CN107316095B - Regional weather drought level prediction method coupled with multi-source data - Google Patents
Regional weather drought level prediction method coupled with multi-source data Download PDFInfo
- Publication number
- CN107316095B CN107316095B CN201610850079.6A CN201610850079A CN107316095B CN 107316095 B CN107316095 B CN 107316095B CN 201610850079 A CN201610850079 A CN 201610850079A CN 107316095 B CN107316095 B CN 107316095B
- Authority
- CN
- China
- Prior art keywords
- drought
- precipitation
- data
- regional
- resolution
- 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
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000001556 precipitation Methods 0.000 claims abstract description 56
- 230000007704 transition Effects 0.000 claims abstract description 21
- 238000012544 monitoring process Methods 0.000 claims description 22
- 238000012417 linear regression Methods 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 5
- 230000008569 process Effects 0.000 claims description 5
- 238000012937 correction Methods 0.000 claims description 4
- 230000008878 coupling Effects 0.000 claims description 4
- 238000010168 coupling process Methods 0.000 claims description 4
- 238000005859 coupling reaction Methods 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims description 4
- 238000006467 substitution reaction Methods 0.000 claims description 4
- 230000015556 catabolic process Effects 0.000 claims description 3
- 238000006731 degradation reaction Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims description 3
- 238000011161 development Methods 0.000 abstract description 6
- 230000015572 biosynthetic process Effects 0.000 abstract description 3
- 230000006355 external stress Effects 0.000 abstract description 3
- 230000007246 mechanism Effects 0.000 abstract description 3
- 239000002243 precursor Substances 0.000 abstract description 2
- 239000011159 matrix material Substances 0.000 description 10
- 238000012546 transfer Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 230000007547 defect Effects 0.000 description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 230000001186 cumulative effect Effects 0.000 description 3
- 238000007418 data mining Methods 0.000 description 3
- 241000039077 Copula Species 0.000 description 2
- 238000007476 Maximum Likelihood Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000010219 correlation analysis Methods 0.000 description 2
- 238000005315 distribution function Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000004907 flux Effects 0.000 description 2
- 230000010355 oscillation Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 239000002352 surface water Substances 0.000 description 2
- 238000012952 Resampling Methods 0.000 description 1
- 238000012300 Sequence Analysis Methods 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000000246 remedial effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000001932 seasonal effect Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 230000035882 stress Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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
- 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—Information and communication technology [ICT] specially adapted for implementation of business processes of 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
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Theoretical Computer Science (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a regional weather drought level prediction method coupled with multi-source data, which comprises the steps of carrying out spatial downscaling on gridded satellite remote sensing precipitation data, constructing a high-resolution regional precipitation spatial database, dividing a drought level by adopting a standardized drought index, introducing a large-scale weather factor reflecting the atmospheric circulation characteristic as a covariate of a drought state transition probability, and constructing a weather drought level prediction model with a time-varying transition probability based on a non-stationary Markov chain model. The invention captures the spatial variability of regional rainfall by using multi-source remote sensing information and basic underlying surface characteristics, makes up for the insufficient rainfall observed by the traditional station, fully utilizes the external stress and precursor signals of drought and waterlogging evolution, namely large-scale meteorological factors capable of reflecting atmospheric circulation characteristics, considers the formation and development mechanism of drought and waterlogging to a certain extent, better fits the dynamic evolution characteristics of a meteorological hydrological system, has stronger scientificity and practicability, and can lay a foundation for building a regional drought and waterlogging early warning and forecasting system.
Description
Technical Field
The invention belongs to the technical field of disaster forecast early warning, and particularly relates to a regional weather drought level prediction method coupled with multi-source data.
Background
Drought is a natural phenomenon of continuous shortage of water, and has the characteristics of high occurrence frequency, long duration, wide spread range and the like. China is in an Asian monsoon climate region, precipitation space and annual distribution are seriously uneven, annual variation of monsoon path and intensity is large, and water heating unevenness caused by factors such as terrain in China and the like cause frequent drought disasters in China, so that the method is one of the most serious countries suffering from drought disasters in the world. The american meteorological institute, based on a summary of various drought definitions, has classified drought into four types, namely, meteorological drought, hydrographic drought, agricultural drought, and socioeconomic drought. Weather drought refers to reduced or no precipitation, and the formation of other types of drought is directly linked with weather drought.
The occurrence and development of drought is significantly more insidious and "creeping" than the "impulsive" consequences of other extreme climatic events such as flooding, typhoon, etc. The initiation and termination times of a drought event are often poorly defined, and once its extent of impact and extent of harm are manifested, countermeasures and remedial measures are often severely lagged behind. Therefore, the drought prediction is timely and accurate, and has great significance for guiding the development of drought-resistant work, strengthening disaster risk emergency management, improving disaster response level and the like.
Drought monitoring is the basis of drought prediction, most of the existing methods are based on actually measured rainfall site data, but the spatial heterogeneity of rainfall distribution is difficult to reflect due to the problem of uneven density and distribution of observation sites; in some remote areas where data is scarce or unavailable, it is more difficult to obtain observation data. In recent years, development of remote sensing technology provides a brand new approach for wide-range drought monitoring. High strength of the extreme[1]The invention discloses a drought monitoring method and a system based on surface water heat flux remote sensing inversion, which are used for estimating regional surface energy and evapotranspiration distribution under different climatic and topographic conditions and providing technical support for regional agricultural drought monitoring. Plum is good and the like[2]The invention discloses a drought monitoring method based on HJ-1A/1B CCD data, which combines MPDI data obtained by HJ-1A/1B CCD remote sensing data and crop growth period to determine the agricultural drought condition; feng Jie et al[3]The drought monitoring method based on data mining is provided, multi-source remote sensing spatial information in drought monitoring is comprehensively considered to perform spatial downscaling on remote sensing rainfall, and a drought monitoring model is constructed by adopting a spatial data mining technology. However, due to the limitation of the current technical level, the drought remote sensing monitoring has the defects of low spatial resolution and the like.
Regional drought and flood phenomena are usually caused by temporary abnormalities of local atmospheric water balance, but due to numerous and interactive factorsThe method is complex, and accurate assessment and prediction of the disaster are difficult to realize at present due to lack of comprehensive knowledge of a catastrophe mechanism. Instead, quantitative estimates of the relative frequency and intensity (grade) of occurrence of drought and flood events can be made. The prediction and forecast of the drought and flood level of the existing region mainly starts from the randomness of drought and flood events, and a time sequence analysis tool is adopted to achieve the purpose of revealing the time-space evolution characteristics of the drought and flood events. Yangxiaoyong, etc[4]The method adopts a two-dimensional Copula function to construct' 28390, and calculates the occurrence probability of drought-waterlogging alternation and drought-waterlogging continuous combined events of all sites by the combined distribution of seasonal precipitation range average percentage sequences of representative meteorological sites of river basin. Song Xin shan et al[5]And (3) calculating the transfer probability, duration, reproduction time and other statistical characteristics of different drought and waterlogging states of the 16 representative stations in the middle and downstream areas of Huang-Huai-Hai in 540 years by using a Markov model. Feng Ping, etc[6]A28390short-term meteorological drought level prediction model of 21 rainfall stations in a river Pan family mouth reservoir control basin is established by adopting a three-dimensional logarithmic linear model, and meteorological drought level prediction with a prediction period of 1 month and 2 months is realized.
According to the method, drought and waterlogging evolution is mostly regarded as a stable process, namely, the statistical characteristics of the method, such as the transition probability of drought and waterlogging states and the like, do not change along with time, and the statistical characteristics can be obtained through statistics of weather or hydrological sequence samples measured in the past. However, the meteorological hydrological system has significant dynamic evolution characteristics subject to interference from climate changes and human activities. In order to effectively cope with the catastrophe risk of regional drought and flood events under dynamic evolution conditions, the development of a regional drought and flood prediction method capable of comprehensively considering internal causes and external stresses is urgently needed.
The references referred to herein are as follows:
[1] high aspiration intensity, drought monitoring method and system based on surface water heat flux remote sensing inversion, and patent number ZL201010623662.6.
[2] Plum, old sea wave, surplus length flood, and the like, a drought monitoring method based on HJ-1A/1B CCD data, patent No. ZL201310379034.1.
[3] Von jew, he shou, yang shiyong, etc. a drought monitoring method based on data mining, publication No. CN105760814A.
[4] Yang Zhi Yong, Yuan Ji, Fang hong Yang, etc., 28390based on Copula function, probability feature analysis of drought and waterlogging combined events in river basin [ J ] in the academic newspaper, 2013,44(5):556 and 569.
[5] Song Xinshan, Yangtonghua, Wang Yuhui, etc. analysis of drought and waterlogging evolution characteristics of east region in Huang-Huai-Hai for 540 years [ J ]. Hydraulic science, 2013,44(12): 1425-1432) based on Markov model.
[6] Von Ping, Hurong, Lijian column weather drought grade prediction research based on three-dimensional logarithmic linear model [ J ]. Water conservancy project, 2014,45(5): 505-.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a regional weather drought level prediction method coupling multi-source data, which comprises the steps of carrying out spatial scale reduction on gridded satellite remote sensing precipitation data, constructing a high-resolution regional precipitation spatial database, dividing the drought level by adopting a standardized drought index, introducing a large-scale weather factor reflecting the atmospheric circulation characteristic as a covariate of the drought state transition probability, and constructing a weather drought level prediction model with the time-varying transition probability based on a non-stable Markov chain model so as to analyze and predict the dynamic evolution rule of the regional drought state.
In order to solve the technical problems, the invention adopts the following technical scheme:
a regional weather drought level prediction method coupled with multi-source data comprises the following steps:
step 1, collecting information of a foundation underlying surface in a region and synchronous long series data of precipitation and large-scale meteorological factors;
step 2, constructing a remote sensing precipitation data space downscaling model corrected by considering factors of a foundation underlying surface, processing low-resolution original remote sensing monitoring precipitation data into higher-resolution precipitation data, and correcting output data by using actually measured precipitation data of a ground observation station to obtain a high-resolution regional precipitation space database;
step 3, sequentially carrying out frequency analysis on the long-series rainfall data of each high-resolution grid in the regional gridding rainfall space database obtained in the step 2, calculating a standardized drought index of the long-series rainfall data, and dividing a table according to the drought level of the drought index to obtain a drought level sequence of each grid;
step 4, according to the drought level sequence obtained in the step 3, the large-scale meteorological factors with certain time lag collected in the step 1 are used as covariates of the drought state transition probability, and a meteorological drought level prediction model with the time-varying transition probability is constructed on the basis of a non-stable Markov chain model;
and 5, adopting the non-stable Markov chain model obtained through optimization in the step 4 to predict the regional weather drought level, so as to obtain the spatial distribution of the whole regional drought level. Further, in the step 2, a spatial downscaling model is constructed by coupling basic underlying surface information by adopting a multivariate linear regression method, and the original remote sensing monitoring rainfall data is subjected to scale degradation.
Further, in the step 2, a geographic information difference method is adopted, and the output high-resolution precipitation data is further corrected by actually measuring precipitation by using the ground observation station.
Further, in step 3, the normalized drought index uses a normalized precipitation index.
Further, in step 4, parameters of each non-stationary Markov chain model are estimated by adopting a generalized cross entropy method.
Further, in step 4, the historical drought level is calculated by retrogradation of a calibrated non-stable Markov chain model, and a final meteorological drought level prediction model is optimized by adopting a Chichi information content criterion.
Compared with the prior art, the invention has the following advantages and effects:
1. the multi-source remote sensing information and the characteristics of the underlying surface of the foundation are fully utilized, the spatial variability of regional rainfall is captured, and the defects that the density and the distribution of rainfall points observed by a traditional station are uneven and the distribution characteristics of the rainfall on the surface cannot be reflected can be overcome.
2. A non-stationary Markov chain model with time-varying characteristics and considering covariates is established, the defect that only the autocorrelation characteristics of a meteorological hydrological sequence are considered in the traditional method is overcome, and the dynamic evolution characteristics of a meteorological hydrological system are better fitted.
3. The method fully utilizes the external stress and precursor signals of the drought and waterlogging evolution, namely the large-scale meteorological factors capable of reflecting the atmospheric circulation characteristics, considers the formation and development mechanism of drought and waterlogging to a certain extent, has stronger scientificity and practicability, and can lay a foundation for building a regional drought and waterlogging early warning and forecasting system.
Drawings
FIG. 1 is a detailed flow diagram of the method of the present invention;
FIG. 2 is a schematic diagram of watershed meshing;
FIG. 3 is a schematic diagram of a precipitation probability distribution fit;
fig. 4 is a schematic diagram of drought level prediction results.
Detailed Description
The technical scheme of the invention is further explained by combining the attached drawings.
FIG. 1 is a detailed flow chart of the method of the present invention, which comprises the following steps:
step 1, collecting information of a foundation underlying surface in the area and synchronous long series data of precipitation and large-scale meteorological factors. This step is conventional in the art.
The basic underlying surface information mainly refers to lattice-point high-resolution Digital Elevation Model (DEM) data, lattice-point high-resolution land cover data (such as Normalized Differential Vegetation Index (NDVI)), and the like; the rainfall data comprises actual measurement of a ground rainfall observation station and grid-like and low-resolution satellite remote sensing monitoring data; the large-scale meteorological factors comprise index data representing various large-scale circulation features in the global range.
Step 2, constructing a remote sensing precipitation data space downscaling model corrected by considering the factors of the underlying surface of the foundation, and processing low-resolution original remote sensing monitoring precipitation data into higher-resolution precipitation data; and correcting the output data by using actually measured precipitation data of the ground observation station to obtain a high-resolution regional precipitation space database.
In the step 2, a spatial downscaling model of the remote sensing rainfall data is constructed by coupling the underlying surface information of the region by adopting a multivariate linear regression method, and the original remote sensing monitoring rainfall data is subjected to scale degradation; and further correcting the output regional precipitation space database with high resolution by using the actual measurement precipitation of the ground observation station by adopting a geographic information difference (GDA) method. Further comprising the substeps of:
(1) according to the spatial resolution of the underlying surface information of the region, the region is spatially dispersed on the basis of a Geographic Information System (GIS) platform and is divided into uniform high-resolution longitude and latitude grids. A schematic diagram of watershed meshing is shown in fig. 2.
(2) And resampling the underlying surface information (a Digital Elevation Model (DEM) and a Normalized Differential Vegetation Index (NDVI)) of the area by adopting a nearest neighbor interpolation method, so that the resolution of the underlying surface information is consistent with that of original remote sensing monitoring rainfall data.
Nearest neighbor interpolation is a technique conventional in the art.
(3) Altitude (h), slope (alpha) and gradient (beta) factors are extracted from a Digital Elevation Model (DEM).
(4) Method for constructing regional underlying surface information and original remote sensing rainfall data by adopting multivariate linear regression method
The empirical relationship of (1):
in the formula: b ═ b0,b1,b2,b3,b4,b5,b6) The parameter matrix of the multiple linear regression model is estimated by a least square method, Z ═ 1, X, Y, h, α is an independent variable matrix, X and Y are respectively the longitude and latitude of the center of a grid point, and the parameters are obtained according to the estimation
Calculating remote sensing precipitation analog value under original low resolution
ZLR＝(1,XLR,YLR,hLR,αLR,βLR,NDVILR) Is an independent variable matrix under low resolution.
The least squares method is conventional in the art.
(5) At low resolution of the original calculation
And
the residual error of (a):
(6) interpolating the low-resolution grid residual by adopting an inverse distance weight interpolation method to obtain a high-resolution residual value
The inverse distance weight interpolation method is conventional in the art.
(7) Applying the parameters of the multiple linear regression model estimated in the step (4) to the high-resolution grid points to obtain an initial high-resolution remote sensing precipitation simulation value
And from the high resolution residual values
Correcting to obtain a remote sensing precipitation analog value under high resolution
ZHR＝(1,XHR,YHR,hHR,αHR,βHR,NDVIHR) Is an independent variable matrix under high resolution.
(8) Calculating each actually measured precipitation of the ground observation station collected in the step 1
Remote sensing precipitation analog value of high resolution grid point containing the site
The residual error between:
(9) station precipitation residual error delta P by adopting inverse distance weight interpolation methodpointInterpolation is carried out to obtain a high-resolution precipitation correction factor value
(10) Precipitation correction factor at high resolution
Adding high-resolution remote sensing precipitation analog value
Obtaining the final high-resolution remote sensing precipitation corrected value
And forming a high-resolution regional precipitation space database.
And 3, sequentially carrying out frequency analysis on the long-series rainfall data of each high-resolution grid in the high-resolution regional rainfall space database obtained in the step 2, calculating a standardized drought index of the long-series rainfall data, and obtaining a drought level sequence of each grid according to a drought index drought level division table.
In step 3, the normalized drought index is the normalized precipitation index (SPI). The calculation steps are as follows:
(1) and (3) fitting the accumulated precipitation of different time scales of each month by using a Gamma distribution line for each corrected high-resolution grid point remote sensing precipitation sequence obtained in the step (2). The probability density function of the Gamma distribution is as follows:
in the formula: a is1And a2Respectively obtaining the shape and scale parameters of Gamma distribution, and estimating by adopting a maximum likelihood method; (. cndot.) is a Gamma function; p is the accumulated precipitation in time periods, such as ten days, months, seasons, years and the like. A precipitation probability distribution fit is shown in figure 3.
Preferably, the present embodiment is based on the month as a time scale, considering that the current remote sensing precipitation products are generally more credible in the month and above scales, but still have larger uncertainty in the smaller scale. With the further improvement of the precision of remote sensing products, the method can be applied to smaller time scales.
The maximum likelihood method is conventional in the art.
(2) Calculating the cumulative probability of precipitation in a certain period of time:
(3) converting the cumulative probability into quantile of standard normal distribution according to an equiprobability principle, namely SPI:
SPI＝Φ-1(G (p)) (9) wherein: phi-1(. cndot.) is the inverse of the standard normal distribution probability distribution function.
In step 3, the drought grade division of the drought index adopts the degree of deviation of the cumulative probability from the normal level (50% quantile).
SPI-based drought rating scale is shown in table 1.
TABLE 1 SPI index drought level Classification Table
As can be seen from the table, SPI can be used not only for drought monitoring, but also for monitoring regional flooding conditions.
And 4, constructing a meteorological drought level prediction model with a time-varying transition probability based on a non-stationary Markov chain model by taking the large-scale meteorological factor with certain time lag as a covariate of the drought state transition probability according to the drought level sequence obtained in the step 3.
Taking the whole research area as a whole system, taking high-resolution grid points in the area as a computing unit, and taking the grid point number of the drought and waterlogging structure of the system at different drought levels at a certain time t as the proportion of the total computing unit number (namely the proportion of the disaster area A (t) ═ A (t) (A))1t,A2t,...,A7t) ) and evolves over time, often described in a Markov chain model.
The stationary Markov chain model assumes the probability pi of converting the drought level state of any one computing unit in the system from i to j at the future t +1 momentij(i, j ═ 1, 2.., 7 in this embodiment) is only relevant to state i, which is known at the current time t, and not to previous states, and does not change over time, i.e.:
in the formula: i. j is the state value, S is the state field, and T is the time field. The evolution process of the proportion of the affected area can be expressed as:
A(t+1)＝A(t)π (11)
in the formula: and pi is a state transition matrix.
In fact, limited observation samples of drought and flood conditions often lead to unreasonable nulls in the experience matrix of the condition transition. Meanwhile, regional precipitation is controlled by various factors and is a typical expression of combined action of a sea-land-air system, the state transfer rule of the regional precipitation is not only related to the current state, but also changes along with the change of an external environment (such as atmospheric stress), therefore, the invention adopts a non-stable time-varying Markov chain model to describe the drought and waterlogging state transfer process, supposes that the transfer probability of the drought and waterlogging state changes along with time, and introduces an external explanation variable to construct the quantitative relation between the external explanation variable and the transfer probability:
πij(t)＝fij(zij(t),ηij) (12)
in the formula: z is a radical ofij(t) is an explanatory variable matrix, ηijAs its regression parameter, fij(. cndot.) is a function of the join covariate and the transition probability. Often, linear regression equations are used to construct empirical relationships between explanatory variables and transition probabilities:
Φ-1(πij(t))＝ηijzij(t) (13)
in the formula: phi-1(. cndot.) is the inverse of the standard normal distribution probability distribution function. The quantile transformation is used to transform the dependent variable from the probability interval [0, 1]]Into a continuous real space.
Step 4, preliminarily selecting the types and the time lags of the large-scale meteorological factors serving as covariates from the alternative large-scale meteorological factors through correlation analysis; estimating parameters of each non-stationary Markov chain model by adopting a generalized cross entropy method; and (3) calculating the historical drought level through a calibrated non-stationary Markov chain model back-substitution, and preferably selecting a final meteorological drought level prediction model by adopting an Akaikeinformation Criterion (AIC) Criterion. Further comprising the substeps of:
(1) setting the maximum Lag Lag of the regional drought level evolution responding to the fluctuation of the large-scale meteorological factors by taking the index data long series data which are collected in the step 1 and represent the large-scale circulation characteristics in the global scope as the candidate large-scale meteorological factor setmaxAnd minimum Lag time LagminGradually increasing all the candidate large-scale meteorological factors from the minimum lag time to the maximum lag time, respectively carrying out correlation analysis and inspection on the precipitation series and the asynchronous sequence of the candidate large-scale meteorological factors by adopting the correlation coefficients according toAnd (4) selecting a plurality of groups of large-scale meteorological factor covariates which have significant influence on the drought and waterlogging evolution process according to the correlation coefficient.
(2) And (4) estimating parameters of each non-stationary Markov chain model by adopting a generalized cross entropy method.
(3) And calculating the historical drought level by a calibrated non-stable Markov chain model back substitution, and preferably selecting a final meteorological drought level prediction model by adopting an AIC criterion.
In step 4, the alternative large-scale meteorological factor set may be one or more of a great north ocean wave index (NAO), a great north wave index (AO), pacific annual oscillation (PDO), a south wave index (SOI), a Multivariable ENSO Index (MEI), great north ocean annual oscillation (AMO), and the like.
In step 4, the implementation takes months as the time step, and the minimum Lag LagminTaking 1 month, the maximum Lag time LagmaxAnd taking for 12 months.
In the step 4, parameters of the non-stationary Markov model are estimated by adopting a generalized cross entropy method, and the problem that the difference between the information quantity carried by the posterior estimation of the model transition probability and the information quantity carried by the prior estimation is minimum can be solved on the premise of fully utilizing all information without increasing redundancy.
(1) Objective function (cross entropy):
in the formula:
is a prior transition probability matrix and provides information for solving the posterior transition probability.
(2) Constraint conditions are as follows:
prior transition probability matrix
Can be replaced by an empirical transfer matrix derived from a stationary Markov chain model.
And 5, predicting the regional weather drought level by adopting the non-stable Markov chain model obtained through optimization in the step 4 to obtain the spatial distribution of the whole regional drought level.
And 5, sequentially applying the drought level prediction model obtained through optimization to each high-resolution grid point to obtain the probability of the drought level of the grid point to be transferred to different levels at the next moment, issuing and forecasting by taking the level state with the maximum transfer probability as the drought level of the grid point at the next moment, and estimating the disaster area of each level. A schematic diagram of the drought level prediction results at a certain time is shown in fig. 4.
The above description is only an embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (6)
1. A regional weather drought level prediction method coupled with multi-source data is characterized by comprising the following steps:
step 1, collecting information of a foundation underlying surface in a region and synchronous long series data of precipitation and large-scale meteorological factors;
step 2, constructing a remote sensing precipitation data space statistics downscaling model with high-resolution basic underlying surface factor correction considered, processing low-resolution original remote sensing monitoring precipitation data into high-resolution remote sensing precipitation data, calculating a residual error correction factor by using ground observation station actual measurement precipitation data, correcting output data to obtain a high-resolution regional gridding precipitation space database, and quantitatively simulating spatial heterogeneity of regional precipitation distribution;
step 3, sequentially carrying out frequency analysis on the long-series rainfall data of each high-resolution grid in the regional gridding rainfall space database obtained in the step 2, calculating a standardized drought index of the long-series rainfall data, and dividing a table according to the drought level of the drought index to obtain a drought level sequence of each grid;
step 4, assuming that the transition probability between the drought classes changes along with time according to the drought class sequence obtained in the step 3, taking the large-scale meteorological factor sequence with certain time lag collected in the step 1 as an external explanatory variable of the drought state transition probability, describing a quantitative function relation between the external explanatory variable and the drought state transition probability based on a non-stationary Markov chain model, and constructing a meteorological drought class prediction model with the time-varying transition probability;
and 5, predicting the weather drought level of each high-resolution grid point in the region by using the non-stationary Markov chain model obtained through optimization in the step 4 to obtain the evolution process of the space distribution of the drought level and the proportion of the disaster area in the whole region.
2. The method for forecasting regional weather drought level coupled with multi-source data as claimed in claim 1, characterized in that:
in the step 2, a spatial downscaling model is constructed by coupling basic underlying surface information by adopting a multivariate linear regression method, and original remote sensing monitoring rainfall data are subjected to scale degradation.
3. The method for forecasting regional weather drought level coupled with multi-source data as claimed in claim 1 or 2, characterized in that:
and step 2, further correcting the output high-resolution precipitation data by actually measuring precipitation by using the ground observation station by adopting a geographic information difference method.
4. The method for forecasting regional weather drought level coupled with multi-source data as claimed in claim 1, characterized in that:
in step 3, the standardized drought index uses a standardized precipitation index.
5. The method for forecasting regional weather drought level coupled with multi-source data as claimed in claim 1, characterized in that:
and 4, estimating parameters of each non-stationary Markov chain model by adopting a generalized cross entropy method.
6. The method for forecasting regional weather drought level coupled with multi-source data as claimed in claim 1 or 5, characterized in that:
and 4, calculating the historical drought level through a calibrated non-stable Markov chain model back substitution, and preferably selecting a final meteorological drought level prediction model by adopting a Chichi pool information quantity criterion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610850079.6A CN107316095B (en) | 2016-09-23 | 2016-09-23 | Regional weather drought level prediction method coupled with multi-source data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610850079.6A CN107316095B (en) | 2016-09-23 | 2016-09-23 | Regional weather drought level prediction method coupled with multi-source data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107316095A CN107316095A (en) | 2017-11-03 |
| CN107316095B true CN107316095B (en) | 2020-09-11 |
Family
ID=60185398
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610850079.6A Active CN107316095B (en) | 2016-09-23 | 2016-09-23 | Regional weather drought level prediction method coupled with multi-source data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107316095B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112131989A (en) * | 2020-09-15 | 2020-12-25 | 河海大学 | Millimeter wave rain measurement model parameter obtaining method based on space rainfall data |
Families Citing this family (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107944387B (en) * | 2017-11-22 | 2021-12-17 | 重庆邮电大学 | Method for analyzing spatial heterogeneity of urban heat island based on semi-variation theory |
| CN108009398B (en) * | 2017-12-12 | 2019-08-06 | 河海大学 | A kind of GCM bearing calibration considering day by day data fluctuation characteristic |
| CN107944219B (en) * | 2017-12-13 | 2021-01-22 | 广东电网有限责任公司电力科学研究院 | Method and device for representing drought and waterlogging disaster-causing characteristics at different time periods |
| CN108154193B (en) * | 2018-01-16 | 2021-10-08 | 黄河水利委员会黄河水利科学研究院 | Long-time sequence precipitation data downscaling method |
| CN110442937B (en) * | 2019-07-24 | 2023-01-24 | 武汉大学 | Drainage basin hydrological simulation method integrating satellite remote sensing and machine learning technology |
| CN111611541B (en) * | 2020-04-26 | 2022-05-20 | 武汉大学 | Method and system for calculating data-free area precipitation data based on Copula function |
| CN111738175A (en) * | 2020-06-24 | 2020-10-02 | 桂林理工大学 | Agricultural drought monitoring system based on remote sensing image and convolutional neural network |
| CN111860974B (en) * | 2020-06-30 | 2022-11-01 | 中国地质大学（武汉） | Drought multistage prediction method based on state space and joint distribution |
| CN112084433B (en) * | 2020-09-14 | 2024-04-16 | 周盛 | Method for carrying out weather modification drought-resistant operation according to regional division |
| CN112649898A (en) * | 2020-11-11 | 2021-04-13 | 中国气象局广州热带海洋气象研究所（广东省气象科学研究所） | Weather drought refined monitoring method |
| CN112734118A (en) * | 2021-01-14 | 2021-04-30 | 华北水利水电大学 | Drought prediction method based on improved CEEMDAN-QR-BL mixed model |
| CN113591033B (en) * | 2021-05-28 | 2022-03-25 | 河海大学 | Arid region underground water ecological burial depth analysis method based on joint probability distribution |
| CN113946796B (en) * | 2021-09-30 | 2024-08-09 | 西安理工大学 | Drought propagation time calculation method based on conditional probability high space-time resolution |
| CN113988673A (en) * | 2021-11-02 | 2022-01-28 | 中山大学 | Drought and waterlogging sudden turning evaluation method |
| CN114677059B (en) * | 2022-05-26 | 2022-08-23 | 水利部交通运输部国家能源局南京水利科学研究院 | Method and system for comprehensively evaluating precision of inversion precipitation product by integrating time-space indexes |
| CN115166874A (en) * | 2022-07-13 | 2022-10-11 | 北京师范大学 | Meteorological drought index SPI construction method based on machine learning |
| CN115409340A (en) * | 2022-08-17 | 2022-11-29 | 西北农林科技大学 | Short-period drought and waterlogging rush index construction method and system |
| CN115639979B (en) * | 2022-09-15 | 2023-05-30 | 河南大学 | High-resolution SPEI data set development method based on random forest regression model |
| CN116881839B (en) * | 2023-07-05 | 2024-08-16 | 江苏省气象台 | Judgment method and system for distinguishing attenuation type of ENSO event |
| CN116680548B (en) * | 2023-08-03 | 2023-10-13 | 南京信息工程大学 | Time sequence drought causal analysis method for multi-source observation data |
| CN117787542B (en) * | 2023-12-22 | 2024-09-06 | 暗物质(北京)智能科技有限公司 | Drought risk assessment method, drought risk assessment device, computer equipment and storage medium |
| CN118520788B (en) * | 2024-07-24 | 2024-09-27 | 江苏禹治流域管理技术研究院有限公司 | Hydrological weather prediction system and method for multiple data sources |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102779391A (en) * | 2012-07-24 | 2012-11-14 | 中国农业科学院农田灌溉研究所 | Drought early-warning method and drought early-warning system |
| CN103577720A (en) * | 2013-11-29 | 2014-02-12 | 民政部国家减灾中心 | Method for estimating regional drought risk |
| CN105550501A (en) * | 2015-12-09 | 2016-05-04 | 中国水利水电科学研究院 | Method for recognizing drought evolution driving mechanism of basin/region |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7347007B2 (en) * | 2000-06-16 | 2008-03-25 | Maguire Stephen B | Low pressure high capacity dryer for resins and other granular and powdery materials |
-
2016
- 2016-09-23 CN CN201610850079.6A patent/CN107316095B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102779391A (en) * | 2012-07-24 | 2012-11-14 | 中国农业科学院农田灌溉研究所 | Drought early-warning method and drought early-warning system |
| CN103577720A (en) * | 2013-11-29 | 2014-02-12 | 民政部国家减灾中心 | Method for estimating regional drought risk |
| CN105550501A (en) * | 2015-12-09 | 2016-05-04 | 中国水利水电科学研究院 | Method for recognizing drought evolution driving mechanism of basin/region |
Non-Patent Citations (2)
| Title |
|---|
| "基于SPEI的北京低频干旱与气候指数关系";苏宏新 等;《生态学报》;20120930;正文第1-2节 * |
| "基于马尔科夫模型的新疆水文气象干旱研究";孙鹏 等;《地理研究》;20140930;正文第3-4节 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112131989A (en) * | 2020-09-15 | 2020-12-25 | 河海大学 | Millimeter wave rain measurement model parameter obtaining method based on space rainfall data |
| CN112131989B (en) * | 2020-09-15 | 2021-07-23 | 河海大学 | Millimeter wave rain measurement model parameter obtaining method based on space rainfall data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107316095A (en) | 2017-11-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107316095B (en) | Regional weather drought level prediction method coupled with multi-source data | |
| CN112070286B (en) | Precipitation forecast and early warning system for complex terrain river basin | |
| Notter et al. | Impacts of environmental change on water resources in the Mt. Kenya region | |
| CN108983320B (en) | Numerical weather forecast-artificial intelligence coupling prediction method for coastal typhoon extreme wind speed | |
| KR102073768B1 (en) | Drought information supply system based on portal | |
| CN110598290B (en) | Method and system for predicting future hydropower generation capacity of basin considering climate change | |
| CN105912836B (en) | A kind of watershed hydrologic cycle analogy method of pure remotely-sensed data driving | |
| Hassan | Climate change impact on groundwater recharge of Umm er Radhuma unconfined aquifer Western Desert, Iraq | |
| Hunter et al. | Development and application of a North American Great Lakes hydrometeorological database—Part I: Precipitation, evaporation, runoff, and air temperature | |
| CN111898660B (en) | Hydrological simulation method for average fusion of multi-source data based on Bayesian mode | |
| CN114019579B (en) | High space-time resolution near-surface air temperature reconstruction method, system and equipment | |
| CN110442937A (en) | A kind of valley water security management method merging satellite remote sensing and machine learning techniques | |
| CN113553782B (en) | Downscaling method for forecasting wind speed | |
| CN110781259A (en) | Hydrological model based on landform unit line | |
| CN112785024A (en) | Runoff calculation and prediction method based on watershed hydrological model | |
| KR20090093520A (en) | Decision support method for integrated river basin water resources management | |
| CN113569438A (en) | Urban flood model construction method based on multi-source rainfall fusion and real-time correction | |
| CN111539597A (en) | Gridding drainage basin social and economic drought assessment method | |
| CN117763970B (en) | Image deep learning-based hydrologic model parameter reconstruction method for scarce data area | |
| CN110543971A (en) | Satellite rainfall and actual rainfall error partition fusion correction method | |
| CN117807917B (en) | Loss function construction method and system based on scene flood disasters | |
| CN116205136A (en) | Large-scale river basin deep learning flood forecasting method based on runoff lag information | |
| Zhang et al. | Quantification of human and climate contributions to multi-dimensional hydrological alterations: A case study in the Upper Minjiang River, China | |
| CN107944466B (en) | Rainfall deviation correction method based on segmentation idea | |
| CN104156776B (en) | A kind of solar energy resources appraisal procedure |
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 |