CN108646319B - Short-time strong rainfall forecasting method and system - Google Patents
Short-time strong rainfall forecasting method and system Download PDFInfo
- Publication number
- CN108646319B CN108646319B CN201810902276.7A CN201810902276A CN108646319B CN 108646319 B CN108646319 B CN 108646319B CN 201810902276 A CN201810902276 A CN 201810902276A CN 108646319 B CN108646319 B CN 108646319B
- Authority
- CN
- China
- Prior art keywords
- optical flow
- fleet
- forecasting
- product
- flow method
- 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.)
- Expired - Fee Related
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/10—Devices for predicting weather conditions
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
-
- 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)
- Environmental & Geological Engineering (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Life Sciences & Earth Sciences (AREA)
- Atmospheric Sciences (AREA)
- Biodiversity & Conservation Biology (AREA)
- Ecology (AREA)
- Environmental Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application relates to a short-time heavy rainfall forecasting method and system. The method comprises the following steps: step a: reading radar base data, and performing radar echo quality control processing on the radar base data; step b: constructing a forecasting system based on a Fleet constraint optical flow method according to the radar base data after the radar echo quality control processing; step c: and constructing an identification tracking system based on the Fleet constrained optical flow method according to the forecasting system based on the Fleet constrained optical flow method, and analyzing data and generating and outputting a Fleet optical flow method product. By constructing a brand-new forecasting system based on a Fleet constraint optical flow method, the problem of echo extrapolation forecasting of locally generated rainfall cloud cluster with the intensity and shape thereof rapidly changing along with time is solved, and the performance of an approach forecasting system for short-time heavy rainfall weather is improved.
Description
Technical Field
The application belongs to the technical field of meteorological service, and particularly relates to a short-time heavy rainfall forecasting method and system.
Background
Short term heavy rainfall is a strong convective weather, meaning that the rainfall in a place exceeds 30 mm in 3 hours. The short-time heavy rainfall has the characteristics of strong burst, violent coming, concentrated rainfall time and the like, and is high in rainfall intensity and concentrated in time, so that urban waterlogging, landslide and other weather-derived disasters are often caused by the short-time heavy rainfall, urban operation is seriously influenced, and even the life is threatened. Short-time heavy rainfall weather often comes in hurry, the forecasting effect by using a conventional meteorological forecasting means is poor, the difficulty of forecasting the weather in advance for 24 hours is high, and meanwhile, due to the fact that the numerical mode is unstable for the initial hours caused by current mode errors, data assimilation and the like, the short-time close forecasting and the demand of the numerical mode are obviously different, and particularly the automatic identification and falling area forecasting and early warning capabilities of the heavy rainfall weather are low.
Take Shenzhen as an example, Shenzhen is an important navy, army and air transportation hub city in China, a coastal city in south China and adjacent to hong Kong. Due to the special geographical position, the Shenzhen city is often faced with sudden weather events such as rainstorms, strong winds, strong waves and the like caused by monsoon and typhoon, and directly influences the security of urban public facilities and the property and life security of people. The demand for accurate and timely fine weather forecast guarantee service drives the rapid development of fine weather forecast. And various major social public activities such as home and abroad large-scale sports events, exhibitions, meetings and the like are frequently held in Shenzhen, which all provide new requirements for the refinement degree of urban weather forecast, particularly rainfall forecast.
Shenzhen is narrow and long in east-west, the narrowest place of the south-north boundary is only about 10 kilometers away, the east-west coast line is as long as 230 kilometers away, meteorological elements are extremely unevenly distributed in space and time, meteorological disasters often have the characteristics of strong outburst, uneven local distribution and easiness in causing disasters, and the situation that rain occurs in the east and west of the sun often occurs. According to historical data statistics, the average difference of the regions with the most rainfall and the least rainfall in Shenzhen calendar year is as high as 1000 mm. Severe losses are often caused by severe weather, such as disasters like urban inland inundation caused by strong rainfall, dangerous slope landslide and the like; the strong wind and heavy rain caused by typhoon can also cause geological disasters, resulting in large-area water immersion. The difficulty of accurately forecasting the precipitation area and the specific influence time is high, and the method is also a research hotspot and difficulty of meteorological workers.
Compared with rainstorm, in the aspect of forecasting, large-scale rainstorm weather caused by laminar cloud is relatively better reported in a falling area, short-time strong precipitation forecasting is always difficult, the falling area is always a point which is scattered one by one, and a forecaster only knows that the short-time strong precipitation with dispersity occurs in a certain area, but the specific point where the short-time strong precipitation occurs is difficult to make accurate forecasting. However, the short-time heavy rainfall causes particularly great harm, and in mountainous areas and areas in the southwest and northwest of China, disasters such as mountain floods, landslides, debris flows and the like are easy to occur due to poor water and soil conditions, and most of the disasters are caused by short-time heavy rainfall.
At present, the automatic extrapolation technology for the convection nowcasting at home and abroad mainly comprises three types: the monomer centroid method, the cross correlation method, and the optical flow method. The cross-correlation method forecast radar echo is developed in the 1990 s in the United states, and the cross-correlation method begins to be applied to the short-term forecast before and after 2008 in China. The single body mass center method is to identify, analyze and track the thunderstorm as a three-dimensional single body, and perform fitting extrapolation on the thunderstorm to perform the approach prediction. The cross correlation method is characterized in that the method for obtaining the optimal spatial correlation of the radar echo is utilized, the moving vector characteristics of the echo in the past are determined by calculating the optimal spatial correlation coefficient of different areas of the continuous time-secondary radar echo in a two-dimensional area, the optimal fitting relation of the radar echo in different times is established, the moving characteristics of the radar echo in the past in a certain area are tracked, and then the future position and shape of the echo are extrapolated through the moving characteristics of the echo. The main advantage of using the cross-correlation algorithm is that the calculation method is simpler, and compared with the method that the single centroid method can only be used for the convection precipitation system, the cross-correlation method can track the convection precipitation system and the lamellar cloud precipitation. Therefore, the cross-correlation method is widely used in the meteorological service department.
However, in daily business use, it is found that in the case of precipitation echoes which are locally generated and whose intensity and shape change rapidly with time, the quality of the motion vector field given by the cross-correlation method is reduced, and the cases of tracking failure are remarkably increased. The cross-correlation method has an inherent defect and needs to introduce a new method. Therefore, it is necessary to carry out more targeted and intensive research on the forecasting technology of the short-time heavy rainfall within 0 to 3 hours, and develop a new approach extrapolation forecasting method suitable for the short-time heavy rainfall weather with severe weather system changes so as to meet the fine demand of the objective rainfall forecast within 0 to 3 hours.
Disclosure of Invention
The application provides a short-term heavy rainfall forecasting method and system, and aims to solve at least one of the technical problems in the prior art to a certain extent.
In order to solve the above problems, the present application provides the following technical solutions:
a short-term heavy rainfall forecasting method comprises the following steps:
step a: reading radar base data, and performing radar echo quality control processing on the radar base data;
step b: constructing a forecasting system based on a Fleet constraint optical flow method according to the radar base data after the radar echo quality control processing;
step c: and constructing an identification tracking system based on the Fleet constrained optical flow method according to the forecasting system based on the Fleet constrained optical flow method, and analyzing data and generating and outputting a Fleet optical flow method product.
The technical scheme adopted by the embodiment of the application further comprises the following steps: in the step a, the radar echo quality control processing on the radar base data specifically includes weighted gaussian filtering, a morphological analysis expansion algorithm, a morphological analysis erosion algorithm, background subtraction of radar data, and radar data regularization processing.
The technical scheme adopted by the embodiment of the application further comprises the following steps: in the step b, the method based on the Fleet constrained optical flow comprises Fleet backward differential extrapolation, Fleet advection extrapolation, cross correlation backward differential extrapolation and cross correlation Lagrange advection extrapolation; the specific algorithm of the Fleet constrained optical flow method comprises the following steps:
step b 1: solving a complex deconvolution of a radar live field according to a motion calculation and estimation theory of a Fleet and Jepson image component;
step b 2: applying a core algorithm for mapping a color point noise field to an optical flow field, performing dynamic color optical flow field and 2D and 3D dynamic color mapping, and dynamically mapping the optical flow field to a 6-channel basic color space with three coordinate axes;
step b 3: converting the optical flow field into a core algorithm based on a color point noise format, and performing two-dimensional display on two-dimensional optical flow fields with different formats;
step b 4: median filtering of radar echo live.
The technical scheme adopted by the embodiment of the application further comprises the following steps: in the step c, the generation and output of the Fleet optical flow method product specifically include:
step c 1: generating a radar echo extrapolation product based on a Fleet constraint optical flow method, and outputting and storing the product;
step c 2: combining Kalman filtering and CRESMAN objective analysis to generate a QPE product based on a Fleet optical flow method, and outputting and storing the QPE product;
step c 3: combining Kalman filtering and CRESMAN objective analysis to generate a QPF product based on a Fleet optical flow method, and outputting and storing the QPF product;
step c 4: and generating a comparison check product of the QPF product based on the Fleet constrained optical flow method and the cross-correlation QPF, and saving the output of the comparison check product.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the step c further comprises the following steps: comprehensively applying a GIS technology and an information visualization technology, constructing a product comprehensive display system based on a fly constraint optical flow method based on WEBGIS, and comprehensively displaying products of the fly optical flow method; and taking the generated Fleet optical flow method product as a set member, respectively calculating the short-term aggregate precipitation probability forecasting results by adopting an averaging method, a correlation method and a Rank method, and comprehensively displaying the short-term aggregate precipitation probability forecasting results on a system platform.
Another technical scheme adopted by the embodiment of the application is as follows: a short-term heavy rainfall forecasting system comprising:
the forecasting system construction module based on the Fleet constraint optical flow method comprises the following steps: the forecasting system is used for reading radar base data, performing radar echo quality control processing on the radar base data, and constructing a forecasting system based on a Fleet constraint optical flow method according to the radar base data after the radar echo quality control processing;
the method comprises the following steps of identifying and tracking a system construction module based on a Fleet constraint optical flow method: and the method is used for constructing an identification tracking system based on the Fleet constrained optical flow method according to the forecasting system based on the Fleet constrained optical flow method, analyzing data and generating and outputting a Fleet optical flow method product.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the radar echo quality control processing of the radar base data by the light flow method forecasting system building module based on the Fleet constraint specifically comprises weighted Gaussian filtering, a morphological analysis expansion algorithm, a morphological analysis erosion algorithm, background subtraction of radar data and radar data regularization processing.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the method comprises the steps of carrying out Fleet backward differential extrapolation, Fleet advection extrapolation, cross correlation backward differential extrapolation and cross correlation Lagrange advection extrapolation on the basis of a Fleet constraint optical flow method; the specific algorithm of the Fleet constrained optical flow method comprises the following steps: solving a complex deconvolution of a radar live field according to a motion calculation and estimation theory of a Fleet and Jepson image component; applying a core algorithm for mapping a color point noise field to an optical flow field, performing dynamic color optical flow field and 2D and 3D dynamic color mapping, and dynamically mapping the optical flow field to a 6-channel basic color space with three coordinate axes; converting the optical flow field into a core algorithm based on a color point noise format, and performing two-dimensional display on two-dimensional optical flow fields with different formats; median filtering of radar echo live.
The technical scheme adopted by the embodiment of the application further comprises the following steps: the method for generating and outputting the Fleet optical flow method product based on the Fleet constrained optical flow method identification and tracking system construction module specifically comprises the following steps: generating a radar echo extrapolation product based on a Fleet constraint optical flow method, and outputting and storing the product; combining Kalman filtering and CRESMAN objective analysis to generate a QPE product based on a Fleet optical flow method, and outputting and storing the QPE product; combining Kalman filtering and CRESMAN objective analysis to generate a QPF product based on a Fleet optical flow method, and outputting and storing the QPF product; and generating a comparison check product of the QPF product based on the Fleet constrained optical flow method and the cross-correlation QPF, and saving the output of the comparison check product.
The technical scheme adopted by the embodiment of the application further comprises the following steps:
the product comprehensive display system construction module based on the Fleet constraint optical flow method comprises the following steps: the system is used for comprehensively applying a GIS technology and an information visualization technology, constructing a product comprehensive display system based on a fly constraint optical flow method based on WEBGIS, and comprehensively displaying products of the fly optical flow method;
the short-term aggregate precipitation probability forecast result calculation module: and the method is used for calculating the short-term aggregate rainfall probability forecasting results by respectively adopting an averaging method, a correlation method and a Rank method by taking the generated Fleet optical flow method product as an aggregate member, and comprehensively displaying the short-term aggregate rainfall probability forecasting results on a system platform.
Compared with the prior art, the embodiment of the application has the advantages that: according to the short-time heavy rainfall forecasting method and system, a brand-new forecasting system based on the Fleet constrained optical flow method is built, the problem of echo extrapolation forecasting of local generation and rapid change of intensity and shape of a rainfall cloud cluster over time is solved, and the performance of a proximity forecasting system for short-time heavy rainfall weather is improved. The method is characterized in that a method combining radar image morphology, characteristic quantity characteristic statistics, pattern recognition and the like is adopted to establish an automatic recognition and approach forecast early warning technology of short-time strong rainfall weather, and an ensemble probability forecast and grade forecast method and a model of the strong convection weather are established on the basis of rainfall probability forecast and grade forecast according to an ensemble forecast theory in a short-time approach ensemble forecast system, so that the early warning capability of the strong convection weather forecast of strong rainfall, hail and the like is improved.
Drawings
FIG. 1 is a flow chart of a method of short-term heavy rainfall forecasting according to an embodiment of the present application;
fig. 2 is a schematic structural diagram of a short-term heavy rainfall forecasting system according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
Please refer to fig. 1, which is a flowchart illustrating a method for forecasting a short-term heavy rainfall according to an embodiment of the present application. The short-time heavy rainfall forecasting method comprises the following steps:
step 100: reading radar base data, and performing radar echo quality control processing on the radar base data;
in step 100, the radar echo quality control processing specifically includes weighted gaussian filtering, morphological analysis expansion algorithm, morphological analysis erosion algorithm, background subtraction of radar data, and radar data regularization processing.
Step 200: constructing a forecasting system based on a Fleet constraint optical flow method according to radar base data after radar echo quality control processing;
in step 200, the optical flow method applied to the meteorological field has the advantages that even if the movement and the shape change of the thunderstorm are severe, the overall movement trend of the thunderstorm can be accurately obtained, the optical flow method uses global smooth constraint to solve a partial differential motion equation, and the overall movement trend of the thunderstorm can be accurately obtained for short-time heavy rainfall with large change. The method specifically comprises the steps of sweet backward differential extrapolation, sweet advection extrapolation, cross correlation backward differential and cross correlation Lagrange advection extrapolation based on the sweet constraint optical flow method. The specific algorithm of the Fleet constrained optical flow method comprises the following steps:
(1) solving a complex deconvolution [ complex deconvolution ] of a radar live field according to a motion calculation and estimation theory of a Fleet and Jepson image component, and realizing the complex deconvolution solution of radar image prediction;
(2) applying a core algorithm for Mapping a Color spot noise field to an optical flow field to realize the optical flow field of Dynamic colors and 2D and 3D Dynamic Color Mapping (Dynamic Color Mapping), and dynamically projecting the optical flow field to a 6-channel basic Color space with three coordinate axes;
(3) converting an optical flow field into a core algorithm based on a color point noise format and performing two-dimensional display on two-dimensional optical flow fields with different formats;
(4) median filtering of radar echo live; the median filtering method is a non-linear smoothing technique, and sets the gray value of each pixel point as the median of all the gray values of the pixel points in a certain neighborhood window of the point. The median filtering is a nonlinear signal processing technology which is based on the ordering statistical theory and can effectively inhibit noise, and the basic principle of the median filtering is to replace the value of one point in a digital image or a digital sequence by the median of all point values in a neighborhood of the point, so that the surrounding pixel values are close to the true values, and isolated noise points are eliminated. The median filtering method is to sort the pixels in the panel according to the size of the pixel value by using a two-dimensional sliding template with a certain structure, and generate a two-dimensional data sequence which is monotonously ascending (or descending). The two-dimensional median filter output is g (x, y) ═ med { f (x-k, y-l), (k, l ∈ W) }, where f (x, y), g (x, y) are the original image and the processed image, respectively. W is a two-dimensional template, typically 2 × 2, 3 × 3 regions, and may be in various shapes such as a line, a circle, a cross, and a circular ring.
Step 300: constructing an identification tracking system based on a Fleet constrained optical flow method, and analyzing data and generating and outputting Fleet optical flow method products;
in step 300, the identification tracking system based on the Fleet constrained optical flow method is deployed at a server side in a distributed mode based on a C/S mode, runs in a multi-thread concurrent mode, mainly completes data analysis and generation and output of Fleet optical flow method products, and specifically comprises the following steps:
(1) generating a radar echo extrapolation product based on a Fleet constraint optical flow method, and outputting and storing the product;
(2) combining Kalman filtering and CRESMAN objective analysis to generate and output a QPE (quantitative precipitation estimation) product based on a Fleet optical flow method, and outputting and storing the output;
(3) and combining Kalman filtering and CRESMAN objective analysis, generating and outputting a QPF (quantitative precipitation prediction) product based on a Fleet optical flow method, and storing the output of the QPF product.
(4) And generating a comparison check product of the QPF product based on the Fleet constrained optical flow method and the cross-correlation QPF, and saving the output of the comparison check product.
Step 400: comprehensively applying a GIS technology and an information visualization technology, constructing a product comprehensive display system based on a fly constraint optical flow method based on WEBGIS (network geographic information system), completing the comprehensive display of the fly optical flow method product, and integrating the product into a nowcasting decision support platform;
in step 400, the product comprehensive display system based on the Fleet constrained optical flow method is based on a B/S framework, and can dynamically display various quantitative estimation forecast, echo live and forecast and other contents based on the Fleet optical flow method on an electronic map in a mottled map mode, so as to provide decision reference for a forecaster. Besides real-time comprehensive display of various types of information, the comprehensive display function module supports the historical playback function of various types of information, and various types of information can be retrieved and checked based on time indexes.
Step 500: taking the generated Fleet optical flow method product as a set member, respectively adopting an averaging method, a correlation method and a Rank method to calculate the short-term aggregate precipitation probability forecasting result, and comprehensively displaying the short-term aggregate precipitation probability forecasting result on a system platform;
in step 500, based on the nonlinearity and complexity of the atmospheric system and the unavoidable uncertainty of the initial value and the mode, the current ensemble prediction theory has been successfully applied to medium-and-short term numerical weather prediction and short-term climate prediction, and permeates into various space and time scales of atmospheric science to explore the application of the ensemble prediction theory in short-term prediction. According to the method and the model, the ensemble probability forecasting method and the ensemble probability forecasting model and the class forecasting model of the strong convection weather are established on the basis of the precipitation probability forecasting and the class forecasting according to the ensemble forecasting theory in the short-time close-to ensemble forecasting system, and the early warning capability of the strong convection weather forecasting such as strong precipitation, hail and the like is improved.
The uncertainty of prediction mainly comes from the uncertainty of the initial state of the atmosphere and the uncertainty of the prediction mode, and the nonlinear characteristic of the atmospheric motion determines that the extremely small error whether from the initial field or the mode is amplified until the larger error is developed in the mode integration process. Ensemble forecasting techniques are currently an effective way to reduce these errors, in which the outcome of each ensemble member represents a possibility of future weather evolution, regardless of how the ensemble member is formed (e.g., by constructing a perturbation field on the basis of the original material that characterizes the error to form the ensemble member, or by selecting a different mode parameterization scheme, varying the values of some physical parameter to form the ensemble member, or taking a different numerical mode as the ensemble member, etc.). The relative probability of a certain weather phenomenon in the future (such as the probability of heavy rain in a certain area) can be obtained by integrating the forecast results of all the set members, and the method has objectivity and quantificationity compared with the conventional probability forecast only by experience or a statistical method, which cannot be realized in the single-mode forecast, so that the method has reference value. Probability forecasting of precipitation does not require forecasting the spatio-temporal distribution of precipitation magnitudes, but rather the probability distribution that occurs when the forecast precipitation intensity reaches a certain magnitude.
According to the method, the light stream method products such as QPF (quick Path fusion) and QPE (quick Path fusion) are used as set members, the data of the set members are unified to a grid with higher resolution by adopting a unified format, unified standard and unified scoring method through analysis or interpolation, the modeling is performed by adopting a cross effective method every day for forecasting, automatic identification and tracking are performed, the forecasting result of the probability of short-term set precipitation is calculated by adopting an averaging method, a correlation method and a Rank method, the result is comprehensively displayed on a system platform, and the forecasting capability of short-term approach is effectively improved. Specifically, the method comprises the following steps:
(1) average method: how to obtain the probability forecast according to the result of each set member, the simplest method is to determine the probability of occurrence of a certain weather phenomenon in the future according to the number of members forecasting the occurrence of the weather phenomenon according to the equal weight of each set member, and if the number of set members is 4, 2 set members forecast that the rainfall reaches the heavy rain in the future for 3h, the probability of the heavy rain in the future for 3h is 50%, which is the most common method at present and called as an averaging method. The weight of each set member in the ensemble forecast is equal to 1/N, N is the number of the set members, and if M set members forecast to generate precipitation of a certain magnitude exist, the probability P of the precipitation of the magnitude in the future is Mx (1/N).
(2) The correlation method comprises the following steps: the method combines a statistical method on the basis of ensemble prediction, and according to accumulated ensemble prediction results of a certain number of times and corresponding actual precipitation conditions, the correlation between the prediction of each ensemble member and the actual precipitation is calculated, and r is assumed to beiThe correlation coefficient between the forecast and the live precipitation of the ith set member is the weight of the set member in the set forecast as follows:
wherein N is the total number of samples; j is the accumulated counter.
If there are M members of the set forecasted to generate a certain magnitude of precipitation, the probability of the generation of the magnitude of precipitation in the future is p, which is the sum of the weights of the M members of the set.
(3) RANK method: the RANK distribution (RANK) of all the members in the set is obtained by statistics of set forecast results for a certain number of times, the probability of a certain weather phenomenon in the future is determined according to the RANK distribution and the forecast results of all the members in the set, and test results show that the forecast effect of the method is better than that of an average method, and the method is called as the RANK method. The RANK method is also a method combining aggregation and statistics. The N set members are arranged in the order of the forecast results from small to large to form N +1 levels (RANK), and the corresponding live events can be located in any level of the N +1 levels. When the forecast results of the live event and all the members of the set are different, it is simple to determine the level of the live event, for example, the forecast precipitation (from small to large) of 8 members of the set is 8,10,11,16,20,22 and 25mm, and if the precipitation of the live event is 21mm, the live event is at the 7 th level of 9 levels. When the forecast results of the live condition and the members in the set are the same, the level of the live condition can be determined only by carrying out technical processing, namely, small-magnitude random quantity is added or subtracted on the forecast results and the live condition values of all the members in the set, the processing does not influence the probability forecast results, the set forecast results are accumulated for a certain number of times, and the results are applied to the level of the live condition corresponding to each forecast. And counting the probability of the live situation at each level, namely the level distribution, and obtaining the precipitation probability forecast according to the forecast result of each member after the distribution exists.
N +1 level distributions, then V occurs in the future1≤V＜V2The precipitation probability P is:
when V is1＝Vi-1,V2When the value is Xi, P is Ri;
when V is1＝0,V2When the number is equal to Xi,
when V is1＝0,0＜V2＜X1When the temperature of the water is higher than the set temperature,
when V is1＝Xi,Xi＜V2≤Xi+1When the temperature of the water is higher than the set temperature,
when Xi < V1≤Xi+1,V2＝Xi+1When the temperature of the water is higher than the set temperature,
generally, the greater the precipitation, the less likely it will occur, and thus more than X will occurNThe distribution of the precipitation probability is not reasonable, the RANK method assumes that the distribution follows Gumbel distribution, F is a Gumbel cumulative distribution function, and then:
when V is1＝XNWhen the temperature of the water is higher than the set temperature,
when V is1≥XNWhen the temperature of the water is higher than the set temperature,
both the correlation method and the RANK method need to accumulate the ensemble prediction results for a certain number of times as statistical samples, and in order to ensure that a certain amount of samples exist, the cross validation technology is adopted, namely, when testing each case, the ensemble prediction results of the other cases are used as the statistical samples.
Please refer to fig. 2, which is a schematic structural diagram of a short-term heavy rainfall forecasting system according to an embodiment of the present application. The short-term heavy rainfall forecasting system comprises a flow constraint optical flow method based forecasting system construction module, a flow constraint optical flow method based identification tracking system construction module, a flow constraint optical flow method based product comprehensive display system construction module and a short-term aggregate rainfall probability forecasting result calculation module.
The forecasting system construction module based on the Fleet constraint optical flow method comprises the following steps: reading radar base data, performing radar echo quality control processing on the radar base data, and constructing a forecasting system based on a Fleet constraint optical flow method according to the radar base data after the radar echo quality control processing; the radar echo quality control processing specifically comprises weighted Gaussian filtering, a morphological analysis expansion algorithm, a morphological analysis erosion algorithm, radar data background subtraction and radar data regularization processing.
The optical flow method is applied to the meteorological field and has the advantages that even if the movement and the shape change of the thunderstorm are severe, the overall movement trend of the thunderstorm can be accurately obtained, the optical flow method uses global smooth constraint to solve a partial differential motion equation, and the overall movement trend of the thunderstorm can be accurately obtained for short-time heavy rainfall with large change. The method specifically comprises a fly backward differential extrapolation method, a fly advection extrapolation method, a cross correlation backward differential method and a cross correlation Lagrange advection extrapolation method. The specific algorithm of the Fleet constrained optical flow method comprises the following steps:
(1) solving a complex deconvolution [ complex deconvolution ] of a radar live field according to a motion calculation and estimation theory of a Fleet and Jepson image component, and realizing the complex deconvolution solution of radar image prediction;
(2) applying a core algorithm for Mapping a Color spot noise field to an optical flow field to realize the optical flow field of Dynamic colors and 2D and 3D Dynamic Color Mapping (Dynamic Color Mapping), and dynamically projecting the optical flow field to a 6-channel basic Color space with three coordinate axes;
(3) converting an optical flow field into a core algorithm based on a color point noise format and performing two-dimensional display on two-dimensional optical flow fields with different formats;
(4) median filtering of radar echo live; the median filtering method is a non-linear smoothing technique, and sets the gray value of each pixel point as the median of all the gray values of the pixel points in a certain neighborhood window of the point. The median filtering is a nonlinear signal processing technology which is based on the ordering statistical theory and can effectively inhibit noise, and the basic principle of the median filtering is to replace the value of one point in a digital image or a digital sequence by the median of all point values in a neighborhood of the point, so that the surrounding pixel values are close to the true values, and isolated noise points are eliminated. The median filtering method is to sort the pixels in the panel according to the size of the pixel value by using a two-dimensional sliding template with a certain structure, and generate a two-dimensional data sequence which is monotonously ascending (or descending). The two-dimensional median filter output is g (x, y) ═ med { f (x-k, y-l), (k, l ∈ W) }, where f (x, y), g (x, y) are the original image and the processed image, respectively. W is a two-dimensional template, typically 2 × 2, 3 × 3 regions, and may be in various shapes such as a line, a circle, a cross, and a circular ring.
The method comprises the following steps of identifying and tracking a system construction module based on a Fleet constraint optical flow method: the method is used for constructing an identification tracking system based on a fly constraint optical flow method, analyzing data and generating and outputting a fly optical flow method product; the identification tracking system based on the Fleet constrained optical flow method is deployed at a server side in a distributed mode based on a C/S mode, runs in a multi-thread concurrent mode, mainly completes data analysis and generation and output of Fleet optical flow method products, and specifically comprises the following steps:
(1) generating a radar echo extrapolation product based on a Fleet constraint optical flow method, and outputting and storing the product;
(2) combining Kalman filtering and CRESMAN objective analysis to generate and output a QPE (quantitative precipitation estimation) product based on a Fleet optical flow method, and outputting and storing the output;
(3) and combining Kalman filtering and CRESMAN objective analysis, generating and outputting a QPF (quantitative precipitation prediction) product based on a Fleet optical flow method, and storing the output of the QPF product.
(4) And generating a comparison check product of the QPF product based on the Fleet constrained optical flow method and the cross-correlation QPF, and saving the output of the comparison check product.
The product comprehensive display system construction module based on the Fleet constraint optical flow method comprises the following steps: the system is used for comprehensively applying a GIS technology and an information visualization technology, constructing a product comprehensive display system based on a fly constraint optical flow method based on WEBGIS, completing the comprehensive display of the fly optical flow method product, and integrating the product comprehensive display system into a nowcasting decision support platform; the product comprehensive display system based on the Fleet constrained optical flow method is based on a B/S framework, can dynamically display various quantitative estimation forecast, echo live and forecast and other contents based on the Fleet optical flow method on an electronic map in a color spot map mode, and provides decision reference for a forecaster. Besides real-time comprehensive display of various types of information, the comprehensive display function module supports the historical playback function of various types of information, and various types of information can be retrieved and checked based on time indexes.
The short-term aggregate precipitation probability forecast result calculation module: the method is used for taking the generated Fleet optical flow method product as a set member, respectively adopting an averaging method, a correlation method and a Rank method to calculate the short-term aggregate precipitation probability forecasting result, and comprehensively displaying the short-term aggregate precipitation probability forecasting result on a system platform; based on the nonlinearity and complexity of an atmospheric system and the unavoidable uncertainty of an initial value and a mode, the current ensemble prediction theory is successfully applied to medium-short term numerical weather prediction and short-term climate prediction, permeates into various space and time scales of atmospheric science, and explores the application of the ensemble prediction theory in short-term prediction. According to the method and the model, the ensemble probability forecasting method and the ensemble probability forecasting model and the class forecasting model of the strong convection weather are established on the basis of the precipitation probability forecasting and the class forecasting according to the ensemble forecasting theory in the short-time close-to ensemble forecasting system, and the early warning capability of the strong convection weather forecasting such as strong precipitation, hail and the like is improved.
The uncertainty of prediction mainly comes from the uncertainty of the initial state of the atmosphere and the uncertainty of the prediction mode, and the nonlinear characteristic of the atmospheric motion determines that the extremely small error whether from the initial field or the mode is amplified until the larger error is developed in the mode integration process. Ensemble forecasting techniques are currently an effective way to reduce these errors, in which the outcome of each ensemble member represents a possibility of future weather evolution, regardless of how the ensemble member is formed (e.g., by constructing a perturbation field on the basis of the original material that characterizes the error to form the ensemble member, or by selecting a different mode parameterization scheme, varying the values of some physical parameter to form the ensemble member, or taking a different numerical mode as the ensemble member, etc.). The relative probability of a certain weather phenomenon in the future (such as the probability of heavy rain in a certain area) can be obtained by integrating the forecast results of all the set members, and the method has objectivity and quantificationity compared with the conventional probability forecast only by experience or a statistical method, which cannot be realized in the single-mode forecast, so that the method has reference value. Probability forecasting of precipitation does not require forecasting the spatio-temporal distribution of precipitation magnitudes, but rather the probability distribution that occurs when the forecast precipitation intensity reaches a certain magnitude.
According to the method, the light stream method products such as QPF (quick Path fusion) and QPE (quick Path fusion) are used as set members, the data of the set members are unified to a grid with higher resolution by adopting a unified format, unified standard and unified scoring method through analysis or interpolation, the modeling is performed by adopting a cross effective method every day for forecasting, automatic identification and tracking are performed, the forecasting result of the probability of short-term set precipitation is calculated by adopting an averaging method, a correlation method and a Rank method, the result is comprehensively displayed on a system platform, and the forecasting capability of short-term approach is effectively improved. Specifically, the method comprises the following steps:
(1) average method: how to obtain the probability forecast according to the result of each set member, the simplest method is to determine the probability of occurrence of a certain weather phenomenon in the future according to the number of members forecasting the occurrence of the weather phenomenon according to the equal weight of each set member, and if the number of set members is 4, 2 set members forecast that the rainfall reaches the heavy rain in the future for 3h, the probability of the heavy rain in the future for 3h is 50%, which is the most common method at present and called as an averaging method. The weight of each set member in the ensemble forecast is equal to 1/N, N is the number of the set members, and if M set members forecast to generate precipitation of a certain magnitude exist, the probability P of the precipitation of the magnitude in the future is Mx (1/N).
(2) The correlation method comprises the following steps: the method combines a statistical method on the basis of ensemble prediction, and according to accumulated ensemble prediction results of a certain number of times and corresponding actual precipitation conditions, the correlation between the prediction of each ensemble member and the actual precipitation is calculated, and r is assumed to beiThe correlation coefficient between the forecast and the live precipitation of the ith set member is the weight of the set member in the set forecast as follows:
if there are M members of the set forecasted to generate a certain magnitude of precipitation, the probability of the generation of the magnitude of precipitation in the future is p, which is the sum of the weights of the M members of the set.
(3) RANK method: the RANK distribution (RANK) of all the members in the set is obtained by statistics of set forecast results for a certain number of times, the probability of a certain weather phenomenon in the future is determined according to the RANK distribution and the forecast results of all the members in the set, and test results show that the forecast effect of the method is better than that of an average method, and the method is called as the RANK method. The RANK method is also a method combining aggregation and statistics. The N set members are arranged in the order of the forecast results from small to large to form N +1 levels (RANK), and the corresponding live events can be located in any level of the N +1 levels. When the forecast results of the live event and all the members of the set are different, it is simple to determine the level of the live event, for example, the forecast precipitation (from small to large) of 8 members of the set is 8,10,11,16,20,22 and 25mm, and if the precipitation of the live event is 21mm, the live event is at the 7 th level of 9 levels. When the forecast results of the live condition and the members in the set are the same, the level of the live condition can be determined only by carrying out technical processing, namely, small-magnitude random quantity is added or subtracted on the forecast results and the live condition values of all the members in the set, the processing does not influence the probability forecast results, the set forecast results are accumulated for a certain number of times, and the results are applied to the level of the live condition corresponding to each forecast. And counting the probability of the live situation at each level, namely the level distribution, and obtaining the precipitation probability forecast according to the forecast result of each member after the distribution exists.
N +1 level distributions, then V occurs in the future1≤V＜V2The precipitation probability P is:
when V is1＝Vi-1,V2When the value is Xi, P is Ri;
when V is1＝0,V2When the number is equal to Xi,
when V is1＝0,0＜V2＜X1When the temperature of the water is higher than the set temperature,
when V is1＝Xi,Xi＜V2≤Xi+1When the temperature of the water is higher than the set temperature,
when Xi < V1≤Xi+1,V2＝Xi+1When the temperature of the water is higher than the set temperature,
generally, the greater the precipitation, the less likely it will occur, and thus more than X will occurNThe distribution of the precipitation probability is not reasonable, the RANK method assumes that the distribution follows Gumbel distribution, F is a Gumbel cumulative distribution function, and then:
when V is1＝XNWhen the temperature of the water is higher than the set temperature,
when V is1≥XNWhen the temperature of the water is higher than the set temperature,
both the correlation method and the RANK method need to accumulate the ensemble prediction results for a certain number of times as statistical samples, and in order to ensure that a certain amount of samples exist, the cross validation technology is adopted, namely, when testing each case, the ensemble prediction results of the other cases are used as the statistical samples.
According to the short-time heavy rainfall forecasting method and system, a brand-new forecasting system based on the Fleet constrained optical flow method is built, the problem of echo extrapolation forecasting of local generation and rapid change of intensity and shape of a rainfall cloud cluster over time is solved, and the performance of a proximity forecasting system for short-time heavy rainfall weather is improved. The method is characterized in that a method combining radar image morphology, characteristic quantity characteristic statistics, pattern recognition and the like is adopted to establish an automatic recognition and approach forecast early warning technology of short-time strong rainfall weather, and an ensemble probability forecast and grade forecast method and a model of the strong convection weather are established on the basis of rainfall probability forecast and grade forecast according to an ensemble forecast theory in a short-time approach ensemble forecast system, so that the early warning capability of the strong convection weather forecast of strong rainfall, hail and the like is improved.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (4)
1. A short-term heavy rainfall forecasting method is characterized by comprising the following steps:
step a: reading radar base data, and performing radar echo quality control processing on the radar base data;
step b: constructing a forecasting system based on a Fleet constraint optical flow method according to the radar base data after the radar echo quality control processing;
step c: constructing an identification tracking system based on the Fleet constrained optical flow method according to the forecasting system based on the Fleet constrained optical flow method, and analyzing data and generating and outputting a Fleet optical flow method product;
in the step a, the radar echo quality control processing on the radar base data specifically comprises weighted gaussian filtering, a morphological analysis expansion algorithm, a morphological analysis erosion algorithm, background subtraction of radar data and radar data regularization processing;
in the step b, the method based on the Fleet constrained optical flow comprises Fleet backward differential extrapolation, Fleet advection extrapolation, cross correlation backward differential extrapolation and cross correlation Lagrange advection extrapolation; the specific algorithm of the Fleet constrained optical flow method comprises the following steps:
step b 1: solving a complex deconvolution of a radar live field according to a motion calculation and estimation theory of a Fleet and Jepson image component;
step b 2: applying a core algorithm for mapping a color point noise field to an optical flow field, performing dynamic color optical flow field and 2D and 3D dynamic color mapping, and dynamically mapping the optical flow field to a 6-channel basic color space with three coordinate axes;
step b 3: converting the optical flow field into a core algorithm based on a color point noise format, and performing two-dimensional display on two-dimensional optical flow fields with different formats;
step b 4: median filtering of radar echo live;
in the step c, the generation and output of the Fleet optical flow method product specifically include:
step c 1: generating a radar echo extrapolation product based on a Fleet constraint optical flow method, and outputting and storing the product;
step c 2: combining Kalman filtering and CRESMAN objective analysis to generate a QPE product based on a Fleet optical flow method, and outputting and storing the QPE product;
step c 3: combining Kalman filtering and CRESMAN objective analysis to generate a QPF product based on a Fleet optical flow method, and outputting and storing the QPF product;
step c 4: generating a QPF product based on a Fleet constraint optical flow method and a cross-correlation QPF comparison inspection product, and outputting and storing the comparison inspection product;
the method comprises the steps of adopting QPF and QPE, namely the Fleet optical flow method products as set members, adopting a uniform format, a uniform standard and a uniform scoring method for data of the set members, unifying the data on a grid with higher resolution through analysis or interpolation, forecasting by adopting a cross effective method every day, automatically identifying and tracking, calculating a forecasting result of the short-term set precipitation probability by adopting an averaging method, a correlation method and a Rank method, and comprehensively displaying the result on a system platform.
2. The method for forecasting short-term heavy rainfall according to claim 1, further comprising after step c: comprehensively applying a GIS technology and an information visualization technology, constructing a product comprehensive display system based on a Fleet constraint optical flow method based on WEBGIS, and comprehensively displaying the Fleet constraint optical flow method product; and taking the generated Fleet constraint optical flow method product as a set member, respectively adopting an average method, a correlation method and a Rank method to calculate the short-term aggregation rainfall probability forecasting result, and comprehensively displaying the short-term aggregation rainfall probability forecasting result on a system platform.
3. A short-term heavy rainfall forecasting system, comprising:
the forecasting system construction module based on the Fleet constraint optical flow method comprises the following steps: the forecasting system is used for reading radar base data, performing radar echo quality control processing on the radar base data, and constructing a forecasting system based on a Fleet constraint optical flow method according to the radar base data after the radar echo quality control processing;
the method comprises the following steps of identifying and tracking a system construction module based on a Fleet constraint optical flow method: the forecasting system based on the fly constraint optical flow method is used for constructing an identification tracking system based on the fly constraint optical flow method according to the forecasting system based on the fly constraint optical flow method, and analyzing data and generating and outputting a product based on the fly constraint optical flow method;
the method comprises the steps that a light stream method forecasting system building module based on the Fleet constraint performs radar echo quality control processing on radar base data, wherein the radar echo quality control processing specifically comprises weighted Gaussian filtering, a morphological analysis expansion algorithm, a morphological analysis erosion algorithm, background subtraction of radar data and radar data regularization processing;
the method comprises the steps of carrying out Fleet backward differential extrapolation, Fleet advection extrapolation, cross correlation backward differential extrapolation and cross correlation Lagrange advection extrapolation on the basis of a Fleet constraint optical flow method; the specific algorithm of the Fleet constrained optical flow method comprises the following steps: solving a complex deconvolution of a radar live field according to a motion calculation and estimation theory of a Fleet and Jepson image component; applying a core algorithm for mapping a color point noise field to an optical flow field, performing dynamic color optical flow field and 2D and 3D dynamic color mapping, and dynamically mapping the optical flow field to a 6-channel basic color space with three coordinate axes; converting the optical flow field into a core algorithm based on a color point noise format, and performing two-dimensional display on two-dimensional optical flow fields with different formats; median filtering of radar echo live;
the method for generating and outputting the Fleet optical flow method product based on the Fleet constrained optical flow method identification and tracking system construction module specifically comprises the following steps: generating a radar echo extrapolation product based on a Fleet constraint optical flow method, and outputting and storing the product; combining Kalman filtering and CRESMAN objective analysis to generate a QPE product based on a Fleet optical flow method, and outputting and storing the QPE product; combining Kalman filtering and CRESMAN objective analysis to generate a QPF product based on a Fleet optical flow method, and outputting and storing the QPF product; generating a QPF product based on a Fleet constraint optical flow method and a cross-correlation QPF comparison inspection product, and outputting and storing the comparison inspection product; the method comprises the steps of adopting QPF and QPE, namely the Fleet optical flow method products as set members, adopting a uniform format, a uniform standard and a uniform scoring method for data of the set members, unifying the data on a grid with higher resolution through analysis or interpolation, forecasting by adopting a cross effective method every day, automatically identifying and tracking, calculating a forecasting result of the short-term set precipitation probability by adopting an averaging method, a correlation method and a Rank method, and comprehensively displaying the result on a system platform.
4. The system of claim 3, further comprising:
the product comprehensive display system construction module based on the Fleet constraint optical flow method comprises the following steps: the system is used for comprehensively applying a GIS technology and an information visualization technology, constructing a product comprehensive display system based on a fly constraint optical flow method based on WEBGIS, and comprehensively displaying products of the fly constraint optical flow method;
the short-term aggregate precipitation probability forecast result calculation module: and the method is used for calculating the short-term aggregate rainfall probability forecasting results by respectively adopting an averaging method, a correlation method and a Rank method by taking the generated Fleet constrained optical flow method product as an aggregate member, and comprehensively displaying the short-term aggregate rainfall probability forecasting results on a system platform.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810902276.7A CN108646319B (en) | 2018-08-09 | 2018-08-09 | Short-time strong rainfall forecasting method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810902276.7A CN108646319B (en) | 2018-08-09 | 2018-08-09 | Short-time strong rainfall forecasting method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108646319A CN108646319A (en) | 2018-10-12 |
| CN108646319B true CN108646319B (en) | 2021-11-09 |
Family
ID=63760917
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810902276.7A Expired - Fee Related CN108646319B (en) | 2018-08-09 | 2018-08-09 | Short-time strong rainfall forecasting method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108646319B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110888186A (en) * | 2019-11-29 | 2020-03-17 | 天津大学 | Method for forecasting hail and short-time heavy rainfall based on GBDT + LR model |
| CN111145245B (en) * | 2019-12-31 | 2023-07-21 | 哈尔滨工业大学（深圳） | Short-time adjacent rainfall forecasting method, system and computer-readable storage medium |
| CN111638565A (en) * | 2020-06-09 | 2020-09-08 | 云南电网有限责任公司电力科学研究院 | Method for monitoring rainstorm in mountainous area |
| CN112184753A (en) * | 2020-09-21 | 2021-01-05 | 上海眼控科技股份有限公司 | Target optical flow method determination method and device, computer equipment and storage medium |
| CN112363140B (en) * | 2020-11-05 | 2024-04-05 | 南京叁云科技有限公司 | Thermodynamic constraint extrapolation objective correction method based on cyclic neural network |
| CN113360854B (en) * | 2021-08-10 | 2021-11-05 | 中国人民解放军国防科技大学 | Data assimilation method based on adaptive covariance expansion |
| CN117518299B (en) * | 2024-01-05 | 2024-03-22 | 南京大学 | Classified strong convection proximity probability forecasting method, system, equipment and terminal |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003288603A (en) * | 2002-03-28 | 2003-10-10 | Nippon Telegr & Teleph Corp <Ntt> | Method, device, program, and recording medium for image scene prediction |
| CN103136596A (en) * | 2013-01-25 | 2013-06-05 | 成都信息工程学院 | Mountain flood disaster monitoring and early-warning business platform system in medium or small rivers |
| CN104977584A (en) * | 2015-06-29 | 2015-10-14 | 深圳市气象台 | Convective weather approach prediction method and system |
| CN105842756A (en) * | 2016-03-30 | 2016-08-10 | 广州市积雨云信息科技有限公司 | Method for forecasting thunderstorm information |
| CN107015229A (en) * | 2017-05-22 | 2017-08-04 | 新疆维吾尔自治区人工影响天气办公室 | Artificial Hail Suppression operation command method based on dual-polarization weather radar |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103675791B (en) * | 2013-12-05 | 2015-12-30 | 北京师范大学 | Based on the Mie scattering lidar cloud recognition methods of numeric distribution equalization |
| CN103745216B (en) * | 2014-01-02 | 2016-10-26 | 中国民航科学技术研究院 | A kind of radar image clutter suppression method based on Spatial characteristic |
| CN107728115B (en) * | 2017-09-11 | 2020-08-11 | 电子科技大学 | SVM-based background interference suppression method after radar target imaging |
-
2018
- 2018-08-09 CN CN201810902276.7A patent/CN108646319B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003288603A (en) * | 2002-03-28 | 2003-10-10 | Nippon Telegr & Teleph Corp <Ntt> | Method, device, program, and recording medium for image scene prediction |
| CN103136596A (en) * | 2013-01-25 | 2013-06-05 | 成都信息工程学院 | Mountain flood disaster monitoring and early-warning business platform system in medium or small rivers |
| CN104977584A (en) * | 2015-06-29 | 2015-10-14 | 深圳市气象台 | Convective weather approach prediction method and system |
| CN105842756A (en) * | 2016-03-30 | 2016-08-10 | 广州市积雨云信息科技有限公司 | Method for forecasting thunderstorm information |
| CN107015229A (en) * | 2017-05-22 | 2017-08-04 | 新疆维吾尔自治区人工影响天气办公室 | Artificial Hail Suppression operation command method based on dual-polarization weather radar |
Non-Patent Citations (2)
| Title |
|---|
| 短期集合降水概率预报试验;王晨稀;《应用气象学报》;20050228;第16卷(第1期);第78-81页 * |
| 精细化预报预警平台在深圳的试验应用;兰红平 等;《气象科技进展》;20131231;第3卷(第6期);第22页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108646319A (en) | 2018-10-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108646319B (en) | Short-time strong rainfall forecasting method and system | |
| CN112070286B (en) | Precipitation forecast and early warning system for complex terrain river basin | |
| Mecklenburg et al. | Improving the nowcasting of precipitation in an Alpine region with an enhanced radar echo tracking algorithm | |
| Cyriac et al. | Variability in Coastal Flooding predictions due to forecast errors during Hurricane Arthur | |
| Antolini et al. | A daily high-resolution gridded climatic data set for Emilia-Romagna, Italy, during 1961-2010. | |
| Hardy et al. | A method for probabilistic flash flood forecasting | |
| Goudenhoofdt et al. | Generation and verification of rainfall estimates from 10-yr volumetric weather radar measurements | |
| Liang et al. | A composite approach of radar echo extrapolation based on TREC vectors in combination with model-predicted winds | |
| CN103529492B (en) | Based on storm body position and the form prediction method of Doppler radar reflectivity image | |
| CN112764129A (en) | Method, system and terminal for thunderstorm short-term forecasting | |
| CN113344291B (en) | Urban inland inundation range forecasting method, device, medium and equipment | |
| Friedland et al. | Isotropic and anisotropic kriging approaches for interpolating surface-level wind speeds across large, geographically diverse regions | |
| CN109407177B (en) | Machine learning and conventional meteorological observation-based fog identification system and application method | |
| Squitieri et al. | On the forecast sensitivity of MCS cold pools and related features to horizontal grid spacing in convection-allowing WRF simulations | |
| Gaughan et al. | Evaluating nighttime lights and population distribution as proxies for mapping anthropogenic CO2 emission in Vietnam, Cambodia and Laos | |
| Bernet et al. | Characterizing precipitation events leading to surface water flood damage over large regions of complex terrain | |
| CN115201938B (en) | Strong convection weather proximity forecasting method and system based on thunderstorm high-voltage analysis | |
| CN115691049A (en) | Convection birth early warning method based on deep learning | |
| Wetterhall et al. | Effects of temporal resolution of input precipitation on the performance of hydrological forecasting | |
| Jessup et al. | Organization of flash-flood-producing precipitation in the northeast United States | |
| CN111340668A (en) | Typhoon disaster evaluation system | |
| CN113075751A (en) | Method and system for fusing observation data in short-term forecasting | |
| CN112379467A (en) | Method and system for fusing weather live situations of mobile terminal based on position | |
| Case et al. | An objective technique for verifying sea breezes in high-resolution numerical weather prediction models | |
| CN116679355A (en) | Precipitation prediction correction method based on cascade Attention-U-Net |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20211109 |