Short mid-term plane rainfall forecast level evaluation method

Abstract

The invention discloses a short and medium-term surface rainfall forecast level evaluation method, which comprises the following steps of S1, collecting short and medium-term forecast results of different surface rainfall forecast products in recent years in a drainage basin, and sorting the forecast results; s2, formulating level evaluation standards of short-term surface rainfall forecast and mid-term surface rainfall forecast under different magnitudes; s3, comprehensively evaluating short-term and medium-term forecasting levels of different rainfall forecasting products; s4, analyzing interannual changes of forecast levels of different rainfall forecast products, and distinguishing forecast level change trends and application conditions of various forecast products; according to the method, the evaluation accuracy and rationality are considered, and a new thought is provided for the rainfall forecast level evaluation problem.

Description

Technical Field

[0001] The present invention relates to the technical field of hydrology, and particularly to a method for evaluating the level of short - and medium - term areal rainfall forecasting. Background Art

[0002] Rainfall is the main driving factor for flood and drought disasters. Due to measurement errors in meteorological data, imperfections in forecasting methods, and the existence of random factors, rainfall forecasting has uncertainty. The errors of different forecasting methods also vary under different conditions. Accurately quantifying the errors of different forecasting methods and selecting the most suitable forecasting method within the basin for accurate rainfall forecasting is of great significance for flood and drought disaster forecasting and early warning, smart city management, and agricultural water resource allocation.

[0003] With the improvement of the forecasting quality of numerical forecast products such as the synoptic situation field and physical quantity field, numerical rainfall forecast products have been successively developed and put into operational use. However, due to the uneven forecasting performance of various numerical forecast products and the imperfect numerical forecast interpretation and application technology, there are still obvious deviations between the forecast results and the actual situation. Therefore, doing a good job in the inspection, comparison, and analysis of the effects of existing rainfall numerical forecast products is of great significance for further improving the accuracy of rainfall forecasting, especially heavy rainfall forecasting. Currently, commonly used rainfall forecasting evaluation indicators include TS score, false alarm rate, misclassification rate, relative error, mean error, mean absolute error, bias score, root mean square error, and mean bias, etc. These indicators are not obvious enough in reflecting the differences in forecast biases under different rainfall magnitudes and are difficult to comprehensively evaluate the short - and medium - term forecasting levels. Summary of the Invention

[0004] The purpose of the present invention is to overcome the above - mentioned deficiencies and provide a method for evaluating the level of short - and medium - term areal rainfall forecasting. For short - and medium - term rainfall processes of different magnitudes, a scoring system for evaluating the level of short - and medium - term rainfall forecasting is proposed to comprehensively evaluate the short - and medium - term forecasting levels of rainfall forecast products.

[0005] To solve the above - mentioned technical problems, the technical solution adopted by the present invention is: a method for evaluating the level of short - and medium - term areal rainfall forecasting, including the following steps:

[0006] S1. Collect the short - and medium - term forecast results of different areal rainfall forecast products in the basin in recent years and sort out the forecast results;

[0007] S2. Formulate evaluation criteria for short - term areal rainfall forecasting and medium - term areal rainfall forecasting at different magnitudes;

[0008] S3. Comprehensively evaluate the short - and medium - term forecasting levels of different rainfall forecast products;

[0009] S4. Analyze the inter-annual variations in the forecasting levels of different rainfall forecasting products, and distinguish the changing trends and applicable situations of the forecasting levels of various forecasting products.

[0010] Preferably, the S1 is specifically as follows:

[0011] S11. Data collection: Collect the short-term and medium-term forecasting results of different rainfall forecasting products in the basin in recent years.

[0012] S12. Medium-term forecasting result collation: Select the rainfall processes in the basin in recent years where the areal rainfall reaches or exceeds the moderate rainfall level for at least two consecutive days and the corresponding medium-term forecasting results as the medium-term rainfall calculation samples, and select the actual short-term rainfall and short-term rainfall forecasting results within the corresponding time period of the medium-term forecasting results as the short-term rainfall calculation samples to ensure the nesting of short-term forecasting and medium-term forecasting in terms of time scale.

[0013] Preferably, in the S12, the given actual rainfall process is numbered as t n , t n = 1, 2, …, t N ; where N is the total number of actual rainfall processes; the time corresponding to the medium-term forecasting samples is [t n - 6, t n - 3], and the corresponding lead times are [7, 4], and the time corresponding to the short-term forecasting samples is [t n - 2, t n , and the lead times are [3, 1].

[0014] Preferably, the S2 includes:

[0015] S21. Establish the evaluation criteria for short-term daily rainfall forecasts at different magnitudes: Divide the short-term daily rainfall according to the rainfall amount. Rainfall less than 9.9 mm is light rain; rainfall of 10 - 24.9 mm is moderate rain; rainfall of 25 - 49.9 mm is heavy rain; rainfall of 50 - 99.9 mm is rainstorm; rainfall above 100 mm is heavy rainstorm or extreme rainstorm; establish the evaluation criteria for short-term rainfall at different rainfall levels respectively.

[0016] More preferably, the specific establishment of the evaluation criteria for short-term rainfall at different rainfall levels in the S21 is as follows:

[0017] S211. Let R1 be the short-term actual rainfall, and R x be the short-term forecast rainfall, and the units of R1 and R x are both mm; denote different forecast rainfall ranges as M i , i = 1, 2, Λ, 11; where M1 is R x = 0; M2 is 0 < R x < 4.9; M3 is 5 < R x< 9.9; M4 is 5 < R x < 14.9; M5 is 10 < R x < 19.9; M6 is 15 < R x < 24.9; M7 is 20 < R x < 39.9; M8 is 30 < R x < 49.9; M9 is 40 < R x < 59.9; M 10 is 50 < R x < 99.9; M 11 is R x > 100; Let S1 be the short-term forecast score; then the short-term rainfall forecast evaluation criteria under different magnitudes are as follows:

[0018] S2111. Light rain forecast evaluation criteria:

[0019] When R1 = 0, if i = 1, S1 = 100, if i = 2, S1 = 80, if i > 2, S1 = 0;

[0020] When 0 < R1 < 4.9, if i = 2, S1 = 100, if i = 3, S1 = 90, if i = 1, 4, 5, S1 = 80, if i = 6, S1 = 60, if i > 6, S1 = 0;

[0021] When 5 < R1 < 9.9, if i = 3, 4, S1 = 100, if i = 5, S1 = 90, if i = 2, 6, S1 = 80, if i = 7, S1 = 60, if i > 7 or i = 1, S1 = 0;

[0022] S2112. Moderate rain forecast evaluation criteria:

[0023] When 10 < R1 < 14.9, if i = 4, 5, S1 = 100, if i = 3, 6, S1 = 90, if i = 7, S1 = 80, if i = 2, 8, S1 = 60, if i > 8 or i = 1, S1 = 0;

[0024] When 15 < R1 < 19.9, if i = 5, 6, S1 = 100, if i = 4, 7, S1 = 90, if i = 3, 8, S1 = 80, if i = 9, S1 = 60, if i = 2, S1 = 40, if i > 9 or i = 1, S1 = 0;

[0025] When 20 < R1 < 24.9, if i = 6, 7, S1 = 100, if i = 5, 8, S1 = 90, if i = 4, 9, S1 = 80, if i = 3, 10, S1 = 60, if i < 3 or i = 11, S1 = 0;

[0026] S2113. Heavy rain forecast evaluation criteria:

[0027] When 25 < R1 < 29.9, if i = 7, S1 = 100; if i = 5, 6, 8, 9, S1 = 90; if i = 10, S1 = 80; if i = 4, S1 = 60; if i = 3, 11, S1 = 40; if i < 3, S1 = 0;

[0028] When 30 < R1 < 39.9, if i = 7, 8, S1 = 100; if i = 6, 9, 10, S1 = 90; if i = 5, S1 = 80; if i = 11, S1 = 60; if i = 4, S1 = 40; if i < 4, S1 = 0;

[0029] When 40 < R1 < 49.9, if i = 8, 9, S1 = 100; if i = 7, 10, S1 = 90; if i = 6, 11, S1 = 80; if i = 5, S1 = 60; if i < 5, S1 = 0;

[0030] S2114. Evaluation criteria for rainstorm forecasts:

[0031] When 50 < R1 < 59.9, if i = 9, 10, S1 = 100; if i = 7, 8, 11, S1 = 90; if i = 6, S1 = 60; if i = 5, S1 = 40; if i < 5, S1 = 0;

[0032] When 60 < R1 < 79.9, if i = 10, S1 = 100; if i = 8, 9, 11, S1 = 90; if i = 7, S1 = 80; if i = 6, S1 = 40; if i < 6, S1 = 0;

[0033] When 80 < R1 < 99.9, if i = 10, 11, S1 = 100; if i = 9, S1 = 90; if i = 7, 8, S1 = 80; if i = 6, S1 = 20; if i < 6, S1 = 0;

[0034] S2115. Evaluation criteria for heavy rainstorm or extremely heavy rainstorm forecasts:

[0035] When R1 > 100, if i = 11, S1 = 100; if i = 9, 10, S1 = 90; if i = 8, S1 = 80; if i = 7, S1 = 60; if i < 7, S1 = 0.

[0036] More preferably, the S2 further includes:

[0037] S22. Formulate evaluation criteria for medium-term daily rainfall forecasts at different levels: Light rainfall: The total rainfall is between 25 mm and 49.9 mm; Moderate rainfall: The total rainfall is between 50 mm and 99.9 mm; Heavy rainfall: The total rainfall is between 100 mm and 249.9 mm; Extremely heavy rainfall: The total rainfall is above 250 mm; Formulate evaluation criteria for medium-term rainfall at different rainfall levels respectively;

[0038] More preferably, the evaluation criteria for medium-term rainfall of different rainfall levels in S22 are specifically as follows:

[0039] S221. Let R2 be the actual medium-term rainfall, and R y be the medium-term predicted rainfall, with the units of R2 and R y both being mm; different predicted rainfall ranges are denoted as N j , j = 1, 2, Λ, 9; where N1 is 0 < R y < 19.9; N2 is 10 < R y < 34.9; N3 is 20 < R y < 39.9; N4 is 30 < R y < 54.9; N5 is 40 < R y < 69.9; N6 is 50 < R y < 79.9; N7 is 70 < R y < 99.9; N8 is 100 < R y < 249.9; N9 is R y > 250; let S2 be the medium-term prediction score, then the evaluation criteria for medium rainfall prediction under different magnitudes are:

[0040] S2211. Evaluation criteria for light rainfall prediction:

[0041] When 25 < R2 < 29.9, if j = 2, 3, S2 = 100, if j = 1, 4, S2 = 80, if j = 5, S2 = 60, if j = 6, S2 = 20, if j > 6, S2 = 0;

[0042] When 30 < R2 < 39.9, if j = 4, S2 = 100, if j = 3, S2 = 90, if j = 2, 5, S2 = 80, if j = 6, S2 = 60, if j = 1, S2 = 20, if j > 6, S2 = 0;

[0043] When 40 < R2 < 49.9, if j = 4, 5, S2 = 100, if j = 3, 6, S2 = 80, if j = 2, S2 = 60, if j = 7, S2 = 40, if j > 7 or j = 1, S2 = 0;

[0044] S2212. Evaluation criteria for moderate rainfall prediction:

[0045] When 50 < R2 < 69.9, if j = 5, 6, S2 = 100, if j = 4, S2 = 90, if j = 7, S2 = 80, if j = 3, 8, S2 = 40, if j = 1, 2, 9, S2 = 0;

[0046] When 70 < R2 < 99.9, if j = 7, S2 = 100; if j = 6, S2 = 90; if j = 8, S2 = 80; if j = 5, S2 = 60; if j = 4 or 9, S2 = 40; if j < 4, S2 = 0.

[0047] S2213, Evaluation criteria for heavy rainfall forecasts:

[0048] When 100 < R2 < 249.9, if j = 8, S2 = 100; if j = 7 or 9, S2 = 80; if j = 6, S2 = 40; if j = 5, S2 = 20; if j < 5, S2 = 0.

[0049] S2214, Evaluation criteria for extremely heavy rainfall forecasts:

[0050] When R2 > 250, if j = 9, S2 = 100; if j = 8, S2 = 80; if j = 7, S2 = 40; if j = 6, S2 = 20; if j < 6, S2 = 0.

[0051] Preferably, the specific value of S3 is as follows:

[0052] S31, Evaluation of short-term rainfall forecast level: Calculate the average value of the short-term rainfall forecast scores S1 corresponding to each forecast product. The larger the value, the better the short-term rainfall forecast effect of the corresponding forecast product.

[0053] S32, Evaluation of medium-term rainfall forecast level: Calculate the average value of the medium-term rainfall forecast scores S2 corresponding to each forecast product. The larger the value, the better the medium-term rainfall forecast effect of the corresponding forecast product.

[0054] S33, Comprehensive evaluation of short- and medium-term forecast levels: Take The average value as the comprehensive score of the short- and medium-term forecast levels The larger the value, the better the short- and medium-term rainfall forecast effect of the corresponding forecast product.

[0055] Preferably, S4 includes:

[0056] S41, Analyze the interannual variation of the forecast levels of different rainfall forecast products. Divide the historical actual rainfall data, short-term rainfall forecast data, and medium-term rainfall forecast data by year. Calculate the average values of the short-term rainfall forecast scores for each year of different rainfall forecast products The average values of the medium-term rainfall forecast scores And the comprehensive scores of the short- and medium-term forecast levels

[0057]

[0058] More preferably, S4 further includes:

[0059] S42. Analyze the inter-annual variations in the forecasting levels of different rainfall forecasting products, and discuss the changing trends of the forecasting levels of various forecasting products and their applicable situations.

[0060] Advantages of the present invention: The method of the present invention proposes a short- and medium-term rainfall forecasting level evaluation score system for short- and medium-term rainfall processes of different magnitudes, and analyzes the inter-annual changing trends of the forecasting levels of different rainfall forecasting products and their applicable situations. This method takes into account both evaluation accuracy and rationality, and provides a new idea for the problem of evaluating rainfall forecasting levels. Detailed implementation manners

[0061] The present invention will be further described in detail below in conjunction with specific embodiments.

[0062] Embodiment 1: A method for evaluating the short- and medium-term areal rainfall forecasting level, comprising the following steps:

[0063] S1. Taking the upper reaches of the Yangtze River as an example, collect the short- and medium-term forecasting results of different rainfall forecasting products in the upper reaches of the Yangtze River Basin in recent years, and organize the forecasting results.

[0064] S11. Data collection: Collect the short-term and medium-term forecasting results of different rainfall forecasting products in the upper reaches of the Yangtze River Basin in recent years.

[0065] S12. Organization of medium-term forecasting results: Select the rainfall processes in which the areal rainfall in the upper reaches of the Yangtze River Basin reaches or exceeds the moderate rainfall level for at least two consecutive days in recent years and the corresponding medium-term forecasting results as the medium-term rainfall calculation samples, and select the actual short-term rainfall amounts and short-term rainfall forecasting results within the medium-term forecasting period as the short-term rainfall calculation samples.

[0066] Number the given actual rainfall process as t n , t n = 1, 2, …, t N ; where N is the total number of actual rainfall processes; the time corresponding to the medium-term forecasting samples is [t n - 6, t n - 3], and the corresponding lead times are [7, 4], and the time corresponding to the short-term forecasting samples is [t n - 2, t n , and the lead times are [3, 1].

[0067] S2. Formulate the evaluation criteria for short-term rainfall forecasting and medium-term rainfall forecasting levels under different magnitudes.

[0068] S21. Establish the evaluation criteria for short-term daily rainfall forecasts at different magnitudes: Classify short-term rainfall according to the rainfall amount. Rainfall less than 9.9 mm is light rain; rainfall of 10 - 24.9 mm is moderate rain; rainfall of 25 - 49.9 mm is heavy rain; rainfall of 50 - 99.9 mm is rainstorm; rainfall above 100 mm is heavy rainstorm or extreme rainstorm; establish evaluation criteria for short-term rainfall at different rainfall levels respectively;

[0069] S211. Let R1 be the short-term actual rainfall amount, and R x be the short-term forecast rainfall amount. The units of R1 and R x are both mm. Denote different forecast rainfall intervals as M i , where i = 1, 2, Λ, 11; among them, M1 is when R x = 0; M2 is when 0 < R x < 4.9; M3 is when 5 < R x < 9.9; M4 is when 5 < R x < 14.9; M5 is when 10 < R x < 19.9; M6 is when 15 < R x < 24.9; M7 is when 20 < R x < 39.9; M8 is when 30 < R x < 49.9; M9 is when 40 < R x < 59.9; M 10 is when 50 < R x < 99.9; M 11 is when R x > 100; Let S1 be the short-term forecast score. Then the evaluation criteria for short-term rainfall forecasts at different magnitudes are:

[0070] S2111. Evaluation criteria for light rain forecasts:

[0071] When R1 = 0, if i = 1, S1 = 100; if i = 2, S1 = 80; if i > 2, S1 = 0;

[0072] When 0 < R1 < 4.9, if i = 2, S1 = 100; if i = 3, S1 = 90; if i = 1, 4, 5, S1 = 80; if i = 6, S1 = 60; if i > 6, S1 = 0;

[0073] When 5 < R1 < 9.9, if i = 3, 4, S1 = 100; if i = 5, S1 = 90; if i = 2, 6, S1 = 80; if i = 7, S1 = 60; if i > 7 or i = 1, S1 = 0;

[0074] S2112. Evaluation criteria for moderate rain forecasts:

[0075] When 10 < R1 < 14.9, if i = 4, 5, S1 = 100; if i = 3, 6, S1 = 90; if i = 7, S1 = 80; if i = 2, 8, S1 = 60; if i > 8 or i = 1, S1 = 0;

[0076] When 15 < R1 < 19.9, if i = 5, 6, S1 = 100; if i = 4, 7, S1 = 90; if i = 3, 8, S1 = 80; if i = 9, S1 = 60; if i = 2, S1 = 40; if i > 9 or i = 1, S1 = 0;

[0077] When 20 < R1 < 24.9, if i = 6, 7, S1 = 100; if i = 5, 8, S1 = 90; if i = 4, 9, S1 = 80; if i = 3, 10, S1 = 60; if i < 3 or i = 11, S1 = 0;

[0078] S2113. Heavy rain forecast evaluation criteria:

[0079] When 25 < R1 < 29.9, if i = 7, S1 = 100; if i = 5, 6, 8, 9, S1 = 90; if i = 10, S1 = 80; if i = 4, S1 = 60; if i = 3, 11, S1 = 40; if i < 3, S1 = 0;

[0080] When 30 < R1 < 39.9, if i = 7, 8, S1 = 100; if i = 6, 9, 10, S1 = 90; if i = 5, S1 = 80; if i = 11, S1 = 60; if i = 4, S1 = 40; if i < 4, S1 = 0;

[0081] When 40 < R1 < 49.9, if i = 8, 9, S1 = 100; if i = 7, 10, S1 = 90; if i = 6, 11, S1 = 80; if i = 5, S1 = 60; if i < 5, S1 = 0;

[0082] S2114. Rainstorm forecast evaluation criteria:

[0083] When 50 < R1 < 59.9, if i = 9, 10, S1 = 100; if i = 7, 8, 11, S1 = 90; if i = 6, S1 = 60; if i = 5, S1 = 40; if i < 5, S1 = 0;

[0084] When 60 < R1 < 79.9, if i = 10, S1 = 100; if i = 8, 9, 11, S1 = 90; if i = 7, S1 = 80; if i = 6, S1 = 40; if i < 6, S1 = 0;

[0085] When 80 < R1 < 99.9, if i = 10, 11, S1 = 100; if i = 9, S1 = 90; if i = 7, 8, S1 = 80; if i = 6, S1 = 20; if i < 6, S1 = 0;

[0086] S2115. Forecast evaluation criteria for heavy rainstorms or extremely heavy rainstorms:

[0087] When R1 > 100, if i = 11, S1 = 100; if i = 9, 10, S1 = 90; if i = 8, S1 = 80; if i = 7, S1 = 60; if i < 7, S1 = 0;

[0088] S22. Formulate the forecast evaluation criteria for medium - term daily rainfall under different magnitudes: Light rainfall: The total rainfall is between 25 mm and 49.9 mm; Moderate rainfall: The total rainfall is between 50 mm and 99.9 mm; Heavy rainfall: The total rainfall is between 100 mm and 249.9 mm; Extremely heavy rainfall: The total rainfall is above 250 mm; Formulate evaluation criteria for medium - term rainfall at different rainfall levels respectively;

[0089] S221. Let R2 be the actual medium - term rainfall and R y be the forecast medium - term rainfall, and the units of R2 and R y are both mm. Denote different forecast rainfall ranges as N j , j = 1, 2, Λ, 9; among them, N1 is 0 < R y < 19.9; N2 is 10 < R y < 34.9; N3 is 20 < R y < 39.9; N4 is 30 < R y < 54.9; N5 is 40 < R y < 69.9; N6 is 50 < R y < 79.9; N7 is 70 < R y < 99.9; N8 is 100 < R y < 249.9; N9 is R y > 250; Let S2 be the medium - term forecast score, then the forecast evaluation criteria for medium - term rainfall under different magnitudes are:

[0090] S2211. Forecast evaluation criteria for light rainfall:

[0091] When 25 < R2 < 29.9, if j = 2, 3, S2 = 100; if j = 1, 4, S2 = 80; if j = 5, S2 = 60; if j = 6, S2 = 20; if j > 6, S2 = 0;

[0092] When 30 < R2 < 39.9, if j = 4, S2 = 100; if j = 3, S2 = 90; if j = 2 or 5, S2 = 80; if j = 6, S2 = 60; if j = 1, S2 = 20; if j > 6, S2 = 0;

[0093] When 40 < R2 < 49.9, if j = 4 or 5, S2 = 100; if j = 3 or 6, S2 = 80; if j = 2, S2 = 60; if j = 7, S2 = 40; if j > 7 or j = 1, S2 = 0;

[0094] S2212. Medium rainfall forecast evaluation criteria:

[0095] When 50 < R2 < 69.9, if j = 5 or 6, S2 = 100; if j = 4, S2 = 90; if j = 7, S2 = 80; if j = 3 or 8, S2 = 40; if j = 1, 2 or 9, S2 = 0;

[0096] When 70 < R2 < 99.9, if j = 7, S2 = 100; if j = 6, S2 = 90; if j = 8, S2 = 80; if j = 5, S2 = 60; if j = 4 or 9, S2 = 40; if j < 4, S2 = 0;

[0097] S2213. Heavy rainfall forecast evaluation criteria:

[0098] When 100 < R2 < 249.9, if j = 8, S2 = 100; if j = 7 or 9, S2 = 80; if j = 6, S2 = 40; if j = 5, S2 = 20; if j < 5, S2 = 0;

[0099] S2214. Extremely heavy rainfall forecast evaluation criteria:

[0100] When R2 > 250, if j = 9, S2 = 100; if j = 8, S2 = 80; if j = 7, S2 = 40; if j = 6, S2 = 20; if j < 6, S2 = 0;

[0101] S3. Conduct a comprehensive evaluation of the short - and medium - term forecast levels for different rainfall forecast products;

[0102] S31. Evaluation of short - term rainfall forecast level: Calculate the average value S1 of the short - term rainfall forecast scores S1 corresponding to each forecast product. The larger S1 is, the better the short - term rainfall forecast effect of the corresponding forecast product. Taking forecast products 1 and 2 as examples, the average value S1 of the short - term forecast scores is shown in Table 1.

[0103] Table 1 Short - term forecast score table for forecast products 1 and 2

[0104]

[0105] S32. Medium-term rainfall forecast level evaluation: Calculate the average value of the medium-term rainfall forecast scores S2 corresponding to each forecast product. The larger it is, the better the medium-term rainfall forecast effect of the corresponding forecast product. The average values of the medium-term forecast scores of forecast products 1 and 2 are shown in Table 2.

[0106] Table 2 Medium-term forecast score table of forecast products 1 and 2

[0107]

[0108] S33. Comprehensive evaluation of short- and medium-term forecast levels: Take the mean value as the comprehensive score of the short- and medium-term forecast level The larger it is, the better the short- and medium-term rainfall forecast effect of the corresponding forecast product.

[0109] S4. Analyze the interannual variations in the forecast levels of different rainfall forecast products, and discuss the changing trends and applicable situations of the forecast levels of various forecast products.

[0110] S41. Analyze the interannual variations in the forecast levels of different rainfall forecast products. Divide the historical actual rainfall data, short-term rainfall forecast data, and medium-term rainfall forecast data by year. Calculate the mean values of the short-term rainfall forecast scores for each year of different rainfall forecast products the mean values of the medium-term rainfall forecast scores and the comprehensive scores of the short- and medium-term forecast levels The average values of the short- and medium-term forecast scores of forecast products 1 and 2 are shown in Table 3.

[0111] Table 3 Short- and medium-term forecast score table of forecast products 1 and 2

[0112]

[0113] S42. Analyze the interannual variation of the forecasting levels of different rainfall forecasting products, and discuss the changing trends of the forecasting levels of various forecasting products and their applicable situations. In terms of short-term forecasting, from 2012 to 2023, the fluctuation range of the short-term forecasting level of Forecasting Product 1 is smaller than that of Forecasting Product 2. In 2013-2014, the forecasting level of Forecasting Product 2 was significantly lower than that of Forecasting Product 1, and in 2015, the forecasting level of Forecasting Product 2 was significantly higher than that of Forecasting Product 1. Overall, the short-term forecasting effect of Forecasting Product 1 is more stable, but the upper limit of the forecasting accuracy of Forecasting Product 2 is higher. In terms of medium-term forecasting, there is no obvious advantage or disadvantage in the overall medium-term forecasting levels of Forecasting Product 1 and Forecasting Product 2. The medium-term forecasting level of Forecasting Product 1 was higher than that of Forecasting Product 2 in 6 years and lower than that of Forecasting Product 2 in 4 years from 2012 to 2023. In terms of the comprehensive evaluation of short- and medium-term forecasting, generally speaking, the short- and medium-term forecasting level of Forecasting Product 1 is more stable, and the short- and medium-term forecasting scores are all above 90 during 2012-2023. In summary, in this embodiment, the short- and medium-term forecasting level of Forecasting Product 1 is higher.

[0114] The above embodiments are only the preferred technical solutions of the present invention and should not be regarded as limitations on the present invention. The protection scope of the present invention shall be the technical solutions recorded in the claims, including the equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, the equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. A method for evaluating the level of short- and medium-term surface rainfall forecasting, characterized in that: It includes the following steps: S1. Collect the short-term and medium-term forecast results of different areal rainfall forecast products in the basin in recent years, and organize the forecast results. S2. Formulate the evaluation criteria for short-term areal rainfall forecasts and medium-term areal rainfall forecast levels under different magnitudes. S3. Conduct a comprehensive evaluation of the short-term and medium-term forecast levels of different rainfall forecast products. S4. Analyze the inter-annual changes in the forecast levels of different rainfall forecast products, and distinguish the changing trends and applicable situations of the forecast levels of various forecast products.

2. A short- and medium-term surface rainfall forecast level evaluation method according to claim 1, characterized in that: The specific content of S1 is as follows: S11. Data collection: Collect the short-term and medium-term forecast results of different rainfall forecast products in the basin in recent years. S12. Medium-term forecast result collation: Select the rainfall processes in the basin in recent years where the areal rainfall has reached or exceeded the moderate rainfall level for at least two consecutive days and the corresponding medium-term forecast results as the medium-term rainfall calculation samples, and select the actual short-term rainfall and short-term rainfall forecast results within the corresponding time period of the medium-term forecast results as the short-term rainfall calculation samples to ensure the nesting of short-term forecasts and medium-term forecasts in terms of time scale.

3. A short- and medium-term surface rainfall prediction level evaluation method according to claim 1, characterized in that: In S12, the given actual rainfall process is numbered as t n , t n = 1, 2, …, t N ; where N is the total number of actual rainfall processes; the time corresponding to the medium-term forecast sample is [t n - 6, t n - 3], and the corresponding lead times are [7, 4]; the time corresponding to the short-term forecast sample is [t n - 2, t n , and the lead times are [3, 1].

4. A short- and medium-term surface rainfall forecast level evaluation method according to claim 1, characterized in that: S2 includes: S21. Formulate the evaluation criteria for short-term daily rainfall forecasts under different magnitudes: Divide the short-term daily rainfall according to the rainfall amount. Rainfall less than 9.9 mm is light rain; rainfall of 10 - 24.9 mm is moderate rain; rainfall of 25 - 49.9 mm is heavy rain; rainfall of 50 - 99.9 mm is rainstorm; rainfall above 100 mm is heavy rainstorm or extreme rainstorm; formulate evaluation criteria for short-term rainfall of different rainfall levels respectively.

5. A method for evaluating the short- and medium-term surface rainfall forecasting level according to claim 4, characterized in that: The specific content of formulating the evaluation criteria for short-term rainfall of different rainfall levels in S21 is as follows: S211. Let R1 be the short-term actual rainfall, and R x be the short-term forecast rainfall. The units of R1 and R x are both mm. Denote different forecast rainfall intervals as M i , where i = 1, 2, Λ, 11. Among them, M1 is when R x = 0; M2 is when 0 < R x < 4.9; M3 is when 5 < R x < 9.9; M4 is when 5 < R x < 14.9; M5 is when 10 < R x < 19.9; M6 is when 15 < R x < 24.9; M7 is when 20 < R x < 39.9; M8 is when 30 < R x < 49.9; M9 is when 40 < R x < 59.9; M 10 is when 50 < R x < 99.9; M 11 is when R x > 100. Let S1 be the short-term forecast score. Then the short-term rainfall forecast evaluation criteria at different magnitudes are as follows: S2111. Evaluation criteria for light rain forecasts: When R1 = 0, if i = 1, S1 = 100; if i = 2, S1 = 80; if i > 2, S1 = 0. When 0 < R1 < 4.9, if i = 2, S1 = 100; if i = 3, S1 = 90; if i = 1, 4, 5, S1 = 80; if i = 6, S1 = 60; if i > 6, S1 = 0. When 5 < R1 < 9.9, if i = 3, 4, S1 = 100; if i = 5, S1 = 90; if i = 2, 6, S1 = 80; if i = 7, S1 = 60; if i > 7 or i = 1, S1 = 0. S2112. Evaluation criteria for moderate rain forecasts: When 10 < R1 < 14.9, if i = 4, 5, S1 = 100; if i = 3, 6, S1 = 90; if i = 7, S1 = 80; if i = 2, 8, S1 = 60; if i > 8 or i = 1, S1 = 0. When 15 < R1 < 19.9, if i = 5, 6, S1 = 100; if i = 4, 7, S1 = 90; if i = 3, 8, S1 = 80; if i = 9, S1 = 60; if i = 2, S1 = 40; if i > 9 or i = 1, S1 = 0. When 20 < R1 < 24.9, if i = 6, 7, S1 = 100; if i = 5, 8, S1 = 90; if i = 4, 9, S1 = 80; if i = 3, 10, S1 = 60; if i < 3 or i = 11, S1 = 0. S2113. Evaluation criteria for heavy rain forecasts: When 25 < R1 < 29.9, if i = 7, S1 = 100; if i = 5, 6, 8, 9, S1 = 90; if i = 10, S1 = 80; if i = 4, S1 = 60; if i = 3, 11, S1 = 40; if i < 3, S1 = 0; When 30 < R1 < 39.9, if i = 7, 8, S1 = 100; if i = 6, 9, 10, S1 = 90; if i = 5, S1 = 80; if i = 11, S1 = 60; if i = 4, S1 = 40; if i < 4, S1 = 0; When 40 < R1 < 49.9, if i = 8, 9, S1 = 100; if i = 7, 10, S1 = 90; if i = 6, 11, S1 = 80; if i = 5, S1 = 60; if i < 5, S1 = 0; S2114. Heavy rain forecast evaluation criteria: When 50 < R1 < 59.9, if i = 9, 10, S1 = 100; if i = 7, 8, 11, S1 = 90; if i = 6, S1 = 60; if i = 5, S1 = 40; if i < 5, S1 = 0; When 60 < R1 < 79.9, if i = 10, S1 = 100; if i = 8, 9, 11, S1 = 90; if i = 7, S1 = 80; if i = 6, S1 = 40; if i < 6, S1 = 0; When 80 < R1 < 99.9, if i = 10, 11, S1 = 100; if i = 9, S1 = 90; if i = 7, 8, S1 = 80; if i = 6, S1 = 20; if i < 6, S1 = 0; S2115. Severe rainstorm or extremely severe rainstorm forecast evaluation criteria: When R1 > 100, if i = 11, S1 = 100; if i = 9, 10, S1 = 90; if i = 8, S1 = 80; if i = 7, S1 = 60; if i < 7, S1 = 0.

6. A method for evaluating the short- and medium-term surface rainfall forecasting level according to claim 4, characterized in that: The said S2 further includes: S22. Formulate medium - term daily rainfall forecast evaluation criteria for different magnitudes: Light rainfall: total rainfall is between 25mm and 49.9mm; Moderate rainfall: total rainfall is between 50mm and 99.9mm; Heavy rainfall: total rainfall is between 100mm and 249.9mm; Extremely heavy rainfall: total rainfall is above 250mm; Formulate evaluation criteria for medium - term rainfall of different rainfall levels respectively.

7. A short- and medium-term surface rainfall forecast level evaluation method according to claim 6, characterized in that: The specific evaluation criteria for medium - term rainfall of different rainfall levels formulated in S22 are as follows: S221. Let R2 be the actual rainfall in the medium term, and R y be the predicted rainfall in the medium term. The units of R2 and R y are both mm. Denote different predicted rainfall ranges as N j , where j = 1, 2, Λ, 9. Among them, N1 is 0 < R y < 19.9; N2 is 10 < R y < 34.9; N3 is 20 < R y < 39.9; N4 is 30 < R y < 54.9; N5 is 40 < R y < 69.9; N6 is 50 < R y < 79.9; N7 is 70 < R y < 99.9; N8 is 100 < R y < 249.9; N9 is R y > 250. Let S2 be the medium-term prediction score. Then the evaluation criteria for medium rainfall prediction under different magnitudes are as follows: S2211. Light rainfall forecast evaluation criteria: When 25 < R2 < 29.9, if j = 2, 3, S2 = 100; if j = 1, 4, S2 = 80; if j = 5, S2 = 60; if j = 6, S2 = 20; if j > 6, S2 = 0; When 30 < R2 < 39.9, if j = 4, S2 = 100; if j = 3, S2 = 90; if j = 2, 5, S2 = 80; if j = 6, S2 = 60; if j = 1, S2 = 20; if j > 6, S2 = 0; When 40 < R2 < 49.9, if j = 4, 5, S2 = 100; if j = 3, 6, S2 = 80; if j = 2, S2 = 60; if j = 7, S2 = 40; if j > 7 or j = 1, S2 = 0; S2212. Medium rainfall forecast evaluation criteria: When 50 < R2 < 69.9, if j = 5, 6, S2 = 100; if j = 4, S2 = 90; if j = 7, S2 = 80; if j = 3, 8, S2 = 40; if j = 1, 2, 9, S2 = 0; When 70 < R2 < 99.9, if j = 7, S2 = 100; if j = 6, S2 = 90; if j = 8, S2 = 80; if j = 5, S2 = 60; if j = 4, 9, S2 = 40; if j < 4, S2 = 0; S2213. Heavy rainfall forecast evaluation criteria: When 100 < R2 < 249.9, if j = 8, S2 = 100; if j = 7, 9, S2 = 80; if j = 6, S2 = 40; if j = 5, S2 = 20; if j < 5, S2 = 0; S2214. Extremely heavy rainfall forecast evaluation criteria: When R2 > 250, if j = 9, S2 = 100; if j = 8, S2 = 80; if j = 7, S2 = 40; if j = 6, S2 = 20; if j < 6, S2 = 0.

8. A method for evaluating the level of short- and medium-term surface rainfall forecasting according to claim 1, characterized in that: The specific S3 is as follows: S31. Short-term rainfall forecast level evaluation: Calculate the average value of the short-term rainfall forecast scores S1 corresponding to each forecast product The larger it is, the better the short-term rainfall forecast effect of the corresponding forecast product is proved; S32. Medium-term rainfall forecast level evaluation: Calculate the average value of the medium-term rainfall forecast scores S2 corresponding to each forecast product The larger it is, the better the medium-term rainfall forecast effect of the corresponding forecast product is proved; S33. Comprehensive evaluation of short- and medium-term forecast level: Take the mean value as the comprehensive score of the short- and medium-term forecast level The larger it is, the better the short- and medium-term rainfall forecast effect of the corresponding forecast product is proven.

9. A short- and medium-term surface rainfall forecast level evaluation method according to claim 1, characterized in that: The S4 includes: S41. Analyze the interannual variation of the forecasting levels of different rainfall forecasting products, and divide the historical actual rainfall data, short-term rainfall forecasting data, and medium-term rainfall forecasting data by year; calculate the average short-term rainfall forecasting scores for each year of different rainfall forecasting products respectively Average medium-term rainfall forecasting score and the comprehensive score of short- and medium-term forecasting levels 10. A method for evaluating the level of short- and medium-term surface rainfall forecasting, according to claim 9, characterized in that: The S4 also includes: S42. Analyze the interannual variations of the forecast levels of different rainfall forecast products, and discuss the changing trends of the forecast levels of various forecast products and their applicable situations.

Images