El Nino event identification and La Nina conversion early warning method, equipment and medium

Abstract

The invention relates to the field of atmospheric science, and discloses an El Nino event identification and Lanina conversion early warning method and device and a medium. According to the method, sea surface temperature and near-ground 2-meter air temperature data in a fifth-generation atmospheric reanalysis product ERA5 of the European mid-term weather forecast center are utilized; calculating a time-lag cross correlation coefficient between the sea temperature grid point and the air temperature grid point, and constructing a directed network between the global air temperature and the sea temperature of the equatorial mid-east Pacific Ocean and the equatorial mid-Pacific Ocean; the importance of temperature nodes is quantified by using average link strength, key areas with advanced response to El Nino are identified in combination with spatial and temporal changes of the link number and the link strength, and early warning is performed on the transformation from El Nino events to La Nina events. According to the method, the influence of the El Nino event on the global climate system can be deeply understood, and the prediction accuracy of the Lanina event is improved, so that a scientific basis is provided for disaster prevention and coping strategy formulation.

Description

Technical Field

[0001] The present invention relates to the field of atmospheric science, and in particular to a method, device and medium for early warning of El Nino event identification and transition to La Nina. Background Art

[0002] The El Niño-Southern Oscillation (ENSO) is one of the core manifestations of the interannual oscillation in the tropical Pacific Ocean and one of the most significant ocean-atmosphere coupled climate phenomena in the world, which has a profound impact on the global climate system. Among them, El Niño events can be divided into eastern and central types. There are significant differences between the two in the distribution and evolution of sea temperature anomalies, and their impacts on the global climate are also unique. However, no matter what type of El Niño event it is, it has a significant impact on global temperature, precipitation patterns, fishery resources, food security and economic development. Therefore, in-depth revelation of the response mechanism of the atmosphere to different types of El Niño events is of great significance for improving climate prediction capabilities and formulating disaster prevention strategies. Summary of the invention

[0003] The purpose of the present invention is to propose a method, device and medium for identifying El Niño events and warning of their transition to La Niña, so as to deeply reveal the response mechanism of the atmosphere to different types of El Niño events, improve climate forecasting capabilities, and formulate corresponding disaster prevention strategies.

[0004] Specifically, the present invention provides a method for identifying El Nino events and early warning of their transition to La Nina, comprising the following steps:

[0005] S1. Data collection: Collect sea surface temperature and 2-meter air temperature data from the fifth-generation atmospheric reanalysis product ERA5 of the European Centre for Medium-Range Weather Forecasts;

[0006] S2. Data preprocessing: Using the data collected in step S1, calculate the daily average sea surface temperature and air temperature, define 1979-2019 as the reference state, subtract the obtained daily average sea surface temperature and air temperature from the reference state temperature, and obtain the anomaly time series;

[0007] S3. Construction of climate network based on mutual correlation coefficient: Combined with the anomaly time series obtained in step S2, the temperature grid points and the sea temperature grid points are divided into two sets. For each pair of nodes i and j in different sets, the lagged mutual correlation coefficient between them is calculated; the peak value of the absolute value of the mutual correlation coefficient and the corresponding time lag are used to represent the strength and direction of the link, and a directed network of air temperature and sea temperature is constructed;

[0008] S4. Quantification of network link characteristics: Set appropriate mutual correlation coefficient thresholds to remove accidental links that are insensitive to El Niño events; focus on links from temperature to sea temperature, and calculate the number of links and average link strength at temperature grid points;

[0009] S5. Quantification of atmospheric response to the eastern and central El Niño: combining the methods described in step S3 and step S4, calculating the frequency distribution of the link strength of the global temperature-equatorial central and eastern Pacific sea temperature network, the spatiotemporal distribution of the number of links and the link strength, identifying key areas with advanced responses to El Niño events, and identifying early warning signals of the transition of El Niño events to La Niña events based on the link strength of key areas;

[0010] S6. Quantification of atmospheric response to central-regional El Niño type I and type II: Combined with the methods described in step S3 and step S4, the frequency distribution of the link strength of the global temperature-equatorial central Pacific sea temperature network, the spatiotemporal distribution of the number of links and the link strength are calculated to identify key areas that have an advanced response to the two types of CP El Niño events, and based on the link strength of key areas, the early warning signals of two consecutive La Niña events after the El Niño event are identified.

[0011] A storage medium stores instructions and data for implementing an El Nino event identification and an early warning method for its transition to La Nina.

[0012] A device for identifying El Nino events and warning of their transition to La Nina comprises: a processor and the storage medium; the processor loads and executes instructions and data in the storage medium to implement a method for identifying El Nino events and warning of their transition to La Nina.

[0013] The beneficial effects provided by the present invention are:

[0014] (1) This paper reveals the interaction between the atmosphere and El Niño events and their key areas through complex network analysis, which not only deepens the understanding of the impact of El Niño events on the global climate system, but also provides scientific support for governments and international organizations to formulate more efficient climate change adaptation plans.

[0015] (2) The present invention proposes a method for warning of a possible La Niña event during an El Niño event in the current year or the next year, which helps to improve the accuracy of La Niña event prediction and provide a scientific basis for formulating disaster prevention and response strategies, thereby effectively reducing the losses caused by climate anomalies.

[0016] (3) By constructing an air temperature and sea temperature network based on different sea temperature regions, the present invention deeply explores the atmospheric advance response to different types of El Niño events, provides a new perspective for climate science research, and promotes further research and understanding of El Niño events and climate change. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic flow chart of the method of the present invention;

[0018] Figure 2 The frequency distribution diagram of link strength in normal years and the difference diagram of link strength frequency distribution between five El Niño years and normal years;

[0019] Figure 3 The spatial distribution of the grid-point link strength of temperature in normal years and five El Niño years;

[0020] Figure 4 The spatial distribution of the number of grid point links for temperature in normal years and five El Niño years;

[0021] Figure 5 It is a three-point smoothed graph of the sum of link strengths in different ranges of the El Niño key area over time and a schematic diagram of the duration of EP and CP El Niño events;

[0022] Figure 6 It is a three-point smoothed graph of the average link strength changing with time and a schematic diagram of the duration of EP, CP-I, CP-II El Niño and La Niña events;

[0023] Figure 7 The spatial distribution map of the temperature grid connection strength in normal years and CP-I and CP-II El Niño years;

[0024] Figure 8 The spatial distribution map of the temperature grid link number in normal years and CP-I and CP-II El Niño years;

[0025] Fig. 9 It is a three-point smoothed graph of the variation of the sum of link strengths over time in different ranges of the key area of ​​CP-II El Niño and a schematic diagram of the duration of EP, CP-I, CP-II El Niño and La Niña events;

[0026] Fig.10 It is a working schematic diagram of the hardware device of the present invention. DETAILED DESCRIPTION

[0027] To make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0028] Before formally describing the present invention, the scheme of the present invention is first generally described for easy understanding.

[0029] Please refer to Figure 1 The present invention provides a method for identifying El Nino events and early warning of their transition to La Nina, comprising the following steps:

[0030] S1. Data collection: Collect sea surface temperature and 2-meter air temperature data from the fifth-generation atmospheric reanalysis product ERA5 of the European Centre for Medium-Range Weather Forecasts;

[0031] In step S1, the collected ERA5 dataset covers the period from January 1979 to December 2020, and the spatial resolution of the data is selected as 5° × 5°.

[0032] S2, data preprocessing: using the data collected in step S1, calculate the daily average sea surface temperature and air temperature, define 1979-2019 as the reference state, subtract the obtained daily average sea surface temperature and air temperature from the reference state temperature, and obtain the anomaly time series;

[0033] In step S2, the formula for calculating the anomaly time series of sea surface temperature and air temperature is as follows:

[0034]

[0035] in, is the temperature of the grid point on the i-th day in the y-th year (excluding February 29th of the leap year to ensure that each grid year has 365 days), where y=1 represents 1979, and so on. It means adding up the temperatures of the i-th day in all the years of the reference state and averaging them.

[0036] S3, climate network construction based on mutual correlation coefficients; combining the anomaly time series obtained in step S2, the temperature grid points and the sea temperature grid points are divided into two sets, and for each pair of nodes i and j in different sets, the lagged mutual correlation coefficient between them is calculated; the peak value of the absolute value of the mutual correlation coefficient and the corresponding time lag are used to represent the strength and direction of the link, and a directed network of air temperature and sea temperature is constructed;

[0037] Step S3 is as follows:

[0038] The anomaly time series obtained in step S2 is divided into two sets, one containing all global temperature grid points (a total of 2664), and the other containing all sea temperature grid points in the selected ocean area (a total of 55); for each pair of nodes i and j in different sets, the mutual correlation coefficient between them is calculated, and the formula is as follows:

[0039]

[0040] in is the time series T i (d) The standard deviation, τ is the time interval, τ∈[0,τ max ],τ max = 200 days, y represents the starting time of the time series when each window interval is 0, and c i,j (τ) = C j,i (-τ). The time lag corresponding to the maximum absolute value of the cross-correlation coefficient is expressed as The sign represents the direction of the link between nodes i and j. When , the link is from i to j. The absolute value of the mutual correlation coefficient represents the link strength. Thus, a directed climate network of air temperature and sea temperature is established based on the mutual correlation coefficient.

[0041] S4. Quantify the characteristics of network links; set appropriate mutual correlation coefficient thresholds to remove accidental links that are insensitive to El Niño events; focus on links from temperature to sea temperature, and calculate the number of links and average link strength at temperature grid points;

[0042] Step S4 is specifically as follows:

[0043] An appropriate threshold Δ is introduced to exclude links with weaker connections to the El Niño event. The adjacency matrix defined by the threshold is as follows:

[0044]

[0045] Where H(X) is the Heaviside step function, defined as H(X≥0)=1 and H(x<0)=0.

[0046] Consider the link from air temperature to sea temperature, that is The degree formula of the grid point is as follows:

[0047]

[0048] That is, the sum of all the correlation coefficients of a grid point that exceed the threshold. The average link strength of the grid point is defined as follows:

[0049]

[0050] That is, the ratio of the degree of the grid point to the number of links of the grid point. is the number of links from the temperature grid to the sea temperature grid, and its formula is as follows:

[0051]

[0052] S5, quantifying the atmospheric response of the Eastern Type (EP) and the Central Type (CP) El Niño; combining the methods described in step S3 and step S4, calculating the frequency distribution of the link strength of the global temperature-equatorial central and eastern Pacific Ocean temperature network, the spatiotemporal distribution of the number of links and the link strength, identifying the key areas with advanced response to the El Niño event, and identifying the early warning signal of the transition of the El Niño event to the La Niña event based on the link strength of the key areas;

[0053] Step S5 is specifically as follows:

[0054] Step S51: Select the sea temperature grid points in the equatorial central and eastern Pacific region (170°W-120°W, 10°S-10°N) to establish a network; define the occurrence and extinction of El Niño events according to the Nino3.4 index. If the Nino3.4 index is >0.5℃ and lasts for at least 5 months, it is an El Niño event; if the Nino3.4 index is <-0.5℃ and lasts for at least 5 months, it is a La Niña event. The time and value when the absolute value of the Nino3.4 index reaches the maximum value are defined as the peak time and peak intensity of the event, respectively. When the peak intensity of the event is >0.5℃ and <1.3℃, the event is defined as a weak intensity event; ≥1.3℃ and <2℃ is a moderate intensity event; ≥2℃ and <2.5℃ is a strong intensity event; ≥2.5℃ is an extremely strong intensity event;

[0055] Step S52: Considering the different sea temperature anomaly areas and their impacts between different types of El Niño events, the NEP index and NCP index are used to distinguish between EP type and CP type El Niño events. The calculation formulas of these two indices are as follows:

[0056] NEP=N3-αN4

[0057] NCP=N4-αN3

[0058] Where N3 is the regional average of the sea surface temperature anomalies in the Nino3 region (150°W-90°W, 5°S-5°N), and N4 is the regional average of the sea surface temperature anomalies in the Nino4 region (160°E-150°W, 5°S-5°N). When N3*N4 is greater than 0, α is equal to 0.4, otherwise α is equal to 0. The distinction rules are as follows: (1) If the absolute value of the NEP index reaches or exceeds 0.5℃ and lasts for at least three months, it is an EP type event; (2) If the absolute value of the NCP index reaches or exceeds 0.5℃ and lasts for at least three months, it is a CP type event; (3) If an event meets both of the above points, the type of event peak shall prevail;

[0059] Step S53: Select five El Nino events of medium or above intensity for analysis based on the Nino3.4 index. Analyze the changes in the number and intensity of links during the El Nino event through the frequency distribution histogram of link intensity, the graph of link number over time, and the probability density distribution graph of link time interval; identify the key areas with advanced response to El Nino events through the spatiotemporal distribution graph of link intensity and link number, and determine the early warning signal of the transition from El Nino event to La Nina event through the time evolution analysis of link intensity in the key areas.

[0060] S6. Quantify the atmospheric response to central El Niño type I and type II; combine the methods described in step S3 and step S4 to calculate the frequency distribution of the link strength of the global temperature-equatorial central Pacific sea temperature network, the spatiotemporal distribution of the number of links and the link strength, identify the key areas that have an advanced response to the two types of CP El Niño events, and identify the early warning signals of two consecutive La Niña events after the El Niño event based on the link strength of the key areas.

[0061] Step S6 is specifically as follows:

[0062] Step S61: Select the sea temperature grid points in the equatorial central Pacific region (160°E-150°W, 10°S-10°N) to establish a network; use the EMI index to distinguish between EP type and CP type El Niño events, and the calculation formula is as follows:

[0063] EMI=[SSTA] C -0.5×[SSTA] W -0.5×[SSTA] E

[0064] [SSTA] C , [SSTA] W , [SSTA] E They represent the average SST anomalies in the C region (165°E-140°W, 10°S-10°N), the W region (125°E-145°E, 10°S-20°N), and the E region (110°W-70°W, 15°S-5°N). The MII index is used to further distinguish between the Central Part El Niño Type I (CP-I) and Central Part El Niño Type II (CP-II) events. When the MII index is greater than 1 standard deviation, it is determined to be a CP-II El Niño event, otherwise it is a CP-I event;

[0065] Step S62: Select eight CP-type El Niño events for analysis based on EMI; analyze the characteristics of CP-type El Niño links and their differences from EP-type El Niño through the link strength frequency distribution diagram and the link time interval probability density distribution diagram; identify the key areas of CP-I and CP-II El Niño through the spatiotemporal distribution diagram of link strength and link number; finally, further explore the signal of two consecutive La Niña events after the El Niño event through the time evolution analysis of link strength in the key areas.

[0066] As an implementation case, the present invention is further described. The implementation case is used to illustrate the present invention, but this case does not limit the scope of application of the present invention. It is also applicable to different regions and other time periods.

[0067] The implementation flow chart of the method of the present invention for studying the influence of atmosphere on El Nino events based on complex network method is as follows: Figure 1 As shown, the following is explained in conjunction with the accompanying drawings:

[0068] (1) Data collection

[0069] This implementation case uses the hourly sea surface temperature and hourly 2-meter atmospheric temperature data from the fifth generation atmospheric reanalysis product (ERA5) of the European Centre for Medium-Range Weather Forecasts (ECMWF). The time range of the dataset is from January 1979 to December 2020, with a spatial resolution of 5°×5°.

[0070] (2) Data preprocessing

[0071] Based on the data described in step (1), the daily average sea surface temperature and atmospheric temperature are calculated. 1979-2019 is defined as the reference state. The multi-year average reference state daily temperature is calculated. The daily average temperature of the reference state is subtracted from the daily average temperature to obtain the anomaly time series. This case selects two ocean regions for research: one is the equatorial central and eastern Pacific region (170°W-120°W, 10°S-10°N), and the other is the equatorial central Pacific region (160°E-150°W, 10°S-10°N). The former mainly studies the characteristics of general El Niño events, while the latter mainly studies the characteristics of the two types of CP-type El Niño events.

[0072] (3) Climate network construction based on mutual correlation coefficients

[0073] The anomaly time series are divided into two sets, one containing all global temperature grid points (a total of 2664), and the other containing all sea temperature grid points in the selected ocean area (a total of 55); the time window is set to 365 days, and the time lag is set to 0-200 days. For each pair of nodes i and j in different sets, their mutual correlation coefficients in each time window are calculated; the absolute value peak of the mutual correlation coefficient represents the link strength, and the corresponding time lag represents the direction of the link between nodes i and j.

[0074] (4) Quantification of network link characteristics

[0075] Next, we introduce a threshold to exclude links that are insensitive to El Niño events: By analyzing the frequency distribution of link strength and the time evolution of the number of links under different thresholds, we find that the number of links with an absolute value of link strength less than 0.2 is very small, and its change in El Niño years is not obvious (e.g. Figure 2 ); In addition, we not only focus on strong links, but also on the changes in weak links. When the threshold is 0.2, the fluctuation of the number of links is large, so the threshold is 0.2 in this case. This case focuses on the link from air temperature to sea temperature, that is, The degree of the temperature grid point is defined as the sum of all the correlation coefficients of the grid point that exceed the threshold in a time window. The number of links of each grid point in the window is calculated, and the ratio of the degree of the grid point in the window to its number of links is calculated, that is, the average link strength.

[0076] (5) Quantification of atmospheric response to Eastern and Central El Niño

[0077] The sea surface temperature grid points in the equatorial central and eastern Pacific (170°W-120°W, 10°S-10°N) were selected to establish a network with the global temperature grid points to analyze the atmospheric response to the eastern type (EP) and central type (CP) El Niño events.

[0078] In this implementation case, five moderate and above El Niño events from 1979 to 2020 were selected for specific analysis based on the Nino3.4 index (1982, 1987, 1997, 2009, and 2015), and EP and CP type events were determined based on the NEP and NCP indices, as shown in Table 1:

[0079] Table 1 Five El Niño events of moderate or above intensity

[0080]

[0081] like Figure 2 In general, in normal years, the link strength is weak (in the range of [-0.6, 0.6]). In the five El Niño years, the number of weak links is significantly reduced compared with normal years, while the number of strong links (the absolute value of link strength is greater than 0.5) increases, and the reduction in weak links is much greater than the increase in strong links. This may be because as the El Niño event develops, the direction of most links changes from sea temperature to air temperature. Analyzing the frequency changes of link strength in different El Niño years, it is found that the EP El Niño event ( Figure 2 b, c, d) occurs, the positive link decreases more than the negative link; while the CP type El Niño event ( Figure 2 e) The number of positive and negative links decreased during the period was basically the same. In addition, the three super El Niño events ( Figure 2 b, d, f) The increase in negative links is significantly greater, while the increase in the number of positive and negative links in the two moderate El Niño events is basically the same, and is much smaller than the increase in the number of negative links in the super El Niño event.

[0082] like Figure 3 As shown in Figure 2, the link strength in some areas is significantly higher and more concentrated in El Niño years. Specifically, there are high-value areas of negative link strength in the tropical Atlantic and northern and eastern Australia, while there are high-value areas of positive link strength in the equatorial central and southeastern Pacific. Figure 3 b, d, and f), there is a strong negative link between the tropical Atlantic and Indonesia, while only one of the two moderate El Niño years has a strong link. Figure 4 In normal years, atmospheric temperature is generally linked to the sea temperature in the equatorial central and eastern Pacific, but most temperature grid points have few links. In El Niño years, some atmospheric regions are linked to most or even all (55) grid points of the ocean, while other regions are completely disconnected from the ocean. Figure 3 It was found that the areas with high link numbers basically overlap with the areas with high link strength, indicating that in El Niño years, more high-strength links will be established at the atmospheric grid points in these areas, and the advanced response of the atmosphere to the ocean is highly localized. Further analysis of the changes in the average link strength and number of links in these areas identified the key areas that have a greater impact on the sea temperature in the equatorial central and eastern Pacific, namely the equatorial eastern Pacific, the tropical Atlantic, and the northern Australia.

[0083] from Figure 5 It can be seen that in the early and middle stages of the El Niño event, a large number of weak links are disconnected, and in the later stages, weak links are reestablished. The total strength of weak links in CP-type El Niño events reaches an extremely low value earlier than that in EP-type El Niño events. Figure 5 It can be found that during the El Niño event, the total strength of strong links generally showed an upward trend. The change amplitude of the CP-type El Niño event was smaller than that of the EP-type El Niño event, and the change trend was inconsistent. This may indicate that the above-mentioned area is not the key area of ​​the CP-type El Niño event. Therefore, in the next step, we will focus on analyzing the CP-type El Niño and identifying its key areas. Further analysis found that when the average link strength of the key area during the El Niño period exceeds a certain threshold (0.47), a La Niña event is likely to occur in the year or the next year after the end of the El Niño.

[0084] (6) Quantification of atmospheric response to CP El Niño Type I and II

[0085] The sea temperature grid points in the equatorial central Pacific region (160°E-150°W, 10°S-10°N) were selected to establish a network with the global temperature grid points to analyze the advanced response of the atmosphere to the two types of CP-type El Niño events. In this step, the EMI index is used to determine the CP-type El Niño event, and the MII index is used to distinguish between CP-I and CP-II El Niño events.

[0086] like Figure 6As shown in the figure, during the EP-type El Niño event, the average link strength shows an upward trend and reaches a peak at the end of the event. This peak is slightly smaller than the average link strength calculated in the equatorial central and eastern Pacific, which may indicate that the atmosphere has a greater impact on the equatorial central and eastern Pacific during the EP-type El Niño event. During the CP-I and CP-II El Niño events, the average link strength also shows an upward trend, and the increase is positively correlated with the intensity of the event itself. For La Niña events, if it occurs immediately after the end of a strong EP-type El Niño event or after the last La Niña event, the average link strength decreases monotonically; this situation indicates that the strong link between the atmosphere and the sea temperature in the equatorial central Pacific basin is largely disconnected. If the La Niña event occurs after a CP-type El Niño event or a weak EP-type El Niño event, its average link strength increases first and then decreases.

[0087] like Figure 7 As shown in Figure 2, both CP-I and CP-II El Niño events have strong positive links in the equatorial central and eastern Pacific. In CP-II El Niño years, there is also an obvious strong negative link area in the tropical Atlantic. Overall, compared with CP-I events, CP-II El Niño years have a stronger link between atmospheric temperature and sea temperature grid points. Figure 8 As shown in Figure 2, compared with normal years, the number of links in the CP-I El Niño event in the extra-equatorial region has increased significantly. Figure 7 It can be seen that for CP-I El Niño events, although the number of links in the equatorial grid points is not high, there are also strong links between the temperature grid points in the equatorial central and eastern Pacific region, which may also indicate that the strong link between the atmospheric temperature in the El Niño year and the sea temperature grid points in the equatorial central Pacific basin is local. For CP-II El Niño events, overall, most of the high link number areas are concentrated in the tropical region, and the link strength is strong. There are also a certain number of strong negative links in the Atlantic region. Further analysis of the spatial distribution of the average link strength and the number of links determines the key areas that have a greater impact on the sea temperature in the equatorial central Pacific, namely the equatorial central Pacific, the northern tropical Atlantic, and the northern temperate Atlantic.

[0088] like Fig. 9As shown in (a), during the CP-I El Niño event, the sum of weak link strengths first decreases and then increases, reaching a peak at the end of the event. During the CP-II El Niño event, the sum of weak link strengths shows a decreasing trend, and may slightly rebound in the later period. For La Niña events, the sum of weak link strengths mostly shows a downward trend. However, in the two La Niña events after the El Niño events in 1997 and 2009, the sum of weak link strengths increased instead of decreasing. During the El Niño event, when the sum of weak link strengths in the key area is lower than a certain threshold, and in the La Niña event that follows, the sum of weak links in the key area shows an increasing trend, it is very likely that a La Niña event will occur again after it ends.

[0089] like Fig. 9 As shown in (b) in Figure 1, during the CP-I El Niño event, the sum of the strong link strengths first reaches a peak and then drops rapidly in the later stages of the event. During the CP-II El Niño, the sum of the strong link strengths continues to rise and reaches a peak after the event. For the La Niña event immediately following the El Niño event (in the same year), the sum of the link strengths first increases and then decreases. In addition, after most La Niña events end, the sum of the strong link strengths will increase again before the next El Niño event begins, and then drop rapidly.

[0090] See also Fig.10 , Fig.10 4 is a schematic diagram of the working of the hardware device of an embodiment of the present invention, wherein the hardware device specifically comprises: an El Nino event recognition and La Nina transition warning device 401, a processor 402 and a storage medium 403.

[0091] An El Nino event identification and early warning device 401 for transition to La Nina: The El Nino event identification and early warning device 401 for transition to La Nina implements the El Nino event identification and early warning method for transition to La Nina.

[0092] Processor 402: The processor 402 loads and executes the instructions and data in the storage medium 403 to implement the method for identifying El Niño events and warning of their transition to La Niña.

[0093] Storage medium 403: The storage medium 403 stores instructions and data; the storage medium 403 is used to implement the El Niño event identification and La Niña transition warning method.

[0094] The beneficial effects of the present invention are:

[0095] (1) This paper reveals the interaction between the atmosphere and El Niño events and their key areas through complex network analysis, which not only deepens the understanding of the impact of El Niño events on the global climate system, but also provides scientific support for governments and international organizations to formulate more efficient climate change adaptation plans.

[0096] (2) The present invention proposes a method for warning of a possible La Niña event during an El Niño event in the current year or the next year, which helps to improve the accuracy of La Niña event prediction and provide a scientific basis for formulating disaster prevention and response strategies, thereby effectively reducing the losses caused by climate anomalies.

[0097] (3) By constructing an air temperature and sea temperature network based on different sea temperature regions, the present invention deeply explores the atmospheric advance response to different types of El Niño events, provides a new perspective for climate science research, and promotes further research and understanding of El Niño events and climate change.

[0098] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for identifying El Nino events and early warning of transition to La Nina, characterized in that: The method comprises the following steps: S1. Data collection: Collect sea surface temperature and 2-meter air temperature data from the fifth-generation atmospheric reanalysis product ERA5 of the European Centre for Medium-Range Weather Forecasts; S2. Data preprocessing: Using the data collected in step S1, calculate the daily average sea surface temperature and air temperature, define 1979-2019 as the reference state, subtract the obtained daily average sea surface temperature and air temperature from the reference state temperature, and obtain the anomaly time series; S3, climate network construction based on mutual correlation coefficient: Combined with the anomaly time series obtained in step S2, the temperature grid points and the sea temperature grid points are divided into two sets, and for each pair of nodes i and j in different sets, the lagged mutual correlation coefficient between them is calculated; The peak value of the absolute value of the mutual correlation coefficient and the corresponding time lag are used to represent the strength and direction of the link, and a directed network of air temperature and sea temperature is constructed; S4. Quantification of network link characteristics: Set appropriate mutual correlation coefficient thresholds to remove accidental links that are insensitive to El Niño events; focus on links from temperature to sea temperature, and calculate the number of links and average link strength at temperature grid points; S5. Quantification of atmospheric response to the eastern and central El Niño: combining the methods described in step S3 and step S4, calculating the frequency distribution of the link strength of the global temperature-equatorial central and eastern Pacific sea temperature network, the spatiotemporal distribution of the number of links and the link strength, identifying key areas with advanced responses to El Niño events, and identifying early warning signals of the transition of El Niño events to La Niña events based on the link strength of key areas; S6. Quantification of atmospheric response to central-regional El Niño type I and type II: Combined with the methods described in step S3 and step S4, the frequency distribution of the link strength of the global temperature-equatorial central Pacific sea temperature network, the spatiotemporal distribution of the number of links and the link strength are calculated to identify key areas that have an advanced response to the two types of CP El Niño events, and based on the link strength of key areas, the early warning signals of two consecutive La Niña events after the El Niño event are identified.

2. The method for identifying El Nino events and early warning of transition to La Nina according to claim 1, characterized in that: In step S2, the daily average temperature is calculated by combining the hourly sea temperature and air temperature data in step S1; 1979-2019 is defined as the reference state, and the multi-year average daily temperature of the reference state is calculated; the reference state temperature is subtracted from the daily average temperature to obtain the anomaly time series, and the formula is as follows: in, is the temperature of the grid point on the i-th day in the y-th year, where y=1 represents 1979, and so on; It means adding up the temperatures of the i-th day in all the years of the reference state and averaging them.

3. The method for identifying El Nino events and early warning of transition to La Nina according to claim 2, characterized in that: Step S3 is as follows: The anomaly time series obtained in step S2 is divided into two sets, one containing all global temperature grid points and the other containing all sea temperature grid points in the selected ocean area; for each pair of nodes i and j in different sets, the lagged mutual correlation coefficient between them is calculated as follows: in is the time series T i (d) The standard deviation, τ is the time interval, τ∈[0,τ max ],τ max = 200 days, y represents the starting time of the time series when each window time interval is 0, and C i,j (τ) = C j,i (-τ); the time lag corresponding to the maximum absolute value of the cross-correlation coefficient is expressed as The sign represents the direction of the link between nodes i and j. When , the link points from i to j; the absolute value of the mutual correlation coefficient represents the link strength, so a directed climate network of air temperature and sea temperature is established based on the mutual correlation coefficient.

4. The method for identifying El Nino events and early warning of transition to La Nina according to claim 3, characterized in that: Step S4 is specifically as follows: A threshold Δ is introduced to exclude links that are insensitive to El Niño events. The adjacency matrix defined by the threshold is as follows: Where H(X) is the Heaviside step function, defined as H(X≥0)=1 and H(X<0)=0; Consider the link from air temperature to sea temperature, that is The degree formula of the grid point is as follows: That is, the sum of all the mutual correlation coefficients of a grid point that exceed the threshold; the average link strength of the grid point is defined as follows: That is, the ratio of the degree of the grid point to the number of links to the grid point; is the number of links from the temperature grid to the sea temperature grid, and its formula is as follows:

5. The method for identifying El Nino events and early warning of transition to La Nina according to claim 4, characterized in that: Step S5 is specifically as follows: Step S51: Select sea temperature grid points in the equatorial central and eastern Pacific region to establish a network; define the occurrence and disappearance of El Nino events according to the Nino3.4 index. When the Nino3.4 index is greater than a first preset value ℃ and the duration of the event is greater than a second preset value, it is an El Nino event; when the Nino3.4 index is less than a third preset value and continues to be greater than the second preset value, it is a La Nina event; The time and value when the absolute value of the Nino3.4 index reaches the maximum are defined as the peak time and peak intensity of the event respectively; according to the peak event and peak intensity, the event is divided into weak intensity event, medium intensity event, strong intensity event and super strong intensity event; Step S52: Considering the different sea temperature anomaly areas and their different impacts between different types of El Nino events, the Eastern El Nino Index and the Central El Nino Index are used to distinguish the Eastern and Central El Nino events. The calculation formulas of these two indices are as follows: NEP=N3-αN4 NCP=N4-αN3 Where N3 is the regional average of the sea surface temperature anomaly in the Nino3 region, and N4 is the regional average of the sea surface temperature anomaly in the Nino4 region; when N3*N4>0, α is equal to 0.4, otherwise α is equal to 0; the distinction rules are as follows: (1) If the absolute value of the NEP index reaches or exceeds 0.5℃ and lasts for at least three months, it is an EP type event; (2) If the absolute value of the NCP index reaches or exceeds 0.5℃ and lasts for at least three months, it is a CP type event; (3) If an event meets both of the above points, the type of peak event shall prevail; Step S53: Select five El Niño events of medium intensity or above for analysis based on the Nino3.4 index; analyze the changes in the number and intensity of links during the El Niño event through the frequency distribution histogram of link intensity, the graph of link number variation over time, and the probability density distribution graph of link time interval; identify the key areas with advanced response to El Niño events through the spatiotemporal distribution graph of link intensity and link number, and determine the early warning signal of the transition from El Niño event to La Niña event by analyzing the time evolution of link intensity in the key areas.

6. A method for identifying El Nino events and early warning of transition to La Nina as claimed in claim 5, characterized in that: When the peak intensity of an event is >0.5℃ and <1.3℃, the event is defined as a weak intensity event; ≥1.3℃ and <2℃, a moderate intensity event; ≥2℃ and <2.5℃, a strong intensity event; and ≥2.5℃, a very strong intensity event.

7. A method for identifying El Nino events and early warning of transition to La Nina as claimed in claim 6, characterized in that: Step S6 is specifically as follows: Step S61: Select sea temperature grid points in the equatorial central Pacific region to establish a network; use the El Nino-Modoki index to identify CP-type El Nino events, and the calculation formula is as follows: EMI=[SSTA] C -0.5×[SSTA] W -0.5×[SSTA] E [SSTA] C , [SSTA] W , [SSTA] e Represent the average of the SST anomalies in the C, W, and E regions, respectively. The El Niño-Modoki II index is used to further distinguish between the central El Niño type I and central El Niño type II events. When the MII index is greater than 1 standard deviation, it is defined as a CP-II El Niño event, otherwise it is a CP-I event. Step S62: Select eight CP-type El Niño events for analysis based on EMI; analyze the characteristics of the central-type El Niño link and its difference from the eastern-type El Niño through the link strength frequency distribution diagram and the link time interval probability density distribution diagram; identify the key areas of CP-I and CP-II El Niño through the spatiotemporal distribution diagram of link strength and link number; finally, explore the signal of two consecutive La Niña events after the El Niño event through the time evolution analysis of link strength in the key areas.

8. A storage medium, characterized in that: The storage medium stores instructions and data for implementing an El Niño event identification and La Niña transition early warning method as described in any one of claims 1 to 8.

9. An El Nino event identification and early warning device for transition to La Nina, characterized in that: include: A processor and a storage medium; the processor loads and executes instructions and data in the storage medium to implement a method for identifying El Niño events and warning of transition to La Niña as described in any one of claims 1 to 8.

Images