Method for typical processing of annual meteorological data

Abstract

The invention belongs to the technical field of meteorological data processing, and particularly relates to a typical processing method for annual meteorological data, which comprises the following steps of: 1, acquiring annual hourly meteorological element observation basic data of at least thirty years; 2, determining the date of typical weather days year by year, and revising the date based on Chinese traditional twenty-four solar terms through a modern ''Qi setting method''; step 3, carrying out typical processing on the year-by-year typical meteorological day meteorological element data to construct a solar term day typical meteorological day meteorological element data sequence; and step 4, processing a meteorological hypothetical year sequence of the typical meteorological day of the solar term based on the meteorological element data sequence of the typical meteorological day of the solar term. According to the method, 24 solar terms are used as time nodes of typical meteorological days to directly reflect staged changes of earth revolution and solar radiation energy, so that the generated typical meteorological day sequence has natural time coherence, annual solar radiation periodic changes can be accurately described, and time continuity and periodicity are ensured; the solar term is a climate phase change critical point, and key turning points are prevented from being ignored.

Description

Technical Field

[0001] This invention belongs to the field of meteorological data processing technology, specifically relating to a method for typical processing of accumulated meteorological data. Background Technology

[0002] Because meteorological conditions can vary significantly from year to year, and to avoid the random changes in weather each year and the lack of sufficient indicativeness of climate characteristics, it is necessary in some research fields to establish typical annual meteorological datasets that can represent the long-term typical meteorological characteristics of a certain place.

[0003] Commonly used meteorological data include Typical Meteorological Year (TMY), Japanese standard meteorological data, Chinese Standard Year (CSWD), and Chinese Typical Meteorological Year (CNTMY). Due to the large amount of TMY data, in practice, characteristic meteorological day data sequences are usually used to replace TMY data for specific application scenarios to simplify the simulation process.

[0004] There are two main methods for generating characteristic meteorological day data sequences: one is to use the Finkelstein-Schafer (FS) statistical method to select typical meteorological days (TMDs) as calculation days based on TMY data; the other is to summarize weather patterns by performing principal component analysis clustering on daily meteorological elements and then selecting typical meteorological days that can represent different weather patterns. The common advantage of both methods is that they use characteristic meteorological day data instead of annual meteorological data, greatly simplifying the simulation process and increasing efficiency.

[0005] However, the drawback is that the characteristic meteorological days generated by Method 1 lack corresponding date information, and the time interval between each meteorological day is unknown, which cannot cope with application scenarios that require accurate description of the periodic changes in solar radiation.

[0006] Method 2 suffers from an unstable data base, lacking correlation and continuity among characteristic meteorological days, thus failing to form a continuous data sequence. Furthermore, neither method strictly defines the time scale for the selected years of observation data, resulting in an inconsistent time scale. Existing findings often target extreme meteorological days or specific short time scales. For example, the document ("Study on Summer Microclimate Adaptation Design of Civic Squares in Hot Summer and Cold Winter Regions Based on Numerical Simulation," Zhou Kairui, *China Excellent Master's Thesis Full-Text Database (Basic Science Series)*, 2021, No. 02, A009-285) discloses a method for selecting typical solar term days, but it uses climate data from Hefei City over the past three years (2016-2018) as its research object, summarizing the general patterns of annual climate change in Hefei. The WMO clearly stipulates that 30 years is the minimum sample size for climate statistics (WMO-No. 1203). Arbitrarily selecting short-term periods may result in average values ​​containing anomalous signals of current climate change, and the typical data generated from short-term climate data is insufficient to describe the general patterns of long-term typical meteorological characteristics in a given area.

[0007] In addition, existing traditional methods often use random selection or monthly averaging to select typical meteorological days, and often use cumulative distribution functions to calculate the probability distribution similarity statistics between typical days and long-term data. However, the selection of typical days has problems such as high statistical probability but weak climatic significance, ignoring key climate turning points, and unevenness of meteorological data time series.

[0008] This invention provides a method for typicalizing accumulated (≥30 years) meteorological data to generate a simplified meteorological element data sequence describing the long-term typical meteorological characteristics of a certain location. This dataset consists of meteorological element data from 24 typical solar terms after thinning, with a certain degree of continuity between the data groups. This dataset can provide representative meteorological conditions with the highest probability of occurrence for studies on a long-term scale, avoiding the shortcomings caused by the randomness or discretization of the research period. Summary of the Invention

[0009] To address the aforementioned problems in existing technologies, this invention provides a method for typifying and processing long-term meteorological data. This method aims to solve the technical problems of existing technologies that use random selection or monthly averaging to select typical meteorological days, often employing cumulative distribution functions to calculate the probability distribution similarity statistics between typical days and long-term data. These methods suffer from high statistical probability but weak climatic significance, neglect of key climatic turning points, and unevenness in the time series of meteorological data.

[0010] To achieve the above objectives, the present invention provides the following technical solution, comprising: Step 1: acquiring basic observational data of hourly meteorological elements for at least thirty years; Step 2: determining the dates of typical meteorological days each year, the dates being based on the traditional Chinese twenty-four solar terms and the modern "fixed-solar-term method"; Step 3: performing typical processing on the meteorological element data of the typical meteorological days each year to construct a meteorological element data sequence for typical meteorological days of the solar terms; Step 4: processing a hypothetical meteorological year sequence for typical meteorological days of the solar terms based on the meteorological element data sequence for typical meteorological days of the solar terms.

[0011] Furthermore, the cumulative hourly meteorological element observation data over the past thirty years is based on the World Meteorological Organization (WMO) standard climate period, with priority given to the period from 1991 to 2020 as the basic research period.

[0012] Furthermore, the determination of typical meteorological days for each year includes using the twenty-four solar terms as time nodes. The twenty-four solar terms include Minor Cold, Major Cold, Beginning of Spring, Rain Water, Awakening of Insects, Spring Equinox, Pure Brightness, Grain Rain, Beginning of Summer, Grain Buds, Grain in Ear, Summer Solstice, Minor Heat, Major Heat, Beginning of Autumn, End of Heat, White Dew, Autumn Equinox, Cold Dew, Frost's Descent, Beginning of Winter, Minor Snow, Major Snow, and Winter Solstice.

[0013] Furthermore, the typification process includes using a cumulative distribution function distance metric to determine typical meteorological days for solar terms.

[0014] Furthermore, the cumulative distribution function distance metric includes: Step 1: Calculate the 30-year average and standard deviation of each meteorological parameter for all solar terms; Step 2: Standardize the meteorological parameter data and calculate the average value of each standardized meteorological parameter; Step 3: Perform a weighted summation on the standardized meteorological parameter data to obtain the weighted summation value DS; Step 4: Select the solar term with the smallest DS value as the representative solar term.

[0015] Furthermore, the typification process includes using a probability density function and a kernel density estimate to measure the density value.

[0016] Furthermore, the meteorological element data sequence for the typical meteorological days of the solar term includes meteorological parameters selected from the maximum dry-bulb temperature, minimum dry-bulb temperature, average dry-bulb temperature, minimum relative humidity, average relative humidity, maximum wind speed, average wind speed, and daily cumulative radiation value of the horizontal surface.

[0017] Furthermore, the weights of the daily cumulative radiation values ​​on the horizontal surface are set to 10 and 20, and the weights of other meteorological parameters are allocated proportionally.

[0018] Furthermore, the processing of the typical meteorological days of the solar terms and the hypothetical meteorological year sequence includes connecting the typical meteorological days of the twenty-four solar terms one after the other, and smoothing the connection through curve fitting.

[0019] Furthermore, the hypothetical meteorological year sequence uses the solar term Minor Cold as the first solar term day of the statistical year.

[0020] Compared with the prior art, the beneficial effects of the present invention are:

[0021] 1. In this invention, the twenty-four solar terms (such as Lesser Cold and Beginning of Spring) are used as the time nodes for typical meteorological days. Based on the 15-degree variation in the sun's ecliptic longitude, the solar terms directly reflect the phased changes in the Earth's revolution and solar radiation energy. For example, the sun is directly overhead at the equator at the vernal equinox, and the sun's altitude angle is lowest at the winter solstice. The 24 solar terms are equidistantly distributed throughout the year, giving the generated typical meteorological day sequence (STTMD) natural temporal continuity. This accurately describes the periodic changes in solar radiation (such as the longest daylight hours at the summer solstice and the shortest at the winter solstice), ensuring temporal continuity and periodicity. The solar terms are critical points of climate phase transition (such as the Awakening of Insects marking a sudden rise in temperature and Frost's Descent indicating a shift in low temperatures), covering periods of high gradient changes in meteorological parameters, avoiding the omission of key turning points, thus strengthening the climatological and physical significance. At the same time, the phenological characteristics of the solar terms are preserved (such as the arrival of wheat harvest at Lesser Fullness and the birth of mantises at Grain in Ear), supporting agricultural planning, ecological research, and historical and cultural analysis, increasing cross-disciplinary application value.

[0022] 2. Existing technologies often arbitrarily select short-term periods (such as data from the past 3 years), and the average values ​​contain abnormal signals of global warming; moreover, they do not meet the minimum sample size of WMO (30 years), resulting in typical data failing to reflect general climate patterns (such as missing gradual inflection points). In this invention, the meteorological data is based on cumulative hourly observation data of at least 30 years, and the WMO standard climate period (such as 1991 to 2020) is preferred to be used to cover key climate turning points such as the period of accelerated global warming. The 30-year data can eliminate short-term weather anomalies (such as El Niño events) and generate climate characteristic sequences representing the "maximum probability of occurrence", ensuring long-term typicality; the use of the WMO official baseline period (1991-2020) allows the results to be directly compared with international reports (such as IPCC assessments), supporting global warming research and increasing international comparability; and it minimizes climate noise interference (such as cold waves in random years), ensuring the stability of the output results and improving data robustness.

[0023] 3. Existing technologies rely on a single cumulative distribution function, neglecting the joint distribution of multidimensional parameters, such as the coupling relationship between temperature, humidity, and radiation; typical days are selected based on "high statistical probability but weak climatic significance," such as ignoring the phenological response of solar terms by using monthly averages. In this invention, based on the idea of ​​probability distribution matching, typical meteorological day sequences are constructed using metrics such as the cumulative distribution function CDF distance measure, probability density function PDF, and KDE density value measure. By comparing joint probability distributions, the main areas of meteorological parameter combinations throughout the year are covered, such as the "Great Heat" pattern of high temperature and high humidity; high weighting highlights the dominance of radiation parameters, which is in line with the essence of solar energy scales for solar terms; and it is compatible with multiple statistical quantities such as CDF, PDF, and KDE, adapting to the needs of different climate zones. Attached Figure Description

[0024] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0025] Figure 1 This is a phenological characteristic diagram of the 24 solar terms in this invention;

[0026] Figure 2 This is a graph showing the meteorological element parameters and weights in this invention;

[0027] Figure 3 This is a flowchart of the present invention;

[0028] Figure 4 This is a flowchart of the algorithm for generating typical meteorological days for various solar terms in this invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] Example 1:

[0031] Please see Figures 1-4 This embodiment provides the following technical solution: Step 1: Obtain basic observation data of meteorological elements over a minimum of thirty years; Step 2: Determine the dates of typical meteorological days each year, based on the traditional Chinese twenty-four solar terms; Step 3: Perform typical processing on the meteorological element data of typical meteorological days each year to construct a meteorological element data sequence of typical meteorological days of solar terms; Step 4: Process the hypothetical meteorological year sequence of typical meteorological days of solar terms based on the meteorological element data sequence of typical meteorological days of solar terms.

[0032] Specifically, by acquiring 30 years of data, determining the dates of solar terms, typifying and processing data, and generating hypothetical year sequences, the generated typical day sequences (STTMD) based on the 24 solar terms naturally possess temporal correlation, solving the data discontinuity problem in existing technologies and ensuring temporal continuity. The minimum 30-year data base covers climate change inflection points (such as the period of accelerated global warming), and the output results represent the climate characteristics with the highest probability of occurrence in a certain location, strengthening long-term typicality. The final generated 24-day hypothetical year sequence (accounting for only 6.5% of the total annual data) significantly simplifies simulation calculations and is suitable for long-term studies such as building energy consumption and agricultural planning, improving application efficiency.

[0033] The basic data of hourly meteorological element observations over the past thirty years are based on the World Meteorological Organization (WMO) standard climate period, with priority given to the period from 1991 to 2020 as the basic research period.

[0034] Specifically, the WMO standard climate period (1991 to 2020) is adopted first. By binding it to the WMO official baseline period, the results can be directly compared with international reports such as the IPCC, supporting the formulation of global climate change policies. The period from 1991 to 2020 covers key turning points in climate warming, avoids short-term data noise interference (such as El Niño events), and ensures the stability of the sequence.

[0035] Determining the dates of typical meteorological days each year involves using the 24 solar terms as time nodes. These 24 solar terms include: Minor Cold, Major Cold, Beginning of Spring, Rain Water, Awakening of Insects, Spring Equinox, Pure Brightness, Grain Rain, Beginning of Summer, Grain Buds, Grain in Ear, Summer Solstice, Minor Heat, Major Heat, Beginning of Autumn, End of Heat, White Dew, Autumn Equinox, Cold Dew, Frost's Descent, Beginning of Winter, Minor Snow, Major Snow, and Winter Solstice. Typicality processing involves using a cumulative distribution function distance metric to determine the typical meteorological days for each solar term.

[0036] Specifically, by using the 24 solar terms as time nodes and employing CDF distance measurement, the phased changes in solar radiation are naturally captured through solar term nodes (such as the summer solstice marking the longest daylight), avoiding random selection that ignores phenological responses (such as the warming transition at the time of the Awakening of Insects), thus making the climatological and physical significance explicit; the CDF distance measurement formula is as follows:

[0037]

[0038] This makes the joint distribution of typical daily meteorological parameters most closely resemble the overall characteristics over 30 years.

[0039] The cumulative distribution function distance metric includes: Step 1: Calculate the 30-year average and standard deviation of meteorological parameters for all solar terms; Step 2: Standardize the meteorological parameter data and calculate the average of the standardized parameters; Step 3: Perform a weighted summation on the standardized meteorological parameter data to obtain the weighted sum value DS; Step 4: Select the solar term with the smallest DS value as the representative solar term. Specifically, the selection is based on minimizing the DS value, using the following formula:

[0040]

[0041] By assigning the highest weight (10, 20) to the daily cumulative radiation value of the horizontal surface, the solar radiation characteristics of typical days are consistent with the solar term energy scale (such as the high heat radiation during Lesser Heat), thus strengthening the core driving force of radiation; by integrating the weighted calculation of 8 parameters such as temperature, humidity, and wind speed, the system covers the main areas of meteorological combinations throughout the year (such as the "high temperature, high humidity and heavy rain" pattern during Greater Heat), which facilitates multi-parameter coordinated control.

[0042] Typical processing includes using probability density functions and kernel density estimation to measure density values.

[0043] The meteorological element data sequence for typical meteorological days of the solar term is constructed, including meteorological parameters selected from the maximum and minimum dry-bulb temperatures, average dry-bulb temperatures, minimum and average relative humidity, maximum and average wind speeds, and daily cumulative radiation on the horizontal surface. The weights of the daily cumulative radiation on the horizontal surface are set to 10 and 20, and the weights of other meteorological parameters are allocated proportionally.

[0044] Specifically, specific parameter combinations (Tmax, Tmin, RHmin, etc.) and radiation weights of 10 and 20 are used through a radiation-dominated parameter system (such as... Figure 2 Weighting allows STTMD data to synchronously reflect the phenological characteristics of solar terms (such as the high temperature and humidity requirements during the "wheat ripening period" of Grain in Ear); parameters such as dry bulb temperature and humidity directly match the input requirements of building energy consumption simulation, avoiding secondary data processing and increasing engineering adaptability.

[0045] The hypothetical meteorological year sequence for processing typical meteorological days of the twenty-four solar terms involves connecting the first and last typical meteorological days of each solar term and smoothing the connection through curve fitting. The hypothetical meteorological year sequence uses the solar term Minor Cold as the first solar term day of the statistical year.

[0046] Specifically, the 24 solar terms are connected end to end, with Minor Cold as the first solar term day, and the curve fitting is smooth. Through the hypothetical annual sequence starting from Minor Cold, the annual cycle path of solar radiation is fully covered (from the lowest point at the winter solstice to the highest point at the summer solstice). By using curve fitting, abrupt changes between adjacent solar terms (such as temperature jumps from the Beginning of Spring to Rain Water) are eliminated, meeting the continuity requirements of outdoor thermal environment simulation.

[0047] The working principle of this invention is as follows: First, based on the World Meteorological Organization (WMO) standards, annual hourly meteorological observation data for ≥30 years (e.g., 1991-2020) is acquired to cover the inflection points of gradual climate change and eliminate short-term anomaly interference. Then, the traditional Chinese 24 solar terms (such as Lesser Cold and Beginning of Spring) are used as time nodes, as the essence of the solar terms is a scale of radiation energy change every 15° of solar ecliptic longitude (e.g., the summer solstice corresponds to the highest solar altitude angle, and the winter solstice to the lowest), directly linking them to meteorological parameter drivers. Next, the meteorological elements of each solar term day (including eight parameters such as dry-bulb temperature Tavg, Tmax, Tmin, relative humidity RHmin, RHavg, wind speed Vavg, Vmax, and daily cumulative radiation SW on the horizontal surface) are typicalized. The core method uses the cumulative distribution function (CDF) distance metric: the parameters are first standardized, and the formula is: The weighted summation is then used to calculate the DS value, using the formula: DS = ∑ i K i ·|η i，s，y |, see weighting Figure 2Among them, SW has the highest weights of 10 and 20. The solar term day with the smallest DS value is selected as the typical meteorological day (STTMD) to ensure that the joint distribution of its meteorological parameters is closest to the overall characteristics of 30 years. Finally, the 24 STTMDs are connected end to end according to the solar term sequence (starting from Xiaohan). The abrupt changes in adjacent day data are smoothed by curve fitting to form a continuous hypothetical meteorological year sequence.

[0048] This process is compatible with probability distribution matching methods such as probability density function (PDF) and kernel density estimation (KDE) (documentation). Figure 2 This process enhances the phenological response of solar terms by weighting parameters dominated by solar radiation (e.g., 50% weighting of SW). For example, high temperature and humidity during the Grain in Ear solar term promote wheat ripening, and high heat radiation during the Great Heat solar term leads to torrential rain. This results in the generated STTMD sequences possessing temporal continuity (equidistant distribution of solar term nodes), climatic typicality (30-year statistical steady state), and high application efficiency (data volume is only 6.5% of the annual total). It systematically addresses the shortcomings of existing technologies, such as the lack of date correlation for characteristic meteorological days (e.g., FS statistical method), omission of inflection points in short-term data (e.g., ignoring gradual changes in 3-year studies), and the weakening of climatic significance by statistical methods (ignoring radiation-driven factors by monthly averages). [Document] Figures 1-2 These are the method flowchart, CDF calculation steps, and multi-algorithm parallel logic, which should be handled according to the guidelines when they need to be embedded.

[0049] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for typicalizing and processing accumulated meteorological data, characterized in that: include: Step 1: Obtain basic observational data of meteorological elements for at least thirty years of cumulative annual observations; Step 2: Determine the dates of the 24 solar terms for each year in all basic data. These dates are based on the traditional Chinese 24 solar terms and have been revised using the modern "solar term determination method". Step 3: The meteorological element data of the solar terms for each year are processed to typicality, typical solar terms are selected, and a sequence of meteorological element data for typical solar terms is constructed. Step 4: Process the meteorological element data sequence of typical meteorological days of the solar term into a hypothetical meteorological year sequence.

2. The method for typicalizing accumulated meteorological data according to claim 1, characterized in that: The cumulative hourly meteorological data for the past thirty years is based on the World Meteorological Organization (WMO) standard climate period, with priority given to the period from 1991 to 2020 as the basic research period.

3. The method for typicalizing accumulated meteorological data according to claim 1, characterized in that: The determination of typical meteorological days each year includes using the 24 solar terms as time nodes. The 24 solar terms include Lesser Cold, Greater Cold, Beginning of Spring, Rain Water, Awakening of Insects, Spring Equinox, Pure Brightness, Grain Rain, Beginning of Summer, Lesser Fullness, Grain in Ear, Summer Solstice, Lesser Heat, Greater Heat, Beginning of Autumn, End of Heat, White Dew, Autumn Equinox, Cold Dew, Frost's Descent, Beginning of Winter, Lesser Snow, Greater Snow, and Winter Solstice. Unlike the traditional understanding of solar terms, the order of the solar terms is readjusted according to the Gregorian meteorological year, with Lesser Cold as the first solar term and Winter Solstice as the 24th solar term.

4. The method for typicalizing accumulated meteorological data according to claim 1, characterized in that: The typicalization process includes using a cumulative distribution function distance metric to determine typical meteorological days for solar terms.

5. The method for typicalizing accumulated meteorological data according to claim 4, characterized in that: The cumulative distribution function distance metric includes: Step 1: Calculate the average and standard deviation of meteorological parameters for all solar terms over 30 years; Step 2: Standardize the meteorological parameter data and calculate the average value of each meteorological parameter after standardization; Step 3: Perform a weighted summation on the standardized meteorological parameter data to obtain the weighted sum value DS; Step 4: Select the solar term day with the smallest DS value as the representative solar term day.

6. The method for typicalizing accumulated meteorological data according to claim 1, characterized in that: The typicalization process may also include using probability density functions and kernel density estimation to measure density values, which can yield similar statistical results.

7. The method for typicalizing accumulated meteorological data according to claim 1, characterized in that: The constructed typical meteorological element data sequence for the solar term includes meteorological parameters selected from the maximum dry-bulb temperature, minimum dry-bulb temperature, average dry-bulb temperature, minimum relative humidity, average relative humidity, maximum wind speed, average wind speed, and daily cumulative radiation value on the horizontal surface, which can be adjusted according to different research objectives and objects.

8. The method for typicalizing accumulated meteorological data according to claim 7, characterized in that: The weights of the daily cumulative radiation values ​​on the horizontal surface are set to 10 and 20, and the weights of other meteorological parameters are allocated proportionally. These weights can be adjusted according to the different targets and objects.

9. The method for typicalizing accumulated meteorological data according to claim 1, characterized in that: The processed typical meteorological day meteorological hypothetical year sequence of the solar term days includes connecting the first and last typical meteorological days of the twenty-four solar term days and smoothing the connection through curve fitting.

10. The method for typicalizing accumulated meteorological data according to claim 1, characterized in that: The hypothetical meteorological year sequence uses the solar term Minor Cold as the first solar term day of the statistical year; Methods based on the idea of ​​probability distribution matching, including but not limited to CDF, PDF, KDE, CTYW, etc., are used for the optimal selection of solar terms.

Images