Wind resource prediction method based on multi-source observation data quality control assimilation

Abstract

The invention relates to a wind resource prediction method based on quality control assimilation of multi-source observation data, and the method comprises the steps: integrating the multi-source observation data, including a multi-band radar reflectivity factor, a radial speed, a ground observation station, and wind profile radar data; quality control processing such as mottle removal, data filling, speed folding removal and discontinuous data removal is carried out, so that the accuracy and continuity of the data are improved; then, through lattice point interpolation and three-dimensional variation assimilation technologies, the processed observation data are fused into an initial background field, and an analysis field closer to the actual atmospheric state is generated, so that the reliability of severe convective weather risk assessment is remarkably improved, and the accuracy of wind energy prediction is also enhanced; and more comprehensive, detailed and stable information support is provided for weather forecast and energy management.

Description

Technical Field

[0001] Multiple embodiments of this specification relate to the field of wind resource prediction, and specifically to a wind resource prediction method based on quality control assimilation of multi-source observation data. Background Art

[0002] Wind energy is one of the clean energy sources with the largest total amount available. It is estimated that the global wind energy is about 2.24×10 9 MW, of which the available wind energy is about 2×10 2 MW is 10 times greater than the total amount of exploitable hydropower on Earth. However, wind energy is highly unstable, and the ability to effectively predict wind energy production directly determines whether wind energy can be effectively utilized.

[0003] Severe convective weather (such as typhoons, heavy rainfall, hail, thunderstorms, and downbursts) often poses serious risks to wind turbines, wind turbines, and power grids, severely impacting normal life and production. Effectively forecasting severe convective weather can increase the controllability and reliability of wind power systems, improve the efficiency of wind turbines, and reduce their operation and maintenance costs.

[0004] Therefore, there is an urgent need for a method that can accurately predict severe convective weather and wind energy. Summary of the Invention

[0005] The embodiments of this specification provide a wind resource prediction method based on quality control assimilation of multi-source observation data, which can accurately predict severe convective weather and wind energy.

[0006] The technical solution is as follows:

[0007] The embodiments of this specification provide a wind resource prediction method based on quality control assimilation of multi-source observation data, including:

[0008] Acquiring multi-source observation data, including multi-band radar reflectivity factor data, multi-band radar radial velocity data, ground station data, and wind profiler radar data;

[0009] Perform despeckling and data filling processing on the multi-band radar reflectivity factor data to obtain the multi-band radar reflectivity factor data after quality control;

[0010] The multi-band radar radial velocity data is subjected to despeckle processing, velocity unfolding processing, discontinuous data removal processing, and data filling processing to obtain the multi-band radar radial velocity data after quality control;

[0011] Perform spatiotemporal consistency checks and discontinuous data removal on ground station data and wind profiler radar data to obtain continuous ground station data and quality-controlled wind profiler radar data.

[0012] Based on the relative distance information between each ground station, the continuous ground station data are fused to obtain the ground station data after quality control;

[0013] Perform grid point interpolation processing on the quality-controlled multi-band radar reflectivity factor data, the quality-controlled multi-band radar radial velocity data, the quality-controlled ground station data, and the quality-controlled wind profiler radar data in the model grid to obtain gridded multi-band radar reflectivity factor data, gridded multi-band radar radial velocity data, gridded ground station data, and gridded wind profiler radar data;

[0014] Obtain an initial background field and, based on the three-dimensional variational assimilation system and the initial background field, assimilate gridded multi-band radar reflectivity factor data, gridded multi-band radar radial velocity data, gridded ground station data, and gridded wind profiler radar data. During the assimilation process, adjust the initial background field to obtain the analysis field.

[0015] Severe convection risk assessment and wind energy forecast are performed based on the obtained analysis field.

[0016] As a preferred solution, the velocity unfolding processing of the multi-band radar radial velocity data includes:

[0017] Perform removal and marking operations on the multi-band radar radial velocity data to remove invalid observation information in the multi-band radar radial velocity data and mark the near-zero radial velocity sectors in the multi-band radar radial velocity data;

[0018] The vertical wind profile is obtained based on the multi-band radar radial velocity data after removal and marking operations;

[0019] Velocity unfolding processing is performed on the multi-band radar radial velocity data after removal and marking operations based on the vertical wind profile.

[0020] As a preferred solution, a removal operation is performed on the multi-band radar radial velocity data to remove invalid observation information in the multi-band radar radial velocity data, including:

[0021] Remove isolated observation points and empty range circles caused by range folding in the multi-band radar radial velocity data to remove invalid observation information in the multi-band radar radial velocity data;

[0022] When the number of observation points in a range circle does not exceed a first preset number, it is considered as an empty range circle.

[0023] As a preferred solution, a marking operation is performed on the multi-band radar radial velocity data to mark the near-zero radial velocity sector in the multi-band radar radial velocity data, including:

[0024] When the number of observation points in the analysis window whose radial velocity is less than the first preset velocity exceeds a preset ratio, all observation points in the analysis window are marked;

[0025] When the number of observation points in the distance circle whose radial velocity is less than a second preset velocity exceeds a second preset number, all observation points in the distance circle are marked;

[0026] When the radial speed of a third preset number of observation points in a range circle is less than the second preset speed, all observation points in the range circle are marked;

[0027] Based on the above marking method, a marking operation is performed on the multi-band radar radial velocity data to mark the near-zero radial velocity sectors in the multi-band radar radial velocity data.

[0028] As a preferred solution, the vertical wind profile is obtained based on the multi-band radar radial velocity data after the removal operation and the marking operation, including:

[0029] Get the radar observed radial velocity V r The expression:

[0030] v r =v h cosαcosβ+v f sinα (2)

[0031] Among them, V h is the background horizontal wind speed, V f is the particle falling velocity, α is the radar detection elevation angle, and β is the radar detection azimuth angle;

[0032] Assuming that the background horizontal wind speed and particle falling speed in the observation area are uniform and constant, and the radial wind speed presents a sinusoidal change, then:

[0033] When β=0, v r1 =v f sinα+v h cosα (3)

[0034] When β=π, v r2 =v f sinα-v h cosα (4)

[0035] Among them, v r1 It represents the radial velocity detected by the radar when the radar detection azimuth is 0 degrees, v r2 It indicates the radar observed radial velocity detected when the radar detection azimuth angle is 180 degrees;

[0036] Combining the above equations (3) and (4), we can get:

[0037]

[0038] Based on equation (5) and the multi-band radar radial velocity data after removal and marking operations, the background horizontal wind speed corresponding to each radar detection elevation angle is calculated, and then the vertical wind profile is obtained.

[0039] As a preferred solution, the velocity unfolding processing is performed on the multi-band radar radial velocity data after the removal operation and the marking operation based on the vertical wind profile, including:

[0040] Get the reference radial velocity on a range circle when interpolating the vertical wind profile to a given radar detection elevation angle. The expression is:

[0041]

[0042] Wherein, a0 represents the average value of all zero-order harmonic terms in the horizontal area covered by the radar observation radial velocity, u0 represents the average value of all first observed horizontal wind components in the horizontal area covered by the radar observation radial velocity in the first horizontal direction, v0 represents the average value of all second observed horizontal wind components in the horizontal area covered by the radar observation radial velocity in the second horizontal direction, α represents the radar detection elevation angle, β represents the radar detection azimuth angle, and the first horizontal direction is perpendicular to the second horizontal direction;

[0043] Obtain the relationship between the zero-order harmonic term, the first observed horizontal wind component, the second observed horizontal wind component, the radar detection elevation angle, and the radar detection azimuth angle:

[0044]

[0045] a represents the zero-order harmonic term, r represents the radius of the range circle, w represents the vertical velocity of the observed particle, and w T represents the final falling velocity of the hydrometeor, u represents the first observed horizontal wind component, v represents the second observed horizontal wind component, represents the horizontal wind divergence in the first horizontal direction, represents the horizontal wind divergence in the second horizontal direction;

[0046] Calculate the velocity folding constant N from the reference radial velocity:

[0047]

[0048] Where vr represents the radar observed radial velocity, v N represents the Nyquist velocity of the radar, Indicates leaving nearest integer;

[0049] When N=0, no velocity folding occurs. When N≠0, the radar observed radial velocity is proportional to v r Adjust to v r +2Nv N ;

[0050] Based on the above adjustment method, velocity unfolding processing is performed on the multi-band radar radial velocity data after the removal and marking operations.

[0051] As a preferred solution, the multi-band radar reflectivity factor data and the multi-band radar radial velocity data are filled with data, both using the least squares local quadratic fitting method, and the fitting function is f = ax 2 +bx+c;

[0052] in:

[0053]

[0054] Where x represents the position of the multi-band radar reflectivity factor or the multi-band radar radial velocity on the corresponding azimuth and radial range circle, y represents the multi-band radar reflectivity factor or the multi-band radar radial velocity, and a, b, and c are fitting parameters;

[0055] The loss function during fitting is: And let:

[0056]

[0057] Among them, y i represents the i-th multi-band radar reflectivity factor or multi-band radar radial velocity, L represents the loss value, f i represents the fitting function value corresponding to the i-th multi-band radar reflectivity factor or multi-band radar radial velocity, x i represents the position of the i-th multi-band radar reflectivity factor or multi-band radar radial velocity on the corresponding azimuth and radial range circle, and n represents the total number of multi-band radar reflectivity factors and multi-band radar radial velocities.

[0058] As a preferred solution, the continuous ground station data are fused based on the relative distance information between the ground stations to obtain the ground station data after quality control. The fusion process is performed using the following formula:

[0059]

[0060] in, represents the fused ground station data corresponding to ground station l, L represents the total number of ground stations, m represents the total number of ground stations in the area with radius R centered on ground station l, and si represents the observation data of the i-th ground station in the area with a radius of R and a ground station l as the center, d i is the distance between the i-th ground station and the ground station l in the area with a radius of R and centered on the ground station l.

[0061] As a preferred solution, the gridded multi-band radar reflectivity factor data are assimilated based on the three-dimensional variational assimilation system and the initial background field, including:

[0062] According to the radar scattering principle, the multi-band radar reflectivity factor y in the gridded multi-band radar reflectivity factor data is expressed as:

[0063] y=yq r +yq s +yq h (14)

[0064] Among them, yq r 、yq s 、yq h They are respectively expressed as rain reflectivity factor, snow reflectivity factor, and hail reflectivity factor;

[0065] The empirical formula between the hydrometeor reflectivity factor and the hydrometeor mixing ratio is obtained:

[0066] yq x =a x ρq x 1.75 (15)

[0067] Where x is one of r, s, and h, ρ is the air density, and q x is the mixing ratio of x types of hydrometeors, a x is a coefficient determined by the dielectric constant, the density of x type of hydrometeor and the phase parameters;

[0068] Get the expression for the hydrometeor reflectivity factor:

[0069] yq x =y·C x (16)

[0070] Among them, C x is the contribution of x types of hydrometeors to the multi-band radar reflectivity factor;

[0071] Substituting formula (16) into formula (15) yields the formula for calculating the hydrometeoric mixing ratio, namely:

[0072]

[0073] Based on the three-dimensional variational assimilation system, the initial background field and the hydrometeor mixing ratio calculation formula, the gridded multi-band radar reflectivity factor data are assimilated.

[0074] As a preferred solution, based on the three-dimensional variational assimilation system and the initial background field, when assimilating the gridded multi-band radar radial velocity data, the forward observation operator of the radar observed radial velocity is used to convert the model wind field data into the radar observed radial velocity;

[0075] The forward observation operator of the radar observed radial velocity is:

[0076]

[0077] Among them, v r3 represents the radar observed radial velocity converted from the model wind field data, r represents the distance from the target particle to the radar, h represents the height of the target particle above the ground, s represents the distance from the target particle to the radar on the earth's curved surface, β represents the radar detection azimuth, u1 represents the horizontal wind component of the first model, v1 represents the horizontal wind component of the second model, and w1 represents the vertical velocity of the model particle.

[0078] The beneficial effects of the technical solutions provided by some embodiments of this specification include at least:

[0079] By integrating multi-source observation data, including multi-band radar reflectivity factors, radial velocity, ground station and wind profiler radar data, and performing quality control processing such as de-speckling, data filling, velocity unfolding, and discontinuous data removal, the accuracy and continuity of the data are improved; then, through grid interpolation and three-dimensional variational assimilation technology, the processed observation data are integrated into the initial background field to generate an analysis field that is closer to the actual atmospheric state. This not only significantly improves the reliability of severe convective weather risk assessment, but also enhances the accuracy of wind energy forecasts, providing more comprehensive, detailed and stable information support for weather forecasting and energy management. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0081] Figure 1 This is a flow chart of a wind resource prediction method based on quality control assimilation of multi-source observation data provided in an embodiment of this specification.

[0082] Figure 2 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this specification. DETAILED DESCRIPTION

[0083] The technical solutions in the embodiments of this specification will be described clearly and completely below in conjunction with the drawings in the embodiments of this specification.

[0084] Throughout this specification, the claims, and the accompanying drawings, the terms "first," "second," "third," and the like are used to distinguish between different items, not to describe a particular order. Furthermore, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements is not limited to the listed steps or elements but may optionally include steps or elements not listed, or may include other steps or elements inherent to the process, method, product, or apparatus.

[0085] The following description provides examples and does not limit the scope, applicability, or examples set forth in the claims. Changes may be made to the functions and arrangements of the elements described without departing from the scope of this specification. Various examples may appropriately omit, replace, or add various processes or components. For example, the described methods may be performed in an order different from the order described, and various steps may be added, omitted, or combined. Furthermore, features described with respect to some examples may be combined in other examples.

[0086] Reference Figure 1 As shown, Figure 1 A flow chart of a wind resource prediction method based on multi-source observation data quality control assimilation provided in an embodiment of this specification may at least include:

[0087] Step 102: Acquire multi-source observation data, wherein the multi-source observation data includes multi-band radar reflectivity factor data, multi-band radar radial velocity data, ground station data, and wind profiler radar data;

[0088] Step 104: performing despeckling and data filling processing on the multi-band radar reflectivity factor data to obtain quality-controlled multi-band radar reflectivity factor data, wherein the despeckling and data filling processing are performed sequentially;

[0089] Step 106: performing despeckling, velocity unfolding, discontinuous data removal, and data padding on the multi-band radar radial velocity data to obtain quality-controlled multi-band radar radial velocity data, wherein the despeckling, velocity unfolding, discontinuous data removal, and data padding are performed in sequence.

[0090] Step 108: performing a spatiotemporal consistency check and discontinuous data elimination process on the ground station data and the wind profiler radar data to obtain continuous ground station data and quality-controlled wind profiler radar data;

[0091] Step 110: Based on the relative distance information between the ground stations, the continuous ground station data are fused to obtain the ground station data after quality control;

[0092] Step 112: performing grid interpolation processing on the quality-controlled multi-band radar reflectivity factor data, the quality-controlled multi-band radar radial velocity data, the quality-controlled ground station data, and the quality-controlled wind profiler radar data in the model grid to obtain gridded multi-band radar reflectivity factor data, gridded multi-band radar radial velocity data, gridded ground station data, and gridded wind profiler radar data;

[0093] Step 114: Obtain an initial background field (the initial background field is constructed based on the WRF (Weather Research and Forecasting Model) numerical model, and the initial background field generation module is provided with initial conditions and boundary conditions by the global model). Based on the three-dimensional variational assimilation system and the initial background field, assimilate the gridded multi-band radar reflectivity factor data, gridded multi-band radar radial velocity data, gridded ground station data, and gridded wind profiler radar data. During the assimilation process, adjust the initial background field to obtain the analysis field.

[0094] Step 116: Perform severe convection risk assessment and wind energy forecast based on the obtained analysis field.

[0095] By integrating multi-source observation data, including multi-band radar reflectivity factors, radial velocity, ground station and wind profiler radar data, and performing quality control processing such as de-speckling, data filling, velocity unfolding, and discontinuous data removal, the accuracy and continuity of the data are improved; then, through grid interpolation and three-dimensional variational assimilation technology, the processed observation data are integrated into the initial background field to generate an analysis field that is closer to the actual atmospheric state. This not only significantly improves the reliability of severe convective weather risk assessment, but also enhances the accuracy of wind energy forecasts, providing more comprehensive, detailed and stable information support for weather forecasting and energy management.

[0096] Weather radar is the primary means of detecting severe convective systems and one of the main tools for monitoring and providing early warning of extreme weather events (typhoons, heavy rainfall, hail, thunderstorms, and strong winds). my country's new generation of weather radars is Doppler radar (CINRAD), which operates in S-band (wavelength approximately 10 cm), C-band (wavelength approximately 5 cm), and X-band (wavelength approximately 3 cm). S-band and C-band radars are primarily in operational use, while X-band radars are occasionally used to detect terrain-affected blind spots and provide local weather services. The high temporal and spatial resolution of this new generation of weather radars makes them unmatched in mesoscale meteorological research compared to other observational data.

[0097] Weather radars emit pulsed electromagnetic waves. When these pulses encounter precipitation (such as raindrops, snowflakes, and hail), most of the energy continues on its way, while a small portion is scattered in all directions by the precipitation. The backscattered energy returns to the radar antenna and is received by the radar. The characteristics of the precipitation system's echoes received by the radar can be used to identify the characteristics of the precipitation system (such as precipitation intensity, the presence of hail, tornadoes, and strong winds). In addition to measuring radar echo intensity (radar reflectivity), new-generation weather radars can also measure the velocity of precipitation targets along the radar's radial direction (called radial velocity) and the velocity spectrum width (a measure of velocity pulsation). These basic and derived products enable operational personnel to effectively monitor and provide early warnings for important weather events or processes, including severe convection (such as short-duration heavy rainfall, hail, thunderstorms, strong winds, and tornadoes), tropical cyclones, and heavy rain, as well as to estimate precipitation.

[0098] 1. Regarding the quality control of multi-band radar reflectivity factor data and multi-band radar radial velocity data, the specific details are as follows:

[0099] (1) The despeckling process of the multi-band radar reflectivity factor data and the despeckling process of the multi-band radar radial velocity data are performed as follows:

[0100] If there are less than three non-default data points in the 3×3 grid around a data point, the data point is marked as a mottled point and removed. Then, a median filter is used to denoise and smooth the data. The median filter formula is as follows:

[0101] h v =median p v-k ,...,p v ,...,p v+k ,v∈B (1)

[0102] Among them, h v is the filtered data corresponding to the vth grid point, p vrepresents the original data corresponding to the vth grid point, k is the filter window size, and B is the total number of grid points.

[0103] Because radars of different bands (S and C bands) and different standards (CA, CB, CC, CD, SA, SB, SC) have different maximum unambiguous velocities, it is necessary to determine the radar's Nyquist velocity to improve the quality of unfolding. Therefore, the present invention employs a unfolding algorithm based on Velocity Azimuth Display (VAD) technology.

[0104] (2) The velocity unfolding process of the multi-band radar radial velocity data includes:

[0105] Perform removal and marking operations on the multi-band radar radial velocity data to remove invalid observation information in the multi-band radar radial velocity data and mark the near-zero radial velocity sectors in the multi-band radar radial velocity data;

[0106] The vertical wind profile is obtained based on the multi-band radar radial velocity data after removal and marking operations;

[0107] Velocity unfolding processing is performed on the multi-band radar radial velocity data after removal and marking operations based on the vertical wind profile.

[0108] (3) Among them, the multi-band radar radial velocity data is removed to remove invalid observation information in the multi-band radar radial velocity data, including:

[0109] Remove isolated observation points and empty range circles caused by range folding in the multi-band radar radial velocity data to remove invalid observation information in the multi-band radar radial velocity data;

[0110] When the number of observation points in a range ring does not exceed a first preset number, it is considered an empty range ring. The first preset number may be, but is not limited to, 4.

[0111] (4) The marking operation is performed on the multi-band radar radial velocity data to mark the near-zero radial velocity sector in the multi-band radar radial velocity data, including:

[0112] When the number of observation points within the analysis window whose radial velocity is less than a first preset velocity (e.g., 1.5 m / s) exceeds a preset ratio (e.g., 75%), all observation points within the analysis window are marked;

[0113] When the number of observation points within the distance circle whose radial velocity is less than a second preset velocity (e.g., 0.5 m / s) exceeds a second preset number (e.g., 20), all observation points within the distance circle are marked;

[0114] When there are a third preset number (e.g., 10) of observation points within the range circle whose radial velocities are less than the second preset velocities, all observation points within the range circle are marked.

[0115] Based on the above marking method, a marking operation is performed on the multi-band radar radial velocity data to mark the near-zero radial velocity sectors in the multi-band radar radial velocity data.

[0116] (5) The vertical wind profile is obtained based on the multi-band radar radial velocity data after the removal operation and the marking operation, including:

[0117] Get the radar observed radial velocity V r The expression:

[0118] v r =v h cosαcosβ+v f sinα (2)

[0119] Among them, V h is the background horizontal wind speed, V f is the particle falling velocity, α is the radar detection elevation angle, and β is the radar detection azimuth angle;

[0120] Assuming that the background horizontal wind speed and particle falling speed in the observation area are uniform and constant, and the radial wind speed presents a sinusoidal change, then:

[0121] When β=0, v r1 =v f sinα+v h cosα (3)

[0122] When β=π, v r2 =v f sinα-v h cosα (4)

[0123] Among them, v r1 It represents the radial velocity detected by the radar when the radar detection azimuth is 0 degrees, v r2 It indicates the radar observed radial velocity detected when the radar detection azimuth angle is 180 degrees;

[0124] Combining the above equations (3) and (4), we can get:

[0125]

[0126] Based on equation (5) and the multi-band radar radial velocity data after removal and marking operations, the background horizontal wind speed corresponding to each radar detection elevation angle is calculated, and then the vertical wind profile is obtained.

[0127] (6) The process of performing velocity unfolding on the multi-band radar radial velocity data after removal and marking based on the vertical wind profile includes:

[0128] Get the reference radial velocity on a range circle when interpolating the vertical wind profile to a given radar detection elevation angle. The expression is:

[0129]

[0130] Where a0 represents the average value of all zero-order harmonic terms in the horizontal area covered by the radar observation radial velocity, u0 represents the average value of all first-observed horizontal wind components in the horizontal area covered by the radar observation radial velocity in the first horizontal direction, v0 represents the average value of all second-observed horizontal wind components in the horizontal area covered by the radar observation radial velocity in the second horizontal direction, α represents the radar detection elevation angle,

[0131] Indicates the radar detection azimuth, where the first horizontal direction is perpendicular to the second horizontal direction;

[0132] Obtain the relationship between the zero-order harmonic term, the first observed horizontal wind component, the second observed horizontal wind component, the radar detection elevation angle, and the radar detection azimuth angle:

[0133]

[0134] a represents the zero-order harmonic term, r represents the radius of the range circle, w represents the vertical velocity of the observed particle, and w T represents the final falling velocity of the hydrometeor, u represents the first observed horizontal wind component, v represents the second observed horizontal wind component, represents the horizontal wind divergence in the first horizontal direction, represents the horizontal wind divergence in the second horizontal direction;

[0135] Calculate the velocity folding constant N from the reference radial velocity:

[0136]

[0137] Where vr represents the radar observed radial velocity, v N represents the Nyquist velocity of the radar, Indicates leaving nearest integer;

[0138] When N=0, no velocity folding occurs. When N≠0, the radar observed radial velocity is proportional to v r Adjust to v r +2Nv N ;

[0139] Based on the above adjustment method, velocity unfolding processing is performed on the multi-band radar radial velocity data after the removal and marking operations.

[0140] (7) Among them, the multi-band radar reflectivity factor data and the multi-band radar radial velocity data are subjected to data filling processing (to fill the data area eliminated by the above steps), both using the least squares local quadratic fitting method, and the fitting function is f = ax 2 +bxc;

[0141] in:

[0142]

[0143] Where x represents the position of the multi-band radar reflectivity factor or the multi-band radar radial velocity on the corresponding azimuth and radial range circle, y represents the multi-band radar reflectivity factor or the multi-band radar radial velocity, and a, b, and c are fitting parameters;

[0144] The loss function during fitting is: And let

[0145]

[0146] Among them, y i represents the i-th multi-band radar reflectivity factor or multi-band radar radial velocity, L represents the loss value, f i represents the fitting function value corresponding to the i-th multi-band radar reflectivity factor or multi-band radar radial velocity, x i represents the position of the i-th multi-band radar reflectivity factor or multi-band radar radial velocity on the corresponding azimuth and radial range circle, and n represents the total number of multi-band radar reflectivity factors and multi-band radar radial velocities.

[0147] At this point, the radar data quality control step is completed, and the radar data after quality control is saved in the form of a three-dimensional coordinate system of radar elevation, azimuth, and radial libraries.

[0148] Next, the radar base data after quality control needs to be interpolated onto the model grid for subsequent assimilation simulation tests. The first step is to interpolate the single-station radar base data from the polar coordinate system of the radar azimuth and radial libraries to the horizontal grid points of the model. Although the distances of the radial libraries of the S and C band radars are different (500m for the C band and 1km for the S band), they need to be interpolated to the model grid of the same horizontal resolution to achieve the subsequent reflectivity puzzle task. The second step is to interpolate the data at the vertical elevation angle to the vertical grid points of the model. This step also uses the least squares local quadratic fitting method.

[0149] At this point, the preprocessing of multi-band radar data is completed and can be used in subsequent radar data assimilation experiments. It can also be applied to the output of single-station radar radial velocity and reflectivity factor images, multi-radar reflectivity factor mosaics, and wind field inversion.

[0150] 2. Quality control and interpolation of ground station data and wind profiler radar data are as follows:

[0151] Over the past five years, the layout of my country's ground meteorological observation station network has been optimized and intensified. A total of more than 76,000 ground-based automatic meteorological observation stations have been built across the country, achieving full coverage of all towns and villages. Due to the popularization and improvement of these intensified ground observation networks, the availability of real-time, high-resolution ground observation data has increased significantly. These data have high temporal and spatial resolution (spatial resolution is less than 10 km, and temporal resolution is about 5 minutes), so they can be used to monitor rapidly changing near-surface small and medium-scale features. These small and medium-scale features are of great significance for the forecast and early warning of severe convective weather. Assimilating these high-temporal and spatial resolution ground observation data can effectively compensate for the inaccuracy of the low-level initial model field caused by the low-level blind spots of radar detection, thereby producing more accurate analysis and forecast fields.

[0152] However, it is difficult to extract large-scale, three-dimensional wind field information using only ground-based meteorological observation networks (with a limited number of vertical layers) and radar observations of radial winds (due to the lack of low-level detection data due to the presence of the lowest elevation angle). Wind profiler radar data, with its high vertical resolution, can effectively extract information about the vertical structure of wind fields within the troposphere and boundary layer, thus compensating for the shortcomings of conventional observational data and, after assimilation, effectively improving the accuracy of numerical forecasts.

[0153] After reading in the data, you first need to convert the wind direction and speed data into horizontal wind components:

[0154] u=-spd·sin dir

[0155] v=-spd·cos dir (11)

[0156] Where u represents the first observed horizontal wind component, v represents the second observed horizontal wind component, spd is the wind speed, and dir is the wind direction.

[0157] Afterward, the ground station data and wind profiler radar data were checked for spatiotemporal consistency and discontinuous data were removed to obtain continuous ground station data and quality-controlled wind profiler radar data. Data with significant differences before and after acquisition time and data with severe spatial discontinuities were removed. The discontinuous data threshold was determined empirically.

[0158] Finally, the continuous ground station data and the wind profiler radar data after quality control are fused and grid interpolated. Since the influence of the relative distance between stations is not considered in the assimilation process of the continuous ground station data, we need to fuse the continuous ground station data first, that is, combine the station data within a certain range according to the relative distance weight, and then use the least squares local quadratic fitting method to interpolate in the horizontal direction. The specific fusion process is performed using the following formula:

[0159]

[0160] in, represents the fused ground station data corresponding to ground station l, L represents the total number of ground stations, m represents the total number of ground stations in the area with radius R centered on ground station l, and s i represents the observation data of the i-th ground station in the area with a radius of R and a ground station l as the center, d i is the distance between the i-th ground station and the ground station l in the area with a radius of R and centered on the ground station l.

[0161] For quality-controlled wind profiler radar data, after temporal and spatial consistency checks, only vertical interpolation is performed. Due to the presence of small and medium-scale turbulence in the lower and middle atmosphere, wind profiler radar data can contain high-frequency pulsations, which can lead to significant errors during the assimilation process. Here, we eliminate this error by temporally averaging the radar wind profiler data from three consecutive moments.

[0162] 3. Assimilation and application of multi-source observation data:

[0163] The present invention realizes the assimilation application of multi-source observation data based on the three-dimensional variational assimilation system and cloud analysis system framework.

[0164] The three-dimensional variational assimilation system uses an incremental form of a loss function that includes background error, observation error, and equation constraints. The radar reflectivity factor is assimilated indirectly. After quality control, the radar reflectivity factor first enters the cloud analysis system, where it is decomposed into pseudo-hydrometeor variables (rain, snow, and hail) by the radar reflectivity factor's backward observation operator. It is then assimilated and analyzed together with other variables. A direct assimilation approach is adopted for the radar radial velocity. The three-dimensional wind field in the model is mapped to the radar radial velocity through linear calculations to facilitate direct comparison with the radial velocity. The mass divergence equation is used as a weak constraint, and the temperature adjustment scheme used by the cloud analysis system is a moist adiabatic scheme.

[0165] A 3D variational assimilation system starts with an initial guess or background field, typically provided by a numerical forecast model, and adjusts the initial background field as observational data are assimilated. The final analysis field is a combination of the initial background field and the observations. This process makes some assumptions (such as a Gaussian distribution), is subject to some physical constraints (such as divergence), and requires numerical methods to minimize a cost function. The cost function of the 3D variational method can be expressed as:

[0166]

[0167] Where g represents the analysis field, g b represents the background field, g, g b Each contains its own corresponding three-dimensional wind component, potential temperature, air pressure, water vapor mixing ratio and pseudo hydrometeor variables (rain, snow and hail) obtained by the cloud analysis system. The entire cost function consists of three parts. The first is the background term B -1 gg b , he represents the analysis field g and the background field g b The deviation between them is given by the transpose of the background error covariance matrix B as a weight factor. The second term is the observation term [H gf 0 ] T R -1 [H gf 0 ], represents the relationship between the analysis field variable and the observation variable f 0 Among them, H represents the forward observation operator, which is used to convert the pattern variable into the observation variable, and R -1 It represents the transpose of the observation error covariance matrix, which includes the inherent error of the observation instrument and the representative error of the observation. c g represents dynamic constraints or other equation constraints, which is very important for the assimilation of mesoscale observations.

[0168] For the indirect assimilation of radar reflectivity factor, we need to use the backward observation operator of reflectivity factor to decompose it into model-resolvable hydrometeor variables (rainwater q r , snow q s and hail q h ).

[0169] Furthermore, based on the three-dimensional variational assimilation system and the initial background field, the gridded multi-band radar reflectivity factor data are assimilated, including:

[0170] According to the radar scattering principle, the multi-band radar reflectivity factor y in the gridded multi-band radar reflectivity factor data is expressed as:

[0171] y=yq r +yq s +yq h (14)

[0172] Among them, yq r 、yq s 、yq h They are respectively expressed as rain reflectivity factor, snow reflectivity factor, and hail reflectivity factor;

[0173] The empirical formula between the hydrometeor reflectivity factor and the hydrometeor mixing ratio is obtained:

[0174] yq x =a x ρq x 1.75 (15)

[0175] Where x is one of r, s, and h, ρ is the air density, and q x is the mixing ratio of x types of hydrometeors, a x is a coefficient determined by the dielectric constant, the density of x type of hydrometeor and the phase parameters;

[0176] Get the expression for the hydrometeor reflectivity factor:

[0177] yq x =y·C x (16)

[0178] Among them, C x is the contribution of x types of hydrometeors to the multi-band radar reflectivity factor;

[0179] Substituting formula (16) into formula (15) yields the formula for calculating the hydrometeoric mixing ratio, namely:

[0180]

[0181] Based on the three-dimensional variational assimilation system, the initial background field and the hydrometeor mixing ratio calculation formula, the gridded multi-band radar reflectivity factor data are assimilated.

[0182] Among them, based on the three-dimensional variational assimilation system and the initial background field, when the gridded multi-band radar radial velocity data are assimilated, the forward observation operator of the radar observed radial velocity is used to convert the model wind field data into the radar observed radial velocity;

[0183] The forward observation operator of the radar observed radial velocity is:

[0184]

[0185] Among them, v r3 represents the radar observed radial velocity converted from the model wind field data, r represents the distance from the target particle to the radar, h represents the height of the target particle above the ground, s represents the distance from the target particle to the radar on the earth's curved surface, β represents the radar detection azimuth, u1 represents the horizontal wind component of the first model, v1 represents the horizontal wind component of the second model, and w1 represents the vertical velocity of the model particle.

[0186] In addition, ground meteorological station data and wind profiler radar data have been converted into model variables after quality control and grid interpolation, so they can directly participate in the assimilation analysis. Moreover, considering the need to assimilate observational data of different spatial scales and types, the present invention sets four assimilation channels in each assimilation step, each channel containing different types of observational data and using filters of different scales. The scale of the filter is determined by the spatial density of each type of observational data. The first channel is used to analyze the encrypted ground meteorological observation data, and the horizontal and vertical influence radius are set to 10km and 0.5km, respectively. The second channel is used to analyze wind profiler radar data, and its horizontal influence radius is set to 30km. Radar radial velocity and pseudo-hydrometeor variables are analyzed in the third and fourth channels, respectively, and the influence radius of both in the horizontal and vertical directions is set to the distance of 2 model grid points.

[0187] Some more detailed solutions in the present invention can be found in the following:

[0188] First, based on the designed model grid configuration, the model was preprocessed using the WPS (WRF preprocessing system). This included defining the model projection, region range, and nesting relationships; reprojecting the data to extract the required meteorological parameters from the GRIB global data; and performing spatiotemporal interpolation of surface parameters and meteorological data, applying the meteorological parameters to the simulation domain. The MODIS_30s high-resolution static terrain data for this period was used as input.

[0189] Background information is provided by global forecast models or reanalysis data. The present invention selects the NCEP Global Forecast System (GFS) global model forecast data and the ERA5 (the fifth-generation ECMWF global reanalysis) reanalysis data. Other data can also be selected according to actual conditions. GFS is a global weather forecast model developed by the National Centers for Environmental Prediction (NCEP) of the United States. It can generate data on dozens of atmospheric and land variables, including temperature, wind speed, precipitation, soil moisture and atmospheric ozone concentration. The system couples four independent models (atmospheric model, ocean model, land / soil model and sea ice model) together to accurately describe weather conditions. The data in the present invention is its 0.5°×0.5° forecast product. The ERA5 reanalysis data is a new generation of reanalysis data set developed by the European Centre for Medium-Range Weather Forecasts (ECMWF), covering the period from January 1940 to the present. The horizontal grid resolution of ERA5's atmospheric and surface variables is 0.25°×0.25°, with 37 layers in the vertical direction and hourly resolution in time.

[0190] After being processed by the WPS module, the meteorological data is then passed to the REAL module for vertical interpolation of variables and further consistency checks. Finally, the initial and boundary conditions for the WRF model are generated. Using these initial and boundary conditions, the WRF-ARW model can perform integration operations.

[0191] Due to certain errors in the initial field, there are imbalances or mismatches between the model's initial field variables and between them and external conditions. Therefore, after the model is started, it needs to run for a period of time to reach the desired "equilibrium state." The time required to reach this state is called the spin-up time. Before the spin-up is complete, the model output is often unreliable. The time required for the model to warm up varies depending on the research object or model. For small and mesoscale meteorological models, such as WRF, the initial field variables, especially the humidity field, are often overly smooth. This leads to a lack of consistency between the model's initial divergence field, diabatic heating, and humidity fields, and a mismatch between dynamic and thermodynamic processes and the water vapor cycle. Furthermore, small and mesoscale models generally begin integration from a cloud-free initial field, so generating cloud water information after a cold start often requires some time. Based on this, we set a two-hour model warm-up time to allow the various physical variables to reach equilibrium.

[0192] In the second step, after receiving the multi-source observation data, it enters a preprocessing system for quality control and interpolation fusion. It then enters a three-dimensional variational assimilation system for analysis along with the spun-up model background variables. After this analysis, the various model physical variables will reach a new equilibrium state, closer to the actual state observed in real-world observations.

[0193] In the third step, the analysis field after data assimilation is used as the new initial field for model integral forecasting. At this point, the single assimilation and prediction steps are completed. If multiple cyclic assimilations are required, the above steps need to be repeated, and the assimilated model forecast field is used as the initial background field for assimilation at the next time, until the last moment of the assimilation time window. Generally, when performing convective-scale rapid cyclic assimilation, radar data assimilation usually requires an assimilation time window of at least one hour, and S-band radar data is input at 6-minute intervals for cyclic assimilation. As needed, encrypted ground observation data and wind profiler radar data can be assimilated at different times. According to the present invention, the ground observation data is temporally interpolated at 5-minute intervals based on the time of radar data assimilation, and rapidly assimilated at 6-minute intervals; according to the properties of the wind profiler radar data, it is time-averaged hourly and assimilated at 1-hour intervals. According to tests, the assimilation scheme under this cyclic structure has the best balance between the various physical variables and the smallest error in the wind field in the boundary layer.

[0194] The fourth step is to perform multi-scale forecasting using the final analysis field from the assimilation window. Depending on the research and operational objectives and the different observational data, assimilation can be performed on different grids. In this method, radar and ground observations are assimilated on the innermost grid of the model, corresponding to the update of storm-scale information; wind profiler radar data are assimilated on the second-layer grid, corresponding to the update of mesoscale environmental information. This completes the multi-source observation data assimilation and numerical forecasting tasks.

[0195] The fifth step is to conduct a severe convection risk assessment and wind power forecast. Based on the numerical forecast output results obtained in the fourth step, we first analyze the future extreme weather potential (whether convective weather will form around the wind farm or power grid system in question) and use this to issue risk warning information; then we evaluate the wind speed data output by the numerical forecast and calculate the future wind power potential. The wind speed-wind power conversion formula is as follows:

[0196] P=0.5ρAv 3 η (19)

[0197] Where P is the wind turbine's output power, ρ is the air density, A is the rotor blade area, v is the full wind speed, and η is the wind turbine's power generation efficiency. In this step, future wind turbine operation and maintenance can be arranged based on the actual wind speed and the wind turbine's maximum load. For example, in a strong typhoon, to reduce the risk of wind turbine damage, wind turbines should be shut down as much as possible and adjusted according to the wind direction and speed outside the typhoon to minimize damage. In a weak typhoon, wind turbine direction can be adjusted according to the wind direction and speed outside the typhoon to enhance wind power output and achieve efficient wind energy utilization.

[0198] The foregoing description of this specification describes specific embodiments. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0199] See also Figure 2 A schematic structural diagram of an electronic device provided in an embodiment of this specification is shown.

[0200] like Figure 2 As shown, the electronic device 200 may include: at least one processor 201 , at least one network interface 204 , a user interface 203 , a memory 205 and at least one communication bus 202 .

[0201] The communication bus 202 may be used to implement connection and communication among the above components.

[0202] The user interface 203 may include buttons, and the optional user interface may also include a standard wired interface or a wireless interface.

[0203] The network interface 204 may include, but is not limited to, a Bluetooth module, an NFC module, a Wi-Fi module, and the like.

[0204] Among them, the processor 201 may include one or more processing cores. The processor 201 uses various interfaces and lines to connect the various parts of the entire electronic device 200, and executes various functions of the electronic device 200 and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 205, and calling data stored in the memory 205. Optionally, the processor 201 can be implemented in at least one hardware form of DSP, FPGA, PLA. The processor 201 can integrate one or a combination of CPU, GPU and modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to handle wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 201, but may be implemented separately through a chip.

[0205] The memory 205 may include RAM or ROM. Optionally, the memory 205 includes a non-transitory computer-readable medium. The memory 205 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 205 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 205 may also be at least one storage device located away from the aforementioned processor 201. The memory 205 as a computer storage medium may include an operating system, a network communication module, a user interface module and a wind resource prediction application. The processor 201 may be used to call the wind resource prediction application stored in the memory 205 and execute the steps of the wind resource prediction method mentioned in the above-mentioned embodiment.

[0206] The embodiments of this specification also provide a computer-readable storage medium storing instructions that, when executed on a computer or processor, cause the computer or processor to perform one or more steps of the aforementioned wind resource prediction method embodiment. If the component modules of the aforementioned electronic device are implemented as software functional units and sold or used as independent products, they may be stored in the computer-readable storage medium.

[0207] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiments of this specification is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted via the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrations. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a digital versatile disc (DVD)), or a semiconductor medium (eg, a solid state disk (SSD)).

[0208] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When executed, the program can include the processes of the above-described method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks. The technical features of this embodiment and the implementation scheme can be combined in any manner unless they conflict.

[0209] The embodiments described above are merely preferred embodiments of this specification and are not intended to limit the scope of this specification. Without departing from the design spirit of this specification, various modifications and improvements made to the technical solutions of this specification by ordinary technicians in this field should fall within the scope of protection determined by the claims of this specification.

Claims

1. A wind resource prediction method based on quality control assimilation of multi-source observation data, characterized in that: include: Acquiring multi-source observation data, including multi-band radar reflectivity factor data, multi-band radar radial velocity data, ground station data, and wind profiler radar data; Perform despeckling and data filling processing on the multi-band radar reflectivity factor data to obtain the multi-band radar reflectivity factor data after quality control; The multi-band radar radial velocity data is subjected to despeckle processing, velocity unfolding processing, discontinuous data removal processing, and data filling processing to obtain the multi-band radar radial velocity data after quality control; Perform spatiotemporal consistency checks and discontinuous data removal on ground station data and wind profiler radar data to obtain continuous ground station data and quality-controlled wind profiler radar data. Based on the relative distance information between each ground station, the continuous ground station data are fused to obtain the ground station data after quality control; Perform grid point interpolation processing on the quality-controlled multi-band radar reflectivity factor data, the quality-controlled multi-band radar radial velocity data, the quality-controlled ground station data, and the quality-controlled wind profiler radar data in the model grid to obtain gridded multi-band radar reflectivity factor data, gridded multi-band radar radial velocity data, gridded ground station data, and gridded wind profiler radar data; Obtain an initial background field and, based on the three-dimensional variational assimilation system and the initial background field, assimilate gridded multi-band radar reflectivity factor data, gridded multi-band radar radial velocity data, gridded ground station data, and gridded wind profiler radar data. During the assimilation process, adjust the initial background field to obtain the analysis field. Severe convection risk assessment and wind energy forecast are performed based on the obtained analysis field.

2. The wind resource prediction method based on multi-source observation data quality control assimilation according to claim 1 is characterized in that: The velocity unfolding processing of the multi-band radar radial velocity data includes: Perform removal and marking operations on the multi-band radar radial velocity data to remove invalid observation information in the multi-band radar radial velocity data and mark the near-zero radial velocity sectors in the multi-band radar radial velocity data; The vertical wind profile is obtained based on the multi-band radar radial velocity data after removal and marking operations; Velocity unfolding processing is performed on the multi-band radar radial velocity data after removal and marking operations based on the vertical wind profile.

3. The wind resource prediction method based on multi-source observation data quality control assimilation according to claim 2 is characterized in that: Perform removal operations on the multi-band radar radial velocity data to remove invalid observation information in the multi-band radar radial velocity data, including: Remove isolated observation points and empty range circles caused by range folding in the multi-band radar radial velocity data to remove invalid observation information in the multi-band radar radial velocity data; When the number of observation points in a range circle does not exceed a first preset number, it is considered as an empty range circle.

4. The wind resource prediction method based on multi-source observation data quality control assimilation according to claim 2 is characterized in that: Perform a marking operation on the multi-band radar radial velocity data to mark the near-zero radial velocity sector in the multi-band radar radial velocity data, including: When the number of observation points in the analysis window whose radial velocity is less than the first preset velocity exceeds a preset ratio, all observation points in the analysis window are marked; When the number of observation points in the distance circle whose radial velocity is less than a second preset velocity exceeds a second preset number, all observation points in the distance circle are marked; When the radial speed of a third preset number of observation points in a range circle is less than the second preset speed, all observation points in the range circle are marked; Based on the above marking method, a marking operation is performed on the multi-band radar radial velocity data to mark the near-zero radial velocity sectors in the multi-band radar radial velocity data.

5. The wind resource prediction method based on multi-source observation data quality control assimilation according to claim 2 is characterized in that: The vertical wind profile is obtained based on the multi-band radar radial velocity data after the removal operation and the marking operation, including: Get the radar observed radial velocity V r The expression: in r =in h cosαcosβ+v f sinα (2) Among them, V h is the background horizontal wind speed, V f is the particle falling velocity, α is the radar detection elevation angle, and β is the radar detection azimuth angle; Assuming that the background horizontal wind speed and particle falling speed in the observation area are uniform and constant, and the radial wind speed presents a sinusoidal change, then: When β = 0, v r1 = v f sinα + v h cosα (3) When β=π, v r2 =v f sinα-v h thing (4) Among them, v r1 It represents the radial velocity detected by the radar when the radar detection azimuth is 0 degrees, v r2 It indicates the radar observed radial velocity detected when the radar detection azimuth angle is 180 degrees; Combining the above equations (3) and (4), we can get: Based on equation (5) and the multi-band radar radial velocity data after removal and marking operations, the background horizontal wind speed corresponding to each radar detection elevation angle is calculated, and then the vertical wind profile is obtained.

6. The wind resource prediction method based on multi-source observation data quality control assimilation according to claim 2 is characterized in that: The method of performing velocity unfolding processing on the multi-band radar radial velocity data after the removal operation and the marking operation based on the vertical wind profile includes: Get the reference radial velocity on a range circle when interpolating the vertical wind profile to a given radar detection elevation angle. The expression is: Wherein, a0 represents the average value of all zero-order harmonic terms in the horizontal area covered by the radar observation radial velocity, u0 represents the average value of all first-observed horizontal wind components in the horizontal area covered by the radar observation radial velocity in the first horizontal direction, v0 represents the average value of all second-observed horizontal wind components in the horizontal area covered by the radar observation radial velocity in the second horizontal direction, α represents the radar detection elevation angle, and represents the radar detection azimuth angle, where the first horizontal direction is perpendicular to the second horizontal direction. Obtain the relationship between the zero-order harmonic term, the first observed horizontal wind component, the second observed horizontal wind component, the radar detection elevation angle, and the radar detection azimuth angle: a represents the zero-order harmonic term, r represents the radius of the range circle, w represents the vertical velocity of the observed particle, and w T represents the final falling velocity of the hydrometeor, u represents the first observed horizontal wind component, v represents the second observed horizontal wind component, represents the horizontal wind divergence in the first horizontal direction, represents the horizontal wind divergence in the second horizontal direction; Calculate the velocity folding constant N from the reference radial velocity: Where vr represents the radar observed radial velocity, v N represents the Nyquist velocity of the radar, Indicates leaving nearest integer; When N=0, no velocity folding occurs. When N≠0, the radar observed radial velocity is proportional to v r Adjust to v r +2Nv N ; Based on the above adjustment method, velocity unfolding processing is performed on the multi-band radar radial velocity data after the removal and marking operations.

7. The wind resource prediction method based on multi-source observation data quality control assimilation according to claim 2 is characterized in that: The data filling process of multi-band radar reflectivity factor data and multi-band radar radial velocity data is carried out by using the least squares local quadratic fitting method, and the fitting function is f = ax 2 +bx+c; in: Where x represents the position of the multi-band radar reflectivity factor or the multi-band radar radial velocity on the corresponding azimuth and radial range circle, y represents the multi-band radar reflectivity factor or the multi-band radar radial velocity, and a, b, and c are fitting parameters; The loss function during fitting is: And let: Among them, y i represents the i-th multi-band radar reflectivity factor or multi-band radar radial velocity, L represents the loss value, f i represents the fitting function value corresponding to the i-th multi-band radar reflectivity factor or multi-band radar radial velocity, x i represents the position of the i-th multi-band radar reflectivity factor or multi-band radar radial velocity on the corresponding azimuth and radial range circle, and n represents the total number of multi-band radar reflectivity factors and multi-band radar radial velocities.

8. The wind resource prediction method based on multi-source observation data quality control assimilation according to claim 1 is characterized in that: Based on the relative distance information between the ground stations, the continuous ground station data are fused to obtain the ground station data after quality control. The fusion process is performed using the following formula: in, represents the fused ground station data corresponding to ground station l, L represents the total number of ground stations, m represents the total number of ground stations in the area with radius R centered on ground station l, and s i represents the observation data of the i-th ground station in the area with a radius of R and a ground station l as the center, d i is the distance between the i-th ground station and the ground station l in the area with a radius of R and centered on the ground station l.

9. The wind resource prediction method based on multi-source observation data quality control assimilation according to claim 1 is characterized in that: Based on the three-dimensional variational assimilation system and the initial background field, the gridded multi-band radar reflectivity factor data are assimilated, including: According to the radar scattering principle, the multi-band radar reflectivity factor y in the gridded multi-band radar reflectivity factor data is expressed as: y=yq r +yq s +yq h (14) Among them, yq r 、yq s 、yq h They are respectively expressed as rain reflectivity factor, snow reflectivity factor, and hail reflectivity factor; The empirical formula between the hydrometeor reflectivity factor and the hydrometeor mixing ratio is obtained: y q x =a x ρq x 1.75 (15) Where x is one of r, s, and h, ρ is the air density, and q x is the mixing ratio of x types of hydrometeors, a x is a coefficient determined by the dielectric constant, the density of x type of hydrometeor and the phase parameters; Get the expression for the hydrometeor reflectivity factor: y q x =y·C x (16) Among them, C x is the contribution of x types of hydrometeors to the multi-band radar reflectivity factor; Substituting formula (16) into formula (15) yields the formula for calculating the hydrometeoric mixing ratio, namely: Based on the three-dimensional variational assimilation system, the initial background field and the hydrometeor mixing ratio calculation formula, the gridded multi-band radar reflectivity factor data are assimilated.

10. The wind resource prediction method based on multi-source observation data quality control assimilation according to claim 1, characterized in that: Based on the three-dimensional variational assimilation system and the initial background field, when assimilating the gridded multi-band radar radial velocity data, the forward observation operator of the radar observed radial velocity is used to convert the model wind field data into the radar observed radial velocity; The forward observation operator of the radar observed radial velocity is: Among them, v r3 represents the radar observed radial velocity converted from the model wind field data, r represents the distance from the target particle to the radar, h represents the height of the target particle above the ground, s represents the distance from the target particle to the radar on the earth's curved surface, β represents the radar detection azimuth, u1 represents the horizontal wind component of the first model, v1 represents the horizontal wind component of the second model, and w1 represents the vertical velocity of the model particle.

Images