La Nina event prediction method based on sea level anomaly in the eastern Indian Ocean

Abstract

The present invention relates to a La Nina event prediction method based on the abnormal sea level height in the East Indian Ocean, using the marker parameters of the abnormal rising Kelvin wave in the East Indian Ocean during the dipole event, such as the abnormal sea level height or the abnormal subsurface sea temperature, including using the near-equatorial sea level height or the abnormal subsurface sea temperature in the Indonesian Sea or the Western Pacific Ocean, or the East Indian Ocean Kelvin wave obtained by the Indian Ocean wind field, and using a regression prediction model to predict the start-up process and outbreak intensity of the La Nina event. The present invention can break through the spring barrier and successfully predict the outbreak of the La Nina event one year in advance.

Description

Technical Field

[0001] The present invention relates to the field of ocean and climate monitoring and forecasting, and in particular to a short-term forecasting method for ENSO cold phase La Nina events. Background Art

[0002] ENSO events are the strongest interannual climate anomalies in the world, and have a significant impact on global climate, society, and economic development. Therefore, the forecast of ENSO is particularly important. At present, ENSO forecast models are mainly divided into statistical models, dynamic models, and methods combining statistics and dynamics. Regardless of which model is adopted, there is a certain deviation in the accuracy of ENSO forecasts. How to improve the accuracy of ENSO forecasts is an urgent problem to be solved today. ENSO spring forecast obstacles are one of the main reasons for the large uncertainty in ENSO forecasts.

[0003] The ENSO "spring forecast barrier" refers to the phenomenon that when most climate models predict ENSO events, the model and continuity forecast skills decline rapidly around April and May, resulting in greater uncertainty in the forecast results. Specifically, The exponential lagged autocorrelation coefficient drops rapidly to close to 0 after crossing the spring. At present, the international climate forecasting business departments can only successfully predict the outbreak of ENSO events up to six months in advance, and there is still great uncertainty in the forecast of abnormal sea surface temperature in the cold tongue area of ​​the eastern Pacific. La Niña events are the cold phase of ENSO events, and their forecast is also subject to the forecast limitations of ENSO. At present, the international community can only successfully predict the outbreak of ENSO up to six months in advance, and it has received little attention. Summary of the invention

[0004] The purpose of the present invention is to provide a linear regression forecasting model based on the anomaly of sea level height in the eastern Indian Ocean during the Indian Ocean Dipole as a precursor factor in response to the above problems.

[0005] To achieve the above object, the present invention provides the following technical solution: a method for predicting La Nina events based on the abnormal sea level height in the eastern Indian Ocean, comprising the following steps:

[0006] A forecast model is constructed based on historical ocean and atmospheric data to represent the relationship between precursor predictors and forecast parameters;

[0007] Based on the precursor prediction factors at the set time, the forecast model is used to predict the La Niña event in the following year.

[0008] The ocean atmosphere historical data includes ocean historical data and atmospheric historical data;

[0009] The ocean historical data include: sea level anomaly or ocean surface temperature anomaly or sub-surface temperature anomaly;

[0010] The ocean historical data include, but are not limited to, those obtained through satellite observation, Argo profiling buoy observation, oceanographic survey ship observation, or climate reanalysis or prediction based on existing ocean historical data sets;

[0011] Atmospheric historical data include: offshore wind fields;

[0012] The atmospheric historical data include, but are not limited to, data obtained through weather stations, satellite observations, climate reanalysis or prediction based on existing sea-atmosphere historical data sets.

[0013] The precursor predictor is one of the following types:

[0014] Abnormal sea level height or abnormal ocean surface temperature or abnormal subsurface temperature in the eastern Indian Ocean;

[0015] The near-equatorial sea level anomaly or the surface sea temperature anomaly or the subsurface sea temperature anomaly in the Indonesian Sea;

[0016] The near-equatorial sea level anomaly or surface sea temperature anomaly or subsurface sea temperature anomaly in the western Pacific;

[0017] East Indian Ocean Kelvin waves obtained from the Indian Ocean atmospheric wind field;

[0018] The forecast parameters are for the sea surface temperature anomaly in the cold tongue area of ​​the eastern Pacific index.

[0019] The prediction parameters are Any of the indices.

[0020] The method of constructing a forecast model for representing the relationship between precursor forecast factors and forecast parameters based on the historical marine atmospheric data comprises the following steps:

[0021] Obtain any one of the sea level height anomaly, sea surface temperature anomaly, subsurface temperature anomaly in the ocean historical data, and Kelvin wave obtained according to the sea surface wind field in the atmospheric historical data as a precursor prediction factor;

[0022] The precursor prediction factors and prediction parameters are fitted to obtain the relationship between the precursor prediction factors and prediction parameters as the prediction model.

[0023] The forecast model is as follows:

[0024] Y predt =a+b*x

[0025] Among them, b is the regression coefficient, a is the constant term, x is the average precursor factor in the autumn of the year when the dipole occurs, and Y predtThese are the forecast parameters for each month after the dipole occurs, from June of the second year to March of the third year after the dipole occurs.

[0026] The precursor prediction factor in the prediction model is selected from any one of the anomaly of sea level height in the eastern Indian Ocean, the anomaly of sea surface temperature, the anomaly of subsurface temperature, and the Kelvin wave obtained according to the sea surface wind field. The prediction parameter is and Any of the indices.

[0027] A La Nina event prediction device based on the abnormal sea level height in the eastern Indian Ocean comprises a memory and a processor; the memory is used to store a computer program; the processor is used to implement the La Nina event prediction method based on the abnormal sea level height in the eastern Indian Ocean when executing the computer program.

[0028] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for predicting La Niña events based on the anomaly of sea level height in the eastern Indian Ocean is implemented.

[0029] The present invention has the following beneficial effects and advantages:

[0030] 1. The present invention utilizes the "ocean channel" mechanism between the Indo-Pacific Oceans and uses the abnormal sea level height in the eastern Indian Ocean during the Indian Ocean Dipole as a precursor forecasting factor, which can break through the "spring barrier" and successfully predict the outbreak of the La Niña event one year in advance.

[0031] 2. Based on the sea level anomaly in the eastern Indian Ocean, it is possible to predict the La Nina phenomenon that will occur in the following year for all dipoles. This is far better than the forecast results of international climate forecasting departments that do not consider the Indian Ocean signal.

[0032] 3. The advantage of this invention is that it grasps the key dynamic process of the Indo-Pacific "ocean channel" mechanism - the propagation of Kelvin waves, which can overcome the "spring barrier" and successfully predict the outbreak and intensity of La Niña events one year in advance with a very high success rate. We use the anomaly of the sea level height in the eastern Indian Ocean in autumn to predict the autumn of the following year. The index, with a correlation of 0.99, breaks through the spring barrier of ENSO forecast for this type of La Niña. The persistence of the index itself shows that the correlation coefficient of ENSO forecast across spring is below 0.2. The forecast results produced by the present invention are much more accurate than the maximum predictability of the spring forecast obstacle.

[0033] 4. The technical solution of the present invention is very simple. It uses a linear regression model, which is completely consistent with the dynamic mechanism of Indo-Pacific Kelvin wave propagation. Its high efficiency and high accuracy can easily extend the prediction results to other fields such as society or economics. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is the regression model diagram;

[0035] Figure 2 This is the forecast result diagram;

[0036] Figure 3 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0037] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0038] The present invention relates to a linear regression forecasting model based on the marker parameters of the abnormally rising Kelvin waves in the Indian Ocean during the Indian Ocean Dipole as precursor factors, which can overcome the "spring barrier" of ENSO forecasting and successfully forecast the start-up process and outbreak intensity of La Niña events one year in advance.

[0039] The precursor factors in the model include the characteristic dynamic parameters of the abnormal Kelvin wave in the East Indian Ocean during the dipole period, such as the sea level anomaly (including dynamic height anomaly) or the ocean surface or subsurface temperature anomaly, and the near-equatorial (5°S-5°N) sea level anomaly (including dynamic height anomaly) or subsurface sea temperature anomaly in the Indonesian Sea or the western Pacific (west of the International Date Line). Since the above Kelvin waves are mainly driven by abnormal wind fields, the relevant precursor factors also include the East Indian Ocean Kelvin waves obtained by calculating the Indian Ocean wind field.

[0040] The marine atmospheric historical data include marine atmospheric data from 1993 to 2021. The marine atmospheric historical data include marine historical data and atmospheric historical data;

[0041] The ocean historical data include: sea level anomaly or ocean surface temperature anomaly or sub-surface temperature anomaly;

[0042] The ocean historical data include, but are not limited to, those obtained through satellite observation, Argo profiling buoy observation, oceanographic survey ship observation, or climate reanalysis or prediction based on existing ocean historical data sets;

[0043] Atmospheric historical data include: offshore wind fields;

[0044] The atmospheric historical data include, but are not limited to, data obtained through weather stations, satellite observations, climate reanalysis or prediction based on existing sea-atmosphere historical data sets.

[0045] In this embodiment, the ocean historical data is obtained through satellite altimeter observations, and the atmospheric near-sea surface wind field historical data is obtained through ERA5 reanalysis data.

[0046] When the interannual anomaly of the sea surface height in the eastern Indian Ocean in autumn observed by satellite altimeters is used for regression and prediction, the average value in the region (0, 10°S), (90°E-110°E) is taken as the precursor factor.

[0047] The forecast parameters are mainly for the sea surface temperature anomaly in the cold tongue area of ​​the eastern Pacific Ocean. The index is representative and can be extended to other representative indexes, including 1, 2, 3, 4 and other indexes. The forecast parameters also include forecast variables extending from the cold tongue sea surface temperature anomaly, such as atmospheric near-sea surface temperature, or other highly relevant physical, social and economic parameters in other regions or fields. The forecast accuracy of the index is the main reference indicator to evaluate the forecast effect.

[0048] The linear regression model established by the present invention is used to carry out statistical forecasting of the La Nina event in the year following the dipole, focusing on the amplitude of the 12-month lagging forecast, and the forecast time is from June of the second year of the dipole to March of the third year. This model can predict the occurrence, development and extinction process of the La Nina event 7-15 months in advance, and its effectiveness can reach the spring after the La Nina peak, thus breaking through the "spring forecast barrier" and predicting the outbreak of the La Nina event one year in advance. The lag time can be set artificially, but the successful forecast of the La Nina event by crossing the "spring barrier" is the main assessment indicator.

[0049] The regression model is based on the lagged correlation between the precursory predictor and the sea surface temperature anomaly in the cold tongue area of ​​the eastern Pacific Ocean. The correlation relationship may introduce other variables, but must include one of the above-mentioned precursory predictor types. The correlation relationship may be a linear correlation coefficient, or a partial correlation coefficient or a linear (including multiple) regression relationship or other similar statistical relationships. The forecast model has a high reliability in predicting La Nina events at a lead time of about 11-12 months. It also includes physical and non-physical parameters that have a high causal relationship or high correlation with the predicted La Nina events on a longer time scale.

[0050] A forecast model for expressing the relationship between precursor predictors and forecast parameters is constructed based on historical ocean and atmospheric data, including the following steps:

[0051] Obtain sea level height from historical ocean and atmospheric data (ocean surface and subsurface temperatures, Kelvin waves calculated from sea surface wind fields);

[0052] right The sea temperature index is fitted to obtain the relationship between the precursory prediction factor and the prediction parameter, which is used as the prediction model.

[0053] The forecast model is as follows: predt =a+b*x

[0054] Among them, b is the regression coefficient, a is the constant term, x is the average precursor prediction factor in the autumn (September-November) of the year when the dipole occurs, Y predt is the forecast parameter for each month after the dipole occurs, and the time can be from June of the second year to March of the third year after the dipole occurs. Here, the sea level anomaly in the eastern Indian Ocean is selected as the precursor forecast factor, and the regression parameter is The SST index, regression coefficient b and constant term a are shown in Table 1. Among them, the model has the best prediction effect for October and November of the following year of the dipole, with a correlation coefficient of 0.99.

[0055] Table 1. Sea level anomaly in the eastern Indian Ocean in autumn of the year of the dipole event (yr(0)) and from June of the second year (yr(+1)) to March of the third year (yr(+2)) The linear fit regression coefficient b and constant term a of the exponential.

[0056] Table 1

[0057]

[0058] Figure 1 The autumn sea level anomaly in the eastern Indian Ocean and the one-year-lagged eastern Pacific Regression model between index anomalies. It shows that the sea level anomaly in the eastern Indian Ocean is used as a precursor factor to successfully predict a strong La Nina event one year in advance. Where yr(0) represents the autumn of the year when the dipole occurs, and yr(+1) represents the second year when the dipole occurs. The straight line is the fitting line, and the hollow dots are the forecasts for 2020 in the autumn of 2019 using the retrospective model of the present invention. The index is abnormal (the predicted value for 2020 in the figure), and the result is highly consistent with the observed value in 2019. Figure 1 As shown, the La Niña event prediction model based on the eastern Indian Ocean sea level anomaly involved in the present invention uses the interannual anomaly of the eastern Indian Ocean sea level in autumn observed by satellite altimeters as a precursor factor to carry out statistical forecasting of the La Niña event in the year following the dipole.

[0059] Figure 2 To use the regression model to predict the eastern Pacific in different months after the dipole The index anomaly shows that the development process of the strong La Niña event predicted by the regression model is in good agreement with the observation. The forecast can break through the "spring barrier" and successfully predict the outbreak of the La Niña event one year in advance. Among them, yr(+1) represents the second year after the dipole occurs, and yr(+2) represents the third year after the dipole occurs. The solid line is the observation (Obs in the legend) The dotted line is the predicted value (Pre in the legend), and the gray column represents the root mean square error. Figure 2 As shown in the figure, the workflow of the entire model is:

[0060] When the Indian Ocean dipole event occurs, we closely monitor the changes in the sea level in the eastern Indian Ocean. In the autumn when the dipole matures, we take the average value of the anomaly of the sea level of the satellite altimeter in the region (0, 10°S) and (90°E-110°E) as the precursor factor, input it into the regression model, and calculate the forecast for 7-15 months later. Then, short-term climate forecasts of other relevant parameters one year in advance are issued based on this.

[0061] like Figure 3 As shown, the present invention provides a statistical forecasting method for ultra-long-term forecasting of a type of La Niña event 6 to 15 months in advance. Through satellite altimeter observations, the average sea level height anomaly in the eastern Indian Ocean in autumn is calculated, and the maximum value of the abnormally rising Kelvin wave in the Indian Ocean during the mature period of the Indian Ocean dipole is captured. This is used as a precursor forecasting factor. According to the action route of the Indo-Pacific "ocean channel" mechanism, that is, the Indian Ocean Kelvin wave passes through the Indonesian Sea to the western Pacific Ocean, and propagates eastward to the cold tongue area of ​​the eastern Pacific Ocean, causing significant tropical Pacific ocean-air coupling variations, and then causing the dynamic process of the outbreak of the La Niña event in the following year, a linear regression statistical model is established. Analysis of observational data shows that this statistical regression model can overcome the spring barrier and predict the outbreak of La Niña events more than one year in advance, and has a very high The forecast accuracy of the index is 0.99, which is consistent with the relevant research results of the Indo-Pacific "ocean channel" mechanism. The forecast model and method have the advantages of simple model, small calculation amount, high accuracy, and long prediction time lead time, avoiding the complex model calculation and dynamic process characterization of La Niña event forecast; it can also realize automatic forecast and early warning through computer programming. If the index is less than -0.5℃ for three consecutive months, it will warn that a La Niña event is about to occur; otherwise, it is predicted that a La Niña event will not occur.

[0062] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of embodiments in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0063] The present invention is described with reference to the method flow according to the embodiment of the present application. It should be understood that the flow in the method flow can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for implementing the functions specified in one or more flows.

[0064] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device that implements the functions specified in one or more processes.

[0065] These computer program instructions may also be loaded onto a computer or other programmable data processing device so that a series of operational steps are performed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes.

[0066] The above is only the most simplified implementation of the present invention. Without departing from the basic premise of the inventive concept of the present invention, several deformations and improved regression models can be made, and these changes all fall within the protection scope of the present invention.

Claims

1. A La Nina event prediction method based on the anomaly of sea level in the eastern Indian Ocean, characterized in that: The following steps are involved: A forecast model is constructed based on historical ocean and atmospheric data to represent the relationship between precursor predictors and forecast parameters; According to the precursor forecast factors at the set time, the forecast model is used to predict the La Nina event in the following year; The precursor predictor is one of the following types: Abnormal sea level height or abnormal ocean surface temperature or abnormal subsurface temperature in the eastern Indian Ocean; The near-equatorial sea level anomaly or the surface sea temperature anomaly or the subsurface sea temperature anomaly in the Indonesian Sea; The near-equatorial sea level anomaly or surface sea temperature anomaly or subsurface sea temperature anomaly in the western Pacific; East Indian Ocean Kelvin waves obtained from the Indian Ocean atmospheric wind field; The forecast parameter is the Niño index for the sea surface temperature anomaly in the cold tongue area of ​​the eastern Pacific Ocean; The forecast parameter is any one of the Niño3.4, Niño 1, Niño 2, Niño 3, and Niño 4 indices; The method of constructing a forecast model for representing the relationship between precursor forecast factors and forecast parameters based on the historical marine atmospheric data comprises the following steps: Obtain any one of the sea level height anomaly, sea surface temperature anomaly, subsurface temperature anomaly in the ocean historical data, and Kelvin wave obtained according to the sea surface wind field in the atmospheric historical data as a precursor prediction factor; Fitting the precursor prediction factors and prediction parameters to obtain the relationship between the precursor prediction factors and prediction parameters as a prediction model; The forecast model is as follows: Y predt =a+b*x Among them, b is the regression coefficient, a is the constant term, x is the average precursor factor in the autumn of the year when the dipole occurs, and Y predt These are the forecast parameters for each month after the dipole occurs, from June of the second year to March of the third year after the dipole occurs.

2. The method for predicting La Nina events based on the anomaly of sea level in the eastern Indian Ocean according to claim 1, characterized in that: The ocean atmosphere historical data includes ocean historical data and atmospheric historical data; The ocean historical data include: sea level anomaly or ocean surface temperature anomaly or sub-surface temperature anomaly; The ocean historical data includes: obtained through satellite observation, Argo profiling buoy observation, ocean survey ship observation, or climate reanalysis or prediction based on existing ocean historical data sets; Atmospheric historical data include: offshore wind fields; The atmospheric historical data include those obtained through meteorological stations, satellite observations, climate reanalysis or prediction based on existing sea-atmosphere historical data sets.

3. The method for predicting La Nina events based on the anomaly of sea level in the eastern Indian Ocean according to claim 1, characterized in that: The precursor prediction factor in the prediction model is selected from any one of the anomaly of sea level height in the eastern Indian Ocean, the anomaly of ocean surface temperature, the anomaly of subsurface temperature, and the Kelvin wave obtained according to the sea surface wind field, and the prediction parameter is any one of the Niño3.4, Niño 1, Niño 2, Niño 3 and Niño 4 indices.

4. A La Nina event prediction device based on the anomaly of sea level height in the eastern Indian Ocean, characterized in that: It comprises a memory and a processor; the memory is used to store a computer program; the processor is used to implement the La Niña event prediction method based on the anomaly of sea level height in the eastern Indian Ocean as described in any one of claims 1 to 3 when executing the computer program.

5. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, the method for predicting La Niña events based on the anomaly of sea level height in the eastern Indian Ocean as described in any one of claims 1 to 3 is implemented.

Images