Comprehensive evaluation method and device for climate mode prediction data and storage medium

Abstract

The invention relates to the technical field of weather numerical mode / climate mode prediction and evaluation, in particular to a comprehensive evaluation method and device for climate mode prediction data and a storage medium. The method comprises the following steps: acquiring prediction data and corresponding observation data of different climate modes; calculating evaluation indexes of different climate modes based on the prediction data and the observation data, wherein the evaluation indexes comprise deviation, a spatial correlation coefficient, a root-mean-square error, a time correlation coefficient and a TS score; calculating the weight of each evaluation index according to the information entropy of each evaluation index; and according to the weighted sum of each evaluation index and the corresponding weight, determining a comprehensive evaluation result of the different climate modes. Therefore, according to the method, mode prediction precision, uncertainty estimation, prediction deviation and extreme event prediction capability are evaluated from time and space, one-sidedness of a single index is avoided, and an evaluation result is more objective and credible.

Description

Technical Field

[0001] The present invention relates to the technical field of meteorological numerical model / climate model prediction and evaluation, and in particular to a comprehensive evaluation method, device and storage medium for climate model prediction data. Background Art

[0002] Climate models are complex computational models used to simulate and predict the behavior of the Earth's climate system. With the deepening of climate research and the advancement of computer technology, climate models have become one of the important tools for understanding and predicting climate change. Climate models can simulate the interactions between the atmosphere, ocean, land, and ice and snow systems, and use numerical methods to predict future climate change and extreme weather events.

[0003] However, due to the complexity of the climate system and the mutual influence of various factors in the earth system, the computational model, initial field, boundary conditions, parameterization schemes, etc. will introduce prediction errors. The existence of chaos theory makes it difficult for numerical models to make long-term accurate predictions. Therefore, the prediction data of numerical models often have systematic deviations from actual observations, that is, the prediction results of climate models are often affected by multiple factors and have a certain degree of uncertainty. In order to improve the prediction accuracy and reliability of climate models, they need to be comprehensively evaluated and verified. However, the current climate model prediction evaluation methods have many limitations. They cannot fully reflect their prediction performance. Summary of the invention

[0004] In view of this, the present invention provides a comprehensive evaluation method, device and storage medium for climate model prediction data to solve the problem that existing climate model evaluation methods have great limitations.

[0005] In a first aspect, the present invention provides a comprehensive evaluation method for climate model prediction data, the method comprising: obtaining prediction data and corresponding observation data of different climate models; calculating evaluation indicators corresponding to different climate models based on the prediction data and the observation data, the evaluation indicators including deviation, spatial correlation coefficient, root mean square error, time correlation coefficient and TS score; calculating the weight of each evaluation indicator according to the information entropy of each evaluation indicator; and determining the comprehensive evaluation results of different climate models according to the weighted sum of each evaluation indicator and the corresponding weight.

[0006] The comprehensive evaluation method of climate model prediction data provided by the embodiment of the present invention calculates multiple evaluation indicators of different climate models based on the prediction data and the observation data, calculates the weight of each evaluation indicator based on the information entropy of each evaluation indicator, and determines the comprehensive evaluation results of different climate models based on the weighted sum of each evaluation indicator and the corresponding weight. Thus, the method evaluates the model prediction accuracy, uncertainty estimation, prediction deviation, and extreme event prediction ability in time and space, avoids the one-sidedness of a single indicator, and makes the evaluation results more objective and credible.

[0007] In an optional embodiment, after obtaining the prediction data of different climate models, the method also includes: for the same climate model, formatting the prediction data; using spatial interpolation to interpolate the prediction data of different resolutions into the same grid; performing temporal interpolation on the prediction data after spatial interpolation; and processing the prediction data after temporal interpolation into globally uniform grid data with a uniform time step.

[0008] In this embodiment, for the same climate model, the forecast data is subjected to format processing, spatial interpolation and temporal interpolation, thereby obtaining globally uniform grid data with a uniform time step, thereby improving processing efficiency.

[0009] In an optional embodiment, after obtaining the prediction data of different climate models, the method also includes: extracting data of preset meteorological variables contained in multiple sets in preset time periods and preset areas in the prediction data of the same climate model, generating multiple data files corresponding to the multiple sets, the data of the multiple sets are the prediction data of the same climate model under different conditions; merging the multiple data files corresponding to the multiple sets to obtain a single data file; processing the single data file to obtain daily average data, monthly average data, multi-year climate state data or abnormal field data of the same climate model.

[0010] In this embodiment, by obtaining multiple sets of prediction data of the same climate model, the data of preset areas, preset time periods and preset meteorological variables in the multiple sets are extracted and merged to facilitate subsequent evaluation of different climate model data.

[0011] In an optional embodiment, data of preset meteorological variables contained in multiple sets in preset time periods and preset areas in the prediction data of the same climate model are extracted to generate multiple data files corresponding to the multiple sets, including: for each climate model, using a cshell script to read the prediction data of the preset meteorological variables contained in each set in each year in turn; segmenting the prediction data of the preset area in the read prediction data; extracting the prediction data of the preset time period in the segmented prediction data, to generate multiple data files corresponding to the multiple sets.

[0012] In this embodiment, cshell script is used to extract data and generate data files, which significantly improves data processing efficiency and consistency.

[0013] In an optional embodiment, the weight of each evaluation indicator is calculated according to the information entropy of each evaluation indicator, including: standardizing each evaluation indicator to obtain a standardized value of each evaluation indicator in each climate mode; calculating the information entropy of each evaluation indicator according to the standardized value of each evaluation indicator in each climate mode and the number of climate modes; determining the weight of each evaluation indicator according to the proportion of the variability value of each evaluation indicator and the total variability value, the variability value being the difference between one and the information entropy.

[0014] In this embodiment, the weight is determined based on the variability of each indicator. The greater the variability of the indicator, the greater the weight. The variability is determined by calculating the information entropy value of each indicator. The larger the entropy value, the higher the uncertainty of the indicator.

[0015] In an optional embodiment, the comprehensive evaluation result includes the comprehensive evaluation results corresponding to the prediction data of different climate models; the method also includes: selecting the prediction data with the best forecast effect according to the comprehensive evaluation results corresponding to the prediction data of different climate models; and improving different climate models according to the evaluation indicators and the comprehensive evaluation results.

[0016] In this embodiment, after the comprehensive evaluation result is calculated, the forecast data with the best forecast effect is determined through the comprehensive evaluation result, which provides a data basis for making various decisions; by improving the model, the performance and forecasting ability of the model are improved.

[0017] In the second aspect, the present invention provides a comprehensive evaluation device for climate model prediction data, the device comprising: a data acquisition module for acquiring prediction data and corresponding observation data of different climate models; an indicator calculation module for calculating evaluation indicators corresponding to different climate models based on the prediction data and observation data, the evaluation indicators including deviation, spatial correlation coefficient, root mean square error, time correlation coefficient and TS score; a weight calculation module for calculating the weight of each evaluation indicator according to the information entropy of each evaluation indicator; a comprehensive evaluation module for determining the comprehensive evaluation results of different climate models according to the weighted sum of each evaluation indicator and the corresponding weight.

[0018] In a third aspect, the present invention provides a computer device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the comprehensive evaluation method for climate model prediction data of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0019] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the comprehensive evaluation method for climate model prediction data of the above-mentioned first aspect or any corresponding embodiment thereof.

[0020] In a fifth aspect, the present invention provides a computer program product, comprising computer instructions for causing a computer to execute the comprehensive evaluation method for climate model prediction data of the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 is a flow chart of a comprehensive evaluation method for climate model prediction data according to an embodiment of the present invention;

[0023] Figure 2 is a flow chart of another comprehensive evaluation method for climate model prediction data according to an embodiment of the present invention;

[0024] Figure 3 It is a schematic diagram of TS scoring of the forecast data of the daily precipitation (unit: mm / day) in summer from lead 1month to 6month in the historical return test from 2003 to 2022 according to an embodiment of the present invention;

[0025] Figure 4 is a structural block diagram of a comprehensive evaluation device for climate model prediction data according to an embodiment of the present invention;

[0026] Figure 5 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0028] According to an embodiment of the present invention, an embodiment of a comprehensive evaluation method for climate model prediction data is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in an order different from that shown here.

[0029] In this embodiment, a comprehensive evaluation method for climate model prediction data is provided, which can be used in electronic devices such as computers, mobile phones, tablet computers, etc. Figure 1 is a flow chart of a comprehensive evaluation method for climate model prediction data according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0030] Step S101, obtain the prediction data and corresponding observation data of different climate models. Among them, climate model refers to a series of global climate models (Global Climate Models, GCMs), which are developed by different countries and research institutions for climate prediction. The global climate model is a kind of numerical model, which is a tool that uses mathematical equations and numerical methods to simulate and predict the behavior of complex systems. Used to simulate and predict the behavior of complex systems. The global climate model is specifically used to simulate and predict the changes and behaviors of the earth's climate system. The output results of the global climate model may include various meteorological variables, such as precipitation, temperature, air pressure and other prediction data.

[0031] Since the global climate model includes forecast data for different meteorological variables around the world in different time periods, the obtained observation data and forecast data correspond to each other. For example, if the forecast data is the precipitation forecast data for the Yangtze River Basin in the summer of 2022, the observation data is the precipitation observation data for the Yangtze River Basin in the summer of 2022. For example, the observation data can be selected from the GPCP (Global Precipitation Climatology Project) dataset.

[0032] Step S102, based on the predicted data and the observed data, the evaluation indicators corresponding to different climate models are calculated, and the evaluation indicators include deviation, spatial correlation coefficient, root mean square error, time correlation coefficient and TS (Threat Score) score. Specifically, the current model prediction evaluation method has many limitations. For example, the singleness of the indicators, the evaluation methods of many numerical models mainly rely on a few indicators, such as root mean square error or a certain correlation coefficient, to evaluate the prediction accuracy of the numerical model, which cannot fully reflect the performance of the numerical model. A model performs well in some indicators, but may not be ideal in other indicators. This step calculates multiple evaluation indicators including deviation, spatial correlation coefficient, root mean square error, time correlation coefficient and TS score, covering the evaluation of the prediction data in time and space, and comprehensively considering the prediction accuracy, spatiotemporal resolution, uncertainty estimation, prediction deviation, extreme event prediction ability, etc.

[0033] Bias refers to the average difference or offset between the predicted data and the observed data. Bias can assess whether the model has systematic bias over a period of time, that is, whether it tends to overestimate or underestimate the observed value. Bias has the advantages of being simple and intuitive to calculate, and can quickly make a preliminary assessment of the accuracy of the predicted data; it is applicable to a variety of meteorological variables and time scales. However, bias may mask the systematic bias in the model, and it is impossible to distinguish the source of the error, resulting in inaccurate assessment of extreme events.

[0034] Specifically, the deviation is calculated using the following formula:

[0035]

[0036] In the formula To predict data, is the observed data, and Bias represents the deviation between the two.

[0037] The Pattern Correlation Coefficient (PCC) is an indicator used to compare the similarity between two spatial patterns (such as climate model prediction data and observation data). That is, it evaluates the consistency between the prediction data and the observation data. The spatial correlation coefficient can evaluate the accuracy of the model's simulation of spatial distribution and can be used to compare the similarity between the spatial distribution simulated by the model and the observation data. However, it is difficult to handle inconsistencies in large and small scales, and is affected by the uneven spatial distribution of observation data.

[0038] Specifically, the spatial correlation coefficient is calculated using the following formula:

[0039]

[0040] The value of the spatial correlation coefficient ranges from -1 to 1. If the value is close to 1, it means that there is a strong positive correlation between the two fields, that is, their spatial patterns are very similar. If the value is close to -1, it means that there is a strong negative correlation between the two fields, that is, their spatial patterns have opposite trends. If the value is close to 0, it means that there is almost no correlation between the two fields, that is, their spatial patterns are almost dissimilar.

[0041] The root mean square error (RMSE) is an indicator used to measure the deviation between the predicted data and the actual observed data. It calculates the root mean square value of the difference between the predicted data and the actual observed data, thus reflecting the average degree of error between the predicted data and the actual observed data. The root mean square error can provide an average measure of the model prediction error. However, the root mean square error is more sensitive to outliers and may be affected by extreme values, so it is not robust enough. And because it is the square root of the sum of squares of the errors, it is more sensitive to large errors and may overemphasize the impact of large errors.

[0042] The root mean square error is calculated by first calculating the difference (residual) between each predicted data and the corresponding observed data; then squaring each difference; then calculating the average of all squared differences; and finally calculating the square root of the average squared difference. Specifically, the root mean square error is calculated using the following formula:

[0043]

[0044] In the formula, original i Represents the i-th observation data, predicted i Represents the i-th predicted data, and n represents the total amount of data. RMSE represents the average deviation between the predicted data and the observed data. The smaller the RMSE value, the higher the regression accuracy, which means the more accurate the forecast.

[0045] Temporal correlation coefficient (TCC) refers to a statistical indicator used to measure the correlation between time series data. It is usually used to analyze the correlation between adjacent time points in time series data to understand the pattern and trend of data changes over time. The temporal correlation coefficient can be used to measure the model's predictive ability for time series data, and can reflect the model's predictive ability for trends and periodicity of time changes. However, it may not perform well for non-periodic data; it only considers the correlation in the time dimension and may ignore important information on spatial distribution.

[0046] Specifically, the time correlation coefficient is calculated using the following formula:

[0047]

[0048] Among them, the time correlation coefficient ranges from -1 to 1. When TCC is close to 1, it means that there is a strong positive correlation between the model prediction and the observation, that is, the change of the model prediction is consistent with the change trend of the observation. When TCC is close to -1, it means that there is a strong negative correlation between the model prediction and the observation, that is, the change of the model prediction is opposite to the change trend of the observation. When TCC is close to 0, it means that there is no linear correlation between the model prediction and the observation. In physical interpretation, TCC is an indicator to measure the time series correlation between model prediction and observation. High TCC means that the model has a strong ability to predict the change trend of the observation value, while low TCC means that there is a large difference between the model prediction and the actual observation.

[0049] The TS score is a meteorological forecast scoring method used to evaluate the consistency between forecast events and observed events, ranging from 0 to 1, where 0 means no forecast skill. The TS score is a scoring indicator that comprehensively considers the prediction accuracy and error rate of the model. It can take into account the hit and false alarm rates of the model at the same time, so it has a strong ability to evaluate extreme events. However, the TS score is sensitive to the selection of different forecast thresholds and may be affected by the threshold selection.

[0050] Specifically, the TS score is calculated using the following formula:

[0051]

[0052] In the formula, Hits, Misses, and False Alarms represent the number of hits, misses, and false alarms, respectively. A hit means that both the observed event and the predicted event occurred. Misses mean that the observed event occurred, but the predicted event did not occur. False alarms mean that the predicted event occurred, but the observed event did not occur.

[0053] Step S103, the weight of each evaluation indicator is calculated according to the information entropy of each evaluation indicator. Specifically, information entropy is used to describe the uncertainty of various possible events in the information source. The average amount of information after excluding redundancy in the information is called "information entropy". The introduction of information entropy solves the problem of quantitative measurement of information, making it possible to describe and quantify the amount of information in mathematical language. Therefore, the weight of the evaluation indicator is calculated by the information entropy of the evaluation indicator, so that the calculated weight includes the information contained in the evaluation indicator.

[0054] Step S104, determining the comprehensive evaluation results of different climate modes according to the weighted sum of each evaluation index and the corresponding weight. Specifically, the comprehensive evaluation result is determined by weighting and calculating the evaluation index and the weight, so that the evaluation result combines the characteristics of each evaluation index. At the same time, by selecting the deviation, spatial correlation coefficient, root mean square error, time correlation coefficient and TS score as the evaluation index, the one-sidedness of different indicators can also be balanced. In this embodiment, the evaluation and verification of the climate mode can be achieved by determining the comprehensive evaluation result.

[0055] The comprehensive evaluation method of climate model prediction data provided by the embodiment of the present invention calculates multiple evaluation indicators of different climate models based on the prediction data and the observation data, calculates the weight of each evaluation indicator based on the information entropy of each evaluation indicator, and determines the comprehensive evaluation results of different climate models based on the weighted sum of each evaluation indicator and the corresponding weight. Thus, the method evaluates the model prediction accuracy, uncertainty estimation, prediction deviation, and extreme event prediction ability in time and space, avoids the one-sidedness of a single indicator, and makes the evaluation results more objective and credible.

[0056] In this embodiment, a comprehensive evaluation method for climate model prediction data is provided, and the process includes the following steps:

[0057] Step S201, obtain the prediction data and corresponding observation data of different climate models. Figure 1 Step S101 of the illustrated embodiment will not be described in detail here.

[0058] Step S202, for the same climate model, format the prediction data; use spatial interpolation to interpolate the prediction data of different resolutions into the same grid; perform temporal interpolation on the prediction data after spatial interpolation; process the prediction data after temporal interpolation into globally uniform grid data with a uniform time step.

[0059] Specifically, in order to adapt to the complex terrain of the earth's surface, climate models often use irregular geographical networks, that is, denser grid points are arranged in areas with active convection (such as near the equator) and complex terrain (such as near mountain ranges), and sparser grid points are used in other areas. However, the positions of the data points in the resulting prediction data on the earth's surface may be irregular. Therefore, the prediction data needs to be processed before evaluation. First, the obtained prediction data is formatted to ensure that the data format is correct. For example, the prediction data output by statistical models and artificial intelligence models are usually in different formats. For unified evaluation, all prediction data need to be processed into unified format and easy to read data; then the data is spatially interpolated to interpolate the prediction data of different resolutions into the same grid; then temporal interpolation is performed to match the required output time step, and finally the original data output by the climate model is processed into globally uniform grid data with a uniform time step.

[0060] Step S203, extracting data of preset meteorological variables contained in multiple sets in preset time periods and preset areas from the prediction data of the same climate model, generating multiple data files corresponding to the multiple sets, where the data of the multiple sets are prediction data of the same climate model under different conditions; merging the multiple data files corresponding to the multiple sets to obtain a single data file; processing the single data file to obtain daily average data, monthly average data, multi-year climate state data or abnormal field data of the same climate model.

[0061] Specifically, the globally uniform grid data with a uniform time step after processing in step S202 is full-variable, multi-set data. In order to facilitate subsequent evaluation and analysis, operations such as variable extraction, time merging, and set merging need to be performed on the data.

[0062] In an optional implementation, extracting data of preset meteorological variables contained in multiple sets in preset time periods and preset areas from the prediction data of the same climate model, and generating multiple data files corresponding to the multiple sets includes:

[0063] Step a1, for each climate model, use a cshell script to read the forecast data of the preset meteorological variables contained in each set in each year in turn; wherein, the C Shell (cshell) script is a command line interpreter and scripting language on Unix and Unix-like operating systems, and its syntax is similar to the C language. A CShell script is a text file containing a series of CShell commands, which is executed by the C Shell interpreter and is used for automated system tasks, file management and data processing. Script files are usually saved with a ".csh" or ".tcsh" extension, and the script can contain variable definitions, conditional statements, loop structures, etc. C Shell scripts can be used to automate data, file operations and other tasks in climate model prediction data processing, significantly improving processing efficiency and consistency.

[0064] In this embodiment, the prediction data includes the results of different climate model outputs, and the output results of the same climate model include multiple sets. Wherein, an ensemble refers to a set of prediction results obtained by running the same climate model multiple times under different initial conditions (such as different initial weather conditions) or different parameters (such as physical parameterization schemes). Each member in the set can include prediction results of multiple meteorological elements (such as precipitation, temperature, wind speed, etc.). For example, for a certain time period, each set member will output the prediction data of different elements in the time period. Each set member will provide prediction data of multiple elements. These results can help quantify the uncertainty and variability in the prediction, thereby improving the reliability of the prediction results.

[0065] When using cshell script to process data, first define a list of processing years, set numbers, and variable names, then select the specified variables from the forecast data for each year and each set number, extract data within a specific date range, and finally generate a new data file. For example, for a certain climate model, the forecast data includes data from 2003 to 2022. First, read the data of each year in turn. Since the data of each year includes data from multiple sets, read the data of the preset meteorological variables (such as the average daily precipitation rate) from each set, so as to obtain the forecast data of the preset meteorological variables contained in each set in each year.

[0066] Step a2, segmenting the prediction data of the preset area in the read prediction data; specifically, for the data from which the preset meteorological variables are extracted, since the data contains global data, in order to facilitate the analysis of a certain area, the data of the preset area is segmented out, such as segmenting the precipitation prediction data of the Yangtze River Basin from the global precipitation data.

[0067] Step a3, extracting the forecast data of the preset time period from the segmented forecast data, and generating multiple data files corresponding to the multiple sets. Specifically, through the processing of the above steps a1 and a2, the extraction of the preset areas and preset meteorological variables in the multiple sets is realized. For example, if the preset area is the Yangtze River Basin and the preset meteorological variable is the daily average precipitation rate, the data files obtained through steps a1 and a2 include:

[0068] Data on daily average precipitation rate in the Yangtze River Basin in 2003_Collection 1

[0069] …

[0070] Daily average precipitation rate data for the Yangtze River Basin in 2003_Set 24 (if this numerical model sets 24 sets) ......

[0072] Average daily precipitation rate data for the Yangtze River Basin in 2022 - Collection 1 ......

[0074] Average daily precipitation rate data for the Yangtze River Basin in 2022 - Collection 24

[0075] Each data file is stored separately according to different ensembles.

[0076] When conducting an evaluation, it is usually processed for a certain month or a certain season. Therefore, the above data files can be further processed according to the preset time period. For example, to extract the daily average precipitation rate data of the Yangtze River Basin in the summer of 2003-2022, the summer data can be extracted from each of the above sets and merged. The resulting data file can be expressed as:

[0077] Average daily precipitation rate in the Yangtze River Basin in summer 2003-2022_Collection 1

[0078] …

[0079] Average daily precipitation rate in the Yangtze River Basin in summer 2003-2022_Collection 24

[0080] In an optional implementation, multiple data files corresponding to multiple sets are merged to obtain a single data file, including the following steps: first, multiple data files are loaded, which contain different sets of the same meteorological variables (i.e., the outputs of different set forecasts); then the loaded files are processed, including reading time, longitude and latitude and other information, and converting them into floating point format. At the same time, the output data of each set is stored in different data; a new variable is created, which is a four-dimensional array containing four dimensions of time, longitude, latitude and set members. Then, the output data of each loaded set is merged into this new variable according to the time and member dimensions; the merged data is written into a new data file, retaining the dimensional information such as time, longitude, latitude and set members, and the merged variable is stored.

[0081] Therefore, by merging multiple data files, the merging of multiple sets is achieved. For example, the 2003-2022 summer Yangtze River Basin precipitation data set 1 to set 24, a total of 24 data, are merged into one file. If the original data is (24*360, longitude, latitude), the processed data becomes (24*360, longitude, latitude, number of sets), so that the 24 separate files obtained in the previous step are written into one file.

[0082] Step S204, based on the predicted data and the observed data, the evaluation indicators corresponding to different climate models are calculated. The evaluation indicators include deviation, spatial correlation coefficient, root mean square error, temporal correlation coefficient and TS score. For details, please refer to Figure 1 Step S102 of the illustrated embodiment will not be described in detail here.

[0083] Step S205, calculating the weight of each evaluation indicator according to the information entropy of each evaluation indicator;

[0084] Specifically, the above step S205 includes:

[0085] Step S2051, standardize each evaluation index to obtain a standardized value of each evaluation index in each climate mode; specifically, by standardizing the evaluation index, the dimensions can be made consistent. The standardization method used in this embodiment is the min-max standardization method, and the min-max standardization formula is as follows:

[0086]

[0087] Among them, x` is the standardized value and x is the original indicator value.

[0088] Step S2052, calculating the information entropy of each evaluation indicator according to the standardized value of each evaluation indicator in each climate mode and the number of climate modes; specifically, the information entropy is determined using the following formula:

[0089]

[0090] Among them, p ij is the normalized value of the jth evaluation indicator in the i-th climate model, m is the number of modes, E j is the information entropy value of the jth evaluation indicator.

[0091] Step S2053, determine the weight of each evaluation indicator according to the proportion of the variability value of each evaluation indicator and the total variability value, and the variability value is the difference between one and the information entropy. Specifically, in this embodiment, the weight is determined based on the variability value of each indicator. The larger the variability value of the indicator, the greater the weight. The variability value is determined by the difference between one and the information entropy, and thus, the weight is calculated using the following formula:

[0092]

[0093] Among them, ω j is the weight of the jth evaluation metric.

[0094] Step S206, determining the comprehensive evaluation results of different climate modes according to the weighted sum of each evaluation index and the corresponding weight. Specifically, the comprehensive evaluation results are expressed by the following formula:

[0095]

[0096] Among them, ω i is the weight of the i-th evaluation indicator, x i ′ is the standardized value of the ith evaluation index, and the final comprehensive score Comprehensive Index can be used to evaluate the overall performance of the climate model.

[0097] In this embodiment, a comprehensive evaluation method for climate model prediction data is provided, and the method comprises the following steps:

[0098] Step S301, obtain the prediction data and corresponding observation data of different climate models; see Figure 1 Step S101 of the illustrated embodiment will not be described in detail here.

[0099] Step S302: Calculate the evaluation indicators corresponding to different climate models based on the predicted data and the observed data. The evaluation indicators include deviation, spatial correlation coefficient, root mean square error, temporal correlation coefficient and TS score. For details, please refer to Figure 1Step S102 of the illustrated embodiment will not be described in detail here.

[0100] Step S303, calculate the weight of each evaluation indicator according to the information entropy of each evaluation indicator; see Figure 1 Step S103 of the illustrated embodiment will not be described in detail here.

[0101] Step S304: Determine the comprehensive evaluation results of different climate modes based on the weighted sum of each evaluation index and the corresponding weight. Figure 1 Step S104 of the illustrated embodiment will not be described in detail here.

[0102] Step S305, selecting the prediction data with the best forecast effect according to the comprehensive evaluation results corresponding to the prediction data of different climate models; specifically, the prediction data of all climate models can be compared and screened according to the comprehensive scores in the comprehensive evaluation results, so as to determine the prediction data with the best forecast effect.

[0103] Step S306: Improve different climate modes according to the evaluation index and the comprehensive evaluation result. Specifically, this embodiment can achieve the following improvements:

[0104] 1. Identify model deficiencies. By analyzing the calculation results of each evaluation indicator, identify which indicators perform poorly (for example, some models have high root mean square errors or low time correlation coefficients). By comparing the performance of different models, identify specific deficiencies, such as systematic deviations in the model's simulation of precipitation or inaccurate predictions of certain meteorological variables.

[0105] 2. Feedback and improvement of model parameters. By feeding back the comprehensive evaluation results to the model developers or research team, the model parameters can be adjusted. For example, the physical parameterization scheme can be adjusted, the initial condition setting can be improved, the boundary conditions can be optimized, etc. Targeted improvements can be made for specific problems identified. For example, if the precipitation forecast deviation is large, the precipitation parameterization scheme can be improved.

[0106] 3. Introduce new data and methods. Specifically, new observation data and higher quality input data can be introduced to improve the initial conditions and boundary conditions of the model. New numerical methods and algorithms can be applied to improve the calculation accuracy and efficiency of the model. For example, finer grid division and improved numerical solution algorithms can be used.

[0107] 4. Repeated iteration and verification. Specifically, the improved model can be re-evaluated and various indicators can be recalculated using the comprehensive scoring method. Through iterative improvement, the performance of the model is continuously optimized until satisfactory prediction accuracy and reliability are achieved.

[0108] The comprehensive evaluation method for climate model prediction data provided by the embodiment of the present invention can verify the accuracy of the model by calculating the comprehensive evaluation results, that is, the accuracy of the meteorological model can be verified by the comprehensive evaluation results, and it is determined whether the predicted data is consistent with the actual observed data. This helps to determine the reliability and applicability of the model and improve the accuracy and credibility of the forecast; the model parameters and algorithms can also be improved, for example, the use of comprehensive evaluation results can help identify the deficiencies in the model parameters and algorithms. Based on the evaluation results, the parameters of the model can be adjusted or the algorithm can be improved to improve the performance and forecasting ability of the model; through the comprehensive evaluation results, the scope of application and limitations of the model can also be determined, so as to better understand the significance and reliability of the model output, so as to effectively apply it in the fields of meteorological forecasting, climate research, etc., and improve the application value; in addition, accurate meteorological forecasts are of great significance for various decision-making, such as responding to extreme weather events, agricultural production arrangements, transportation planning, etc. Evaluation of model output can provide more reliable information and help make better decisions; and the evaluation of model output is part of the development of meteorological science. Through continuous evaluation and improvement, the progress and development of numerical models can be promoted, and the level of meteorological forecasting and climate simulation can be improved.

[0109] As a specific application example of the embodiment of the present invention, Figure 2 As shown in Figure 1, the comprehensive evaluation method of climate model prediction data specifically includes the following process: using the data collection and processing module to obtain the prediction data output by different climate models, and processing it into globally uniform grid data with a unified time step; then using the result evaluation module to calculate the bias (Bias), PCC, RMSE, TCC, TS and other evaluation indicators, and using the formula and Calculate a comprehensive score, determine the optimal prediction data set for a certain area based on the comprehensive score, or improve the climate model. Specifically, this embodiment takes the seasonal precipitation in the Yangtze River Basin as an example, evaluates the prediction accuracy of a certain model (such as ModelA) for the seasonal precipitation in the Yangtze River Basin in summer (June, July, and August), and generates a set of optimal data sets for the prediction of seasonal precipitation in the Yangtze River Basin in summer based on the evaluation results. The observation data selected is the GPCP (Global Precipitation Climatology Project) daily precipitation data set. For ModelA, the result evaluation module is used to calculate the values ​​of evaluation indicators such as Bias, PCC, RMSE, TCC, and TS for the Yangtze River Basin, which are 0.2, 0.8, 2.5, 0.7, and 0.5, respectively.

[0110] Among them, this embodiment collects the difference distribution data between the average summer precipitation in June, July and August in the Yangtze River Basin predicted one month in advance and the observed data. The data shows that the assessment deviation of the climate state in the Yangtze River Basin is distributed between -2mm / day and 2mm / day, with a small deviation.

[0111] This embodiment also collects the summer climate state distribution data of the Yangtze River Basin for the observed data and the predicted data, wherein the PCC value of the predicted data and the observed data is 0.892, indicating that the spatial simulation of the model for the summer seasonal precipitation prediction in the Yangtze River Basin is good and is consistent with the spatial distribution of the observed data. In addition to the Yangtze River Basin, the PCC values ​​of other regions where the model outputs the predicted data are also calculated, as shown in Table 1. From Table 1, it can be clearly seen that the spatial simulation effect of the model for different regions reflects the model simulation effect.

[0112] Table 1 PCC (spatial correlation coefficient) in different regions

[0113]

[0114] In Table 1, Global means global, i.e., the spatial correlation coefficient between the global prediction data and the observation data output by the model; EA means east Asia, i.e., the spatial correlation coefficient between the prediction data and the observation data output by the model in East Asia; YRB means Yangtze River Basin, UYRB means the upper Yangtze River Basin; M / LRB means the middle / low river basin. The numbers and letters in brackets indicate longitude and latitude, such as EA(4:53N,73:150E), which means 4-53 degrees north latitude and 73-150 degrees east longitude in East Asia.

[0115] This embodiment also collects the spatial distribution data of the root mean square error (RMSE) between the forecast results of the summer precipitation anomaly (unit: mm / day) in the historical feedback experiment in the Yangtze River Basin from 2003 to 2022, which starts one month in advance and the GPCP (observation data) results. The data show that the RMSE in the Yangtze River Basin is stable between 0.8-1.2, indicating that the average deviation between the predicted data and the observed data is small, indicating that the model has a certain simulation ability for the summer precipitation in the Yangtze River Basin, and has a stronger simulation ability for the middle and lower reaches of the Yangtze River.

[0116] This embodiment also collects the spatial distribution data of TCC of the forecast results of summer precipitation anomaly and GPCP (observation data) results of the historical feedback test in the Yangtze River Basin from 2003 to 2022, which is reported one month in advance. The data show that the forecast data and observation data of summer abnormal precipitation in the middle and lower reaches of the Yangtze River are highly correlated, and the TCC distribution is above 0.4 and many areas have passed the 95% significance test. There are also some skills in the upper reaches of the Yangtze River. The invention can intuitively reflect the prediction effect of the numerical model by calculating TCC.

[0117] Figure 3 The TS scores of light rain for the June, July and August precipitation are given in December (lead 1month), January (lead 2month), February (lead 3month), March (lead 4month), April (lead 5month) and May (lead 6month). It can be seen from the figure that regardless of the start time, the TS scores of the Yangtze River Basin as a whole and its upper and middle and lower reaches are stable between 0.32 and 0.33. It is particularly noteworthy that the TS score of the middle and lower reaches of the Yangtze River is higher than that of the other two regions. This shows that in the middle and lower reaches of the Yangtze River, the forecast model has a more outstanding ability to predict the occurrence of precipitation weather events. Judging from the results of the TS score, the climate model has a certain degree of forecasting skills, but there is still room for improvement. The model can capture the trend of precipitation events in some cases, but it still has limitations in predicting some subtle or specific situations. Therefore, although the TS score is around 0.3, the overall performance of the model is acceptable for precipitation forecasting in the Yangtze River Basin, but there is still room for improvement.

[0118] In addition, the evaluation index values ​​corresponding to ModeB and ModelC can also be calculated using the result evaluation module and visualized in the above manner. Therefore, the evaluation index values ​​of the Yangtze River Basin for climate models ModelA, ModeB and ModelC can be shown in Table 2 below:

[0119] Table 2

[0120] model Bias PCC RMSE TCC TS Model A 0.2 0.8 2.5 0.7 0.5 Model B 0.1 0.7 3.0 0.6 0.6 Model C 0.15 0.75 2.8 0.65 0.55

[0121] After determining the evaluation index value, the comprehensive score can be calculated. The calculation process can be implemented using the following process:

[0122] 1. Standardized processing.

[0123] This embodiment uses the min-max normalization formula shown below to calculate the normalized value of each indicator.

[0124]

[0125] Among them, for the evaluation index Bias, according to Table 1 above, min = 0.1, max = 0.2. Therefore, the standardized value of ModelA is The normalized value of ModelB is The standardized value of ModelA is For the PCC evaluation index, according to Table 1 above, min = 0.7, max = 0.8. Therefore, the standardized value of Model A is The normalized value of ModelB is The standardized value of ModelA is For the RMSE evaluation index, according to Table 1 above, min = 2.5, max = 3.0. Therefore, the standardized value of Model A is The normalized value of ModelB is The standardized value of ModelA is For the TCC evaluation index, according to Table 1 above, min = 0.6, max = 0.7. Therefore, the standardized value of Model A is The normalized value of ModelB is The standardized value of ModelA is For the TS evaluation index, according to Table 1 above, min = 0.5, max = 0.6. Therefore, the standardized value of Model A is The normalized value of ModelB is The standardized value of ModelA is Based on the above calculation results, the standardized evaluation indicators are shown in Table 3 below:

[0126] Table 3

[0127] model Bias' PCC' RMSE' TCC' TS' Model A 0.2 0.8 2.5 0.7 0.5 Model B 0.1 0.7 3.0 0.6 0.6 Model C 0.15 0.75 2.8 0.65 0.55

[0128] 2. Calculate the indicator weights.

[0129] (1) The information entropy E of the jth index is calculated using the following formula j :

[0130]

[0131] Among them, p ij is the normalized value of the jth evaluation indicator in the i-th climate model, m is the number of modes, and in this embodiment, m=3.

[0132] Among them, for the evaluation indicator Bias, its information entropy can be calculated according to Table 2 above:

[0133]

[0134] For the PCC evaluation indicator, its information entropy can be calculated according to Table 2 above:

[0135] E PCC =0.3466

[0136] For the RMSE evaluation indicator, its information entropy can be calculated according to Table 2 above:

[0137] E RMSE =0.3466

[0138] For the TCC evaluation indicator, its information entropy can be calculated according to Table 2 above:

[0139] E TCC =0.3466

[0140] For the TS evaluation indicator, its information entropy can be calculated according to Table 2 above:

[0141] E TS =0.3466

[0142] (2) The weight of the jth indicator is calculated using the following formula:

[0143]

[0144] Among them, for the evaluation indicator Bias, its weight is calculated using the following formula:

[0145]

[0146] Similarly, the calculation process of PCC, RMSE, TCC, and TS is the same as the weight calculation process of Bias, and each weight is 0.2.

[0147] 3. Calculate the comprehensive score.

[0148] The comprehensive score of each mode is calculated using the following formula:

[0149]

[0150] The ModelA comprehensive score is calculated using the following formula:

[0151] Comprehensive score = 0.2 × 1.0 + 0.2 × 1.0 + 0.2 × 0.0 + 0.2 × 1.0 + 0.2 × 0.0 = 0.6

[0152] The ModelB comprehensive score is calculated using the following formula:

[0153] Comprehensive score = 0.2 × 0.0 + 0.2 × 0.0 + 0.2 × 1.0 + 0.2 × 0.0 + 0.2 × 1.0 = 0.4

[0154] The ModelB comprehensive score is calculated using the following formula:

[0155] Comprehensive score = 0.2 × 0.5 + 0.2 × 0.5 + 0.2 × 0.6 + 0.2 × 0.5 + 0.2 × 0.5 = 0.52

[0156] Thus, ModelA has a comprehensive score of 0.6, ModelB has a comprehensive score of 0.4, and ModelC has a comprehensive score of 0.52. According to the comprehensive score, ModelA performs best, followed by ModelC, and finally ModelB.

[0157] In this embodiment, a comprehensive evaluation device for climate model prediction data is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0158] This embodiment provides a comprehensive evaluation device for climate model prediction data, such as Figure 4 As shown, including:

[0159] A data acquisition module 81 is used to obtain prediction data and corresponding observation data of different climate models;

[0160] An indicator calculation module 82 is used to calculate evaluation indicators corresponding to different climate models based on the predicted data and the observed data, the evaluation indicators including deviation, spatial correlation coefficient, root mean square error, temporal correlation coefficient and TS score;

[0161] A weight calculation module 83 is used to calculate the weight of each evaluation indicator according to the information entropy of each evaluation indicator;

[0162] The comprehensive evaluation module 84 is used to determine the comprehensive evaluation results of different climate modes according to the weighted sum of various evaluation indicators and corresponding weights.

[0163] In an optional embodiment, the device also includes: a data processing module, which is specifically used to format the prediction data for the same climate model; use spatial interpolation to interpolate the prediction data of different resolutions into the same grid; perform temporal interpolation on the prediction data after spatial interpolation; and process the prediction data after temporal interpolation into globally uniform grid data with a uniform time step.

[0164] In an optional embodiment, the device also includes: a data extraction module, which is used to extract data of preset meteorological variables contained in multiple sets in preset time periods and preset areas in the prediction data of the same climate model, and generate multiple data files corresponding to the multiple sets, and the data of the multiple sets are the prediction data of the same climate model under different conditions; a file merging module, which is used to merge the multiple data files corresponding to the multiple sets to obtain a single data file; and a file processing module, which is used to process the single data file to obtain daily average data, monthly average data, multi-year climate state data or abnormal field data of the same climate model.

[0165] In an optional embodiment, the data extraction module, for each climate model, uses a cshell script to sequentially read the prediction data of the preset meteorological variables contained in each set in each year; divides the prediction data of the preset area in the read prediction data; extracts the prediction data of the preset time period in the divided prediction data, and generates multiple data files corresponding to multiple sets.

[0166] In an optional embodiment, the weight calculation module is specifically used to: standardize each evaluation indicator to obtain a standardized value of each evaluation indicator in each climate mode; calculate the information entropy of each evaluation indicator based on the standardized value of each evaluation indicator in each climate mode and the number of climate modes; determine the weight of each evaluation indicator based on the proportion of the variability value of each evaluation indicator and the total variability value, the variability value being the difference between one and the information entropy.

[0167] In an optional embodiment, the comprehensive evaluation result includes the comprehensive evaluation results corresponding to the prediction data of different climate models; the device also includes: an improvement module, which is specifically used to select the prediction data with the best forecast effect according to the comprehensive evaluation results corresponding to the prediction data of different climate models; and improve different climate models according to the evaluation indicators and the comprehensive evaluation results.

[0168] The further functional description of each of the above modules is the same as that of the above corresponding embodiments and will not be repeated here.

[0169] The embodiment of the present invention also provides a computer device having the above Figure 4 The comprehensive evaluation device of climate model prediction data is shown.

[0170] See also Figure 5 , Figure 5 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Figure 5As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 5 A processor 10 is taken as an example.

[0171] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0172] The memory 20 stores instructions executable by at least one processor 10, so that at least one processor 10 executes the method shown in the above embodiment.

[0173] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the use of a computer device based on the presentation of a small program landing page, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0174] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0175] The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0176] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0177] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of the computer program instruction in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc., and accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0178] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the attached application.

Claims

1. A comprehensive evaluation method for climate model prediction data, characterized in that: The method comprises: Obtain prediction data and corresponding observation data from different climate models; Calculate evaluation indicators corresponding to different climate models based on the predicted data and the observed data, wherein the evaluation indicators include deviation, spatial correlation coefficient, root mean square error, temporal correlation coefficient and TS score; Calculate the weight of each evaluation indicator according to the information entropy of each evaluation indicator; The comprehensive evaluation results of different climate models are determined based on the weighted sum of various evaluation indicators and corresponding weights.

2. The method according to claim 1, characterized in that After obtaining the prediction data of different climate models, the method further includes: For the same climate model, formatting the prediction data; Using spatial interpolation, the prediction data of different resolutions are interpolated into the same grid; Perform temporal interpolation on the predicted data after spatial interpolation; The temporally interpolated prediction data are processed into globally uniform grid data with a uniform time step.

3. The method according to claim 1, characterized in that After obtaining the prediction data of different climate models, the method further includes: Extracting data of preset meteorological variables contained in multiple sets in preset time periods and preset areas from the prediction data of the same climate model, and generating multiple data files corresponding to the multiple sets, wherein the data of the multiple sets are prediction data of the same climate model under different conditions; Merge multiple data files corresponding to multiple sets to obtain a single data file; The single data file is processed to obtain daily average data, monthly average data, multi-year climate state data or abnormal field data of the same climate model.

4. The method according to claim 1, characterized in that: Extracting data of preset meteorological variables contained in multiple sets in preset time periods and preset areas from the prediction data of the same climate model, and generating multiple data files corresponding to the multiple sets, including: For each climate model, the cshell script is used to read the forecast data of the preset meteorological variables contained in each set in each year in turn; Splitting the prediction data of the preset area in the read prediction data; The prediction data of a preset time period is extracted from the segmented prediction data to generate multiple data files corresponding to the multiple sets.

5. The method according to claim 1, characterized in that The weight of each evaluation indicator is calculated according to the information entropy of each evaluation indicator, including: Each evaluation index is standardized to obtain the standardized value of each evaluation index in each climate model; The information entropy of each evaluation indicator is calculated based on its standardized value in each climate mode and the number of climate modes; The weight of each evaluation indicator is determined according to the proportion of the variability value of each evaluation indicator to the total variability value, and the variability value is the difference between one and the information entropy.

6. The method according to claim 1, characterized in that The comprehensive assessment result includes comprehensive assessment results corresponding to prediction data of different climate models; the method further includes: Select the forecast data with the best forecast effect based on the comprehensive evaluation results corresponding to the forecast data of different climate models; Different climate models are improved based on the evaluation indicators and comprehensive evaluation results.

7. A comprehensive evaluation device for climate model prediction data, characterized in that: The device comprises: Data acquisition module, used to obtain prediction data and corresponding observation data of different climate models; An indicator calculation module, used to calculate evaluation indicators corresponding to different climate models based on the predicted data and the observed data, wherein the evaluation indicators include deviation, spatial correlation coefficient, root mean square error, temporal correlation coefficient and TS score; A weight calculation module is used to calculate the weight of each evaluation indicator according to the information entropy of each evaluation indicator; The comprehensive evaluation module is used to determine the comprehensive evaluation results of different climate modes based on the weighted sum of various evaluation indicators and corresponding weights.

8. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the comprehensive evaluation method for climate model prediction data according to any one of claims 1 to 6 by executing the computer instructions.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the comprehensive evaluation method for climate model prediction data according to any one of claims 1 to 6.

10. A computer program product, characterized in that The method comprises computer instructions for causing a computer to execute the comprehensive evaluation method for climate model prediction data according to any one of claims 1 to 6.

Images