Hourly scale extreme high temperature identification method for air temperature sequence data

Abstract

This application relates to a method for identifying hourly-scale extreme high temperatures using temperature series data. The method includes: preprocessing hourly temperature data to obtain an hourly-scale temperature series, and calculating a daily-scale average temperature series based on a meteorological daily window. Based on the hourly-scale temperature series and historical reference periods, determining hourly-scale extreme high temperature thresholds for different hours and identifying hourly-scale extreme high temperatures. Based on the daily-scale average temperature series and historical reference periods, determining daily-scale extreme high temperature thresholds for different dates and identifying daily-scale extreme high temperature days and non-daily-scale extreme high temperature days. Extracting the corresponding times that meet the hourly-scale extreme high temperature criteria from both daily-scale and non-daily-scale extreme high temperature days to obtain the corresponding hourly high temperature identifier sequences. This method enables accurate identification of the specific time periods and distribution characteristics of extreme high temperature events, avoiding the omission of high-temperature events.

Description

Technical Field

[0001] This application relates to the field of climate data processing technology, and in particular to a method for identifying hourly-scale extreme high temperatures in temperature series data. Background Technology

[0002] Extreme heat is one of the most severe extreme weather events affecting human health, energy systems, the ecological environment, and urban operational safety under the background of climate change. Effective identification of extreme heat events can improve the ability to conduct refined assessments and early warnings of heat risks, providing crucial information support for extreme heat disaster early warning and emergency response, public health heat exposure analysis, climate change impact analysis, urban disaster prevention and mitigation, and energy dispatch and management.

[0003] Current research on extreme high temperatures primarily relies on diurnal temperature indicators for identification and analysis, with identification methods mainly falling into two categories. One category is based on daily average temperature, which typically compares a given day's average temperature with historical statistical thresholds for the same period. If the daily average temperature exceeds the threshold, an extreme high-temperature event is considered to have occurred that day. This method can characterize the combined thermal state of day and night to some extent, but it can only determine whether extreme high temperatures occur throughout the day, failing to identify the specific time of day when extreme high temperatures occur, such as morning, noon, or night. This results in the loss of information about the high-temperature process, hindering applications in meteorological monitoring, human heat exposure assessment, and medical emergency response, and overlooking short-term high-temperature events masked by diurnal indicators. The other category is based on daily maximum or minimum temperatures, using these as the identification indicator. While this can differentiate between day and night temperatures to some extent, it essentially still uses the "day" as the smallest unit of analysis, failing to meet the need for time-period differentiation and providing an incomplete description of the high-temperature process. Summary of the Invention

[0004] Therefore, it is necessary to provide an hourly-scale extreme high-temperature identification method for temperature sequence data that can accurately identify the specific time period and distribution characteristics of extreme high-temperature events and avoid missing high-temperature processes, in order to address the above-mentioned technical problems.

[0005] A method for identifying hourly-scale extreme high temperatures in temperature series data includes the following steps: S1: Preprocess the acquired hourly temperature data to obtain an hourly temperature sequence; S2: Based on the hourly temperature series and the preset meteorological daily window, calculate the daily average temperature to obtain the daily average temperature series; S3: Based on the hourly temperature sequence, the preset historical reference period and the hourly temperature value corresponding to the historical reference period, determine the hourly extreme high temperature threshold at different hourly times; S4: Based on the hourly extreme high temperature threshold, identify the hourly extreme high temperature for each hourly temperature sequence; S5: Based on the daily average temperature sequence, the preset historical reference period and the daily average temperature value corresponding to the historical reference period, determine the daily extreme high temperature threshold for different dates; S6: Based on the daily-scale extreme high temperature threshold, daily-scale extreme high temperature is identified for each daily-scale average temperature sequence to determine daily-scale extreme high temperature days and non-daily-scale extreme high temperature days; S7: Extract the corresponding times that satisfy the hourly extreme high temperature on the daily-scale extreme high temperature day to obtain the hourly high temperature identifier sequence that occurs on the daily-scale extreme high temperature day; extract the corresponding times that satisfy the hourly extreme high temperature on the non-daily-scale extreme high temperature day to obtain the hourly high temperature identifier sequence that occurs on the non-daily-scale extreme high temperature day.

[0006] Preferably, the preprocessing process is as follows: The first scenario: For hourly temperature data that are continuously missing for less than the first preset time threshold, time series interpolation is used to fill in the missing data. The second scenario: For missing data exceeding the second preset time threshold, but with multiple other locations and continuous hourly temperature data, spatial interpolation is used to fill in the gaps. Hourly temperature data that does not meet the first or second criteria will be considered invalid. The hourly temperature data includes spatial location coordinates and geographic reference system, recording date and time, and the corresponding hourly temperature value; Summer periods are defined based on the climate characteristics of the regions corresponding to hourly temperature sequences.

[0007] Preferably, in S3, for an hourly temperature series, let the first... Summer of the year day The hourly temperature is Based on the hourly temperature values ​​corresponding to the preset historical reference period, the hourly-scale extreme high temperature threshold is calculated for different hourly times. The expression is:

[0008] in, This indicates taking the 90th percentile. Indicates the historical reference period. Indicates the starting year of the historical reference period. Indicates the end year of the historical reference period. and They represent the first The seven days before and the seven days after the day; Judge the first day Hourly temperature Is it greater than or equal to the corresponding hourly-scale extreme high temperature threshold? If so, then the first Extreme high temperatures on an hourly scale occur within hours; otherwise, the... No hourly-scale extreme high temperatures occurred within hours, satisfying the expression:

[0009] in, For the first Summer of the year day The results of identifying extreme high temperatures on an hourly scale, where "1" indicates the hour. Extreme high temperatures on an hourly scale occurred within hours, with "0" indicating the hour. No extreme high temperatures on an hourly scale occurred within 24 hours.

[0010] Preferably, in S5, for a diurnal temperature series, let the first... Summer of the year The average daily temperature of the day is The daily-scale extreme high temperature threshold for different dates is calculated based on the hourly temperature values ​​corresponding to the preset historical reference period. The expression is:

[0011] Judge the first Daily average temperature Is it greater than or equal to the corresponding diurnal extreme high temperature threshold? If so, then the day is a day with a daily-scale extreme high temperature, and is therefore a day with a daily-scale extreme high temperature. Otherwise, the day is a day without a daily-scale extreme high temperature, and is therefore a day without a daily-scale extreme high temperature, satisfying the expression:

[0012] in, For the first Summer of the year The results of the daily-scale extreme high temperature identification, where "1" indicates the first day. An extreme high temperature on a daily scale occurred, with "0" indicating the [number]th [day]. No extreme high temperatures on a daily scale have occurred.

[0013] Preferably, the process of S7 is as follows: When the Results of daily extreme high temperature identification When the value is "1", it indicates that an extreme high temperature on an hourly scale occurred during that day. Time is extracted to obtain the hourly high temperature identifier sequence that occurred on a diurnal extreme high temperature day, satisfying the expression: When the first Results of daily extreme high temperature identification When the value is "0", it indicates that an extreme high temperature occurred on an hourly scale during that day. Time is extracted to obtain hourly high temperature identifier sequences that occur on non-diurnal extreme high temperature days, satisfying the expression: .

[0014] Preferably, the method further includes: evaluating the effectiveness of identifying hourly-scale extreme high temperatures within a target time period based on hourly high temperature identifier sequences occurring on diurnal-scale extreme high temperature days and hourly high temperature identifier sequences occurring on non-diurnal-scale extreme high temperature days.

[0015] Preferably, the process of evaluating the effectiveness of identifying hourly-scale extreme high temperatures within the target time period includes: Based on the identification results of extreme high temperature on an hourly scale Calculate the total number of hourly extreme high temperatures during the target period. The expression is:

[0016] in, The total number of days in the target period; Hourly high temperature identifier sequence within a day based on diurnal extreme high temperature Calculate the total number of hourly extreme high temperatures within the day of the daily extreme high temperature during the target period. The expression is:

[0017] Hourly high temperature identifier sequence within non-diurnal extreme high temperature days Calculate the total number of non-diurnal extreme high temperatures per hour during the target period. The expression is:

[0018] Based on the total number of hourly extreme high temperatures Total number of hourly extreme high temperatures within a day on a daily scale Total number of hourly extreme high temperatures within a non-solar scale day The recognition rate of extreme high temperatures on an hourly scale within the target time period was calculated respectively. and omission rate The expression is:

[0019]

[0020] Using the recognition rate The efficiency of daily-scale extreme high temperature identification in characterizing hourly-scale extreme high temperatures is quantified by utilizing the omission rate. The improvement in fully capturing the high-temperature process by quantifying the recognition of extreme high temperatures on an hourly scale compared to that on a daily scale.

[0021] Preferably, the high-incidence windows of hourly-scale extreme high temperatures within the target period are evaluated. These high-incidence windows include the total hourly extreme high-incidence windows within the target period, the hourly extreme high-incidence windows of daily extreme high-temperature days within the target period, and the hourly extreme high-incidence windows of non-daily extreme high-temperature days within the target period. The process includes: Based on the hourly high temperature identifier sequences occurring on diurnal extreme high temperature days and on non-diurnal extreme high temperature days, the diurnal distribution characteristics of hourly extreme high temperatures in the target period are identified. The process is as follows: Based on the identification results of extreme high temperature on an hourly scale Calculate the total hourly-scale extreme high temperature at different times within the target time period. Frequency of occurrence The expression is:

[0022] in, The total number of days in the target period; The period when the frequency of extreme high temperatures on an hourly scale is higher than the preset extreme high temperature frequency threshold is defined as the hourly extreme high temperature high incidence window; based on the total hourly extreme high temperature high incidence window within the target period, the moments when extreme high temperatures occur in a concentrated manner within the target period are identified.

[0023] Preferably, the high-incidence windows of hourly extreme high temperatures within the target time period on daily extreme high temperature days are evaluated. These high-incidence windows include the total hourly extreme high temperature high-incidence windows within the target time period, the hourly extreme high temperature high-incidence windows on daily extreme high temperature days within the target time period, and the hourly extreme high temperature high-incidence windows on non-daily extreme high temperature days within the target time period. The process includes: Based on the hourly high temperature indicator sequence within the diurnal extreme high temperature day The statistical analysis of hourly extreme high temperatures on daily extreme high-temperature days within the target period at different times. Frequency of occurrence The expression is: ; The period when the frequency of hourly extreme high temperatures exceeds the preset extreme high temperature frequency threshold is defined as the hourly extreme high temperature high incidence window; based on the hourly extreme high temperature high incidence window of the daily extreme high temperature day, the moments when extreme high temperatures occur in a concentrated manner during the daily extreme high temperature day are identified.

[0024] Preferably, the high-incidence windows of hourly extreme high temperatures within the target time period on non-diurnal extreme high temperature days are evaluated. These high-incidence windows include the total hourly extreme high temperature high-incidence windows within the target time period, the hourly extreme high temperature high-incidence windows on diurnal extreme high temperature days within the target time period, and the hourly extreme high temperature high-incidence windows on non-diurnal extreme high temperature days within the target time period. The process includes: Based on the hourly high temperature indicator sequence within non-daily extreme high temperature days The statistical target period includes the hourly extreme high temperatures of non-daily-scale extreme high-temperature days at different times. Frequency of occurrence The expression is: ; The period when the frequency of hourly extreme high temperatures exceeds the preset extreme high temperature frequency threshold is defined as the hourly extreme high temperature high incidence window; based on the hourly extreme high temperature high incidence window of non-daily extreme high temperature days, the time when extreme high temperatures occur that is overlooked in non-daily extreme high temperature days is identified.

[0025] Compared with the prior art, the beneficial effects of the technical solution of the present invention are: This invention proposes a method for identifying hourly-scale extreme high temperatures based on temperature series data. Based on hourly-scale temperature series, a preset historical reference period, and the corresponding hourly temperature values ​​for that historical reference period, hourly-scale extreme high temperature thresholds are determined for different hourly times. Based on these thresholds, hourly-scale extreme high temperature identification is performed on each hourly-scale temperature series. Similarly, based on daily-scale average temperature series, a preset historical reference period, and the corresponding daily average temperature values ​​for that historical reference period, daily-scale extreme high temperature thresholds are determined for different dates. Based on these thresholds, daily-scale extreme high temperature identification is performed on each daily-scale average temperature series to identify daily-scale extreme high temperature days and non-daily-scale extreme high temperature days. On daily-scale extreme high temperature days, the corresponding times that satisfy hourly-scale extreme high temperatures are extracted to obtain hourly high temperature identifier sequences occurring within those days, representing the duration of high temperatures within those days. On non-daily-scale extreme high temperature days, the corresponding times that satisfy hourly-scale extreme high temperatures are extracted to obtain hourly high temperature identifier sequences occurring within those days, representing short-duration high temperature events within those days. This invention can precisely quantify the timing and duration of extreme high temperatures, enabling accurate identification of the specific time periods and distribution characteristics of extreme high-temperature events, avoiding omissions of high-temperature processes, and accurately capturing short-term high-temperature processes. It plays an important role in scenarios such as high-temperature risk monitoring and early warning, urban thermal environment health risk assessment, and urban climate change response. Attached Figure Description

[0026] Figure 1This is a flowchart illustrating an hourly-scale extreme high-temperature identification method for temperature series data in one embodiment. Figure 2 This is a diagram showing the results of identifying extreme high temperatures on an hourly and daily scales within a target time period in one embodiment. Figure 3 This is a graph showing the results of the analysis of the correspondence between hourly and daily extreme high temperatures during the target time period in one embodiment. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0028] Example 1 This embodiment provides a method for identifying hourly-scale extreme high temperatures in temperature series data. The flowchart of this method can be found here. Figure 1 This includes the following steps: S1: Preprocess the acquired hourly temperature data to obtain an hourly temperature sequence; S2: Based on the hourly temperature series and the preset meteorological daily window, calculate the daily average temperature to obtain the daily average temperature series; S3: Based on the hourly temperature sequence, the preset historical reference period and the hourly temperature value corresponding to the historical reference period, determine the hourly extreme high temperature threshold at different hourly times; S4: Based on the hourly extreme high temperature threshold, identify the hourly extreme high temperature for each hourly temperature sequence; S5: Based on the daily average temperature sequence, the preset historical reference period and the daily average temperature value corresponding to the historical reference period, determine the daily extreme high temperature threshold for different dates; S6: Based on the daily-scale extreme high temperature threshold, daily-scale extreme high temperature is identified for each daily-scale average temperature sequence to determine daily-scale extreme high temperature days and non-daily-scale extreme high temperature days; S7: Extract the corresponding times that satisfy the hourly extreme high temperature on the daily-scale extreme high temperature day to obtain the hourly high temperature identifier sequence that occurs on the daily-scale extreme high temperature day; extract the corresponding times that satisfy the hourly extreme high temperature on the non-daily-scale extreme high temperature day to obtain the hourly high temperature identifier sequence that occurs on the non-daily-scale extreme high temperature day.

[0029] In this embodiment, the acquired hourly temperature data is preprocessed to obtain an hourly temperature sequence. In this embodiment, the local time from 20:00:00 to 19:59:59 the next day is preferably used as a meteorological day window. Based on the hourly temperature sequence, the daily average temperature is calculated to obtain the daily average temperature sequence.

[0030] Based on hourly temperature sequences, preset historical reference periods, and corresponding hourly temperature values ​​for those periods, hourly extreme high-temperature thresholds are determined for different hourly times. Based on these thresholds, extreme high-temperature days are identified for each hourly temperature sequence. Similarly, based on daily average temperature sequences, preset historical reference periods, and corresponding daily average temperature values ​​for those periods, daily extreme high-temperature thresholds are determined for different dates. Based on these thresholds, daily extreme high-temperature days are identified for each daily average temperature sequence, distinguishing between daily extreme high-temperature days and non-daily extreme high-temperature days. The thresholds used in this embodiment are relative thresholds based on historical distributions, rather than fixed absolute temperature thresholds, thus adapting to differences in regional climate backgrounds.

[0031] During days of extreme high temperatures on a diurnal scale, the corresponding times that satisfy the criteria for extreme high temperatures on an hourly scale are extracted to obtain hourly high temperature identifier sequences occurring within those days. These sequences represent the duration of high temperatures during such days. Similarly, during days of extreme high temperatures on a non-diurnal scale, the corresponding times that satisfy the criteria for extreme high temperatures on an hourly scale are extracted to obtain hourly high temperature identifier sequences occurring within those days. These sequences represent short-duration high temperature events during such days.

[0032] Compared to traditional extreme high-temperature identification methods based solely on daily-scale data, this embodiment demonstrates the advantages of hourly-scale extreme high-temperature identification methods based on temperature series data over traditional daily-scale methods through dual identification and correlation analysis at both hourly and daily scales. Hourly-scale extreme high-temperature identification can precisely quantify the timing and duration of extreme high-temperature events, enabling accurate identification of the specific time periods and distribution characteristics of extreme high-temperature events, avoiding omissions of high-temperature processes, and accurately capturing short-term high-temperature events. The unified analysis framework at both hourly and daily scales provides a new technology for characterizing extreme high-temperature events, which can be effectively applied to scenarios such as high-temperature risk monitoring and early warning, urban thermal environment health risk assessment, and urban climate change response.

[0033] Example 2 In this embodiment, the preprocessing process for the acquired hourly temperature data is as follows: The first scenario: For hourly temperature data that are continuously missing for less than the first preset time threshold, time series interpolation is used to fill in the missing data. The second scenario: For missing data exceeding the second preset time threshold, but with multiple other locations and continuous hourly temperature data, spatial interpolation is used to fill in the gaps. Hourly temperature data that does not meet the first or second criteria will be considered invalid. In this embodiment, the acquired hourly temperature data refers to the hourly temperature data of the target study area, used to identify extreme high temperatures on hourly and daily scales. This includes spatial location coordinates and geographic reference system, recording date and time, and the corresponding hourly temperature value. Specifically, this embodiment preferably uses a first preset time threshold of 1-2 hours, indicating a relatively short period of continuous data loss; and a second preset time threshold of 1-2 days, indicating a longer period of data loss for a certain temperature sequence. When too much hourly temperature data is missing, the data is considered invalid.

[0034] After handling missing data, a continuous hourly temperature series was obtained for subsequent calculations and analysis.

[0035] Summer periods are defined based on the climate characteristics of the corresponding regions according to hourly temperature sequences. In the Northern Hemisphere, it is generally defined as June to August (92 days in total), and in the Southern Hemisphere, it is generally defined as December to February of the following year (90 days in total).

[0036] In S3, for an hourly temperature series, let the first... Summer of the year day The hourly temperature is Based on the hourly temperature values ​​corresponding to the preset historical reference period, the hourly-scale extreme high temperature threshold is calculated for different hourly times. The expression is:

[0037] in, This indicates taking the 90th percentile. Indicates the historical reference period. Indicates the starting year of the historical reference period. Indicates the end year of the historical reference period. and They represent the first The seven days before and the seven days after the day; Judge the first day Hourly temperature Is it greater than or equal to the corresponding hourly-scale extreme high temperature threshold? If so, then the first Extreme high temperatures on an hourly scale occur within hours; otherwise, the... No hourly-scale extreme high temperatures occurred within hours, satisfying the expression:

[0038] in, For the first Summer of the year day The results of identifying extreme high temperatures on an hourly scale, where "1" indicates the hour. Extreme high temperatures on an hourly scale occurred within hours, with "0" indicating the hour. No extreme high temperatures on an hourly scale occurred within 24 hours.

[0039] Specifically, a corresponding historical reference period is selected based on the time period of the hourly temperature data. The historical reference period is generally selected as 1981-2010, or selected according to the actual data situation, with a minimum of 20 years to ensure sufficient historical temperature samples to support the threshold calculation results. This example uses 1981-2010 as the reference period and June-August as the summer period to calculate the hourly extreme high temperature threshold at 8:00 AM on June 1st. We collected the temperature values ​​at 8:00 AM from May 25th to June 8th for each period between 1981 and 2010, resulting in 450 samples over 30 years and 15 days. After sorting these samples, we calculated the 90th percentile, which is the hourly extreme high temperature threshold at 8:00 AM on June 1st. .

[0040] It should be noted that the hourly extreme high temperature threshold is established separately for each hour, rather than calculating a unified threshold by mixing all 24 hours of a day. That is, the threshold for 0 o'clock corresponds to the threshold for 0 o'clock, the threshold for 1 o'clock corresponds to the threshold for 1 o'clock, and so on, thereby eliminating the interference of the diurnal temperature variation background on the threshold setting.

[0041] In S5, for a diurnal temperature series, let the first... Summer of the year The average daily temperature of the day is The daily-scale extreme high temperature threshold for different dates is calculated based on the hourly temperature values ​​corresponding to the preset historical reference period. The expression is:

[0042] Judge the first Daily average temperature Is it greater than or equal to the corresponding diurnal extreme high temperature threshold? If so, then the day is a day with a daily-scale extreme high temperature, and is therefore a day with a daily-scale extreme high temperature. Otherwise, the day is a day without a daily-scale extreme high temperature, and is therefore a day without a daily-scale extreme high temperature, satisfying the expression:

[0043] in, For the first Summer of the year The results of the daily-scale extreme high temperature identification, where "1" indicates the first day. An extreme high temperature on a daily scale occurred, with "0" indicating the [number]th [day]. No extreme high temperatures on a daily scale have occurred.

[0044] This embodiment uses 1981-2010 as the reference period and June-August as the summer period to calculate the diurnal extreme high temperature threshold on June 1st. We collected temperature values ​​from May 25th to June 8th for each period between 1981 and 2010, resulting in 450 samples (30 years x 15 days). After sorting these samples, we calculated the 90th percentile, which represents the diurnal extreme high temperature threshold for June 1st. .

[0045] The process of S7 is as follows: When the Results of daily extreme high temperature identification When the value is "1", it indicates that an extreme high temperature on an hourly scale occurred during that day. Time is extracted to obtain the hourly high temperature identifier sequence that occurred on a diurnal extreme high temperature day, satisfying the expression: When the first Results of daily extreme high temperature identification When the value is "0", it indicates that an extreme high temperature occurred on an hourly scale during that day. Time is extracted to obtain hourly high temperature identifier sequences that occur on non-diurnal extreme high temperature days, satisfying the expression: .

[0046] Specifically, when the Results of daily extreme high temperature identification A value of "1" indicates that the day is a day of extreme high temperature on a daily scale. This refers to the results of identifying day-scale extreme high temperatures within the 24 hours of that day. Hours marked as "1" are extracted to obtain an hourly high-temperature identifier sequence for days with extreme high temperatures on a diurnal scale. This sequence identifies the specific time when the hourly extreme high temperature occurs on each diurnal extreme high-temperature day. The number of hourly high temperatures on a diurnal extreme high-temperature day represents the actual duration of the extreme high temperature on that day, thus solving the problem of unclear occurrence time and duration of extreme high temperatures in existing diurnal scale identification methods.

[0047] When the Results of daily extreme high temperature identification A value of "0" indicates that the day is not a day of extreme high temperature on a daily scale. If any hour on that day meets the criteria for extreme high temperature on an hourly scale... A value of "1" indicates that although the day is not a day of extreme high temperature on a daily scale, there is a short-duration extreme high temperature process that was missed by the daily extreme high temperature identification method. Therefore, the hourly extreme high temperature event that occurred on that day should be considered. Time is extracted to obtain an hourly high temperature identifier sequence that occurs on non-diurnal scale extreme high temperature days. This sequence identifies the specific time when the hourly extreme high temperature occurs on each non-diurnal scale extreme high temperature day, solving the problem of missing short-term high temperature processes on non-diurnal scale extreme high temperature days in existing technologies.

[0048] In this embodiment, a method for identifying hourly-scale extreme high temperatures based on temperature series data further includes: evaluating the effectiveness of identifying hourly-scale extreme high temperatures within a target time period based on hourly high temperature identifier sequences occurring on days with daily extreme high temperatures and hourly high temperature identifier sequences occurring on days with non-daily extreme high temperatures.

[0049] The process of evaluating the effectiveness of identifying hourly-scale extreme high temperatures within a target time period includes: Based on the identification results of extreme high temperature on an hourly scale Calculate the total number of hourly extreme high temperatures during the target period. The expression is:

[0050] in, The total number of days in the target period; Hourly high temperature identifier sequence within a day based on diurnal extreme high temperature Calculate the total number of hourly extreme high temperatures within the day of the daily extreme high temperature during the target period. The expression is:

[0051] Hourly high temperature identifier sequence within non-diurnal extreme high temperature days Calculate the total number of non-diurnal extreme high temperatures per hour during the target period. The expression is:

[0052] Based on the total number of hourly extreme high temperatures Total number of hourly extreme high temperatures within a day on a daily scale Total number of hourly extreme high temperatures within a non-solar scale day The recognition rate of extreme high temperatures on an hourly scale within the target time period was calculated respectively. and omission rate The expression is:

[0053]

[0054] Using the recognition rate The efficiency of daily-scale extreme high temperature identification in characterizing hourly-scale extreme high temperatures is quantified by utilizing the omission rate. The improvement in fully capturing the high-temperature process by quantifying the recognition of extreme high temperatures on an hourly scale compared to that on a daily scale.

[0055] Specifically, the recognition rate of extreme high temperatures on an hourly scale. The lower the omission rate The higher the value, the worse the daily-scale extreme high temperature identification effect is on capturing hourly-scale extreme high temperature processes. The region corresponding to this temperature series needs to use hourly-scale extreme high temperature identification methods to improve the monitoring accuracy of extreme high temperature events in the region.

[0056] The assessment of high-incidence windows for hourly extreme high temperatures within the target time period is conducted. These high-incidence windows include the total hourly extreme high-incidence windows within the target time period, the hourly extreme high-incidence windows for daily extreme high-incidence days within the target time period, and the hourly extreme high-incidence windows for non-daily extreme high-incidence days within the target time period. The process includes: Based on the hourly high temperature identifier sequences occurring on diurnal extreme high temperature days and on non-diurnal extreme high temperature days, the diurnal distribution characteristics of hourly extreme high temperatures in the target period are identified. The process is as follows: Based on the identification results of extreme high temperature on an hourly scale Calculate the total hourly-scale extreme high temperature at different times within the target time period. Frequency of occurrence The expression is:

[0057] in, The total number of days in the target period; The period when the frequency of extreme high temperatures on an hourly scale is higher than the preset extreme high temperature frequency threshold is defined as the hourly extreme high temperature high incidence window; based on the total hourly extreme high temperature high incidence window within the target period, the moments when extreme high temperatures occur in a concentrated manner within the target period are identified.

[0058] Specifically, the frequency of extreme high temperatures on a total hourly scale within the target time period. It reflects the frequency distribution of hourly extreme high temperatures in the corresponding region over a 24-hour period, characterizing its intraday variation pattern.

[0059] In this embodiment, the average hourly extreme high temperature frequency over 24 hours is preferably used as a preset extreme high temperature frequency threshold, and the total hourly extreme high temperature peak window within the target time period is defined as follows. The expression is:

[0060] The process involves assessing the high-incidence windows of hourly extreme high temperatures within the target time period on daily extreme high-temperature days. These high-incidence windows include the total hourly extreme high-incidence windows within the target time period, the hourly extreme high-incidence windows on daily extreme high-temperature days within the target time period, and the hourly extreme high-incidence windows on non-daily extreme high-temperature days within the target time period. The process includes: Based on the hourly high temperature indicator sequence within the diurnal extreme high temperature day The statistical analysis of hourly extreme high temperatures on daily extreme high-temperature days within the target period at different times. Frequency of occurrence The expression is: ; The period when the frequency of hourly extreme high temperatures exceeds the preset extreme high temperature frequency threshold is defined as the hourly extreme high temperature high incidence window; based on the hourly extreme high temperature high incidence window of the daily extreme high temperature day, the moments when extreme high temperatures occur in a concentrated manner during the daily extreme high temperature day are identified.

[0061] The process involves assessing the high-incidence windows of hourly extreme high temperatures within the target time period on non-diurnal extreme high-temperature days. These high-incidence windows include the total hourly extreme high-incidence windows within the target time period, the hourly extreme high-incidence windows on diurnal extreme high-temperature days within the target time period, and the hourly extreme high-incidence windows on non-diurnal extreme high-temperature days within the target time period. The process includes: Based on the hourly high temperature indicator sequence within non-daily extreme high temperature days The statistical target period includes the hourly extreme high temperatures of non-daily-scale extreme high-temperature days at different times. Frequency of occurrence The expression is: ; The period when the frequency of hourly extreme high temperatures exceeds the preset extreme high temperature frequency threshold is defined as the hourly extreme high temperature high incidence window; based on the hourly extreme high temperature high incidence window of non-daily extreme high temperature days, the time when extreme high temperatures occur that is overlooked in non-daily extreme high temperature days is identified.

[0062] Specifically, by obtaining the differences in intraday distribution characteristics and high-incidence windows of hourly extreme high temperatures under the two scenarios of daily extreme high temperature days and non-daily extreme high temperature days, it is possible to identify at which times hourly high temperatures occur in concentrated events on daily extreme high temperature days, and at which times hourly high temperatures are most neglected on non-daily extreme high temperature days, providing intraday pattern support for accurate early warning and response to extreme high temperatures.

[0063] Example 3 This embodiment uses hourly temperature data from June to August of the summer months of 1981-2023 in a certain area as the raw data to identify extreme high temperatures on hourly and daily scales. The specific implementation process is as follows: The raw hourly temperature data is preprocessed to obtain a complete and continuous hourly scale for the target time period. Based on a pre-defined meteorological daily window, a daily temperature sequence is calculated. For the hourly and daily temperature data from the historical reference period of 1981-2010, the percentile method is used to calculate the hourly and daily extreme high temperature thresholds, respectively. Based on the hourly extreme high temperature thresholds, each hourly temperature sequence is used to identify hourly extreme high temperatures. If the hourly temperature at a certain moment is higher than the corresponding hourly extreme high temperature threshold, then an hourly extreme high temperature is determined to have occurred at that moment. Based on the daily extreme high temperature thresholds, each daily average temperature sequence is used to identify daily extreme high temperatures. If the daily average temperature of a certain day is higher than the daily extreme high temperature threshold for that day, then that day is determined to have occurred on a daily extreme high temperature day; otherwise, it is a non-daily extreme high temperature day. After completing the extreme high temperature identification, a correlation analysis of hourly and daily extreme high temperature events is conducted. Indicators such as the duration and frequency of hourly extreme high temperatures within daily extreme high temperature days and non-daily extreme high temperature days are statistically analyzed, and the results are output as visual charts.

[0064] The results of identifying extreme high temperatures on hourly and daily scales at this location from June 14 to 28, 2023 are as follows: Figure 2 As shown in the diagram. The solid yellow line represents the daily average temperature series, the dashed yellow line represents the daily extreme high temperature threshold, and the yellow area represents the daily extreme high temperature. The solid red line represents the hourly temperature series, the dashed red line represents the hourly extreme high temperature threshold, and the red area represents the hourly extreme high temperature. From... Figure 2 It is known that the daily average temperature series can only identify June 16-17, June 21-25, and June 27 as daily extreme high-temperature days, failing to reflect the fine changes in intraday temperature. In contrast, the hourly temperature series can clearly capture non-daily extreme high-temperature days, such as June 14-15, June 19, and June 28. Short-duration extreme high-temperature events in the afternoon and evening (corresponding to the red area) are missed by traditional daily extreme high-temperature identification methods because the daily maximum temperature does not reach the daily extreme high-temperature threshold. Furthermore, within daily extreme high-temperature days, the hourly extreme high-temperature identification results can accurately and completely show the duration of extreme high-temperature events, such as the high-temperature event that started on the morning of June 21 and ended at noon on June 25, providing more accurate methodological support for characterizing and responding to extreme high-temperature events.

[0065] The correlation analysis results of hourly and daily extreme high temperature events in this area are as follows: Figure 3 As shown. In Figure 3In the left-hand annular plot, the average duration of hourly extreme high temperatures within diurnal extreme high-temperature days (high-temperature days) is generally above 19.8 hours / day, while the average hourly high-temperature frequency within diurnal extreme high-temperature days (non-high-temperature days) is approximately 5.7 hours / day. Furthermore, among all hourly extreme high temperatures during this period, 79.8% can be identified by the diurnal extreme high-temperature method, with an omission rate of 20.2%, reflecting the inadequacy of traditional diurnal extreme high-temperature methods in characterizing short-duration high-temperature processes. Figure 3 In the frequency line graph on the right, the black circular line represents the total hourly high temperature, the red square line represents the hourly extreme high temperature on high-temperature days, and the blue star-shaped line represents the hourly extreme high temperature on non-high-temperature days. The black area represents the high-incidence window of the total hourly extreme high temperature, the red area represents the high-incidence window of the hourly extreme high temperature on high-temperature days, and the blue area represents the high-incidence window of the hourly extreme high temperature on non-high-temperature days. This further shows that the frequency of hourly extreme high temperatures on high-temperature days is significantly higher during the daytime (07:00-19:00) than on non-high-temperature days, and it almost covers the entire daytime period; while the hourly extreme high temperatures on non-high-temperature days are concentrated in the nighttime period (16:00-06:00), indicating that the omission of short-term high temperatures by daily-scale extreme high temperatures mainly occurs during the nighttime period, underestimating the risk of high temperatures at night.

Claims

1. A method for identifying hourly-scale extreme high temperatures in temperature series data, characterized in that, Includes the following steps: S1: Preprocess the acquired hourly temperature data to obtain an hourly temperature sequence; S2: Based on the hourly temperature series and the preset meteorological daily window, calculate the daily average temperature to obtain the daily average temperature series; S3: Based on the hourly temperature sequence, the preset historical reference period and the hourly temperature value corresponding to the historical reference period, determine the hourly extreme high temperature threshold at different hourly times; S4: Based on the hourly extreme high temperature threshold, identify the hourly extreme high temperature for each hourly temperature sequence; S5: Based on the daily average temperature sequence, the preset historical reference period and the daily average temperature value corresponding to the historical reference period, determine the daily extreme high temperature threshold for different dates; S6: Based on the daily-scale extreme high temperature threshold, daily-scale extreme high temperature is identified for each daily-scale average temperature sequence to determine daily-scale extreme high temperature days and non-daily-scale extreme high temperature days; S7: Extract the corresponding times that satisfy the hourly extreme high temperature on the daily-scale extreme high temperature day to obtain the hourly high temperature identifier sequence that occurs on the daily-scale extreme high temperature day; extract the corresponding times that satisfy the hourly extreme high temperature on the non-daily-scale extreme high temperature day to obtain the hourly high temperature identifier sequence that occurs on the non-daily-scale extreme high temperature day.

2. The method for identifying hourly-scale extreme high temperatures based on temperature series data according to claim 1, characterized in that, The preprocessing process is as follows: The first scenario: For hourly temperature data that are continuously missing for less than the first preset time threshold, time series interpolation is used to fill in the missing data. The second scenario: For missing data exceeding the second preset time threshold, but with multiple other locations and continuous hourly temperature data, spatial interpolation is used to fill in the gaps. Hourly temperature data that does not meet the first or second criteria will be considered invalid. The hourly temperature data includes spatial location coordinates and geographic reference system, recording date and time, and the corresponding hourly temperature value; Summer periods are defined based on the climate characteristics of the regions corresponding to hourly temperature sequences.

3. The method for identifying hourly-scale extreme high temperatures based on temperature series data according to claim 1, characterized in that, In S3, for an hourly temperature series, let the first... Summer of the year day The hourly temperature is Based on the hourly temperature values ​​corresponding to the preset historical reference period, the hourly-scale extreme high temperature threshold is calculated for different hourly times. The expression is: in, This indicates taking the 90th percentile. Indicates the historical reference period. Indicates the starting year of the historical reference period. Indicates the end year of the historical reference period. and They represent the first The seven days before and the seven days after the day; Judge the first day Hourly temperature Is it greater than or equal to the corresponding hourly-scale extreme high temperature threshold? If so, then the first Extreme high temperatures on an hourly scale occur within hours; otherwise, the... No hourly-scale extreme high temperatures occurred within hours, satisfying the expression: in, For the first Summer of the year day The results of identifying extreme high temperatures on an hourly scale, where "1" indicates the hour. An extreme high temperature occurred on an hourly scale, with "0" indicating the hour. j No extreme high temperatures on an hourly scale occurred within 24 hours.

4. The method for identifying hourly-scale extreme high temperatures based on temperature series data according to claim 3, characterized in that, In S5, for a diurnal temperature series, let the first... Summer of the year The average daily temperature of the day is The daily-scale extreme high temperature threshold for different dates is calculated based on the hourly temperature values ​​corresponding to the preset historical reference period. The expression is: Judge the first Daily average temperature Is it greater than or equal to the corresponding diurnal extreme high temperature threshold? If so, then the day is a day with a daily-scale extreme high temperature, and is therefore a day with a daily-scale extreme high temperature. Otherwise, the day is a day without a daily-scale extreme high temperature, and is therefore a day without a daily-scale extreme high temperature, satisfying the expression: in, For the first Summer of the year The results of the daily-scale extreme high temperature identification for the day, "1" indicates the first... An extreme high temperature on a daily scale occurred, with "0" indicating the first time. No extreme high temperatures on a daily scale have occurred.

5. The method for identifying hourly-scale extreme high temperatures based on temperature series data according to claim 4, characterized in that, The process of S7 is as follows: When the Results of daily extreme high temperature identification When it is "1", it indicates that an extreme high temperature on an hourly scale occurred during that day. Time is extracted to obtain the hourly high temperature identifier sequence that occurred on a diurnal extreme high temperature day, satisfying the expression: When the first Results of daily extreme high temperature identification When the value is "0", it indicates that an extreme high temperature occurred on an hourly scale during that day. Time is extracted to obtain hourly high temperature identifier sequences that occur on non-diurnal extreme high temperature days, satisfying the expression: .

6. A method for identifying hourly-scale extreme high temperatures based on temperature series data according to any one of claims 3 to 5, characterized in that, The method further includes: evaluating the effectiveness of identifying hourly-scale extreme high temperatures within a target time period based on hourly high temperature identifier sequences occurring on diurnal-scale extreme high temperature days and hourly high temperature identifier sequences occurring on non-diurnal-scale extreme high temperature days.

7. The method for identifying hourly-scale extreme high temperatures in temperature series data according to claim 6, characterized in that, The process of evaluating the effectiveness of identifying hourly-scale extreme high temperatures within a target time period includes: Based on the identification results of extreme high temperature on an hourly scale Calculate the total number of hourly extreme high temperatures during the target period. The expression is: in, The total number of days in the target period; Hourly high temperature identifier sequence within a day based on diurnal extreme high temperature Calculate the total number of hourly extreme high temperatures within the day of the daily extreme high temperature during the target period. The expression is: Hourly high temperature identifier sequence within non-diurnal extreme high temperature days Calculate the total number of non-diurnal extreme high temperatures per hour during the target period. The expression is: Based on the total number of hourly extreme high temperatures Total number of hourly extreme high temperatures within a day on a daily scale Total number of hourly extreme high temperatures within a non-solar scale day The recognition rate of extreme high temperatures on an hourly scale within the target time period was calculated respectively. and omission rate The expression is: Using the recognition rate The efficiency of daily-scale extreme high temperature identification in characterizing hourly-scale extreme high temperatures is quantified by utilizing the omission rate. The improvement in fully capturing the high-temperature process by quantifying the recognition of extreme high temperatures on an hourly scale compared to that on a daily scale.

8. The method for identifying hourly-scale extreme high temperatures based on temperature series data according to claim 6, characterized in that, The assessment of high-incidence windows for hourly extreme high temperatures within the target time period is conducted. These high-incidence windows include the total hourly extreme high-incidence windows within the target time period, the hourly extreme high-incidence windows for daily extreme high-incidence days within the target time period, and the hourly extreme high-incidence windows for non-daily extreme high-incidence days within the target time period. The process includes: Based on the hourly high temperature identifier sequences occurring on diurnal extreme high temperature days and on non-diurnal extreme high temperature days, the diurnal distribution characteristics of hourly extreme high temperatures in the target period are identified. The process is as follows: Based on the identification results of extreme high temperature on an hourly scale Calculate the total hourly-scale extreme high temperature at different times within the target time period. Frequency of occurrence The expression is: in, The total number of days in the target period; The period when the frequency of extreme high temperatures on an hourly scale is higher than the preset extreme high temperature frequency threshold is defined as the hourly extreme high temperature high incidence window; based on the total hourly extreme high temperature high incidence window within the target period, the moments when extreme high temperatures occur in a concentrated manner within the target period are identified.

9. The method for identifying hourly-scale extreme high temperatures based on temperature series data according to claim 6, characterized in that, The process involves assessing the high-incidence windows of hourly extreme high temperatures within the target time period on daily extreme high-temperature days. These high-incidence windows include the total hourly extreme high-incidence windows within the target time period, the hourly extreme high-incidence windows on daily extreme high-temperature days within the target time period, and the hourly extreme high-incidence windows on non-daily extreme high-temperature days within the target time period. The process includes: Based on the hourly high temperature indicator sequence within the diurnal extreme high temperature day The statistical analysis of hourly extreme high temperatures on daily extreme high-temperature days within the target period at different times. Frequency of occurrence The expression is: ; The period when the frequency of hourly extreme high temperatures exceeds the preset extreme high temperature frequency threshold is defined as the hourly extreme high temperature high incidence window; based on the hourly extreme high temperature high incidence window of the daily extreme high temperature day, the moments when extreme high temperatures occur in a concentrated manner during the daily extreme high temperature day are identified.

10. The method for identifying hourly-scale extreme high temperatures based on temperature series data according to claim 6, characterized in that, The process involves assessing the high-incidence windows of hourly extreme high temperatures within the target time period on non-diurnal extreme high-temperature days. These high-incidence windows include the total hourly extreme high-incidence windows within the target time period, the hourly extreme high-incidence windows on diurnal extreme high-temperature days within the target time period, and the hourly extreme high-incidence windows on non-diurnal extreme high-temperature days within the target time period. The process includes: Based on the hourly high temperature indicator sequence within non-daily extreme high temperature days The statistical target period includes the hourly extreme high temperatures of non-daily-scale extreme high-temperature days at different times. Frequency of occurrence The expression is: ; The period when the frequency of hourly extreme high temperatures exceeds the preset extreme high temperature frequency threshold is defined as the hourly extreme high temperature high incidence window; based on the hourly extreme high temperature high incidence window of non-daily extreme high temperature days, the time when extreme high temperatures occur that is overlooked in non-daily extreme high temperature days is identified.

Images