Method and device for correcting short-term numerical weather prediction data

Abstract

The application relates to a short-time numerical weather forecast data correction method and device, which comprises the following steps: acquiring a first infrared forecast cloud picture for forecasting weather in a first preset time period and a second infrared forecast cloud picture for forecasting weather in a second preset time period; acquiring an actual observation cloud picture obtained by observing the weather in the first preset time period; determining first optical flow field evolution information corresponding to the second preset time period based on the first infrared forecast cloud picture and the actual observation cloud picture; and correcting the second infrared forecast cloud picture by using the first optical flow field evolution information to obtain target weather forecast data. According to the application, the first optical flow field evolution information capable of reflecting the position and intensity deviation of a cloud cluster is used to correct the second infrared forecast cloud picture, so that the target weather forecast data obtained finally is more accurate, and the accuracy of the weather forecast data is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of short-term numerical weather prediction, and in particular to a short-term numerical weather prediction data correction method and device. BACKGROUND

[0002] Short-term numerical prediction generally refers to 0-12 hour numerical weather prediction. In areas covered by Doppler weather radar data, such as land, radar data can be used to correct 0-6 hour short-term numerical precipitation prediction results to improve precipitation prediction accuracy. However, in areas such as the open sea and highlands without radar data, such work cannot be carried out.

[0003] With the economic and social and military development of the open sea, highlands and other regions, the demand for accurate precipitation prediction is increasing. Improving the accuracy of short-term precipitation prediction in these regions still has very important research and application value. SUMMARY

[0004] In order to solve the above technical problems or at least partially solve the above technical problems, the present application provides a short-term numerical weather prediction data correction method and device.

[0005] In a first aspect, the present application provides a short-term numerical weather prediction data correction method, comprising:

[0006] obtaining a first infrared forecast cloud image for forecasting weather in a first preset time period and a second infrared forecast cloud image for forecasting weather in a second preset time period, the first preset time period being before the current time, and the second preset time period being after the current time;

[0007] obtaining an actual observation cloud image obtained by observing the weather in the first preset time period;

[0008] determining first optical flow field evolution information corresponding to the second preset time period based on the first infrared forecast cloud image and the actual observation cloud image;

[0009] correcting the second infrared forecast cloud image using the first optical flow field evolution information to obtain target weather prediction data.

[0010] Optionally, determining the first optical flow field evolution information corresponding to the second preset time period based on the first infrared forecast cloud image and the actual observation cloud image comprises:

[0011] determining a plurality of evolution intermediate points in the first preset time period;

[0012] For each of the evolution intermediate points, a brightness temperature deviation of the first infrared forecast cloud picture relative to the actual observation cloud picture in the first preset time period is calculated by using an optical flow method, to obtain second optical flow field evolution information corresponding to the first preset time period;

[0013] The first optical flow field evolution information corresponding to the second preset time period is determined based on the second optical flow field evolution information.

[0014] Optionally, determining the first optical flow field evolution information corresponding to the second preset time period based on the second optical flow field evolution information comprises:

[0015] For brightness temperature, the brightness temperature optical flow field evolution information corresponding to the second preset forecast time period is determined based on the second optical flow field evolution information;

[0016] For precipitation, the precipitation optical flow field evolution information corresponding to the second preset time period is determined based on the brightness temperature optical flow field evolution information;

[0017] The brightness temperature optical flow field evolution information and the precipitation optical flow field evolution information are determined as the first optical flow field evolution information.

[0018] Optionally, for brightness temperature, the brightness temperature optical flow field evolution information corresponding to the second preset forecast time period is determined based on the second optical flow field evolution information, comprising:

[0019] For brightness temperature, the second optical flow field evolution information is interpolated and extrapolated to obtain third optical flow field evolution information corresponding to a third preset time period, the third preset time period being located in the second preset time period;

[0020] For brightness temperature, a mean value of the third optical flow field evolution information is calculated as fourth optical flow field evolution information corresponding to a fourth preset time period, the fourth preset time period being located in the second preset time period, and the fourth preset time period being later than the third preset time period;

[0021] A combination of the third optical flow field evolution information and the fourth optical flow field evolution information is determined as the brightness temperature optical flow field evolution information corresponding to the second preset forecast time period.

[0022] Optionally, for precipitation, the precipitation optical flow field evolution information corresponding to the second preset time period is determined based on the brightness temperature optical flow field evolution information, comprising:

[0023] For precipitation, the brightness temperature optical flow field evolution information is multiplied by the first preset weight to obtain fifth optical flow field evolution information corresponding to a precipitation area;

[0024] For precipitation, the brightness temperature optical flow field evolution information is multiplied by the second preset weight to obtain sixth optical flow field evolution information corresponding to a non-precipitation area;

[0025] The combination of the fifth optical flow field evolution information and the sixth optical flow field evolution information is determined as the precipitation optical flow field evolution information corresponding to the second preset time period.

[0026] Optionally, the second infrared forecast cloud image is corrected using the first optical flow field evolution information to obtain target weather forecast data, including:

[0027] The brightness temperature forecast data in the second infrared forecast cloud image is corrected using the brightness temperature optical flow field evolution information in the first optical flow field evolution information to obtain first correction data;

[0028] The precipitation forecast data in the second infrared forecast cloud image is corrected using the precipitation optical flow field evolution information in the first optical flow field evolution information to obtain second correction data;

[0029] The combination of the first correction data and the second correction data is determined as the target weather forecast data.

[0030] Optionally, a first infrared forecast cloud image forecasting weather of a first preset time period is obtained, including:

[0031] A global numerical weather prediction product corresponding to the first preset time period and geographical data sets are input into a pre-processing module of a weather research and forecast (WRF) model to obtain an initial field and side boundary conditions of the WRF model;

[0032] Based on the initial field and the side boundary conditions, the WRF model is run to obtain atmospheric environment data;

[0033] The atmospheric environment data is input into a rapid radiative transfer (RTTOV) model, and satellite coefficient files and satellite cloud coefficient files contained in the RTTOV model are input into the RTTOV model, and the RTTOV model is run to obtain an infrared forecast cloud image.

[0034] In a second aspect, the present application provides a short-term numerical weather prediction data correction device, including:

[0035] A first obtaining module is configured to obtain a first infrared forecast cloud image forecasting weather of a first preset time period and a second infrared forecast cloud image forecasting weather of a second preset time period, the first preset time period being before a current time, and the second preset time period being after the current time;

[0036] A second obtaining module is configured to obtain an actual observation cloud image obtained by observing weather of the first preset time period.

[0037] determining, based on the first infrared forecast cloud picture and the actual observation cloud picture, first optical flow field evolution information corresponding to the second preset time period;

[0038] correcting the second infrared forecast cloud picture using the first optical flow field evolution information to obtain target weather forecast data.

[0039] Optionally, the determining module comprises:

[0040] a first determining unit configured to determine a plurality of evolution intermediate points within the first preset time period;

[0041] a calculating unit configured to, for each of the evolution intermediate points, calculate, using an optical flow method, a brightness temperature deviation of the first infrared forecast cloud picture relative to the actual observation cloud picture within the first preset time period to obtain second optical flow field evolution information corresponding to the first preset time period;

[0042] a second determining unit configured to determine, based on the second optical flow field evolution information, first optical flow field evolution information corresponding to the second preset time period.

[0043] Optionally, the second determining unit comprises:

[0044] a first determining sub-unit configured to, for brightness temperature, determine, based on the second optical flow field evolution information, brightness temperature optical flow field evolution information corresponding to the second preset time period;

[0045] a second determining sub-unit configured to, for precipitation, determine, based on the brightness temperature optical flow field evolution information, precipitation optical flow field evolution information corresponding to the second preset time period;

[0046] a third determining sub-unit configured to determine the brightness temperature optical flow field evolution information and the precipitation optical flow field evolution information as the first optical flow field evolution information.

[0047] Optionally, the first determining sub-unit is further configured to:

[0048] for brightness temperature, perform interpolation and extrapolation processing on the second optical flow field evolution information to obtain third optical flow field evolution information corresponding to a third preset time period, the third preset time period being within the second preset time period;

[0049] for brightness temperature, calculate a mean value of the third optical flow field evolution information as fourth optical flow field evolution information corresponding to a fourth preset time period, the fourth preset time period being within the second preset time period, the fourth preset time period being later than the third preset time period;

[0050] The combination of the third optical flow field evolution information and the fourth optical flow field evolution information is determined as the brightness temperature optical flow field evolution information corresponding to the second preset prediction time period.

[0051] Optionally, the second determination subunit is further configured to:

[0052] For precipitation, the brightness temperature optical flow field evolution information is multiplied by the first preset weight to obtain fifth optical flow field evolution information corresponding to a precipitation area;

[0053] For precipitation, the brightness temperature optical flow field evolution information is multiplied by the second preset weight to obtain sixth optical flow field evolution information corresponding to a non-precipitation area;

[0054] The combination of the fifth optical flow field evolution information and the sixth optical flow field evolution information is determined as the precipitation optical flow field evolution information corresponding to the second preset time period.

[0055] Optionally, the correction module comprises:

[0056] A first correction unit is configured to correct brightness temperature prediction data in the second infrared prediction cloud image by using brightness temperature optical flow field evolution information in the first optical flow field evolution information to obtain first correction data;

[0057] A second correction unit is configured to correct precipitation prediction data in the second infrared prediction cloud image by using precipitation optical flow field evolution information in the first optical flow field evolution information to obtain second correction data;

[0058] A third determination unit is configured to determine the combination of the first correction data and the second correction data as the target weather prediction data.

[0059] Optionally, the first acquisition module comprises:

[0060] An input unit is configured to input a global numerical weather prediction product corresponding to a first preset time period and geographical data sets into a pre-processing module of a weather research and prediction (WRF) model to obtain an initial field and side boundary conditions of the WRF model;

[0061] A first running unit is configured to run the WRF model based on the initial field and the side boundary conditions to obtain atmospheric environmental data;

[0062] A second running unit is configured to input the atmospheric environmental data into a rapid radiative transfer (RTTOV) model, and input a satellite coefficient file and a satellite cloud coefficient file contained in the RTTOV model into the RTTOV model, and run the RTTOV model to obtain an infrared prediction cloud image.

[0063] In a third aspect, the present application provides an electronic device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete mutual communication through the communication bus.

[0064] The memory is used for storing a computer program.

[0065] The processor is used for executing the program stored on the memory, and realizes the short-time numerical weather prediction data correction method according to any one of the first aspect.

[0066] In a fourth aspect, the present application provides a computer readable storage medium, wherein the computer readable storage medium stores a short-time numerical weather prediction data correction method program, and the short-time numerical weather prediction data correction method program is executed by a processor to realize the steps of the short-time numerical weather prediction data correction method according to any one of the first aspect.

[0067] Compared with the prior art, the above technical solution provided by the embodiments of the present application has the following advantages:

[0068] The first light flow field evolution information reflecting the position and intensity deviation of the cloud cluster is used to correct the second infrared forecast cloud picture of the second preset time period, so that the final target weather prediction data is more accurate, and the accuracy of the weather prediction data is improved. Compared with the conventional forecast data correction technology based on radar data, the present technical solution is based on satellite data and has the characteristics of wide coverage of space range, and can be used in radar data-free areas such as the open sea and the plateau, and has strong application value. BRIEF DESCRIPTION OF DRAWINGS

[0069] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present application and, together with the specification, serve to explain the principles of the application.

[0070] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced as follows. Obviously, for those skilled in the art, other drawings can also be obtained based on these drawings without creative labor.

[0071] Figure 1 A flowchart of a short-time numerical weather prediction data correction method provided by the embodiments of the present application;

[0072] Figure 2 A flowchart of a first infrared forecast cloud picture acquisition method provided by the embodiments of the present application;

[0073] Figure 3 A structural diagram of a short-time numerical weather forecast data correction device provided in an embodiment of this application;

[0074] Figure 4 This is a structural diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0075] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0076] With the economic, social, and military development of regions such as the open sea and plateaus, the demand for accurate precipitation forecasts is becoming increasingly strong. Improving the accuracy of short-term precipitation forecasts in these regions remains of great research and application value. Therefore, this application provides a method and apparatus for correcting short-term numerical weather prediction data.

[0077] Figure 1 This application provides a method for correcting short-term numerical weather forecast data, the method comprising:

[0078] Step S101: Obtain a first infrared forecast cloud image for forecasting the weather in a first preset time period and a second infrared forecast cloud image for forecasting the weather in a second preset time period.

[0079] In this embodiment of the application, the first infrared forecast cloud image is a numerical forecast cloud image for forecasting the weather within a first preset time period. Further, the first infrared forecast cloud image is an FY-4 infrared forecast cloud image (brightness temperature value). In one embodiment of this application, obtaining the first infrared forecast cloud image for forecasting the weather within the first preset time period includes:

[0080] The GFS global numerical weather prediction product and geographic dataset corresponding to the first preset time period are input into the preprocessing module of the Weather Research and Forecasting (WRF) model to obtain the initial field and lateral boundary conditions of the WRF model. The GFS global numerical weather prediction product is produced by the Global Forecast Systems (GFS) of the US NCEP. Based on the initial field and the lateral boundary conditions, the WRF model is run to carry out refined numerical weather prediction and obtain atmospheric environmental data, including forecast results for elements such as precipitation. Based on the atmospheric environmental data, diagnostic methods are used to produce cloud cover, cloud base height, cloud top height, and cloud classification forecast results. The satellite zenith angle is obtained according to the latitude and longitude information of the simulated area grid. The distribution of fresh and saltwater is determined according to the land-sea distribution data. The atmospheric environment data is input into the Rapid Radiative Transfer for TOVS (RTTOV) mode, and the satellite coefficient files and satellite cloud coefficient files contained in the RTTOV mode are also input into the RTTOV mode. The satellite coefficient files and satellite cloud coefficient files can be the coefficient files and cloud coefficient files of the FY-4 satellite. The RTTOV mode is run to obtain the first infrared forecast cloud image.

[0081] The calculation methods for cloud cover, cloud top height, and cloud base height are as follows:

[0082] 1) Calculate cloud cover based on relative humidity (RH);

[0083] RH threshold (RH) _00 At sea

[0084]

[0085] The RH threshold on land is

[0086]

[0087] Among them, gridkm is the grid spacing in kilometers.

[0088] If the temperature t > 261.16 K, then the cloud cover is:

[0089]

[0090] If the temperature t < 261.16 K and t > 203.16 K, and RH > RH _land The cloud cover is:

[0091]

[0092] 2) Calculate the cloud top height;

[0093] For the cloud cover product obtained in 1), find the layer with cloud cover CLDFRA>0 for the first time from the top layer downwards, and read the geopotential height of that layer as the cloud top height; if CLDFRA≥4, add 0.3 times the difference between the geopotential height of the previous layer and that layer as a correction amount based on the existing height.

[0094] 3) Calculate the cloud base height;

[0095] For the cloud cover product obtained in 1), find the layer where the cloud cover CLDFRA>0 for the first time from the bottom layer upwards, and read the geopotential height of that layer as the cloud base height; if CLDFRA≥4, then reduce the geopotential height difference between that layer and the next layer by 0.3 times as the correction amount based on the existing height.

[0096] Cloud classification and satellite zenith angle can be calculated using methods found in related technologies.

[0097] In this embodiment of the application, the second infrared forecast cloud image is a numerical forecast cloud image for forecasting the weather within a second preset time period. Further, the second infrared forecast cloud image is an FY-4 infrared forecast cloud image (brightness temperature value). The second infrared forecast cloud image is obtained by processing the input data of the second preset time period in WRF mode combined with RTTOV mode. The acquisition method of the second infrared forecast cloud image is similar to that of the first infrared forecast cloud image, and will not be described again here.

[0098] The first preset time period is located before the current time, and the second preset time period is located after the current time. Furthermore, the maximum boundary point of the first preset time period is the current time, and the minimum boundary point of the second preset time period is the current time. That is, the first preset time period consists of two adjacent time periods with the current time as the dividing point. Furthermore, the length of the second preset time period is twice the length of the first preset time period. For example, the first preset time period is -2 to 0 hours (where 0 hours represents the current time), and the second preset time period is 0 to 4 hours.

[0099] Step S102: Obtain the actual observed cloud image obtained by observing the weather during the first preset time period;

[0100] In this embodiment of the application, the actual observed cloud image is the observed cloud image actually observed within the first preset time period. Furthermore, the second infrared forecast cloud image is the FY-4 observed cloud image (brightness temperature value).

[0101] After obtaining the actual observed cloud image, the root mean square error and correlation coefficient of the first infrared forecast cloud image can be statistically tested to ensure that the first infrared forecast cloud image is within the normal value range and improve the accuracy of subsequent calculations.

[0102] Figure 2 This is a flowchart illustrating a method for obtaining a first infrared forecast cloud image in a practical application, as provided in an embodiment of this application.

[0103] Step S103: Determine the first optical flow field evolution information corresponding to the second preset time period based on the first infrared forecast cloud map and the actual observed cloud map;

[0104] The first infrared forecast cloud image is used to forecast the weather for a first preset time period, while the actual observation cloud image is obtained by observing the weather for the first preset time period. Therefore, in this step, the brightness temperature deviation (i.e., optical flow field) between the infrared forecast cloud image and the actual observation cloud image can be calculated using the optical flow method. Then, the optical flow field evolution information for the first preset time period can be determined based on the brightness temperature deviation. Then, based on the first optical flow field evolution information, the second optical flow field evolution information corresponding to the second preset time period can be obtained using a preset calculation method. The second optical flow field evolution information for the second preset time period can be accurately predicted based on the first optical flow field evolution information for the first preset time period. For details, please refer to the following embodiments.

[0105] Step S104: Use the first optical flow field evolution information to correct the second infrared forecast cloud image to obtain target weather forecast data.

[0106] Since the location and intensity of cloud clusters affect the location and intensity of precipitation, and there are deviations in the location and intensity of forecasted cloud clusters, this will be reflected in the optical flow field of the forecasted and measured brightness temperatures. When using the optical flow field to correct the brightness temperature (location and intensity) of forecasted cloud clusters, this information can also be used to correct the morphological distribution of forecasted precipitation. Through comparative experiments, a correction method that has a positive impact on forecasted precipitation can be found. Therefore, in this embodiment, the first optical flow field evolution information includes brightness temperature optical flow field evolution information and precipitation optical flow field evolution information. The second infrared forecast cloud image is corrected using the brightness temperature optical flow field evolution information and the precipitation optical flow field evolution information, respectively.

[0107] In one embodiment of this application, step S104 uses the first optical flow field evolution information to correct the second infrared forecast cloud image to obtain target weather forecast data, including:

[0108] The brightness temperature forecast data in the second infrared forecast cloud image is corrected using the brightness temperature optical flow field evolution information in the first optical flow field evolution information to obtain the first corrected data; the precipitation forecast data in the second infrared forecast cloud image is corrected using the precipitation optical flow field evolution information in the first optical flow field evolution information to obtain the second corrected data; the combination of the first corrected data and the second corrected data is determined as the target weather forecast data.

[0109] This application embodiment obtains first optical flow field evolution information based on a first infrared forecast cloud image and an actual observed cloud image for a first preset time period. Then, it corrects the second infrared forecast cloud image for a second preset time period based on the first optical flow field evolution information. By using the first optical flow field evolution information, which can reflect the position and intensity deviation of cloud clusters, to correct the second infrared forecast cloud image, the final target weather forecast data is more accurate, and the accuracy of the weather forecast data is improved.

[0110] In another embodiment of this application, step S103, based on the first infrared forecast cloud image and the actual observed cloud image, determines the first optical flow field evolution information corresponding to the second preset time period, including:

[0111] Step S201: Determine multiple evolution intermediate points within the first preset time period;

[0112] In this embodiment of the application, the evolution midpoint refers to the midpoint of the evolution process. For example, the first preset time period is -2 to 0 hours, and the evolution midpoint can be -2, -1, or 0. The more evolution midpoints there are, the greater the evolution density, and the more accurate the final optical flow field evolution information is. In practical applications, different evolution midpoint densities can be selected according to actual needs.

[0113] Step S202: For each of the evolution intermediate points, the brightness temperature deviation of the first infrared forecast cloud image relative to the actual observed cloud image within the first preset time period is calculated using the optical flow method to obtain the second optical flow field evolution information corresponding to the first preset time period.

[0114] For example, for brightness temperature, the brightness temperature deviation (i.e., optical flow field) of the first infrared forecast cloud image relative to the actual observed cloud image can be obtained hourly within the time frame of -2 to 0 hours using the optical flow method, and the optical flow field corresponding to each hour can be combined to obtain the second optical flow field evolution information.

[0115] Optical flow is analogous to the velocity vector of a rigid body. A velocity vector is generally expressed using two components, u and v. Assuming a point moves to a new position in a short time, and the image's grayscale value remains constant during this time, an equivalence relationship between grayscale values ​​and time / position can be established. This is then expanded using Taylor's formula to obtain the optical flow constraint equation. Since this equation contains u and v components, solving it requires certain constraints; therefore, this application uses the HS global constraint scheme.

[0116] The main calculation steps are as follows:

[0117] J = J O +α 2 J HS 1)

[0118] JO =∫∫(I x u+I y v+I t ) 2 dxdy 2)

[0119]

[0120] Equation 1) is the HS global constraint equation. α is the smoothing coefficient; J O J is the gray-scale conservation term; HS For smoothing constraint terms; I x I y and I t The spatial gradient and temporal variability of the image grayscale are given respectively, which can be calculated from the grayscale values ​​of two adjacent images (i.e., two images at the same time in the first infrared forecast cloud image and the actual observation cloud image); (u,v) is the optical flow.

[0121] By using variational methods and recursive algorithms to solve the global constraint equations, a recursive solution to (u,v) can be obtained:

[0122]

[0123]

[0124] Where k is the number of iterations, u (0) and u (0) This is the initial estimate of the optical flow, which can be set to zero.

[0125] Step S203: Determine the first optical flow field evolution information corresponding to the second preset time period based on the second optical flow field evolution information.

[0126] Since the evolution information of the first optical flow field includes the evolution information of the brightness temperature optical flow field and the evolution information of the precipitation optical flow field, the evolution information of the brightness temperature optical flow field and the evolution information of the precipitation optical flow field corresponding to the second preset time period can be determined based on the evolution information of the brightness temperature optical flow field and the evolution information of the precipitation optical flow field in the evolution information of the first optical flow field.

[0127] In one embodiment of this application, step S203, which determines the first optical flow field evolution information corresponding to the second preset time period based on the second optical flow field evolution information, includes:

[0128] Step S301: For brightness temperature, determine the brightness temperature optical flow field evolution information corresponding to the second preset prediction time period based on the second optical flow field evolution information;

[0129] In one embodiment of this application, for brightness temperature, determining the brightness temperature optical flow field evolution information corresponding to a second preset prediction time period based on the second optical flow field evolution information includes:

[0130] Step S401: For brightness temperature, the second optical flow field evolution information is interpolated and extrapolated to obtain the third optical flow field evolution information corresponding to the third preset time period.

[0131] In this embodiment of the application, the third preset time period is located within the second preset time period; for example, it is assumed that the second preset time period is 0 to 4 hours and the third preset time period is 0 (excluding) to 2 hours.

[0132] For example, the interpolation and extrapolation processes can be performed by referring to the formulas in Table 1 below for the period from 0 (excluding) to 2 hours.

[0133] Step S402: For brightness temperature, calculate the mean value of the third optical flow field evolution information, and use it as the fourth optical flow field evolution information corresponding to the fourth preset time period.

[0134] In this embodiment of the application, the fourth preset time period is located within the second preset time period, and the fourth preset time period is later than the third preset time period; for example, it is assumed that the second preset time period is 0 to 4 hours, and the fourth preset time period is 2 (excluding) to 4 hours.

[0135] An example of the mean value of the third optical flow field evolution information can be processed by referring to the formula in Table 1 below for 2 (excluding) to 4 hours.

[0136] Step S403: The combination of the third optical flow field evolution information and the fourth optical flow field evolution information is determined as the brightness temperature optical flow field evolution information corresponding to the second preset prediction time period.

[0137] For example, the scheme for obtaining the brightness temperature optical flow field evolution information for 0 to 4 hours (the second preset forecast time period) is as follows: the optical flow field for -2 to 0 hours is obtained by using a time-based linear interpolation and extrapolation formula; the average value of the optical flow field for 0 to 2 hours is calculated and set as the optical flow field for 2 to 4 hours; the two are combined to obtain the brightness temperature optical flow field evolution information corresponding to the second preset forecast time period.

[0138] Table 1. Optical Flow (OF) Acquisition Scheme for Brightness Temperature from 0 to 4 Hours

[0139] Time (h) Protocol -2 OF -2 OF -2 <!-- 8 -->]]> -1 [IN -1 = IN -1 ]]> 0 [Of0 = Of0] 1 [CD AT OF1 = (OF0 - OF -2 ) / 2 + OF0]] 2 [CD AT OF2 = (OF1 - OF -1 ) / 2 + OF1]] 3 [Of3 = (Of3 + Of2 + Of1) / 3] 4 [Of4 = (Of3 + Of2 + Ofi) / 3]

[0140] Step S302: For precipitation, determine the precipitation optical flow field evolution information corresponding to the second preset time period based on the brightness temperature optical flow field evolution information;

[0141] In one embodiment of this application, for precipitation, determining the precipitation photoflow field evolution information corresponding to the second preset time period based on the brightness temperature photoflow field evolution information includes:

[0142] Step S501: For precipitation, multiply the brightness temperature photoflow field evolution information by the first preset weight to obtain the fifth photoflow field evolution information corresponding to the precipitation area.

[0143] In this embodiment of the application, the first preset weight is a pre-set value, for example, the first preset value is 1.0.

[0144] Step S502: For precipitation, the brightness temperature optical flow field evolution information is multiplied by the second preset weight to obtain the sixth optical flow field evolution information corresponding to the non-precipitation area.

[0145] In this embodiment of the application, the second preset weight is a pre-set value, for example, the second preset value is 0.5.

[0146] Step S503: The combination of the fifth optical flow field evolution information and the sixth optical flow field evolution information is determined as the precipitation optical flow field evolution information corresponding to the second preset time period.

[0147] For example, the scheme for obtaining precipitation optical flow field evolution information for 0-4 hours (second preset forecast period) is as follows: based on the brightness temperature optical flow field evolution information for 0-4 hours, use 1.0 times the optical flow field information for areas with predicted precipitation (precipitation rate > 0.1 mm / h); use 0.5 times the optical flow field information for areas with predicted no precipitation; combine the two to obtain precipitation optical flow field evolution information corresponding to the second preset forecast period.

[0148] Step S303: The brightness temperature optical flow field evolution information and the precipitation optical flow field evolution information are determined as the first optical flow field evolution information.

[0149] The embodiments of this application can automatically determine the corresponding optical flow field evolution information for brightness temperature and precipitation, thereby obtaining the first optical flow field evolution information, which facilitates the use of the first optical flow field evolution information to correct the second infrared forecast cloud image and improve the accuracy of the target weather forecast data.

[0150] In another embodiment of this application, an implementation method and results in practical application are also provided.

[0151] (1) Data

[0152] Data used: The study utilized global numerical weather prediction products produced by the NCEP Global Forecast Systems (GFS) from 12:00 UTC on September 13 to 12:00 UTC on September 14, 2021, observational data from the FY-4A satellite imager, and CMORPH precipitation data obtained from multi-source data fusion analysis provided by a certain organization.

[0153] (2) Test Results

[0154] 1) FY-4 Forecast Cloud Image Production Technology. Forecast cloud images were produced for four channels: mid-infrared (low), water vapor (6.70 μm), far-infrared (11.2 μm), and far-infrared (12.0 μm). Case studies showed that the average root mean square error and correlation coefficient for each channel from 0 to 24 hours were 20.9K and 0.41, 8.4K and 0.52, 23.7K and 0.45, and 23.5K and 0.45, respectively, with significant correlations observed for each infrared channel.

[0155] 2) Correction techniques for forecast cloud image errors in numerical precipitation forecasts. A correction study was conducted on 0–4 hour forecast brightness temperature and precipitation forecasts. Case studies showed that after correction of the 0–4 hour forecast brightness temperature, the average root mean square error decreased by 5.4 K, the average absolute error decreased by 3.4 K, and the average correlation coefficient increased by 18.6%. Compared to the uncorrected forecasts, the TS score for 0–4 hour precipitation forecasts improved by 11.3%, 10.6%, and 73.0% for light rain, moderate rain, and heavy rain, respectively.

[0156] Experiments show that the method proposed in this application can effectively improve the accuracy of short-term numerical forecast cloud images and precipitation in areas such as offshore and plateau regions, and has significant application value.

[0157] In another embodiment of this application, a correction device for short-term numerical weather forecast data is also provided, such as... Figure 3 As shown, it includes:

[0158] The first acquisition module 11 is used to acquire a first infrared forecast cloud image for forecasting the weather for a first preset time period and a second infrared forecast cloud image for forecasting the weather for a second preset time period, wherein the first preset time period is before the current time and the second preset time period is after the current time.

[0159] The second acquisition module 12 is used to acquire the actual observed cloud image obtained by observing the weather during the first preset time period;

[0160] The determination module 13 is used to determine the evolution information of the first optical flow field corresponding to the second preset time period based on the first infrared forecast cloud map and the actual observed cloud map;

[0161] The correction module 14 is used to correct the second infrared forecast cloud image using the first optical flow field evolution information to obtain target weather forecast data.

[0162] Optionally, the determining module includes:

[0163] The first determining unit is used to determine multiple evolution intermediate points within the first preset time period;

[0164] The calculation unit is used to calculate the brightness temperature deviation of the first infrared forecast cloud image relative to the actual observed cloud image within the first preset time period for each of the evolution intermediate points using the optical flow method, so as to obtain the second optical flow field evolution information corresponding to the first preset time period.

[0165] The second determining unit is used to determine the first optical flow field evolution information corresponding to the second preset time period based on the second optical flow field evolution information.

[0166] Optionally, the second determining unit includes:

[0167] The first determining subunit is used to determine the brightness temperature optical flow field evolution information corresponding to the second preset prediction time period based on the second optical flow field evolution information;

[0168] The second determining subunit is used to determine the precipitation optical flow field evolution information corresponding to the second preset time period based on the brightness temperature optical flow field evolution information.

[0169] The third determining subunit is used to determine the brightness temperature optical flow field evolution information and the precipitation optical flow field evolution information as the first optical flow field evolution information.

[0170] Optionally, the first determining subunit is further configured to:

[0171] For brightness temperature, the second optical flow field evolution information is interpolated and extrapolated to obtain the third optical flow field evolution information corresponding to the third preset time period, which is located within the second preset time period;

[0172] For brightness temperature, the mean value of the optical flow field evolution information of the third optical flow field evolution information is calculated as the fourth optical flow field evolution information corresponding to the fourth preset time period. The fourth preset time period is located within the second preset time period and is later than the third preset time period.

[0173] The combination of the third optical flow field evolution information and the fourth optical flow field evolution information is determined as the brightness temperature optical flow field evolution information corresponding to the second preset prediction time period.

[0174] Optionally, the second determining subunit is further configured to:

[0175] For precipitation, the brightness temperature optical flow field evolution information is multiplied by the first preset weight to obtain the fifth optical flow field evolution information corresponding to the precipitation area;

[0176] For precipitation, the brightness temperature optical flow field evolution information is multiplied by the second preset weight to obtain the sixth optical flow field evolution information corresponding to the non-precipitation area;

[0177] The combination of the fifth and sixth optical flow field evolution information is determined as the precipitation optical flow field evolution information corresponding to the second preset time period.

[0178] Optionally, the correction module includes:

[0179] The first correction unit is used to correct the brightness temperature forecast data in the second infrared forecast cloud map by using the brightness temperature optical flow field evolution information in the first optical flow field evolution information, so as to obtain the first corrected data.

[0180] The second correction unit is used to correct the precipitation forecast data in the second infrared forecast cloud image by using the precipitation optical flow field evolution information in the first optical flow field evolution information, so as to obtain the second corrected data.

[0181] The third determining unit is used to determine the combination of the first corrected data and the second corrected data as the target weather forecast data.

[0182] Optionally, the first acquisition module includes:

[0183] The input unit is used to input the GFS global numerical weather prediction product and geographic dataset corresponding to the first preset time period into the preprocessing module of the meteorological research and forecasting WRF model to obtain the initial field and lateral boundary conditions of the WRF model.

[0184] The first operating unit is used to run the WRF mode based on the initial field and the lateral boundary conditions to obtain atmospheric environment data;

[0185] The second operating unit is used to input the atmospheric environment data into the fast radiative transfer (RTTOV) mode, input the satellite coefficient file and satellite cloud coefficient file contained in the RTTOV mode into the RTTOV mode, run the RTTOV mode, and obtain the infrared forecast cloud image.

[0186] In another embodiment of this application, an electronic device is also provided, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus.

[0187] Memory, used to store computer programs;

[0188] The processor, when executing a program stored in memory, implements the method for correcting short-term numerical weather forecast data as described in any of the foregoing method embodiments.

[0189] The electronic device provided in this embodiment of the invention allows the processor to execute a program stored in a memory to obtain first optical flow field evolution information based on a first infrared forecast cloud image and an actual observed cloud image within a first preset time period. Then, based on the first optical flow field evolution information, the processor corrects a second infrared forecast cloud image within a second preset time period. By using the first optical flow field evolution information, which reflects the position and intensity deviation of cloud clusters, to correct the second infrared forecast cloud image, the final target weather forecast data becomes more accurate, thereby improving the precision of the weather forecast data.

[0190] The communication bus 1140 mentioned in the above-mentioned electronic device can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus 1140 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0191] The communication interface 1120 is used for communication between the above-mentioned electronic device and other devices.

[0192] The memory 1130 may include random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0193] The processor 1110 mentioned above can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0194] In another embodiment of this application, a computer-readable storage medium is provided, on which a program for a method of correcting short-term numerical weather forecast data is stored. When the program for correcting short-term numerical weather forecast data is executed by a processor, it implements the steps of the method for correcting short-term numerical weather forecast data as described in any of the foregoing method embodiments.

[0195] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0196] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method of correcting short-range numerical weather prediction data, characterized by, The method comprises the following steps: obtaining a first infrared forecast cloud picture for forecasting weather in a first preset time period and a second infrared forecast cloud picture for forecasting weather in a second preset time period, the first preset time period being before a current time, and the second preset time period being after the current time; obtaining an actual observation cloud picture obtained by observing weather in the first preset time period; determining first optical flow field evolution information corresponding to the second preset time period based on the first infrared forecast cloud picture and the actual observation cloud picture; determining first optical flow field evolution information corresponding to the second preset time period based on the first infrared forecast cloud picture and the actual observation cloud picture comprises the following steps: determining a plurality of evolution intermediate points in the first preset time period; for each evolution intermediate point, calculating a brightness temperature deviation of the first infrared forecast cloud picture relative to the actual observation cloud picture in the first preset time period by using an optical flow method to obtain second optical flow field evolution information corresponding to the first preset time period; determining first optical flow field evolution information corresponding to the second preset time period based on the second optical flow field evolution information; determining first optical flow field evolution information corresponding to the second preset time period based on the second optical flow field evolution information comprises the following steps: for brightness temperature, determining brightness temperature optical flow field evolution information corresponding to the second preset forecast time period based on the second optical flow field evolution information; for precipitation, determining precipitation optical flow field evolution information corresponding to the second preset time period based on the brightness temperature optical flow field evolution information; determining the brightness temperature optical flow field evolution information and the precipitation optical flow field evolution information as the first optical flow field evolution information; correcting the second infrared forecast cloud picture by using the first optical flow field evolution information to obtain target weather forecast data; correcting the second infrared forecast cloud picture by using the first optical flow field evolution information to obtain target weather forecast data comprises the following steps: correcting brightness temperature prediction data in the second infrared forecast cloud picture by using brightness temperature optical flow field evolution information in the first optical flow field evolution information to obtain first correction data; correcting precipitation prediction data in the second infrared forecast cloud picture by using precipitation optical flow field evolution information in the first optical flow field evolution information to obtain second correction data; determining a combination of the first correction data and the second correction data as the target weather forecast data.

2. The method of claim 1, wherein, For brightness temperature, determining brightness temperature optical flow field evolution information corresponding to the second preset forecast time period based on the second optical flow field evolution information comprises the following steps: for brightness temperature, performing interpolation and extrapolation processing on the second optical flow field evolution information to obtain third optical flow field evolution information corresponding to a third preset time period, the third preset time period being in the second preset time period; for brightness temperature, calculating a mean value of the third optical flow field evolution information as fourth optical flow field evolution information corresponding to a fourth preset time period, the fourth preset time period being in the second preset time period, and the fourth preset time period being later than the third preset time period. The combination of the third optical flow field evolution information and the fourth optical flow field evolution information is determined as the brightness temperature optical flow field evolution information corresponding to a second preset prediction time period.

3. The method of claim 1, wherein, For precipitation, the brightness temperature optical flow field evolution information is multiplied by a first preset weight to obtain fifth optical flow field evolution information corresponding to a precipitation area; For precipitation, the brightness temperature optical flow field evolution information is multiplied by a second preset weight to obtain sixth optical flow field evolution information corresponding to a non-precipitation area; The combination of the fifth optical flow field evolution information and the sixth optical flow field evolution information is determined as the precipitation optical flow field evolution information corresponding to the second preset time period. The first infrared forecast cloud image for predicting the weather in the first preset time period is obtained, including:

4. The method of claim 1, wherein, The global numerical weather prediction product and geographical data set corresponding to the first preset time period are input into a pre-processing module of a weather research and forecast (WRF) model to obtain an initial field and a side boundary condition of the WRF model; The WRF model is run based on the initial field and the side boundary condition to obtain atmospheric environment data; The infrared forecast cloud image is obtained by inputting the atmospheric environment data into a rapid radiative transfer (RTTOV) model and inputting a satellite coefficient file and a satellite cloud coefficient file contained in the RTTOV model into the RTTOV model and running the RTTOV model. The first infrared forecast cloud image for predicting the weather in the first preset time period and the second infrared forecast cloud image for predicting the weather in the second preset time period are obtained, the first preset time period being before the current time, and the second preset time period being after the current time; 5. A correction device for short-range numerical weather prediction data, characterized in that The actual observation cloud image obtained by observing the weather in the first preset time period is obtained; The first optical flow field evolution information corresponding to the second preset time period is determined based on the first infrared forecast cloud image and the actual observation cloud image. The first optical flow field evolution information corresponding to the second preset time period is determined based on the second optical flow field evolution information. The first optical flow field evolution information corresponding to the second preset time period is determined based on the second optical flow field evolution information. The first optical flow field evolution information corresponding to the second preset time period is determined based on the second optical flow field evolution information. The first optical flow field evolution information corresponding to the second preset time period is determined based on the second optical flow field evolution information. The first optical flow field evolution information corresponding to the second preset time period is determined based on the second optical flow field evolution information. The first optical flow field evolution information corresponding to the second preset time period is determined based on the second optical flow field evolution information. ​ ​ ​ ​ A correction module is configured to correct the second infrared forecast cloud image by using the first optical flow field evolution information, so as to obtain target weather forecast data. The correction module comprises: A first correction unit is configured to correct brightness temperature forecast data in the second infrared forecast cloud image by using brightness temperature optical flow field evolution information in the first optical flow field evolution information, so as to obtain first correction data. A second correction unit is configured to correct precipitation forecast data in the second infrared forecast cloud image by using precipitation optical flow field evolution information in the first optical flow field evolution information, so as to obtain second correction data. A third determination unit is configured to determine a combination of the first correction data and the second correction data as the target weather forecast data.

6. An electronic device, comprising: The device comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory are in communication with each other through the communication bus. The memory is configured to store a computer program. The processor is configured to execute the program stored in the memory, so as to implement the short-term numerical weather forecast data correction method according to any one of claims 1-4.

7. A computer readable storage medium characterized in that, The computer readable storage medium stores a program of the short-term numerical weather forecast data correction method, and the program of the short-term numerical weather forecast data correction method is executed by the processor to implement the steps of the short-term numerical weather forecast data correction method according to any one of claims 1-4.

Images