CN114385600B - Downscaling correction method and device for ECMWF temperature data - Google Patents
Downscaling correction method and device for ECMWF temperature data Download PDFInfo
- Publication number
- CN114385600B CN114385600B CN202210284987.9A CN202210284987A CN114385600B CN 114385600 B CN114385600 B CN 114385600B CN 202210284987 A CN202210284987 A CN 202210284987A CN 114385600 B CN114385600 B CN 114385600B
- Authority
- CN
- China
- Prior art keywords
- data
- ecmwf
- resolution
- grid data
- initial high
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000012937 correction Methods 0.000 title claims abstract description 13
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 26
- 238000013135 deep learning Methods 0.000 claims abstract description 25
- 238000005070 sampling Methods 0.000 claims abstract description 25
- 230000004927 fusion Effects 0.000 claims abstract description 17
- 238000012545 processing Methods 0.000 claims abstract description 17
- 238000012795 verification Methods 0.000 claims description 22
- 238000000605 extraction Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 6
- 238000010219 correlation analysis Methods 0.000 claims description 6
- 238000004891 communication Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 5
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 230000035945 sensitivity Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000000802 evaporation-induced self-assembly Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000001556 precipitation Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Image Processing (AREA)
Abstract
The invention provides a downscaling correction method and a downscaling correction device for ECMWF temperature data, which relate to the technical field of data processing and comprise the following steps: acquiring ECMWF fine mesh data and CLDAS temperature data of an area to be processed, and determining target data in the ECMWF fine mesh data, wherein the resolutions of the ECMWF fine mesh data are all first resolutions; performing multi-channel fusion on target data to obtain four-dimensional grid data; the four-dimensional grid data are input into a deep learning network model to obtain initial high-resolution grid data, the deep learning network model is a network model formed by a residual error network with a regularization layer removed and a sub-pixel sampling algorithm, the resolution of the initial high-resolution grid data is a second resolution, and the second resolution is higher than the first resolution, so that the technical problems that the resolution and the prediction accuracy of the existing ECMWF temperature data are low are solved.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a downscaling correction method and device for ECMWF temperature data.
Background
High-resolution grid refinement is an important technical support for professional meteorological services, and meteorological data used in the current meteorological services have different time and spatial resolutions and are low in resolution.
Due to the fact that the resolution ratio is low, accuracy of temperature elements in the high-resolution grid is low, and a large image is generated for professional meteorological services, and therefore the high-resolution grid needs to be refined urgently.
No effective solution has been proposed to the above problems.
Disclosure of Invention
In view of this, the present invention provides a method and an apparatus for downscaling and correcting ECMWF temperature data, so as to alleviate the technical problems of low resolution and low prediction accuracy of the existing ECMWF temperature data.
In a first aspect, an embodiment of the present invention provides a downscaling correction method for ECMWF temperature data, including: acquiring ECMWF fine grid data and CLDAS temperature data of an area to be processed, and determining target data in the ECMWF fine grid data, wherein the target data comprises: the system comprises ECMWF temperature data, ECMWF wind speed data and terrain data, wherein the resolution of the ECMWF fine grid data is a first resolution; performing multi-channel fusion on the target number and the CLDAS temperature data to obtain four-dimensional grid data; inputting the four-dimensional grid data into a deep learning network model to obtain initial high-resolution grid data, wherein the deep learning network model is a network model formed by a residual error network without a regularization layer and a sub-pixel sampling algorithm, the resolution of the initial high-resolution grid data is a second resolution, and the second resolution is higher than the first resolution; verifying the initial high-resolution mesh data, and determining the initial high-resolution mesh data as target high-resolution mesh data when the verification is passed.
Further, determining target data in the ECMWF fine mesh data includes: determining ECMWF temperature data in the ECMWF fine grid data; determining feature data in the ECMWF fine grid data based on a correlation analysis algorithm, wherein the feature data is data with the highest correlation and contribution degree with the ECMWF temperature data, and the feature data comprises: the ECMWF wind speed data and the terrain data.
Further, inputting the four-dimensional grid data into a deep learning network model to obtain initial high-resolution grid data, including: performing feature extraction on the four-dimensional grid data by using a residual error network without a regularization layer to obtain high-frequency feature data of the four-dimensional grid data; superposing the high-frequency characteristic data to the four-dimensional grid data to obtain middle four-dimensional grid data; and sampling the intermediate four-dimensional grid data by utilizing a sub-pixel sampling algorithm to obtain the initial high-resolution grid data.
Further, validating the initial high resolution mesh data comprises: calculating target parameters of the initial high-resolution mesh data, wherein the target parameters include: a root mean square error and a peak signal-to-noise ratio between the ECMWF temperature data and the CLDAS temperature data in the initial high resolution grid data; if the root mean square error is smaller than a first preset threshold and the peak signal-to-noise ratio is larger than a second preset threshold, interpolating the ECMWF temperature data in the initial high-resolution grid data to a meteorological ground station of the area to be processed by using a bilinear interpolation algorithm, and determining an average error between the ECMWF temperature data interpolation in the initial high-resolution grid data and the meteorological ground station of the station; and if the average error is smaller than a third preset threshold value, the verification is passed.
In a second aspect, an embodiment of the present invention further provides a downscaling correction apparatus for ECMWF temperature data, including: the system comprises an acquisition unit, a fusion unit, a processing unit and a verification unit, wherein the acquisition unit is used for acquiring ECMWF fine mesh data and CLDAS temperature data of an area to be processed and determining target data in the ECMWF fine mesh data, and the target data comprises: the system comprises ECMWF temperature data, ECMWF wind speed data and terrain data, wherein the resolution of the ECMWF fine grid data is a first resolution; the fusion unit is used for performing multi-channel fusion on the target number and the CLDAS temperature data to obtain four-dimensional grid data; the processing unit is configured to input the four-dimensional grid data into a deep learning network model to obtain initial high-resolution grid data, where the deep learning network model is a network model formed by a residual network from which a regularization layer is removed and a sub-pixel sampling algorithm, and the resolution of the initial high-resolution grid data is a second resolution higher than the first resolution; the verification unit is configured to verify the initial high-resolution mesh data, and determine the initial high-resolution mesh data as target high-resolution mesh data when the verification passes.
Further, the obtaining unit is configured to: determining target data in the ECMWF fine grid data, and determining ECMWF temperature data in the ECMWF fine grid data; determining feature data in the ECMWF fine grid data based on a correlation analysis algorithm, wherein the feature data is data with the highest correlation and contribution degree with the ECMWF temperature data, and the feature data comprises: the ECMWF wind speed data and the terrain data.
Further, the processing unit is configured to: performing feature extraction on the four-dimensional grid data by using a residual error network without a regularization layer to obtain high-frequency feature data of the four-dimensional grid data; superposing the high-frequency characteristic data to the four-dimensional grid data to obtain middle four-dimensional grid data; and sampling the intermediate four-dimensional grid data by utilizing a sub-pixel sampling algorithm to obtain the initial high-resolution grid data.
Further, target parameters of the initial high resolution mesh data are calculated, wherein the target parameters include: a root mean square error and a peak signal to noise ratio between the ECMWF temperature data and the CLDAS temperature data in the initial high resolution grid data; if the root mean square error is smaller than a first preset threshold and the peak signal-to-noise ratio is larger than a second preset threshold, interpolating the ECMWF temperature data in the initial high-resolution grid data to a meteorological ground station of the area to be processed by using a bilinear interpolation algorithm, and determining an average error between the ECMWF temperature data interpolation in the initial high-resolution grid data and the meteorological ground station of the station; and if the average error is smaller than a third preset threshold value, the verification is passed.
In a third aspect, an embodiment of the present invention further provides an electronic device, including a memory and a processor, where the memory is used to store a program that supports the processor to execute the method in the first aspect, and the processor is configured to execute the program stored in the memory.
In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium.
In the embodiment of the invention, by acquiring ECMWF fine mesh data and CLDAS temperature data of an area to be processed, target data in the ECMWF fine mesh data is determined, wherein the target data comprises: the system comprises ECMWF temperature data, ECMWF wind speed data and terrain data, wherein the resolution of the ECMWF fine grid data is a first resolution; performing multi-channel fusion on the target number and the CLDAS temperature data to obtain four-dimensional grid data; inputting the four-dimensional grid data into a deep learning network model to obtain initial high-resolution grid data, wherein the deep learning network model is a network model formed by a residual error network with a regularization layer removed and a sub-pixel sampling algorithm, the resolution of the initial high-resolution grid data is a second resolution, and the second resolution is higher than the first resolution; and verifying the initial high-resolution grid data, and determining the initial high-resolution grid data as target high-resolution grid data under the condition that the verification is passed, so that the aim of performing fine processing on the ECMWF grid temperature data is fulfilled, the technical problem that the existing ECMWF temperature data is low in resolution and prediction accuracy is solved, and the technical effect of improving the resolution and prediction accuracy of the ECMWF temperature data is achieved.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flowchart of a downscaling correction method for ECMWF temperature data according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a downscaling correction apparatus for ECMWF temperature data according to an embodiment of the present invention;
fig. 3 is a schematic diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
To make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The first embodiment is as follows:
according to an embodiment of the present invention, there is provided an embodiment of a method for downscaling ECMWF temperature data, it being noted that the steps illustrated in the flowchart of the figures may be performed in a computer system, such as a set of computer-executable instructions, and that, although a logical order is illustrated in the flowchart, in some cases, the steps illustrated or described may be performed in an order different than here.
Fig. 1 is a flowchart of a downscaling method for correcting ECMWF temperature data according to an embodiment of the present invention, as shown in fig. 1, the method includes the following steps:
step S102, obtaining ECMWF fine grid data and CLDAS temperature data of an area to be processed, and determining target data in the ECMWF fine grid data, wherein the target data comprises: the system comprises ECMWF temperature data, ECMWF wind speed data and terrain data, wherein the resolution of the ECMWF fine grid data is a first resolution;
note that the ECMWF fine mesh data is fine mesh data with a resolution of 0.05 × 0.05 at European Center (ECMWF), and the CLDAS temperature data is temperature data provided by the CMA land data assimilation system.
Typically, CLDAS temperature data is 0.01 x 0.01.
ECMWF includes more than 20 features, including: temperature, humidity, terrain, rainfall, wind speed, wind direction, convective precipitation, surface air pressure, visibility, etc.
Step S104, performing multi-channel fusion on the target number and the CLDAS temperature data to obtain four-dimensional grid data;
step S106, inputting the four-dimensional grid data into a deep learning network model to obtain initial high-resolution grid data, wherein the deep learning network model is a network model formed by a residual error network with a regularization layer removed and a sub-pixel sampling algorithm, the resolution of the initial high-resolution grid data is a second resolution, and the second resolution is higher than the first resolution;
step S108, the initial high-resolution grid data are verified, and under the condition that the verification is passed, the initial high-resolution grid data are determined to be target high-resolution grid data.
In the embodiment of the invention, by acquiring ECMWF fine mesh data and CLDAS temperature data of an area to be processed, target data in the ECMWF fine mesh data is determined, wherein the target data comprises: the system comprises ECMWF temperature data, ECMWF wind speed data and terrain data, wherein the resolution of the ECMWF fine grid data is a first resolution; performing multi-channel fusion on the target number and the CLDAS temperature data to obtain four-dimensional grid data; inputting the four-dimensional grid data into a deep learning network model to obtain initial high-resolution grid data, wherein the deep learning network model is a network model formed by a residual error network with a regularization layer removed and a sub-pixel sampling algorithm, the resolution of the initial high-resolution grid data is a second resolution, and the second resolution is higher than the first resolution; and verifying the initial high-resolution grid data, and determining the initial high-resolution grid data as target high-resolution grid data under the condition that the verification is passed, so that the aim of performing fine processing on the ECMWF grid temperature data is fulfilled, the technical problem that the existing ECMWF temperature data is low in resolution and prediction accuracy is solved, and the technical effect of improving the resolution and prediction accuracy of the ECMWF temperature data is achieved.
In the embodiment of the present invention, step S102 includes the following steps:
step S11, determining ECMWF temperature data in the ECMWF fine mesh data;
step S12, determining feature data in the ECMWF fine grid data based on a correlation analysis algorithm, where the feature data is data with the highest correlation and contribution degree with the ECMWF temperature data, and the feature data includes: the ECMWF wind speed data and the terrain data.
In the embodiment of the present invention, the ECMWF fine mesh data includes 20 features, so that first, the ECMWF temperature data in the ECMWF fine mesh data is determined, and then, all the features of the ECMWF are considered in the deep learning network, and the deviation possibly in the training is more difficult to converge, so that we need to select an element with high temperature correlation from the 20 features. The temperature is very sensitive to terrain, the higher the tropospheric atmosphere is from the ground, the less ground long-wave radiant energy is absorbed. The air temperature therefore decreases with increasing altitude. The temperature is reduced by 0.65 ℃ on average every 100m of the whole tropospheric elevation, so that the topographic data is considered to be very necessary in the training of the model. Because the correlation between each feature in the weather is different, for the temperature element, the correlation analysis is performed by using Lasson, and finally 2 feature elements with the highest temperature correlation and contribution degree are selected from the ECMWF data: and combining the ECMWF wind speed data and the topographic data with the CLDAS temperature data corresponding to the same time in the previous day, so as to form a data combination of 4 characteristics, and finally performing multi-channel fusion on the target data to obtain four-dimensional grid data.
In the embodiment of the present invention, step S106 includes the following steps:
step S21, performing feature extraction on the four-dimensional grid data by using the residual error network without the regularization layer to obtain high-frequency feature data of the four-dimensional grid data;
step S22, superposing the high-frequency characteristic data to the four-dimensional grid data to obtain middle four-dimensional grid data;
step S23, sampling the intermediate four-dimensional grid data by using a sub-pixel sampling algorithm to obtain the initial high-resolution grid data.
In the embodiment of the invention, the residual error network without the regularization layer is used for extracting the high-frequency characteristics of the four-dimensional grid data, then the extracted high-frequency characteristics data are superposed into the input four-dimensional grid data to obtain intermediate four-dimensional grid data, and finally, a sub-pixel up-sampling algorithm is adopted for sampling and restoring the intermediate four-dimensional grid data to obtain the initial high-resolution grid data.
In the feature extraction, ResNet50 is adopted to extract features, in a residual error network, a BN (Batch Normalization) layer is removed, because the Batch regularization (BN) layer mainly has a standardized feature effect, the Batch regularization (BN) has an excellent effect on network convergence, gradient propagation can be smoothly performed, certain flexibility of the extracted features is lost, the calculated amount is increased, and because super-resolution reconstruction belongs to the problem of more basic problem research (Low level) compared with the problems of image classification and the like, the Batch regularization (BN) layer is removed from a residual error module, and specific details of the performance improved by the Batch regularization (BN) layer are removed.
Finally, sub-pixel up-sampling is adopted for the middle four-dimensional grid data, sub-pixels obtained through sampling are added into the network, the detail quality of the image can be improved, the accuracy after interpolation is improved, a large number of 0 complementing areas exist in general deconvolution, and the result is possibly harmful. Therefore, the pixel shuffle realizes the reconstruction from a low-resolution image to a high-resolution image by means of sub-pixel convolution.
Since the details and the edge parts of the terrain are abrupt and change rapidly, wherein the details and the edge parts represent high-frequency components on a frequency domain, high-frequency feature data extracted from the ECMWF can represent the details in the terrain, and then the high-frequency feature data are superposed into the four-dimensional grid data, namely the information of the terrain features is superposed on the ECMWF data to obtain the four-dimensional grid data on which the terrain feature information is superposed.
In the embodiment of the present invention, step S108 includes the following steps:
step S31, verifying the initial high resolution mesh data, including:
step S32, calculating target parameters of the initial high resolution mesh data, wherein the target parameters include: a root mean square error and a peak signal to noise ratio between the ECMWF temperature data and the CLDAS temperature data in the initial high resolution grid data;
step S33, if the root mean square error is smaller than a first preset threshold and the peak signal-to-noise ratio is greater than a second preset threshold, interpolating the ECMWF temperature data in the initial high resolution grid data to the meteorological ground station of the area to be processed by using a bilinear interpolation algorithm, and determining an average error between the ECMWF temperature data interpolation in the initial high resolution grid data and the meteorological ground station of the station;
and step S33, if the average error is smaller than a third preset threshold, the verification is passed.
In the embodiment of the invention, the initial high-resolution grid data is verified, wherein two modes are adopted for verification, one mode is that the initial high-resolution grid data (0.01 × 0.01 resolution) is verified, and the verified indexes mainly comprise two aspects of root mean square error and peak signal-to-noise ratio; the second is to interpolate the initial high resolution grid data to the site and compare it with the meteorological ground station in the area to be processed to check the average error.
The root mean square error is the square root of the sum of the squares of the ECMWF temperature data and the CLDAS temperature data in the initial high resolution grid data, and the root mean square error is well reflective of the deviation between the initial high resolution grid data and the CLDAS temperature data. The following formula isRoot mean square error of
Is the average of the ECMWF temperature data for the ith grid in the initial high resolution grid data,
is the square of the temperature data for the ith grid in the CLDAS temperature data.
The peak signal-to-noise ratio is the most common and most widely used image objective evaluation index, but is based on the error between corresponding pixel points, namely, the image quality evaluation based on error sensitivity. In image processing, PSNR calculation is often required for objective evaluation of an image. PSNR is an objective measure of image distortion or noise level. The larger the PSNR value between 2 images, the more similar. The general standard is 30dB, and the image degradation below 30dB is obvious. Since the visual characteristics of human eyes (the human eyes have high sensitivity to contrast differences with low spatial frequency, the human eyes have high sensitivity to luminance contrast differences, and the human eyes have high chroma, the perception result of one region by the human eyes is affected by the surrounding adjacent regions, and the like) are not considered, the situation that the evaluation result is inconsistent with the subjective feeling of the human often occurs.
Wherein the content of the first and second substances,
for peak signal-to-noise ratio, MAX represents the maximum value of image color, and MSE is the average error.
Interpolating the initial high-resolution grid data to a station and comparing the station with a meteorological ground station of the area to be processed, and checking an average error, wherein the method specifically comprises the following steps:
interpolating to a ground station, considering a vertical decrement rate of the temperature of the terrain, when the grid temperature is detected based on the observation of a meteorological ground station in a region to be processed, sometimes the station is at the top of a mountain, surrounding grid points are at the feet of the mountain, or the station is at the valley, surrounding grids are at the waist or the top of the mountain, and at the moment, the initial high-resolution grid data is interpolated to the station by adopting bilinear interpolation, so that a systematic deviation exists. During temperature detection, the deviation is deducted, so that the initial high-resolution grid data is compared with an interpolation method, and for illustration, features of the terrain and elements with strong correlation are fused in deep learning, so that the terrain complex region can be improved.
Since the 0.05 × 0.05 resolution data of the ECMWF has a problem of a terrain error when the data is scaled to 0.01 × 0.01 resolution data, in the embodiment of the present invention, the feature extraction is performed on the data of the low-resolution ECMWF through the deep learning convolutional neural network, and the up-sampling is performed on the data of the low-resolution ECMWF by combining the data such as geographic information, so as to obtain the 0.01 × 0.01-resolution ECMWF data.
Example two:
the embodiment of the present invention further provides a downscaling correcting device for ECMWF temperature data, where the downscaling correcting device for ECMWF temperature data is used to execute the downscaling correcting method for ECMWF temperature data provided in the foregoing embodiments of the present invention, and the following is a specific description of the downscaling correcting device for ECMWF temperature data provided in the embodiments of the present invention.
As shown in fig. 2, fig. 2 is a schematic diagram of the downscaling correction device for ECMWF temperature data, where the downscaling correction device for ECMWF temperature data includes: an acquisition unit 10, a fusion unit 20, a processing unit 30 and a verification unit 40.
The acquiring unit 10 is configured to acquire ECMWF fine mesh data and CLDAS temperature data of an area to be processed, and determine target data in the ECMWF fine mesh data, where the target data includes: the system comprises ECMWF temperature data, ECMWF wind speed data and terrain data, wherein the resolution of the ECMWF fine grid data is a first resolution;
the fusion unit 20 is configured to perform multi-channel fusion on the target number and the CLDAS temperature data to obtain four-dimensional grid data;
the processing unit 30 is configured to input the four-dimensional mesh data into a deep learning network model to obtain initial high-resolution mesh data, where the deep learning network model is a network model formed by a residual network from which a regularization layer is removed and a sub-pixel sampling algorithm, and the resolution of the initial high-resolution mesh data is a second resolution higher than the first resolution;
the verification unit 40 is configured to verify the initial high-resolution mesh data, and in a case that the verification passes, determine the initial high-resolution mesh data as target high-resolution mesh data.
In the embodiment of the invention, by acquiring ECMWF fine mesh data and CLDAS temperature data of an area to be processed, target data in the ECMWF fine mesh data is determined, wherein the target data comprises: the system comprises ECMWF temperature data, ECMWF wind speed data and terrain data, wherein the resolution of the ECMWF fine grid data is a first resolution; performing multi-channel fusion on the target number and the CLDAS temperature data to obtain four-dimensional grid data; inputting the four-dimensional grid data into a deep learning network model to obtain initial high-resolution grid data, wherein the deep learning network model is a network model formed by a residual error network without a regularization layer and a sub-pixel sampling algorithm, the resolution of the initial high-resolution grid data is a second resolution, and the second resolution is higher than the first resolution; and verifying the initial high-resolution grid data, and determining the initial high-resolution grid data as target high-resolution grid data under the condition that the verification is passed, so that the aim of performing fine processing on the ECMWF grid temperature data is fulfilled, the technical problem that the resolution and the prediction accuracy of the conventional ECMWF temperature data are low is solved, and the technical effect of improving the resolution and the prediction accuracy of the ECMWF temperature data is realized.
Example three:
an embodiment of the present invention further provides an electronic device, which includes a memory and a processor, where the memory is used to store a program that supports the processor to execute the method in the first embodiment, and the processor is configured to execute the program stored in the memory.
Referring to fig. 3, an embodiment of the present invention further provides an electronic device 100, including: the device comprises a processor 50, a memory 51, a bus 52 and a communication interface 53, wherein the processor 50, the communication interface 53 and the memory 51 are connected through the bus 52; the processor 50 is arranged to execute executable modules, such as computer programs, stored in the memory 51.
The Memory 51 may include a high-speed Random Access Memory (RAM) and may also include a non-volatile Memory (non-volatile Memory), such as at least one disk Memory. The communication connection between the network element of the system and at least one other network element is realized through at least one communication interface 53 (which may be wired or wireless), and the internet, a wide area network, a local network, a metropolitan area network, and the like can be used.
The bus 52 may be an ISA bus, PCI bus, EISA bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one double-headed arrow is shown in FIG. 3, but this does not indicate only one bus or one type of bus.
The memory 51 is used for storing a program, the processor 50 executes the program after receiving an execution instruction, and the method executed by the apparatus defined by the flow process disclosed in any of the foregoing embodiments of the present invention may be applied to the processor 50, or implemented by the processor 50.
The processor 50 may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware or instructions in the form of software in the processor 50. The Processor 50 may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; the Integrated Circuit may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic device, or discrete hardware components. The various methods, steps and logic blocks disclosed in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present invention may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software modules may be located in ram, flash, rom, prom, or eprom, registers, etc. as is well known in the art. The storage medium is located in the memory 51, and the processor 50 reads the information in the memory 51 and completes the steps of the method in combination with the hardware thereof.
Example four:
the embodiment of the present invention further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the computer program performs the steps of the method in the first embodiment.
In addition, in the description of the embodiments of the present invention, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of description and simplicity of description, but do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only one logical division, and there may be other divisions when actually implemented, and for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of devices or units through some communication interfaces, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.
Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present invention, which are used for illustrating the technical solutions of the present invention and not for limiting the same, and the protection scope of the present invention is not limited thereto, although the present invention is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: those skilled in the art can still make modifications or changes to the embodiments described in the foregoing embodiments, or make equivalent substitutions for some features, within the scope of the disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the embodiments of the present invention, and they should be construed as being included therein. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (8)
1. A downscaling correction method for ECMWF temperature data is characterized by comprising the following steps:
acquiring ECMWF fine grid data and CLDAS temperature data of an area to be processed, and determining target data in the ECMWF fine grid data, wherein the target data comprises: the system comprises ECMWF temperature data, ECMWF wind speed data and terrain data, wherein the resolution of the ECMWF fine grid data is a first resolution;
performing multi-channel fusion on the target data and the CLDAS temperature data to obtain four-dimensional grid data;
inputting the four-dimensional grid data into a deep learning network model to obtain initial high-resolution grid data, wherein the deep learning network model is a network model formed by a residual error network with a regularization layer removed and a sub-pixel sampling algorithm, the resolution of the initial high-resolution grid data is a second resolution, and the second resolution is higher than the first resolution;
verifying the initial high-resolution grid data, and determining the initial high-resolution grid data as target high-resolution grid data under the condition that the verification is passed;
wherein validating the initial high resolution mesh data comprises:
calculating target parameters of the initial high-resolution mesh data, wherein the target parameters include: a root mean square error and a peak signal to noise ratio between the ECMWF temperature data and the CLDAS temperature data in the initial high resolution grid data;
if the root mean square error is smaller than a first preset threshold and the peak signal-to-noise ratio is larger than a second preset threshold, interpolating the ECMWF temperature data in the initial high-resolution grid data to a meteorological ground station of the area to be processed by using a bilinear interpolation algorithm, and determining an average error between the ECMWF temperature data interpolation in the initial high-resolution grid data and the meteorological ground station of the station;
and if the average error is smaller than a third preset threshold value, the verification is passed.
2. The method of claim 1, wherein determining target data in the ECMWF fine mesh data comprises:
determining ECMWF temperature data in the ECMWF fine grid data;
determining feature data in the ECMWF fine mesh data based on a correlation analysis algorithm, wherein the feature data are data with the highest correlation and contribution degrees with the ECMWF temperature data, and the feature data comprise: the ECMWF wind speed data and the terrain data.
3. The method of claim 2, wherein inputting the four-dimensional mesh data into a deep learning network model resulting in initial high resolution mesh data comprises:
performing feature extraction on the four-dimensional grid data by using a residual error network without a regularization layer to obtain high-frequency feature data of the four-dimensional grid data;
superposing the high-frequency characteristic data to the four-dimensional grid data to obtain middle four-dimensional grid data;
and sampling the intermediate four-dimensional grid data by utilizing a sub-pixel sampling algorithm to obtain the initial high-resolution grid data.
4. An apparatus for downscaling correction of ECMWF temperature data, comprising: an acquisition unit, a fusion unit, a processing unit and a verification unit, wherein,
the acquiring unit is configured to acquire ECMWF fine mesh data and CLDAS temperature data of an area to be processed, and determine target data in the ECMWF fine mesh data, where the target data includes: the system comprises ECMWF temperature data, ECMWF wind speed data and terrain data, wherein the resolution of the ECMWF fine grid data is a first resolution;
the fusion unit is used for performing multi-channel fusion on the target data and the CLDAS temperature data to obtain four-dimensional grid data;
the processing unit is configured to input the four-dimensional grid data into a deep learning network model to obtain initial high-resolution grid data, where the deep learning network model is a network model formed by a residual network from which a regularization layer is removed and a sub-pixel sampling algorithm, and a resolution of the initial high-resolution grid data is a second resolution higher than the first resolution;
the verification unit is used for verifying the initial high-resolution grid data, and determining the initial high-resolution grid data as target high-resolution grid data under the condition that the verification is passed;
wherein the verification unit is configured to:
calculating target parameters of the initial high-resolution mesh data, wherein the target parameters include: a root mean square error and a peak signal to noise ratio between the ECMWF temperature data and the CLDAS temperature data in the initial high resolution grid data;
if the root mean square error is smaller than a first preset threshold and the peak signal-to-noise ratio is larger than a second preset threshold, interpolating the ECMWF temperature data in the initial high-resolution grid data to a meteorological ground station of the area to be processed by using a bilinear interpolation algorithm, and determining an average error between the ECMWF temperature data interpolation in the initial high-resolution grid data and the meteorological ground station of the station;
and if the average error is smaller than a third preset threshold value, the verification is passed.
5. The apparatus of claim 4, wherein the obtaining unit is configured to:
determining target data in the ECMWF fine grid data, and determining ECMWF temperature data in the ECMWF fine grid data;
determining feature data in the ECMWF fine grid data based on a correlation analysis algorithm, wherein the feature data is data with the highest correlation and contribution degree with the ECMWF temperature data, and the feature data comprises: the ECMWF wind speed data and the terrain data.
6. The apparatus of claim 5, wherein the processing unit is configured to:
performing feature extraction on the four-dimensional grid data by using a residual error network without a regularization layer to obtain high-frequency feature data of the four-dimensional grid data;
superposing the high-frequency characteristic data to the four-dimensional grid data to obtain middle four-dimensional grid data;
and sampling the intermediate four-dimensional grid data by utilizing a sub-pixel sampling algorithm to obtain the initial high-resolution grid data.
7. An electronic device comprising a memory for storing a program that enables a processor to perform the method of any of claims 1 to 3 and a processor configured to execute the program stored in the memory.
8. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method according to any one of the claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210284987.9A CN114385600B (en) | 2022-03-23 | 2022-03-23 | Downscaling correction method and device for ECMWF temperature data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210284987.9A CN114385600B (en) | 2022-03-23 | 2022-03-23 | Downscaling correction method and device for ECMWF temperature data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114385600A CN114385600A (en) | 2022-04-22 |
| CN114385600B true CN114385600B (en) | 2022-07-19 |
Family
ID=81205575
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210284987.9A Active CN114385600B (en) | 2022-03-23 | 2022-03-23 | Downscaling correction method and device for ECMWF temperature data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114385600B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116663432B (en) * | 2023-07-28 | 2023-10-27 | 中国气象局公共气象服务中心(国家预警信息发布中心) | Hundred-meter height wind speed forecast correction downscaling method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110619605A (en) * | 2019-09-17 | 2019-12-27 | 中国气象局公共气象服务中心(国家预警信息发布中心) | Parallel meteorological element downscaling method and device, electronic equipment and storage medium |
| CN112541654A (en) * | 2020-11-09 | 2021-03-23 | 华能新能源股份有限公司 | Regional wind energy resource fine evaluation method combining regional climate, local daily change and multi-scale interaction characteristics |
| CN113392365A (en) * | 2021-05-12 | 2021-09-14 | 中国电力科学研究院有限公司 | High-resolution meteorological grid data generation method and system |
| CN113536576A (en) * | 2021-07-20 | 2021-10-22 | 中国气象局气象干部培训学院四川分院(四川省信息工程学校) | Method and system for correcting distance integral power statistics of numerical forecast product |
| CN113807432A (en) * | 2021-09-16 | 2021-12-17 | 成都卡普数据服务有限责任公司 | Air temperature forecast data correction method based on deep learning |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110738252B (en) * | 2019-10-14 | 2020-08-14 | 广州地理研究所 | Space autocorrelation machine learning satellite precipitation data downscaling method and system |
-
2022
- 2022-03-23 CN CN202210284987.9A patent/CN114385600B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110619605A (en) * | 2019-09-17 | 2019-12-27 | 中国气象局公共气象服务中心(国家预警信息发布中心) | Parallel meteorological element downscaling method and device, electronic equipment and storage medium |
| CN112541654A (en) * | 2020-11-09 | 2021-03-23 | 华能新能源股份有限公司 | Regional wind energy resource fine evaluation method combining regional climate, local daily change and multi-scale interaction characteristics |
| CN113392365A (en) * | 2021-05-12 | 2021-09-14 | 中国电力科学研究院有限公司 | High-resolution meteorological grid data generation method and system |
| CN113536576A (en) * | 2021-07-20 | 2021-10-22 | 中国气象局气象干部培训学院四川分院(四川省信息工程学校) | Method and system for correcting distance integral power statistics of numerical forecast product |
| CN113807432A (en) * | 2021-09-16 | 2021-12-17 | 成都卡普数据服务有限责任公司 | Air temperature forecast data correction method based on deep learning |
Non-Patent Citations (3)
| Title |
|---|
| Ruian TIE 等.CLDASSD: Reconstructing Fine Textures of the Temperature Field Using Super-Resolution Technology.《ADVANCES IN ATMOSPHERIC SCIENCES》.2022,第39卷全文. * |
| 董玲等.ECMWF细网格2 m温度在抚州市预报能力检验.《江西科学》.2020,(第02期), * |
| 齐铎等.基于格点的中国东北中北部2 m温度数值预报检验及偏差订正.《干旱气象》.2020,(第01期), * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114385600A (en) | 2022-04-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109345456B (en) | Generation countermeasure network training method, image processing method, device, and storage medium | |
| CN112001374B (en) | Cloud detection method and device for hyperspectral image | |
| CN109360190B (en) | Building damage detection method and device based on image superpixel fusion | |
| TW200835351A (en) | Structured smoothing for superresolution of multispectral imagery based on registered panchromatic image | |
| CN114385600B (en) | Downscaling correction method and device for ECMWF temperature data | |
| CN114972191A (en) | Method and device for detecting farmland change | |
| CN109522899B (en) | Detection method and device for ripe coffee fruits and electronic equipment | |
| US20220198710A1 (en) | Methods for analysis of an image and a method for generating a dataset of images for training a machine-learned model | |
| CN111931744B (en) | Method and device for detecting change of remote sensing image | |
| CN109801212A (en) | Fish-eye image splicing method based on SIFT features | |
| CN114332182B (en) | SAR image registration method, equipment and medium based on multi-feature constraint | |
| CN105488759A (en) | Image super-resolution reconstruction method based on local regression model | |
| CN115512232A (en) | Crop seed germination condition identification model, construction method and application thereof | |
| Yu et al. | Performance evaluation of edge-directed interpolation methods for noise-free images | |
| Sustika et al. | Generative adversarial network with residual dense generator for remote sensing image super resolution | |
| CN111723753A (en) | Satellite remote sensing image strip removing method and device and electronic equipment | |
| CN108776958B (en) | Mix the image quality evaluating method and device of degraded image | |
| JP2017103756A5 (en) | ||
| CN112950479B (en) | Image gray level region stretching algorithm | |
| Erfanifard et al. | A robust approach to generate canopy cover maps using UltraCam-D derived orthoimagery classified by support vector machines in Zagros woodlands, West Iran | |
| CN114820376A (en) | Fusion correction method and device for stripe noise, electronic equipment and storage medium | |
| CN115082993A (en) | Face biopsy method and device based on mouth opening action | |
| CN110136061B (en) | Resolution improving method and system based on depth convolution prediction and interpolation | |
| CN114488070B (en) | Radar echo extrapolation method and device based on deep learning model | |
| CN112529956B (en) | Blackbody position detection method, electronic device and computer readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: A downscaling correction method and device for ECMWF temperature data Granted publication date: 20220719 Pledgee: Beijing first financing Company limited by guarantee Pledgor: Beijing Hongxiang Technology Co.,Ltd. Registration number: Y2024980012246 |