Short temporary rainfall forecasting method and device based on dynamic weight determination

Abstract

The invention provides a short temporary rainfall forecasting method and device based on dynamic weight determination, and solves the technical problem that the accuracy and stability cannot be realized by comprehensively utilizing the data of each forecasting source according to the forecasting performance change. The method comprises the following steps: forming a dynamic evaluation index of forecasting accuracy of each forecasting mechanism in the same area according to the frequency of accurate rainfall forecasting in the same area in the past time period of the current forecasting moment; and arbitrating the forecast data of the future time period at the current forecast moment by each forecast mechanism by using the dynamic evaluation index to form forecast of the rainfall event, dynamically quantifying the rainfall weight of the forecast data according to the dynamic evaluation index, and forming forecast of the rainfall according to the weight and the forecast data. And a measurement reference for forecasting performance change and forecasting accuracy of the forecasting source in a spatial scale under the same real-time time scale is established. And forming a consensus basis for arbitration aiming at an event trend and a data source dynamic fixed weight for event quantitative analysis. And the comprehensive forecasting result is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of weather forecasting, and particularly to a short-term and imminent precipitation forecasting method and device based on dynamic weight determination. Background Art

[0002] Short-term and imminent precipitation forecasting generally refers to the prediction of the probability and intensity of rainfall events within the next few hours, which has important practical significance for fields such as flood control, traffic management, and agricultural planning. In urban areas, short-term and imminent precipitation forecasting can help relevant departments make drainage preparations in advance and reduce the risk of urban waterlogging; in the agricultural field, timely precipitation forecasting can guide farmers to reasonably arrange irrigation and harvesting and reduce losses. Therefore, improving the accuracy of short-term and imminent precipitation forecasting is of great significance for reducing the impact of natural disasters and ensuring the safety of people's lives and property. Due to the rapid changes and uncertainties of weather systems, short-term and imminent precipitation forecasting has always been a challenge in meteorological forecasting. With the increasing demand for meteorological services in society, it has become particularly important to improve the accuracy and timeliness of short-term and imminent precipitation forecasting.

[0003] With the development of meteorological prediction technology, many current forecasting agencies can provide rainfall forecasts at different time scales and spatial scales. However, the forecasting accuracy of a single forecasting agency is often affected by various factors and it is difficult to ensure accurate forecasts in all cases. Therefore, in order to overcome the limitations of a single weather forecasting model and improve the accuracy of rainfall forecasting, it is necessary to integrate the forecasting information of multiple forecasting agencies. There are forecasting differences in the forecasting information between different forecasting models or services. How to integrate these differences to provide a stable and reliable forecasting process to optimize the accuracy and adaptability of the forecasting results is a technical challenge. Summary of the Invention

[0004] In view of the above problems, embodiments of the present invention provide a short-term and imminent precipitation forecasting method and device based on dynamic weight determination, which solve the technical problem of being unable to comprehensively utilize the forecasting data of each forecasting source according to the change of forecasting performance to achieve accuracy and stability.

[0005] The short-term and imminent precipitation forecasting method based on dynamic weight determination according to embodiments of the present invention includes:

[0006] Forming a dynamic evaluation index of the forecasting accuracy of each forecasting agency in the same region according to the frequency of accurate precipitation forecasting in the same region during the past period at the current forecasting moment;

[0007] Arbitrating the forecasting data of each forecasting agency for the future period at the current forecasting moment by using the dynamic evaluation index to form a forecast of precipitation events, dynamically quantifying the rainfall weight of the forecasting data according to the dynamic evaluation index, and comprehensively forecasting the data according to the rainfall weight to form a forecast of rainfall.

[0008] In one embodiment of the present invention, the formation of the dynamic evaluation index includes:

[0009] Normalize the forecast information and observation information of each forecasting agency and convert them into forecast data;

[0010] Based on the forecast data of the same region in the past period, form the TS score of the accurate precipitation forecast of each forecasting agency in the same region every day;

[0011] Count the numerical repetition frequency of the TS scores of each forecasting agency in the same past period in the same region;

[0012] Based on the TS score, numerical repetition frequency and numerical threshold, form the dynamic evaluation index of each forecasting agency in the same past period in the same region.

[0013] In one embodiment of the present invention, the past period is 180 days or one year.

[0014] In one embodiment of the present invention, the past period also includes the same seasons or the same months of different years.

[0015] In one embodiment of the present invention, the formation of the precipitation event forecast includes:

[0016] Based on the forecast data of each forecasting agency for a specific future period, determine the positive and negative sides that support the occurrence of the precipitation event;

[0017] Use the cumulative sum of the dynamic evaluation indexes of the positive and negative sides as the arbitration basis to form the forecast result of the precipitation event.

[0018] In one embodiment of the present invention, the formation of the rainfall forecast includes:

[0019] Based on the dynamic evaluation index, form the rainfall weight of each forecasting agency;

[0020] Based on the rainfall weight, comprehensively form the forecast result of the rainfall according to the forecast data of the positive forecasting agencies.

[0021] In one embodiment of the present invention, it further includes:

[0022] Based on the difference between the forecast and the observation of the rainfall, form the quantification of the forecast error and the optimization of the evaluation.

[0023] The short-term and imminent precipitation forecasting device based on dynamic weight determination according to the embodiment of the present invention includes:

[0024] A memory for storing the program code of the processing process of the above-mentioned short-term and imminent precipitation forecasting method based on dynamic weight determination;

[0025] A processor for executing the program code.

[0026] The short-term and imminent precipitation forecasting device based on dynamic weight determination according to an embodiment of the present invention includes:

[0027] A dynamic evaluation setting module, configured to form a dynamic evaluation index of the forecasting accuracy of each forecasting agency in the same region according to the frequency of accurate precipitation forecasting in the same region during the past period at the current forecasting moment;

[0028] A dynamic weight determination forecasting module, configured to arbitrate the forecasting data of each forecasting agency for the future period at the current forecasting moment by using the dynamic evaluation index to form a forecast of a precipitation event, dynamically quantify the rainfall weight of the forecasting data according to the dynamic evaluation index, and form a forecast of the rainfall amount by comprehensively forecasting the data according to the rainfall weight.

[0029] In an embodiment of the present invention, it further includes:

[0030] An error feedback optimization module, configured to form a forecast error quantification and evaluation optimization according to the difference between the forecast and the observation of the rainfall amount.

[0031] The short-term and imminent precipitation forecasting method and device based on dynamic weight determination according to the embodiment of the present invention establish a measurement benchmark for the change of the forecasting performance of the forecasting source and the forecasting accuracy within the spatial scale under the same real-time time scale by forming a dynamic evaluation index for the forecasting accuracy of each forecasting source in the past period. By using the timeliness and dynamics of the evaluation index, a consensus basis for arbitrating the event trend and a data source dynamic weight determination for event quantitative analysis are formed. The optimal weight configuration is obtained by calculating the weights of each forecasting source in real time, so that the comprehensive forecasting result is optimized. It can adapt to different weather conditions and changes in the performance of the forecasting source, and improve the forecasting accuracy and adaptability by using the complementarity of different prediction tools. It realizes the optimal combination of decision-making sources, balances the importance and reliability of different forecasting agencies, and improves the forecasting accuracy and stability. Description of the Drawings

[0032] Figure 1 The figure shows a schematic flowchart of a short-term and imminent precipitation forecasting method based on dynamic weight determination according to an embodiment of the present invention.

[0033] Figure 2 The figure shows a schematic architecture diagram of a short-term and imminent precipitation forecasting device based on dynamic weight determination according to an embodiment of the present invention. Detailed Embodiments

[0034] To make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0035] An embodiment of the present invention is a short-term and imminent precipitation forecasting method based on dynamic weight determination as follows Figure 1 shown. In Figure 1 , this embodiment includes:

[0036] Step 100: According to the frequency of accurate precipitation forecasts in the same region during the past period at the current forecast moment, form a dynamic evaluation index for the forecast accuracy of each forecasting agency in the same region.

[0037] Those skilled in the art can understand that the current forecast moment refers to the time point at which, for a future determined period or a future determined moment, the current prediction result is made using existing observation data, telemetry data, remote sensing data, and empirical data. This time point and the basic data used can be time-sequentially identified through a time stamp formed by the clock of the National Time Service Center. Different forecasting agencies usually form independent regional divisions and set up meteorological observation stations within the same geographical terrain area. The forecasting agency uses independent prediction tools (such as information processing algorithms and data analysis models, etc.) to analyze the basic data to form independent forecast data to show the forecast results of short-term and imminent precipitation. The past period refers to a determined-duration time period that extends to the current forecast moment before reaching the current forecast moment. The forecast data formed during the past period can be processed with time-sequential identification.

[0038] Generally, the forecasting agency conducts short-term and imminent precipitation forecasts at unit time intervals. The forecast results for precipitation events are quantified for accuracy through on-site observation results and quantified into the following four states:

[0039]

[0040]

[0041] By statistically counting the frequencies of accurate forecast results of precipitation forecasts by each forecasting agency in the same region during the past period, an evaluation index for the quantitative evaluation of the forecast accuracy of each forecasting agency in the same region can be formed. As the current forecast moment progresses, the evaluation index changes dynamically.

[0042] Step 200: Arbitrate the forecast data of each forecasting agency for the future period at the current forecast moment using the dynamic evaluation index to form a forecast of precipitation events. Dynamically quantify the rainfall weight of the forecast data according to the dynamic evaluation index, and synthesize the forecast data according to the rainfall weight to form a forecast of rainfall.

[0043] The future time period refers to a certain definite time period or time point after the current forecast moment. The forecast data includes at least precipitation events confirmed by the forecasting agency and the estimated rainfall amounts in the precipitation events. The arbitration behavior can be embodied as the principle of the minority obeying the majority with rules, and dynamic evaluation indicators can be used as references in the rules. The dynamic evaluation indicators are iteratively formed from regional and time-series forecast data. By establishing a dynamic association of weights for the estimated rainfall amounts of each forecasting agency in a determined area through the dynamic evaluation indicators, the real-time nature of the weight association is formed by using the time-series dynamics of the dynamic evaluation indicators. The comprehensive estimated result of the rainfall amount is obtained through the weight association.

[0044] The short-term and imminent precipitation forecasting method and device based on dynamic weight determination in the embodiments of the present invention establish a measurement benchmark for the change in the forecasting performance of forecasting sources and the forecasting accuracy within the spatial scale at the same real-time time scale by forming dynamic evaluation indicators for the forecasting accuracy of each forecasting source in the past time period. By using the timeliness and dynamics of the evaluation indicators, a consensus basis for arbitrating event trends and a data source dynamic weight determination for event quantitative analysis are formed. The optimal weight configuration is obtained by calculating the weights of each forecasting source in real time, so that the comprehensive forecasting result is optimized. It can adapt to different weather conditions and changes in the performance of forecasting sources, and improve the forecasting accuracy and adaptability by using the complementarity of different prediction tools. It realizes the optimal combination of decision-making sources, balances the importance and reliability of different forecasting agencies, and improves the forecasting accuracy and stability.

[0045] As Figure 1 shown, in an embodiment of the present invention, step 100 includes:

[0046] Step 110: Normalize the forecasting information and observation information of each forecasting agency and convert them into forecast data.

[0047] The data sources for each forecasting agency to conduct precipitation event and rainfall amount forecasts include the observation information of rain gauges arranged in meteorological observation stations set by the forecasting agency in a determined area. There are differences in the regional division and meteorological observation station settings of each forecasting agency. Each forecasting agency often forms differential data storage for forecasting information and observation information based on a general data structure.

[0048] Normalizing the forecasting information and observation information is beneficial to meeting the consistency and comparability of data, and comprehensively using the forecast data of multiple data sources to improve the data analysis ability. The normalization process includes, but is not limited to, data processing at levels such as time-series intervals, regional division, and format definition. The normalization process mainly includes:

[0049] Data cleaning: Remove outliers and missing data to ensure data quality.

[0050] Format Definition Conversion: The product update frequencies and data element units of different forecasting agencies vary. Unifying the data formats of different data sources facilitates subsequent processing.

[0051] Cycle Adaptation: Convert the time series of forecast information and observation information at fixed time intervals. For example, adjust the update frequency with an hourly time interval.

[0052] Region Matching: The rainfall data observed by the rain gauges of meteorological observation stations is the rainfall amount at known specific regional locations, while the forecast data provided by forecasting agencies are mostly grid meteorological products. It is necessary to match the forecast data (including the corresponding observation data) of different data sources in terms of region to ensure the consistency of the regional coordinate space of the data.

[0053] The forecast data after normalization by each forecasting agency is stored in a relational database for incremental update of the forecast data with the continuously formed new forecast data.

[0054] Step 120: Based on the forecast data of the same region in the past period, form the TS score of the accurate precipitation forecast of each forecasting agency in the same region every day.

[0055] In an embodiment of the present invention, the past period is a duration of 180 days forward from the current forecast time.

[0056] In an embodiment of the present invention, the Threat Score (TS) mechanism is used to evaluate the accuracy of precipitation forecasts. The value range of the TS score is from 0 to 1, and the higher the value, the better the accuracy of the forecast. A perfectly accurate forecast will get a TS score of 1, while a completely unfounded forecast (i.e., a random forecast) usually has a TS score close to 0.

[0057] In an embodiment of the present invention, the forecast data of each forecasting agency is hourly forecast data, that is, a day includes 24 sets of forecast data.

[0058] In an embodiment of the present invention, there are four forecast results: correctly forecasting rain, correctly forecasting no rain, falsely forecasting rain (actually no rain), and missing rain (actually there is rain but forecasting no rain). In a day, the frequencies of the four forecast results are denoted as TP (True Positive), TN (True Negative), FP (False Positive), and FN (False Negative).

[0059] The value of the TS score is:

[0060]

[0061] where i is the identifier of the forecasting agency, j is the identifier of the forecast region, and k is the date in the past period, that is, TSi,j,k It is the TS score for the accurate precipitation forecast of the forecasting agency i on the k-th day in the forecasting area j.

[0062] When TN = 24, TP = FP = FN = 0 (actually no rain and the forecast is also no rain), the TS score is recorded as 0.5.

[0063] Step 130: Count the numerical repetition frequency of the TS scores of each forecasting agency in the same past period in the same region.

[0064] The repetition frequency of the same TS score value in the past period (180 days) is f i,j (TS i,j,k ), that is, the number of times each TS score value TS i,j,k appears in 180 TS scores (corresponding to 180 days).

[0065] Step 140: Form the dynamic evaluation index of each forecasting agency in the same past period in the same region according to the TS score, the numerical repetition frequency, and the numerical threshold.

[0066] In an embodiment of the present invention, the dynamic evaluation index is as follows:

[0067] S i,j = ∑ TS≥.5 TS i,j,k × f i,j (TS i,j,k ) (2)

[0068] where TS ≥ 0.5 is the numerical threshold of the TS score, and TS i,j,k × f i,j (TS i,j,k ) is the product of the TS score value and the repetition frequency. The dynamic evaluation index S i,j is the sum of the products of the TS score values of the forecasting agency i in the forecasting area j that are greater than 0.5. The dynamic evaluation index S i,j changes accordingly with the adjustment of the current forecast time, the duration of the past period, and the numerical threshold of the TS score. As time progresses, the dynamic evaluation index is in dynamic update.

[0069] The short-term and imminent precipitation forecasting method based on dynamic weight determination in the embodiment of the present invention feeds back the reliability and accuracy of past forecast data through the dynamic evaluation index. It uses the TS score to quantify the accuracy measurement within a specific duration, uses the accumulation of the product of the TS scores to amplify the scale to improve the measurement accuracy, and uses the dynamic nature of the TS score over time to obtain a real-time measurement of the improvement of the forecast performance. The dynamic evaluation index establishes a real-time basis for the forecast accuracy, the improvement of the forecast performance, and the quantitative comparison among agencies within the past period of the current forecast time.

[0070] Taking S i,jAs a dynamic evaluation index of institution i at weather station j, the higher the index score, the more accurate the forecast result of the institution at the location of the weather station. When comprehensively evaluating the forecast products later, the weight of the institution is greater.

[0071] After the forecast data of each forecasting institution have been fully accumulated and enriched, the selection of the past time period can be further optimized as a adjustment parameter. In an embodiment of the present invention, the past time period can be extended to the annual scale, or the same seasons, the same months or the same dates of different years can be additionally selected, and the data of the same period in previous years are incorporated into the calculation of the comprehensive score to consider the influence of seasonal changes on the forecast accuracy.

[0072] As Figure 1 shown, in an embodiment of the present invention, step 200 includes:

[0073] Step 210: Determine the pros and cons that support the occurrence of a precipitation event according to the forecast data of each forecasting institution for a specific future time period.

[0074] Those skilled in the art can understand that predicting whether a precipitation event will occur is to predict a switch state, and there are clear conclusions in the forecast data of each forecasting institution. According to the clear conclusions, the forecasting institutions that support the occurrence of the precipitation event can be determined for both the pros and the cons.

[0075] Step 220: Use the sum of the dynamic evaluation indexes of the pros and cons as the arbitration basis to form the forecast result of the precipitation event.

[0076] The dynamic evaluation index reflects the forecast reliability of the forecasting institution in the determined area. The arbitration basis is formed by the following difference of the sums and the forecast of the precipitation event is made:

[0077] ∑ 有雨 S i,j -∑ 无雨 S i,j ≥0, precipitation;

[0078] ∑ 有雨 S i,j -∑ 无雨 S i,j <0, no precipitation.

[0079] Step 230: Form the rainfall weights of each forecasting institution according to the dynamic evaluation index.

[0080] When the forecasting institutions on the positive side have an arbitration advantage, the calculation of the rainfall weights is carried out.

[0081] For the dynamic evaluation indexes of the participating forecasting institutions, the rainfall weights W of each forecasting institution on the positive side are formed by the following formula i,j :

[0082]

[0083] That is, S i,j represents the dynamic evaluation index of the regional agency i, and the denominator is the sum of the dynamic evaluation indexes of all participating agencies. This ensures that the sum of the weights of all agencies is 1, thus reflecting the relative importance of each agency's forecast in the comprehensive rainfall prediction.

[0084] Step 240: Form a rainfall forecast result by comprehensively forecasting data of the positive forecasting agencies according to the rainfall weights.

[0085] Through the rainfall weight W i,j Integrate the predicted rainfall from different forecasting agencies with the following formula to form the weighted rainfall PZ i,j,m :

[0086]

[0087] where P i,j,m represents the predicted rainfall of agency i at the location of weather station j for the next m hours, and the weighted rainfall PZ i,j,m integrates the predictions of different agencies to obtain a more accurate rainfall forecast.

[0088] The short-term and impending precipitation forecasting method based on dynamic weight determination in the embodiments of the present invention uses the dynamics of dynamic evaluation indexes to form dynamic weight determination. It reduces the influence of historical results on the current weight, improves the accuracy of weight evaluation. By continuously adjusting the weight, the difference between the prediction result and the actual observation data is minimized. In short-term and impending precipitation forecasting, the weights of different data sources and forecasting methods can be optimized through a dynamic process, making the comprehensive forecast result more accurate.

[0089] As Figure 1 shown, in an embodiment of the present invention, it further includes:

[0090] Step 300: Form a forecast error quantification and evaluation optimization according to the difference between the forecast and the observation of rainfall.

[0091] By taking the difference between the predicted rainfall and the observed rainfall as the error metric data for measuring the short-term and nowcasting precipitation prediction process in this embodiment, the systematic error or the system loss function is calculated. Commonly used error metric methods for the loss function include the root mean square error (RMSE) and the standard deviation (STD), which can be used to iteratively quantify and reduce the difference between the predicted output of the complete prediction process and the actual precipitation data. Through the error metric, a continuous optimization process for the formation process of the key parameters in the short-term and nowcasting precipitation prediction process in this embodiment is formed. Furthermore, the dynamic optimization and regular evaluation of the prediction model are formed. This helps to timely detect performance bottlenecks and adjust the optimization strategy according to the evaluation results. The evaluation of the model performance can be carried out through various metrics, including but not limited to the accuracy, stability, and reliability of the prediction. By quantifying the error in the prediction process, an iterative optimization of the technical parameters and state parameters in the prediction process is formed until the systematic error or the system loss function stabilizes within an acceptable threshold range.

[0092] An embodiment of the present invention provides a short-term and nowcasting precipitation prediction device based on dynamic weight determination, comprising:

[0093] A memory for storing the program code of the processing process of the short-term and nowcasting precipitation prediction method based on dynamic weight determination in the above embodiment;

[0094] A processor for executing the program code of the processing process of the short-term and nowcasting precipitation prediction method based on dynamic weight determination in the above embodiment.

[0095] The processor may adopt a DSP (Digital Signal Processor) digital signal processor, an FPGA (Field-Programmable Gate Array) field programmable gate array, an MCU (Microcontroller Unit) system board, an SoC (system on a chip) system board, or a minimum system of a PLC (Programmable Logic Controller) including I / O.

[0096] An embodiment of the short-term and nowcasting precipitation prediction device based on dynamic weight determination of the present invention is as Figure 2 shown. In Figure 2 this, this embodiment includes:

[0097] A dynamic evaluation setting module 10 for forming a dynamic evaluation index of the prediction accuracy of each prediction agency in the same region according to the frequency of accurate precipitation prediction in the same region during the past period at the current prediction moment;

[0098] The dynamic weight determination and prediction module 20 is used to arbitrate the prediction data of each prediction agency for a future period at the current prediction time by using dynamic evaluation indicators to form a prediction of precipitation events, dynamically quantify the rainfall weights of the prediction data according to the dynamic evaluation indicators, and form a prediction of rainfall by synthesizing the prediction data according to the rainfall weights.

[0099] As Figure 2 shown, in an embodiment of the present invention, the dynamic evaluation setting module 10 includes:

[0100] The data preprocessing unit 11 is used to normalize the prediction information and observation information of each prediction agency and convert them into prediction data;

[0101] The basic TS score unit 12 is used to form the TS score of accurate precipitation prediction of each prediction agency every day in the same region according to the prediction data in the past period in the same region;

[0102] The TS score statistics unit 13 is used to count the numerical repetition frequency of the TS scores of each prediction agency in the same past period in the same region;

[0103] The evaluation index formation unit 14 is used to form the dynamic evaluation index of each prediction agency in the same past period in the same region according to the TS score, the numerical repetition frequency and the numerical threshold.

[0104] As Figure 2 shown, in an embodiment of the present invention, the dynamic weight determination and prediction module 20 includes:

[0105] The precipitation event classification unit 21 is used to determine the positive and negative sides that support the occurrence of precipitation events according to the prediction data of each prediction agency for a specific future period;

[0106] The precipitation event arbitration unit 22 is used to form a prediction result of precipitation events with the accumulated sum of the dynamic evaluation indicators of the positive and negative sides as the arbitration basis;

[0107] The weight dynamic determination unit 23 is used to form the rainfall weights of each prediction agency according to the dynamic evaluation indicators;

[0108] The weighted rainfall prediction unit 24 is used to form a prediction result of rainfall by synthesizing the prediction data of the positive prediction agency according to the rainfall weights.

[0109] As Figure 2 shown, in an embodiment of the present invention, it further includes:

[0110] The error feedback and optimization module 30 is used to form prediction error quantification and evaluation optimization according to the difference between the prediction and the observation of rainfall.

[0111] As described above, it is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A short-term precipitation forecasting method based on dynamic weighting, characterized in that: include: Based on the frequency of accurate precipitation forecasts in the same area in the past period at the current forecast time, a dynamic evaluation index of the forecast accuracy of each forecast agency in the same area is formed; Dynamic evaluation indicators are used to arbitrate the forecast data of each forecasting agency for the future time period at the current forecast time to form a forecast of precipitation events. The rainfall weights of the forecast data are dynamically quantified based on the dynamic evaluation indicators, and the rainfall forecast is formed based on the rainfall weights and integrated forecast data.

2. The short-term precipitation forecasting method based on dynamic weighting according to claim 1, characterized in that: The formation of the dynamic evaluation index includes: Normalize the forecast information and observation information of each forecast agency and convert them into forecast data; The TS score of each forecasting agency's accurate daily precipitation forecast in the same region is formed based on the forecast data of the same region in the previous period; Count the repetition frequency of TS scores of various forecasting agencies in the same region in the same past period; Based on the TS score, numerical repetition frequency and numerical threshold, dynamic evaluation indicators of each forecasting agency in the same region in the same past period are formed.

3. The short-term precipitation forecasting method based on dynamic weighting according to claim 2 is characterized in that: The previous period is 180 days or a year.

4. The short-term precipitation forecasting method based on dynamic weighting according to claim 3 is characterized in that: The past time periods also include the same season or the same month in different years.

5. The short-term precipitation forecasting method based on dynamic weighting according to claim 1, characterized in that: The forecast of precipitation events includes: Determine the positive and negative sides supporting the occurrence of precipitation events based on the forecast data of each forecast agency for a specific future period; The cumulative sum of positive and negative dynamic evaluation indicators is used as the arbitration basis to form the forecast results of precipitation events.

6. The method for forecasting short-term precipitation based on dynamic weighting according to claim 1, characterized in that: The precipitation forecast comprises: Form rainfall weights for each forecasting agency based on dynamic evaluation indicators; The rainfall forecast result is formed by integrating the forecast data of the square forecasting agency according to the rainfall weight.

7. The short-term precipitation forecasting method based on dynamic weighting according to claim 1, characterized in that: Also includes: The forecast error is quantified and evaluated based on the difference between the forecast and observation of rainfall.

8. A short-term precipitation forecasting device based on dynamic weighting, characterized in that: include: A memory, used to store a program code for a processing process of a method for short-term precipitation forecasting based on dynamic weighting as claimed in any one of claims 1 to 7; A processor is used to execute the program code.

9. A short-term precipitation forecasting device based on dynamic weighting, characterized in that: include: A dynamic evaluation setting module is used to form a dynamic evaluation index of the forecast accuracy of each forecast agency in the same area based on the frequency of accurate precipitation forecasts in the same area in the past period of the current forecast time; The dynamic weighted forecast module is used to arbitrate the forecast data of each forecast agency for the future time period at the current forecast time using dynamic evaluation indicators to form a forecast of precipitation events, dynamically quantify the rainfall weight of the forecast data according to the dynamic evaluation indicators, and form a rainfall forecast based on the rainfall weight integrated forecast data.

10. The short-term precipitation forecasting device based on dynamic weighting according to claim 9, characterized in that: Also includes: The error feedback optimization module is used to quantify and evaluate the forecast error based on the difference between the forecast and observation rainfall.

Images