammonium salt, phosphate, silicate and partial pressure of carbon dioxide on the surface of sea water;
and correcting the prediction result according to the real-time observation information.
Further, the method further comprises the following steps: and correcting and updating the trained LSTM model based on the DNN framework according to the real-time observation information.
Another embodiment of the present invention provides a seawater acidification early warning and forecasting system, including:
the data subsystem comprises a data control module, a field buoy observation data receiving module, an operation module, a first processing module and a first storage module;
the prediction subsystem comprises an early warning control module, a prediction control module, a correction module and a second storage module;
The application subsystem comprises an adjacent early warning evaluation module, a second processing module, a manufacturing module and an archiving module;
and the controller is used for controlling at least one of the data subsystem, the prediction subsystem and the application subsystem to realize the method.
Another embodiment of the present invention provides an electronic device, including:
memory, a processor and a computer program stored on the memory and executable on the processor, which processor implements the above method when executing the program.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. according to the seawater acidification early warning and forecasting method, the early warning and forecasting of seawater acidification is completed based on the LSTM model through a plurality of judgment indexes, so that the early warning and forecasting of seawater acidification in a continuous and future long-time scale in a full seawater space range can be realized, and the requirements of fusion observation of seawater acidification early warning and forecasting are met;
2. according to the seawater acidification early warning and forecasting system, the early warning and forecasting of seawater acidification is completed based on the LSTM model through a plurality of judgment indexes, so that the early warning and forecasting of seawater acidification in a continuous and future long-time scale in a full seawater space range can be realized, and the requirements of fusion observation of seawater acidification early warning and forecasting are met;
3. According to the electronic equipment, through a plurality of judgment indexes, the pre-warning prediction of the seawater acidification is completed based on the LSTM model, the pre-warning prediction of the seawater acidification in a continuous and future long-time scale in a full seawater space range can be realized, and the requirements of fusion observation of the seawater acidification pre-warning prediction are met.
Drawings
The drawings are included to provide a better understanding of the present invention and are not to be construed as limiting the invention. Wherein:
FIG. 1 is a schematic diagram of a seawater acidification early warning method according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a seawater acidification early warning system according to an embodiment of the present invention;
FIG. 3 is a flow diagram of an early warning control module according to one embodiment of the invention;
FIG. 4 is a flow diagram of a correction module according to one embodiment of the invention;
FIG. 5 is a schematic diagram of an electronic device according to one embodiment of the invention;
FIG. 6 is a schematic diagram of an uncorrected ocean pH 24-hour predictive simulated visual product in accordance with one embodiment of the invention;
FIG. 7 is a schematic representation of a 24 hour predicted simulated visualization of corrected ocean pH in accordance with one embodiment of the invention.
Reference numerals:
100-seawater acidification early warning and forecasting system, 101-data subsystem, 1011-data control module, 1012-field buoy observation data receiving module, 1013-operation module, 1014-first processing module, 1015-first storage module, 102-prediction subsystem, 1021-early warning and control module, 1022-prediction and control module, 1023-correction module, 1024-second storage module, 103-application subsystem, 1031-proximity early warning and evaluation module, 1032-second processing module, 1033-manufacturing module, 1034-archiving module, 104-controller, 200-electronic device, 201-processor, 202-memory.
Detailed Description
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative and intended to explain the present invention and should not be construed as limiting the invention.
As shown in FIG. 1, the invention provides a seawater acidification early warning and forecasting method, which comprises the following steps:
acquiring early warning forecast data; the early warning forecast data comprise at least one of wind speed, atmospheric carbon dioxide partial pressure, seawater temperature, salinity, phosphate, silicate, seawater surface carbon dioxide partial pressure and seawater pH;
and inputting the early warning and forecasting data to a trained LSTM model based on a TensorFlow framework to finish early warning and forecasting of seawater acidification and obtain a forecasting result.
In some embodiments, as shown in fig. 1, a technical scheme of coupling an Off-line model and artificial intelligence is provided for the scene of the early warning and forecasting of the acidification of the seawater in the south China sea.
As the seawater continuously absorbs atmospheric carbon dioxide, the chemical system of the seawater carbonate changes due to the dissolution of the carbon dioxide in the seawater, so that the seawater acidification phenomenon becomes serious, and the marine ecological communities such as coral reefs and the like are further influenced. According to the invention, by utilizing txt standardized historical and real-time early warning and forecasting data comprising key elements such as wind speed, atmospheric carbon dioxide partial pressure, sea water temperature, salinity, nutrient salt, sea water surface carbon dioxide partial pressure, sea water pH and the like, the rapid early warning and forecasting of sea water acidification is carried out by using an LSTM deep learning method based on a Tensorflow framework, and the south sea water acidification early warning and forecasting system 100 based on the Off-line model and artificial intelligence coupling is formed. The aim is to solve the defect that the current seawater acidification early warning period is short and the defect that the current seawater acidification early warning is limited by an observation means based on the daily automatic business operation of an autonomously established physical-biochemical numerical forecasting system in the south China sea; based on various atmospheric, physical and chemical variables such as sea temperature, salinity, nitrate, ammonium salt, phosphate, silicate, sea water carbon dioxide partial pressure, atmospheric carbon dioxide partial pressure and the like, an Off-line sea water acidification model is established, so that the blank of the current sea water acidification prediction technology and long-term change evaluation can be filled; the method has the advantages that through predicting a plurality of indexes such as sea water pH, calcite saturation, aragonite saturation and the like, the pre-warning products of the sea water acidification degree in different prediction periods are comprehensively given, and the defect of singleness of the current sea water acidification pre-warning products can be overcome; the comprehensive visual production system for the seawater acidification from the key area to the south China sea area provides more visual prediction products for future changes of the seawater acidification for users or governments, and knows the conditions of the seawater acidification degree, duration and the like.
In some embodiments, further comprising a method of forming the trained LSTM model, comprising:
constructing an input factor according to a historical data set, wherein the historical data set comprises at least one of wind speed, atmospheric carbon dioxide partial pressure, sea water temperature, salinity, nutrient salt, sea water surface carbon dioxide partial pressure and sea water pH;
constructing a preliminary LSTM model based on a TensorFlow framework;
and training the preliminary LSTM model through a time sequence inverse transfer algorithm according to the input factors to obtain a trained LSTM model.
In the embodiment, the LSTM deep learning method is utilized, the constructed seawater acidification quick prediction model is based on field buoy observation data, the problem of low timeliness of a numerical model is solved, and whether the seawater is acidified or not can be quickly early-warned in the first time.
Wherein the historical dataset may comprise data of at least 20 years of south China sea buoy observation, ship survey, satellite remote sensing, numerical simulation.
In some embodiments, historical data such as wind speed, atmospheric carbon dioxide partial pressure, sea water temperature, salinity, phosphate, silicate, sea water surface carbon dioxide partial pressure, sea water pH and the like of the buoy observation, ship investigation, satellite remote sensing, numerical simulation of the south China sea in the last 23 years from 2000 to 2022 are utilized, a time sequence inverse transfer algorithm is adopted to carry out historical data set simulation training based on an LSTM deep learning network model, an input factor is physical and chemical element data of a predicted date and two days before the predicted date, an LSTM model is output as a calcification rate of the predicted date, an LSTM model optimal super parameter is determined through simulation learning, and a sea water acidification rapid assessment prediction model is established.
Among them, the time series inverse transfer algorithm (Backpropagation Through Time, BPTT) is a training algorithm used in a Recurrent Neural Network (RNN). RNNs are a type of neural network with cyclic connections that can process sequence data, such as time series, speech, natural language, etc.
The main idea of the BPTT algorithm is to calculate the gradient of the parameters in the RNN by means of a back-propagation algorithm and use the gradient for updating the parameters. Since RNNs have cyclic connections, all time steps in the sequence need to be considered when calculating gradients. The BPTT algorithm updates the parameters in the RNN by expanding all time steps in the sequence into a forward pass, then calculating the gradient for each time step, and back-propagating the gradient back to the previous time step.
Specifically, the BPTT algorithm includes the steps of:
1. forward transfer: for each time step t, the input x_t is passed into the RNN and the hidden state h_t and the output y_t are calculated.
2. Back propagation: for each time step t, the gradient of the loss function l_t is calculated and the gradient is back propagated to the previous time step. All time steps in the sequence need to be considered when calculating the gradient.
3. Parameter updating: parameters in the RNN, such as the weight matrix W and the bias vector b, are updated using the calculated gradients.
The BPTT algorithm can enable the RNN to learn the long-term dependency relationship in the sequence data and improve the prediction capability of the model.
Wherein the input factor of the LSTM model is a subset of the historical dataset of the predicted date and the previous 2 days, and the output of the trained LSTM model is an acidification or calcification index of the predicted date.
In some embodiments, the calcification rates of the plurality of predicted dates form a calcification rate sample library, and based on the calcification rate sample library, each real-time early warning and forecasting site is evaluated for the seawater acidification phenomenon, wherein the seawater acidification threshold is adopted for evaluation.
The technical scheme of one implementation scheme of the invention is as follows: historical data such as wind speed, atmospheric carbon dioxide partial pressure, sea water temperature, salinity, phosphate, silicate, sea water surface carbon dioxide partial pressure, sea water pH and the like of the last twenty three years of south China sea buoy observation, ship investigation, satellite remote sensing and numerical simulation in the year 2000 to 2022 are utilized, based on an LSTM deep learning network model, a time sequence inverse transfer algorithm is applied to carry out historical data set simulation training, input factors are physical and chemical element data of a predicted date and the first two days, an LSTM model is output as a calcification rate of the predicted date, the optimal super parameter of the LSTM model is determined through simulation learning, and a sea water acidification rapid assessment prediction model is established.
When the device is used, data such as wind speed, atmospheric carbon dioxide partial pressure, sea water temperature, salinity, phosphate, silicate, sea water surface carbon dioxide partial pressure, sea water pH and the like which are monitored in real time are input, and prediction data of related elements are input, and based on an established LSTM rapid evaluation calculation model, the physical and chemical element data of the prediction date and the first two days are utilized to perform prediction calculation on the sea water acidification index of the prediction date, so that a future sea water acidification index sample library is obtained. Indicators of seawater acidification or calcification may include seawater pH, aragonite saturation, calcite saturation, and the like.
In some embodiments, based on the calcification rate sample library and the seawater acidification threshold, the seawater acidification phenomenon of each real-time monitoring site is evaluated, and the python software is used for manufacturing rapid seawater acidification early warning information and visualized products.
Further comprises:
acquiring a plurality of sea water area positions;
setting an acidification warning value set according to the positions of all the sea water areas;
and sending seawater acidification proximity warning information according to the acidification warning value and the prediction result.
In addition, the seawater acidification prediction data in the seawater acidification early warning prediction method can be utilized to provide seawater acidification approaching early warning information for local government or related departments to take action in advance according to different acidification warning values set by seawater characteristics or ecosystems of different areas when the warning values are reached.
Further comprises:
obtaining standard layer lattice point data according to the prediction result;
and updating the trained LSTM model into a seawater pH value calculation model according to the standard layer lattice point data.
The ocean serves as a huge carbon reservoir for the earth's CO2 system circulation, and the global ocean absorbs about 30% -40% of CO2 concentration emitted by humans. The ocean absorbs atmospheric CO2 while moderating global warming, also destroys the balance of the chemical structure of seawater, resulting in an increase in seawater acidity. The seawater acidification not only affects the seawater chemical characteristics of ocean carbon chemistry, nutrient salts, trace elements and the like, but also further affects the ocean phytoplankton, microorganisms, various large animals and coral reefs, and even damages the whole ocean ecosystem.
Further, the standard layer lattice point data can be based on a cubic spline interpolation method, and for a south China sea area, the data under an original sigma coordinate system is interpolated to the layer depth of a surface layer, ten meters, fifty meters, one hundred meters, two hundred meters and five hundred meters under the standard layer, and a netcdf file in a standard format is output.
The specific calculation method of the seawater pH value calculation model is as follows:
Hsws＝αH/fH (1)
where αh is the hydrogen ion activity, fH is the activity coefficient of the hydrogen ion h+, and the specific calculation equation is as follows:
αH＝TA-CAlk-BAlk-OH-PAlk-SiAlk+Hfree+HSO4+HF (2)
fH＝ln10.*(TC.*K1F.*H.*(H.*H+K1F.*K2F+4.*H.*K2F)./Denom./Denom
+BAlk.*H./(KBF+H)+OH+H)
(3)
Where TA is the total input alkalinity, TC is the total input inorganic carbon, and the other various parts of the equation are calculated: calk=tc k1. (h+2.×k2)/Denom; denom= (h.h+k1.h+k1.k2); h=10 (-pHx); balk=tb KB./(kb+h); oh= KW./H; palk=tp..photostop./photosboost; sialk=tsi..ksi./(ksi+h); photostop=kp1..kp2..h+2..kp: KP2 KP 3-H; phosboost=h.h. +kp1.h. +kp1.kp2.h+kp2.kp3; hfre=h./FREEtoTOT; fretotot= (1+ts./KS); hso4= TS./(1+ks./Hfree); hf= TF./(1+kf./Hfree).
K1 and K2 are carbonic acid first dissociation constant and carbonic acid second dissociation constant,
K1＝10(A1-A2/Tk-log(Tk)*A3+Salt*(A4-Salt*A5)) (4)
wherein Tk is Kelvin temperature, salt is salinity, and A1, A2, A3, A4 and A5 are related parameters, as shown in Table 1.
TABLE 1 parameters in the formula for the first dissociation constant of carbonic acid
K2＝10(B1-B2/Tk+Salt*(B3-Salt*B4)) (5)
Wherein, B1, B2, B3 and B4 are related parameters, and are shown in Table 2.
TABLE 2 parameters in the formula for the second dissociation constant of carbonic acid
KP1, KP2, KP3, KSi, KW, KB, KS, KF are all coefficients related to temperature, salinity. TSi and TP are the input silicate and phosphate, respectively.
Further comprises:
acquiring real-time observation information, wherein the real-time observation information comprises at least one of sea temperature, salinity, nitrate, ammonium salt, phosphate, silicate and partial pressure of carbon dioxide on the surface of sea water;
And correcting the prediction result according to the real-time observation information.
The buoy observation is used as a machine learning true value, and the numerical value of the nearest numerical mode lattice point and the numerical mode lattice point extending outwards of the buoy is used as the input of online correction. Buoy observation refers to a method for performing ocean observation by using a buoy as an observation platform. In the field of oceanography, a buoy is a floating device for collecting marine environment data, and can continuously and fixedly observe marine environment parameters such as water quality, temperature, salinity, ocean current and the like for a long time and transmit the data to a shore station or a satellite in real time. The data has important significance in the aspects of marine science research, climate prediction, fishery resource management and the like. Buoy observation technology is an important non-invasive measurement method in marine science research, and is widely applied to the fields of marine ecosystems, climate change, environmental protection and the like.
As shown in fig. 6 and 7, fig. 6 is a 24-hour predicted simulated visual product of the ocean pH without correction, and fig. 7 is a 24-hour predicted product of the seawater pH after correction, and the corrected seawater pH can be found more intuitively and accurately.
Further comprises: and correcting and updating the trained LSTM model based on the DNN framework according to the real-time observation information.
In an exemplary technical scheme, by utilizing a DNN machine learning method, based on early warning and forecasting data received in real time, the problem of how to integrate observation and numerical forecasting can be solved by performing real-time rolling correction on seawater acidification indexes predicted by an Off-line seawater acidification model. Specifically, the correction input data is obtained through a control file written based on shell script, and the correction input data comprises real-time monitoring data of txt standardized wind speed, atmospheric carbon dioxide partial pressure, seawater temperature, salinity, nutrient salt, seawater surface carbon dioxide partial pressure and other factors; obtaining sea temperature, salinity, nitrate, ammonium salt, phosphate, silicate and sea water surface carbon dioxide partial pressure prediction data of seven days in the standard layer lattice point data; aiming at the LSTM model, based on a DNN deep learning framework, the correction prediction is carried out on the subsequent seawater acidification numerical prediction by utilizing real-time observation information such as sea temperature, salinity, nitrate, ammonium salt, phosphate, silicate, partial pressure of carbon dioxide on the surface of seawater and the like between the starting time of seawater acidification numerical prediction and the release time of a seawater acidification prediction product.
Wherein the DNN framework represents a deep neural network framework, which is a software tool or library for implementing and training deep neural networks. Deep neural networks are a class of machine learning models that typically contain multiple layers of neural networks, the connections between which are represented by weights that are learned through a training process to extract features from the data and perform complex pattern recognition. In the present invention, considering the corrective action, one or more of the following DNN frameworks may be selected: tensorFlow, pyTorch, keras, caff, MXNet, theano, CNTK.
Compared with the prior art, the invention has the following advantages:
1. the seawater acidification early warning evaluation integrates the influence of a plurality of seawater physical, chemical and atmospheric element indexes;
2. the early warning prediction system can release the forecast product of the sea water acidification of the south China sea in real time every day, thereby effectively ensuring the timeliness and pertinence of forecast;
3. based on LSTM deep learning, seawater acidification early warning information can be rapidly given;
4. the DNN machine learning method is utilized, so that real-time rolling correction can be carried out on the seawater acidification product subjected to numerical forecasting;
5. the forecasting system of the invention can realize full-space real-time continuous sea acidification assessment and prediction of the whole south China sea area or key area, can enable the state/local government to know the condition of the sea acidification degree from seven days in a short period to three months in the future, comprehensively influences the sea ecosystems such as coral reefs and the like, and carries out sea ecological restoration and protection work according to specific characteristic implementation specific scheme
The invention also provides a seawater acidification early warning and forecasting system 100, which comprises:
the data subsystem 101 comprises a data control module 1011, a field buoy observation data receiving module 1012, an operation module 1013, a first processing module 1014 and a first storage module 1015;
A predictive subsystem 102 including an early warning control module 1021, a predictive control module 1022, a correction module 1023, and a second storage module 1024;
the application subsystem 103 comprises an adjacent early warning evaluation module 1031, a second processing module 1032, a manufacturing module 1033 and an archiving module 1034;
a controller 104 for implementing the above method by controlling at least one of the data subsystem 101, the prediction subsystem 102 and the application subsystem 103.
In some embodiments, a block diagram of an exemplary system suitable for implementing embodiments of the present invention is shown. The system shown in fig. 2 is merely an example, and should not be construed as limiting the functionality and scope of use of embodiments of the present invention.
The system 100 for pre-warning and pre-warning the acidification of the south sea water based on the Off-line model and artificial intelligence coupling mainly comprises three parts, namely a data subsystem 101, a pre-warning subsystem 102 and an application subsystem 103. Wherein the operation module 1013 of the data subsystem 101 can be used for predicting the numerical value of the south China sea area, and the first processing module 1014 can process the key data; the early warning control module 1021 of the forecasting subsystem 102 can realize the rapid early warning control of the seawater acidification; the predictive control module 1022 may implement Off-line based seawater acidification prediction; the correction module 1023 can realize seawater acidification rolling correction prediction; the second processing module 1032 of the application subsystem 103 may implement predictive data processing; the manufacturing module 1033 can realize visual product manufacturing; the archive module 1034 may implement predictive product archiving.
Specifically, the data control module 1011 controls the automatic operation of each module of the subsystem based on the shell script file, and the user can check the operation status through the log file; the on-site buoy observation data receiving module 1012 is mainly used for receiving historical and real-time data such as wind speed, atmospheric carbon dioxide partial pressure, sea water temperature, salinity, nutrient salt, sea water surface carbon dioxide partial pressure, sea water pH and the like which are early-warning and forecast for a south China sea buoy, a tide station and the like; the operation module 1013 is used for realizing numerical forecasting of the south China sea area and forecasting and predicting key elements such as sea temperature, salinity, nitrate, ammonium salt, phosphate, silicate, partial pressure of carbon dioxide on the surface of sea water and the like. The module mainly comprises two sub-modules, wherein the sub-modules automatically run every day to conduct routine prediction of a week in the future; and secondly, if the day is fifteen days per month through time judgment, automatically downloading atmospheric prediction data of three months in the future, and carrying out automatic operation of a physical-biological geochemical numerical forecasting system of the south China sea to predict key elements of three months in the future. A user can check whether the operation is normal or not through log files; the key element data processing module mainly comprises two sub-modules: firstly, aiming at the historical and real-time early warning and forecasting data collected by the field buoy observation data receiving module 1012, carrying out quality control on the early warning and forecasting data based on a data quality control system, carrying out standardized processing on the early warning and forecasting data by using a python script, and outputting a txt file; secondly, interpolating data under an original sigma coordinate system to layer depths of a surface layer, ten meters, fifty meters, one hundred meters, two hundred meters and five hundred meters under a standard layer in a south China sea area by utilizing key element prediction data obtained by operation of an operation module 1013 based on a cubic spline interpolation method, outputting netcdf files in a standard format, and interpolating data under the original sigma coordinate system to layer depths of the surface layer, one meter, two meters, five meters, ten meters and twenty meters under the standard layer in a local key area such as a Guangdong and Australia large bay area, and outputting netcdf files in the standard format; the first storage module 1015 is configured to store the early warning forecast and the forecast data obtained after the normalization process in a classified manner according to the date.
In use, the run module 1013 determines two problems according to the south China sea area physical-biological geochemical numerical model calculations: firstly, confirming whether the time is 15 days per month; if yes, carrying out calculation operation of three months in the future, and if not, carrying out treatment. A seven day future prediction is then made. The standard layer data of key elements of the key attention area of seawater acidification is extracted by the first processing module 1014 through cubic spline interpolation, and then the data are classified and stored by the first storage module 1015 after standardized processing. The on-site buoy observation data receiving module 1012 also stores on-site early warning and forecasting data such as the ocean level station of the marine buoy in the south China through the first storage module 1015.
The prediction subsystem 102, as shown in fig. 3, wherein the early warning control module 1021 is a script control file written based on a shell, and has the following main functions: a1. the data subsystem 101 and the seawater acidification rapid early warning subsystem are linked; a2. the historical early warning forecast data which is standardized by txt and contains factors such as wind speed, atmospheric carbon dioxide partial pressure, sea water temperature, salinity, nutrient salt, sea water surface carbon dioxide partial pressure, sea water pH and the like and is acquired by the first storage module 1015 in the link data subsystem 101 are subjected to normalization, data segmentation and training set division; a3. training historical early warning and forecasting data based on an LSTM deep learning frame to obtain a seawater acidification rapid early warning model; a4. after the model training is completed, txt standardized real-time early warning and forecasting data comprising wind speed, atmospheric carbon dioxide partial pressure, sea water temperature, salinity, nutrient salt, sea water surface carbon dioxide partial pressure and other factors are obtained in a first storage module 1015 in the linked data subsystem 101, and based on a sea water acidification fast early warning model, the acidification indexes such as sea water pH, aragonite saturation, calcite saturation and the like are rapidly predicted.
The prediction control module 1022, which may be an Off-line seawater acidification prediction subsystem control module, is a control file written based on a shell script, and has the following main functions: a1. the data subsystem 101 and the Off-line seawater acidification prediction subsystem are linked, and key element standard layer lattice point data acquired by a first storage module 1015 module in the data subsystem 101 are linked; a2. automatically calling sea temperature, salinity, nitrate, ammonium salt, phosphate, silicate and sea water surface carbon dioxide partial pressure prediction data of a netcdf format stored by the data subsystem 101 every day, calculating acidification indexes such as sea water pH, aragonite saturation, calcite saturation and the like based on an OFF-line sea water acidification calculation module, and performing short-term prediction on sea water acidification; a3. if the time of the day is fifteen days of the present month, calling key element data such as sea temperature, salinity, nitrate, ammonium salt, phosphate, silicate, partial pressure of sea surface carbon dioxide, partial pressure of atmospheric carbon dioxide and the like of three months in the future under the netcdf format, and carrying out long-term prediction of three months aiming at acidification indexes such as sea water pH, aragonite saturation, calcite saturation and the like; a4. and carrying out netcdf standard format processing filing on the seawater acidification index prediction data.
The correction module 1023 may be a seawater acidification rolling correction prediction subsystem control module, as shown in fig. 4, and is a control file written based on shell script, and has the following main functions: a1. the data subsystem 101 and the seawater acidification rolling correction prediction subsystem are linked; a2. the txt standardized real-time early warning and forecasting data comprising wind speed, atmospheric carbon dioxide partial pressure, sea water temperature, salinity, nutrient salt, sea water surface carbon dioxide partial pressure and other factors are acquired in a first storage module 1015 in the link data subsystem 101; a3. linking sea temperature, salinity, nitrate, ammonium salt, phosphate, silicate, sea water surface carbon dioxide partial pressure prediction data of the netcdf format stored by the data subsystem 101 for seven days in the future; a3. linking an Off-line seawater acidification prediction subsystem in the seawater acidification early warning prediction subsystem 102; a4. based on DNN deep learning framework, correcting and predicting data for the subsequent seawater acidification numerical forecast by utilizing real-time observation information such as sea temperature, salinity, nitrate, ammonium salt, phosphate, silicate, partial pressure of carbon dioxide on the surface of seawater and the like between the starting time of Off-line seawater acidification numerical forecast and the release time of a seawater acidification forecast product, taking buoy observation as a machine learning true value, and taking the numerical values of the nearest numerical mode lattice point and the outwards extending numerical mode lattice point of the buoy as the input of online correction.
The second storage module 1024 is configured to process the fast prediction data of the early warning control module 1021, the short-term and long-term prediction data of the prediction control module 1022, and the prediction data after the real-time rolling correction of the correction module 1023 into a netcdf standard format file, and store and archive the same.
The application subsystem 103, wherein the proximity warning evaluation module 1031 may be a seawater acidification prediction data of a rapid seawater acidification warning subsystem in the seawater acidification prediction subsystem 102, and when reaching the warning value, provides seawater acidification proximity warning information for local government or related departments according to different regional seawater characteristics or different acidification warning values set by an ecological system; the second processing module 1032 may be a script control file written based on shell, and main functions include: the sea water acidification prediction data of different time scales calculated by the sea water acidification early warning prediction subsystem 102 are processed, firstly, the time dimension is processed from the world time to the Beijing time, and secondly, sea water acidification data of heavy-point ecological protection areas such as Guangdong and Australian bay areas, coral reefs and the like in key areas of the south China are extracted by a cubic spline interpolation method; the manufacturing module 1033 may be configured to visually manufacture the degree and duration of acidification of the sea water based on python software using the sea water acidification data of the south China sea area and the key area obtained by the second processing module 1032; the archiving module 1034 may archive and store the date corresponding to the seawater acidification pre-alarm visualized product generated by the approaching pre-alarm evaluation module 1031 and the manufacturing module 1033.
The system described above is used to represent in the form of a general purpose computing device. The components of the system may also include, but are not limited to: one or more processors or processing units, a system memory, and a bus that connects the various system components (including the system memory and processing units), collectively referred to as a controller.
Various implementations of the systems and techniques described above in this invention may be implemented in digital electronic circuitry, integrated circuit systems, FPGAs (FieldProgrammable Gate Array, field programmable gate arrays), ASICs (Application-Specific Integrated Circuit, application-specific integrated circuits), ASSPs (Application Specific Standard Product, application-specific standard products), SOCs (System On Chip, complex programmable logic devices), CPLDs (Complex Programmable Logic Device), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.
To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display ) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.
An electronic device 200, comprising:
memory 202, processor 201, and a computer program stored on memory 202 and executable on processor 201, which processor 201 implements the above-described methods when executing the program.
It should be noted that, as shown in fig. 5, the electronic device 200 in this example is represented in the form of a general-purpose computing device. The components of the electronic device 200 may include, but are not limited to: one or more processors 201 or processing units, a system memory 202, and a bus connecting different system components (including the system memory 202 and processing units).
The electronic device 200 includes a computing unit that can perform various appropriate actions and processes according to a computer program stored in a ROM (Read-Only Memory 202) or a computer program loaded from a storage unit into a RAM (Random Access Memory, random access Memory 202). In the RAM, various programs and data required for the operation of the device may also be stored. The computing unit, ROM and RAM are connected to each other by a bus. An I/O (Input/Output) interface is also connected to the bus.
Various components in the electronic device 200 are connected to the I/O interface, including: an input unit such as a keyboard, a mouse, etc.; an output unit such as various types of displays, speakers, and the like; a storage unit such as a magnetic disk, an optical disk, or the like; and communication units such as network cards, modems, wireless communication transceivers, and the like. The communication unit allows the device to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.
The exemplary electronic device 200 used to implement embodiments of the present invention is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. Electronic device 200 may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smart phones, wearable devices, and other similar computing devices. The components shown in this disclosure, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed in this disclosure.
The processor 201 is a computing unit that may be a variety of general and/or special purpose processing components with processing and computing capabilities. Some examples of processor 201 include, but are not limited to, a CPU (Central Processing Unit ), GPU (Graphic Processing Units, graphics processing unit), various dedicated AI (Artificial Intelligence ) computing chips, various computing units running machine learning model algorithms, DSPs (Digital Signal Processor, digital signal processor 201), and any suitable processor 201, controller 104, microcontroller 104, etc. The computing unit performs the various methods and processes described above, such as an imaging method based on a focal plane fit. For example, in some embodiments, the focal plane fitting-based imaging method may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and/or installed onto the device via the ROM and/or the communication unit. One or more steps of the methods described above may be performed when the computer program is loaded into RAM and executed by a computing unit. Alternatively, in other embodiments, processor 201 may be configured to perform the aforementioned methods by any other suitable means (e.g., by means of firmware).
In the context of the present invention, memory 202 may be a machine-readable medium, which may be a tangible medium, that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may also be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, RAM, ROM, EPROM (Electrically Programmable Read-Only-Memory 202) or flash Memory 202, an optical fiber, a CD-ROM (Compact Disc Read-Only Memory 202), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
Program code for carrying out methods of the present invention may be written in any combination of one or more programming languages. These program code may be provided to the processor 201 or controller 104 of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the program code, when executed by the processor 201 or controller 104, causes the functions/operations specified in the flowchart and/or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. according to the seawater acidification early warning and forecasting method provided by the invention, the early warning and forecasting of seawater acidification is completed based on the LSTM model through a plurality of judgment indexes, so that the seawater acidification early warning and forecasting can be realized in a full seawater space range continuously based on the future long time scale, and the requirements of fusion observation of seawater acidification early warning and forecasting are met;
2. according to the seawater acidification early warning and forecasting system provided by the invention, the early warning and forecasting of seawater acidification is completed based on the LSTM model through a plurality of judgment indexes, so that the seawater acidification early warning and forecasting can be realized in a full seawater space range continuously based on the future long time scale, and the requirements of fusion observation of seawater acidification early warning and forecasting are met;
3. according to the electronic equipment provided by the invention, through a plurality of judgment indexes, the early warning prediction of the seawater acidification is completed based on the LSTM model, the early warning prediction of the seawater acidification in a continuous and future long-time scale can be realized aiming at the full seawater space range, and the requirements of fusion observation of the seawater acidification early warning prediction are met.
In the several embodiments provided in the present invention, it should be understood that the disclosed technology may be implemented in other manners. The above-described embodiments of the apparatus are merely exemplary, and the division of the units, for example, may be a logic function division, and may be implemented in another manner, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interfaces, units or modules, or may be in electrical or other forms.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.
Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In the description of the present invention, the meaning of "plurality" means at least two, for example, two, three, etc., unless specifically defined otherwise.
Claims (10)
1. A seawater acidification early warning and forecasting method comprises the following steps: acquiring early warning forecast data; it is characterized in that the method comprises the steps of,
the early warning forecast data comprise at least one of wind speed, atmospheric carbon dioxide partial pressure, seawater temperature, salinity, phosphate, silicate, seawater surface carbon dioxide partial pressure and seawater pH;
and inputting the early warning and forecasting data to a trained LSTM model based on a TensorFlow framework to finish early warning and forecasting of seawater acidification and obtain a forecasting result.
2. The seawater acidification early warning method according to claim 1, further comprising a method for forming the trained LSTM model, comprising:
constructing an input factor according to a historical data set, wherein the historical data set comprises at least one of wind speed, atmospheric carbon dioxide partial pressure, sea water temperature, salinity, nutrient salt, sea water surface carbon dioxide partial pressure and sea water pH;
constructing a preliminary LSTM model based on a TensorFlow framework;
and training the preliminary LSTM model through a time sequence inverse transfer algorithm according to the input factors to obtain a trained LSTM model.
3. The method of claim 2, wherein the historical dataset comprises data from at least 20 years of national sea buoy observation, ship survey, satellite remote sensing, numerical modeling.
4. The seawater acidification early warning method according to claim 1, wherein the input factor of the LSTM model is a subset of the historical dataset of a predicted date and 2 days before the predicted date, and the output of the trained LSTM model is an acidification or calcification index of the predicted date.
5. The seawater acidification early warning method according to claim 1, further comprising:
acquiring a plurality of sea water area positions;
setting an acidification warning value set according to the positions of all the sea water areas;
and sending seawater acidification proximity warning information according to the acidification warning value and the prediction result.
6. The seawater acidification early warning method according to claim 1, further comprising:
obtaining standard layer lattice point data according to the prediction result;
and updating the trained LSTM model into a seawater pH value calculation model according to the standard layer lattice point data.
7. The seawater acidification early warning method according to claim 1, further comprising:
acquiring real-time observation information, wherein the real-time observation information comprises at least one of sea temperature, salinity, nitrate, ammonium salt, phosphate, silicate and partial pressure of carbon dioxide on the surface of sea water;
And correcting the prediction result according to the real-time observation information.
8. The seawater acidification early warning method according to claim 7, further comprising: and correcting and updating the trained LSTM model based on the DNN framework according to the real-time observation information.
9. A seawater acidification early warning forecasting system (100), characterized by comprising:
a data subsystem (101) comprising a data control module (1011), a field buoy observation data receiving module (1012), an operation module (1013), a first processing module (1014) and a first storage module (1015);
a prediction subsystem (102) comprising an early warning control module (1021), a prediction control module (1022), a correction module (1023) and a second storage module (1024);
an application subsystem (103) comprising an adjacent pre-warning evaluation module (1031), a second processing module (1032), a production module (1033) and an archiving module (1034);
a controller (104) for implementing the method according to any of claims 1-8 by controlling at least one of the data subsystem (101), the prediction subsystem (102) and the application subsystem (103).
10. An electronic device (200), characterized by comprising:
memory (202), a processor (201) and a computer program stored on the memory (202) and executable on the processor (201), the processor (201) implementing the method according to any one of claims 1-8 when executing the program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311263362.5A CN117253344A (en) | 2023-09-27 | 2023-09-27 | Seawater acidification early warning and forecasting method, system and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311263362.5A CN117253344A (en) | 2023-09-27 | 2023-09-27 | Seawater acidification early warning and forecasting method, system and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117253344A true CN117253344A (en) | 2023-12-19 |
Family
ID=89127572
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311263362.5A Pending CN117253344A (en) | 2023-09-27 | 2023-09-27 | Seawater acidification early warning and forecasting method, system and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117253344A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117457096A (en) * | 2023-12-26 | 2024-01-26 | 山东省海洋资源与环境研究院(山东省海洋环境监测中心、山东省水产品质量检验中心) | Dynamic monitoring and adjusting system for simulating carbon dioxide dissolution in ocean acidification device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107132530A (en) * | 2017-05-13 | 2017-09-05 | 西安费斯达自动化工程有限公司 | My god/the integral monitoring system of sky/ships and light boats/buoy/islands and reefs base marine site multisensor |
| CN116502394A (en) * | 2023-02-17 | 2023-07-28 | 国家海洋环境监测中心 | Evaluation method of seawater acidification control factors |
-
2023
- 2023-09-27 CN CN202311263362.5A patent/CN117253344A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107132530A (en) * | 2017-05-13 | 2017-09-05 | 西安费斯达自动化工程有限公司 | My god/the integral monitoring system of sky/ships and light boats/buoy/islands and reefs base marine site multisensor |
| CN116502394A (en) * | 2023-02-17 | 2023-07-28 | 国家海洋环境监测中心 | Evaluation method of seawater acidification control factors |
Non-Patent Citations (2)
| Title |
|---|
| 刘晓辉;孙丹青;黄备;王健鑫;: "东海沿岸海域表层海水酸化趋势及影响因素研究", 海洋与湖沼, no. 02, 15 March 2017 (2017-03-15) * |
| 许崇轩: "《基于相空间重构和双向门控循环单元神经网络的海水指标预测研究》", 中国优秀学位论文, 1 May 2023 (2023-05-01), pages 1 - 67 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117457096A (en) * | 2023-12-26 | 2024-01-26 | 山东省海洋资源与环境研究院(山东省海洋环境监测中心、山东省水产品质量检验中心) | Dynamic monitoring and adjusting system for simulating carbon dioxide dissolution in ocean acidification device |
| CN117457096B (en) * | 2023-12-26 | 2024-03-22 | 山东省海洋资源与环境研究院(山东省海洋环境监测中心、山东省水产品质量检验中心) | Dynamic monitoring and adjusting system for simulating carbon dioxide dissolution in ocean acidification device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114493052B (en) | Multi-model fusion self-adaptive new energy power prediction method and system | |
| CN117253344A (en) | Seawater acidification early warning and forecasting method, system and electronic equipment | |
| KR20160060793A (en) | Typhoon surge automatic forecasting method using active data collection type script and numerical model | |
| CN112949178A (en) | Sea surface wind field wind speed intelligent prediction forecasting system based on deep learning, computer equipment and storage medium | |
| CN114399073A (en) | Ocean surface temperature field prediction method based on deep learning | |
| CN116306318B (en) | Three-dimensional ocean thermal salt field forecasting method, system and equipment based on deep learning | |
| CN113806349A (en) | Spatiotemporal missing data completion method, device and medium based on multi-view learning | |
| Xie et al. | Developing a deep learning-based storm surge forecasting model | |
| Jiang et al. | [Retracted] Ocean Observation Data Prediction for Argo Data Quality Control Using Deep Bidirectional LSTM Network | |
| Wang et al. | Customized deep learning for precipitation bias correction and downscaling | |
| Meng et al. | Physical knowledge-enhanced deep neural network for sea surface temperature prediction | |
| Meng et al. | Tropical cyclone intensity probabilistic forecasting system based on deep learning | |
| Fu et al. | Simulation and forecasting of fishery weather based on statistical machine learning | |
| CN112016744B (en) | Forest fire prediction method and device based on soil moisture and storage medium | |
| CN114139590A (en) | Method for estimating ocean temperature | |
| CN115206040A (en) | Biological invasion early warning method, device and terminal for nuclear power water intake | |
| CN110930024B (en) | Intelligent analysis system and method for deep sea emergency situation based on Bayesian network | |
| CN114943189B (en) | XGboost-based acoustic velocity profile inversion method and system | |
| CN114202551B (en) | Grading drawing method and grading drawing device for karst stony desertification | |
| CN113888380A (en) | Method, device, equipment and medium for predicting man overboard trajectory | |
| Wang et al. | Dynamic graphs attention for ocean variable forecasting | |
| CN111325384A (en) | NDVI prediction method combining statistical characteristics and convolutional neural network model | |
| KR101568819B1 (en) | Coastal erosion automatic forecasting method using active data collection type script and numerical model | |
| CN116978191B (en) | Ocean disaster prediction and early warning method and system | |
| CN112508204A (en) | Short-term forecasting method, device, equipment and medium for sea surface algae |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |