A method and device for diagnosing uncertainty in weather forecast
Through the sensitivity analysis of ensemble forecasts and the Tubing clustering algorithm, the sensitive areas of weather forecasts are identified and classified, and the problem of insufficient automated diagnosis in the prior art is solved, and the accuracy and reliability of weather forecasts are improved.
Patent Information
- Application Number
- CN202410283458.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-03-13
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-03-13
AI Technical Summary
In the prior art, weather forecast uncertainty diagnosis relies on manual intervention, the degree of automation is low and the diagnostic accuracy and objectivity need to be improved.
Sensitivity analysis and Tubing clustering algorithm based on ensemble forecasts are used to determine key sensitive areas and classify members. By comparing central and Tubes members, the physical process leading to the development of uncertainty in weather forecasts is diagnosed.
It realizes automated diagnosis of weather forecasts, improves the accuracy and reliability of forecasts, and identifies sensitive areas in real time and classifies them, improving the accuracy of forecasts.
Smart Images

Figure CN118260666B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of information processing technology, and in particular to a method and device for diagnosing uncertainty in weather forecasts. Background Art
[0002] Ensemble forecast is a multi-member forecast constructed from physical and numerical aspects through multi-model, multi-disturbance, multi-forcing and other methods. Unlike single forecast, ensemble forecast provides an average forecast state while also providing the uncertainty range or probability forecast of the forecast. The difference in ensemble members affects the results of weather forecast.
[0003] In related technologies, the diagnostic method for ensemble forecast uncertainty relies on human intervention, has a low degree of automation, and the diagnostic accuracy and objectivity need to be further improved. Summary of the invention
[0004] The present application provides a method and device for weather forecast uncertainty diagnosis to solve the problems existing in the related art.
[0005] In a first aspect, the present invention provides a method for diagnosing uncertainty in weather forecasts, comprising determining key sensitive areas based on forecast variables and state variables of members in an ensemble forecast using preset ensemble sensitivity analysis rules; classifying members of the key sensitive areas based on a Tubing clustering algorithm to obtain central class members and Tubes class members; and comparing extreme members of the central class members and the Tubes class members to diagnose the physical process that leads to the development of weather forecast uncertainty based on the comparison results.
[0006] Optionally, based on the forecast variables and state variables of the members in the ensemble forecast, determining the key sensitive areas using preset ensemble sensitivity analysis rules includes: obtaining from a data source the forecast variables of the ensemble forecast members in the preset area at the first preset time, and the state variables of each grid point at the second preset time under the model forecast method; performing a sensitivity analysis based on the linear correlation between the forecast variables and the state variables; determining the connected domain that passes the significance test in the sensitivity analysis results based on a connected domain identification algorithm; and determining the key sensitive areas based on preset parameters in the connected domain.
[0007] Optionally, classifying the members of the sensitive area based on the Tubing clustering algorithm to obtain central class members and Tubes class members includes: determining the distance between each member and the set average of the forecast calculation set; and clustering based on the distance to classify the central class members and Tubes class members.
[0008] Optionally, clustering is determined based on the distance to classify central class members, including: sorting the distances from small to large, and classifying the two members with the smallest distances as central class members; based on the relationship between the variance of the distances between the central class members and the total variance, continuously classifying the member with the smallest current distance into the central class members until the variance of the distances between the central class members and the total variance satisfy a preset relationship, wherein the distance between the central class member finally classified and the set average is used as the clustering radius.
[0009] Optionally, the clustering determined based on the distance to classify the Tubes class members includes: for a set of members that fall outside the central class members, classifying the members with the largest distance as extreme members of the tube class, that is, the tube ends of the tube class; using the line connecting the tube ends of the tube class and the set average as the axis and the cluster radius as the radius to obtain a tube column; judging whether the members fall into the tube column in descending order of the distance; if they do not fall into the tube column, defining the corresponding members as extreme members of the new tube class, determining a new tube column based on the extreme members of the new tube class and the set average, and judging whether the remaining members fall into the new tube column; iterating the above steps until all remaining members participate in the classification.
[0010] Optionally, the central class members and extreme members of the Tubes class members are compared to diagnose the physical process leading to the development of weather forecast uncertainty based on the comparison results, including: calculating the member average value of the obtained central class members; comparing the extreme members of the Tubes class members with the member average value of the central class members to diagnose the physical process leading to the development of weather forecast uncertainty.
[0011] In a second aspect, the present invention provides a device for diagnosing uncertainty in weather forecasts, comprising a key sensitive area determination unit, configured to determine key sensitive areas based on forecast variables and state variables of members in an ensemble forecast using preset ensemble sensitivity analysis rules; a clustering unit, configured to classify members of the sensitive areas based on a Tubing clustering algorithm to obtain central class members and Tubes class members; and a diagnosis unit, configured to compare extreme members of the central class members and the Tubes class members to diagnose the physical process that leads to the development of weather forecast uncertainty based on the comparison results.
[0012] Optionally, the key sensitive area is configured to: obtain from a data source the forecast variables of the ensemble forecast members in the preset area at the first preset time, and the state variables of each grid point at the second preset time under the model forecasting method; perform a sensitivity analysis based on the linear correlation between the forecast variables and the state variables; determine the connected domain that passes the significance test in the sensitivity analysis results based on a connected domain identification algorithm; and determine the key sensitive area based on preset parameters within the connected domain.
[0013] In a third aspect, the present invention provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the method described in any implementation of the first aspect is implemented.
[0014] In a fourth aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method provided in any implementation of the first aspect when executing the program.
[0015] The present invention discloses a method and device for diagnosing uncertainty in weather forecasts, wherein the method includes determining key sensitive areas based on forecast variables and state variables of members in an ensemble forecast using preset ensemble sensitivity analysis rules; classifying members of the key sensitive areas based on a Tubing clustering algorithm to obtain central class members and Tubes class members; and comparing extreme members of the central class members and the Tubes class members to diagnose the physical process that causes the development of weather forecast uncertainty based on the comparison results. The ensemble sensitivity analysis and Tubing clustering analysis techniques are fully combined to identify sensitive areas of weather forecasts in real time and perform forecast classification, more accurately diagnose sources of uncertainty in weather forecasts, and thus improve the accuracy and reliability of forecasts. The defect in the related art that the degree of automation of weather forecast uncertainty diagnosis needs to be improved is overcome. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0017] Figure 1 A flow chart of the method for weather forecast uncertainty diagnosis used in this application;
[0018] Figure 2 A method corresponding to the present application Figure 1 Schematic diagram of an electronic device. DETAILED DESCRIPTION
[0019] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.
[0020] The following is combined with Figure 1 The scenario shown is an exemplary description of the method for weather forecast uncertainty diagnosis in the present application. The method comprises the following steps:
[0021] Step 101: Based on the forecast variables and state variables of the members in the ensemble forecast, the key sensitive areas are determined using the preset ensemble sensitivity analysis rules.
[0022] In this embodiment, the forecast variables P of the n members of the ensemble forecast at the preset time t1 and the preset area are intercepted, and the state variables X of each grid point at the preset time t2 are obtained from the relevant data source. According to the linear correlation between the forecast variables of the ensemble members and the state variables, sensitivity correlation analysis is performed. The state variables may include dynamic factors (such as the potential height on the isobaric surface, zonal wind speed, meridional wind speed, vorticity, divergence field, etc.), thermal factors (such as the temperature on the isobaric surface, pseudo equivalent potential temperature, etc.), and water vapor factors (such as specific humidity, relative humidity, water vapor flux, water vapor flux divergence, atmospheric precipitable water, etc. on the isobaric surface).
[0023] As an optional implementation method of this embodiment, based on the forecast variables and state variables of the members in the ensemble forecast, determining the key sensitive area using the preset ensemble sensitivity analysis rules includes: obtaining from the data source the forecast variables of the ensemble forecast members in the preset area at the first preset time, and the state variables of each grid point at the second preset time under the model forecast method; performing a sensitivity analysis based on the linear correlation between the forecast variables and the state variables; determining the connected domain that passes the significance test in the sensitivity analysis result based on the connected domain identification algorithm; and determining the key sensitive area based on the preset parameters in the connected domain.
[0024] In this optional implementation, sensitivity correlation analysis is performed based on the linear correlation between the predicted variables and the state variables of the set members. The linear correlation between the regional average predicted variable P and the state variable X at each grid point can be represented by the Pearson correlation coefficient between the two, where n represents the number of set members:
[0025]
[0026] According to this formula, the spatial distribution of the Pearson correlation coefficient of all grid points on the prediction variable P and the state variable X field can be calculated.
[0027] Furthermore, a confidence level of 95% is used as the standard for passing the significance test, and a connected domain recognition algorithm is used to identify connected domains that pass the significance test in the Pearson correlation coefficient field as sensitive areas based on the connectivity between adjacent pixels or grid points in the spatial data. The longitude and latitude information of the circumscribed matrix of each connected domain is extracted and stored in the memory.
[0028] Furthermore, based on the external matrix information of the identified connected domain, the absolute maximum value of the Pearson correlation coefficient in each connected domain is calculated, and the connected domain where the maximum absolute value of the Pearson correlation coefficient is located is identified as the key sensitive area of the state variable to the predicted variable.
[0029] Based on the above method, this embodiment realizes real-time diagnosis of weather forecast sensitive areas and improves the ability to respond to sudden meteorological events.
[0030] Step 102: Classify the members of the key sensitive area based on the Tubing clustering algorithm to obtain central class members and Tubes class members.
[0031] In this embodiment, the Tubing clustering algorithm is used to classify the members according to the discrete degree of the distance between the state variables of different members and the ensemble average value in the key sensitive area, and obtain the central class members and Tubes members.
[0032] As an optional implementation method of this embodiment, the members of the sensitive area are classified based on the Tubing clustering algorithm to obtain central class members and Tubes class members, including: determining the distance between each member and the set average of the forecast calculation set; and clustering is determined based on the distance to classify the central class members and Tubes class members.
[0033] As an optional implementation method of this embodiment, the clustering determined based on the distance to classify the central class members includes: sorting the distances from small to large, and classifying the two members with the smallest distances as central class members; based on the relationship between the variance of the distances between the central class members and the total variance, continuously classifying the members with the current smallest distance into the central class members until the variance of the distances between the central class members and the total variance meet a preset relationship, wherein the distance between the last central class member and the set average is used as the clustering radius.
[0034] In the above optional implementation, assuming that the number of members is n, the set sample is represented by f = {f1, f2, ..., f nFirst, the set mean is calculated, and then the distances dm = {dm(1), dm(2), …, dm(n)} between different members and the set mean are calculated.
[0035] Furthermore, dm is sorted from small to large, and the two members with the smallest distance to the set average are first classified as central class members. The variance between the distances of the central class members is calculated. When the variance is less than 50% of the total variance, the member with the closest distance is further classified as the central class. When the central class variance is not less than 50% of the total variance, the increase of central class members is stopped. The distance dm(i) between the member i that enters the central class last and the set average is defined as the cluster radius r, where the central class radius is the same as the tube class radius, both r.
[0036] As an optional implementation method of this embodiment, clustering is determined based on the distance to classify Tubes class members, including: for a set of members that fall outside the central class members, the members with the largest distance are classified as extreme members of the tube class, that is, the tube ends of the tube class; a line connecting the tube ends of the tube class and the set average value is used as the axis and the cluster radius is used as the radius to obtain a tube column; judging whether a member falls into the tube column in descending order of the distance; if the member does not fall into the tube column, defining the corresponding member as an extreme member of a new tube class, determining a new tube column based on the extreme members of the new tube class and the set average value, and judging whether the remaining members fall into the new tube column; iterating the above steps until all remaining members participate in the classification.
[0037] In this optional implementation, a set of members falling outside the central class is obtained, and the member farthest from the set average (i.e., the member corresponding to the maximum dm value) is classified as the extreme member of the first pipe class, i.e., the pipe end of the first pipe class, and a pipe string with a radius of r and the line connecting the set average and the position of the first pipe end as the axis is obtained. And according to the order of dm values from large to small, it is judged whether the member falls into the known pipe string. If not, it is defined as an extreme member of a new pipe class. Assuming that the extreme member of the known pipe string is j, the basis for judging whether the set member i falls into the pipe string defined by member j is as follows:
[0038]
[0039] dx 2 (i,j)=dm 2 (i)-dd 2 (i,j)
[0040] Wherein, dm(i) and dm(j) respectively represent the distances between the set members i and j and the set average value, d(i,j) represents the distance between the members i and j, dd(i,j) represents the projection of the member i on the axis determined by the member j and the set average value, and dx(i,j) represents the distance from the member i to the axis determined by the member j and the set average value. If both dx(i,j) < r and dd(i,j) > 0 are satisfied, then the member i falls into the pipe column determined by the member j and is a member of this pipe type; otherwise, i is an extreme member of a new pipe type. Continuously iterate this step until all the remaining members have participated in the classification.
[0041] Step 103: Compare the extreme members among the central class members and the Tubes class members to diagnose the physical processes that lead to the development of weather forecast uncertainty based on the comparison results.
[0042] As an optional implementation manner of this embodiment, comparing the extreme members among the central class members and the Tubes class members to diagnose the physical processes that lead to the development of weather forecast uncertainty includes: calculating the member average value of the obtained central class members; comparing the extreme members among the Tubes class members with the member average value of the central class members to diagnose the physical processes of the development of weather forecast uncertainty.
[0043] In this embodiment, the central class members are obtained by clustering Tubing and the member average calculation is performed. The obtained result represents the predictable component of the weather forecast, and the most extreme members of the Tubes class represent the uncertainty information of the weather system characterized by this type of Tube. By comparing the differences between the extreme members of each type of Tube and the central class average, the key physical processes that lead to the development of weather forecast uncertainty in the preset area are diagnosed.
[0044] This embodiment fully combines the ensemble sensitivity analysis and Tubing clustering analysis techniques to real-time identify the sensitive areas of weather forecasts and perform forecast typing, more accurately diagnose the sources of weather forecast uncertainty, and thus improve the accuracy and reliability of the forecast.
[0045] The above is the method provided by one or more embodiments of the present application. Based on the same idea, the present application also provides a corresponding device for diagnosing weather forecast uncertainty, including a key sensitive area determination unit configured to determine the key sensitive area based on the forecast variables and state variables of the members in the ensemble forecast and using the preset ensemble sensitivity analysis rules;
[0046] A clustering unit configured to classify the members in the sensitive area based on the Tubing clustering algorithm to obtain central class members and Tubes class members;
[0047] The diagnosis unit is configured to compare the extreme members of the central class members and the Tubes class members to diagnose the physical process that causes the development of weather forecast uncertainty based on the comparison result.
[0048] As an optional implementation method of this embodiment, the key sensitive area is configured as follows: obtaining from a data source the forecast variables of the ensemble forecast members in the preset area at the first preset time, and the state variables of each grid point at the second preset time under the model forecast method; performing a sensitivity analysis based on the linear correlation between the forecast variables and the state variables; determining the connected domain that passes the significance test in the sensitivity analysis results based on a connected domain identification algorithm; and determining the key sensitive area based on preset parameters within the connected domain.
[0049] As an optional implementation method of this embodiment, the members of the sensitive area are classified based on the Tubing clustering algorithm to obtain central class members and Tubes class members, including: determining the distance between each member and the set average of the forecast calculation set; and clustering is determined based on the distance to classify the central class members and Tubes class members.
[0050] As an optional implementation method of this embodiment, the clustering determined based on the distance to classify the central class members includes: sorting the distances from small to large, and classifying the two members with the smallest distances as central class members; based on the relationship between the variance of the distances between the central class members and the total variance, continuously classifying the members with the current smallest distance into the central class members until the variance of the distances between the central class members and the total variance meet a preset relationship, wherein the distance between the last central class member and the set average is used as the clustering radius.
[0051] As an optional implementation method of this embodiment, the clustering determined based on the distance to classify the Tubes class members includes: for a set of members that fall outside the central class members, classifying the members with the largest distance as extreme members of the tube class, that is, the tube ends of the tube class; using the line connecting the tube ends of the tube class and the set average as the axis and the cluster radius as the radius to obtain a tube column; judging whether the members fall into the tube column in descending order of the distance; if they do not fall into the tube column, defining the corresponding members as extreme members of the new tube class, determining a new tube column based on the extreme members of the new tube class and the set average, and judging whether the remaining members fall into the new tube column; iterating the above steps until all remaining members participate in the classification.
[0052] As an optional implementation method of this embodiment, the central class members and the extreme members of the Tubes class members are compared to diagnose the physical process that causes the development of weather forecast uncertainty based on the comparison results, including: calculating the member average value of the obtained central class members; comparing the extreme members of the Tubes class members with the member average value of the central class members to diagnose the physical process that causes the development of weather forecast uncertainty.
[0053] The present application also provides a computer readable medium, which stores a computer program, which can be used to execute the above Figure 1 Based on the method provided.
[0054] Figure 2 A schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as Figure 2 As shown, the electronic device 50 includes: a processor 501 (processor), a memory 502 (memory) and a bus 503;
[0055] The processor 501 and the memory 502 communicate with each other via the bus 503;
[0056] The processor 501 is used to call the program instructions in the memory 502 to execute the methods provided by the above-mentioned method embodiments.
[0057] This embodiment provides a non-transitory computer-readable storage medium, which stores computer instructions. The computer instructions enable a computer to execute the methods provided by the above method embodiments.
[0058] A person skilled in the art can understand that all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk, etc., various storage media that can store program codes.
[0059] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative effort.
[0060] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiment.
[0061] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.
Claims
1. A method for weather forecast uncertainty diagnosis, characterized in that: include: Based on the forecast variables and state variables of the members in the ensemble forecast, the key sensitive areas are determined using the preset ensemble sensitivity analysis rules, including: obtaining the forecast variables of the ensemble forecast members in the preset area at the first preset time from the data source, and the state variables of each grid point at the second preset time under the model forecast mode; performing sensitivity analysis based on the linear correlation between the forecast variables and the state variables; determining the connected domain that passes the significance test in the sensitivity analysis results based on the connected domain identification algorithm; and determining the key sensitive area based on the preset parameters in the connected domain; Classifying the members of the key sensitive area based on the Tubing clustering algorithm to obtain central class members and Tubes class members, including: determining the distance between each member and the set average of the set; clustering based on the distance to classify the central class members and Tubes class members, Comparing the central class members and the extreme members in the Tubes class members to diagnose the physical process leading to the development of weather forecast uncertainty based on the comparison results, including: calculating the member average value of the obtained central class members; comparing the extreme members in the Tubes class members with the member average value of the central class members to diagnose the physical process leading to the development of weather forecast uncertainty; Wherein, the clustering is determined based on the distance to classify the central class members, including: sorting the distances from small to large, and classifying the two members with the smallest distances as central class members; based on the relationship between the variance of the distances between the central class members and the total variance, continuously classifying the member with the smallest current distance into the central class members, until the variance of the distances between the central class members and the total variance meet the preset relationship, wherein the distance between the central class members and the set average is finally used as the clustering radius; Clustering is determined based on the distance to classify the Tubes class members, including: for a set of members that fall outside the central class members, the members with the largest distance are classified as extreme members of the tube class, that is, the tube ends of the tube class; a line connecting the tube ends of the tube class and the set average value is used as the axis and the cluster radius is used as the radius to obtain a tube column; judging whether a member falls into the tube column in descending order of the distance; if the member does not fall into the tube column, the corresponding member is defined as an extreme member of a new tube class, and a new tube column is determined based on the extreme members of the new tube class and the set average value, and judging whether the remaining members fall into the new tube column; iterating the above steps until all remaining members participate in the classification.
2. A device for diagnosing uncertainty in weather forecasts, characterized in that: include: The key sensitive area determination unit is configured to determine the key sensitive area based on the forecast variables and state variables of the members in the ensemble forecast using a preset ensemble sensitivity analysis rule, including: obtaining from a data source the forecast variables of the ensemble forecast members in the preset area at the first preset time, and the state variables of each grid point at the second preset time in the model forecast mode; performing sensitivity analysis based on the linear correlation between the forecast variables and the state variables; determining the connected domain that passes the significance test in the sensitivity analysis result based on a connected domain identification algorithm; and determining the key sensitive area based on preset parameters in the connected domain; The clustering unit is configured to classify the members of the sensitive area based on the Tubing clustering algorithm to obtain central class members and Tubes class members, including: determining the distance between each member and the set average of the set; and performing clustering based on the distance to classify the central class members and Tubes class members; The diagnosis unit is configured to compare the central class members and the extreme members in the Tubes class members to diagnose the physical process that causes the development of weather forecast uncertainty based on the comparison result, including: calculating the member average value of the obtained central class members; comparing the extreme members in the Tubes class members with the member average value of the central class members to diagnose the physical process that causes the development of weather forecast uncertainty; Wherein, the clustering is determined based on the distance to classify the central class members, including: sorting the distances from small to large, and classifying the two members with the smallest distances as central class members; based on the relationship between the variance of the distances between the central class members and the total variance, continuously classifying the member with the smallest current distance into the central class members, until the variance of the distances between the central class members and the total variance meet the preset relationship, wherein the distance between the central class members and the set average is finally used as the clustering radius; Clustering is determined based on the distance to classify the Tubes class members, including: for the member set falling outside the central class members, the member with the largest distance is classified as the extreme member of the tube class, that is, the tube end of the tube class; the line connecting the tube end of the tube class and the set average value is used as the axis and the cluster radius is used as the radius to obtain the tube column; according to the order of the distance from large to small, whether the member falls into the tube column is judged; if it does not fall into the tube column, the corresponding member is defined as the extreme member of the new tube class, and a new tube column is determined based on the extreme member of the new tube class and the set average value, and whether the remaining members fall into the new tube column is judged; iterate the above steps until all remaining members participate in the classification.
3. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to claim 1 is implemented.
4. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method according to claim 1 is implemented.
Citation Information
Patent Citations
A method and system for selecting members of a weather forecast set
CN109325633A
A hierarchical clustering method for set numerical weather forecast members
CN110245692A
Typhoon multivariable set sensitivity analysis method, typhoon forecasting method and typhoon forecasting system
CN110472781A
GRAPES-REPS region ensemble forecasting method and system
CN114417623A
Minimizing Uncertainty Envelopes in Trajectories of Evolving Ensemble Members
US20150310137A1
