CN116796799A - Method for creating small-river basin flood rainfall threshold model in area without hydrologic data - Google Patents
Method for creating small-river basin flood rainfall threshold model in area without hydrologic data Download PDFInfo
- Publication number
- CN116796799A CN116796799A CN202210177755.3A CN202210177755A CN116796799A CN 116796799 A CN116796799 A CN 116796799A CN 202210177755 A CN202210177755 A CN 202210177755A CN 116796799 A CN116796799 A CN 116796799A
- Authority
- CN
- China
- Prior art keywords
- model
- data
- basin
- river basin
- flood
- 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
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000012549 training Methods 0.000 claims abstract description 27
- 238000012360 testing method Methods 0.000 claims abstract description 21
- 238000011156 evaluation Methods 0.000 claims abstract description 12
- 238000005070 sampling Methods 0.000 claims description 13
- 238000004458 analytical method Methods 0.000 claims description 9
- 238000010276 construction Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 8
- 238000013527 convolutional neural network Methods 0.000 claims description 5
- 239000002689 soil Substances 0.000 claims description 5
- 238000009825 accumulation Methods 0.000 claims description 4
- 238000013136 deep learning model Methods 0.000 claims description 3
- 238000011176 pooling Methods 0.000 claims description 3
- 101100421300 Schizosaccharomyces pombe (strain 972 / ATCC 24843) set8 gene Proteins 0.000 claims 2
- 231100000656 threshold model Toxicity 0.000 abstract 1
- 238000013135 deep learning Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 238000004088 simulation Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 201000009409 embryonal rhabdomyosarcoma Diseases 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
Classifications
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/211—Schema design and management
- G06F16/212—Schema design and management with details for data modelling support
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
- Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping
Abstract
The invention provides a method for creating flood rainfall threshold models in different reproduction periods of small watershed in a region without hydrologic data, which comprises the following steps: step 1, establishing a drainage basin database; step 2, determining the input and output of the model; step 3, creating a model sample data set; step 4, constructing a model network; step 5, dividing a sample training set and a test set; step 6, model training; step 7, model testing and evaluation; and 8, optimizing the model. Aiming at the problem that the difficulty of estimating the rainfall threshold value of the flood critical surface by applying the traditional hydrologic method is high for the middle and small watershed in the area without hydrologic data, the method for creating the flood rainfall threshold value model of the middle and small watershed in the area without hydrologic data in different reproduction periods is provided, and an effective technical method is provided for hydrologic forecasting of the middle and small watershed in the area without hydrologic data.
Description
Technical Field
The invention relates to a flood rainfall threshold model creation method, in particular to a flood rainfall threshold model creation method for small watershed in different reproduction periods in a non-hydrologic data area.
Background
Flood disasters are one of the most important meteorological disasters in China. Flood monitoring, forecasting and early warning are important non-engineering measures for preventing and reducing flood disaster. The threshold value of the rainfall of the critical flood disaster surface is an important index for monitoring and early warning flood disasters, and small and medium rivers often lack enough hydrologic information and data due to low distribution density of hydrologic sites, and the areas are often key objects for carrying out flood control and flood prevention work in all areas.
At present, a traditional hydrologic model based on a physical mechanism or a physical concept is established for a river basin by a parameterized method for determining a rainfall threshold of a critical flood disaster surface in flood forecast, but the model is difficult to operate, the parameter adjustment is complex, and the application of the hydrologic model is more difficult for small and medium-sized river basins with deficient hydrologic data. Meanwhile, the conditions of meteorological hydrology, topography elevation, land utilization type, vegetation type and the like of the middle and small watershed are complex and various, so that the production and convergence process of the watershed is complex and various, and the traditional hydrologic model can not meet the development requirements of the current hydrologic forecast. With the progress of scientific information, the data acquisition modes are gradually diversified, and hydrological process simulation and prediction methods based on data driving technology are also increasingly developed.
For the problem that the rainfall threshold value of the disaster-causing face of the flood in the small river basin in the area without the hydrological data is difficult to estimate or has larger error by using the traditional method, the space variability of the rainfall and the underlying surface characteristics of the river basin can be better considered by using the deep learning technology, and the rainfall threshold value of the disaster-causing face of the flood in the small river can be accurately estimated by considering the associated factors such as vegetation, soil and the like.
Disclosure of Invention
The invention aims to provide a method for creating a flood rainfall threshold model of different reproduction periods of small and medium-sized watercourses in a non-hydrologic data area, which can be used for estimating the flood disaster critical surface rainfall threshold of different reproduction periods of the small and medium-sized watercourses in the non-hydrologic data area.
In order to achieve the above purpose, the invention adopts the following technical scheme:
the invention provides a method for creating a flood rainfall threshold model of small and medium-sized watershed in different reproduction periods in a region without hydrologic data, which sequentially comprises the following steps:
step 1, establishing a drainage basin database: acquiring a rainfall grid data set of a watershed in a region without hydrologic data, grid data of sub-watersheds of different grades, flood rainfall threshold data of different reproduction periods and underlying surface data; the different-level sub-river basin grid data sets are respectively obtained by using a hydrological analysis toolbox of ArcGIS based on river network confluence accumulation thresholds of three-level, four-level and five-level river basins.
Step 2, determining input and output of a model: the model is input into a sublevel surface characteristic of a river basin, the sublevel surface characteristic of the river basin comprises 6 factors of digital elevation model DEM data, gradient data, slope data, land utilization data, normalized vegetation indexes and temperature vegetation drought indexes as characteristic variables X of the river basin, and flood rainfall thresholds of the river basins of different grades in each recurrence period are output as target variables Y.
Step 3, creating a model sample data set: the input of the model is the underlying surface characteristic of the river basin, the output is the flood rainfall threshold values of each reproduction period of the sub-river basin with different grades, and the sample data set of the river basin is subjected to standardized processing and then is subjected to sample collection;
step 4, sample training set and test set division: dividing the collected sample data set into a training set and a testing set;
step 5, model network construction: constructing flood rainfall threshold prediction models of different reproduction periods of the drainage basin based on a convolutional neural network Squeeze Net;
step 6, model training: training a model by using a training data set to obtain model parameters, and establishing flood rainfall threshold prediction models of different reproduction periods;
step 7, model test and evaluation: inputting the test sample into the trained model, and calculating an evaluation index to evaluate the precision of the test model;
step 8, optimizing a model: and modifying model parameters according to model evaluation indexes, and obtaining an optimal deep learning model for simulating rainfall of flood critical surfaces of different reproduction periods of the river basin through repeated training and testing.
Further, in step 1, the different-level sub-basin raster data sets are sub-basin raster data sets corresponding to three different-level basins respectively obtained by inputting river network confluence accumulation amount thresholds of the three-level, four-level and five-level basins based on a basin digital elevation model by using a hydrological analysis tool box of an ArcGIS.
Further, in step 1, the Flood rainfall threshold data of different reproduction periods of the basin is a Flood rainfall threshold data set of different reproduction periods of the basin corresponding to three different levels of sub-basins by using a GIS Flood Tool and using a basin digital elevation model and rainfall grid data as basic data and estimating the basin through hydrologic analysis and a regional regression equation.
Further, in step 2, the drainage basin underlying surface data includes drainage basin digital elevation model DEM data, gradient data, slope data, land use type data, land use 8 OLI remote sensing data, wherein the land use type data is converted into land use degree, and a normalized vegetation index NDVT and a temperature vegetation drought index TVDI of the drainage basin are calculated based on the land use 8 OLI data in ENVI5.3 to represent the situations of vegetation coverage and soil humidity of the drainage basin respectively.
Further, in step 3, the 6 different factors X of the basin are normalized to between 0 and 100. And then carrying out sample collection through a program based on 6 different factors X, different levels of sub-basin raster data B and different reappearance period flood rainfall threshold data Y after basin standardization processing.
Further, in step 3, three sample collection methods are implemented: whole sampling, multiple reproduction period random sampling of the same drainage basin, multiple present period random sampling of different drainage basins.
Further, in step 3, the sample data set includes data sets obtained by three sample collection modes, and each sampling mode includes four combination cases of single-drainage-single-reproduction period, single-drainage-multiple-reproduction period, multiple-drainage-single-reproduction period and multiple-drainage-multiple-reproduction period.
Further, in step 5, the construction model adopts a convolutional neural network Squeeze Net construction model, and comprises an input layer, an implied layer and an output layer, wherein the implied layer comprises a convolutional layer, a pooling layer, a full-connection layer, a Softmax layer and a classification layer, the input layer and the first convolutional layer of the modified network are used as the number of rows and columns of the input network by identifying the number of rows and columns of sample images, so that the input size and the number of channels of the input layer can be customized according to the characteristics of the input image, the number of channels of the first convolutional layer is enabled to be consistent with the number of image wave bands, the full-connection layer, the Softmax layer and the classification layer are modified, the output size of the full-connection layer is enabled to be 1, the Softmax layer and the classification layer are replaced by the regression layer, and then the training sample input model is iterated, and the training model is obtained.
Further, in step 6, the model parameters include: maximum generation number, learning rate, minimum batch number, and solver.
Further, in step 7, the model evaluation index includes: root mean square error, relative deviation, average absolute error and correlation coefficient.
The beneficial effects of the invention are as follows: based on ArcGIS and deep learning technology, the method for creating the flood rainfall threshold model of the small and medium-sized river basins in the area without hydrologic data in different reproduction periods is provided, so that the rainfall threshold of the flood disaster critical surface rainfall threshold of the small and medium-sized river basins can be estimated more accurately by considering the space variability of the rainfall and the underlying surface characteristics of the small and medium-sized river basins and considering the associated factors such as vegetation, soil and the like.
(1) The method utilizes a deep learning technology to extract the characteristics of the sublevel surface of the river basin, comprehensively considers the influence of associated factors such as vegetation, soil, land utilization degree and the like on flood rainfall thresholds in different reproduction periods.
(2) The invention can solve the problems of large number of small watershed in the non-data area, low data processing speed, complex processing process and the like by using the deep learning technology.
Drawings
Fig. 1: the model of the present invention creates a flow chart.
Detailed Description
The invention is further described with reference to the accompanying drawings:
as shown in fig. 1, the present invention includes the steps of:
(1) Establishing a drainage basin database:
and acquiring a rainfall grid data set of the watershed in the area without the hydrologic data, grid data of different levels of sub-watersheds, flood rainfall threshold data of different reproduction periods and underlying surface data. The rainfall grid data set and the underlying data can be directly obtained by downloading through resource environment science and data centers of China academy of sciences and a geospatial data cloud website respectively, and the different-level sub-basin grid data and the different-reproduction-period Flood rainfall threshold data are indirectly obtained through the processing analysis of a hydrological analysis toolbox of an ArcGIS and a GIS Flood Tool respectively. The rainfall grid data are average rainfall grid data of each meteorological site of the river basin for a plurality of years, the underlying surface data comprise river basin digital elevation model DEM data, gradient data, slope data, land utilization type data and Landset8 OLI remote sensing data, the land utilization type data are converted into land utilization degree, and the normalized vegetation index NDVT and the temperature vegetation drought index TVDI of the river basin are obtained in ENVI5.3 based on Landset8 OLI data through calculation. The hydrologic analysis toolbox for ArcGIS for the grid data of the sub-drainage basins with different grades respectively obtains grid data sets of the sub-drainage basins corresponding to the three different grades based on river network confluence accumulation thresholds of the three-grade, four-grade and five-grade drainage basins. And using a GIS Flood Tool to estimate Flood rainfall threshold data sets of different reproduction periods of the basin corresponding to the three different levels of sub-basins by using a hydrologic analysis and regional regression equation by taking the basin digital elevation model DEM and the rainfall grid data as basic data. Wherein the reproduction period includes 7 kinds, which are 5 years first, 10 years first, 15 years first, 20 years first, 30 years first, 50 years first, and 100 years first.
(2) Model input and output determination:
the model is input into a sublevel surface characteristic of a river basin, the sublevel surface characteristic of the river basin comprises 6 factors of digital elevation model DEM data, gradient data, slope data, land utilization data, normalized vegetation indexes and temperature vegetation drought indexes as characteristic variables X of the river basin, and flood rainfall thresholds of the river basins of different grades in each recurrence period are output as target variables Y.
(3) Creating a model sample dataset:
and determining that the input of the model is the underlying surface characteristic of the river basin, and outputting the model as flood rainfall thresholds of each reproduction period of the sub-river basin of different grades. Firstly, 6 different factors X of the watershed are standardized to 0-100, and the raster data types are converted into floating point types. And then carrying out sample collection through a program based on 6 different factors X, different levels of sub-basin raster data B and different reappearance period flood rainfall threshold data Y after basin standardization processing. The three sample collection modes are as follows: whole sampling, multiple reproduction period random sampling of the same drainage basin, multiple present period random sampling of different drainage basins. Each sampling form comprises four combination modes of single drainage basin-single reproduction period, single drainage basin-multiple reproduction period, multiple drainage basins-single reproduction period and multiple drainage basins-multiple reproduction period, so that the influence of factors such as different terrains, vegetation, land utilization degree and the like of the drainage basins is considered while the number of samples is ensured.
Firstly, 6 kinds of drainage basin characteristic factors X, three levels of drainage basin grid data sets B and different reproduction period rainfall threshold values Y corresponding to the three levels of drainage basins are loaded, wherein X and B data types are in grid formats, Y data are in Excel table formats, and each table comprises 7 kinds of reproduction period rainfall threshold values. When a single drainage basin is used, one drainage basin and a corresponding table are selected, when a plurality of drainage basins are used, a Q input 1 is a single reproduction period, a Q input 2 is a multiple present period, and each sampling mode corresponds to four combination modes.
(4) Sample training set and test set partitioning:
and (3) dividing all sample data sets obtained in the previous step into a training set and a testing set respectively, wherein the dividing ratio of the training set to the testing set is 8:2.
(5) Model construction:
and establishing a flood rainfall threshold simulation model of different reproduction periods of the river basin based on deep learning. The construction model adopts a convolutional neural network squeze Net construction model and comprises an input layer, an implicit layer and an output layer, wherein the implicit layer comprises a convolutional layer, a pooling layer, a full-connection layer, a Softmax layer and a classification layer. The input layer and the first convolution layer of the network are modified by identifying the number of rows and columns of the sample image as the number of rows and columns of the input network, so that the input size and the number of channels of the input layer can be defined according to the input image characteristics, the number of channels of the first convolution layer is consistent with the number of image wave bands, the full-connection layer, the Softmax layer and the classification layer are modified, the output size of the full-connection layer is 1, the Softmax layer and the classification layer are replaced by a regression layer, and then the training sample input model is iterated repeatedly, so that the training model is obtained.
(6) Model training:
training the model by using a training data set to obtain model parameters, and establishing flood rainfall threshold simulation models in different reproduction periods.
(7) Model test and evaluation:
and inputting the test sample into the trained model, and calculating an evaluation index to evaluate the precision of the test model. 3 statistical indexes, namely Root Mean Square Error (RMSE), relative deviation BIAS and average absolute error (MAE), are selected for evaluating the model prediction result. The specific calculation formula is as follows:
ERMS represents the deviation between the predicted value and the actual value, the range of values is (0, + -infinity), and the closer to 0, the better the prediction effect. BIAS describes the magnitude of the deviation of the predicted value from the actual value, the range of values being-100%, the closer to 0 the smaller the degree of deviation from the actual value. MAE is the average value of absolute errors, range [0, + ], equal to 0 when the predicted value completely coincides with the true value, i.e., a perfect model; the larger the error, the larger the value. The smaller the MAE value, the better the accuracy of the predictive model.
(8) And (3) an optimal model:
and modifying model parameters according to model evaluation indexes, and obtaining an optimal deep learning model for simulating rainfall of flood critical surfaces of different reproduction periods of the river basin through repeated training and testing.
Claims (10)
1. A method for creating a threshold model of rainfall in a small river basin in a hydrologic data-free area is characterized by comprising the following steps:
step 1, establishing a drainage basin database: acquiring a rainfall grid data set of a watershed in a region without hydrologic data, grid data of sub-watersheds of different grades, flood rainfall threshold data of different reproduction periods and underlying surface data;
step 2, determining input and output of a model: the model is input into a sublevel surface characteristic of a river basin, the sublevel surface characteristic of the river basin comprises 6 factors of digital elevation model DEM data, gradient data, slope data, land utilization data, normalized vegetation indexes and temperature vegetation drought indexes as characteristic variables X of the river basin, and flood rainfall thresholds of the river basins of different grades in each recurrence period are output as target variables Y.
Step 3, creating a model sample data set: the input of the model is the underlying surface characteristic of the river basin, the output is the flood rainfall threshold values of each reproduction period of the sub-river basin with different grades, and the sample data set of the river basin is subjected to standardized processing and then is subjected to sample collection;
step 4, sample training set and test set division: dividing the collected sample data set into a training set and a testing set;
step 5, model network construction: constructing flood rainfall threshold prediction models of different reproduction periods of the drainage basin based on a convolutional neural network Squeeze Net;
step 6, model training: training a model by using a training data set to obtain model parameters, and establishing flood rainfall threshold prediction models of different reproduction periods;
step 7, model test and evaluation: inputting the test sample into the trained model, and calculating an evaluation index to evaluate the precision of the test model;
step 8, optimizing a model: and modifying model parameters according to model evaluation indexes, and obtaining an optimal deep learning model for simulating rainfall of flood critical surfaces of different reproduction periods of the river basin through repeated training and testing.
2. The method for creating the small-river basin flood rainfall threshold model in the hydrologic material-free region according to claim 1, which is characterized in that: in the step 1, the different-level sub-basin grid data sets are respectively obtained by using a hydrological analysis tool box of an ArcGIS, and inputting river network confluence accumulation quantity thresholds of three-level, four-level and five-level basins based on a basin digital elevation model.
3. The method for creating the small-river basin flood rainfall threshold model in the hydrologic material-free region according to claim 1, which is characterized in that: in the step 1, the Flood rainfall threshold data of the different reproduction periods of the river basin are based on a river basin digital elevation model and rainfall grid data by using a GIS Flood Tool, and the Flood rainfall threshold data sets of the different reproduction periods of the river basin corresponding to the three different levels of sub-river basins are estimated through hydrologic analysis and regional regression equations.
4. The method for creating the small-river basin flood rainfall threshold model in the hydrologic material-free region according to claim 1, which is characterized in that: in step 2, the river basin underlying surface data includes river basin digital elevation model DEM data, gradient data, slope data, land use type data, and land set8 OLI remote sensing data, wherein the land use type data is converted into land use degree, and a normalized vegetation index NDVT and a temperature vegetation drought index TVDI of the river basin are calculated based on the land set8 OLI data in ENVI5.3 to represent the condition of river basin vegetation coverage and soil humidity, respectively.
5. The method for creating the small-river basin flood rainfall threshold model in the hydrologic material-free region according to claim 1, which is characterized in that: in step 3, 6 different factors X of the basin are normalized to between 0 and 100. And then carrying out sample collection through a program based on 6 different factors X, different levels of sub-basin raster data B and different reappearance period flood rainfall threshold data Y after basin standardization processing.
6. The method for creating the small-river basin flood rainfall threshold model in the hydrologic material-free region according to claim 1, which is characterized in that: in step 3, three sample collection sequential modes are implemented: whole sampling, multiple reproduction period random sampling of the same drainage basin, multiple present period random sampling of different drainage basins.
7. The method for creating the small-river basin flood rainfall threshold model in the hydrologic material-free region according to claim 1, which is characterized in that: the sample data set in the step 3 comprises data sets obtained by three sample acquisition modes, and each sampling mode comprises four combination conditions of single-drainage-single-reproduction period, single-drainage-multiple-reproduction period, multiple-drainage-single-reproduction period and multiple-drainage-multiple-reproduction period.
8. The method for creating the small-river basin flood rainfall threshold model in the hydrologic material-free region according to claim 1, which is characterized in that: in step 5, the construction model adopts a convolutional neural network Squeeze Net construction model, and comprises an input layer, an implicit layer and an output layer, wherein the implicit layer comprises a convolutional layer, a pooling layer, a full-connection layer, a Softmax layer and a classification layer, the input layer and the first convolutional layer of the network are modified to identify the number of rows and columns of a sample image as the number of rows and columns of the input network, so that the input size and the number of channels of the input layer can be customized according to the characteristics of the input image, the number of channels of the first convolutional layer is enabled to be consistent with the number of image wave bands, the full-connection layer, the Softmax layer and the classification layer are modified, the output size of the full-connection layer is enabled to be 1, the Softmax layer and the classification layer are replaced by the regression layer, and then the training sample input model is iterated to obtain the training model.
9. The method for creating the small-river basin flood rainfall threshold model in the hydrologic material-free region according to claim 1, which is characterized in that: in step 6, the model parameters include: maximum generation number, learning rate, minimum batch number, and solver.
10. The method for creating the small-river basin flood rainfall threshold model in the hydrologic material-free region according to claim 1, which is characterized in that: in step 7, the model evaluation index includes: root mean square error, relative deviation, mean absolute error, correlation coefficient 4 index.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210177755.3A CN116796799A (en) | 2022-02-24 | 2022-02-24 | Method for creating small-river basin flood rainfall threshold model in area without hydrologic data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210177755.3A CN116796799A (en) | 2022-02-24 | 2022-02-24 | Method for creating small-river basin flood rainfall threshold model in area without hydrologic data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116796799A true CN116796799A (en) | 2023-09-22 |
Family
ID=88044223
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210177755.3A Pending CN116796799A (en) | 2022-02-24 | 2022-02-24 | Method for creating small-river basin flood rainfall threshold model in area without hydrologic data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116796799A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117152623A (en) * | 2023-11-01 | 2023-12-01 | 中铁水利信息科技有限公司 | Flood forecasting method, device and medium based on big data |
-
2022
- 2022-02-24 CN CN202210177755.3A patent/CN116796799A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117152623A (en) * | 2023-11-01 | 2023-12-01 | 中铁水利信息科技有限公司 | Flood forecasting method, device and medium based on big data |
| CN117152623B (en) * | 2023-11-01 | 2024-01-09 | 中铁水利信息科技有限公司 | Flood forecasting method, device and medium based on big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wu et al. | Depth prediction of urban flood under different rainfall return periods based on deep learning and data warehouse | |
| Wu et al. | Evaluating uncertainty estimates in distributed hydrological modeling for the Wenjing River watershed in China by GLUE, SUFI-2, and ParaSol methods | |
| Nourani et al. | A wavelet based data mining technique for suspended sediment load modeling | |
| Afshar et al. | Particle swarm optimization for automatic calibration of large scale water quality model (CE-QUAL-W2): Application to Karkheh Reservoir, Iran | |
| CN107463730B (en) | A kind of streamflow change attribution recognition methods considering Spatio-temporal Evolution of Land Use | |
| Greve et al. | Quantifying the ability of environmental parameters to predict soil texture fractions using regression-tree model with GIS and LIDAR data: The case study of Denmark | |
| Nasiri et al. | Simulation of water balance equation components using SWAT model in Samalqan Watershed (Iran) | |
| Dumedah et al. | Selecting model parameter sets from a trade-off surface generated from the non-dominated sorting genetic algorithm-II | |
| Kumar et al. | Regional flood frequency analysis using soft computing techniques | |
| Hollaway et al. | The challenges of modelling phosphorus in a headwater catchment: Applying a ‘limits of acceptability’uncertainty framework to a water quality model | |
| Han et al. | Bayesian uncertainty analysis in hydrological modeling associated with watershed subdivision level: a case study of SLURP model applied to the Xiangxi River watershed, China | |
| CN113176393B (en) | HASM model-based three-dimensional estimation method and system for soil organic carbon reserves | |
| CN108733952B (en) | Three-dimensional characterization method for spatial variability of soil water content based on sequential simulation | |
| Chen et al. | A spatiotemporal estimation method for hourly rainfall based on F-SVD in the recommender system | |
| Roushangar et al. | Exploring the multiscale changeability of precipitation using the entropy concept and self-organizing maps | |
| Zhou et al. | Reconstruction of missing spring discharge by using deep learning models with ensemble empirical mode decomposition of precipitation | |
| Choudhary et al. | Effect of root zone soil moisture on the SWAT model simulation of surface and subsurface hydrological fluxes | |
| Moghadam et al. | Investigating the performance of data mining, lumped, and distributed models in runoff projected under climate change | |
| CN116796799A (en) | Method for creating small-river basin flood rainfall threshold model in area without hydrologic data | |
| Gao et al. | A framework for automatic calibration of SWMM considering input uncertainty | |
| Gull et al. | Modelling streamflow and sediment yield from two small watersheds of Kashmir Himalayas, India | |
| Darvishi Salookolaei et al. | Application of grey system theory in rainfall estimation | |
| Ogale et al. | Modelling and short term forecasting of flash floods in an urban environment | |
| Ji et al. | Implementing generative adversarial network (GAN) as a data-driven multi-site stochastic weather generator for flood frequency estimation | |
| Zhang et al. | A bootstrap method to estimate the influence of rainfall spatial uncertainty in hydrological simulations |
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 |