Dry-heat composite event identification method under influence of climate change

Abstract

The invention relates to the technical field of extreme climate recognition, in particular to a dry-heat composite event recognition method under the influence of climate change, which comprises the following steps of: collecting temperature and humidity data of a current meteorological station, screening the temperature and humidity data, and removing any abnormal value or data missing condition to obtain cleaned real-time meteorological data. According to the method, real-time temperature data and humidity data are collected and screened, the real-time data and historical data in the same period are matched and compared, climate deviation characteristics and historical trends of the dry heat events are comprehensively analyzed, the accurate dry heat event condition likelihood ratio is scientifically calculated, a dynamic threshold mechanism for the real-time environment condition change trend is constructed, and the real-time environment condition change trend is determined. The accuracy of meteorological condition change analysis and the timeliness of dry-heat composite event identification are improved, and more accurate and real-time dry-heat composite event monitoring and early warning are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of extreme climate identification, and in particular to a method for identifying dry-hot composite events under the influence of climate change. Background Art

[0002] Global warming has exacerbated the instability of the climate system, namely, the emergence of complex and concurrent extreme climate events composed of multiple events, whose impact is greater than the sum of the impacts of individual events. Among them, complex extreme dry heat events in which high temperature and drought occur simultaneously and in the same place in inland areas occur more frequently.

[0003] In the existing technology, the complex and nonlinear interaction between environmental parameters and historical data is not fully considered, which makes it easy for event identification and judgment to be biased, resulting in insufficient sensitivity and accuracy of early warning, especially in the case of rapid changes or sudden abnormalities in real-time climate data, resulting in delays in the implementation of actual emergency measures or misjudgment. Therefore, improvements are needed. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings in the prior art and propose a method for identifying dry-heat composite events under the influence of climate change.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: a method for identifying dry-heat composite events under the influence of climate change, comprising the following steps: Collect temperature and humidity data of the current weather station, filter the temperature and humidity data, remove any abnormal values ​​or data missing, and obtain cleaned real-time weather data; perform statistical analysis on the cleaned real-time weather data, calculate the deviation value from the historical data of the same period, and obtain the deviation analysis result; Based on the deviation analysis results, the probability ratio of the occurrence and non-occurrence of the dry heat event under the current meteorological conditions is calculated to obtain a conditional likelihood ratio; using the conditional likelihood ratio, combined with a set threshold, it is determined whether the conditions for the occurrence of a dry heat composite event are currently met, and a dry heat composite event determination result is generated; Dynamically adjust the dry-heat composite event determination result, analyze the current climate trend, adjust the warning sensitivity, and obtain an adjusted threshold; based on the adjusted threshold, re-evaluate the current dry-heat composite event and generate updated dry-heat composite event information; Based on the updated dry-heat composite event information, response measures are determined, including evacuation suggestions and decision support information for emergency resource allocation, to obtain a decision suggestion output.

[0006] Preferably, the steps for obtaining the cleaned real-time meteorological data are: Compare the real-time collected temperature data and humidity data one by one to see if they are within the extreme value range, remove abnormal temperature data and abnormal humidity data that are out of range, and generate real-time meteorological data after removing abnormal values; Based on the real-time meteorological data after removing the abnormal values, determine whether the temperature data and the humidity data are missing or incompletely recorded, remove the temperature data and the humidity data with missing data or incompletely recorded, and generate the real-time meteorological data after removing the missing data; Based on the real-time meteorological data after removing missing data, the temperature data and the humidity data are sorted in chronological order and the consistency of the data is verified to generate the cleaned real-time meteorological data.

[0007] Preferably, the steps of obtaining the deviation analysis result are: Call the temperature data and humidity data in the cleaned real-time meteorological data, call the temperature data and humidity data in the historical meteorological data of the same period, establish the corresponding relationship between the real-time and historical data of the same period one by one, perform the matching check of the real-time and historical temperature data and humidity data item by item, remove the data items that have not been matched successfully, and form a meteorological data pair after successful matching; According to the meteorological data pairs after successful matching, the deviation analysis results are calculated using the following formula: ; in, represents the deviation analysis results, Represents the temperature data in the real-time meteorological data after cleaning, Represents the temperature data in the historical meteorological data for the same period. Represents the humidity data in the real-time meteorological data after cleaning, Represents the humidity data in the historical meteorological data for the same period.

[0008] Preferably, the step of obtaining the conditional likelihood ratio is: Based on the cleaned real-time meteorological data, the temperature data and humidity data of the same period in history are matched one by one and the difference between the real-time data and the historical data is calculated to form the real-time meteorological data after the standard deviation test; According to the deviation analysis results and the real-time meteorological data after the standard deviation test, the conditional likelihood ratio is calculated, and the calculation formula is: ; in, represents the conditional likelihood ratio, represents the deviation analysis results, Represents the mean of the deviation analysis results when dry and hot events occurred during the same period in history, Represents the temperature data change rate in the current real-time meteorological data. Represents the humidity in the current real-time meteorological data. Represents the average humidity when dry and hot events occurred during the same period in history. Represents the rate of change of humidity data in the current real-time meteorological data. Represents the current ambient air pressure data, Represents the average of air pressure data under historical environmental conditions during the same period.

[0009] Preferably, the step of obtaining the dry heat composite event determination result is: Match the daily temperature and humidity amplitudes corresponding to the occurrence of dry and hot events in the same period of history one by one, and count the meteorological amplitude difference characteristics when the historical dry and hot events occurred and did not occur, forming a historical meteorological amplitude feature set; According to the historical meteorological amplitude feature set, the dry heat event determination threshold is calculated, and the calculation formula is: ; in, is the threshold for determining dry heat events, Represents the average amplitude of daily temperature variation during the same period in history, Represents the daily temperature variation amplitude in the current real-time meteorological data, Represents the average amplitude of daily humidity changes during the same period in history. Represents the daily amplitude of humidity variation in the current real-time meteorological data. Represents the solar radiation intensity in the current real-time meteorological data. Represents the average intensity of solar radiation during the same period of history. Represents the dew point temperature in the current real-time meteorological data. Represents the average dew point temperature corresponding to the occurrence of dry and hot events in the same period of history. Represents the atmospheric saturated water vapor pressure in the current real-time meteorological data. Represents the average atmospheric saturated water vapor pressure corresponding to the occurrence of dry and hot events in the same period of history; The conditional likelihood ratios are compared with the dry heat event determination thresholds one by one to determine whether the conditional likelihood ratios exceed the dry heat event determination thresholds, thereby forming a dry heat composite event determination result.

[0010] Preferably, the step of obtaining the adjusted threshold value is: Based on the dry-heat composite event determination result, the temperature data, humidity data and wind speed data in the real-time meteorological data are called one by one, the occurrence frequency of the dry-heat composite event in the current continuous period is calculated, and the correlation characteristics between the event frequency and the real-time meteorological data are analyzed to form a dry-heat event correlation feature set; According to the dry-heat event correlation feature set, the fluctuation degree of temperature data, the decline amplitude of humidity data and the change trend of wind speed data in the current real-time meteorological data are analyzed one by one to determine the extreme degree of current meteorological conditions and form a real-time meteorological extreme degree classification result; According to the real-time meteorological extreme degree classification result, the adjusted threshold is calculated, and the calculation formula is: ; in, is the adjusted threshold, Represents the standard deviation of the difference between the temperature data in real-time meteorological data and the historical temperature data. Represents the variance of the rate of change of air humidity data in real-time meteorological data, Represents the peak value of wind speed data in real-time meteorological data. Represents the daily variation amplitude of precipitation in real-time meteorological data, Represents the rate of change of solar radiation intensity in real-time meteorological data.

[0011] Preferably, the step of acquiring the updated dry heat composite event information is: Calling the adjusted threshold and the dry heat composite event determination result, comparing the relationship between the conditional likelihood ratio in the dry heat composite event determination result and the adjusted threshold one by one, marking the data record exceeding the adjusted threshold, and forming a real-time dry heat composite event trigger record; Based on the real-time dry-heat composite event trigger record, the time interval and continuous frequency of occurrence are statistically recorded one by one, and the time distribution and event continuity characteristics are analyzed to form updated dry-heat composite event information.

[0012] Preferably, the steps of obtaining the decision suggestion output are: Based on the updated dry-heat composite event information, the preset event level response standards are compared to determine the type of response measures, including personnel evacuation, emergency resource deployment and information release, and generate decision-making recommendation outputs.

[0013] Compared with the prior art, the advantages and positive effects of the present invention are: The present invention collects and screens real-time temperature and humidity data, matches and compares real-time data with historical data of the same period, comprehensively analyzes climate deviation characteristics and historical trends of dry-heat events, scientifically calculates accurate conditional likelihood ratios of dry-heat events, and constructs a dynamic threshold mechanism for real-time environmental condition change trends, thereby improving the accuracy of meteorological condition change analysis and the timeliness of dry-heat composite event identification, achieving more accurate and real-time monitoring and early warning of dry-heat composite events, improving the sensitivity and reliability of dry-heat composite event prediction and early warning response, reducing early warning delays, and enhancing the accuracy and response speed of response measures decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a schematic diagram of the steps of the present invention. DETAILED DESCRIPTION

[0015] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0016] See also Figure 1 The present invention provides a technical solution, a method for identifying dry-heat composite events under the influence of climate change, comprising the following steps: Collect the temperature and humidity data of the current weather station, filter the temperature and humidity data, remove any abnormal values ​​or missing data, and obtain the cleaned real-time weather data; perform statistical analysis on the cleaned real-time weather data, calculate the deviation value from the historical data of the same period, and obtain the deviation analysis results; Based on the deviation analysis results, the probability ratio of the occurrence and non-occurrence of dry heat events under the current meteorological conditions is calculated to obtain the conditional likelihood ratio; the conditional likelihood ratio is used in combination with the set threshold to determine whether the conditions for the occurrence of dry heat composite events are met and generate the dry heat composite event determination result; Dynamically adjust the results of dry-heat composite event determination, analyze the current climate trend, adjust the sensitivity of the warning, and obtain the adjusted threshold; based on the adjusted threshold, re-evaluate the current dry-heat composite event and generate updated dry-heat composite event information; Based on the updated dry-heat composite event information, the response measures are judged, including evacuation suggestions and decision support information for emergency resource allocation, and decision-making recommendation outputs are obtained.

[0017] The steps to obtain cleaned real-time meteorological data are as follows: Compare the real-time collected temperature data and humidity data one by one to see if they are within the extreme value range, remove abnormal temperature data and abnormal humidity data that are out of range, and generate real-time meteorological data after removing abnormal values; Based on the real-time meteorological data after removing abnormal values, determine whether the temperature data and humidity data are missing or incompletely recorded, remove the temperature data and humidity data with missing data or incompletely recorded, and generate the real-time meteorological data after removing the missing data; Based on the real-time meteorological data after removing missing data, the temperature data and humidity data are sorted in chronological order and the consistency of the data is verified to generate the cleaned real-time meteorological data.

[0018] Specifically, for the reference content of the extreme value range, we can first count the lowest and highest temperatures of historical meteorological samples, and combine them with the upper and lower limits of humidity recorded in different regions and seasons to set temperature thresholds and humidity thresholds. These values ​​are mainly based on extreme cases obtained from long-term observations and supplemented by linear regression to adjust the possible upper and lower limits. In the specific implementation process, the temperature and humidity values ​​received in real time are compared with the above-mentioned intervals in turn. If the temperature or humidity of each record exceeds the threshold, it is judged as abnormal, and then these abnormal records are deleted. If it is found that the difference between the abnormal record and the adjacent record is too large during deletion (for example, the temperature difference with the previous and next data exceeds 10°C or the humidity difference is more than 50%), an additional review is performed and the final number of eliminations is counted. After this process, all records that meet the defined interval and have not been reviewed and eliminated are retained, and finally real-time meteorological data after removing abnormal values ​​is generated.

[0019] Based on the real-time meteorological data after removing outliers, each temperature and humidity record is checked for integrity. If the temperature field or humidity field in a record is empty or cannot be identified, it is considered a missing record. Further considering that some records may appear to have values ​​but the collection results are unreliable due to failures, the temperature and humidity change trends in consecutive time periods can be compared first: for example, if the difference in temperature data collected three times in a row is less than 2°C and the fourth collection suddenly obtains a value much greater than 90°C or less than 0°C, the record is marked as abnormal and classified as missing. In addition, the humidity record can also be checked. The records are compared in context. If the humidity drops by 60% in a short interval between two data points, they are also marked and classified as missing. All records with missing marks are deleted according to the set standard steps. Here, a rule can be set as an example: if the proportion of missing records exceeds 5% within a certain period of time, a comprehensive re-inspection is performed, and then the records that are determined to be deleted are removed from the data set to form the final remaining record set. The 5% threshold used in this rule is empirically set based on the comprehensive stability of each detection equipment in the past year. Finally, after executing all deletion operations, real-time meteorological data after removing missing data is obtained.

[0020] Based on the real-time meteorological data after removing missing data, it is necessary to sort the time sequence and check the consistency of temperature and humidity data. The sorting steps here include sorting each valid record by the collection timestamp first. If it is found that the timestamps of a small number of records conflict with the order of the previous and subsequent records, they are listed separately and the reasons are checked. For example, if there are multiple records at the same time point, the first collected data is taken as the main reference. When checking the consistency, first check whether the changes in temperature and humidity at adjacent time points are consistent with the previously obtained interval judgment. For example, the temperature rise within a one-minute interval must be between 0.1℃ and 2.0℃. If it exceeds this range, it is marked as suspicious. Then, combined with historical statistical data, the reasonable fluctuation range between adjacent records is further checked. If it is confirmed that there is an obvious abnormality, the corresponding record will be eliminated. During the process, each elimination operation will be compared at the numerical level to exclude accidental deletion and update the sequence structure after elimination. After completing all time series corrections, the record order can be confirmed to be completely consistent in the last traversal, and finally the cleaned real-time meteorological data is generated.

[0021] The steps to obtain the deviation analysis results are: Call the temperature data and humidity data in the cleaned real-time meteorological data, call the temperature data and humidity data in the historical meteorological data of the same period, establish the corresponding relationship between the real-time and historical data of the same period one by one, perform the matching check of the real-time and historical temperature data and humidity data item by item, remove the data items that have not been matched successfully, and form a meteorological data pair after successful matching; According to the meteorological data pairs after successful matching, the deviation analysis results are calculated using the following formula: ; in, represents the deviation analysis results, Represents the temperature data in the real-time meteorological data after cleaning, Represents the temperature data in the historical meteorological data for the same period. Represents the humidity data in the real-time meteorological data after cleaning, Represents the humidity data in the historical meteorological data for the same period.

[0022] Specifically, based on the correlation between the cleaned real-time meteorological data and the historical meteorological data of the same period, in order to determine whether the two can correspond one to one, it is necessary to first read the cleaned real-time temperature and humidity records, and then read the historical temperature and humidity records of the same period, and then perform one-to-one matching according to the timestamp and compare them one by one. For example, the real-time temperature record is compared with the effective range of 0°C to 90°C, and the real-time humidity record is compared with the effective range of 0% to 100%. For each real-time record within the above range, the numerical difference is compared with the corresponding historical record. If the temperature difference is within 2°C and the humidity difference is within 10%, it is considered to be a successful match, otherwise it is considered to be unmatched. Here, 2°C and 10 The % threshold is obtained based on the statistics of meteorological observation results accumulated in the past two years. We can first calculate the average daily temperature difference and humidity difference from the data of previous years and perform deviation analysis on them. Then, after comprehensively considering the climate fluctuations in different seasons, we will float the average difference upward by a certain proportion, and finally determine the two empirical thresholds of 2°C and 10%. During the process, if a certain real-time data or historical record has a serious disorder in the timestamp, it will be skipped and the next one will be matched. If a large batch of matches fail, it is necessary to recheck the consistency of the time series. All records that meet the above standards are marked as available and summarized into new sequences in turn. Finally, after checking the correlation of the data in each time period as a whole, a pair of meteorological data with successful matching is formed.

[0023] formula The benefit is that by doubly measuring the temperature and humidity differences and introducing different operations such as cubic root and square to measure multi-dimensional differences, it can take into account the numerical relative relationship between temperature and humidity in the same expression, while weakening the impact of extreme values ​​on the overall result, thereby providing a more layered difference representation in the subsequent judgment of dry-heat composite events.

[0024] The steps for obtaining the parameters are as follows: first, continuously collect temperature data within 24 hours at the weather station, then summarize the average temperature records of each hour according to the hourly time period, and then remove the abnormal values ​​of these records and complete the time series correction. , select the time interval data comparable to the historical period from the confirmed valid hourly temperature records, and perform arithmetic averaging on all these temperature values ​​to obtain a real-time average temperature, recorded as In the specific implementation, the average temperature of the current 30 days can also be used. For example, if the hourly temperature records of the day include 21.3℃, 22.7℃, 24.2℃, etc., the 24 records can be averaged to get .

[0025] The steps for obtaining the parameters are to first select the same meteorological data as above in the historical meteorological data of the same period. The temperature data in the time period that matches the corresponding date or season is collected in a 24-hour recording mode, and then the average of all the hourly data is obtained. If the historical database contains temperature records for several years, we can further calculate the average values ​​by year and then perform a quadratic curve fitting on this series of average values ​​to obtain the temperature average value closest to the current period as For example, when retrieving data for the same day from 2019 to 2022, if the average temperature at the hour of the day is 22.1℃, 23.6℃, ​​24.0℃, etc., the data can be summarized as .

[0026] The steps to obtain the parameters are as follows: Select 24 consecutive hours of humidity data from the weather station humidity records in the same time period, compare the reasonable range of 0% to 100% one by one, remove abnormal values, and then calculate the arithmetic average of the remaining records. At the same time, in order to ensure the stability of the quantification process, the humidity data of the same hourly period can be extracted in the past 30 days to cross-check with the real-time records. When the number of abnormalities exceeds the preset 4%, the equipment status is rechecked to obtain a reliable humidity data set and finally calculate the average. For example, if the 24-hour humidity record of a single day includes 71%, 73%, 74%... and other data in sequence, the final result is .

[0027] The steps to obtain the parameters are as follows: Humidity records of the same historical time period are summarized. On the premise of ensuring that the data is complete and without missing, the humidity values ​​at the hour of the day are averaged. If there are humidity records for multiple years, the humidity means of the same date in each year are summed up and averaged, and then the humidity level that best represents the current date is selected as the humidity level. For example, after processing the humidity records of the same day from 2019 to 2022, multiple groups of average humidity values ​​such as 70%, 72%, 75%, etc. may be obtained, and these values ​​are then aggregated to determine the average. .

[0028] Calculation process: Calculate the temperature difference first , if the previous , ,but ; Then calculate the humidity difference ,like , ,but ; Divide it into two parts according to the formula: The first part of the molecule is ; The denominator of the first part is = ; The second part of the molecule is = ; The denominator of the second part is: = ; Add the two parts together and take the cube root of the whole to get , the specific multi-level formula examples are as follows: ; Part I ; Part 2 ; Addition result ; ; The results show that the calculated It is about 1.093. If a specific reference standard is set in combination with the corresponding temperature difference and humidity difference, the larger the value, the more obvious the difference between the current real-time temperature and the historical temperature, and the humidity difference is relatively larger. The smaller the value, the closer the two are. Different boundaries can be set according to the difference. For example, if the experience is set in advance As a boundary, When it is greater than 1.2, it means that the temperature and humidity deviation is more significant. When it is lower than 1.2, it means that the temperature and humidity deviation is relatively small.

[0029] The steps to obtain the conditional likelihood ratio are: Based on the cleaned real-time meteorological data, the temperature and humidity data of the same period in history are matched one by one and the difference between the real-time and historical data is calculated to form the real-time meteorological data after the standard deviation test; According to the deviation analysis results and the real-time meteorological data after the standard deviation test, the conditional likelihood ratio is calculated. The calculation formula is: ;

[0030] in, represents the conditional likelihood ratio, represents the deviation analysis results, Represents the mean of the deviation analysis results when dry and hot events occurred during the same period in history, Represents the temperature data change rate in the current real-time meteorological data. Represents the humidity in the current real-time meteorological data. Represents the average humidity when dry and hot events occurred during the same period in history. Represents the rate of change of humidity data in the current real-time meteorological data. Represents the current ambient air pressure data, Represents the average of air pressure data under historical environmental conditions during the same period.

[0031] Specifically, based on the matching between the cleaned real-time meteorological data and the historical meteorological data of the same period, it is necessary to find the corresponding historical temperature and humidity information for each real-time record one by one, and then calculate the numerical difference between the two. In the process, the temperature and humidity records of the current real-time data can be read first and compared with the pre-established effective range. For example, the temperature value is limited to between 0°C and 90°C, and the humidity value is limited to between 0% and 100%. If a real-time temperature exceeds this range, it will not be included in the matching process. Then the real-time records and historical data that meet the range are compared one by one. Whenever the temperature and humidity in the same time period or the same observation frequency are matched, the difference is calculated. Specifically, a difference set can be obtained through the method of temperature difference = real-time temperature - historical temperature, humidity difference = real-time humidity - historical humidity. In order to ensure that these differences are statistically significant, a standard deviation test can be performed on the difference set based on a sufficient number of records. The mean of the difference set is first calculated and compared with the mean and each The degree of deviation of the difference is used to judge its distribution in the whole group of differences. If a difference is more than 3 times the standard deviation from the mean, it is marked as abnormal, and the distribution proportion of the abnormal items is continued to be counted. If the abnormal items exceed 5%, the reliability of the batch of data during the collection period is reviewed again. If the abnormal proportion is less than 5%, the overall difference distribution can be regarded as stable and the difference data that have passed the test will be retained to form real-time meteorological data after the standard deviation test. This 5% threshold is derived from the data observation of the stability of meteorological monitoring equipment in previous years and is determined in combination with the error range statistics of the sensor under various environmental conditions. The specific calculation can be carried out by first observing the same time period on multiple devices in multiple locations and summarizing the cumulative distribution of the errors, and then using quantile statistics on the distribution to obtain a suitable percentage threshold and record it. If the test result passes, it means that most of the real-time and historical differences are within the predetermined range. Continuing to summarize these differences that have passed the test can obtain real-time meteorological data after the standard deviation test.

[0032] formula The benefit is that it takes into account the combined effects of the numerical differences between real-time and historical deviation analysis results, the temperature change rate, the humidity rate, and the air pressure level. It demonstrates the sensitivity to the interrelationships of multidimensional meteorological parameters through multiple calculations such as cube root and square root, thus meeting the needs of simultaneous measurement of multiple factors such as temperature, humidity, and air pressure in dry-heat composite event analysis.

[0033] parameter The steps to obtain are as follows, which has been calculated previously.

[0034] parameter The acquisition steps are as follows: select the temperature and humidity data corresponding to the current monitoring period during the period of historical dry heat events, and calculate a set of value, and then for this group Worth doing and getting In order to obtain a reliable mean, the temperature and humidity records during no less than ten dry heat events can be collected, and the same deviation analysis method can be used to calculate the data of each event. Finally, all the results are summarized and the average is recorded as .

[0035] parameter The acquisition steps are as follows: first, continuously monitor the temperature change rate of the current real-time meteorological data, that is, record a number of temperature data within a set time interval and calculate the difference change rate of adjacent data. The temperature change rate of each time period is recorded as , and then accumulate or sum these rate values ​​multiple times to finally get an overall rate of change , to quantify this process, we can use: ,in is the number of monitoring intervals, Indicates The temperature change rate is calculated by recording the temperature once every hour for 24 hours and averaging the change rate of the two consecutive hours. If the temperature rises steadily from 20°C to 28°C within 24 hours, and the hourly temperature rise is between 0.3°C and 0.5°C in most periods, the overall average is .

[0036] parameter The acquisition steps are as follows: directly read the air humidity value from the current real-time meteorological data and obtain the average humidity value after cleaning and verification. First, remove the abnormal values ​​of the humidity data within 24 hours and average it by hour. Assuming that the statistical results show that the humidity records are concentrated in the range of 50% to 70%, the humidity data is finally sorted out. =62%.

[0037] parameter The steps to obtain the humidity are to find the average humidity level in the meteorological records when the historical dry and hot events occurred, and this average value is called , usually the hourly humidity data of several consecutive days during the dry and hot event are selected for calculation, and the daily average is taken as the overall average to obtain , the value range is also 0% to 100%. For a certain place, if the average humidity obtained in the dry and hot events over the years is mostly around 40%, it can be calculated according to the actual situation. About 40%.

[0038] parameter The acquisition steps are to pay attention to the change rate of the current real-time humidity data, that is, to calculate the difference of humidity values ​​within a fixed time step, summarize these difference values ​​and average them to obtain If the humidity drops by 1% to 2% every 2 hours during a continuous 12-hour observation, the arithmetic average is =1.5%.

[0039] parameter The acquisition steps are to collect the current ambient air pressure value of the weather station, and average the data within 24 hours after correction. If the average air pressure statistical value for the whole day is around 1010hPa, then record =1010.

[0040] parameter The acquisition steps are to collect air pressure records under historical environmental conditions and use them with parameters The same time period and method are used to calculate the average. If the average air pressure after analyzing a period of historical data is about 1005hPa, then =1005.

[0041] Calculation process: First, according to the above obtained parameter values, for example , , , , , , , Bring in: ; ; ; ; ; ; ; Then the numerator of the formula is: ; The denominator of the formula is: ; From this we can get: ; The results show that when When the value is about 0.0945, it indicates that there is a certain gap between the current environmental deviation and the historical dry heat event conditions. The internal threshold can be combined to determine whether it has entered the high-risk area of ​​dry heat composite events. If a demarcation value is set empirically or statistically in the subsequent execution process, for example ,when Greater than 0.3 indicates a closer characteristic of dry heat events. If it is less than 0.3, it means that the overall conditions are more different from the distribution of dry and hot events.

[0042] The steps for obtaining the dry heat composite event determination results are as follows: Match the daily temperature and humidity amplitudes corresponding to the occurrence of dry and hot events in the same period of history one by one, and count the meteorological amplitude difference characteristics when the historical dry and hot events occurred and did not occur, forming a historical meteorological amplitude feature set; According to the historical meteorological amplitude feature set, the dry heat event determination threshold is calculated using the following formula: ; in, is the threshold for determining dry heat events, Represents the average amplitude of daily temperature variation during the same period in history, Represents the daily temperature variation amplitude in the current real-time meteorological data, Represents the average amplitude of daily humidity changes during the same period in history. Represents the daily amplitude of humidity variation in the current real-time meteorological data. Represents the solar radiation intensity in the current real-time meteorological data. Represents the average intensity of solar radiation during the same period of history. Represents the dew point temperature in the current real-time meteorological data. Represents the average dew point temperature corresponding to the occurrence of dry and hot events in the same period of history. Represents the atmospheric saturated water vapor pressure in the current real-time meteorological data. Represents the average atmospheric saturated water vapor pressure corresponding to the occurrence of dry and hot events in the same period of history; The conditional likelihood ratios are compared with the dry heat event determination threshold one by one to determine whether the conditional likelihood ratio exceeds the dry heat event determination threshold, thereby forming a dry heat composite event determination result.

[0043] Specifically, in order to match the daily temperature variation amplitude and humidity variation amplitude values ​​corresponding to the occurrence of dry and hot events in the same period of history one by one, the temperature and humidity observation records in the past three to five years in the same geographical area can be retrieved first, and the maximum and minimum values ​​of daily temperature and humidity within 24 hours are subtracted to obtain the daily variation amplitude. These amplitudes are recorded and compared with the dates marked as dry and hot events or non-dry and hot events, and a comparison data set containing dry and hot states and non-dry and hot states is generated. Then, the time period that can clearly correspond to the occurrence of dry and hot events is selected, and the daily temperature variation amplitude and humidity variation amplitude in these time periods are recorded separately. When recording the daily temperature variation amplitude, a valid range can be set first, such as the temperature range of 0℃ to 90℃, and the daily maximum and minimum temperature differences are recorded in this range. If the amplitude data of individual dates exceed this range, the calibration status of the corresponding meteorological sensor is checked. If it is confirmed to be normal, the records out of range are classified as extreme exceptions and the reasons are attached. The daily humidity variation amplitude is tested in the same way with reference to the humidity value range of 0% to 100% to ensure that all amplitude data are within a valid range. The data set was then divided into a group where dry heat events occurred and a group where they did not occur. The difference in amplitude between the two groups was determined by comparing the data distribution of the two groups. The number of samples in each group (for example, the dry heat group contained 200 dates and the non-dry heat group contained 600 dates) was combined. If the average daily amplitude of several heat groups was significantly greater than that of the non-dry heat group, the difference value was counted. At the same time, the difference between the temperature and humidity amplitudes of each date was compared. For example, dates with a temperature amplitude greater than 15°C and a humidity amplitude greater than 30% were classified as potential dry heat groups, and then compared with actual occurrence dates. The dry and hot event records are matched and checked. If the overlap rate of the two exceeds 50%, it means that the amplitude range is significantly associated with dry and hot events. When further refining the setting methods of various thresholds, data with longer observation periods can be introduced for segmented calculations. For example, in extremely high temperature seasons, the range of the daily temperature variation amplitude can be moderately increased to above 20°C, and in extremely dry seasons, the range of the daily humidity variation amplitude can be increased to above 40%. All the values ​​after matching and difference analysis are integrated together to finally form the historical meteorological amplitude feature set required for this section.

[0044] formula The benefit of this method is that it can integrate multiple factors such as the daily variation of temperature and humidity, solar radiation intensity, dew point temperature, and atmospheric saturated water vapor pressure into the same expression, and use operations such as cube root and absolute value to make a hierarchical measurement of extreme values ​​and differences, taking into account the multidimensional attributes of meteorological elements, which is conducive to further determining the conditions for the occurrence of dry and hot events.

[0045] The steps to obtain the parameters are as follows: first, collect the daily temperature change amplitude records in multiple years of the same period in history, and limit them to the same geographical area and similar seasonal periods. The difference between the highest and lowest temperatures in each day is taken as the amplitude value. Then, the extreme records caused by faulty equipment are eliminated from these amplitude values. Then, the remaining amplitudes are summed and divided by the corresponding number of days to obtain an average amplitude recorded as In this example, if 300 days of valid data are collected in summer, the total temperature is about 4200℃, then we can calculate .

[0046] The steps for obtaining the parameters are as follows: select the temperature records for 24 consecutive hours from the current real-time meteorological data, count the maximum and minimum temperatures, and calculate the difference to obtain the daily variation amplitude. If the maximum temperature on that day is 30°C and the minimum temperature is 16°C, then .

[0047] The steps for obtaining the parameters are to collect the daily humidity amplitude in the same time period in the dry heat event history library, find the humidity amplitude data marked as the dry heat event, and calculate the average of these amplitudes one by one. For example, if the humidity amplitude data for the 100 days when a dry heat event occurs is retrieved and the total is 3200%, then .

[0048] The steps to obtain the parameters are as follows: record the daily humidity amplitude of the current day or multiple days in the current real-time meteorological data, accumulate the difference between the highest humidity and the lowest humidity every day or compare them on a single day to obtain the real-time humidity amplitude, and then perform an average operation to obtain the humidity amplitude. For example, for 5 consecutive days, the highest humidity is 85%, 81%, 78%, 76%, and 80%, and the lowest humidity is 50%, 49%, 47%, 45%, and 51%, respectively. Then the daily amplitude is 35%, 32%, 31%, 31%, and 29%, respectively. The average is .

[0049] The steps to obtain the parameters are to record the solar radiation intensity within a certain period of time from the current meteorological observation. The measured values ​​within a day can be taken at multiple time points for weighted average or direct average. If an observation station measures multiple groups of radiation intensities during 8 hours of sunshine, such as 800W / m 2 、820W / m 2 、750W / m 2 etc., we can first do the arithmetic average to get After comprehensive statistics over multiple days, the average value is about 780W / m 2 , then record it as .

[0050] The steps to obtain the parameters are to call the solar radiation intensity data when the historical dry and hot events occurred in the same period, and also use the time-division measurement and averaging method to generate the daily radiation intensity, and then summarize the average value after multiple days to get If the statistical results over the years show that the average radiation intensity is about 600W / m 2 You can .

[0051] The step of obtaining the parameters is to record the dew point temperature in the current real-time meteorological data. The dew point calculation formula in meteorology can be combined with meteorological elements such as temperature and humidity: ; in , represents temperature, Represents humidity, and is the coefficient, usually taken , , and the result is averaged after multiple measurements. For example, when the average temperature is 25°C and the average humidity is 60%, the dew point temperature is obtained by substituting the formula into .

[0052] The steps to obtain the parameters are to query the dew point temperature records corresponding to the historical dry heat events during the same period, and ensure that The calculation method is consistent with that of the previous one. The stored dew point temperature data can also be directly retrieved from the historical database. The arithmetic average of multiple days is calculated and summarized. The dew point temperature during several thermal events is generally in the range of 10℃ to 15℃, which can be finally obtained. .

[0053] The steps to obtain the parameters are as follows: based on the temperature and humidity in the current real-time meteorological data, the approximate calculation method of the atmospheric saturated water vapor pressure is used: ,in is the temperature, For humidity, calculate the daily or hourly saturated water vapor pressure and take the average value to get If the average temperature in a day is 25℃ and the humidity is 60%, the saturated water vapor pressure is approximately 14.9hPa.

[0054] The steps to obtain the parameters are to retrieve the atmospheric saturated water vapor pressure recorded during the historical dry and hot events of the same period. If the data has been directly stored, it can be read directly. Otherwise, the above formula can also be used to process temperature and humidity, and the average value of the results of multiple days is obtained. For example, if the average temperature during a hot period is 23°C and the humidity is 50%, then the result is about 11.5hPa, which can be recorded. .

[0055] Calculation process: Set the values ​​of each parameter: , , , , , , , , , , when the humidity amplitude is regarded as an integer, it can be recorded as 32, 31.6. Here, the unit % can be understood as a pure numerical ratio in the calculation, and is substituted into the formula: ; ; The molecular result is approximately , Calculate the denominator first: ; ; Look again , ; So the denominator totals , and finally calculate: ; The results show that the dry heat event determination threshold It is approximately 26.78. In the subsequent comparison stage, when the likelihood ratio value of other conditions exceeds this threshold, it can be determined that the dry heat complex event has entered a highly likely state. If it is lower than this value, it means that the current conditions are not highly consistent with the historical dry heat amplitude characteristics, and it can be used as a reference to determine the subsequent disposal method.

[0056] After obtaining the dry heat event determination threshold, when comparing the conditional likelihood ratio with the aforementioned threshold one by one, it is necessary to first record the conditional likelihood ratio Q value calculated in all time periods and compare it with the established F value one by one. If the Q value is greater than the F value, it is determined that the risk level of the dry heat composite event has been reached, otherwise it is still regarded as a normal state. To ensure the continuity of the comparison process, a recording rule can be set. For example, a Q value is generated every 30 minutes and compared with the latest F value. If the Q value is greater than the F value for three consecutive times, the period is listed as a suspected dry heat interval. If the Q value then falls below the F value, the mark is removed. The Q value is obtained from the previous step. The calculation includes a deviation check on temperature and humidity, as well as a comprehensive measurement of multiple factors such as air pressure and radiation. The F value is the threshold for determining dry heat events obtained in the previous paragraph. If a 24-hour continuous monitoring is carried out in a certain place, and the F value is immediately compared after the current Q value is calculated in each monitoring period, the sequences with Q values ​​greater than or less than the F value can be gradually collected, and the cumulative duration and frequency of Q values ​​greater than the F value can be checked. When the cumulative duration exceeds 3 hours or the frequency exceeds 5 times, a phased conclusion can be recorded. The phased conclusion can be connected to the relevant monitoring system or manual on-duty personnel, and the judgment result can be updated to the dry heat composite event judgment result.

[0057] The steps to obtain the adjusted threshold are: Based on the determination results of dry-heat composite events, the temperature data, humidity data and wind speed data in the real-time meteorological data are called one by one to calculate the frequency of dry-heat composite events in the current continuous period, and the correlation characteristics between the event frequency and the real-time meteorological data are analyzed to form a set of correlation characteristics of dry-heat events; According to the dry and hot event correlation feature set, the fluctuation degree of temperature data, the decline amplitude of humidity data and the change trend of wind speed data in the current real-time meteorological data are analyzed one by one to determine the extreme degree of current meteorological conditions and form a real-time meteorological extreme degree classification result; According to the real-time meteorological extreme degree classification results, the adjusted threshold is calculated using the following formula: ; in, is the adjusted threshold, Represents the standard deviation of the difference between the temperature data in real-time meteorological data and the historical temperature data. Represents the variance of the rate of change of air humidity data in real-time meteorological data, Represents the peak value of wind speed data in real-time meteorological data. Represents the daily variation amplitude of precipitation in real-time meteorological data, Represents the rate of change of solar radiation intensity in real-time meteorological data.

[0058] Specifically, based on the determination result of dry-heat composite events, first continuously extract all time periods that have been determined as dry-heat composite events and record them in the corresponding index, then call the temperature, humidity and wind speed values ​​in the real-time meteorological data, and compare the temperature range of 0°C to 90°C, the humidity range of 0% to 100%, and the wind speed range of 0m / s to 50m / s for each record in turn. If a certain data meets the above range, it will be retained for subsequent statistics on the frequency of dry-heat events in the time period. If it does not meet the above range, check whether it is a sensor failure or an abnormal mark. If a failure is confirmed, skip the record. If an abnormality is confirmed and the number exceeds 5%, it is necessary to test the robustness of the acquisition device and record the abnormality rate. After obtaining all the records that meet the valid range and are associated with the dry-heat event index, check the time difference of adjacent records one by one and determine whether the dry-heat event still exists in the adjacent time. If several thermal event marks appear 3 or more times in a row, it is considered that the period is a relatively stable one. The dry and hot stage is included in further statistics, and then the historical data series of temperature and humidity are compared. The cumulative value of the number of dry and hot composite events in different time periods is correlated with the value range or fluctuation amplitude of temperature, humidity and wind speed. For example, when dry and hot events occur more than 5 times within 24 hours, the relationship between event frequency and real-time meteorological data is summarized by comparing whether the average temperature is above 30℃ or whether the wind speed is above 20m / s, and whether the humidity is lower than 30%. If it is found that the frequency of dry and hot events is significantly increased when high temperature, low humidity and strong wind are superimposed, such high-frequency segments can be marked in the dry and hot event association feature set. If the wind speed fluctuates less during the analysis, its feature can be marked as a continuous low wind speed segment. If dry and hot events still occur frequently, it can be explained that they are mainly driven by temperature and humidity at this time. These statistical results are written into the association set to form a dry and hot event association feature set.

[0059] According to the dry heat event associated feature set, it is necessary to analyze the temperature data, humidity data and wind speed data in the current real-time meteorological data one by one, and calculate the temperature fluctuation degree, humidity drop amplitude and wind speed change trend to determine the extreme degree of meteorological conditions. When executing it specifically, first set an observation window (for example, record once every 30 minutes), count the maximum and minimum temperature values ​​in this window and calculate their difference. If the temperature difference of multiple consecutive windows always exceeds 10℃, it is considered that the temperature fluctuation is large. At the same time, it is necessary to confirm that these temperature values ​​are between 0℃ and 90℃ to exclude abnormal measurements. Next, the humidity is processed in the same way. If the humidity continues to drop for a period of time and the drop is greater than 20%, it can be marked as a humidity drop section, and then compared with the dry heat event associated feature set obtained previously. If the frequency of occurrence of the corresponding dry heat event is significantly higher than the previous average level, it is marked in this step. The wind speed data is calculated according to its highest The peak intensity is judged by the large value. For example, the maximum wind speed exceeding 30m / s is regarded as a high wind speed trend. If the wind speed is still within the normal range, the original classification is maintained. After the analysis is completed, the temperature fluctuation, humidity drop, and wind speed are combined to obtain an extreme degree score. The specific score can be determined according to the position of each measurement value in the corresponding interval. For example, a temperature fluctuation of 10℃ corresponds to a high score, a humidity drop of more than 25% corresponds to a higher score, and a wind speed of more than 30m / s corresponds to an even higher score. After accumulating three sub-items, the final classification is made with reference to the frequency of dry heat events. According to the measured combined score, it is divided into several levels such as extreme, high, and moderate. These levels have corresponding average reference values ​​and upper and lower limits in the local monitoring data for many years. If all three indicators are high and dry heat events occur frequently, this period will be classified as an extreme degree classification result. If only local indicators are prominent, they will be placed in the moderate or general category, and finally a real-time meteorological extreme degree classification result is formed.

[0060] formula The benefit is that it takes into account multiple influencing factors such as temperature difference distribution, humidity change rate, wind speed peak, precipitation amplitude and solar radiation change rate, and reduces the distortion of extreme values ​​through piecewise processing of cube roots and square roots, while achieving the measurement of the comprehensive effects of multiple factors.

[0061] The steps for obtaining the parameters are as follows: first, extract the temperature sequence of the same observation period from the current real-time meteorological data and the historical temperature data, then calculate the point-by-point difference between the two to form a difference sequence, then subtract the mean from the difference sequence to obtain a zero-mean sequence, square each item on the basis of this sequence and accumulate it, finally divide it by the sequence length to obtain the variance, and then take the square root of the variance as the standard deviation and record it as , which is usually between 1°C and 10°C. A reasonable range can be determined through historical monitoring data. For example, in a certain place, 24 hourly data are collected every day in the past 7 days. The difference is accumulated with the historical average temperature for many years and then the variance is calculated. If the standard deviation is about 5.3°C, then .

[0062] The steps for obtaining the parameters are as follows: after measuring the humidity at minute or hour intervals in the current real-time meteorological data, the humidity difference between each adjacent interval is calculated to obtain a change rate sequence, and then the same variance operation is performed on the sequence, that is, all change rates are first averaged, and then the sum of the square differences between each rate and the mean is calculated and divided by the sequence length. Finally, the variance is recorded as The value range can be between 0 and 10 depending on the region and season, or it may be higher or lower. If a place collects humidity every hour for 24 hours and finds that the rate of change is mostly between 2% and 5%, with a variance of about 1.8, record .

[0063] The parameter acquisition step is to extract the maximum instantaneous value of wind speed from the current real-time meteorological data and record it as the peak wind speed value. The value can be between 0m / s and 50m / s. If the wind speed measurement in the past 12 hours finds that it has reached 28m / s, it will be recorded as .

[0064] The steps to obtain the parameters are as follows: first, record the hourly precipitation in the current meteorological data, find the maximum and minimum precipitation in the data of 24 hours in a day, and define the difference between the two as the daily variation amplitude and record it as If the area has a maximum precipitation of 6mm and a minimum precipitation of 0mm throughout the day, then .

[0065] The steps for obtaining the parameters are to collect the radiation intensity within a certain time interval from the current solar radiation monitoring, calculate the intensity change rate once an hour on the same day, and obtain a rate sequence. The values ​​of each rate in the sequence are recorded and summarized. If the highest change rate is very different from the lowest change rate, the difference between the adjacent time periods can be calculated and then divided by the number of measurements to obtain an average change rate, and finally form If we count the radiation intensity over the whole day, we will find that the radiation intensity increases faster from morning to noon, with an average rate of about 20W / m 2 / h, you can mark .

[0066] Calculation process: Enter the value: , , , , , calculate the numerator first: ; ; The first part of the denominator: ; ; The second part of the denominator: ; ; Add the two denominators together: , and finally concluded: ; This result shows that the adjusted threshold It is about 0.134. This method can be used when a more sensitive or robust judgment is needed for the dry heat composite event judgment process. By comparing with other judgment indicators one by one, if the conditional likelihood ratio or other measurement results of certain monitoring periods exceed 0.134, it means that the current meteorological conditions have tended to a higher degree of extreme. If it is lower than 0.134, it is temporarily judged that the strong dry heat critical value has not been reached.

[0067] The steps to obtain updated dry heat composite event information are: The adjusted threshold and the dry-heat composite event determination result are called, and the relationship between the conditional likelihood ratio in the dry-heat composite event determination result and the adjusted threshold is compared one by one, and the data record exceeding the adjusted threshold is marked to form a real-time dry-heat composite event trigger record; Based on the real-time trigger records of dry-heat composite events, the time intervals and continuous frequencies of occurrence are statistically recorded item by item, and the time distribution and event continuity characteristics are analyzed to form updated dry-heat composite event information.

[0068] Specifically, when calling the adjusted threshold and the dry-heat composite event determination result, the previously obtained conditional likelihood ratio list can be taken out one by one, and the values ​​can be compared one by one against the adjusted threshold obtained previously. If the value of a conditional likelihood ratio is greater than the threshold, the record is marked as an over-limit record. If it is less than the threshold, the normal mark is maintained. In order to clearly set the source of this threshold, it can be determined in combination with actual experience and monitoring data. For example, thousands of meteorological records are collected during a one-year observation period and the corresponding conditional likelihood ratios are calculated. After sorting their distribution characteristics, the upper 10% quantile is selected as the threshold. This ratio can also be appropriately adjusted up or down based on climate differentiation. Threshold sequences for different seasons and regions are formed. During the execution process, all records are first traversed to detect whether there are abnormal records or missing records and exclude them. After the data integrity is determined, each conditional likelihood ratio is compared with the adjusted threshold relationship in turn. If it exceeds the threshold, it is marked in the trigger list, and the timestamp, corresponding temperature and humidity and other auxiliary information are recorded. If multiple records are found to exceed the threshold in succession, it means that the dry heat event has continuity or further intensification signs. These continuous records can be classified in this link, depending on the situation, whether it is a single event or multiple events coexisting. Finally, all records exceeding the threshold are summarized to obtain real-time dry heat compound event trigger records.

[0069] Based on the exceeding threshold mark information in the real-time dry-heat composite event trigger record, it is necessary to retrieve the occurrence time of each record one by one and calculate the interval between each record. If the time interval between two adjacent mark records is too short (for example, less than 10 minutes), they will be regarded as the continuation of the same dry-heat event, and the number of triggers of the event will be counted. When the number of consecutive times exceeds 3, it means that there is a strong dry-heat composite signal in this period. In order to ensure that the 10-minute threshold is set reasonably, the shortest interval of similar events in the past year can be retrieved for statistics. For example, statistics show that the intervals of dry-heat events in high-incidence areas are mostly concentrated in 8 to 12 minutes. In this way, 10 minutes can be used as the main judgment basis, and then fluctuate up and down according to the actual situation. After all records are sorted on the timeline, the number of consecutive marks is counted. For example, within a certain period of time, at 1:00, 1: 08. Three triggers at 1:15 am are considered a complete dry heat event and the trigger frequency is recorded. If the time between trigger points exceeds 1 hour, it is considered a different event and the next event is counted again. Then the cumulative duration of each dry heat event is counted. If an event lasts from 1:00 to 2:30 am, the duration is 1.5 hours. After comparing the duration and interval distribution of different events, an event distribution sequence can be formed. Combined with the previously obtained information such as temperature, humidity, wind speed, etc., the environmental change characteristics when the event occurs can be observed, such as whether the high temperature is maintained above 30°C, whether the humidity is lower than 20%, and whether the wind speed exceeds 20 m / s. A cross-analysis is performed with previous events. Finally, based on these statistical data and triggering conditions, an up-to-date time distribution summary and event duration characteristic information are obtained, and updated dry heat composite event information is formed.

[0070] The steps to obtain the decision recommendation output are: Based on the updated dry-heat composite event information, the preset event level response standards are compared to determine the type of response measures, including personnel evacuation, emergency resource allocation and information release, and generate decision-making recommendation outputs.

[0071] Specifically, based on the updated dry-heat composite event information, it is necessary to compare it with the preset event level response standards. First, according to the previously obtained indicators such as duration, trigger frequency, temperature and humidity thresholds, the dry-heat composite events can be roughly divided into three categories: mild, moderate, and severe. If it is known through long-term meteorological data analysis in the early stage that mild events are usually triggered less than 5 times within 2 hours and the temperature range is mainly between 25℃ and 30℃, moderate events may be triggered 5 to 10 times within 4 hours and the temperature is mostly maintained between 30℃ and 35℃, and severe events are often triggered more than 10 times within 6 hours and the temperature is more than 40℃. If the temperature is above 35℃, the current event information can be matched according to these classification conditions. If it is confirmed that a certain dry heat event is severe, high-priority actions will be selected in the response measures type in this link, including evacuation of personnel in the risk area and simultaneous deployment of emergency materials or resources. If a shelter is needed, the area can be selected according to the on-site environment and warning signs can be set up. If the event is only mild after comparison, information release reminders will be given priority, and emergency resource allocation will be arranged as appropriate. Finally, all event levels are summarized and the respective measures types are summarized, including evacuation of personnel, deployment of emergency resources and information release, to form a decision-making recommendation output.

[0072] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for identifying dry-hot composite events under the influence of climate change, characterized in that: The following steps are involved: Collect the temperature and humidity data of the current weather station, filter the temperature and humidity data, remove any abnormal values ​​or missing data, and obtain the cleaned real-time weather data; Performing statistical analysis on the cleaned real-time meteorological data, calculating deviation values ​​from historical data of the same period, and obtaining deviation analysis results; Based on the deviation analysis results, the probability ratio of the occurrence and non-occurrence of the dry heat event under the current meteorological conditions is calculated to obtain the conditional likelihood ratio; Using the conditional likelihood ratio and a set threshold, determining whether a condition for a dry heat composite event to occur is currently met, and generating a dry heat composite event determination result; Dynamically adjust the determination result of the dry-heat composite event, analyze the current climate trend, adjust the sensitivity of the warning, and obtain an adjusted threshold; re-evaluating the current dry heat composite event based on the adjusted threshold value to generate updated dry heat composite event information; Based on the updated dry-heat composite event information, response measures are determined, including evacuation suggestions and decision support information for emergency resource allocation, to obtain a decision suggestion output.

2. The method for identifying dry-heat composite events under the influence of climate change according to claim 1, characterized in that: The steps for obtaining the cleaned real-time meteorological data are as follows: Compare the real-time collected temperature data and humidity data one by one to see if they are within the extreme value range, remove abnormal temperature data and abnormal humidity data that are out of range, and generate real-time meteorological data after removing abnormal values; Based on the real-time meteorological data after removing the abnormal values, determine whether the temperature data and the humidity data are missing or incompletely recorded, remove the temperature data and the humidity data with missing data or incompletely recorded, and generate the real-time meteorological data after removing the missing data; Based on the real-time meteorological data after removing missing data, the temperature data and the humidity data are sorted in chronological order and the consistency of the data is verified to generate the cleaned real-time meteorological data.

3. The method for identifying dry-heat composite events under the influence of climate change according to claim 1, characterized in that: The steps for obtaining the deviation analysis results are: Call the temperature data and humidity data in the cleaned real-time meteorological data, call the temperature data and humidity data in the historical meteorological data of the same period, establish the corresponding relationship between the real-time and historical data of the same period one by one, perform the matching check of the real-time and historical temperature data and humidity data item by item, remove the data items that have not been matched successfully, and form a meteorological data pair after successful matching; According to the meteorological data pairs after successful matching, the deviation analysis results are calculated using the following formula: ； in, represents the deviation analysis results, Represents the temperature data in the real-time meteorological data after cleaning, Represents the temperature data in the historical meteorological data for the same period. Represents the humidity data in the real-time meteorological data after cleaning, Represents the humidity data in the historical meteorological data for the same period.

4. The method for identifying dry-heat composite events under the influence of climate change according to claim 1, characterized in that: The steps for obtaining the conditional likelihood ratio are: Based on the cleaned real-time meteorological data, the temperature data and humidity data of the same period in history are matched one by one and the difference between the real-time data and the historical data is calculated to form the real-time meteorological data after the standard deviation test; According to the deviation analysis results and the real-time meteorological data after the standard deviation test, the conditional likelihood ratio is calculated, and the calculation formula is: ； in, represents the conditional likelihood ratio, represents the deviation analysis results, Represents the mean of the deviation analysis results when dry and hot events occurred during the same period in history, Represents the temperature data change rate in the current real-time meteorological data. Represents the humidity in the current real-time meteorological data. Represents the average humidity when dry and hot events occurred during the same period in history. Represents the rate of change of humidity data in the current real-time meteorological data. Represents the current ambient air pressure data, Represents the average of air pressure data under historical environmental conditions during the same period.

5. The method for identifying dry-heat composite events under the influence of climate change according to claim 1, characterized in that: The steps for obtaining the dry heat composite event determination result are as follows: Match the daily temperature and humidity amplitudes corresponding to the occurrence of dry and hot events in the same period of history one by one, and count the meteorological amplitude difference characteristics when the historical dry and hot events occurred and did not occur, forming a historical meteorological amplitude feature set; According to the historical meteorological amplitude feature set, the dry heat event determination threshold is calculated, and the calculation formula is: ； in, is the threshold for determining dry heat events, Represents the average amplitude of daily temperature variation during the same period in history, Represents the daily temperature variation amplitude in the current real-time meteorological data, Represents the average amplitude of daily humidity changes during the same period in history. Represents the daily amplitude of humidity variation in the current real-time meteorological data. Represents the solar radiation intensity in the current real-time meteorological data. Represents the average intensity of solar radiation during the same period of history. Represents the dew point temperature in the current real-time meteorological data. Represents the average dew point temperature corresponding to the occurrence of dry and hot events in the same period of history. Represents the atmospheric saturated water vapor pressure in the current real-time meteorological data. Represents the average atmospheric saturated water vapor pressure corresponding to the occurrence of dry and hot events in the same period of history; The conditional likelihood ratios are compared with the dry heat event determination thresholds one by one to determine whether the conditional likelihood ratios exceed the dry heat event determination thresholds, thereby forming a dry heat composite event determination result.

6. The method for identifying dry-heat composite events under the influence of climate change according to claim 1, characterized in that: The steps for obtaining the adjusted threshold are: Based on the dry-heat composite event determination result, the temperature data, humidity data and wind speed data in the real-time meteorological data are called one by one, the occurrence frequency of the dry-heat composite event in the current continuous period is calculated, and the correlation characteristics between the event frequency and the real-time meteorological data are analyzed to form a dry-heat event correlation feature set; According to the dry-heat event correlation feature set, the fluctuation degree of temperature data, the decline amplitude of humidity data and the change trend of wind speed data in the current real-time meteorological data are analyzed one by one to determine the extreme degree of current meteorological conditions and form a real-time meteorological extreme degree classification result; According to the real-time meteorological extreme degree classification result, the adjusted threshold is calculated, and the calculation formula is: ； in, is the adjusted threshold, Represents the standard deviation of the difference between the temperature data in real-time meteorological data and the historical temperature data. Represents the variance of the rate of change of air humidity data in real-time meteorological data, Represents the peak value of wind speed data in real-time meteorological data. Represents the daily variation amplitude of precipitation in real-time meteorological data, Represents the rate of change of solar radiation intensity in real-time meteorological data.

7. The method for identifying dry-heat composite events under the influence of climate change according to claim 1, characterized in that: The steps for obtaining the updated dry heat composite event information are as follows: Calling the adjusted threshold and the dry heat composite event determination result, comparing the relationship between the conditional likelihood ratio in the dry heat composite event determination result and the adjusted threshold one by one, marking the data record exceeding the adjusted threshold, and forming a real-time dry heat composite event trigger record; Based on the real-time dry-heat composite event trigger record, the time interval and continuous frequency of occurrence are statistically recorded one by one, and the time distribution and event continuity characteristics are analyzed to form updated dry-heat composite event information.

8. The method for identifying dry-heat composite events under the influence of climate change according to claim 1, characterized in that: The steps for obtaining the decision suggestion output are: Based on the updated dry-heat composite event information, the preset event level response standards are compared to determine the type of response measures, including personnel evacuation, emergency resource deployment and information release, and generate decision-making recommendation outputs.

Images