CN105893329A - Monthly-scale-based tide level data consistency correction method - Google Patents
Monthly-scale-based tide level data consistency correction method Download PDFInfo
- Publication number
- CN105893329A CN105893329A CN201610249068.2A CN201610249068A CN105893329A CN 105893329 A CN105893329 A CN 105893329A CN 201610249068 A CN201610249068 A CN 201610249068A CN 105893329 A CN105893329 A CN 105893329A
- Authority
- CN
- China
- Prior art keywords
- item
- point
- tidal level
- tide gauge
- moon
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
- G06F17/148—Wavelet transforms
-
- 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
- G06N3/10—Interfaces, programming languages or software development kits, e.g. for simulating neural networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Operations Research (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Other Liquid Machine Or Engine Such As Wave Power Use (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a monthly-scale-based tide level data consistency correction method. The method comprises the steps of establishing a relation between an actual measurement tide level series and trend terms, periodic terms as well as random factor items; determining positions where abnormal points and abrupt change points appear by means of comparing residual errors of different scales of wavelet inspection and calculating expectation of modules of wavelet decomposition coefficients in two wavelet scales; applying an average value to replace the abnormal value to eliminate the interference of the random factor items; eliminating the interference of the periodic terms by means of moving average processing; establishing a correlation among the trend terms, tide level data obtained after the periodic terms are eliminated and actual measurement data; selecting lunar tide level data of the recent five years to train a BP neural network, and calculating by utilizing the network obtained by training to obtain a corrected value. The monthly-scale-based tide level data consistency correction method solves the problem of monthly-scale tide level data consistency correction, and eliminates the interference of the random factor items in a correction process; the corrected tide level data is calculated by means of a trained BP neural network model, so that the reliability and accuracy of correction are improved.
Description
Technical field
The present invention relates to hydrological data revision field, especially relate to a kind of tide based on moon yardstick
Position Data coordinating modification method.
Background technology
When carrying out the planning of relevant hydraulic engineering construction, generally we think such as water level, tide
Environment produced by the data such as position, rainfall is constant, thus the hydrology we studied is wanted
Prime sequences regards simple stochastic variable as, and the concordance of its factor of influence makes data system also
There is relevant coherency accordingly, it is simple to our research.
For the tide gauge of each sea outfall door, along with coastal area expanding economy, people
Class activity has had a strong impact on the land surface condition that tidal level is formed, especially for sanidal transformation
Engineering etc., above-mentioned impact change in tide gauge dependency.Carrying out tidal level impact ground
The hydraulic condition of river simulation in district or river course calculation and the tidal level to non-avaible mouth door carry out forecast and grind
When studying carefully, tide gauge is particularly important accurately.
Current research mainly carries out general correction on year yardstick to tide gauge；Repairing
Serial sudden change value or exceptional value are not processed during just；Pass through method of least square
It is fitted can not well reflecting tide gauge variation tendency and measured value to revising series
Between relation.
In a word, tide gauge tool is of great significance, and either the river course of tidal reach is built
If, or urban flood defence, a set of tide level data after revising can help us more accurate
True forecasts the water level being likely to occur in river course, accomplishes to prevent trouble before it happens.At present to tidal level system
The method that row are modified is little, the most fairly simple, is often both for a year tide gauge and enters
Row is revised, and the method suitability is preferable, obtains revised moon tide gauge, can be with upper
Trip river course water situation carries out more preferable combination research.
Summary of the invention
The present invention devises a kind of tide gauge consistent correction method based on moon yardstick, and it solves
Technical problem is that certainly carries out general correction to tide gauge at present on year yardstick, is revising
During not sudden change value or exceptional value to series process, by method of least square pair
Revise series be fitted well reflecting tide gauge variation tendency and measured value it
Between relation, and ultimately result in both relation distortions.
In order to solve the technical problem of above-mentioned existence, present invention employs below scheme:
A kind of tide gauge consistent correction method based on moon yardstick, comprises the following steps:
Step 1, the moon tide gauge of a certain website is sequentially arranged into tidal level series；Institute
Stating tide gauge is actual measurement moon highest water level data or actual measurement moon lowest water level data；
Step 2, the catastrophe point eliminated in tidal level series and abnormity point:
Step 3, the periodic term factor eliminated in tidal level series: step 3.1, to eliminate random because of
The tidal level series of subitem carries out wavelet analysis, obtains wavelet variance diagram, analysis wavelet variogram institute
Show the main energy cycle of series, so that it is determined that the primary period n, n of tidal level series are natural number；Step
Rapid 3.2, tidal level series proceeded by moving average from n/2 item process, each moving average relates to
And n item, result W (t) after the periodic term that is eliminated；If n/2 is non-integer, then take n/2
For [n]+1, the integer part i.e. taking n adds 1 for n/2；
Step 4, the tide gauge of nearly N may be considered generation under tale quale, and N is
The natural number of >=5, calculates the moon tide gauge of nearly a year and averages A, and meansigma methods A represents existing
Trend term under the conditions of shape, actual measurement item and the difference of meansigma methods A under calculating tale quale, but recently
The data of 1 year are not involved in calculating, and only calculate the actual measurement item of N-1 and the difference of meansigma methods A；
Step 5, calculating chart to N near under tale quale collect statistics, but do not include
The moon data of nearest 1 year, statistical items includes that under tale quale, each time point eliminates abnormity point and dashes forward
Tidal level value X (t) of height, eliminate result W (t) after periodic term, actual measurement item and meansigma methods A it
Difference α (t)；Wherein,
α (t)=A-X (t)；
Step 6, choose BP neutral net and ask for correction value, choose the moon data of nearly N, no
Including the data of nearest 1 year；(N-1) * 12 samples are as training set altogether, and input item is for disappearing
Except the result after periodic term and the difference of actual measurement item and meansigma methods A, output item for eliminate abnormity point and
The tidal level value of catastrophe point, and the neutral net after must training；
Step 7, the correction value of each time point of neural computing obtained by step 6 training,
Input item is the result after eliminating periodic term and difference α surveying item and meansigma methods At, output is
The correction value in each year；
Step 8, the tide gauge of nearest a year use measured value to be correction value, and historical summary initiates
Correction value before n/2 all replaces by n/2 correction value, and n is natural number, if n/2 is non-whole
Number, then taking n/2 is [n]+1, and the integer part i.e. taking n adds 1 for n/2.
Further, by step 1 tide gauge survey sequence regard as periodic term and random because of
The combination of subitem, formula is:
Z (t)=f [A (t), P (t), R (t)],
Wherein Z (t) represents actual measurement tidal level sequence, and what A (t), P (t), R (t) represented respectively is to become
Gesture item, periodic term and random factor item, t represents the time.
Further, step 2 is made up of step by step following:
Step 2.1, ask taken tidal level series meansigma methods；
Step 2.2, utilization wavelet analysis method, to the plan that tidal level is serial under two wavelet scales
Close residual error and carry out online wavelet decomposition；
Under step 2.3, two wavelet scales of calculating, the mould of coefficient of wavelet decomposition, calculates difference and obtains
The expectation E of eacht；
Step 2.4, detection abnormity point and catastrophe point；
Step 2.5, choose the time that abnormity point and catastrophe point occur according to detection principle, with average
Value replaces exceptional value or sudden change value.
Further, if step 2.2 does not occurs modulus maximum point, and not having in step 2.3
There is sudden change, illustrate at this neither abnormity point is not catastrophe point；If step 2.2 occurs mould
Not sudden change in maximum point, and step 2.3, illustrates it is abnormity point at this；If step
Modulus maximum point occurs in 2.2, and step 2.3 has sudden change, illustrate it is catastrophe point at this.
Further, the meansigma methods of tidal level series of the value at abnormity point and catastrophe point is replaced, logical
Cross the process to abnormity point and catastrophe point, eliminate random factor item R (t) and tidal level series is done
Disturb, be expressed as:
X (t)=g [A (t), P (t)]；What A (t), P (t), R (t) represented respectively is trend term, week
Phase item and random factor item；X (t) represents the tidal level series eliminated after random factor item, t generation
The table time.
A kind of application using above-mentioned modification method reduction tide gauge, it is characterised in that: pass through
Revise and eliminate the changing environment impact on data, reduce the concordance of tide gauge, it is ensured that
Use tide gauge to carry out the reliability of hydrologic(al) frequency analysis, use revised moon yardstick
Tide gauge provides reliable foundation for urban flood defence.Such as: limit flood bank builds height.
Should tide gauge consistent correction method based on moon yardstick have the advantages that
(1) the inventive method has solved and has repaiied tide gauge currently without on moon yardstick
Positive problem, in method, the exceptional value to tide gauge is processed, and solves exceptional value pair
The impact of correction result.
(2) the inventive method uses bp neural network model to forecast correction value, solves and repaiies
Positive error is big, the problems such as degree of accuracy is the highest.
(3) tide level data to moon yardstick that the present invention is initiative has carried out consistent correction, builds
Found the relation between actual measurement tidal level series and trend term, periodic term and random factor item, passed through
Coefficient of wavelet decomposition under the residual error of contrast different scale small echo inspection and two wavelet scales of calculating
The expectation of mould, determine the position that abnormity point and catastrophe point occur, use meansigma methods to replace this different
Constant value eliminates the interference of random factor item.The interference eliminating periodic term is processed by moving average.
Tide gauge after trend term, elimination periodic term and field data are set up dependency relation, choosing
Taking nearly tide gauge training on 5 days BP neutral net, the network calculations using training to obtain obtains
Correction value.The problem solving the tide gauge consistent correction of moon yardstick, in makeover process
Eliminate the interference of random factor item, calculate revised by the BP neural network model of training
Tide gauge, improves reliability and the precision of correction.
Detailed description of the invention
Below in conjunction with embodiment, the present invention will be further described:
Step 1, by a certain website actual measurement highest water level data or lowest water level data according to year
Order statistics arranges, and in year, the order according to the moon arranges.
Step 2, the catastrophe point eliminated in tidal level series and abnormity point:
(1) meansigma methods of taken tidal level series is sought.
(2) using wavelet analysis method, under two wavelet scales, the matching to tidal level series is residual
Difference carries out online wavelet decomposition, can be chosen at line small echo recursive decomposition method and decompose small echo.
(3) calculate the mould of coefficient of wavelet decomposition under two wavelet scales, calculate difference and obtain each
The expectation E of itemt。
(4) detection abnormity point and catastrophe point
1) if (2) do not occur modulus maximum point, and not sudden change, the explanation in (3)
Neither abnormity point is not catastrophe point at this.
2) if (2) occur modulus maximum point, and the not sudden change in (3), this is described
Place is abnormity point.
3) if (2) occur modulus maximum point, and (3) there is sudden change, illustrate at this
It it is catastrophe point.
4) meansigma methods of tidal level series of the value at abnormity point and catastrophe point is replaced, by different
Often point and the process of catastrophe point, can eliminate random factor item R (t) interference to tidal level series,
It is expressed as: X (t)=g [A (t), P (t)].
Step 3, the periodic term factor eliminated in tidal level series:
3.1, the tidal level series eliminating random factor item is carried out wavelet analysis, obtain wavelet variance
Figure, the main energy cycle of series shown in analysis wavelet variogram, so that it is determined that the master of tidal level series
Cycle n.
3.2, from n/2 item, tidal level series is proceeded by moving average to process, each moving average
Relate to n item, tidal level series W (t) of the periodic term that is eliminated.
Step 4, the tide gauge of nearly 5 years may be considered generation under tale quale, calculate
Average A, A of the moon tide gauge of nearly 1 year can represent the trend term under tale quale, meter
Actual measurement item and the difference of meansigma methods A under calculation tale quale:
α (t)=A-X (t).
Step 5, the calculating chart of nearly 5 years under tale quale is collected statistics (do not include
The moon data of nearly 1 year), statistical items includes that under tale quale, each time point eliminates abnormity point and dashes forward
Tidal level value X (t) of height, eliminate result W (t) after periodic term, actual measurement item and meansigma methods A it
Difference α (t).
Step 6, choose BP neutral net and ask for correction value, choose nearly 5 years and (do not include recently
The data of 1 year) moon data, totally 48 samples are as training set, and input item is the elimination cycle
Result W (t) after Xiang and actual measurement item and difference α (t) of meansigma methods A, output item is for eliminating abnormity point
Tidal level value X (t) with catastrophe point.
Step 7, the correction of each time point of BP neural computing obtained by step 6 training
Value, input item is result W (t) after eliminating periodic term and surveys item and meansigma methods A and actual measurement item
Difference α with meansigma methods At, output is the correction value in each year.
Step 8, the data of nearest a year use measured value to be correction value, the historical summary starting year
Correction value before n/2 all replaces by n/2 correction value.
Above in conjunction with embodiment, the present invention is carried out exemplary description, it is clear that the reality of the present invention
Now it is not subject to the restrictions described above, as long as have employed method design and the technical scheme of the present invention
The various improvement carried out, or the most improved design by the present invention and technical scheme directly apply to
Other occasion, the most within the scope of the present invention.
Claims (6)
1. a tide gauge consistent correction method based on moon yardstick, comprises the following steps:
Step 1, the moon tide gauge of a certain website is sequentially arranged into tidal level series；Institute
Stating tide gauge is actual measurement moon highest water level data or actual measurement moon lowest water level data；
Step 2, the catastrophe point eliminated in tidal level series and abnormity point；
Step 3, the periodic term factor eliminated in tidal level series: step 3.1, to eliminate random because of
The tidal level series of subitem carries out wavelet analysis, obtains wavelet variance diagram, analysis wavelet variogram institute
Show the main energy cycle of series, so that it is determined that the primary period n, n of tidal level series are natural number；Step
Rapid 3.2, tidal level series proceeded by moving average from n/2 item process, each moving average relates to
And n item, result W (t) after the periodic term that is eliminated；If n/2 is non-integer, then take n/2
For [n]+1, the integer part i.e. taking n adds 1 for n/2；
Step 4, the tide gauge of nearly N may be considered generation under tale quale, and N is
The natural number of >=5, calculates the moon tide gauge of nearly a year and averages A, and meansigma methods A represents existing
Trend term under the conditions of shape, actual measurement item and the difference of meansigma methods A under calculating tale quale, but recently
The data of 1 year are not involved in calculating, and only calculate the actual measurement item of N-1 and the difference of meansigma methods A；
Step 5, calculating chart to N near under tale quale collect statistics, but do not include
The moon data of nearest 1 year, statistical items includes that under tale quale, each time point eliminates abnormity point and dashes forward
Tidal level value X (t) of height, eliminate result W (t) after periodic term, actual measurement item and meansigma methods A it
Difference α (t)；Wherein,
α (t)=A-X (t)；
Step 6, choose BP neutral net and ask for correction value, choose the moon data of nearly N, no
Including the data of nearest 1 year；(N-1) * 12 samples are as training set altogether, and input item is for disappearing
Except the result after periodic term and the difference of actual measurement item and meansigma methods A, output item for eliminate abnormity point and
The tidal level value of catastrophe point, and the neutral net after must training；
Step 7, the correction value of each time point of neural computing obtained by step 6 training,
Input item is the result after eliminating periodic term and difference α surveying item and meansigma methods At, output is
The correction value in each year；
Step 8, the tide gauge of nearest a year use measured value to be correction value, and historical summary initiates
Correction value before n/2 all replaces by n/2 correction value, and n is natural number, if n/2 is non-whole
Number, then taking n/2 is [n]+1, and the integer part i.e. taking n adds 1 for n/2.
Tide gauge consistent correction method based on moon yardstick the most according to claim 1,
It is characterized in that: by step 1 tide gauge survey sequence regard as periodic term and random because of
The combination of subitem, formula is:
Z (t)=f [A (t), P (t), R (t)],
Wherein Z (t) represents actual measurement tidal level sequence, and what A (t), P (t), R (t) represented respectively is to become
Gesture item, periodic term and random factor item, t represents the time.
Tide gauge consistent correction method based on moon yardstick the most according to claim 1,
It is characterized in that:
Step 2 is made up of step by step following:
Step 2.1, ask taken tidal level series meansigma methods；
Step 2.2, utilization wavelet analysis method, to the plan that tidal level is serial under two wavelet scales
Close residual error and carry out online wavelet decomposition；
Under step 2.3, two wavelet scales of calculating, the mould of coefficient of wavelet decomposition, calculates difference and obtains
The expectation E of eacht；
Step 2.4, detection abnormity point and catastrophe point；
Step 2.5, choose the time that abnormity point and catastrophe point occur according to detection principle, with average
Value replaces exceptional value or sudden change value.
Tide gauge consistent correction method based on moon yardstick the most according to claim 3,
It is characterized in that: if step 2.2 does not occurs modulus maximum point, and not having in step 2.3
There is sudden change, illustrate at this neither abnormity point is not catastrophe point；If step 2.2 occurs mould
Not sudden change in maximum point, and step 2.3, illustrates it is abnormity point at this；If step
Modulus maximum point occurs in 2.2, and step 2.3 has sudden change, illustrate it is catastrophe point at this.
5. consistent according to tide gauge based on moon yardstick in any of the one of claim 1-4
Property modification method, it is characterised in that: to flat by tidal level series of the value at abnormity point and catastrophe point
Average replaces, and by the process to abnormity point and catastrophe point, eliminates random factor item R (t) right
The interference of tidal level series, is expressed as:
X (t)=g [A (t), P (t)]；What A (t), P (t), R (t) represented respectively is trend term, week
Phase item and random factor item；X (t) represents the tidal level series eliminated after random factor item, t generation
The table time.
6. use an application for modification method reduction tide gauge described in claim 1-5, its
It is characterized by revising the impact eliminating changing environment to data, reduces tide gauge
Concordance, it is ensured that using tide gauge to carry out the reliability of hydrologic(al) frequency analysis, use is repaiied
After just the moon yardstick tide gauge be urban flood defence provide reliable foundation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610249068.2A CN105893329B (en) | 2016-04-20 | 2016-04-20 | A kind of tide gauge consistent correction method based on moon yardstick |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610249068.2A CN105893329B (en) | 2016-04-20 | 2016-04-20 | A kind of tide gauge consistent correction method based on moon yardstick |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105893329A true CN105893329A (en) | 2016-08-24 |
| CN105893329B CN105893329B (en) | 2017-08-25 |
Family
ID=56705249
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610249068.2A Active CN105893329B (en) | 2016-04-20 | 2016-04-20 | A kind of tide gauge consistent correction method based on moon yardstick |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105893329B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109190254A (en) * | 2018-09-05 | 2019-01-11 | 上海交通大学 | A kind of tidal level variation forecasting procedure based on real-time measuring data |
| CN113806959A (en) * | 2021-09-26 | 2021-12-17 | 上海市水利工程设计研究院有限公司 | High-tide-level quantitative research method for estuary design under future situation |
| CN114693002A (en) * | 2022-05-23 | 2022-07-01 | 中国海洋大学 | Tide level prediction method, device, electronic equipment and computer storage medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102214262A (en) * | 2010-04-02 | 2011-10-12 | 上海海洋大学 | Tide predicting method |
-
2016
- 2016-04-20 CN CN201610249068.2A patent/CN105893329B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102214262A (en) * | 2010-04-02 | 2011-10-12 | 上海海洋大学 | Tide predicting method |
Non-Patent Citations (3)
| Title |
|---|
| 包为民，元晓华，陈斌，张小琴: "新三江闸天文潮预报及修正方法研究", 《水电能源科学》 * |
| 罗毅，阳洋: "一种遗传神经网络潮位预测算法", 《电力科学与工程》 * |
| 肖淼元: "长江口潮位一致性检验与修正方法研究", 《长江科学院院报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109190254A (en) * | 2018-09-05 | 2019-01-11 | 上海交通大学 | A kind of tidal level variation forecasting procedure based on real-time measuring data |
| CN109190254B (en) * | 2018-09-05 | 2023-07-25 | 上海交通大学 | Tidal level change forecasting method based on real-time measurement data |
| CN113806959A (en) * | 2021-09-26 | 2021-12-17 | 上海市水利工程设计研究院有限公司 | High-tide-level quantitative research method for estuary design under future situation |
| CN113806959B (en) * | 2021-09-26 | 2024-03-15 | 上海市水利工程设计研究院有限公司 | Quantitative research method for estuary design high tide level in future situation |
| CN114693002A (en) * | 2022-05-23 | 2022-07-01 | 中国海洋大学 | Tide level prediction method, device, electronic equipment and computer storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105893329B (en) | 2017-08-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Le Cozannet et al. | Approaches to evaluate the recent impacts of sea-level rise on shoreline changes | |
| Nkuna et al. | Filling of missing rainfall data in Luvuvhu River Catchment using artificial neural networks | |
| CN106814378B (en) | A kind of GNSS location time series cyclophysis method for digging | |
| CN103268525B (en) | A kind of Hydrological Time Series simulating and predicting method based on WD-RBF | |
| CN105893329A (en) | Monthly-scale-based tide level data consistency correction method | |
| Sahu et al. | Density-based spatial clustering of application with noise approach for regionalisation and its effect on hierarchical clustering | |
| Prass et al. | Comparison of forecasts of mean monthly water level in the Paraguay River, Brazil, from two fractionally differenced models | |
| Bailly et al. | Modelling spatial variability along drainage networks with geostatistics | |
| Kowalewski et al. | Sensitivity of the Baltic Sea level prediction to spatial model resolution | |
| Chen et al. | A high speed method of SMTS | |
| CN117113854B (en) | Salt tide forecasting method based on ConvLSTM and three-dimensional numerical simulation | |
| CN104268662A (en) | Settlement prediction method based on step-by-step optimization quantile regression | |
| Sertel et al. | Estimating daily mean sea level heights using artificial neural networks | |
| Rajabi-Kiasari et al. | Forecasting of absolute dynamic topography using deep learning algorithm with application to the Baltic Sea | |
| Shahzad et al. | Flood forecasting for River Mekong with data‐based models | |
| CN116861955A (en) | Method for inverting submarine topography by machine learning based on topography unit partition | |
| CN106403908A (en) | Water depth prediction method and system based on time sequence | |
| Kim | High-spatial-resolution streamflow estimation at ungauged river sites or gauged sites with missing data using the National Hydrography Dataset (NHD) and US Geological Survey (USGS) streamflow data | |
| Yakubu et al. | Modelling Uncertainties in Differential Global Positioning System Dataset | |
| CN113627465A (en) | Rainfall data space-time dynamic fusion method based on convolution long-short term memory neural network | |
| Shi et al. | A Grey Model for Short-Term Prediction of the Ionospheric TEC | |
| CN116204836B (en) | Method for screening suitable forest areas of mangrove forest on estuary beach | |
| Morovati et al. | Utilizing hybrid machine learning techniques and gridded precipitation data for advanced discharge simulation in under-monitored river basins | |
| Jahanmard et al. | Quantification of hydrodynamic model sea level bias utilizing deep learning and synergistic integration of data sources | |
| Chu et al. | Optimization scheme for coastal morphodynamic model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |