Rainfall diagnosis method based on multivariable scale decomposition

Abstract

The invention relates to a precipitation diagnosis method based on multivariable scale decomposition. The method comprises the following steps: firstly, obtaining high temporal-spatial resolution reanalysis data lattice point data and precipitation measurement satellite observation data; preprocessing rainfall measurement satellite observation data to generate a rainfall daily change data set; on the basis of the high-temporal-spatial-resolution reanalysis data, water vapor revenue and expenditure items are constructed, and a water vapor revenue and expenditure daily change data set is generated; based on the rainfall daily change data set and the water vapor revenue and expenditure daily change data set, establishing a water vapor revenue and expenditure equation of a daily time scale whole-layer integral containing a nonlinear term; finally, time scale decomposition is carried out on multivariable related to the key water vapor income and expenditure nonlinear term, contributions of the water vapor income and expenditure term acted by different time scale variables are calculated, and relative contributions of the different time scale variables in the water vapor income and expenditure nonlinear term are calculated; according to the method, water vapor income and expenditure equation diagnosis and a multivariable time scale decomposition method are combined, and the relative importance of influence of a multi-scale climate mode on rainfall daily change can be quantitatively revealed.

Description

TECHNICAL FIELD

[0001] The present application relates to a precipitation diagnostic method based on multivariate time scale decomposition, belonging to the technical field of meteorological forecasting. BACKGROUND

[0002] Precise precipitation diagnosis is the core link of meteorological forecasting, water resource management and disaster prevention and reduction. With the development of social economy, various industries have put forward unprecedentedly high requirements for fine and quantitative precipitation information. However, precipitation is a complex physical process with significant multi-scale characteristics, from large-scale water vapor transport, weather scale system lifting, to mesoscale convective triggering and microphysical processes, which jointly determine the falling area, intensity and type of precipitation.

[0003] Traditional diagnostic techniques are relatively single, such as empirical analysis based on stations or radar echoes, linear diagnosis of individual physical quantities, which are difficult to clearly separate the contributions of different scale systems and physical processes, leading to unclear explanation of the formation mechanism of key weather such as extreme precipitation and persistent rainstorm, thus becoming a bottleneck restricting the improvement of forecasting ability. The existing technical methods often regard precipitation as a whole phenomenon, or only focus on a single dominant scale, lacking effective tools to clearly and quantitatively separate and correlate the "organizing force" of large-scale circulation background, the "carrier" role of mesoscale convective system and the "feedback" effect of cumulus scale in a precipitation event.

[0004] Traditional precipitation diagnostic methods have inherent limitations in dealing with increasingly complex precipitation events, especially extreme heavy precipitation. These methods rely on single variable and isolated analysis paradigm, which is difficult to effectively reveal how atmospheric dynamics, thermodynamics and water vapor processes interact; at the same time, they lack the ability to analyze multi-scale nonlinear interactions, and cannot quantitatively separate the respective contributions of weather scale system and mesoscale convection to precipitation, leading to a blurred understanding of precipitation mechanism, especially the triggering mechanism of extreme events.

[0005] In addition, although traditional research has confirmed that large-scale climate modes are the main driving force of seasonal and interannual precipitation variability, there is a significant cognitive gap and methodological deficiency in revealing how they specifically affect precipitation diurnal variation, a key link. This deficiency is due to a fundamental "scale gap": climate modes are slow-varying signals of planetary and intraseasonal scales, while precipitation diurnal variation is dominated by local circulation, topography and cumulus convection, etc. Traditional diagnostic methods cannot effectively "bridge" this gap and cannot quantitatively answer. SUMMARY

[0006] The present application is to solve the problems in the prior art, and provides a precipitation diagnostic method combining water vapor budget equation diagnosis and multivariate time scale decomposition.

[0007] In order to achieve the above object, the technical scheme of the present application is as follows: a precipitation diagnosis method based on multivariate scale decomposition, comprising the following steps: S1, obtaining high spatiotemporal resolution reanalysis data grid data and precipitation measurement satellite observation data; S2, preprocessing the precipitation measurement satellite observation data to generate a precipitation diurnal variation data set; S3, constructing a water vapor budget term based on high spatiotemporal resolution reanalysis data to generate a water vapor budget diurnal variation data set; S4, establishing a water vapor budget equation of diurnal time scale integral containing nonlinear terms based on the precipitation diurnal variation data set and the water vapor budget diurnal variation data set; S5, performing time scale decomposition on the multivariate involved in the nonlinear term to calculate the contribution of the water vapor budget term of different time scale variables, and obtaining key water vapor budget nonlinear terms; S6, calculating the relative contribution of different time scale variables in the water vapor budget nonlinear term to obtain key contribution variables.

[0008] The high spatiotemporal resolution reanalysis data grid data includes horizontal wind field and specific humidity data, the spatial range of which is global, the spatial resolution is 1.25°x1.25°, and the temporal resolution is 3h; Evaporation data, the spatial range of which is global, the spatial resolution is 1.25°x1.25°, and the temporal resolution is 1h; The precipitation measurement satellite observation data includes global precipitation measurement multi-satellite precipitation joint inversion data set, the spatial range of which is global 60°S-60°N, the spatial resolution is 0.1°x0.1°, and the temporal resolution is 3h.

[0009] In the step S2, the precipitation measurement satellite observation data is preprocessed by converting the recording mode of the precipitation measurement satellite observation data from universal time to local time.

[0010] In the step S3, the method for constructing the water vapor budget term includes: calculating the water vapor convergence divergence term in the vertical integral of the water vapor budget equation and the horizontal water vapor advection term , the formulas are as follows: , ; wherein, is a horizontal gradient operator, is specific humidity, is a horizontal wind vector field, represents the integral of the whole layer from 1000 hPa to 100 hPa; the universal time is represented by and characterized by local time and , generating water vapor budget diurnal variation data set.

[0011] The water vapor budget equation is: ; wherein, is the precipitation rate, is the evaporation term, is the vertically integrated water vapor convergence divergence term, is the vertically integrated horizontal water vapor advection term; and are nonlinear terms.

[0012] In the step S5, the method for calculating the contribution of the water vapor budget term of different time scale variables includes: The nonlinear term , The variables and q used for calculation are recorded as , and is decomposed into time scales as follows: wherein, , and represent components of different time scales, respectively.

[0013] In the step S6, the method for calculating the relative contribution of different variables in the water vapor budget nonlinear term is: Based on the total differential relationship , the relative contribution is estimated, wherein, represents the difference between different stages, , represent different variables of different time scales, respectively.

[0014] The present application has the following beneficial effects: The present application combines water vapor budget equation diagnosis and multivariate time scale decomposition method, can more accurately calculate the contribution of different time scale variables to the nonlinear term of the water vapor budget equation, and the relative contribution of different physical quantities in the water vapor budget nonlinear term, refine different time scale physical processes, and quantitatively reveal the relative importance of multi-scale climate modalities affecting precipitation diurnal variation. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 is a flowchart of the present application Figure 2 is the water vapor budget diagnosis result of the oceanic continental region (10°S-6°N, 95°-150°E) affected by annual cycle in the embodiment of the present application; Figure 3The water vapor budget diagnosis results of the MJO-affected oceanic continent region (10°S-6°N, 95°-150°E) of the embodiment of the present application. DETAILED DESCRIPTION

[0016] The present application will be described in detail below with reference to the accompanying drawings and specific embodiments. EMBODIMENT

[0017] As Figure 1 shown, the present embodiment provides a precipitation diagnosis method based on multivariate scale decomposition, which is to obtain the key physical processes affecting precipitation and the contributions of different time scale variables involved, and the method specifically includes the following steps: S1, obtaining high spatiotemporal resolution reanalysis data grid point data and precipitation measurement satellite observation data; The obtained data includes reanalysis data grid point data and precipitation measurement satellite observation data. The reanalysis data grid point data includes: horizontal wind field and specific humidity, with a global spatial range, a spatial resolution of 1.25°x1.25°, and a temporal resolution of 3h; evaporation, with a global spatial range, a spatial resolution of 1.25°x1.25°, and a temporal resolution of 1h; optionally, the data can be the fifth generation global weather and climate reanalysis data ERA5 from the European Centre for Medium-Range Weather Forecasts (ECMWF). The precipitation measurement satellite observation data includes: Global Precipitation Measurement (GPM) multi-satellite precipitation joint inversion (IMERG) dataset, with a global spatial range of 60°S-60°N, a spatial resolution of 0.1°x0.1°, and a temporal resolution of 3h.

[0018] S2, preprocessing the precipitation measurement satellite observation data to generate precipitation diurnal variation dataset; Convert the GPM precipitation measurement satellite observation data record mode from universal time (UTC) to local time (LST): LST varies with longitude, and based on the longitudinal distance of each grid point from 0°, the universal time of each grid point in the study area is converted to local time: ; wherein, is the longitude, and UTC and LST are the time of universal time and local time, respectively. To convert a set of data from UTC to LST, the data must be converted on each longitude column of the latitude grid.

[0019] Converting 8 UTC time data to 00:00, 03:00, …, 21:00 LST requires creating 8 new time grid data. First, calculate the UTC time corresponding to each longitude from LST. If the time point of LST coincides with a certain time point in UTC, use the data of this time point in UTC at this longitude. Otherwise, linearly interpolate between the two closest time points. For example, if 60°E is 06:00 LST, then the UTC time is 02:00. Therefore, for this longitude, the data can be obtained by linearly interpolating between 00:00 and 03:00 data: .

[0020] S3, based on high spatiotemporal resolution reanalysis data, constructing water vapor budget terms to generate water vapor budget diurnal variation data sets; The specific method for constructing water vapor budget terms includes: Calculating the vertical integration of water vapor convergence and divergence terms in the water vapor budget equation And horizontal water vapor advection terms , the formula is as follows: , ; Wherein, is the horizontal gradient operator, is the specific humidity, is the horizontal wind vector field, representing the vertical integration from 1000 hPa to 100 hPa. According to the method described in step S2, convert UTC-represented and into LST-represented and to generate water vapor budget diurnal variation data sets.

[0021] S4, based on the precipitation diurnal variation data set and the water vapor budget diurnal variation data set, establishing a water vapor budget equation with nonlinear terms for daily time scale integration; The specific water vapor budget diagnostic method includes: Using the daily time scale integration of the water vapor budget equation: ; Wherein, the asterisk represents the diurnal variation, is the precipitation rate, and the left side of the equation is the diurnal variation of the precipitation rate, represents the difference between different stages, such as different seasons, different phases of ENSO or different phases of MJO. is the surface evaporation rate. Since the time tendency term of specific humidity on the right side of the above equation is much smaller than the precipitation rate, it can be ignored, so the simplified water vapor budget equation is obtained: ; is the precipitation rate, is the evaporation term, is the vertically integrated water vapor convergence divergence term, is the vertically integrated horizontal water vapor advection term. It is shown that the diurnal variation of precipitation can be balanced by the diurnal variation of the evaporation and the diurnal variation of the divergence of the atmospheric water vapor flux.

[0022] S5, time scale decomposition of the multi-variables involved in the key water vapor budget nonlinear term is performed to calculate the contribution of the water vapor budget term of different time scale variables; that is, the contribution of the key water vapor budget nonlinear term obtained after time scale decomposition of the physical quantity is calculated, and the key water vapor budget term with larger contribution is obtained.

[0023] The specific method for calculating the contribution of the water vapor budget term of different time scale variables includes: decomposing the nonlinear term , , , , , , , , , , , , , , ,

[0024] , , , ,

[0025] , , ,

[0026] S6, the relative contribution of different time scale variables in the water vapor budget nonlinear term is calculated; that is, the nonlinear term is further split into the relative contribution of different variables through the total differential relationship, and the key contribution variable is obtained.

[0027] Specifically, it includes: based on the total differential relationship , estimate the relative contribution. Where represents the difference between different stages, , respectively represent different variables of different time scales. Therefore and The change in the product can be divided into three parts, the first two parts are respectively attributed to the change of different variables of different time scales, and the third part is the remaining part related to the covariance of different variables of different time scales.

[0028] The relative contribution estimated by the total differential relationship is the main method of analyzing the influence of each factor on the overall change in a multivariate system, so the above total differential equation can be understood as: and The change in the product can be divided into three parts, the first part is attributed to the change of , the second part is attributed to the change of , and the third part is the remaining part related to the covariance of and .

[0029] The method of the present embodiment generates precipitation diurnal variation diagnostic results by constructing a water vapor balance equation as a diagnostic model, inputting satellite observation and reanalysis data into the diagnostic model, specifically: obtain the required precipitation diagnostic three-hourly satellite observation and reanalysis data, and input into the determined water vapor balance diagnostic equation, calculate the most important contribution factor in the regulation of multi-scale climate modes on precipitation diurnal variation, that is, the key contribution variable.

[0030] Figure 2 The water vapor balance diagnostic results of the oceanic continent region (10°S-6°N, 95°-150°E) affected by the annual cycle using the method of the present application, (a)-(c) are respectively the difference between the precipitation budget item diurnal range of the 1000-100 hPa vertical integration in the Northern Hemisphere winter (DJF) and summer (JJA); The time scale decomposition result of the horizontal water vapor convergence term in the precipitation budget; The relative contribution of the wind change of the daily time scale and the seasonal mean state water vapor change to the second item on the right side of (b). It can be seen that the precipitation diurnal variation amplitude is larger in DJF than in JJA, which is mainly attributed to the vertical integration of the horizontal water vapor convergence term ( ) (the key water vapor budget nonlinear term), and the evaporation and horizontal water vapor advection both contribute very little, and the contribution of evaporation is negative; By time scale decomposition of the vertical integration of the horizontal water vapor convergence term, it can be found that the main contribution term is the daily time scale wind convergence seasonal mean water vapor ( ), and the contributions of the other terms are very limited; and it is found by total differential relationship calculation that for the daily time scale wind convergence seasonal mean water vapor ( The main contributor to the difference is the diurnal wind (sea-land breeze, ), which accounts for about 85%, while the difference in the seasonal mean water vapor contributes relatively little.

[0031] Figure 3 For the diagnosis results of the water vapor budget in the MJO-affected western oceanic continental region (8°S-6°N, 95°-120°E) using the method of the present application, (a)-(c) are, respectively, the difference in the 1000-100 hPa vertically integrated precipitation budget term between MJO events P2-P3 and P6-P7; the time scale decomposition results of the horizontal water vapor convergence term in the precipitation budget; and the relative contributions of the diurnal wind variation and the low-frequency water vapor variation to the second term on the right side of (b). It can be seen that P2-P3 has a larger diurnal precipitation difference than P6-P7, which is mainly contributed by the vertically integrated horizontal water vapor convergence term ( ), while the contributions of evaporation and horizontal water vapor advection are negative. Further decomposition of the vertically integrated horizontal water vapor convergence term shows that the dominant term is the diurnal wind horizontal convergence low-frequency water vapor term ( ) (key water vapor budget nonlinear term), while the contributions of other terms are very small; and through full differential relationship calculation, it is found that for the change of the dominant term (diurnal wind convergence low-frequency water vapor term), the change of the diurnal wind (sea-land breeze, ) is the main contributor, accounting for about 92%. The low-frequency water vapor also contributes, but is limited compared to the diurnal wind.

[0032] The present application uses global precipitation measurement satellite observation data and high spatiotemporal resolution reanalysis data, combines water vapor budget equation diagnosis and multivariate time scale decomposition method, can more accurately calculate the contributions of different time scale variables to the water vapor budget equation nonlinear term, and the relative contributions of different physical quantities in the water vapor budget nonlinear term, and refine different time scale physical processes, and provides strong support for quantitatively revealing the relative importance of multi-scale climate modes affecting precipitation diurnal variation. It can effectively supplement the technical method system of atmospheric multi-scale interaction research, thereby improving the accuracy and fineness of diagnosis.

[0033] The technical solutions of the present application are not limited to the above-mentioned embodiments, and any technical solutions obtained by equivalent replacement fall within the scope of the present application.

Claims

1. A precipitation diagnosis method based on multivariate scale decomposition, characterized in that, Includes the following steps: S1. Acquire high spatiotemporal resolution reanalysis grid data and precipitation measurement satellite observation data; S2. Preprocess the satellite observation data for precipitation measurement to generate a daily precipitation variation dataset; S3. Based on high spatiotemporal resolution reanalysis data, construct water vapor balance items and generate a daily variation dataset of water vapor balance; S4. Based on the daily precipitation variation dataset and the daily water vapor balance variation dataset, establish a water vapor balance equation with a daily time-scale integral containing nonlinear terms. S5. Decompose the multiple variables involved in the nonlinear term into time scales, calculate the contribution of the water vapor budget to the water vapor budget under the influence of variables at different time scales, and obtain the key water vapor budget nonlinear term. S6. Calculate the relative contributions of variables at different time scales in the nonlinear term of the water vapor budget to obtain the key contribution variables.

2. The precipitation diagnosis method based on multivariate scale decomposition according to claim 1, characterized in that: The high spatiotemporal resolution reanalysis data gridded data includes: horizontal wind field and specific humidity data. The spatial range of this data is global, with a spatial resolution of 1.25°×1.25° and a temporal resolution of 3h. Evaporation data, with a global spatial range, a spatial resolution of 1.25° × 1.25°, and a temporal resolution of 1 hour; The precipitation measurement satellite observation data includes: a global precipitation measurement multi-satellite precipitation joint inversion dataset, which has a spatial range of 60°S-60°N, a spatial resolution of 0.1°×0.1°, and a temporal resolution of 3h.

3. The precipitation diagnosis method based on multivariate scale decomposition according to claim 2, characterized in that: In step S2, the preprocessing of precipitation measurement satellite observation data involves converting the recording mode of precipitation measurement satellite observation data from UTC to local time.

4. The precipitation diagnosis method based on multivariate scale decomposition according to claim 3, characterized in that: In step S3, the method for constructing the water vapor balance item includes: Calculate the water vapor convergence and divergence term in the vertical integral of the water vapor budget equation. and horizontal water vapor advection The formula is as follows: ， ； in, For the horizontal gradient operator, For wetness, For the horizontal wind vector field, This represents the vertical integral across the entire layer from 1000 hPa to 100 hPa; Representing Universal Time and Converted to local time representation and Generate a dataset of daily water vapor balance variations.

5. The precipitation diagnosis method based on multivariate scale decomposition according to claim 4, characterized in that: The water vapor budget equation is as follows: ； in, For precipitation rate, For evaporation, For the water vapor convergence and divergence term of vertical integration, For the horizontal water vapor advection term of the vertical integral; where, and This is a nonlinear term.

6. The precipitation diagnosis method based on multivariate scale decomposition according to claim 5, characterized in that: In step S5, the method for calculating the contribution of water vapor budget to the effects of variables at different time scales includes: nonlinear terms , Variables used in calculation And 𝑞 is recorded as ,right Decompose the timescale as follows: ,in, , and These represent components at different time scales.

7. The precipitation diagnosis method based on multivariate scale decomposition according to claim 6, characterized in that: In step S6, the method for calculating the relative contributions of different variables in the nonlinear term of the water vapor budget is as follows: Based on total differential relation Estimate the relative contribution, among which, Representing the differences at different stages , These represent different variables at different time scales.

Images