A method for predicting interdecadal variability of el nino and la nina events using warm water volumes

Abstract

The application discloses a method for predicting interdecadal variation of El Nino and La Nina events by using warm water volume, which comprises the following steps: obtaining Nino3.4 index data, sea surface temperature data, warm water volume index data, 20 DEG C isotherm depth data and sea surface zonal wind stress data; selecting El Nino and La Nina events; calculating interdecadal variation of sea surface temperature anomaly; obtaining asymmetry and causal effect of warm water volume in predicting El Nino and La Nina events. The application solves the technical problem that the prior art does not involve interdecadal variation of ENSO events predicted by warm water volume, determines that the interdecadal variation of ENSO events predicted by warm water volume has significant asymmetry in positive and negative events, and the change trend of the key region of the tropical Pacific Ocean air-sea coupling center and the west warm and east cold aggravate the occurrence of continuous La Nina events, which influences the interdecadal variation of El Nino and La Nina events predicted by warm water volume.

Description

Technical Field

[0001] This invention relates to a method for predicting the interdecadal variations of El Niño and La Niña events, and more specifically to a method for predicting the interdecadal variations of El Niño and La Niña events using warm water volume. Background Technology

[0002] El Niño-Southern Oscillation (ENSO) is the most significant interannual variability signal of air-sea interaction in the tropical Pacific. El Niño (positive ENSO events) is typically associated with a significant increase in sea surface temperature (SST) in the central and eastern tropical Pacific, while La Niña (negative ENSO events) is associated with a decrease in SST in the central and eastern Pacific. ENSO events not only have local impacts, but they also influence global climate through atmospheric teleconnections, including precipitation distribution, tropical storm frequency and intensity, global carbon cycle, and ecosystem health. Studies have shown that the accumulation and release of heat in the equatorial Pacific is closely related to the evolution of ENSO events. Existing literature 1 (Jin F F. Anequatorial ocean recharge paradigm for ENSO. Part I: Conceptual model[J]. Journal of the Atmospheric Sciences, 1997, 54(7): 811-829, https: / / doi.org / 10.1175 / 1520-0469(1997)054<0811:AEORPF>2.0.CO;2) proposes a charging and discharging theory. This theory states that when the trade winds strengthen, warm water accumulates in the western Pacific Ocean, while the volume of warm water in the equatorial Pacific Ocean increases, entering a charging phase and accumulating energy for El Niño; when the trade winds weaken, the warm water spreads eastward, while the volume of warm water in the equatorial Pacific Ocean decreases, entering a discharging phase and triggering La Niña events. This cycle drives the alternation of El Niño and La Niña through the accumulation and release of ocean heat, in which the interaction between warm water volume and wind stress anomalies is the key to maintaining the oscillation.This theory not only explains the cyclical mechanism of ENSO events, but also provides a theoretical basis for predicting ENSO events. Reference 2 (Meinen CS and McPhaden M J. Observations of warm water volume changes in the equatorial Pacific and their relationship to El Niño and La Niña[J]. Journal of Climate, 2000, 13(20):3551-3559, https: / / doi.org / 10.1175 / 1520-0442(2000)013<3551:OOWWVC>2.0.CO;2) confirms through observational data that warm water volume can predict El Niño or La Niña events in the next 7 months. Positive warm water volume anomalies (charging) often lead to El Niño, while negative warm water volume anomalies (discharging) often lead to La Niña. However, this study mainly establishes correlations through statistical observational data and has not yet studied the impact of warm water volume on ENSO events through rigorous causal methods.

[0003] Recent studies have revealed a significant climate shift in the Pacific Ocean-atmosphere system around 2000. Correspondingly, the background climate of the Pacific also changed, with stronger equatorial easterly winds and a more pronounced tilt in the oceanic thermocline. Under this changing climate background, both ENSO and warm water volume underwent significant alterations. Firstly, ENSO characteristics shifted towards a higher frequency and smaller amplitude over time, with an increased frequency of El Niño events in the central Pacific. Simultaneously, studies indicate a significantly shortened lead time for warm water volume over ENSO events, but these studies did not differentiate between positive and negative ENSO events in their interdecadal variations. Research shows that El Niño and La Niña, as positive and negative ENSO events respectively, exhibit significant differences in their impact on the climate system. For example, in terms of average intensity, the average amplitude of El Niño events is greater than that of La Niña events. Spatially, strong El Niño events tend to occur in the eastern Pacific and are closer to the equator; strong La Niña events, on the other hand, develop more extensively in the central Pacific, with a wider latitudinal reach. Meanwhile, studies have shown that El Niño events tend to weaken rapidly in the following summer, while La Niña events last longer than El Niño events. However, research has mainly focused on the asymmetry of ENSO events themselves (such as spatial structure and persistence), while insufficient research has been conducted on the asymmetry of warm water volume in predicting El Niño and La Niña and their interdecadal variations. Reference 3 (Liang X S. Unraveling the cause-effect relationship between time series[J]. Physical Review E, 2014, 90(5): 052150,https: / / doi.org / 10.1103 / PhysRevE.90.052150) proposes a rigorous causal analysis method, the Liang-Kleeman information flow method. This study applied the information flow method to the study of the causal relationship between the Indian Ocean Dipole and ENSO. However, this information flow method has not been applied to the study of warm water volume predicting ENSO events.

[0004] In summary, existing technologies have studied the relationship between warm water volume and overall ENSO events, while El Niño and La Niña exhibit significant asymmetry. However, they have not addressed the asymmetry in warm water volume prediction of El Niño and La Niña event changes; the spatiotemporal structure changes in key regions for warm water volume prediction of ENSO events remain unclear; and rigorous causal analysis has not been used to study the prediction of ENSO events by warm water volume. Summary of the Invention

[0005] The purpose of this invention is to provide a method for predicting the interdecadal variations of El Niño and La Niña events using warm water volume. This method solves the technical problem that existing technologies do not address the prediction of interdecadal variations of ENSO events using warm water volume. It establishes that the prediction of interdecadal variations of ENSO events using warm water volume exhibits significant asymmetry between positive and negative events. Changes in key regions of the tropical Pacific air-sea coupling center and the trend of warming in the west and cooling in the east exacerbate the occurrence of consecutive La Niña events, ultimately affecting the prediction of interdecadal variations of El Niño and La Niña events using warm water volume.

[0006] To achieve the above objectives, the present invention provides a method for predicting the interdecadal variations of El Niño and La Niña events using warm water volume, the method comprising:

[0007] Step 1: Obtain Nino 3.4 index data, sea surface temperature data, warm water volume index data, 20°C isotherm depth data, sea surface zonal wind stress data, and ocean temperature data;

[0008] Step 2: Calculate and select El Niño and La Niña events based on the Nino3.4 index; use sea surface temperature data, combined with the selected El Niño and La Niña events, to calculate the composite sea surface temperature anomaly and obtain the interdecadal variation of sea surface temperature anomalies.

[0009] Step 3: Based on the selected El Niño and La Niña events, the interdecadal variation of warm water volume indices and Nino3.4 indexes, the phase method of warm water volume-Nino3.4, and the scatter plot of warm water volume-Nino3.4 are used to obtain the interdecadal variation law of warm water volume in predicting ENSO events, and to obtain the asymmetry of warm water volume in predicting El Niño and La Niña events.

[0010] Step 4: Using the 20°C isotherm depth data as an indicator of local warm water volume, and employing the lead-lag correlation method, the spatiotemporal variation patterns of local warm water volume and zonal wind stress affecting ENSO events are derived.

[0011] Step 5: Using the 20°C isotherm depth data as an indicator of local warm water volume, quantitative causal analysis is conducted using the information flow method to obtain the interdecadal variation of the causal influence of local warm water volume and zonal wind stress on ENSO events.

[0012] Step Six: Using ocean temperature data, we calculate and analyze the spatiotemporal variation of the influence of warm water volume and zonal wind stress on ENSO events, as well as the interdecadal variation of their causal effects. This analysis will reveal the background state variation of ocean temperature and its impact on warm water volume in predicting ENSO events.

[0013] Preferably, in step one, the horizontal resolution of the sea surface temperature data is 2° × 2°, and the warm water volume index data is defined as the amount of water with a temperature above 20°C in the equatorial Pacific Ocean region; the spatial resolution of the 20°C isotherm depth data and the sea surface zonal wind stress data is 0.25° × 0.25°; the ocean temperature data is monthly average ocean temperature data with a spatial resolution of 2° × 2°, calculated based on the global ocean data assimilation system dataset from the U.S. National Center for Weather and Environmental Prediction from January 1980 to December 2023.

[0014] Preferably, the ocean region of the equatorial Pacific Ocean is the ocean region from 120°E to 80°W and from 5°N to 5°S.

[0015] Preferably, in step two, calculating and selecting El Niño and La Niña events refers to calculating the time series of the Nino3.4 index from 1980 to 2023 and performing de-tilting processing. The Nino3.4 index is used as the indicator to define El Niño and La Niña events, and El Niño and La Niña events from 1980 to 1999 and 2000 to 2023 are selected. The El Niño events are positive ENSO events, and the La Niña events are negative ENSO events. The sea surface temperature anomaly synthesis is a synthesis of sea surface temperature anomalies from the El Niño and La Niña events from 1980 to 1999 and 2000 to 2023. The interdecadal variation of the sea surface temperature anomalies is the result of the El Niño and La Niña events in the two phases from 1980 to 1999 and 2000 to 2023. The differences in the spatial distribution characteristics of sea surface temperature anomalies; the Nino3.4 index is the average sea surface temperature anomaly value of the ocean area from 170°W to 120°W and from 5°N to 5°S, and the Nino3.4 index characterizes ENSO; the use of the Nino3.4 index as an indicator to define El Niño and La Niña events means that the Nino3.4 index is detrended, and an interval in which the detrended index exceeds +0.5°C and lasts for at least 5 months is defined as an El Niño event, and an interval in which the detrended index is below -0.5°C and lasts for at least 5 months is defined as a La Niña event; the sea surface temperature anomaly synthesis of ENSO events is obtained by multiplying [El Niño event - La Niña event] by 0.5 for 1980-1999 and 2000-2023.

[0016] Preferably, in step three, the study area for the warm water volume is the ocean region from 120°E to 80°W and from 5°N to 5°S; a warm water volume > 0 indicates that the tropical Pacific Ocean is in a charging state, while a warm water volume < 0 indicates that the tropical Pacific Ocean is in a discharging state; the interdecadal variation pattern is a composite analysis of El Niño and La Niña events selected based on the Nino3.4 index from 1980 to 1999 and from 2000 to 2023. The evolution and phase of the warm water volume-Nino3.4 in the El Niño or La Niña events are synthesized from the El Niño or La Niña events from 1980 to 1999 and from 2000 to 2023, respectively, while the temporal evolution of the ENSO event is obtained by multiplying [El Niño event - La Niña event] by 0.5 from 1980 to 1999 and from 2000 to 2023.

[0017] Preferably, step four includes: using the correlation between the leading and lagging phases of 1980-1999 and 2000-2023, analyzing the spatiotemporal distribution characteristics of the influence of local warm water volume and zonal wind stress on ENSO events in 1980-1999 and 2000-2023, and revealing their impact on ENSO prediction capabilities.

[0018] Preferably, in step five, the quantitative causal analysis is performed using sample covariance; the information flow method includes two time series X2 and X1, and the formula for the information flow from X2 to X1 is expressed as: ;

[0019] The variables are defined as follows:

[0020] C 12 C 11 and C 22 : C 12 C represents the covariance between X1 and X2. 11 C represents the covariance between X1 and X2. 22 This represents the covariance between X2 and X2;

[0021] C 1,d1 and C 2,d1 : C 1,d1 C represents the covariance between X1 and the derivative of X1. 2,d1 This represents the covariance between the derivatives of X2 and X1.

[0022] X1 refers to the Nino3.4 time series;

[0023] X2 refers to the time series of 20°C isotherm data or the time series of zonal wind stress data;

[0024] T 2→1: represents the information flow from X2 to X1, where a positive value indicates that X2 increases the uncertainty of X1, and a negative value indicates that X2 stabilizes X1; T represents the information flow from X2 to X1, assuming the reliability test is passed. 2→1 The larger the absolute value of T, the better. 2→1 The larger the amplitude, the greater the influence of X2 on X1.

[0025] Preferably, in step five, the 20°C isotherm depth data and zonal wind stress data are substituted into time series X2, and the Nino3.4 index is substituted into time series X1; the information flow from warm water volume to Nino3.4 and from zonal wind stress to Nino3.4 in the two periods of 1980-1999 and 2000-2023 are calculated respectively, and the causal effects of warm water volume and zonal wind stress anomalies on ENSO events are compared and evaluated to determine whether there have been changes; and the interdecadal variation in the prediction of ENSO by warm water volume is determined.

[0026] Preferably, in step six, the ocean temperature data refers to the ocean temperature data of the upper 300 meters of the Global Ocean Data Assimilation System dataset from January 1980 to December 2023, compiled by the U.S. National Center for Weather and Environmental Prediction.

[0027] Preferably, in step six, the calculation and analysis refers to calculating the average change of ocean temperature in the upper 300 meters of the tropical Pacific Ocean during the two periods of 1980-1999 and 2000-2023, analyzing the zonal distribution characteristics of the background state of the tropical Pacific Ocean during 1980-1999 and 2000-2023, and exploring its impact on the prediction of warm water volume ENSO.

[0028] This invention provides a method for predicting the interdecadal variations of El Niño and La Niña events using warm water volume. This method addresses the technical problem that existing technologies do not address the interdecadal variations of ENSO events using warm water volume, and offers the following advantages:

[0029] 1. Compared with existing technologies, this invention, for the first time, analyzes the asymmetry in the interdecadal variation of warm water volume in predicting El Niño and La Niña events from three perspectives (i.e., the temporal evolution of the warm water volume index and the Nino3.4 index, the phase method of warm water volume-Nino3.4, and the scatter plot of warm water volume-Nino3.4): In predicting El Niño events, the peak range of warm water volume remains within 8 months of the Nino3.4 peak, indicating that warm water volume is effective in predicting El Niño events 8 months in advance; in predicting La Niña events, the peak range of warm water volume changes from 8 months to 3 months, meaning that warm water volume predicts La Niña events from 8 months in advance to 3 months in advance, revealing the asymmetric pattern in the interdecadal variation of warm water volume in predicting El Niño and La Niña events. This invention further identifies the cause of this predictive asymmetry: since 2000, consecutive La Niña events have dominated, and the tropical Pacific background has shown La Niña-type warming.

[0030] 2. Compared with existing technologies, this invention is the first to utilize the information flow method to determine the causal influence of local warm water volume on ENSO events, identifying warm water volume as an effective dynamic predictor of ENSO events from the perspective of information flow. It calculates the information flow from local warm water volume to the Nino 3.4 index for the periods before and after 2000 (1980-1999 and 2000-2023), revealing the spatial structural variation of the influence of local warm water volume on ENSO events. It determines that the influence of warm water volume on ENSO events increases in the western equatorial Pacific (corresponding to an increase in the negative amplitude of the information flow in the western Pacific) and decreases in the eastern Pacific (corresponding to a decrease in the positive amplitude of the information flow in the eastern Pacific). This invention overcomes the limitations of previous isolated discussions of "decreased predictive timeliness of warm water volume" or "ENSO asymmetry" and simple statistical correlation analysis, providing a mechanistic explanation for the interdecadal variation of ENSO predictability. Attached Figure Description

[0031] Figure 1 This is a time series graph of the Nino3.4 index from 1980 to 2023.

[0032] Figure 2 The power spectrum of the Nino 3.4 exponent around 2000.

[0033] Figure 3 This is a map showing the distribution of sea surface temperature anomalies during the ENSO, El Niño, and La Niña events around 2000.

[0034] Figure 4 Synthetic time evolution of the warm water volume index and Nino3.4 index for ENSO events, El Niño and La Niña events around 2000.

[0035] Figure 5 Synthetic phase diagram of warm water volume and Nino3.4 for El Niño and La Niña events around 2000.

[0036] Figure 6 This is a normalized scatter plot of warm water volume around 2000 versus Nino3.4.

[0037] Figure 7 This is a spatial distribution map showing the leading and lagging correlation between the depth of the 20°C isotherm and the Nino3.4 index around 2000.

[0038] Figure 8 This is a spatial distribution map showing the leading and lagging correlation between zonal wind stress and the Nino3.4 index around 2000.

[0039] Figure 9 This is a graph showing the information flow distribution from the 20°C isotherm depth to the Nino 3.4 index around 2000.

[0040] Figure 10 This is a distribution map of information flow from zonal wind stress to the Nino 3.4 index around 2000.

[0041] Figure 11 This represents the interdecadal variation of the zonal distribution of ocean temperature in the upper 300 meters of the tropical Pacific Ocean around 2000. Detailed Implementation

[0042] The technical solutions in the embodiments of the present invention will be clearly and completely described below. 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.

[0043] Example 1:

[0044] A method for predicting interdecadal variations of El Niño and La Niña events using warm water volume, the method comprising:

[0045] Step 1: Obtain Nino 3.4 index data, sea surface temperature data, warm water volume index data, 20°C isotherm depth data, sea surface zonal wind stress data, and ocean temperature data.

[0046] Obtain Nino3.4 index data, provided by the Climate Prediction Center (CPC) of the National Oceanic and Atmospheric Administration (NOAA). It is defined as the mean sea surface temperature anomaly over the region (170°W–120°W, 5°N–5°S). The Nino3.4 index data was downloaded from https: / / www.cpc.ncep.noaa.gov / products / analysis_monitoring / ensostuff / detrend.nino34.ascii.txt. Obtain sea surface temperature data, derived from monthly mean data from the Extended Reconstructed Sea Surface Temperature Version 5 (ONAS) dataset, with a horizontal resolution of 2° × 10⁻⁶. 2°, data downloaded from https: / / downloads.psl.noaa.gov / Datasets / noaa.ersst.v5; Warm Water Volume Index data obtained, provided by the Australian Bureau of Meteorology National Operations Centre, defined as the amount of water with a temperature above 20°C in the equatorial Pacific Ocean (120°E–80°W, 5°N–5°S), is an important indicator reflecting changes in ocean heat content, Warm Water Volume Index data downloaded from https: / / www.pmel.noaa.gov / tao / wwv / data / wwv.dat; 20°C isotherm depth data and sea surface zonal wind stress data with a spatial resolution of 0.25° × 0.25° obtained, this data comes from the Ocean Reanalysis System 5 of the European Centre for Medium-Range Weather Forecasts (ECMWF). 5. The ORAS5 dataset, downloaded from https: / / cds.climate.copernicus.eu / datasets / reanalysis-oras5?tab=overview, provides monthly mean ocean temperature data with a spatial resolution of 2° × 2°. This data is based on the Global Ocean Data Assimilation System (GODAS) dataset from the U.S. National Center for Weather and Environmental Prediction, spanning from January 1980 to December 2023, and was downloaded from https: / / downloads.psl.noaa.gov / Datasets / godas / ...

[0047] The climatological period referred to in this paper is from January 1980 to December 2023. Outliers were calculated by taking the difference between the original monthly values ​​and the monthly climatological average values ​​during the study period and then standardizing them.

[0048] Step 2: Calculate and select El Niño and La Niña events based on the Nino3.4 index; use sea surface temperature data, combined with the selected El Niño and La Niña events, to calculate the composite sea surface temperature anomaly and obtain the interdecadal variation of sea surface temperature anomalies.

[0049] Based on the Nino3.4 index, the time series of the Nino3.4 index from 1980 to 2023 was calculated and de-tilted. According to the definition of the Climate Prediction Center, the Nino3.4 index was used as the indicator for defining positive and negative ENSO events. El Niño and La Niña events from 1980 to 1999 and from 2000 to 2023 were selected, with positive events representing El Niño events and negative events representing La Niña events. Power spectrum analysis of the de-tilted Nino3.4 index was performed to obtain the interdecadal variation characteristics of ENSO events. Using sea surface temperature data, combined with the selected El Niño and La Niña events, the sea surface temperature anomalies of the 1980-1999 and 2000-2023 El Niño and La Niña events were synthesized to obtain the spatial distribution characteristics of sea surface temperature anomalies of El Niño and La Niña events around 2000.

[0050] The interdecadal variation of sea surface temperature (SST) anomalies is defined as the difference in the spatial distribution characteristics of SST anomalies during the El Niño and La Niña events of 1980–1999 and 2000–2023. The Nino3.4 index is the average SST anomaly value in the region (170°W–120°W, 5°N–5°S), and it characterizes ENSO. Based on the Nino3.4 index, the index is detrended, and an interval with an index exceeding +0.5°C for at least 5 months is defined as an El Niño event, while an interval with an index below -0.5°C for at least 5 months is defined as a La Niña event. The composite of SST anomalies for ENSO events is obtained by multiplying [El Niño event - La Niña event] by 0.5 for the periods of 1980–1999 and 2000–2023.

[0051] like Figure 1 The chart shows the time series of the Nino 3.4 index (blue line) from 1980 to 2023. The red dashed line represents ±0.5°C (unit: °C). Figure 1It can be seen that between 1980 and 2023, a total of 12 El Niño events (1982 / 1983, 1986 / 1987, 1991 / 1992, 1994 / 1995, 1997 / 1998, 2002 / 2003, 2004 / 2005, 2006 / 2007, 2009 / 2010, 2014 / 2015, 2015 / 2016, 2018 / 2019) and 13 La Niña events (1984 / 1985, 1988 / 1989, 1995) were selected. The El Niño events are listed in the following dates: 1996, 1998 / 1999, 1999 / 2000, 2000 / 2001, 2007 / 2008, 2008 / 2009, 2010 / 2011, 2011 / 2012, 2020 / 2021, 2021 / 2022, and 2022 / 2023. Before 2000 (1980-1999), there were 5 El Niño events and 5 La Niña events. After 2000 (2000-2023), there were 7 El Niño events and 8 La Niña events.

[0052] like Figure 2 The image shows the power spectrum of the Nino 3.4 index around 2000, where (a) represents 1980-1999 and (b) represents 2000-2023. The red line represents the 95% confidence level. Figure 2 It can be seen that before 2000, the Nino3.4 index was dominated by a cycle power of 40-50 months, or 3-4 years, while after 2000, the Nino3.4 index was dominated by a cycle power of 15-20 months, or 1-2 years. This shows that the ENSO event cycle has obvious interdecadal variation characteristics before and after 2000.

[0053] like Figure 3 The image shows the sea surface temperature anomaly distribution of ENSO, El Niño, and La Niña events around 2000. (a), (c), and (e) are composite images of ENSO, El Niño, and La Niña events before 2000; (b), (d), and (f) are composite images of ENSO, El Niño, and La Niña events after 2000. (ENSO event data is shown in color, unit: °C). Figure 3 It can be seen that since 2000, ENSO events have not only undergone interdecadal variations in their cycle, but the sea surface temperature anomaly patterns they exhibit have also changed significantly. Figure 3 As shown in (a) and (b), compared to before 2000, the intensity of the ENSO sea surface temperature anomaly has significantly weakened since 2000, and the center of the largest sea surface temperature anomaly has shifted towards the central Pacific. Figure 3As shown in (c) to (f), synthesizing the positive and negative phases of ENSO reveals that sea surface temperature anomalies significantly weakened after 2000 for both El Niño and La Niña events. However, the shift of El Niño's sea surface temperature anomalies towards the central Pacific was more pronounced. Furthermore, this difference in the interdecadal variation of the sea surface temperature anomaly centers of the positive and negative phases of ENSO is related to the changes in the ocean-atmosphere coupling center after 2000, and the fact that the anomaly center of La Niña itself is located in the central Pacific.

[0054] Step 3: Based on the selected El Niño and La Niña events, using the temporal evolution of the warm water volume index and Nino3.4 index, the phase method of warm water volume-Nino3.4, and the scatter plot of warm water volume-Nino3.4, we can obtain the interdecadal variation law of warm water volume in predicting ENSO events and obtain the asymmetry of warm water volume in predicting El Niño and La Niña events.

[0055] This invention employs the evolution of the warm water volume index and the Nino3.4 index, the warm water volume-Nino3.4 phase method, and a warm water volume-Nino3.4 scatter plot to analyze the interdecadal variation of the charging and discharging processes around 2000. A warm water volume > 0 indicates the tropical Pacific is in a charging state, while a warm water volume < 0 indicates the tropical Pacific is in a discharging state. Based on the Nino3.4 index, El Niño and La Niña events selected for analysis around 2000 are synthesized. The evolution and phase of the warm water volume-Nino3.4 in El Niño or La Niña events are synthesized from El Niño or La Niña events around 2000, respectively. The temporal evolution of ENSO events is obtained by multiplying [El Niño event - La Niña event] by 0.5 around 2000. All events include the previous year, the current year, and the following year. Specifically, the year in which the absolute value of the index first exceeds 0.5°C is defined as the current year, the year before is the previous year, and the year after is the following year.

[0056] like Figure 4 The figure shows the composite temporal evolution of the warm water volume index and Nino3.4 index for ENSO events, El Niño, and La Niña events around 2000. The red line represents warm water volume, and the blue line represents Nino3.4. The data has been normalized, and the standard deviation of the warm water volume index is 1.06 × 10⁻⁶. 14 cubic meters, the standard deviation of the Nino3.4 index is 0.887°C; 0 in parentheses on the horizontal axis represents the current year, and 1 represents the following year; (a) ENSO 1980-1999; (b) ENSO 2000-2023; (c) El Niño 1980-1999; (d) El Niño 2000-2023; (e) La Niña 1980-1999; (f) La Niña 2000-2023. (By) Figure 4As can be clearly seen in (a) to (b), the evolution trends of the Nino3.4 index and warm water volume in ENSO events around 2000 are generally similar: the peak value of warm water volume always precedes the peak value of the Nino3.4 index, but there are significant interdecadal differences. Before 2000, the intensity of warm water volume was greater, with a lead time of up to 8 months (peaking in spring); while after 2000, the intensity of warm water volume weakened, the lead time shortened to about 3 months (peaking in autumn), and the discharge process slowed down after the following spring. In addition, the anomalous intensity of the Nino3.4 index also weakened after 2000.

[0057] Overall, the lead time of warm water volume in predicting ENSO events has decreased since 2000, but whether the prediction times of its positive and negative events (El Niño and La Niña events) are consistent or asymmetrical requires further investigation.

[0058] Depend on Figure 4 As shown in (c) to (f), the charging and discharging processes of El Niño and La Niña exhibit significant asymmetry in their interannual variations around 2000. For El Niño ( Figure 4 (c) and (d)) After 2000, the charging and discharging process of El Niño events remained significant. Both involved initial charging by the warm water volume in the tropical Pacific (warm water volume > 0), causing an abnormal increase in sea surface temperature in the Nino 3.4 region, followed by discharging by the warm water volume (warm water volume < 0), leading to an abnormal decrease in sea surface temperature. However, compared to the Nino 3.4 sea surface temperature, the lead time of the warm water volume charging process (warm water volume > 0) remained roughly the same, still occurring approximately 8 months before the event. As for La Niña events (… Figure 4 (e) and (f)) After 2000, the discharge process of warm water volume in La Niña events weakened and the lead time shortened. Its peak value was only about 3 months earlier than Nino3.4. Before 2000, the evolution of Nino3.4 showed that the evolution of La Niña events was mainly the transformation from El Niño to La Niña events, that is, Nino3.4 changed from positive to negative values ​​in the early stage. After 2000, the main evolution form was continuous La Niña events (the Nino3.4 index remained negative from the early stage to the development). In addition, in this year, 3 months before the abnormal drop in sea surface temperature in the Nino3.4 region, there was no obvious discharge process of warm water volume. Instead, it continued to charge.

[0059] From the phase diagram, as Figure 5As shown, this is a composite phase diagram of warm water volume and Nino3.4 for El Niño and La Niña events around 2000. The green line represents the previous year, the red line the current year, and the blue line the following year, all normalized. (a) El Niño 1980-1999; (b) El Niño 2000-2023; (c) La Niña 1980-1999; (d) La Niña 2000-2023, arranged chronologically in a clockwise direction. Figure 5 As shown in (a) and (b), in the phase evolution of El Niño events, whether before or after 2000 years, El Niño events all begin with a small positive warm water volume and near-zero Nino3.4 in the previous year (-1 year), gradually develop to reach their peak in the current year (year 0), and begin to decline in the following year (year 1). And... Figure 5 In the phase evolution of La Niña events (c) and (d), before 2000, they were mainly transformed from El Niño events, with significant charging and discharging processes. The warm water volume reached the discharge peak at the end of -1 year and then gradually decayed. After 2000, La Niña events mainly manifested as continuous La Niña events, with weakened charging and discharging processes and reduced discharge amplitude.

[0060] like Figure 6 As shown, this is a normalized scatter plot of warm water volume versus Nino3.4 around 2000. (a) and (b) show the warm water volume leading the Nino3.4 index by 8 months in 1980-1999 and 2000-2023, respectively; (c) and (d) show the warm water volume leading the Nino3.4 index by 3 months in 1980-1999 and 2000-2023, respectively. Figure 6 As shown in (a) and (c), before 2000, the warm water volume in spring (February-April) had a strong correlation with the Nino3.4 index (overall correlation coefficient 0.75), and its predictive ability for both El Niño and La Niña events was relatively stable (correlation coefficients > 0.5), indicating that warm water volume remained an effective ENSO predictor even 8 months in advance. However, while the warm water volume in summer (July-September) (3 months in advance) had a high overall correlation (0.72), its predictive ability for La Niña events was weak (0.41), showing an asymmetry in its predictive ability between positive and negative events.

[0061] Depend on Figure 6As shown in (b) and (d), after 2000, the predictive power of spring warm water volume significantly decreased (total correlation coefficient 0.43); while the predictive power of summer warm water volume increased (total correlation coefficient 0.74), but this change was mainly driven by La Niña events—its correlation increased sharply from 0.55 to 0.88, while the predictive power of El Niño events decreased (from 0.68 to 0.46). This indicates that the prediction time of warm water volume for Nino3.4 shortened after 2000 (from 8 months to 3 months), mainly because La Niña events rely more on short-term (summer) warm water volume anomalous signals, while the predictive mechanism of El Niño events may be affected by other factors. Furthermore, through... Figure 6 The study found that the correlation between warm water volume and the Nino3.4 index for positive and negative events increased, further revealing the asymmetry in the prediction of El Niño and La Niña events by warm water volume.

[0062] Step 4: Using the 20°C isotherm depth data as an indicator of local warm water volume, and employing the lead-lag correlation method, the spatiotemporal variation patterns of local warm water volume and zonal wind stress affecting ENSO events are derived.

[0063] By utilizing the correlations under different time lead and lag conditions, we analyze the spatiotemporal distribution characteristics of the influence of local warm water volume and zonal wind stress on ENSO events around 2000, and reveal their impact on ENSO prediction capabilities.

[0064] Existing research indicates that zonal wind stress plays a crucial dynamic role in the triggering and maintenance of ENSO events by modulating equatorial undercurrents and thermocline depth. To further explore the relationship between warm water volume and the Nino3.4 index across different decadal periods, lead-lag correlation analyses were conducted for two periods: 1980–1999 and 2000–2023. These analyses investigated the correlation changes between warm water volume and zonal wind stress in February–April, May–June, and July–September, and Nino3.4 in October–December, revealing the interdecadal differences in their spatiotemporal structure. The 20°C isotherm depth data can be used as an indicator of local warm water volume.

[0065] like Figure 7 The image shows the spatial distribution of the leading-lag correlation between the 20°C isotherm depth and the Nino3.4 index around 2000. The left column represents 1980-1999, and the right column represents 2000-2023. The 20°C isotherm depth data are for February-April, May-June, and July-September, and the Nino3.4 index is for October-December. Figure 7As shown in (a) and (b), when the correlation is 8 months ahead, during the period of 1980-1999, the area of ​​high correlation covered 160°E-120°W, with the highest correlation reaching 0.6. However, after 2000, the positive correlation area between spring (February-April) warm water volume and winter Nino3.4 significantly shrank, with the correlation coefficient dropping to around 0.4, and concentrated in the central Pacific region (160°W-120°W). Figure 7 As shown in (c) and (d), when the correlation coefficient is 5 months ahead, before 2000, the positive correlation coefficient further increased to 0.8, and the area of ​​maximum correlation shifted eastward to the eastern Pacific Ocean compared to when the correlation coefficient was 8 months ahead. After 2000, the positive correlation coefficient slightly increased (0.6) and the area of ​​correlation further expanded, but overall it was still less than before 2000. Figure 7 As shown in (e) and (f), when the correlation coefficient is 3 months ahead, before 2000, the region with the largest correlation further expanded, with the region having a correlation coefficient of 0.8 or higher covering 150°W-100°W. After 2000, the correlation coefficient was also significantly higher than before 2000, with the region having a correlation coefficient of 0.8 or higher covering 160°W-100°W.

[0066] like Figure 8 The image shows the spatial distribution of the leading-lag correlation between zonal wind stress and the Nino3.4 index around 2000. The left column represents 1980-1999, and the right column represents 2000-2023. Zonal wind stress is the zonal wind stress from February to April, May to June, and July to September based on the depth data of the 20°C isotherm, and the Nino3.4 index is the Nino3.4 index from October to December. Figure 8 It can be seen that the changes in wind stress and warm water volume are basically consistent, indicating that the ocean-atmosphere coupling center has shifted significantly westward since 2000.

[0067] In summary, after 2000, the timeliness and spatial effectiveness of warm water volume in key ENSO regions decreased significantly from 8 months to 5 months, and the area with the greatest correlation between ocean and air showed shrinkage and weakening, but remained significant when leading by 3 months. This change further led to a decrease in thermocline feedback efficiency 8 to 5 months in advance, and a weakening of ENSO event prediction capability.

[0068] Step 5: Using the 20°C isotherm depth data as an indicator of local warm water volume, quantitative causal analysis is conducted using the information flow method to obtain the interdecadal variation of the causal influence of local warm water volume and zonal wind stress on ENSO events.

[0069] The results in step four indicate that warm water volume and zonal wind stress are correlated with the Nino3.4 index, exhibiting a certain lead-lag characteristic. However, the correlation analysis itself cannot distinguish the direction of influence, i.e., whether changes in warm water volume and zonal wind stress dominate Nino3.4, or whether changes in Nino3.4 lead to changes in warm water volume and zonal wind stress.

[0070] Therefore, this invention further employs the Liang-Kleeman information flow method, rigorously derived based on first principles, for analysis. Compared to correlation methods, the information flow method not only provides the intensity of the influence but also reflects its direction. For two time series X2 and X1, the final formula for the information flow from X2 to X1 (Liang X S. Unraveling the cause-effect relation between time series[J]. Physical Review E, 2014, 90(5):052150, https: / / doi.org / 10.1103 / PhysRevE.90.052150) is expressed as: ;

[0071] The variables are defined as follows:

[0072] C 12 C 11 and C 22 : C 12 C represents the covariance between X1 and X2. 11 C represents the covariance between X1 and X2. 22 This represents the covariance between X2 and X2;

[0073] C 1,d1 and C 2,d1 : C 1,d1 C represents the covariance between X1 and the derivative of X1. 2,d1 This represents the covariance between the derivatives of X2 and X1.

[0074] X1 refers to the Nino3.4 time series;

[0075] X2 refers to the time series of 20°C isotherm data or the time series of zonal wind stress data;

[0076] T 2→1 : represents the information flow from X2 to X1, where a positive value indicates that X2 increases the uncertainty of X1, and a negative value indicates that X2 stabilizes X1; T represents the information flow from X2 to X1, assuming the reliability test is passed. 2→1 The larger the absolute value of T, the better. 2→1 The larger the amplitude, the greater the influence of X2 on X1.

[0077] This formula uses sample covariance for causal analysis, avoiding complex entropy calculations and making causal measurement simpler and applicable to high-dimensional and nonlinear systems. Furthermore, when reliability is tested, regardless of whether the information flow rate is positive or negative, its absolute value reflects the strength of the causal influence.

[0078] Based on the information flow causal analysis method in Reference 3, the 20°C isotherm depth data and zonal wind stress data were substituted into time series X2, and the Nino3.4 index was substituted into time series X1. The information flow from warm water volume to Nino3.4 and from zonal wind stress to Nino3.4 in the two periods of 1980-1999 and 2000-2023 were calculated respectively. The causal effects of warm water volume and zonal wind stress anomalies on ENSO events were compared and evaluated to determine whether there were changes. It was determined that the prediction of ENSO by warm water volume has undergone interdecadal changes.

[0079] like Figure 9 The diagram shows the information flow distribution from the depth of the 20°C isotherm to the Nino 3.4 index around 2000. The information flow from warm water volume to the Nino 3.4 index is as follows: (a) from the depth of the 20°C isotherm to Nino 3.4 (1980-1999); (b) from the depth of the 20°C isotherm to Nino 3.4 (2000-2023). Unit: nanoteslas / month. The marked areas represent areas that passed the 95% confidence test. Figure 9 As shown in (a), from 1980 to 1999, the information flow from warm water volume to Nino3.4 in the western equatorial Pacific (150°E-0°) was negative, reaching a minimum of -0.08 nanoteslas per month, indicating that the western Pacific warm pool has a strong stabilizing effect on ENSO and is an important moderating factor for ENSO events. The information flow in the eastern equatorial Pacific (150°W-100°W) was positive, reaching a maximum of 0.18 nanoteslas per month, indicating that the eastern Pacific warm water has a significant impact on ENSO, but also increases ENSO instability. Figure 9 As shown in (b), from 2000 to 2023, the causal influence in the western equatorial Pacific region strengthened, with the information flow value decreasing to -0.1 nanots / month and expanding southward, indicating that the stabilizing effect of the western Pacific warm pool on ENSO was further enhanced. Meanwhile, the information flow in the eastern equatorial Pacific weakened, with the maximum value decreasing to 0.12 nanots / month. Furthermore, the positive value area narrowed significantly to the north and south, indicating that the influence of the 20°C isotherm depth on ENSO instability decreased in the eastern Pacific and gradually increased in the central Pacific.

[0080] like Figure 10The diagram shows the information flow distribution from zonal wind stress to the Nino 3.4 index around 2000. The information flow from zonal wind stress to the Nino 3.4 index is as follows: (a) from zonal wind stress to Nino 3.4 (1980-1999); (b) from zonal wind stress to Nino 3.4 (2000-2023). Unit: nanotes per month. The marked areas represent areas that passed the 95% confidence test. Figure 10 As shown in (a), through further causal analysis of information flow, this invention found that the zonal wind stress in the central and eastern equatorial Pacific (150°E-160°W) region from 1980 to 1999 had a significantly positive impact on the information flow of Nino3.4, with the highest intensity reaching 0.12 nanoteslas per month. Figure 10 As shown in (b), from 2000 to 2023, the high-value area of ​​information flow shifted westward to the (160°E-170°W) region, and its intensity weakened to approximately 0.08 nanoteslas per month. This indicates that the causal influence of zonal wind stress on ENSO events has weakened, and the wind stress regulation capacity of the equatorial Pacific has decreased.

[0081] In summary, after 2000, the causal effects of warm water volume and zonal wind stress on ENSO weakened in the eastern Pacific but strengthened in the western Pacific, and the area with the highest information flow values ​​showed a southward expansion trend.

[0082] Step Six: By analyzing the spatiotemporal variation of warm water volume and zonal wind stress in relation to ENSO events and the interdecadal variation of their causal effects, and using ocean temperature data, we calculate and analyze the background temperature variation of the upper 300-meter ocean in the tropical Pacific and its impact on warm water volume in predicting ENSO events.

[0083] Around 2000, the performance of ENSO events underwent significant interdecadal variations. The relationship between warm water volume and zonal wind stress and Nino3.4 obtained in steps four and five showed significant interdecadal variations in time, space, and intensity. This indicates that the air-sea coupling center and ocean thermal variability in the tropical Pacific have also undergone significant adjustments. These changes not only affect the occurrence and evolution of ENSO events but also have a profound impact on their predictability.

[0084] Using upper 300-meter ocean temperature data from GODAS, we calculated the multi-year average variation of upper tropical Pacific ocean temperature in two periods: 1980–1999 and 2000–2023. We analyzed the zonal distribution characteristics of the tropical Pacific background temperature around 2000 and further explored its impact on the warm water volume prediction ENSO.

[0085] like Figure 11As shown, this illustrates the interdecadal variation of zonal ocean temperature distribution at the upper 300 meters of the tropical Pacific Ocean around 2000, with the north-south direction representing the average within the range of (2°N-2°S). (a) shows the difference in average ocean temperature, with black dots indicating significant areas at the 99% significance level of the t-test; (b) shows the difference in ocean temperature standard deviation, with black dots indicating areas passing the 99% significance level of the F-test; (c) shows the average ocean temperature (isolines) and standard deviation (shaded). In (a) and (b), the green dashed lines represent the 20°C isotherm outline in GODAS from January 1980 to December 2023, while in (c), the green solid line (blue dashed line) represents the 20°C isotherm outline in GODAS from January 1980 to January 1999 (January 2000 to December 2023), respectively representing the thermocline and its interdecadal variation (colored, unit: °C). Figure 11 As shown in (a), compared to 1980-1999, the average sea surface temperature (SST) in the western tropical Pacific increased significantly from 2000 to 2023, with a maximum increase of 1.4°C, while the SST in the eastern Pacific showed a significant decrease, with a maximum decrease of -0.8°C. Overall, this exhibits a La Niña-like interdecadal variation trend of "warmer in the west and colder in the east." Furthermore, Figure 11 (b) shows that the temperature variability (temperature standard deviation difference) is significantly reduced above the thermocline in the central and western Pacific and below the thermocline in the eastern Pacific, with the largest reduction in variability reaching -1.2°C. Figure 11 (c) shows that the slope of the thermocline further increases, meaning that after 2000 the thermocline exhibited a "deeper in the west and shallower in the east" structure.

[0086] Therefore, overall, the climate variability in the tropical Pacific has decreased since 2000, and the background state has tended towards a "La Niña-like" warming, that is, the ocean temperature structure characteristics of a colder eastern Pacific and a warmer western Pacific have been further strengthened. This trend can sustainably enhance the cold phase tendency of ENSO, support more persistent shallow easterly winds and cold water upwelling, making it easier to trigger and maintain consecutive La Niña events. This has led to a significant shortening of the lead time for predicting La Niña based on warm water volume. The occurrence mechanism of El Niño still depends on the eastward shift of warm water volume in the western Pacific and the feedback of sea surface temperature anomalies. Therefore, although the background state has cooled down, the accumulation of warm water volume still has predictive value for leading seasonal scales.

[0087] Although the present invention has been described in detail through the preferred embodiments above, it should be understood that the above description should not be considered as a limitation of the present invention. Various modifications and substitutions to the present invention will be apparent to those skilled in the art after reading the above description. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A method for predicting the interdecadal variation of El Niño and La Niña events using warm water volume, characterized in that, The method includes: Step 1: Obtain Nino 3.4 index data, sea surface temperature data, warm water volume index data, 20°C isotherm depth data, sea surface zonal wind stress data, and ocean temperature data; Step 2: Calculate and select El Niño and La Niña events based on the Nino3.4 index; use sea surface temperature data, combined with the selected El Niño and La Niña events, to calculate the composite sea surface temperature anomaly and obtain the interdecadal variation of sea surface temperature anomalies. Step 3: Based on the selected El Niño and La Niña events, the interdecadal variation of warm water volume indices and Nino3.4 indexes, the phase method of warm water volume-Nino3.4, and the scatter plot of warm water volume-Nino3.4 are used to obtain the interdecadal variation law of warm water volume in predicting ENSO events, and to obtain the asymmetry of warm water volume in predicting El Niño and La Niña events. Step 4: Using the 20°C isotherm depth data as an indicator of local warm water volume, and employing the lead-lag correlation method, the spatiotemporal variation patterns of local warm water volume and zonal wind stress affecting ENSO events are derived. Step 5: Using the 20°C isotherm depth data as an indicator of local warm water volume, quantitative causal analysis is conducted using the information flow method to obtain the interdecadal variation of the causal influence of local warm water volume and zonal wind stress on ENSO events. Step Six: Using ocean temperature data, we calculate and analyze the spatiotemporal variation of the influence of warm water volume and zonal wind stress on ENSO events, as well as the interdecadal variation of their causal effects. This analysis will reveal the background state variation of ocean temperature and its impact on warm water volume in predicting ENSO events.

2. The method according to claim 1, characterized in that, In step one, the horizontal resolution of the sea surface temperature data is 2° × 2°, and the warm water volume index data is defined as the amount of water with a temperature above 20°C in the equatorial Pacific Ocean region; the spatial resolution of the 20°C isotherm depth data and the sea surface zonal wind stress data is 0.25° × 0.25°; the ocean temperature data is monthly average ocean temperature data with a spatial resolution of 2° × 2°, calculated based on the global ocean data assimilation system dataset from January 1980 to December 2023 of the U.S. National Center for Weather and Environmental Prediction.

3. The method according to claim 2, characterized in that, The ocean region of the equatorial Pacific Ocean is defined as the ocean region from 120°E to 80°W and from 5°N to 5°S.

4. The method according to claim 1, characterized in that, In step two, calculating and selecting El Niño and La Niña events refers to calculating the time series of the Nino3.4 index from 1980 to 2023 and performing de-tilting processing. The Nino3.4 index is used as the indicator to define El Niño and La Niña events, and El Niño and La Niña events from 1980 to 1999 and 2000 to 2023 are selected. El Niño events are positive ENSO events, and La Niña events are negative ENSO events. The sea surface temperature anomaly synthesis is a synthesis of sea surface temperature anomalies from the El Niño and La Niña events from 1980 to 1999 and 2000 to 2023. The interdecadal variation of the sea surface temperature anomalies is the sea surface temperature anomaly from the two El Niño and La Niña events from 1980 to 1999 and 2000 to 2023. The differences in the spatial distribution characteristics of temperature anomalies; the Nino3.4 index is the average sea surface temperature anomaly value of the ocean area from 170°W to 120°W and from 5°N to 5°S, and the Nino3.4 index characterizes ENSO; the use of the Nino3.4 index as an indicator for defining El Niño and La Niña events means that the Nino3.4 index is detrended, and an interval in which the detrended index exceeds +0.5°C and lasts for at least 5 months is defined as an El Niño event, and an interval in which the detrended index is below -0.5°C and lasts for at least 5 months is defined as a La Niña event; the sea surface temperature anomaly synthesis of ENSO events is obtained by multiplying [El Niño event - La Niña event] by 0.5 for 1980-1999 and 2000-2023.

5. The method according to claim 1, characterized in that, In step three, the study area for the warm water volume is the ocean region from 120°E to 80°W and from 5°N to 5°S; a warm water volume > 0 indicates that the tropical Pacific Ocean is in a charging state, while a warm water volume < 0 indicates that the tropical Pacific Ocean is in a discharging state; the interdecadal variation pattern is a composite analysis of El Niño and La Niña events selected based on the Nino3.4 index from 1980 to 1999 and from 2000 to 2023. The evolution and phase of the warm water volume-Nino3.4 in El Niño or La Niña events are synthesized from the El Niño or La Niña events from 1980 to 1999 and from 2000 to 2023, respectively, while the temporal evolution of the ENSO event is obtained by multiplying [El Niño event - La Niña event] by 0.5 from 1980 to 1999 and from 2000 to 2023.

6. The method according to claim 1, characterized in that, Step four includes: using the lead-lag correlation between 1980-1999 and 2000-2023, analyzing the spatiotemporal distribution characteristics of the influence of local warm water volume and zonal wind stress on ENSO events in 1980-1999 and 2000-2023, and revealing their impact on ENSO prediction capabilities.

7. The method according to claim 1, characterized in that, In step five, the quantitative causal analysis is performed using sample covariance; the information flow method includes two time series X2 and X1, and the formula for the information flow from X2 to X1 is expressed as: ; The variables are defined as follows: C 12 C 11 and C 22 : C 12 C represents the covariance between X1 and X2. 11 C represents the covariance between X1 and X2. 22 This represents the covariance between X2 and X2; C 1,d1 and C 2,d1 : C 1,d1 C represents the covariance between X1 and the derivative of X1. 2,d1 This represents the covariance between the derivatives of X2 and X1. X1 refers to the Nino3.4 time series; X2 refers to the time series of 20°C isotherm data or the time series of zonal wind stress data; T 2→1 : represents the information flow from X2 to X1, where a positive value indicates that X2 increases the uncertainty of X1, and a negative value indicates that X2 stabilizes X1; T represents the information flow from X2 to X1, assuming the reliability test is passed. 2→1 The larger the absolute value of T, the better. 2→1 The larger the amplitude, the greater the influence of X2 on X1.

8. The method according to claim 7, characterized in that, Step five also includes: substituting the 20°C isotherm depth data and zonal wind stress data into time series X2, and substituting the Nino3.4 index into time series X1; calculating the information flow from warm water volume to Nino3.4 and from zonal wind stress to Nino3.4 for the two periods of 1980-1999 and 2000-2023 respectively, comparing and evaluating whether the causal effects of warm water volume and zonal wind stress anomalies on ENSO events have changed; and determining that the prediction of ENSO by warm water volume has undergone interdecadal changes.

9. The method according to claim 1, characterized in that, In step six, the ocean temperature data refers to the ocean temperature data of the upper 300 meters of the Global Ocean Data Assimilation System dataset from January 1980 to December 2023, compiled by the U.S. Centers for Weather and Environmental Prediction.

10. The method according to claim 9, characterized in that, In step six, the calculation and analysis refers to calculating the average change of ocean temperature in the upper 300 meters of the tropical Pacific Ocean during the two periods of 1980-1999 and 2000-2023, analyzing the zonal distribution characteristics of the background state of the tropical Pacific Ocean during 1980-1999 and 2000-2023, and exploring its impact on the prediction of warm water volume ENSO.

Images