CN115358134B - River basin medium-long term runoff prediction method based on space-time granulation data scene model - Google Patents
River basin medium-long term runoff prediction method based on space-time granulation data scene model Download PDFInfo
- Publication number
- CN115358134B CN115358134B CN202210151465.1A CN202210151465A CN115358134B CN 115358134 B CN115358134 B CN 115358134B CN 202210151465 A CN202210151465 A CN 202210151465A CN 115358134 B CN115358134 B CN 115358134B
- Authority
- CN
- China
- Prior art keywords
- hydrologic
- river basin
- runoff
- time
- space
- 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.)
- Active
Links
Images
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- 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
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Computer Hardware Design (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Tourism & Hospitality (AREA)
- Game Theory and Decision Science (AREA)
- Geometry (AREA)
- Development Economics (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- General Business, Economics & Management (AREA)
- Medical Informatics (AREA)
Abstract
The invention discloses a long-term runoff prediction method in a river basin based on a space-time granulated data scene model, which comprises the steps of firstly constructing a space-time granulated data scene model oriented to long-term runoff prediction in the river basin, realizing hydrological space-time object representation of different granularity levels, organizing and integrating hydrological object attributes, methods and relations in the river basin scene, and realizing long-term runoff prediction in the river basin based on the space-time granulated data scene model. The invention can realize the data organization and integration of multi-granularity hydrologic space-time objects, and is beneficial to improving the prediction effect and macroscopic judging capability of long-term runoffs in the river basin under the complex environment.
Description
Technical Field
The invention relates to a data organization and medium-long-term runoff prediction method in the technical field of hydrologic big data, in particular to a river basin medium-long-term runoff prediction method based on a space-time granulated data scene model.
Background
The method has the advantages that the high-speed development and wide application of information technologies such as the Internet of things, the GIS, big data and the like are realized, particularly, the formation of an air-to-ground integrated earth observation system is realized, the human observation capability to the air-to-ground is greatly improved, the variety and the magnitude of the wading information are expanded, the rich data of the types such as river basin weather, hydrology, geography and human activities are provided for hydrology research and application, and a data basis is provided for exploring a hydrologic forecasting method based on big data artificial intelligence. However, with the increase of wading data, the existing hydrologic data organization model lacks a single precision feature capable of accurately expressing hydrologic space-time data based on geometric feature elements of points, lines and planes, and can accurately express multi-granularity characterization of the hydrologic space-time data on space-time attribute, state and internal relation, so that the hydrologic space-time object characterization analysis and application requirements of different granularity levels cannot be met, and effective organization and sharing of information related to hydrologic forecast are difficult to implement.
In addition, due to the comprehensive influence of many factors such as climate, underlying surface and human activities, the long-term runoff change process in the river basin has certain space-time uncertainty, and the factors influencing runoff mainly comprise precipitation, air temperature, topography, geology, soil, vegetation, human activities and the like. The current long-term runoff prediction in the river basin is mainly based on the runoff change of some typical sites (single granularity level) in the river basin to predict the condition of the medium-long term runoff change, and the multi-granularity comprehensive analysis is not carried out by combining a plurality of hydrologic space-time objects in a research area from the aspect of the river basin (coarse granularity), so that the accuracy and macroscopic judging capability of the long-term runoff prediction in the river basin in a complex environment are insufficient.
Therefore, how to realize a space-time granulation data scene model and a long-term runoff prediction method in a river basin is the problem to be solved by the invention.
Disclosure of Invention
The invention aims to: aiming at the problems and the shortcomings in the prior art, the invention provides a river basin medium-long-term runoff prediction method based on a space-time granulation data scene model.
The technical scheme is as follows: a long-term runoff prediction method in a river basin based on a space-time granulated data scene model is characterized in that firstly, a space-time granulated data scene model oriented to long-term runoff prediction in the river basin is constructed, the characteristics of hydrologic space-time objects at different granularity (time and space) layers and the organization and integration of hydrologic object attributes, methods and relations in the river basin scene are realized, and then the long-term runoff prediction in the river basin is realized on the basis of the space-time granulated data scene model; the method comprises the following steps:
s1-1 study area determination.
S1-2, searching hydrologic objects related to long-term runoff prediction in the river basin in the determined research area according to the long-term runoff prediction requirement in the river basin.
S1-3 judges whether the relevant hydrologic objects selected in the step S1-2 are coded, if all the relevant hydrologic objects are coded, the process goes to the step S1-5, otherwise, the process goes to the step S1-4.
S1-4, encoding the hydrologic object, and encoding the selected hydrologic object.
S1-5, space-time granularity analysis, namely performing space-time granularity and life cycle analysis on the coded hydrologic object, and obtaining the minimum space-time granularity, particle number and life cycle of the hydrologic object to form an initial state of a space-time granulated data scene.
S1-6, generating a space-time granulating data scene, and organizing and integrating the selected attribute, state and relation of the hydrologic object related to the long-term runoff prediction in the river basin on the basis of the initial state of the space-time granulating data scene to form a space-time granulating data scene model oriented to the long-term runoff prediction in the river basin.
S1-7, constructing a change factor for representing the overall trend of long-term runoffs in the whole river basin at the level of the river basin (coarse granularity) based on multi-granularity space-time granulation data formed by the space-time granulation data scene model, and mainly influencing objects of the change factor, specifically: (1) Constructing a basin runoff overall trend change factor with space-time granulation characteristics and life cycle, which is also a prediction object of an artificial intelligent model so as to represent the trend change condition of long-term runoffs in the basin; (2) The method for constructing the main object influencing the integral trend change factor of the basin runoff is also an input object of an artificial intelligent prediction model, and comprises the steps of calculating the area weighting represented by meteorological stations in the basin by adopting a Thiessen polygon method to construct a precipitation object influencing the integral trend change factor of the basin runoff, screening the climate objects (the number is n atm ) Basin normalization based on SPOT imagesThe vegetation index calculating method constructs vegetation objects which influence the overall trend change factors of the runoffs in the river basin and cover the whole river basin. Then the basin runoff overall trend change factor, the rainfall object, the climate object and the vegetation object are summed up (n atm Observations of +3) subjects for the first 12 months as input candidate features for the intelligent predictive model, totals [ (n) atm +3)×12]The characteristic is selected, and then the characteristic screening method is adopted for the (n) atm +3) subjects, totaling [ (n) atm +3)×12]And screening the features to obtain a key feature set influencing the overall trend change factor of the long-term runoff in the flow domain, and providing input for the artificial intelligent prediction model. And finally outputting a long-term runoff overall trend change factor in the flow domain based on the artificial intelligent prediction model to realize long-term runoff prediction in the flow domain.
In the step S1-1 study area determination, a specific study area is determined through map operation or manually.
In S1-5, the space-time granularity analysis comprises the following steps:
s1-5-1 time granularity analysis is used for analyzing the time granularity of a research object in a space-time granulating data scene model, and selecting the time granularity which is suitable for the problem can improve the solving quality of the problem; the time granularity can be divided into minutes, times, days, months, years and the like, is related to the interval of sampling time, such as short-term runoff prediction, time granularity is time, days, medium-long-term runoff prediction, and time granularity can be ten days, months and years.
S1-5-2 space granularity analysis is used for analyzing the space granularity of a research object in the space-time granulating data scene model, and selecting the space granularity which is suitable for the problem can reduce the complexity of calculation and improve the solving efficiency of the problem; the spatial granularity can be divided into stations, sections, rivers, watercourses, countries and the like, is related to errors of used measuring equipment, such as runoff prediction of single hydrologic stations commonly used in hydrologic fields, is the station, belongs to fine granularity, and the runoff prediction of the whole watercourse, is the whole watercourse, and belongs to coarse granularity.
S1-5-3 life cycle analysis, describing time characteristics of state change of the hydrologic space-time object, including initial time and ending time of the state change, so as to represent the initial time of hydrologic time series data for runoff prediction analysis.
S1-6, organizing and integrating the attributes, methods and relations of the selected hydrologic objects related to the long-term runoff prediction in the river basin to form data description of the hydrologic objects, and constructing a multi-granularity hydrologic space-time object data scene mode oriented to the long-term runoff prediction in the river basin according to the data description.
The data description of the hydrologic object comprises three parts, namely an attribute, a method and a relation, wherein the attribute is used for describing the state, the composition and the characteristics of the hydrologic object, the method is used for describing the behavior characteristics of the hydrologic object, and the relation is used for describing the belonged and associated characteristics of the hydrologic object.
The attribute description of the hydrologic object contains two classes, object attributes and sub-attributes. Wherein: basic attribute description (object itself) of the hydrologic object for describing basic information of the hydrologic object; the state attribute description of the hydrologic object is used for describing the current state information of the hydrologic object; the characteristic attribute description of the hydrologic object is used for describing the business attribute information of the hydrologic object;
the method description of the hydrologic object is a package body composed of data and operation, has a direct corresponding relation with a specific hydrologic object, and aims to acquire attribute information of the hydrologic object, change the state of the hydrologic object and the mapping relation among different objects, extract characteristic attribute information and the like by calling the method.
The relationship description is used to describe the interrelationship of the hydrologic object with other objects as belonging, composition, collection, etc.
The space-time granulated data scene model defines:
defining 1 a hydrological object set in a data scene;
HS＝{O 1 ,O 2 ,…,O w } (1)
wherein: HS represents a hydrologic data scene consisting of the states of w hydrologic objects at a given moment within a given investigation region (range); o (O) 1 Representing the state of the 1 st hydrological object, O 2 Representing the state of the 2 nd hydrologic object, O w Representing the state of the w-th hydrologic object.
Definition 2 multi-granularity characterization of the c-th hydrologic object:
wherein:
O c representing the state of the c-th hydrologic object.
OID c Representing hydrologic object O c And is unique, by a hydrologic object encoding step.
S c (t) represents the hydrologic object O under a specific spatial coordinate system c A set of spatial data types that vary over time,
wherein: />
Representing hydrologic object O c Is a set of spatial data types of (a),
q k represent a sequence of discrete coordinate points, q k ＝{(x 1 ,y 1 ,z 1 ),(x 2 ,y 2 ,z 2 ),…,(x n ,y n ,z n )}。
A c (t) represents hydrologic object O c A set of attributes that vary over time,
wherein: />
Representing hydrologic object O c At t i The j attribute status at time j, j E [1, m]，i∈[1,n]M represents hydrologic object O c N represents the hydrologic object O c Is not shown in the figures).
T c (T s ,T e ) Representing hydrologic object O c Describing the time characteristic of the change in state. Wherein: t (T) s Representing hydrologic object O c Initial time of state change; t (T) e Representing hydrologic object O c End time of state change.
And S1-7, constructing a basin runoff overall trend change factor with multi-time space granularity characteristics and life cycle, specifically, performing time space granularity selection and life cycle selection on all hydrologic site objects in a research area, and constructing the basin runoff overall trend change factor so as to obtain the basin runoff overall trend change factor.
Wherein, the construction method of the basin runoff integral trend change factor is as follows,
in which W is i Weight of ith hydrologic site, Q i Control area percentage of ith hydrologic site, Q j The control area percentage of the jth hydrological site is represented by m, which is the number of hydrological sites with the uniform runoff consistency reaching a preset standard in the river basin, C j The integral trend change factor of the basin runoff of the jth month, C ij The month average diameter flow rate of the jth month of the ith hydrological site.
The Thiessen polygon method is adopted to calculate the area weighting represented by the meteorological site in the river basin, and a precipitation object is constructed, wherein, the Thiessen polygon method has the calculation formula,
in the method, in the process of the invention,
for average precipitation in basin, P i For the i-th observation station to synchronously reduce the water quantity, P 1 The synchronous precipitation amount P of the 1 st observation station 2 For the 2 nd observation station to synchronously reduce the water quantity, P n For the contemporaneous precipitation of the nth observation station, S i For the i-th observation station control area S 1 For the 1 st observation station control area S 2 Control area for the 2 nd observation station, S n The control area is the nth observation station, and S is the total area of the river basin;
if the continuous missing date of the contemporaneous precipitation daily value data is less than 10 days, the contemporaneous precipitation daily value data is replaced by a daily average value of years, and if the continuous missing contemporaneous precipitation daily value data is 10 days or more, the adjustment calculation is carried out by adopting a linear difference method based on the daily average value of years.
The method for screening the climatic objects with strong relevance to the runoff process of the river basin by adopting the relevance coefficient method comprises the following specific steps:
let the runoff object be Y and the climate object be variable X, the correlation coefficient between the runoff object Y and the climate object X is defined as,
wherein r is XY For the correlation coefficient between climate object X and runoff object Y, N is the number of samples of the runoff-climate object, X i For the ith sample value of the climate object X,
is the average value of the climate object X, Y i For the ith sample value of the runoff object Y,
is the average value of the runoff object Y;
r XY the value range is [ -1,1]，|r XY A large value indicates a high linear correlation between the runoff object Y and the climate object X, when |r XY The value of is close to 0, which means that the linear correlation between the runoff object Y and the climate object X is low, when r XY When the I value is 0, the runoff object Y is linearly independent of the climate object X;
obtaining |r with large value XY The climate object corresponding to the I is the climate object with strong correlation with the basin runoff process.
The basin normalization vegetation index calculation method based on the SPOT image constructs a vegetation object covering the whole basin, wherein the vegetation object N constructed based on the basin normalization vegetation index of the SPOT image DVI The calculation method of (a) is as follows,
N DVI ＝0.004×DN-0.1 (7)
wherein N is DVI For NDVI values, DN is a gray scale value between 0 and 250.
The characteristic screening method is adopted to realize the characteristic screening of key objects influencing the overall change trend of the runoff in the river basin, specifically, the characteristic screening is carried out on the overall trend change factor, the rainfall object, the climate object and the vegetation object of the runoff in the river basin by adopting a partial mutual information method, so that a key characteristic set influencing the long-term runoff process change in the river basin is obtained, and the early-stage characteristic screening is provided for the long-term runoff prediction in the river basin.
The artificial intelligent model is based to realize long-term runoff prediction in the river basin, specifically, an artificial intelligent model is constructed, and key feature sets are used as inputs of the artificial intelligent prediction model to predict the long-term runoff change trend in the river basin.
Compared with the prior art, the invention has the following remarkable advantages: the method can accurately express multi-granularity representation of the hydrologic object based on space-time characteristics, can realize organization and integration of the hydrologic object in a watershed scene based on space-time properties, states and internal relations, meets the requirements of hydrologic space-time object representation analysis and application of different granularity levels, completes effective organization and sharing of hydrologic big data, and realizes the goal of long-term runoff prediction in the watershed based on the analysis and the application.
Drawings
FIG. 1 is a flow chart of a method according to an embodiment of the present invention;
FIG. 2 is a diagram of the geographical condition of a river basin according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of the construction of a space-time granular data scene model of a river basin according to an embodiment of the present invention;
FIG. 4 is a diagram showing different spatial granularity in a space-time granulating data scene of a river basin according to an embodiment of the present invention;
FIG. 5 is a graph of different time granularity in a scenario of the time-space granular data of the Yangtze river basin according to an embodiment of the present invention;
FIG. 6 is a diagram illustrating attributes in a space-time granular data scenario for a Yangtze river basin according to an embodiment of the present invention;
fig. 7 is a diagram illustrating a process of organizing hydrographic data objects in a river basin according to an embodiment of the present invention.
Detailed Description
The present invention is further illustrated below in conjunction with specific embodiments, it being understood that these embodiments are meant to be illustrative of the invention only and not limiting the scope of the invention, and that modifications of the invention, which are equivalent to those skilled in the art to which the invention pertains, will fall within the scope of the invention as defined in the claims appended hereto.
A river basin medium-long-term runoff prediction method based on a space-time granulated data scene model comprises the following steps:
s1-1, determining a research area, and determining a specific research area through map operation or manually.
S1-2, searching hydrologic objects related to long-term runoff prediction in the river basin in the selected specific region according to the long-term runoff prediction requirement in the river basin.
S1-3 judges whether the relevant hydrologic objects selected in the step S1-2 are coded, if all the relevant hydrologic objects are coded, the process goes to the step S1-5, otherwise, the process goes to the step S1-4.
S1-4, encoding the hydrologic object, and encoding the selected hydrologic object.
S1-5, space-time granularity analysis, namely performing space-time granularity and life cycle analysis on the coded hydrologic object to obtain the minimum space-time granularity, particle number and life cycle of the coded hydrologic object.
S1-5-1 time granularity analysis, wherein the time granularity can be divided into minutes, carved, time, day, ten days, month, year and the like, and the time granularity is related to the interval of sampling time.
S1-5-2 space granularity analysis, wherein the space granularity can be divided into stations, sections, rivers, river basins, countries and the like, and is related to the error of the used measuring equipment.
S1-5-3 life cycle analysis, describing time characteristics of state change of the hydrologic space-time object, including initial time and end time of state change.
S1-6, organizing and integrating the attributes, methods and relations of the selected hydrologic objects related to the long-term runoff prediction in the river basin to form data description of the hydrologic objects, and constructing a multi-granularity hydrologic space-time object data scene mode oriented to the long-term runoff prediction in the river basin according to the data description.
The data description of the hydrologic object comprises three parts, namely an attribute, a method and a relation, wherein the attribute is used for describing the state, the composition and the characteristics of the hydrologic object, the method is used for describing the behavior characteristics of the hydrologic object, and the relation is used for describing the belonged and associated characteristics of the hydrologic object.
The attribute description of the hydrologic object contains two classes, object attributes and sub-attributes. Wherein: basic attribute description (object itself) of the hydrologic object for describing basic information of the hydrologic object; the state attribute description of the hydrologic object is used for describing the current state information of the hydrologic object; the characteristic attribute description of the hydrologic object is used for describing the business attribute information of the hydrologic object;
the method description of the hydrologic object is a package body composed of data and operation, has a direct corresponding relation with a specific hydrologic object, and aims to acquire attribute information of the hydrologic object, change the state of the hydrologic object and the mapping relation among different objects, extract characteristic attribute information and the like by calling the method.
The relationship description is used to describe the interrelationship of the hydrologic object with other objects as belonging, composition, collection, etc.
Space-time granulated data scene model definition:
defining 1 a hydrological object set in a data scene;
HS＝{O 1 ,O 2 ,…,O w } (1)
wherein: HS represents a hydrologic data scene consisting of the states of w hydrologic objects at a given moment within a given area (range); o (O) 1 Representing the state of the 1 st hydrological object, O 2 Representing the state of the 2 nd hydrologic object, O w Representing the state of the w-th hydrologic object.
Definition 2 multi-granularity characterization of the c-th hydrologic object:
wherein:
O c represents the cThe state of the individual hydrologic object.
OID c Representing hydrologic object O c And is unique.
S c (t) represents the hydrologic object O under a specific spatial coordinate system c A set of spatial data types that vary over time,
wherein: />
Representing hydrologic object O c Is a set of spatial data types of (a),
q k represent a sequence of discrete point coordinate points, q k ＝{(x 1 ,y 1 ,z 1 ),(x 2 ,y 2 ,z 2 ),…,(x n ,y n ,z n )}。
A c (t) represents hydrologic object O c A set of attributes that vary over time,
wherein: />
Representing hydrologic object O c At t i The j attribute status at time j, j E [1, m]，i∈[1,n]。
T c (T s ,T e ) Representing hydrologic object O c Describing the time characteristic of the change in state. Wherein: t (T) s Representing hydrologic object O c Initial time of state change; t (T) e Representing hydrologic object O c End time of state change.
S1-7, predicting medium-long-term runoff in a river basin, comprising the following steps of:
s1-7-1 construction of basin runoff integral trend change factor (runoff object) with multi-time space granularity characteristic and life cycle
S1-7-1-1 space-time granularity selection, wherein the selected time granularity is month, and the space granularity is the elegant river basin aiming at the actual requirement of long-term runoff prediction analysis in the elegant river basin;
s1-7-1-2 life cycle is selected, mainly the data starting time of runoff and influencing objects thereof, the life cycle of the runoff objects in the Yahulling river basin is 1960, 1 month and 2016, the life cycle of the rainfall objects is 1960, 1 month and 2012, 9 months and 30 days, the weather objects are mainly 95 atmospheric circulation indexes, the life cycle is 1951, 1 month and 2017, 2 months, and the life cycle of the vegetation objects is 1998, 4 months and 2008, 7 months.
S1-7-1-3, selecting two estuaries, a mall, a official land and two beach 4 hydrologic sites, and constructing a global trend change factor of runoff of the river basin of the elegance hulling river by formulas (3) and (4) (a reverse weighting method);
wherein: w (W) i Weight of ith hydrologic site, Q i The control area percentage of the ith hydrological site is represented by m, which is the number of hydrological sites with good uniform runoff consistency in the river basin, C j The integral runoff trend change factor of the jth month, C ij Month average diameter flow for the ith hydrographic site, the jth month.
S1-7-2, calculating area weighting represented by meteorological stations in the elegance river basin by adopting a Thiessen polygon method to construct a precipitation object:
s1-7-2-1, 22 meteorological sites of a Yangtze river basin are selected, wherein the meteorological sites comprise Muli, zhaozhaofei, dafu, deger, shikon, hui, huaping, lijiang, xichang, salt source, vietnam, jiulong, rice city, kangding, pond, xinlong, ganzi, bapool, dari, clear water river and Jade tree;
s1-7-2-2, calculating the area represented by each station of 22 meteorological stations through Thiessen polygons;
s1-7-2-3, constructing a precipitation object through a formula (5);
in the method, in the process of the invention,
average precipitation (mm) for basin, P i For the i-th observation station to synchronously reduce the water quantity, P 1 The synchronous precipitation amount P of the 1 st observation station 2 For the 2 nd observation station to synchronously reduce the water quantity, P n For the contemporaneous precipitation of the nth observation station, S i Control area (km) for the ith station 2 )，S 1 For the 1 st measuring station control area S 2 Control area for 2 nd measuring station, S n For nth station controlArea, S is the total area of the river basin (km) 2 )。
If rainfall daily value data is not measured, the corresponding processing modes mainly comprise: (1) If the continuous missing date is less than 10 days, using a perennial average value for replacement; (2) If the measurement data are missing for 10 or more continuous days, the adjustment calculation is performed by adopting a linear difference method based on the average value of years.
S1-7-3, selecting 21 remote related climate objects from 95 atmospheric flow indexes influencing the long-term runoff process change in the elegance river basin through a formula (6), taking the climate objects as alternative input variables of a long-term runoff prediction model in the river basin, wherein the 95 atmospheric flow indexes are shown in a table 1, and the screening result is shown in a table 2.
Wherein r is XY For the correlation coefficient between the climate object X and the forecast object Y, N is the number of samples, X i For the ith sample value of the climate object X,
is the average value of the climate object X, Y i For the i-th sample value of the forecast object Y, -, is given>
Is the mean value of the forecast object Y.
r XY The value range is [ -1,1]，|r XY The larger the i value, the higher the linear correlation between the runoff object Y and the climate object X; when |r XY The closer the i value is to 0, the lower the linear correlation between the runoff object Y and the climate object X; when |r XY When the value is 0, it means that the runoff object Y is linearly independent of the climate object X.
Table 195 climate objects and classifications
TABLE 2 results of screening for remote-related climatic objects in the Yangher river basin
In the step S1-7-4, a vegetation object covering the whole river basin is constructed through a formula (7), the specific situation of the elegance river basin NDVI is shown in figure 5,
N DVI ＝0.004×DN-0.1 (7)
wherein N is DVI For NDVI values, DN is a gray scale value between 0 and 250.
In the step S1-7-5, key characteristics of main objects related to long-term runoff change in the river basin are screened by adopting a partial mutual information method, and early characteristic screening is provided for long-term runoff prediction in the river basin.
S1-7-6, constructing an artificial intelligent model, taking a key feature set as input of the artificial intelligent model, predicting the long-term runoff change trend in the river basin, solving the highly complex nonlinear problem of long-term runoff prediction in the river basin, and meeting the requirements of the long-term runoff change trend prediction and application under different conditions.
S1-8 ends.
Fig. 2 is a diagram showing geographical conditions of the river basin according to the embodiment of the present invention.
As shown in fig. 3, a structure diagram constructed by a space-time granulating data scene model of a river basin according to an embodiment of the invention is shown;
(1) Space-time granulating data scene model for elegance river basin
The hydrologic objects in the natural scene of the elegance river basin comprise the elegance river basin, hydrologic stations, meteorological stations, runoff, meteorology, vegetation, climate, elevation, water quality, underground water, soil, ice edges and the like.
(2) Hydrologic object lookup
According to the actual demand of long-term runoff prediction in the river basin, selecting an influence object related to long-term runoff prediction research in the river basin from the hydrologic objects, wherein the influence object comprises the following components: runoff objects (constructed from two estuaries, mall, official, two beach 4 hydrologic stations), precipitation objects (constructed from 22 meteorological stations in the woody, zhaojun, daffodil, dagger, shikone, congress, huaping, lijiang, wenchang, salt source, vietnam, jiulong, rice city, kanding, pond, new dragon, color da, ganmai, pond, dar day, clear water river, jade tree, etc.), vegetation objects, climate and yahula river basin.
(3) Spatiotemporal granular data scene generation
The space-time granularity data scene generation comprises hydrologic object coding, space-time granularity analysis and data scene generation. The method comprises the steps of selecting a hydrologic object according to a network of things (EPC) coding system, wherein the hydrologic object coding is to carry out WID coding on the selected hydrologic object based on the EPC coding system, and the WID code is unique; the space-time granularity analysis is to analyze the space-time granularity and the life cycle of the coded hydrologic object to obtain the minimum space-time granularity, the particle number and the life cycle; the data scene generation is to assign the time granularity, the space granularity and the life cycle of the hydrologic object according to the actual application requirement, and to organize and integrate the attribute, the state and the relation of the hydrologic object to form a multi-granularity hydrologic space-time object data scene mode, namely a space-time granulated data scene model.
(4) Hydrologic object coding
The hydrologic object coding module codes related hydrologic objects such as runoff, precipitation, vegetation, climate and the like and the elegance river basin, and marks the hydrologic objects as a Wid code which is unique.
As shown in fig. 4, the present invention is directed to different spatial granularity in the space-time granulation data scene of the jacent river basin.
The space granularity is selected from three kinds of granularity, namely fine granularity, medium granularity and coarse granularity. Wherein, the fine granularity corresponds to single sites (points) such as hydrological sites, weather table sites and the like; the middle granularity corresponds to the influence objects (lines) of linear structures such as runoff objects (constructed by two estuaries, a mall, a official, two beaches and 4 hydrologic stations), precipitation objects (constructed by 22 meteorological stations) and the like; coarse granularity corresponds to the influence objects (surfaces) of river basin layers such as elegance river basin, vegetation, climate and the like. The embodiment of the invention selects the elegance river basin (coarse granularity) as the space granularity.
Fig. 5 shows different time granularity in the space-time granulation data scene of the elegance river basin according to the embodiment of the invention.
The time granularity is selected from different time scales such as time scale, day scale, month scale and year scale. The embodiment of the invention selects a month scale as the time granularity.
FIG. 6 is a diagram illustrating attributes in a space-time granular data scenario for a Yangtze river basin according to an embodiment of the present invention;
the attribute of the hydrologic object is used for describing basic information of the hydrologic object, including name, type, longitude and latitude, management department, granularity size, wid code and the like. Because of more hydrological sites, the attribute description in the space-time granulating data scene of the elegant river basin in this embodiment is to select one hydrological site and one weather station site as representatives in each of the elegant river basin, and the whole elegant river basin.
FIG. 7 is a diagram showing the process of organizing hydrographic data objects in the Yangtze river basin according to the embodiment of the invention;
the elegant river basin attribute description comprises basic attributes (WID, names, longitude and latitude and the like), state attributes (current flow, water quality, flood peak situation and the like) and characteristic attributes (management department and the like), the method description comprises the steps of inquiring hydrological meteorological data of the river basin, acquiring state data of the river basin at different moments in real time, hydrological situation evolution, changing relations (such as father, son, collection) between the river basin and other objects and the like, and the relation description comprises relations between the river basin and hydrological sites, relations between the river basin and meteorological stations, relations between the river basin and atmospheric ring flow indexes and the like; the hydrologic site attribute description comprises basic attributes (WID, names, longitude and latitude and the like), state attributes (current flow, water storage level, storage capacity and the like) and characteristic attributes (management department and the like), the method description comprises inquiring hydrologic information, hydrologic information display and the like, and the relationship description comprises a basin, a next-stage hydrologic site and the like; the weather station attribute description comprises basic attributes (WID, names, longitude and latitude, etc.), state attributes (current air pressure, air temperature, wind speed, etc.) and characteristic attributes (management departments, etc.), the method description comprises the steps of inquiring weather information, displaying weather information, etc., and the relationship description comprises the affiliated river basin, the next weather station, etc.;
after the organization and integration of the hydrological object attribute, method and relation under the data scene are completed, the data exchange and sharing of hydrological data and business departments and other departments are realized through the data exchange and sharing function based on the EPC.
Claims (9)
1. A method for predicting long-term runoffs in a river basin based on a space-time granulated data scene model is characterized in that a space-time granulated data scene model facing the long-term runoffs in the river basin is firstly constructed, the characteristics of hydrologic space-time objects at different granularity levels and the organization and integration of hydrologic object attributes, methods and relations in the river basin scene are realized, and then the long-term runoffs in the river basin are predicted on the basis of the space-time granulated data scene model; the method comprises the following steps:
s1-1 study area determination;
s1-2, searching hydrologic objects related to long-term runoff prediction in the river basin in the determined research area according to long-term runoff prediction requirements in the river basin;
s1-3, judging whether the relevant hydrologic objects selected in the step S1-2 are coded, if all the relevant hydrologic objects are coded, turning to the step S1-5, otherwise turning to the step S1-4;
s1-4, encoding the hydrologic object, and encoding the selected hydrologic object;
s1-5, space-time granularity analysis, namely performing space-time granularity and life cycle analysis on the coded hydrologic object, and obtaining the minimum space-time granularity, particle number and life cycle of the hydrologic object to form an initial state of a space-time granulated data scene;
s1-6, generating a space-time granulating data scene, and organizing and integrating the selected attribute, state and relation of the hydrologic object related to the long-term runoff prediction in the river basin on the basis of the initial state of the space-time granulating data scene to form a space-time granulating data scene model oriented to the long-term runoff prediction in the river basin;
s1-7, constructing a change factor for representing the overall trend of long-term runoffs in the whole river basin and a main influence object thereof on the basis of multi-granularity space-time granulation data formed by the space-time granulation data scene model, wherein the change factor comprises the following specific components: firstly, constructing a basin runoff integral trend change factor with space-time granulation characteristics and life cycle as a prediction object of an artificial intelligent model so as to represent the trend change condition of long-term runoffs in the basin; secondly, constructing a method for influencing a main object of long-term runoff prediction in a river basin, and taking the method as an input object of an artificial intelligent prediction model, wherein the method comprises the steps of calculating area weights represented by meteorological stations in the river basin by adopting a Thiessen polygon method, constructing a precipitation object, screening weather objects with strong correlation with the runoff process of the river basin by adopting a correlation coefficient method, and constructing a vegetation object covering the whole river basin by adopting a river basin normalization vegetation index calculation method based on SPOT images; then, a feature screening method is adopted to realize the feature screening of key objects influencing the integral change trend of the runoff in the river basin; and finally, realizing long-term runoff prediction in the flow domain based on the intelligent prediction model.
2. The method for long-term runoff prediction in a river basin based on a scene model of spatio-temporal granular data according to claim 1, wherein in S1-5, the spatio-temporal granular analysis comprises:
s1-5-1 time granularity analysis, which is used for analyzing the time granularity of a study object in a space-time granulating data scene model, wherein the time granularity is related to the interval of sampling time, short-term runoff prediction, the time granularity is time and day, medium-long-term runoff prediction, and the time granularity is ten days, month and year;
s1-5-2 spatial granularity analysis is used for analyzing the spatial granularity of a research object in the space-time granulating data scene model; the space granularity can be divided into stations, sections, rivers, river basins and countries, is related to errors of the used measuring equipment, belongs to fine granularity, and belongs to runoff prediction of the whole river basin, and the space granularity is the whole river basin and belongs to coarse granularity;
s1-5-3 life cycle analysis, describing time characteristics of state change of the hydrologic space-time object, including initial time and ending time of the state change, so as to represent the initial time of hydrologic time series data for runoff prediction analysis.
3. The method for long-term runoff prediction in a river basin based on a space-time granular data scene model according to claim 1, wherein the data description of the hydrologic object comprises three parts, namely an attribute, a method and a relation, wherein the attribute is used for describing the state, the composition and the characteristics of the hydrologic object, the method is used for describing the behavior characteristics of the hydrologic object, and the relation is used for describing the belonged and associated characteristics of the hydrologic object; the attribute description of the hydrologic object comprises two types of object attributes and sub-attributes; basic attribute description of hydrologic object, is used for describing the basic information of hydrologic object; the state attribute description of the hydrologic object is used for describing the current state information of the hydrologic object; the characteristic attribute description of the hydrologic object is used for describing the business attribute information of the hydrologic object; the method description of the hydrologic object is a package body composed of data and operation, has a direct corresponding relation with a specific hydrologic object, and aims to acquire attribute information of the hydrologic object, change the state of the hydrologic object and the mapping relation among different objects and extract characteristic attribute information by calling the method; the relationship description is used to describe the interrelationship of the belongings, compositions, collections between the hydrologic object and other objects.
4. The method for predicting the long-term runoff in the river basin based on the space-time granulation data scene model according to claim 1, wherein the steps S1-7 are characterized in that the step S1-7 is used for constructing a river basin runoff overall trend change factor with multi-space granularity characteristics and life cycle, specifically, carrying out space-time granularity selection and life cycle selection on all hydrologic site objects in a research area, and constructing the river basin runoff overall trend change factor so as to obtain the river basin runoff overall trend change factor;
wherein, the construction method of the basin runoff integral trend change factor is as follows,
in which W is i Weight of ith hydrologic site, Q i Control area hundred for the ith hydrologic sitePercent, Q j The control area percentage of the jth hydrological site is represented by m, which is the number of hydrological sites with the uniform runoff consistency reaching a preset standard in the river basin, C j The integral trend change factor of the basin runoff of the jth month, C ij The month average diameter flow rate of the jth month of the ith hydrological site.
5. The method for predicting long-term runoff in a river basin based on a space-time granulated data scene model as recited in claim 1, wherein the Thiessen polygon method is used for calculating the area weighting represented by meteorological sites in the river basin to construct a precipitation object, and the calculation formula of the Thiessen polygon method is as follows,
in the method, in the process of the invention,
for average precipitation in basin, P i For the i-th observation station to synchronously reduce the water quantity, P 1 The synchronous precipitation amount P of the 1 st observation station 2 For the 2 nd observation station to synchronously reduce the water quantity, P n For the contemporaneous precipitation of the nth observation station, S i For the i-th observation station control area S 1 For the 1 st observation station control area S 2 Control area for the 2 nd observation station, S n The control area is the nth observation station, and S is the total area of the river basin;
if the continuous missing date of the contemporaneous precipitation daily value data is less than 10 days, the contemporaneous precipitation daily value data is replaced by a daily average value of years, and if the continuous missing contemporaneous precipitation daily value data is 10 days or more, the adjustment calculation is carried out by adopting a linear difference method based on the daily average value of years.
6. The method for predicting long-term runoff in a river basin based on a space-time granulated data scene model according to claim 1, wherein the method for screening climatic objects with strong correlation with the runoff process in the river basin by adopting a correlation coefficient method is specifically as follows:
let the runoff object be Y and the climate object be variable X, the correlation coefficient between the runoff object Y and the climate object X is defined as,
wherein r is XY For the correlation coefficient between climate object X and runoff object Y, N is the number of samples of the runoff-climate object, X i For the ith sample value of the climate object X,
is the average value of the climate object X, Y i For the i-th sample value of the runoff object Y, < >>
Is the average value of the runoff object Y; />
r XY The value range is [ -1,1]，|r XY A large value indicates a high linear correlation between the runoff object Y and the climate object X, when |r XY The value of is close to 0, which means that the linear correlation between the runoff object Y and the climate object X is low, when r XY When the I value is 0, the runoff object Y is linearly independent of the climate object X;
obtaining |r with large value XY The climate object corresponding to the I is the climate object with strong correlation with the basin runoff process.
7. The method for predicting long-term runoff in a river basin based on a space-time granulated data scene model as recited in claim 1, wherein the method for calculating a river basin normalized vegetation index based on a SPOT image constructs a vegetation object covering the whole river basin, wherein the vegetation object N constructed based on the river basin normalized vegetation index of the SPOT image DVI The calculation method of (a) is as follows,
N DVI ＝0.004×DN-0.1 (7)
wherein N is DVI For NDVI values, DN is a gray scale value between 0 and 250.
8. The method for predicting long-term runoff in a river basin based on a space-time granulation data scene model according to claim 1, wherein the feature screening method is used for achieving feature screening of key objects influencing the overall change trend of the runoff in the river basin, specifically, feature screening is carried out on the overall trend change factors, precipitation objects, climate objects and vegetation objects of the runoff in the river basin by adopting a partial mutual information method, a key feature set influencing the process change of the long-term runoff in the river basin is obtained, and early feature screening is provided for predicting the long-term runoff in the river basin.
9. The method for predicting long-term runoff in a river basin based on the space-time granulated data scene model according to claim 1, wherein the method is characterized in that the method for predicting long-term runoff in the river basin based on the artificial intelligence model is implemented, specifically, an artificial intelligence model is built, and key feature sets are used as inputs of the artificial intelligence prediction model to predict the long-term runoff change trend in the river basin.
Applications Claiming Priority (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110284456.5A CN113065280A (en) | 2021-03-17 | 2021-03-17 | Method for predicting integral change trend of runoff in drainage basin |
| CN2021102844156 | 2021-03-17 | ||
| CN2021102844565 | 2021-03-17 | ||
| CN202110284415.6A CN113065233A (en) | 2021-03-17 | 2021-03-17 | Method for constructing multi-granularity hydrological space-time object data scene model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115358134A CN115358134A (en) | 2022-11-18 |
| CN115358134B true CN115358134B (en) | 2023-04-21 |
Family
ID=84030427
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210151465.1A Active CN115358134B (en) | 2021-03-17 | 2022-02-18 | River basin medium-long term runoff prediction method based on space-time granulation data scene model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115358134B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113379110A (en) * | 2017-11-21 | 2021-09-10 | 中国水利水电科学研究院 | Medium-and-long-term runoff forecast result trend testing method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10267951B2 (en) * | 2016-05-12 | 2019-04-23 | The Climate Corporation | Statistical blending of weather data sets |
| CN107895221B (en) * | 2017-10-25 | 2021-09-28 | 北京微肯佛莱科技有限公司 | Medium-and-long-term scheduling and maintenance plan optimization method for cascade hydropower station in market environment |
-
2022
- 2022-02-18 CN CN202210151465.1A patent/CN115358134B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113379110A (en) * | 2017-11-21 | 2021-09-10 | 中国水利水电科学研究院 | Medium-and-long-term runoff forecast result trend testing method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115358134A (en) | 2022-11-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wielemaker et al. | Significance and application of the multi-hierarchical landsystem in soil mapping | |
| Chen et al. | Integrating weather observations and local-climate-zone-based landscape patterns for regional hourly air temperature mapping using machine learning | |
| Sayre et al. | Terrestrial ecosystems of south America | |
| CN108764527A (en) | A kind of Soil organic carbon pool space-time dynamic prediction suitable environment Variable Selection method | |
| Silva et al. | Analysis of the response of the Epitácio Pessoa reservoir (Brazilian semiarid region) to potential future drought, water transfer and LULC scenarios | |
| CN115358134B (en) | River basin medium-long term runoff prediction method based on space-time granulation data scene model | |
| Wang et al. | An aquatic ecoregion delineation approach based on GIS and spatial environmental data in Heihe River Basin, Northwestern China | |
| Cartwright et al. | Machine learning based imputation of mountain snowpack depth within an operational Lidar sampling framework in Southwest Alberta | |
| Preeti et al. | Application of GIS in Rainwater Harvesting Research: A Scoping Review | |
| CN113537567A (en) | Ecological tourism resource development value evaluation method | |
| Teerarojanarat | Remotely-Sensed Derived Built-up Area as an Alternative Indicator in the Study of Thailand's Regional Development. | |
| NABUNYA | Simulation of impacts of land-use and land cover changes on water balance in Oued Fez basin (Morocco) | |
| Saddiqi | Modelling Kücük Menderes watershed using soil and water assessment tool (SWAT) | |
| Fang | A GA-Based BP Artificial Neural Network for Estimating Monthly Surface Air Temperature of the Antarctic during 1960–2019 | |
| Lan et al. | Snow cover monitoring by remote sensing and snowmelt runoff calculation in the upper Huanghe River basin | |
| ZHAO et al. | Risk Mapping of Rainstorm-Flood Disaster Based on FAHP-PCA Method: A Case Study of Sanming City, China | |
| Maymand et al. | The application of landslide phase level in increasing the purity of soil map units in geopedology. | |
| Suresh | Utilization of remote sensing and gis for mapping of natural resource-on water resource action plan and prioritization for yagachi watershed hassan district | |
| Supuwiningsih et al. | Forecasting Agricultural Land in West Denpasar Using the Semi Averages Methods Applied in GIS | |
| Hernández et al. | Evaluation of total runoff for the Rio San Pedro sub-basin (Nayarit, Mexico) assessing their hydrologic response units | |
| Hanh et al. | Application of MCE-AHP technical for modelling paddy zoning: A case study in Vietnam | |
| رئيس التحرير | Application of GIS Raster Calculator for determining Rain water harvesting Suitable sites In Wadi Soba-Khartoum-Sudan | |
| Baki et al. | A modified DRASTIC model for groundwater vulnerability assessment using connecting path and analytic hierarchy process methods | |
| MUNDHE et al. | Assessment of Domestic Water Resources for Sustainable Utilization Using Geospatial Techniques. The Case of Pune City, India. | |
| CN113065233A (en) | Method for constructing multi-granularity hydrological space-time object data scene model |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |