Method and system for quantifying and predicting influence of planet orbit change on earth atmospheric water vapor change

Abstract

The invention discloses a method and a system for quantifying and predicting the influence of planet orbit change on earth atmospheric water vapor change. The method comprises the following steps: acquiring an atmospheric water vapor data set and an auxiliary data set; analyzing and determining a plurality of planet orbit changes influencing the atmospheric water vapor change based on the atmospheric water vapor data set and the auxiliary data set; the plurality of planet orbit changes comprise earth rotation, revolution of the moon around the earth, revolution of the earth around the sun and orbit changes of other planets; constructing a dynamic quantitative model, and obtaining influence quantitative values of a plurality of planet orbit changes on atmospheric water vapor changes based on the dynamic quantitative model; determining a contribution value of each planet orbit change based on the influence quantized value; and constructing and training an atmospheric water vapor prediction model, and predicting the atmospheric water vapor content through the atmospheric water vapor prediction model. The method provides a new perspective and theoretical support for constructing a more precise climate model in the future, and is crucial for deeply understanding the complexity of an earth climate system.

Description

Technical Field

[0001] The present invention belongs to the technical field of atmospheric water vapor prediction, and in particular relates to a method and system for quantifying and predicting the impact of planetary orbit changes on changes in earth's atmospheric water vapor. Background Art

[0002] In addition to global warming and changes in the Earth's interior, changes in the planet's orbit are also considered an important factor in changes in atmospheric water vapor content. Gribbin, Zachos and other scholars pointed out that although these orbital changes have relatively small effects, they play a key role in regulating the transfer of heat from the sun to the Earth.

[0003] In recent decades, scientists have gradually begun to understand the connection between planetary orbital changes and the climate system by analyzing historical climate records, geological evidence and numerical simulations. Researchers Cinco, Hinnov and others found that key parameters of the Earth's orbit, such as eccentricity, inclination and ecliptic inclination, change over time, resulting in periodic changes in the distance between the Earth and the Sun and the total amount of solar radiation received. These changes directly affect the seasonal changes and global energy balance of the Earth, which in turn affect the changes and distribution of water vapor in the atmosphere. Therefore, changes in planetary orbits cause seasonal and interannual changes in atmospheric water vapor content on a global scale. This change may trigger a nonlinear response of the climate system, thereby affecting the global climate pattern and precipitation distribution. On the other hand, changes in planetary orbits may also weaken or enhance human-induced warming and water vapor rise in a short period of time, leading to oscillations in the Earth's atmospheric water vapor. Although existing studies have emphasized the important role of celestial dynamics in regulating the Earth's climate system, the specific impact of planetary orbital changes on Earth's water vapor changes still requires further research and quantitative analysis. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes a method and system for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere, so as to solve the problems existing in the above-mentioned prior art.

[0005] To achieve the above-mentioned purpose, the present invention provides a method for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere, comprising:

[0006] Obtain atmospheric water vapor datasets and auxiliary datasets;

[0007] Determine several planetary orbit changes that affect atmospheric water vapor changes based on the atmospheric water vapor data set and the auxiliary data set; the several planetary orbit changes include the rotation of the earth, the revolution of the moon around the earth, the revolution of the earth around the sun and other planetary orbit changes;

[0008] A dynamic quantitative model is constructed, and based on the dynamic quantitative model, quantitative values ​​of the impact of several planetary orbit changes on atmospheric water vapor changes are obtained; and a contribution value of each planetary orbit change is determined based on the quantitative impact values;

[0009] An atmospheric water vapor prediction model is constructed and trained using the atmospheric water vapor dataset, and the atmospheric water vapor content is predicted using the atmospheric water vapor prediction model.

[0010] Optionally, the auxiliary data set includes daily data on day length, daily data on atmospheric carbon dioxide concentration, moon phase information, and daily data on the distance between the sun and the earth, the solar declination angle, and the lunar declination angle.

[0011] Optionally, the dynamic quantitative model includes a dynamic quantitative model of the earth's rotation, a dynamic quantitative model of the moon's revolution around the earth, a dynamic quantitative model of the earth's revolution around the sun, and dynamic quantitative models of other planetary orbit changes.

[0012] Optionally, the process of constructing the dynamic quantitative model of the earth's rotation includes:

[0013] The initial dynamic quantization formula of the earth's rotation is constructed as follows:

[0014]

[0015] The initial earth rotation dynamic quantization formula is transformed to obtain the following formula:

[0016]

[0017] Among them, t l represents the tth day of the mth day of the nth sidereal month l hours, |E| represents the absolute value of the difference in water vapor in the Earth's atmosphere between two consecutive hours, and e is an exponential term. is the optional phase shift of Y1(t), the width of ω1 in the cosine term is determined by the number of hours to complete the periodic motion;

[0018] The continuous constraints are constructed as follows:

[0019]

[0020] Substituting equation (2) and equation (3) into equation (4), and substituting their derivatives into equation (5), we can obtain:

[0021]

[0022] Substituting equations (6) and (7) into equation (3), we can obtain the dynamic quantitative model of the Earth's rotation over multiple days.

[0023] Optionally, the process of constructing a dynamic quantitative model of the moon's revolution around the earth includes:

[0024] A basic dynamic quantification formula for the moon's revolution around the earth is constructed, and a monthly dynamic quantification formula for the moon's revolution around the earth is obtained based on the basic dynamic quantification formula for the moon's revolution around the earth. A weight function driven by atmospheric water vapor data is constructed, and the weight function is fit through historical data, and a dynamic quantification model for the moon's revolution around the earth is obtained based on the fitted weight function and the monthly dynamic quantification formula for the moon's revolution around the earth.

[0025] Optionally, the process of constructing the dynamic quantitative model of the Earth's revolution around the Sun includes:

[0026] The initial dynamic quantitative formula of the Earth's revolution around the Sun is constructed as follows:

[0027]

[0028] where Y3(t) represents the effect of the Earth's revolution on the Earth's atmospheric water vapor, t represents the hour of the year, |S| represents the absolute value of the difference between the monthly averages in adjacent sidereal months, c and s are the exponential term and the optional phase shift, respectively, and the width of ω3 is determined by the number of hours to complete the periodic motion;

[0029] The initial dynamic quantization formula of the earth's revolution around the sun is transformed to obtain the following formula:

[0030]

[0031] In order to make Y3(t) continuous at the intersection of two annual cycles t0, the following two constraints need to be met:

[0032]

[0033] Substituting equations (9) and (10) into equation (11) and substituting their derivatives into equation (18), we can obtain:

[0034]

[0035] Substituting equations (13) and (14) into equation (8), we can obtain the continuous dynamic quantitative model of the Earth's revolution around the Sun.

[0036] Optionally, the dynamic quantitative model of the orbital change of other planets is as follows:

[0037]

[0038] Among them, X(t) represents the impact of other planetary orbit changes on the change of Earth's atmospheric water vapor, and M(t) is a continuous function of the atmospheric water vapor value at time t. represents the change of atmospheric water vapor from year t to year t+1; Y1(t) represents the effect of the Earth's rotation at time t, which changes with time t; Y2(t) represents the effect of the Moon's revolution, which changes with time t; Y3(t) represents the effect of the Earth's revolution, which changes with time t; represents a sliding window, n represents the size of the time window, and j represents the time offset.

[0039] Optionally, the impact quantification values ​​of several planetary orbit changes are summed to obtain the total impact quantification value, and the corresponding contribution value is determined according to the proportion of the impact quantification value of each planetary orbit change in the total impact quantification value.

[0040] Optionally, the process of building and training an atmospheric water vapor prediction model includes:

[0041] An atmospheric water vapor prediction model with a four-layer deep learning network structure is constructed, and the Bayesian optimization method is used to select training data to train the atmospheric water vapor prediction model.

[0042] The present invention also provides a system for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere, comprising:

[0043] Dataset acquisition module, used to acquire atmospheric water vapor dataset and auxiliary dataset;

[0044] A factor analysis module, used for analyzing several planetary orbit changes that affect atmospheric water vapor changes based on the atmospheric water vapor data set and the auxiliary data set;

[0045] A dynamic quantification module is used to construct a dynamic quantification model, and obtain the quantitative values ​​of the impact of several planetary orbit changes on atmospheric water vapor changes based on the dynamic quantification model, and determine the contribution value of each planetary orbit change based on the quantitative impact value;

[0046] The atmospheric water vapor prediction module is used to construct an atmospheric water vapor prediction model and train it through the atmospheric water vapor data set, and predict the atmospheric water vapor content through the trained model.

[0047] Compared with the prior art, the present invention has the following advantages and technical effects:

[0048] The present invention analyzes the impact of different planetary orbit changes on the changes in the earth's atmospheric water vapor through long-term series data analysis. Based on the analysis of the impact of different planetary orbit changes on the earth's atmospheric water vapor in each time scale, a dynamic quantification model based on adjacent differences is constructed to extract the impact of these orbital changes and eliminate the interference of the earth's internal variability, thereby quantifying the impact of different planetary orbits on the changes in the earth's atmospheric water vapor. Finally, an LSTM deep learning model is constructed to predict the changes in the earth's atmospheric water vapor content in the next ten years. The results show that the atmospheric water vapor will show a slow upward trend in the future. The research results not only provide a new perspective for understanding the mechanism of climate change on the earth, but also provide a valuable reference for the study of climate evolution on other planets. The present invention discovered the close connection between the earth's water vapor cycle and the changes in the planetary orbit, providing a new perspective and theoretical support for the construction of a more precise climate model in the future, which is crucial for a deep understanding of the complexity of the earth's climate system, and also provides a scientific basis for the formulation of long-term climate change response strategies and environmental policies. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0050] Figure 1 The method technology roadmap of the embodiment of the present invention is as follows: (A) is a plane diagram of the orbital changes of the main planets in the solar system; (B) is a diagram of the changes in the water vapor in the Earth's atmosphere dominated by the orbital changes of different planets in different time scales; (C) is a quantitative model for the impact of the orbital changes of different planets; and (D) is an LSTM deep learning prediction model and analysis;

[0051] Figure 2 A schematic diagram of the average contribution of different planetary orbit changes to the change of water vapor in the Earth's atmosphere from 1959 to 2022 according to an embodiment of the present invention;

[0052] Figure 3 The kernel density diagram of the influence of different planetary orbit changes on the change of water vapor in the earth's atmosphere according to the embodiment of the present invention; (A) is the influence of the earth's rotation, (B) is the influence of the moon's revolution, (C) is the influence of the earth's revolution, and (D) is the influence of other planetary orbit changes;

[0053] Figure 4 The prediction results of the water vapor in the earth's atmosphere under different time scale models of the embodiments of the present invention are as follows: (A) is the hourly average prediction of the water vapor in the earth's atmosphere, (B) is the daily average prediction of the water vapor in the earth's atmosphere, (C) is the monthly average prediction of the water vapor in the earth's atmosphere, and (D) is the annual average prediction of the water vapor in the earth's atmosphere;

[0054] Figure 5The LSTM deep learning model of the embodiment of the present invention predicts the interannual atmospheric water vapor changes of the earth. (A) is a comparison of the annual average predicted values ​​of the earth's atmospheric water vapor with the original annual average data of atmospheric water vapor at different time scales, and (B) is a prediction of the change of the annual average value of the earth's atmospheric water vapor under the influence of the planet's orbit from 2023 to 2032;

[0055] Figure 6 This is a schematic diagram of the LSTM neural network of an embodiment of the present invention;

[0056] Figure 7 Schematic diagram of the change of model training effect of an embodiment of the present invention, (A) is the change of the loss function of the network training set and the test set over time, and (B) is the relationship between LSTM accuracy and time step. DETAILED DESCRIPTION

[0057] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0058] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0059] Embodiment 1

[0060] like Figure 1 As shown, this embodiment provides a method for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere, including:

[0061] In order to deeply evaluate the impact of planetary orbit changes on the changes in water vapor in the Earth's atmosphere, this embodiment constructs a theoretical framework to explore how different planetary orbit changes affect the changes in water vapor in the Earth's atmosphere. In this study, the Earth is regarded as an integral system, and the natural variability of its internal climate system is considered to be a response to planetary orbital forcing. Therefore, the main atmospheric water vapor changes in the Earth's interior can be regarded as one of the intrinsic manifestations caused by planetary orbital forcing, and the influence of the Earth's internal variability (such as CO2, etc.) can be eliminated as much as possible through neighboring subtraction. Therefore, by constructing a model based on the neighboring difference (differential) method, the internal influence of the Earth can be approximately ignored, and only the impact of different planetary orbital changes on the changes in water vapor in the Earth's atmosphere can be analyzed. Figure 1 A technical roadmap for analyzing, quantifying, and predicting the effects of planetary orbit changes on changes in Earth's atmospheric water vapor is presented. Figure 1A in the figure is a planar diagram of the orbital motion of the planets in the solar system. The figure takes the sun as the center and details the rotation of the earth, the revolution of the moon around the earth, the revolution of the earth around the sun, and the orbits of other planets around the sun. The changes in the orbits of these different planets affect the changes in the water vapor in the earth's atmosphere on different time scales. Figure 1 In B, the influence of the Earth's rotation on the daily variation of the Earth's atmospheric water vapor, the variation of the Earth's atmospheric water vapor caused by the Moon's revolution around the Earth in the sidereal month cycle, and the seasonal variation caused by the Earth's revolution are analyzed based on data of different time scales. In addition, the orbital changes of other planets in the solar system also have an impact on the variation of the Earth's atmospheric water vapor, but due to the long distance between these planets and the Earth, the impact is relatively small, mainly manifested in the local details of the superposition of orbital signals of different planets. The relative motion of other planets and the Earth is relatively complex and has a long period, resulting in the diversity of local details of the variation of the Earth's atmospheric water vapor. The periodic influence of the large orbital changes of other planets is mainly manifested on the interannual and longer time scales, so their influence is taken together for comprehensive consideration. These factors work together to affect the variation of the Earth's atmospheric water vapor. By comprehensively analyzing the atmospheric water vapor data of different time scales, the influence of different planetary orbital changes on the variation of the Earth's atmospheric water vapor is studied.

[0062] Figure 1In C, when describing the effect of the Earth's rotation on the change of the Earth's atmospheric water vapor, since the overall change of the Earth's interior is relatively stable within two consecutive hours, and the effect of the Earth's revolution, the Moon's revolution and the orbital changes of other planets on the change of the Earth's atmospheric water vapor is longer than the effect of the Earth's rotation, the change in a short time (such as two hours) is relatively small. Based on this, the effect of the Earth's internal variability (such as CO2 changes, etc.) and the orbits of other planets with longer periods are approximately eliminated by subtracting the values ​​of the Earth's atmospheric water vapor between adjacent hours, so as to approximately obtain the effect of the Earth's rotation, which is expressed by the harmonic function Y1(t). Since the Earth's rotation period is 24 hours, it completes its own rotation period in one day, and the Earth's revolution around the sun in the same day only accounts for 1 / 365 of its annual revolution period. In addition, the orbital changes of other planets have a very long period of influence on the change of the Earth's atmospheric water vapor. Therefore, when describing the influence of the moon's revolution on the change of water vapor in the earth's atmosphere, the subtraction of the data of adjacent days in the sidereal month cycle can approximately eliminate the influence of the earth's rotation (completed its rotation cycle), the earth's revolution around the sun, the orbits of other planets, and the influence of the earth's interior, so as to approximately obtain the influence of the moon's revolution on the change of water vapor in the earth's atmosphere, which is expressed by the irregular change function Y2(t). Similarly, in each sidereal month, since the earth's rotation and the moon's revolution have completed their respective orbital cycles, and the influence cycle of other planets' orbits is very long relative to the earth's revolution cycle, the monthly average values ​​of adjacent months between sidereal months are subtracted to eliminate the influence of the earth's rotation, the moon's revolution, the orbital changes of other planets, and the interior of the earth, and then approximately obtain the influence of the earth's revolution on the change of water vapor in the earth's atmosphere, which is expressed by the function Y3(t).

[0063] Although the orbital changes of other planets (such as Mercury, Mars, Jupiter, etc.) also affect the changes in the water vapor in the Earth's atmosphere, their influence period is relatively longer, and the main period is manifested on the interannual and longer time scales. Therefore, when analyzing the impact of the orbital changes of other planets on the changes in the water vapor in the Earth's atmosphere, considering that the Earth's rotation, the Moon's revolution and the Earth's revolution have completed their respective orbital cycles on the interannual time scale, by subtracting the annual average values ​​between adjacent interannual periods, the influence of the Earth's rotation, the Moon's revolution, the Earth's revolution around the Sun and the internal factors of the Earth are approximately eliminated, thereby obtaining the impact of the orbital changes of other planets. Finally, the function X(t) is used to represent the superimposed influence of the changes in the orbits of other planets in the solar system except the Earth, the Moon and the Sun on the changes in the water vapor in the Earth's atmosphere. In order to deeply understand the impact of the orbital changes of these planets on the changes in the water vapor in the Earth's atmosphere, the function X(t) is superimposed and applied to the curve Y1(t)-Y3(t) for analysis, and a series of mathematical models based on neighboring differences (differential method) are constructed. The purpose of these models is to quantify the specific contribution of the orbital changes of different planets to the changes in the water vapor in the Earth's atmosphere on different time scales. With the help of the quantitative model proposed in this study, it is possible to evaluate the extent to which the orbital changes of different planets affect the changes in atmospheric water vapor. Figure 1 The long short-term memory (LSTM) deep learning model is used in the D to construct prediction models of different time scales. After adjusting the corresponding parameter settings for training and verification comparative analysis, the future changes in the earth's atmospheric water vapor are further predicted.

[0064] 1. Data:

[0065] In this embodiment, the fifth generation global atmospheric reanalysis dataset (ERA5) provided by the European Centre for Medium-Range Weather Forecasts (ECMWF) is used to obtain atmospheric water vapor data. ERA5 combines model data with observational data from all over the world to create a global dataset. The time resolution of this dataset is 1 hour and the spatial resolution is 0.25°×0.25°. Using data from January 1959 to December 2022, the impact of different planetary orbit changes on the changes in the Earth's atmospheric water vapor on different time scales is analyzed.

[0066] In addition, additional auxiliary data were collected to enhance the depth and breadth of the analysis. These include daily data on length of day (LOD) provided by the International Earth Rotation and Reference System Service Prediction Center (IERS) of the U.S. Naval Observatory, and daily data on atmospheric carbon dioxide concentration from 1959 to 2022 archived by the Scripps Institution of Oceanography. At the same time, moon phase information was obtained from the Fourmilab website in Switzerland, and daily data on the Sun-Earth distance, solar declination, and lunar declination were obtained from the ephemeris of NASA's Jet Propulsion Laboratory (JPL). These auxiliary data sets provide comprehensive background information and greatly improve the accuracy of the analysis of water vapor trends in the Earth's atmosphere.

[0067] 2. The impact of orbital changes of different planets on changes in water vapor in the Earth's atmosphere

[0068] (1) The impact of the Earth’s rotation on changes in atmospheric water vapor

[0069] During the Earth's rotation, it was observed that the Earth's atmospheric water vapor exhibits different changes within a day. Through the analysis of hourly changes within consecutive days and long-term data, it was found that although the Earth's orbital position, the position of the Moon's revolution around the Earth, and the orbital positions of other planets will change in two consecutive days, these changes have relatively little impact on the Earth's atmospheric water vapor within the adjacent 24-hour cycle. Therefore, it can be concluded that the hourly changes in the Earth's atmospheric water vapor within the daily cycle are mainly due to the influence of the Earth's rotation.

[0070] (2) The impact of the moon’s revolution on the changes in water vapor in the Earth’s atmosphere

[0071] The periodic changes in the force of the moon's gravity on the earth's atmospheric water vapor also cause atmospheric tides of 27.3 days and 13.6 days, and have an impact on weather changes. The moon's tidal effects and their periodic changes are well-known phenomena and can be clearly observed in ocean records. The lunar tidal cycle is caused by the moon's gravitational effect on the earth, which regulates ocean currents to a certain extent, thereby directly and indirectly affecting the spatial distribution of water vapor in the earth's atmosphere. The asymmetric alternating changes in the moon's gravity on the solid earth and the ocean lead to periodic oscillations in crustal stress. This periodic oscillation may cause local natural disasters and affect local or even global water vapor changes. In addition, the change in water vapor in the earth's atmosphere is also affected by the reflection and infrared radiation of the moon. Therefore, the impact of the moon's revolution on the change of water vapor in the earth's atmosphere cannot be ignored. The revolution of the moon around the earth is one of the important reasons affecting the daily change of water vapor in the earth's atmosphere.

[0072] (3) The impact of the Earth’s revolution on changes in water vapor in the Earth’s atmosphere

[0073] In the movement of the Earth, the revolution of the Earth around the Sun is one of the important factors that affect the change of water vapor in the Earth's atmosphere. The revolution of the Earth causes periodic changes in the Sun relative to the equatorial plane. When the Earth is in different positions on its elliptical orbit around the Sun, seasonal changes are formed. Different seasons and regions receive different solar radiation, which directly affects the atmospheric temperature and the evaporation and precipitation of water vapor. The impact of the moon's revolution on the daily changes of water vapor in the Earth's atmosphere has been analyzed in Method (II) (2), while the gravitational perturbations of other planetary orbits have produced complex superposition effects on the distribution of water vapor in the Earth's atmosphere on a longer time scale.

[0074] (4) The impact of other planetary orbit changes on the Earth's atmospheric water vapor changes

[0075] From the above analysis, it can be seen that if there are only the sun, the earth and the moon in the solar system, the changes in atmospheric water vapor on the earth will be relatively regular. However, in addition to the rotation of the earth, the revolution of the moon around the earth and the revolution of the earth around the sun, which will affect the changes in the atmospheric water vapor on the earth, other planets in the solar system will also affect the changes in the atmospheric water vapor on the earth. The orbital changes of these planets have a long period, and they move at the same time as the earth, so that the effects of the orbital changes of these planets are superimposed on the effects of the rotation of the earth, the revolution of the moon and the revolution of the earth. Since each planet has its own unique periodicity, this causes the atmospheric water vapor on the earth to show different characteristics on different time scales. In addition, the relationship between the earth, the moon and the sun is relatively stable, and their movement patterns can be regarded as a relatively simple system, that is, one revolves around another. In contrast, the relative motion of other planets and the earth is more complicated, and the relative motion between planets further increases the complexity of the impact of their orbital changes on the atmospheric water vapor on the earth. This orbital change of relative motion makes it difficult to describe its periodic change law with a mathematical model in a short period of time. The superimposed effects of the orbital changes of these different planets make the local changes in the atmospheric water vapor on the earth diverse and uncertain, thus leading to the diversity of climate change.

[0076] In order to more deeply analyze the impact of other planetary orbits on the superimposed signal of water vapor in the Earth's atmosphere, Fourier transform analysis and multiple time scale (MTM) methods are used to analyze long time series data on the interannual scale, aiming to more comprehensively understand the long-term impact of other planetary orbit changes on the Earth's atmospheric water vapor changes. The analysis used ERA5 reanalysis data covering 1959 to 2022, and developed spectral domain methods and harmonic approximation models to identify all periodic oscillations in historical time series and reconstruct the impact of planetary orbit changes on the Earth's atmospheric water vapor. Using frequency domain methods, the period, amplitude, and phase in the time series are calculated to extract related oscillation signals. Spectral analysis methods help to identify periodic components in time series, thereby constructing a harmonic approximation model. The set of oscillation signals of the reconstructed model It can be expressed using the continuous time formula:

[0077]

[0078] Among them, f, A and Represent the frequency, amplitude and phase of the sine wave, respectively, while t represents the year and M is the number of significant peaks extracted. When applying the FFT and MTM methods for frequency analysis, the U values ​​are set to 0.013 and 0.0094, respectively. The planetary orbit change periods obtained by these two spectral analysis methods are listed in Tables 1 and 2, respectively. The results show that since 1959, the global atmospheric water vapor has shown multiple fluctuation periods, including 3.5 years, 8.25 years, 9.15 years, 9.85 years, 13.4 years, 15.25 years, 21.3 years, 32 years, and 62 years. These periods coincide with the astronomical periods of different planetary orbit combinations. The specific analysis results are shown in Tables 1 and 2. This analysis shows that the superposition of different planetary orbits affects the changes in the Earth's atmospheric water vapor. These periodic fluctuations cause changes on an interannual scale in the climate system, affecting key climate elements such as precipitation patterns, temperature and atmospheric circulation.

[0079] Table 1

[0080]

[0081] Table 2

[0082]

[0083]

[0084] (III) Quantifying the impact of orbital changes of different planets on changes in water vapor in the Earth's atmosphere

[0085] (1) Build a quantitative model

[0086] This embodiment studies and analyzes the impact of different planetary orbit changes on the change of water vapor in the Earth's atmosphere at different time scales. In order to quantify the impact of different planetary orbit changes and eliminate the interference of the Earth's internal variability, a dynamic quantitative model based on adjacent difference (differential method) is established to quantify the impact of different planetary orbit changes on the change of water vapor in the Earth's atmosphere.

[0087] F(t)=Y1(t)+Y2(t)+Y3(t)+X(t) (2)

[0088] The function F(t) represents the influence of all the orbital changes of the transformed planets on the change of water vapor in the Earth's atmosphere at time t. F(t) is determined by the sum of the four functions Y1(t), Y2(t), Y3(t), and X(t). Y1(t), Y2(t), and Y3(t) respectively represent the influence of the Earth's rotation, the Moon's revolution, and the Earth's revolution on the change of water vapor in the Earth's atmosphere at time t. X(t) represents the influence of the orbital changes of other planets in the solar system except the Sun, the Moon, and the Earth on the change of water vapor on the Earth at time t.

[0089] During the rotation of the Earth, the daily variation of atmospheric water vapor is periodically affected. For the rotation of the Earth, by subtracting the atmospheric water vapor data within two consecutive hours, the influence of the internal variability of the Earth and the orbital changes of other planets can be theoretically eliminated. Within two consecutive hours, the overall change of the Earth's interior is relatively small, and the movement trajectories of the moon's revolution, the Earth's revolution, and the orbits of other planets are also very small relative to the Earth's rotation. From the perspective of calculus, it is believed that their influence on the atmospheric water vapor of the Earth can be approximately eliminated by subtracting adjacent values. In addition, according to the analysis in (II) (1), the influence of the Earth's rotation on the change of atmospheric water vapor is significant. The atmospheric water vapor change curves for consecutive days show that the daily change trend is similar, and the main factor affecting the change of atmospheric water vapor between two consecutive hours is the rotation of the Earth. Therefore, by calculating the absolute value of the difference in the average atmospheric water vapor values ​​of adjacent hours in a day, the degree of influence of the Earth's rotation on the atmospheric water vapor of the Earth on that day can be approximately expressed (see formula (3)). Using the atmospheric water vapor content values ​​of all hours in the study time range (1959-2022) as the data basis, the influence of the Earth's rotation on the change of atmospheric water vapor can be quantified.

[0090]

[0091] where t l represents the tth day of the mth day of the nth sidereal month l Hours. |E| represents the absolute value of the difference in water vapor in the Earth's atmosphere within two consecutive hours, and e is an exponential term. is an optional phase shift of Y1(t). To improve the convergence of Y1(t), the number of hours of motion needs to be known a priori. The width of ω1 in the cosine term is determined by the number of hours to complete the periodic motion.

[0092] In view of the physical continuity of the Earth's water vapor, it is assumed that the function of the Earth's rotation on the change of the Earth's atmospheric water vapor (i.e., Y1(t)) is continuous in all cases, including the intersection time between two day and night cycles t l0 Here, the continuity of the Earth's rotation Y1(t) function is evaluated. Y1(t) on the mth and m+1th day can be expressed as:

[0093]

[0094] To make Y1(t) cycle in two days l0 The intersection points of are continuous at all times and must satisfy the following two constraints:

[0095]

[0096] Substituting equation (4) and equation (5) into equation (6), and substituting their derivatives into equation (7), we can obtain:

[0097]

[0098] From formula (8), we can get |E m |Represented by |E m+1 |, formula (9) converts e m+1 With e m Similarly, substituting equations (8) and (9) into equation (5), we get the multi-day continuous Y1(t) model. On the mth day (m ≥ 2), the total number of free parameters of the function Y1(t) is m+1 (i.e., the first day is e, arrive ). In this study, the Levenberg-Marquardt minimization algorithm is used to estimate the free parameters, and a general scheme for global optimization is proposed. Except for the first day, the parameters |E| and e for each day are calculated by formula (8) and formula (9), respectively.

[0099] In the quantitative model of the influence of the moon's revolution around the earth on the change of water vapor in the earth's atmosphere, the difference in water vapor in the earth's atmosphere between two adjacent days is used to approximately quantify the influence of the moon's revolution. From the analysis of (II) (2), it is known that the daily change trend of water vapor in the earth's atmosphere is mainly affected by the rotation of the earth. The daily change trend is similar because the earth rotates once a day. Although the daily change trend of water vapor in the earth's atmosphere is the same, the specific values ​​(vertical coordinates) of the daily fluctuation of water vapor in the earth's atmosphere are different within a sidereal month cycle. The difference in these numerical fluctuations is mainly due to the influence of the moon's revolution around the earth on the change of water vapor in the earth's atmosphere. In addition, since the earth completes its rotation cycle every day, and the period of the earth's revolution around the sun in a day is 1 / 365, and the changes in the orbits of other planets are smaller, the difference between two adjacent days can approximately eliminate the effects of the earth's rotation, the earth's revolution around the sun, the changes in the orbits of other planets and the earth's interior (see formula (10)). In quantifying the influence of the moon's revolution on the change of water vapor in the earth's atmosphere, the water vapor values ​​of the earth's atmosphere on all days from 1959 to 2022 are used as the data basis.

[0100]

[0101] Where A2 is the amplitude, c is a constant, I2 is the exponential term, n is the nth sidereal month, ts represents the tth of the nth sidereal month s Theoretically, the moon's revolution in the tth day of the nth sidereal month in the ith year s The influence of the sky on the earth's water vapor can be expressed as n (t s ). However, through the actual analysis of atmospheric water vapor data, it is found that the influence of the moon on the earth's atmospheric water vapor is irregular, so its specific period cannot be expressed by the harmonic function formula (10). Therefore, the indefinite integral of the difference in the earth's atmospheric water vapor for two consecutive days is calculated by formula (11) to quantify the influence of the moon's revolution around the earth on the earth's atmospheric water vapor.

[0102]

[0103] Among them, Y2(t) represents the impact of the moon's revolution on the earth's atmospheric water vapor. t1 and t2 are the t1 and t2 days of the nth sidereal month respectively. W(t) represents the atmospheric water vapor content on the tth day. f(t) is a weight function driven by atmospheric water vapor data, which is used to adjust the impact weights of different time periods. a represents the impact weights in different time periods. In actual calculations, the function f(t) is determined by analyzing and fitting historical data, and the impact weights are determined according to the changes in different time periods.

[0104] According to the analysis of (II) (3), the daily variation trend of the water vapor in the Earth's atmosphere is mainly affected by the rotation of the Earth. In a sidereal month cycle, the specific numerical fluctuation of the water vapor in the Earth's atmosphere is mainly caused by the influence of the moon's revolution around the Earth on the change of atmospheric water vapor. However, the numerical fluctuations of the water vapor in the Earth's atmosphere in two adjacent sidereal months are not the same, and there is no fixed pattern. This is caused by the influence of the Earth's revolution on the change of the water vapor in the Earth's atmosphere. In two adjacent months, the Earth has completed 54 / 365 of its revolution cycle, which makes its influence more significant. In order to quantify the influence of the Earth's revolution on the change of the water vapor in the Earth's atmosphere, the monthly average values ​​of the water vapor in the Earth's atmosphere in adjacent sidereal month cycles are subtracted from each other (see formula (14)) to approximately eliminate the influence of the Earth's interior and the influence of the Earth's rotation and the moon's revolution around the Earth (because they complete the same cycle). In addition, since the influence of the orbits of other planets in a month is relatively small, the influence of the changes in the orbits of other planets can also be approximately eliminated by subtracting the monthly average values ​​in adjacent sidereal month cycles. The monthly average data of Earth's atmospheric water vapor during all sidereal monthly cycles from 1959 to 2022 are used to quantify the impact of the Earth's revolution on the changes in Earth's atmospheric water vapor.

[0105]

[0106] where Y3(t) represents the effect of the Earth's revolution on the Earth's atmospheric water vapor, t represents the hour of the year, and |S| represents the absolute value of the difference between the monthly averages of adjacent sidereal months. c and s are the exponential term and the optional phase shift, respectively. The width of ω3 is determined by the number of hours it takes to complete its periodic motion.

[0107] In order to make the function of the effect of the Earth's revolution on the change of the Earth's atmospheric water vapor (i.e., Y3(t)) continuous in all cases, including the intersection time t0 between two consecutive years. The functional continuity of the Earth's revolution Y3(t) is evaluated, and Y3(t) in the i-th year and the i+1-th year can be expressed as:

[0108]

[0109] In order to make Y3(t) continuous at the intersection of two annual cycles t0, the following two constraints need to be met:

[0110]

[0111] Substituting equations (15) and (16) into equation (17) and substituting their derivatives into equation (18), we can obtain:

[0112]

[0113] Substituting equations (19) and (20) into equation (14) yields a continuous Y3(t) model. The total number of free parameters of function Y3(t) in the i-th year (i≥2) is i+1 (i.e., the first year is s, c1 to c i ), these free parameters are estimated by applying the Levenberg-Marquardt minimization algorithm. |S| and s for each year (except the first year) are calculated by equations (19) and (20), respectively.

[0114] X(t) represents the impact of orbital changes of other planets in the solar system (except the sun, the earth and the moon) on the changes in the water vapor in the earth's atmosphere. The impact of the orbital changes of other planets on the changes in the water vapor in the earth's atmosphere can be determined by a physical function that depends on the orbit of the planet and its position. However, it is difficult to establish such a function because even if the movement of each planet can be strictly calculated, the physical mechanism that links the movement of a single planet with the water vapor in the earth's atmosphere still needs further study. In addition, due to the complexity of the physical process that links the water vapor in the earth's atmosphere to the entire planetary system, the impact of these other planets is considered together here. Although the trends of the changes in the water vapor in the earth's atmosphere are basically similar every day of the year, different fluctuation curves occasionally appear. At the same time, in the three-dimensional diagram of the daily changes in the water vapor in the earth's atmosphere from 1959 to 2022, some irregular fluctuations can also be observed locally, which are due to the impact of the orbital changes of other planets on the changes in the water vapor in the earth's atmosphere. From the analysis in (II) (4), it can be seen that the periodic fluctuation phenomenon of the water vapor in the earth's atmosphere caused by the orbital changes of different planets is mainly manifested on a longer interannual scale and is irregular. Therefore, by subtracting the annual average values ​​of atmospheric water vapor in two consecutive years, we can approximately eliminate the effects of the Earth's internal variability, the Earth's rotation, the Moon's revolution around the Earth, and the Earth's revolution around the Sun, thereby obtaining the effects of other orbital changes, as shown in formula (21).

[0115]

[0116] Where X(t) represents the impact of other planetary orbit changes on the changes in the Earth's atmospheric water vapor, and M(t) is a continuous function of the atmospheric water vapor value at time t. represents the change in atmospheric water vapor from year t to year t+1. Y1(t) represents the effect of the Earth's rotation at time t, which varies with time t. Y2(t) represents the effect of the Moon's revolution, which varies with time t. Y3(t) represents the effect of the Earth's revolution, which varies with time t. Represents a sliding window, which is used to smoothly eliminate the effects of the Earth's rotation, the Earth's revolution, and the Moon's revolution. n represents the size of the time window, and j represents the time offset.

[0117] This embodiment not only quantifies the impact of different planetary orbit changes on the change of Earth's atmospheric water vapor, but also defines the contribution ratio of different planetary orbit changes to the change of Earth's atmospheric water vapor, which is expressed as the ratio of the impact of the change of Earth's atmospheric water vapor caused by the orbital changes of each planet to the sum of the impacts of all planetary orbital changes.

[0118]

[0119] Among them, formula (22) represents the contribution of the Earth's rotation to the change of water vapor in the Earth's atmosphere in the i-th year; formula (23) represents the contribution of the Moon's revolution to the change of water vapor in the Earth's atmosphere in the i-th year; formula (24) represents the contribution of the Earth's revolution to the change of water vapor in the Earth's atmosphere in the i-th year; formula (25) represents the contribution of the orbital changes of other planets to the change of water vapor in the Earth's atmosphere in the i-th year.

[0120] (2) Specific quantification

[0121] In order to quantify the impact of the Earth's rotation on the change of the Earth's atmospheric water vapor within two consecutive hours, the function Y1(t) in formula (3) is used to evaluate the effect of the Earth's rotation on the change of the Earth's atmospheric water vapor. By adopting the Levenberg-Marquardt algorithm, the first-order derivative model is successfully fitted to the time series of the difference in the Earth's atmospheric water vapor for two consecutive hours, and the unknown parameters of the function are solved. At the same time, the average contribution of the Earth's rotation to the change of the Earth's atmospheric water vapor during the study period is calculated using formula (22). The calculation and analysis show that the average contribution of the Earth's rotation to the change of the Earth's atmospheric water vapor during the study period is about 4% (see Figure 2 and Figure 3 A).

[0122] Similarly, the function Y2(t) in formula (12) is used to quantify the effect of the moon's revolution around the earth on the change of water vapor in the earth's atmosphere, and formula (23) is used to calculate the average contribution of the moon's revolution to the change of water vapor in the earth's atmosphere during the study period. The model is used to calculate the change of atmospheric water vapor caused by the moon's revolution around the earth in two consecutive days. The results show that the average contribution of the moon's revolution around the earth to the change of water vapor in the earth's atmosphere is about 10% (see Figure 2 and Figure 3 B).

[0123] When quantifying the effect of the Earth's revolution on the change of water vapor in the Earth's atmosphere, the monthly average atmospheric water vapor difference in adjacent sidereal months was analyzed. By fitting the first-order derivative model of the function Y3(t) in formula (14) to the monthly average atmospheric water vapor difference time series of the sidereal month cycle, the unknown parameters of the relevant function were obtained. Subsequently, formula (24) was used to calculate the average contribution of the Earth's revolution to the change of water vapor in the Earth's atmosphere during the study period. The results show that the average contribution of the Earth's revolution to the change of water vapor in the Earth's atmosphere is about 71% (see Figure 2 and Figure 3 C).

[0124] In addition, the function Z(t) in formula (21) is used to measure the impact of other planetary orbit changes on the change of Earth's atmospheric water vapor, and the average contribution of its impact during the study period 1959-2022 is calculated using formula (25). In-depth analysis shows that the average contribution of other planetary orbit changes to the change of Earth's atmospheric water vapor is about 15% (see Figure 2 and Figure 3 D).

[0125] (IV) Predicting the impact of future orbital changes of different planets on Earth’s water vapor changes

[0126] LSTM (Long Short-Term Memory) is a classic deep learning model (RNN). Its uniqueness lies in its ability to predict future trends based on the network state of the input of the current time step and the output of the previous time step, thereby automatically learning and capturing the long-term dependencies and periodic changes between samples in long time series. Due to its excellent predictive ability, LSTM has been widely used in many fields such as speech recognition, handwriting recognition, and time series prediction. The structure of LSTM consists of a repeated chain of modules, which updates the state through a recurrent network, and can effectively capture the long-term dependencies and periodicity in the sequence. From the analysis in (ii) and (iii), it can be seen that the orbital changes of different planets have a periodic impact on the water vapor in the earth's atmosphere, so this long-term change feature is very consistent with the simulation and prediction capabilities of LSTM.

[0127] The LSTM repetitive unit consists of four key parts, including the memory unit C, the forget gate f t , input gate i t and output gate O t These four parts interact and work together to enable the LSTM network to better process and remember information in the sequence. For more information about its working principle, see Figure 6 .exist Figure 6 In the second repeating module, use p t As the input vector of LSTM. Among them, the gate f t ,i t and O t and memory unit C t ′ are all determined by (p t ,h t-1 ) control. Then, use f t and i t To update the cell state C t . Door O t Determines how much information is propagated to time step t+1. These gates consist of a Sigmoid-type fully connected deep learning layer and a point-by-point multiplication operation. The working mechanism of the gate and the information flow can be expressed as:

[0128] f t =σ(W f ·[h t-1 ,p t ]+b f ) (26)

[0129] i t =σ(W i ·[h t-1 ,p t ]+b i ) (27)

[0130] C' t =tanh(W c ·[h t-1 ,p t ]+b c ) (28)

[0131] C t =f t ⊙C t-1 +i t ⊙C' t (29)

[0132] O t =σ(W o ·[h t-1, p t ]+b o ) (30)

[0133] h t =O t ⊙tanh(C t ) (31)

[0134] Among them, the transformation σ from input to i, f, and O is a Sigmoid function, ⊙ represents point-by-point multiplication, and W and b represent weights and biases, respectively. h is a hidden state, and tanh(l) is the hyperbolic tangent function, which operates piecewise on each element of the vector l. These equations describe the input calculation, activation function, and information flow of the gates in LSTM. This structure allows the LSTM network to effectively learn and capture long-term dependencies in the input sequence.

[0135] In this embodiment, the atmospheric water vapor data of four time scales from 1959 to 2022 are divided into hourly, daily average, monthly average and annual average data. Specifically, the data from 1959 to 2000 are used for model training, and the data from 2001 to 2022 are used for model testing. This data partitioning strategy ensures that the model is fully tested and optimized in both the training and verification stages. In order to improve the performance of the long short-term memory (LSTM) model, the Bayesian optimization method is applied, and 10% of the data is randomly selected from the training data set for basic experiments. The number of LSTM layers was extensively tested, ranging from 1 to 5 layers, and combined experiments were performed with 1 to 3 dense layers. Finally, a 4-layer deep learning network structure containing two layers of LSTM and two layers of dense layers was determined, in which the number of neurons in each layer was 64, 32, 64 and 25, respectively. In order to improve the training stability and convergence effect of the network, batch normalization is used for each hidden layer, and the batch size is set to 32, and the final configuration achieves the best effect.

[0136] The mean square error (MSE) was selected as the loss function to measure the deviation between the model prediction value and the actual value. After testing a variety of gradient descent algorithms including Adam, RMSprop, AdaGrad, Nesterovs, SGD, and Adadelta, the Adam optimizer was finally selected, which performed well on multiple tasks and data sets. Although this study used data of four different time scales (hourly, daily average, monthly average, and annual average), in order to maintain the consistency and comparability of the model, all data were trained and tested using the same LSTM model parameter settings. Therefore, Figure 7 The best results after comprehensive evaluation of four types of data under this model are shown. With the increase of training rounds, the loss function of the model gradually decreases and finally converges to below 0.05, indicating that the constructed LSTM model performs well in predicting the changing trend of water vapor content in the earth's atmosphere. In addition, Figure 7 The coefficient of determination (R 2 ), mean absolute error (MAE) and root mean square error (RMSE) with time step. Through the analysis of these indicators, the optimal time step is determined to be 58, that is, the first 58 input sequences are used to predict the 59th input sequence.

[0137] Finally, the optimized LSTM deep learning model successfully predicted the future trend of the water vapor content in the Earth's atmosphere. The detailed results of the training and prediction are as follows: Figure 4 , Figure 5 As shown in Table 3, the effectiveness and accuracy of the model are further verified.

[0138] Combining the mathematical model constructed in this study (see Methods section (III)) and historical data on atmospheric water vapor, this example aims to determine the contribution of orbital changes of different planets to changes in Earth's atmospheric water vapor. Figure 2 The contribution rates of the Earth's rotation, the Moon's revolution, the Earth's revolution and other planetary orbit changes to the Earth's atmospheric water vapor changes on different time scales calculated based on mathematical models from 1959 to 2022 are 4%, 10%, 71% and 15% respectively. The impact of different planetary orbit changes on the Earth's atmospheric water vapor changes is from large to small: the Earth's revolution, other planetary orbit changes, the Moon's revolution, and the Earth's rotation. Figure 3 The figure shows the impact of the orbital changes of different planets on the change of water vapor in the Earth's atmosphere from 1959 to 2022. Specifically, Figures A to D represent the Earth's rotation, the Moon's revolution, the Earth's revolution, and the orbits of other planets. In the figure, the change of water vapor in the Earth's atmosphere is marked with different colors according to its nuclear density. During this period, the Earth's rotation ( Figure 3 A) and the lunar revolution ( Figure 3 B in the Earth's revolution ( Figure 3 C) and other planetary orbits ( Figure 3 The range of water vapor changes in the Earth's atmosphere caused by D) is 0.01 kg / m 2 -0.04Kg / m 2 , 0.04Kg / m 2 -0.10Kg / m 2 , 0.3Kg / m 2 -0.6Kg / m 2 and 0.1Kg / m 2 -0.5Kg / m 2 The average values ​​are 0.02Kg / m 2 , 0.06Kg / m 2 , 0.5Kg / m 2 and 0.1Kg / m 2 .

[0139] This embodiment uses a long short-term memory network (LSTM) deep learning model to predict the future trend of changes in the water vapor content in the earth's atmosphere. In order to ensure the accuracy of the prediction, the performance of the LSTM (long short-term memory network) prediction model based on data of different time scales is compared. When dividing the data set, the hourly atmospheric water vapor data, daily average atmospheric water vapor data, monthly average atmospheric water vapor data and annual average atmospheric water vapor data from 1959 to 2022 were selected respectively, and the data from 1959 to 2000 were used for model training, and the data from 2000 to 2022 were used for model testing. In the evaluation of the prediction effect of the LSTM model, special attention was paid to the mean square error (MSE), mean absolute error (MAE) and R 2 Statistics and other indicators. Figure 4Table 1 and Table 2 show the results of testing and verifying the LSTM deep learning prediction model based on atmospheric water vapor data at four different time scales (hourly data, daily average data, monthly average data, and annual average data).

[0140] from Figure 4 It can be seen that the performance of the prediction model varies at different time scales. 2 , etc., the prediction performance of hourly, daily, and monthly scale data is relatively excellent. 2 All of them are over 0.95, while the annual scale model is relatively weak. This reason can be explained by the laws of planetary orbits. Changes on the hourly scale are mainly dominated by the laws of the Earth's rotation, changes on the daily scale are mainly dominated by the laws of the Moon's revolution, changes on the monthly scale are mainly dominated by the laws of the Earth's revolution, and changes on the interannual scale are mainly dominated by the laws of the orbits of other planets. Since the Earth's rotation, the Moon's revolution, and the Earth's revolution all have their own obvious laws, the prediction effect is very good. The reason why the interannual scale model performs slightly worse is that the orbital periods of other planets (Venus, Mars, Mercury, Saturn, and Jupiter, etc.) on the interannual scale are relatively long, resulting in the superposition of multiple influences, so that no significant periodic laws are formed in the short term. The laws of planetary orbits are consistent with the results of actual data analysis.

[0141] Table 3

[0142]

[0143] Finally, in Figure 5 In A, the data of four time scales are uniformly applied to the LSTM deep learning model to predict the annual average value, and the results are compared with the original annual average value of atmospheric water vapor. It is found that the prediction effect of the monthly scale data model is the best, followed by the daily scale, hourly scale and annual scale data models. Considering the influence of the planetary orbit and the prediction accuracy of the model, the LSTM model based on monthly scale data is selected to predict the interannual variation trend of the water vapor content in the Earth's atmosphere in the future. Figure 5 Figure B shows the predicted results of the annual average water vapor value of the Earth's atmosphere from 2023 to 2032. In the figure, the red part on the left represents the annual average water vapor value data from 1959 to 2022, the green part is the predicted value from 2023 to 2032, and the blue curve is the fitting curve of the water vapor data from 1959 to 2032. Figure 5 As can be seen from B in Figure 1, although the water vapor content in the Earth's atmosphere will fluctuate in the next decade, the overall trend and the fitting curve both show a slowly rising trend. This trend is mainly attributed to the continuous impact of the combined effect of changes in the orbits of different planets on atmospheric water vapor.

[0144] Embodiment 2

[0145] This embodiment discloses a system for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere, including:

[0146] Dataset acquisition module, used to acquire atmospheric water vapor dataset and auxiliary dataset;

[0147] A factor analysis module, used for analyzing several planetary orbit changes that affect atmospheric water vapor changes based on the atmospheric water vapor data set and the auxiliary data set;

[0148] A dynamic quantification module is used to construct a dynamic quantification model, and obtain the quantitative values ​​of the impact of several planetary orbit changes on atmospheric water vapor changes based on the dynamic quantification model, and determine the contribution value of each planetary orbit change based on the quantitative impact value;

[0149] The atmospheric water vapor prediction module is used to construct an atmospheric water vapor prediction model and train it through the atmospheric water vapor data set, and predict the atmospheric water vapor content through the trained model.

[0150] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere, characterized in that: The following steps are involved: Obtain atmospheric water vapor datasets and auxiliary datasets; Determine several planetary orbit changes that affect atmospheric water vapor changes based on the atmospheric water vapor data set and the auxiliary data set; the several planetary orbit changes include the rotation of the earth, the revolution of the moon around the earth, the revolution of the earth around the sun and other planetary orbit changes; A dynamic quantitative model is constructed, and based on the dynamic quantitative model, quantitative values ​​of the impact of several planetary orbit changes on atmospheric water vapor changes are obtained; based on the quantitative impact values, the contribution value of each planetary orbit change is determined; An atmospheric water vapor prediction model is constructed and trained using the atmospheric water vapor dataset, and the atmospheric water vapor content is predicted using the atmospheric water vapor prediction model.

2. The method for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere according to claim 1, characterized in that: The auxiliary data set includes daily data of day length, daily data of atmospheric carbon dioxide concentration, moon phase information, and daily data of sun-earth distance, solar declination angle, and lunar declination angle.

3. The method for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere according to claim 1, characterized in that: The dynamic quantitative model includes the dynamic quantitative model of the earth's rotation, the dynamic quantitative model of the moon's revolution around the earth, the dynamic quantitative model of the earth's revolution around the sun and the dynamic quantitative model of other planetary orbit changes.

4. The method for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere according to claim 3, characterized in that: The process of constructing the earth rotation dynamic quantitative model includes: The initial dynamic quantization formula of the earth's rotation is constructed as follows: The initial earth rotation dynamic quantization formula is transformed to obtain the following formula: Among them, t l represents the tth day of the mth day of the nth sidereal month l hours, |E| represents the absolute value of the difference in water vapor in the Earth's atmosphere between two consecutive hours, and e is an exponential term. is the optional phase shift of Y1(t), the width of ω1 in the cosine term is determined by the number of hours to complete the periodic motion; The continuous constraints are constructed as follows: Substituting equation (2) and equation (3) into equation (4), and substituting their derivatives into equation (5), we can obtain: Substituting equations (6) and (7) into equation (3), we can obtain the dynamic quantitative model of the Earth's rotation over multiple days.

5. The method for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere according to claim 3, characterized in that: The process of constructing a dynamic quantitative model of the moon's revolution around the earth includes: A basic dynamic quantification formula for the moon's revolution around the earth is constructed, and a monthly dynamic quantification formula for the moon's revolution around the earth is obtained based on the basic dynamic quantification formula for the moon's revolution around the earth. A weight function driven by atmospheric water vapor data is constructed, and the weight function is fit through historical data, and a dynamic quantification model for the moon's revolution around the earth is obtained based on the fitted weight function and the monthly dynamic quantification formula for the moon's revolution around the earth.

6. The method for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere according to claim 3, characterized in that: The process of constructing the dynamic quantitative model of the Earth's revolution around the Sun includes: The initial dynamic quantitative formula of the Earth's revolution around the Sun is constructed as follows: where Y3(t) represents the effect of the Earth's revolution on the Earth's atmospheric water vapor, t represents the hour of the year, |S| represents the absolute value of the difference between the monthly averages in adjacent sidereal months, c and s are the exponential term and the optional phase shift, respectively, and the width of ω3 is determined by the number of hours to complete the periodic motion; The initial dynamic quantization formula of the earth's revolution around the sun is transformed to obtain the following formula: In order to make Y3(t) continuous at the intersection of two annual cycles t0, the following two constraints need to be met: Substituting equations (9) and (10) into equation (11) and substituting their derivatives into equation (18), we can obtain: Substituting equations (13) and (14) into equation (8), we can obtain the continuous dynamic quantitative model of the Earth's revolution around the Sun.

7. The method for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere according to claim 3, characterized in that: The dynamic quantitative model of other planetary orbit changes is as follows: Among them, X(t) represents the impact of other planetary orbit changes on the change of Earth's atmospheric water vapor, and M(t) is a continuous function of the atmospheric water vapor value at time t. represents the change of atmospheric water vapor from year t to year t+1; Y1(t) represents the effect of the Earth's rotation at time t, which changes with time t; Y2(t) represents the effect of the Moon's revolution, which changes with time t; Y3(t) represents the effect of the Earth's revolution, which changes with time t; represents a sliding window, n represents the size of the time window, and j represents the time offset.

8. The method for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere according to claim 1, characterized in that: The impact quantification values ​​of several planetary orbit changes are summed to obtain the total impact quantification value, and the corresponding contribution value is determined according to the proportion of the impact quantification value of each planetary orbit change in the total impact quantification value.

9. The method for quantifying and predicting the impact of planetary orbit changes on changes in water vapor in the Earth's atmosphere according to claim 1, characterized in that: The process of building and training an atmospheric water vapor prediction model includes: An atmospheric water vapor prediction model with a four-layer deep learning network structure is constructed, and the Bayesian optimization method is used to select training data to train the atmospheric water vapor prediction model.

10. A system for quantifying and predicting the effect of planetary orbit changes on changes in water vapor in the Earth's atmosphere, which implements the method described in any one of claims 1 to 9, characterized in that: Dataset acquisition module, used to acquire atmospheric water vapor dataset and auxiliary dataset; A factor analysis module, used for analyzing several planetary orbit changes that affect atmospheric water vapor changes based on the atmospheric water vapor data set and the auxiliary data set; A dynamic quantification module is used to construct a dynamic quantification model, and obtain the quantitative values ​​of the impact of several planetary orbit changes on atmospheric water vapor changes based on the dynamic quantification model, and determine the contribution value of each planetary orbit change based on the quantitative impact value; The atmospheric water vapor prediction module is used to construct an atmospheric water vapor prediction model and train it through the atmospheric water vapor data set, and predict the atmospheric water vapor content through the trained model.

Images