Production of surface shortwave radiation products, long-short-term memory network training methods and devices

Abstract

The present disclosure discloses a surface shortwave radiation product production method and device, and a long short-term memory network training method and device, which relate to the field of aerospace technology. A specific implementation scheme is as follows: extracting a first shortwave radiation data set and a second shortwave radiation data set that are time-aligned from first surface shortwave radiation data collected by a satellite; obtaining a photovoltaic shortwave radiation data set that is time-aligned with the first shortwave radiation data set collected by a ground photovoltaic radiometer; determining an albedo data set based on a satellite albedo product and a surface land cover product; obtaining an input shortwave radiation data sequence based on the first shortwave radiation data set, the second shortwave radiation data set, the albedo data set, the photovoltaic shortwave radiation data set, and the coordinate vector points of the ground photovoltaic radiometer; inputting the input shortwave radiation data sequence into a pre-trained long short-term memory network to obtain an initial surface shortwave radiation product; and obtaining a target surface shortwave radiation product based on the initial surface shortwave radiation product.

Description

Technical Field

[0001] The present invention belongs to the field of aerospace technology, in particular to the field of satellite remote sensing technology, and more particularly to a method and device for producing surface shortwave radiation products, a method and device for training long and short-term memory networks, electronic equipment, and a computer-readable storage medium. Background Art

[0002] Surface shortwave radiation is a crucial component of solar radiation, encompassing ultraviolet, visible, and near-infrared radiation, with wavelengths ranging from 0.2 to 4 microns. The vast majority of solar energy reaches the Earth in the form of shortwave radiation, making it the primary source of energy received by the Earth's surface. Shortwave radiation is a key factor influencing surface temperature, soil evaporation, vegetation photosynthesis, and the Earth's surface energy balance, playing a fundamental role in natural processes and engineering applications.

[0003] In the field of photovoltaic power generation, the intensity of shortwave radiation energy directly determines the power generation efficiency of photovoltaic modules. Its temporal and spatial variability has a significant impact on power generation assessment and grid load forecasting. Especially in the context of high penetration of renewable energy, rapid changes in shortwave radiation can cause large fluctuations in photovoltaic output, increasing the difficulty of grid regulation and even threatening grid operation safety. Therefore, obtaining accurate and stable shortwave radiation data is not only crucial for the design and operation of photovoltaic power plants, but also has important significance for ensuring the stability and security of the power system.

[0004] Currently, surface shortwave radiation data is primarily acquired through ground-based instruments and satellite remote sensing. Ground-based instruments, such as radiometers, can provide high-precision, real-time data, but their deployment costs are high and their coverage is limited. In contrast, satellite remote sensing offers the advantage of wide coverage and high-frequency radiation data, but its spatial resolution is lower. Consequently, existing surface radiation products face a conflict between monitoring accuracy and wide-area coverage.

[0005] The information disclosed in this background technology section is only intended to enhance understanding of the overall background of the invention and should not be regarded as an admission or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the Invention

[0006] The purpose of the present disclosure is to solve the technical problem that existing surface shortwave radiation products cannot take into account both temporal and spatial resolution, and to provide a surface shortwave radiation product production method and device, a long-short-term memory network training method and device, an electronic device, and a computer-readable storage medium.

[0007] A first aspect of the present disclosure provides a method for producing a surface shortwave radiation product, the method comprising: extracting a time-aligned first shortwave radiation dataset and a second shortwave radiation dataset from first surface shortwave radiation data collected by a satellite, the first shortwave radiation dataset comprising at least one first shortwave radiation data, the second shortwave radiation dataset comprising at least one second shortwave radiation data, the first time resolution of the first shortwave radiation data being greater than the second time resolution of the second shortwave radiation data, the first spatial coverage of the first shortwave radiation data being less than the second spatial coverage of the second shortwave radiation data, and the first surface shortwave radiation data having a first spatial resolution; acquiring a photovoltaic shortwave radiation dataset collected by a ground photovoltaic radiometer that is time-aligned with the first shortwave radiation dataset; determining an albedo dataset of pixels having a second spatial resolution based on a satellite albedo product and a surface land cover product, the satellite albedo product The first spatial resolution is smaller than the land type spatial resolution of the surface land cover product, and the second spatial resolution is smaller than the land type spatial resolution and larger than the first spatial resolution; based on the first shortwave radiation dataset, the second shortwave radiation dataset, the albedo dataset, the photovoltaic shortwave radiation dataset, and the coordinate vector points of the ground photovoltaic radiometer, an input shortwave radiation data sequence with a first temporal resolution and a second spatial resolution is obtained; the input shortwave radiation data sequence is input into a pre-trained long short-term memory network to obtain an initial surface shortwave radiation product with a first temporal resolution, a second spatial resolution, and belonging to the second spatial coverage range, and the long short-term memory network is used to predict the surface shortwave radiation product at future moments and belonging to the second spatial coverage range; based on the initial surface shortwave radiation product, a target surface shortwave radiation product with a third spatial resolution is obtained, and the third spatial resolution is larger than the second spatial resolution.

[0008] A second aspect of the present disclosure provides a long short-term memory network training method, the method comprising: obtaining a data training set, the data training set comprising: a photovoltaic shortwave radiation value set, a surface reflectivity value set, a second shortwave radiation value set, and a first shortwave radiation value set aligned at minute moments, the first spatial coverage range of the first shortwave radiation value being smaller than the second spatial coverage range of the second shortwave radiation value, the first shortwave radiation value, the second shortwave radiation value, the surface reflectivity value set, and the photovoltaic shortwave radiation value set all having a first temporal resolution and a second spatial resolution; constructing a plurality of input sequences and output data pairs based on the data training set; training an initial long short-term memory network based on the input sequences and the output data pairs to obtain a trained long short-term memory network, the long short-term memory network being used to preset a surface shortwave radiation product having the first temporal resolution, the second spatial resolution, and belonging to the second spatial coverage range at a future moment.

[0009] According to a third aspect of the present disclosure, there is provided a surface shortwave radiation product production device, which includes: an extraction unit, configured to extract a first shortwave radiation dataset and a second shortwave radiation dataset that are time-aligned from first surface shortwave radiation data collected by a satellite, the first shortwave radiation dataset including at least one first shortwave radiation data, the second shortwave radiation dataset including at least one second shortwave radiation data, the first time resolution of the first shortwave radiation data being greater than the second time resolution of the second shortwave radiation data, and the first spatial coverage of the first shortwave radiation data being less than the second spatial coverage of the second shortwave radiation data, and the first surface shortwave radiation data having a first spatial resolution; an acquisition unit, configured to acquire a photovoltaic shortwave radiation dataset collected by a ground photovoltaic radiometer that is time-aligned with the first shortwave radiation dataset; a determination unit, configured to determine an albedo dataset of pixels having a second spatial resolution based on a satellite albedo product and a surface land cover product, the satellite albedo product having a first spatial resolution. The spatial resolution is smaller than the land type spatial resolution of the surface land cover product, and the second spatial resolution is smaller than the land type spatial resolution and larger than the first spatial resolution; the sequence obtaining unit is configured to obtain an input shortwave radiation data sequence with a first time resolution and a second spatial resolution based on the first shortwave radiation dataset, the second shortwave radiation dataset, the albedo dataset, the photovoltaic shortwave radiation dataset and the ground fiber optic radiometer coordinate vector points; the input unit is configured to input the input shortwave radiation data sequence into a pre-trained long short-term memory network to obtain an initial surface shortwave radiation product with a first time resolution, a second spatial resolution and belonging to a second spatial coverage range, and the long short-term memory network is used to predict the surface shortwave radiation product at a future moment and belonging to the second spatial coverage range; the product obtaining unit is configured to obtain a target surface shortwave radiation product with a third spatial resolution based on the initial surface shortwave radiation product, and the third spatial resolution is larger than the second spatial resolution.

[0010] A fourth aspect of the present disclosure provides a long short-term memory network training device, which includes: a data acquisition unit, configured to acquire a data training set, the data training set including: a photovoltaic shortwave radiation value set, a surface reflectivity value set, a second shortwave radiation value set and a first shortwave radiation value set aligned at minute moments, the first spatial coverage range of the first shortwave radiation value being smaller than the second spatial coverage range of the second shortwave radiation value, the first shortwave radiation value, the second shortwave radiation value, the surface reflectivity value set and the photovoltaic shortwave radiation value set all having a first temporal resolution and a second spatial resolution; a construction unit, configured to construct an input sequence and an output data pair based on the data training set; a training unit, configured to train an initial long short-term memory network based on the input sequence and the output data pair to obtain a trained long short-term memory network, the long short-term memory network being used to preset a surface shortwave radiation product having the first temporal resolution, the second spatial resolution and belonging to the second spatial coverage range at a future moment.

[0011] A fifth aspect of the present disclosure provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the method of the first aspect or the second aspect are implemented.

[0012] A sixth aspect of the present disclosure provides a computer-readable storage medium, which implements the steps of the method of the first aspect or the second aspect when the computer program is executed by a processor.

[0013] Compared with the existing technology, the technical effects achieved by the present invention are as follows: by constructing a fusion framework of multi-satellite and ground radiometer data and adopting a long short-term memory network, high temporal resolution data and data of the second spatial coverage range are effectively integrated, and the limitation of traditional single-satellite payloads that cannot take into account both temporal and spatial resolutions is broken through, and for the first time, standardized product output with global coverage and temporal and spatial resolution reaching the first temporal resolution / second spatial resolution is achieved; land cover data with land type spatial resolution is introduced into the albedo correction process, and the reconstruction and fine correction of sub-kilometer scale albedo are realized, thereby significantly improving the spatial accuracy and resolution capability of surface shortwave radiation; based on the albedo data of the second spatial resolution, data is supplemented for the satellite-uncovered area that generates the first shortwave radiation data set to generate a global shortwave radiation intermediate product of the second spatial resolution; then, terrain radiation correction is performed in combination with the third spatial resolution data to generate a target surface shortwave radiation product of the third spatial resolution, which takes into account both the radiation estimation accuracy and the adaptability to areas with complex terrain. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 is a flow chart of an embodiment of a method for producing a surface shortwave radiation product according to the present disclosure;

[0015] Figure 2It is a schematic diagram of the distribution of surface shortwave radiation values ​​of the surface shortwave radiation product of the first satellite disclosed in the present invention;

[0016] Figure 3 It is a schematic diagram of the distribution of surface shortwave radiation values ​​of the target surface shortwave radiation product disclosed in this disclosure;

[0017] Figure 4 is a flow chart of an embodiment of a long short-term memory network training method according to the present disclosure;

[0018] Figure 5 It is a structural schematic diagram of an embodiment of a surface shortwave radiation product production device according to the present disclosure;

[0019] Figure 6 1 is a schematic structural diagram of an embodiment of a long short-term memory network training device according to the present disclosure;

[0020] Figure 7 It is a block diagram of an electronic device used to implement the surface shortwave radiation product production method or the long short-term memory network training method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0021] Unless expressly stated otherwise, throughout the specification and claims, the term "comprise" or variations such as "include" or "comprising", etc., will be understood to include the stated elements or components but not to exclude other elements or other components.

[0022] The technical solutions of the present invention are described below by means of specific embodiments. It should be understood that one or more steps mentioned in the present invention do not exclude the presence of other methods and steps before and after the combination step, or other methods and steps may be inserted between these explicitly mentioned steps. It should also be understood that these examples are only used to illustrate the present invention and are not used to limit the scope of the present invention. Unless otherwise specified, the numbering of each method step is only for the purpose of identifying each method step, and does not limit the order of arrangement of each method or the scope of implementation of the present invention. Changes or adjustments in their relative relationships can also be regarded as the scope of implementation of the present invention without substantial changes in the technical content.

[0023] The sources of the raw materials and instruments used in the examples are not particularly limited and can be purchased from the market or prepared according to conventional methods known to those skilled in the art.

[0024] Current major surface shortwave radiation products, including mainstream radiation products, have relatively low temporal and spatial resolution, hindering their precise application in scenarios such as photovoltaic power generation. This suggests that existing surface shortwave radiation products and estimates still have shortcomings. First, there is a trade-off between temporal and spatial resolution. Meteorological satellites can provide high-frequency radiation data, but their low spatial resolution makes it difficult to accurately reflect radiation variations in local areas, especially in photovoltaic power plant site selection and assessment, where higher accuracy is required. Conversely, satellite products with higher spatial resolution, due to their long update cycles, cannot meet rapidly changing radiation monitoring needs. Furthermore, surface albedo, a key parameter in surface shortwave radiation estimation, is too coarse. Traditional albedo product estimation methods are unable to account for complex surface albedo variations. This is particularly true in high-albedo regions (such as deserts and snow-covered areas) and complex terrain (such as plateaus and mountainous areas), where surface albedo varies dramatically. Existing technologies struggle to accurately capture its impact on radiation estimates, thus reducing the accuracy of surface shortwave radiation estimates. Furthermore, in addition to issues with spatial resolution and albedo, existing technologies still inadequately consider topographic factors. Topography directly affects the incident angle and intensity of solar radiation, especially in mountainous and plateau areas, but existing radiation estimation methods generally fail to fully incorporate these topographic parameters.

[0025] Currently, combining ground-based monitoring with satellite remote sensing data is considered an important means to improve the accuracy of radiation estimation, but it still faces many technical challenges in data fusion and correction. For example, the differences in the temporal, spatial, and physical characteristics of multi-source data lead to complex fusion algorithms. Especially under complex weather and terrain conditions, how to effectively improve the spatiotemporal resolution and accuracy of radiation estimation remains a key problem in current research. Overall, existing technologies still have much room for improvement in terms of insufficient spatial resolution, interference from meteorological and terrain factors, and lack of multi-source data fusion strategies. There is an urgent need to improve the accuracy and practicality of surface shortwave radiation estimation through technological innovation to provide more reliable data support for applications such as photovoltaic power generation, climate modeling, and agricultural meteorology.

[0026] In response to the deficiencies in the prior art, the present disclosure provides a method for producing surface shortwave radiation products. Figure 1 A process 100 of an embodiment of a method for producing a surface shortwave radiation product is shown. The method for producing a surface shortwave radiation product includes the following steps:

[0027] Step 101 : extracting a time-aligned first shortwave radiation dataset and a second shortwave radiation dataset from first surface shortwave radiation data collected by a satellite.

[0028] In this embodiment, the first shortwave radiation data set includes at least one first shortwave radiation data set, the second shortwave radiation data set includes at least one second shortwave radiation data set, the first temporal resolution of the first shortwave radiation data is greater than the second temporal resolution of the second shortwave radiation data, and the first spatial coverage of the first shortwave radiation data is less than the second spatial coverage of the second shortwave radiation data, and the first surface shortwave radiation data has a first spatial resolution. The values ​​of the first temporal resolution, the second temporal resolution, and the first spatial resolution can be set as required, e.g., the first temporal resolution is 10 minutes, the second temporal resolution is 15 minutes, and the first spatial resolution is 5 km; the first spatial coverage and the second spatial coverage can be determined based on the coverage of different satellites, e.g., the first spatial coverage is East Asia, and the second spatial coverage is global.

[0029] In this embodiment, the first surface shortwave radiation data may be minute-level surface shortwave radiation data. The surface shortwave radiation data refers to the shortwave portion of the solar radiation received by the ground. The surface shortwave radiation data may be data recorded by a satellite. The first surface shortwave radiation data may include at least two types of shortwave radiation data with different time resolutions, such as first shortwave radiation data and second shortwave radiation data. The first shortwave radiation data and the second shortwave radiation data have different spatial coverage.

[0030] In this embodiment, the above-mentioned step 101 includes: obtaining first surface shortwave radiation data collected by the satellite according to a required time period; extracting the L2-level shortwave radiation dataset of the first satellite and the L2-level shortwave radiation dataset of the second satellite from the first surface shortwave radiation data; performing data preprocessing on the L2-level shortwave radiation dataset of the first satellite and the L2-level shortwave radiation dataset of the second satellite, wherein the preprocessing includes: converting the nominal projection into the latitude and longitude projection, and converting the nc format into the tif format, to obtain the first preprocessed dataset and the second preprocessed dataset; and performing time matching on the first preprocessed dataset and the second preprocessed dataset to obtain the first shortwave radiation dataset and the second shortwave radiation dataset aligned in minute time.

[0031] In this embodiment, the first satellite has a first time resolution and a first spatial coverage, and the second satellite has a second time resolution and a second spatial coverage. By processing the L2 level shortwave radiation dataset of the first satellite, the surface shortwave radiation product of the first satellite is obtained, such as the surface shortwave radiation value distribution of the surface shortwave radiation product of the first satellite is shown as follows: Figure 2 As shown, in Figure 2 The larger the surface shortwave radiation value of the medium surface shortwave radiation product, the lighter the color in the image, and the smaller the surface shortwave radiation value, the darker the color in the image.

[0032] Step 102: Acquire a photovoltaic shortwave radiation dataset collected by a ground-based photovoltaic radiometer that is time-aligned with the first shortwave radiation dataset.

[0033] In this embodiment, the above-mentioned step 102 includes: selecting photovoltaic power stations laid out in different regions and different landform types, obtaining minute-level surface shortwave radiation numerical data continuously monitored by ground photovoltaic radiometers configured on each photovoltaic power station, and aligning the time scale with the first shortwave radiation dataset and the interpolated second shortwave radiation dataset to obtain a photovoltaic shortwave radiation dataset.

[0034] Step 103: Determine an albedo dataset of pixels with a second spatial resolution based on the satellite albedo product and the surface land cover product.

[0035] In this embodiment, the spatial resolution of the satellite albedo product is smaller than the land type spatial resolution of the surface land cover product. The second spatial resolution is smaller than the land type spatial resolution and greater than the first spatial resolution. The land type spatial resolution is the spatial resolution of the surface land cover product, which varies depending on the specific surface land cover product, ranging from 10m to 1000m. In this embodiment, the land type spatial resolution can be 30m. The second spatial resolution can be set as needed. For example, the second spatial resolution is 500m, which is the same as the spatial resolution of the satellite albedo product.

[0036] In this embodiment, the above-mentioned step 103 includes: obtaining an albedo product from a satellite with a medium-resolution imaging spectrometer, obtaining a surface land cover product, spatially superimposing the data of the two products, and recalculating the albedo of each pixel of the medium-resolution imaging spectrometer based on the proportion of land cover types covered in the data pixel of each medium-resolution imaging spectrometer to obtain an albedo dataset of pixels with a second spatial resolution.

[0037] Step 104 : obtaining an input shortwave radiation data sequence with a first temporal resolution and a second spatial resolution based on the first shortwave radiation data set, the second shortwave radiation data set, the albedo data set, the photovoltaic shortwave radiation data set, and the coordinate vector points of the ground photovoltaic radiometer.

[0038] In this embodiment, the shortwave radiation data sequence is various types of data to be predicted, such as: a first shortwave radiation data set after converting the spatial resolution, a second shortwave radiation data set after converting the temporal resolution and spatial resolution. The shortwave radiation data sequence is input into a pre-trained long short-term memory network to obtain an initial surface shortwave radiation product output by the long short-term memory network.

[0039] In this embodiment, step 104 includes: resampling the first shortwave radiation dataset and the interpolated second shortwave radiation dataset to a second spatial resolution (500 meters) to achieve spatial downscaling, reducing the spatial resolution of the two datasets from the first spatial resolution to the second spatial resolution, unifying them into the geographic coordinate system WGS84, and obtaining a resampled first shortwave radiation dataset; spatially superimposing the surface reflectance recalculated at the second spatial resolution (500 meters), and registering the interpolated second shortwave radiation dataset and the albedo dataset to the first shortwave radiation dataset using a pixel area weighting method, to obtain a registered second shortwave radiation dataset and a registered albedo dataset, thereby ensuring that the data fusion is continuous and authentic.

[0040] In this embodiment, pixel area weighting applies weighting to geographic data based on the actual ground area represented by each pixel to more accurately reflect the actual ground conditions. Due to the influence of geographic projection, pixels in different locations represent different actual ground areas, with pixels near the equator having larger areas and those near the North Pole having smaller areas. Without area weighting, directly averaging the data will result in inaccurate results.

[0041] In this embodiment, the coordinate vector points of the ground photovoltaic radiometer, the resampled first shortwave radiation dataset, the registered second shortwave radiation dataset, the registered albedo dataset, and the photovoltaic shortwave radiation dataset are spatially superimposed. The coordinate vector points of the ground photovoltaic radiometer are used to extract an input shortwave radiation data sequence of a preset frequency (e.g., 10 minutes) through a point value extraction method. The input shortwave radiation data sequence includes: a first shortwave radiation value set, a second shortwave radiation value set, a surface reflectivity value set, and a photovoltaic shortwave radiation value set. The first shortwave radiation value set is a dataset obtained by extracting the resampled first shortwave radiation dataset in the superimposed space at a preset frequency, the second shortwave radiation value set is a dataset obtained by extracting the registered second shortwave radiation dataset in the superimposed space at a preset frequency, the surface reflectivity value set is a dataset obtained by extracting the registered albedo dataset in the superimposed space at a preset frequency, and the photovoltaic shortwave radiation value set is a dataset obtained by extracting the photovoltaic shortwave radiation dataset in the superimposed space at a preset frequency.

[0042] Step 105: Input the input shortwave radiation data sequence into a pre-trained long short-term memory network to obtain an initial surface shortwave radiation product with a first temporal resolution, a second spatial resolution and a second spatial coverage.

[0043] In this embodiment, the long short-term memory network is used to predict the surface shortwave radiation product at a future time and within the second spatial coverage area.

[0044] In this embodiment, the input shortwave radiation data sequence is combined with the LSTM (Long ShortTerm Memory) of deep learning to predict the first shortwave radiation data belonging to the second spatial resolution, thereby completing the data of the uncovered areas in the first shortwave radiation data and forming an initial surface shortwave radiation product with the first temporal resolution and belonging to the second spatial coverage range.

[0045] Step 106: Obtain a target surface shortwave radiation product with a third spatial resolution based on the initial surface shortwave radiation product.

[0046] In this embodiment, the third spatial resolution is greater than the second spatial resolution. The third spatial resolution may be set based on development requirements. For example, the third spatial resolution is 30 m.

[0047] In this embodiment, step 106 includes downscaling and terrain correcting the initial surface shortwave radiation product using a digital elevation model (DEM) at a third spatial resolution, thereby spatially downscaling and terrain correcting the initial surface shortwave radiation product to obtain a target surface shortwave radiation product. Specifically, step 106 includes calculating the slope, aspect, angle of incidence, sky visible area factor, and sun-earth distance correction factor for each pixel of the initial surface shortwave radiation product using the DEM; and substituting the above calculation results into the equations for direct and diffuse radiation components of the radiation process according to the radiation transfer process to obtain the target surface shortwave radiation product after spatial downscaling and terrain correction.

[0048] The present disclosure provides a method for producing a surface shortwave radiation product. First, a first shortwave radiation dataset and a second shortwave radiation dataset are extracted from first surface shortwave radiation data collected by a satellite, wherein a first time resolution of the first shortwave radiation data is greater than a second time resolution of the second shortwave radiation data, and a first spatial coverage of the first shortwave radiation data is less than a second spatial coverage of the second shortwave radiation data, and the first surface shortwave radiation data has a first spatial resolution; secondly, a photovoltaic shortwave radiation dataset collected by a ground photovoltaic radiometer and time-aligned with the first shortwave radiation dataset is obtained; thirdly, based on a satellite albedo product and a surface land cover product, an albedo dataset of pixels having a second spatial resolution is determined, wherein the spatial resolution of the satellite albedo product is less than the land type spatial resolution of the surface land cover product; The second spatial resolution is smaller than the land type spatial resolution and greater than the first spatial resolution; then, based on the first shortwave radiation dataset, the second shortwave radiation dataset, the albedo dataset, the photovoltaic shortwave radiation dataset and the ground photovoltaic radiometer coordinate vector points, an input shortwave radiation data sequence with a first time resolution and a second spatial resolution is obtained; then, the input shortwave radiation data sequence is input into a pre-trained long short-term memory network to obtain an initial surface shortwave radiation product with a first time resolution, a second spatial resolution and belonging to the second spatial coverage range, and the long short-term memory network is used to predict the surface shortwave radiation product at future moments and belonging to the second spatial coverage range; finally, based on the initial surface shortwave radiation product, a target surface shortwave radiation product with a third spatial resolution is obtained, and the third spatial resolution is greater than the second spatial resolution. Therefore, by constructing a fusion framework of multi-element satellite and ground radiometer data and adopting long short-term memory network, high temporal resolution data and data of the second spatial coverage range are effectively integrated; land cover data of land type spatial resolution are introduced into the albedo correction process, and the reconstruction and fine correction of sub-kilometer scale albedo are realized, thereby significantly improving the spatial accuracy and resolution capability of surface shortwave radiation; based on the albedo data of the second spatial resolution, data complement is performed on the satellite-uncovered areas for generating the first shortwave radiation dataset to generate a global shortwave radiation intermediate product of the second spatial resolution; then, terrain radiation correction is performed in combination with the third spatial resolution data to generate a target surface shortwave radiation product of the third spatial resolution, which takes into account both the radiation estimation accuracy and the adaptability to areas with complex terrain; the present disclosure forms a set of scalable, high-precision, high temporal and spatial resolution global surface shortwave radiation data production methods, which provides a reliable data foundation and technical support for applications such as solar resource assessment, climate modeling, and agricultural meteorological analysis.

[0049] In some optional implementations of the present disclosure, the extracting of the time-aligned first and second shortwave radiation datasets from the first surface shortwave radiation data collected by the satellite includes: extracting the first surface shortwave radiation data from the minute-level surface shortwave radiation data collected by the satellite based on a historical time period; extracting the first satellite dataset and the second satellite dataset from the first surface shortwave radiation data; and time-matching a dataset with a first time resolution in the first satellite dataset with a dataset with a second time resolution in the second satellite dataset to obtain the first and second shortwave radiation datasets aligned in minutes.

[0050] In this optional implementation, the time periods of the first satellite dataset and the second satellite dataset are aligned according to the principle of data synthesis and high-frequency data acquisition. The time series data of the second satellite dataset that do not correspond to the time period of the time series data of the first satellite dataset are interpolated and supplemented using the mean interpolation method.

[0051] (1)

[0052] In formula (1), T is the time period of the first satellite dataset to be interpolated in the second satellite dataset, and t is the median value in this time period. Represents the mean value of the pixel with row and column numbers r, c in time period T, Represents the pixel value of the first satellite dataset with row and column numbers r and c 15 minutes before and after time t. t+15 represents the next 15 minutes, and t-15 represents the first 15 minutes.

[0053] The method for obtaining the first shortwave radiation data and the second shortwave radiation data provided by this optional implementation first extracts the first surface shortwave radiation data from the surface shortwave radiation data, then extracts the first satellite data and the second satellite data from the first surface shortwave radiation data, and obtains the first shortwave radiation dataset and the second shortwave radiation dataset by time-matching the first satellite data and the second satellite data. This provides a reliable implementation method for obtaining the first shortwave radiation dataset and the second shortwave radiation dataset.

[0054] In some optional implementations of the present disclosure, the above-mentioned determination of the albedo dataset of pixels with the second spatial resolution based on the satellite albedo product and the surface land cover product includes: obtaining a satellite albedo product with the second spatial resolution and a surface land cover product with a land type spatial resolution; spatially superimposing the satellite albedo product and the surface land cover product, and calculating the proportion of each land cover type in each pixel; constructing a linear regression model based on the proportion; and using the linear regression model to spatially assign albedo to different land cover types to recalculate the albedo of each pixel to obtain the albedo dataset of pixels with the second spatial resolution.

[0055] In this optional implementation, the above-mentioned acquisition of the satellite albedo product with the second spatial resolution and the surface land cover product with the land type spatial resolution includes: obtaining the satellite broadband directional albedo (Albedo_Broadband_BSA) product with the Moderate Resolution Imaging Spectroradiometer (hereinafter referred to as MODIS), with the time scale aligned with the first shortwave radiation data, and simultaneously obtaining the GlobeLand30 (short for the global land cover dataset) surface land cover raster data, processing the two data into tif format, and unifying the coordinate system into the WGS84 geographic coordinate system.

[0056] In this optional implementation, constructing a linear regression model based on the proportion includes: constructing the following linear regression model for the satellite albedo product of each pixel of the second spatial resolution based on the proportion:

[0057] (2)

[0058] In formula (2), is the raw MODIS albedo; is the area ratio of the i-th land type in the pixels of the second spatial resolution (based on statistics of each land cover type); is the albedo contribution factor corresponding to the i-th land type (solved by least squares regression).

[0059] In this optional implementation, the above-mentioned use of the linear regression model to spatially assign albedo to different land cover types to recalculate the albedo of each pixel to obtain the albedo dataset of the pixel with the second spatial resolution includes: using land cover data with a third spatial resolution to recalculate the albedo of each pixel to obtain the albedo dataset, as shown in formula (3):

[0060] (3)

[0061] This optional implementation provides a method for obtaining an albedo dataset for pixels with a second spatial resolution. For the first time, land cover data with a third spatial resolution are introduced into the albedo correction process. By calculating the proportion of different land types within each pixel, a linear regression model between land type distribution and albedo is constructed, achieving reconstruction and refined correction of sub-kilometer scale albedo, thereby significantly improving the spatial accuracy and resolution of surface shortwave radiation, and effectively solving the problem of insufficient spatial heterogeneity in traditional shortwave radiation estimation based on MODIS albedo products.

[0062] In some optional implementations of the present disclosure, the above-mentioned obtaining of the input shortwave radiation data sequence with the first temporal resolution and the second spatial resolution based on the first shortwave radiation dataset, the second shortwave radiation dataset, the albedo dataset and the coordinate vector points of the ground photovoltaic radiometer includes: resampling the first shortwave radiation dataset and the second shortwave radiation dataset so that the spatial resolutions of the first shortwave radiation dataset and the second shortwave radiation dataset are both the second spatial resolution; using the pixel area weighted method to align the resampled second shortwave radiation dataset and the albedo dataset to the resampled first shortwave radiation dataset to obtain the aligned second shortwave radiation dataset, the surface reflectance dataset and the resampled first shortwave radiation dataset; spatially superimposing the aligned second shortwave radiation dataset, the resampled first shortwave radiation dataset, the photovoltaic shortwave radiation dataset and the surface reflectance dataset to obtain a superimposed dataset; and extracting the input shortwave radiation data sequence with the first temporal resolution from the superimposed dataset using the coordinate vector points of the ground photovoltaic radiometer through a point value extraction method.

[0063] In this optional implementation, the registration method of the resampled second shortwave radiation dataset and the albedo dataset is the same. Taking the resampled second shortwave radiation dataset as an example, the pixel value of the newly registered second shortwave radiation dataset under the pixel of the resampled first shortwave radiation dataset is It can be calculated by the following formula (4):

[0064] (4)

[0065] In formula (4), i, j are the row and column numbers of the covered FY-4B shortwave radiation data pixels, n is the number of covered pixels, represents the pixel value of the pixel in the i-th row and j-th column in the wind and cloud data. is the weight value of the pixel, It can be determined by the following formula (5):

[0066] (5)

[0067] In formula (5), The pixel in the i-th row and j-th column of the resampled second shortwave radiation data occupies the area of ​​one pixel in the spatial resolution of the resampled first shortwave radiation data covering it. It represents the area of ​​one pixel in the spatial resolution of the first shortwave radiation data after resampling.

[0068] In this optional implementation, when extracting the input shortwave radiation data sequence by the point extraction value method, it can be performed according to the preset frequency, wherein the specific value of the preset frequency can be set as needed, for example, the preset frequency is 10 minutes.

[0069] The method for obtaining an input shortwave radiation data sequence provided by this optional implementation method resamples multiple data sets and aligns the sampled data sets using a pixel area weighted method. After superimposing the data sets, the input shortwave radiation data sequence with a first time resolution is extracted from the superimposed data sets using a ground photovoltaic radiometer coordinate vector point passing point value extraction method, thereby providing a reliable implementation means for obtaining the input shortwave radiation data sequence.

[0070] In some optional implementations of the present disclosure, obtaining the target surface shortwave radiation product with a third spatial resolution based on the initial surface shortwave radiation product includes: resampling the second spatial resolution of the initial surface shortwave radiation product to match the data elevation model to obtain a processed surface shortwave radiation product; calculating the direct radiation component based on the data elevation model, the processed surface shortwave radiation product, and the expression of the direct radiation component in the surface shortwave radiation process; calculating the scattered radiation component based on the data elevation model, the processed surface shortwave radiation product, and the expression of the scattered radiation in the surface shortwave radiation process; and summing the direct radiation component and the scattered radiation component to obtain a final surface shortwave radiation product having the third spatial resolution.

[0071] In this optional implementation, the above-mentioned resampling of the second spatial resolution of the initial surface shortwave radiation product to match the data elevation model to obtain the processed surface shortwave radiation product includes: using bilinear interpolation to convert the initial surface shortwave radiation product into a grid corresponding to the second spatial resolution, overlaying the initial surface shortwave radiation product with the data elevation model having the second spatial resolution, and making the grids completely overlap through georeferencing to obtain the processed surface shortwave radiation product.

[0072] In this optional implementation, the slope is calculated using DEM. and slope aspect Slope refers to the degree of inclination of the ground surface at a point or area, usually expressed as an angle or percentage. Aspect refers to the direction of the ground surface's maximum slope at a point or area, usually expressed as an angle ranging from 0° to 360°. Slope and aspect are two very important parameters in terrain analysis, as they help us understand the relief and directional characteristics of the terrain.

[0073] In this optional implementation, different geographical locations at the same time have different solar zenith angles and solar azimuth angles. The solar zenith angle in a large scale range is calculated by using the following formulas (6) and (7): and the solar azimuth .

[0074] (6)

[0075] In formula (6), : solar altitude angle, : The geographical latitude of the observation point, δ: The solar declination angle (related to the date, which can be approximated by the formula , N is the cumulative number of days per year), ω: hour angle (needs to be converted to the time difference between local time and 12 noon - 15° per hour).

[0076] (7)

[0077] In formula (7), : Solar azimuth (usually due north is 0°, rotating clockwise to 360°); : solar altitude angle; : The geographical latitude of the observation point; : Solar declination angle (related to the date, can be approximated by the formula , N is the cumulative number of days per year).

[0078] In this optional implementation, based on slope, aspect, and solar zenith angle and the solar azimuth Sun angle, calculate the angle of incidence:

[0079] (8)

[0080] In formula (8), is the angle of incidence, is the solar zenith angle and solar altitude angle Mutual redundancy, is the slope, is the solar azimuth, For the slope direction.

[0081] Use DEM and visual distance analysis method to calculate the sky view factor (SVF) of each pixel. Use DEM and solar azimuth and solar altitude angles to extract the elevation of the pixel to be calculated in the direction of the sun as the origin, calculate the elevation angle of the pixel and these points, and take the maximum value. Compared with the solar altitude angle, if If the sun's altitude is less than the point's altitude, there is no adjacent terrain blocking the point. =1), otherwise it is blocked by the adjacent terrain ( =0).

[0082] Calculate the correction factor for the distance between the Sun and the Earth based on the Julian day:

[0083] (9)

[0084] In formula (9), , N is the annual cumulative day; according to the radiation transmission process of direct radiation and scattered radiation, the direct radiation component is calculated separately ( ) and diffuse radiation components ( ):

[0085] (10)

[0086] (11)

[0087] In formula (10) and formula (11), is the angle of incidence, is the solar zenith angle, is the terrain shielding factor, SVF is the sky viewing area factor, is the solar shortwave radiation value before terrain correction, is the correction factor for the distance between the Sun and the Earth.

[0088] Finally, the target surface shortwave radiation product is obtained by formula (12): :

[0089] (12)

[0090] like Figure 3 The figure shows a schematic diagram of the distribution of surface shortwave radiation values ​​of the target surface shortwave radiation product. The third spatial resolution of the target surface shortwave radiation product is greater than Figure 2 The first spatial resolution of the surface shortwave radiation product of the first satellite is shown, and the second spatial coverage of the target surface shortwave radiation product is greater than Figure 2 The first spatial coverage of the surface shortwave radiation product of the first satellite is shown, and the target surface shortwave radiation product is Figure 2 The surface shortwave radiation products of the first satellite shown are all of the first time resolution. Figure 3 The larger the surface shortwave radiation value of the medium surface shortwave radiation product, the lighter the color in the image, and the smaller the surface shortwave radiation value, the darker the color in the image.

[0091] This optional implementation provides a method for obtaining the target surface shortwave radiation product. First, the initial surface shortwave radiation product is resampled, then the direct radiation component and the diffuse radiation component are calculated, and finally the final surface shortwave radiation product is obtained through the direct radiation component and the diffuse radiation component. This provides a reliable implementation method for obtaining the final surface shortwave radiation product. By performing terrain correction based on DEM data, it effectively makes up for the deficiency of traditional radiation products in ignoring the influence of terrain, making the final product more physically consistent and geographically adaptable in complex terrain areas such as mountains and plateaus.

[0092] The present disclosure provides a long short-term memory network training method. Figure 4A process 400 of an embodiment of a long short-term memory network training method is shown, and the long short-term memory network training method includes the following steps:

[0093] Step 401: Obtain a data training set.

[0094] In this embodiment, the data training set includes: a photovoltaic shortwave radiation value set, a surface reflectivity value set, a second shortwave radiation value set and a first shortwave radiation value set aligned with minute time. The first spatial coverage range of the first shortwave radiation value is smaller than the second spatial coverage range of the second shortwave radiation value. The first shortwave radiation value, the second shortwave radiation value, the surface reflectivity value set and the photovoltaic shortwave radiation value set all have a first time resolution and a second spatial resolution.

[0095] In this embodiment, the numerical set in the data training set can be a sequence obtained by collecting, resampling, and aligning the first shortwave radiation dataset, the second shortwave radiation dataset, the albedo dataset, and the photovoltaic shortwave radiation dataset according to a preset frequency. Specifically, the process of obtaining the data training set can refer to Figure 1 The process of obtaining the input shortwave radiation data sequence is not repeated here.

[0096] Specifically, the data training set includes: 、 、 as well as , i≥1. Among them, Indicates the The photovoltaic shortwave radiation value set (time series) of the photovoltaic weather station with samples; Indicates the The second shortwave radiation value set (time series) of samples; Indicates the A set of surface albedo values ​​(time series) of samples; Indicates the The first shortwave radiation value set (target value) of samples.

[0097] Step 402: construct multiple input sequence and output data pairs based on the data training set.

[0098] In this embodiment, the above data training sets are combined to form an input sequence, and a sliding window method is used to construct input-output pairs for initial long short-term memory network training. Specifically, the input sequence and output data pairs include: an input feature sequence and a prediction target. The construction results of the input sequence and output data pairs are as follows:

[0099] Input feature sequence: , , ;

[0100] Prediction target: ;

[0101] in To represent the length of the historical time window, the specific value of the historical time window length can be set based on development requirements, for example Indicates the previous hour, with a time interval of 10 minutes.

[0102] Step 403: Based on the input sequence and the output data pair, the initial long short-term memory network is trained to obtain a trained long short-term memory network.

[0103] In this embodiment, the long short-term memory network is used to preset a surface shortwave radiation product with a first temporal resolution, a second spatial resolution and a second spatial coverage range at a future moment.

[0104] In this embodiment, the input features are normalized, missing data are filled, and the input sequence and output data pairs are divided into a training set (e.g., 70% of the total data), a validation set (e.g., 15% of the total data), and a test set (e.g., 15% of the total data) by region.

[0105] In this example, when training the initial LSTM network, the model structure must be predefined. For example, the input dimension (input_dim) must be equal to the number of features (e.g., 3); the hidden layer dimension (hidden_dim) must be 32; the output dimension (output_dim) must be 1; and the number of LSTM layers (num_layers) must be 2. Different epochs (training rounds) and batch sizes can be set for each sample or region to accommodate differences in sample size. Mean squared error (MSE) is used as the loss function, and the Adam optimizer is used.

[0106] In this example, the weights and biases of the LSTM network are randomly initialized. Input data is fed into the LSTM time-step by time-step through forward propagation. At each time step, the LSTM computes the hidden state and cell state for the current time step. Ultimately, the LSTM calculates the output based on the hidden state.

[0107] In this example, during each LSTM training iteration, a defined loss function is used to calculate the error between the model's predicted value and the true value, i.e., the LSTM loss. The network's weights and biases are updated by calculating the gradient of the loss function with respect to each parameter. The gradient is then propagated back to each layer of the network using the chain rule. Based on the calculated gradient, the network parameters are updated using an optimizer. Model performance is evaluated on a validation set, and hyperparameters (such as the learning rate, number of hidden units, and number of layers) are adjusted to optimize the model. The final model performance is evaluated on a test set to ensure that the LSTM network has good generalization capabilities.

[0108] The long short-term memory network training method provided by the present disclosure first obtains a data training set; then, based on the data training set, constructs multiple input sequences and output data pairs; finally, trains an initial long short-term memory network based on the input sequences and output data pairs to obtain a trained long short-term memory network. Thus, the spatial coverage range of multiple shortwave radiation values ​​is collected to train the long short-term memory network, so that the trained long short-term memory network has stronger prediction performance, thereby improving the reliability and accuracy of the long short-term memory network training.

[0109] In some optional implementations of the present disclosure, obtaining the data training set includes: extracting a first satellite dataset and a second satellite dataset from minute-level surface shortwave radiation data collected by a satellite, the surface shortwave radiation data having a first spatial resolution; temporally matching a dataset with a first temporal resolution in the first satellite dataset with a dataset with a second temporal resolution in the second satellite dataset to obtain a first shortwave radiation dataset and a second shortwave radiation dataset aligned at minute moments; obtaining a photovoltaic shortwave radiation dataset collected by a ground-based photovoltaic radiometer aligned at minute moments with the first shortwave radiation dataset; determining an albedo dataset of pixels at a second spatial resolution based on a satellite albedo product and a surface land cover product, the satellite spatial resolution of the satellite albedo product being smaller than the land type spatial resolution of the surface land cover product, and the second spatial resolution being smaller than the land type spatial resolution; and obtaining a photovoltaic shortwave radiation value set, a surface reflectance value set, a second shortwave radiation value set, and a first shortwave radiation value set aligned at minute moments based on the first shortwave radiation dataset, the second shortwave radiation dataset, the albedo dataset, the photovoltaic shortwave radiation dataset, and the coordinate vector points of the ground-based photovoltaic radiometer.

[0110] It should be noted that the photovoltaic shortwave radiation value set, the surface reflectivity value set, the second shortwave radiation value set and the first shortwave radiation value set constitute the input shortwave radiation data sequence, and the acquisition of the input shortwave radiation data sequence is described in detail in the above embodiment and will not be repeated here.

[0111] Further references Figure 5 As an implementation of the methods shown in the above figures, the present disclosure provides an embodiment of a surface shortwave radiation product production device. Figure 1 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0112] Figure 5As shown, the surface shortwave radiation product production device 500 provided in this embodiment includes: an extraction unit 501, a photovoltaic acquisition unit 502, a determination unit 503, a sequence acquisition unit 504, an input unit 505, and a product acquisition unit 506. The extraction unit 501 can be configured to extract a time-aligned first shortwave radiation dataset and a second shortwave radiation dataset from first surface shortwave radiation data collected by a satellite. The first shortwave radiation dataset includes at least one first shortwave radiation data item, the second shortwave radiation dataset includes at least one second shortwave radiation data item, the first temporal resolution of the first shortwave radiation data item is greater than the second temporal resolution of the second shortwave radiation data item, the first spatial coverage of the first shortwave radiation data item is less than the second spatial coverage of the second shortwave radiation data item, and the first surface shortwave radiation data item has the first spatial resolution. The photovoltaic acquisition unit 502 can be configured to acquire a photovoltaic shortwave radiation dataset collected by a ground-based photovoltaic radiometer that is time-aligned with the first shortwave radiation dataset. The determination unit 503 may be configured to determine an albedo dataset for pixels having a second spatial resolution based on a satellite albedo product and a surface land cover product, wherein the satellite spatial resolution of the satellite albedo product is less than the land type spatial resolution of the surface land cover product, and the second spatial resolution is less than the land type spatial resolution but greater than the first spatial resolution. The sequence acquisition unit 504 may be configured to acquire an input shortwave radiation data sequence having a first temporal resolution and a second spatial resolution based on the first shortwave radiation data set, the second shortwave radiation data set, the albedo data set, the photovoltaic shortwave radiation data set, and the coordinate vector points of the ground fiber optic radiometer. The input unit 505 may be configured to input the input shortwave radiation data sequence into a pre-trained long short-term memory network to acquire an initial surface shortwave radiation product having the first temporal resolution, the second spatial resolution, and belonging to the second spatial coverage range. The long short-term memory network is used to predict surface shortwave radiation products at future moments that belong to the second spatial coverage range. The product obtaining unit 506 may be configured to obtain a target surface shortwave radiation product of a third spatial resolution based on the initial surface shortwave radiation product, where the third spatial resolution is greater than the second spatial resolution.

[0113] In this embodiment, the specific processing and technical effects of the surface shortwave radiation product production device 500: the extraction unit 501, the photovoltaic acquisition unit 502, the determination unit 503, the sequence acquisition unit 504, the input unit 505, and the product acquisition unit 506 can be referred to respectively. Figure 1 The relevant descriptions of step 101, step 102, step 103, step 104, step 105, and step 106 in the corresponding embodiment are not repeated here.

[0114] In some embodiments of the present disclosure, the extraction unit 501 is configured to: extract first surface shortwave radiation data from minute-level surface shortwave radiation data collected by satellites based on a historical time period; extract a first satellite dataset and a second satellite dataset from the first surface shortwave radiation data; and perform time matching on a dataset with a first time resolution in the first satellite dataset and a dataset with a second time resolution in the second satellite dataset to obtain the first shortwave radiation dataset and the second shortwave radiation dataset aligned at minute moments.

[0115] In some embodiments of the present disclosure, the above-mentioned determination unit 503 is configured to: obtain a satellite albedo product with a second spatial resolution and a surface land cover product with a land type spatial resolution; spatially superimpose the satellite albedo product and the surface land cover product, and calculate the proportion of each land cover type in each pixel; based on the proportion, construct a linear regression model; use the linear regression model to spatially assign albedo to different land cover types to recalculate the albedo of each pixel, and obtain an albedo dataset of pixels with the second spatial resolution.

[0116] In some embodiments of the present disclosure, the sequence obtaining unit 504 may be configured to: resample the first shortwave radiation dataset and the second shortwave radiation dataset so that the spatial resolutions of the first shortwave radiation dataset and the second shortwave radiation dataset are both the second spatial resolution; register the resampled second shortwave radiation dataset and the albedo dataset to the resampled first shortwave radiation dataset using a pixel area weighted method to obtain a registered second shortwave radiation dataset, a surface reflectance dataset, and a resampled first shortwave radiation dataset; spatially superimpose the registered second shortwave radiation dataset, the resampled first shortwave radiation dataset, the photovoltaic shortwave radiation dataset, and the surface reflectance dataset to obtain a superimposed dataset; and extract an input shortwave radiation data sequence with the first temporal resolution from the superimposed dataset using a ground photovoltaic radiometer coordinate vector point passing point value extraction method.

[0117] In some embodiments of the present disclosure, the product obtaining unit 506 is configured to: resample the second spatial resolution of the initial surface shortwave radiation product to match the data elevation model to obtain a processed surface shortwave radiation product; calculate the direct radiation component based on the data elevation model, the processed surface shortwave radiation product, and the expression of the direct radiation component in the surface shortwave radiation process; calculate the scattered radiation component based on the data elevation model, the processed surface shortwave radiation product, and the expression of the scattered radiation in the surface shortwave radiation process; sum the direct radiation component and the scattered radiation component to obtain the final surface shortwave radiation product P3 belonging to the third spatial resolution.

[0118] The surface shortwave radiation product production device provided by the embodiments of the present disclosure constructs a fusion framework of multi-satellite and ground radiometer data and adopts a long short-term memory network to effectively integrate high-temporal resolution data and data of the second spatial coverage range, and breaks through the limitation of traditional single-satellite payloads that cannot take into account both temporal and spatial resolutions. For the first time, it achieves global coverage and standardized product output with temporal and spatial resolutions of the first temporal resolution / second spatial resolution; introduces land cover data with land type spatial resolution into the albedo correction process, realizes the reconstruction and fine correction of sub-kilometer scale albedo, thereby significantly improving the spatial accuracy and resolution capability of surface shortwave radiation; based on the albedo data of the second spatial resolution, the satellite coverage area for generating the first shortwave radiation data set is supplemented with data to generate a global shortwave radiation intermediate product of the second spatial resolution; then, terrain radiation correction is performed in combination with the third spatial resolution data to generate a target surface shortwave radiation product of the third spatial resolution, which takes into account both the radiation estimation accuracy and the adaptability to areas with complex terrain.

[0119] Further references Figure 6 As an implementation of the methods shown in the above figures, the present disclosure provides an embodiment of a long short-term memory network training device. Figure 2 In the electronic device corresponding to the method embodiment shown.

[0120] Figure 6 As shown, the long short-term memory network training device 600 provided in this embodiment includes: a data acquisition unit 601, a construction unit 602, and a training unit 603. The data acquisition unit 601 can be configured to acquire a data training set, the data training set including: a minute-aligned photovoltaic shortwave radiation value set, a surface reflectivity value set, a second shortwave radiation value set, and a first shortwave radiation value set, wherein the first shortwave radiation value set has a first spatial coverage range smaller than the second spatial coverage range of the second shortwave radiation value set, and the first shortwave radiation value set, the second shortwave radiation value set, the surface reflectivity value set, and the photovoltaic shortwave radiation value set all have a first temporal resolution and a second spatial resolution. The construction unit 602 can be configured to construct an input sequence and an output data pair based on the data training set. The training unit 603 can be configured to train an initial long short-term memory network based on the input sequence and the output data pair, thereby obtaining a trained long short-term memory network, wherein the long short-term memory network is used to preset a surface shortwave radiation product with the first temporal resolution, the second spatial resolution, and the second spatial coverage range at a future time.

[0121] In this embodiment, the specific processing of the data acquisition unit 601, the construction unit 602, and the training unit 603 and the technical effects thereof can be referred to in the respective embodiments. Figure 2The relevant descriptions of step 201, step 202, and step 203 in the corresponding embodiment are not repeated here.

[0122] In some embodiments of the present disclosure, the data acquisition unit 601 is configured to: extract a first satellite dataset and a second satellite dataset from minute-level surface shortwave radiation data collected by a satellite, where the surface shortwave radiation data has a first spatial resolution; perform time matching on a dataset with a first temporal resolution in the first satellite dataset and a dataset with a second temporal resolution in the second satellite dataset to obtain a first shortwave radiation dataset and a second shortwave radiation dataset aligned at minute times; obtain a photovoltaic shortwave radiation dataset collected by a ground-based photovoltaic radiometer that is aligned at minute times with the first shortwave radiation dataset; determine an albedo dataset of pixels at a second spatial resolution based on a satellite albedo product and a surface land cover product, where the satellite spatial resolution of the satellite albedo product is smaller than the land type spatial resolution of the surface land cover product, and the second spatial resolution is smaller than the land type spatial resolution; and obtain a photovoltaic shortwave radiation value set, a surface reflectance value set, a second shortwave radiation value set, and a first shortwave radiation value set aligned at minute times based on the first shortwave radiation dataset, the second shortwave radiation dataset, the albedo dataset, the photovoltaic shortwave radiation dataset, and the coordinate vector points of the ground-based photovoltaic radiometer.

[0123] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0124] Figure 7 A schematic block diagram of an example electronic device 700 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their modes are provided for example only and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0125] like Figure 7As shown, electronic device 700 includes a computing unit 701, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 702 or a computer program loaded from a storage unit 708 into a random access memory (RAM) 703. Various programs and data required for the operation of electronic device 700 may also be stored in RAM 703. Computing unit 701, ROM 702, and RAM 703 are interconnected via a bus 704. An input / output (I / O) interface 705 is also connected to bus 704.

[0126] Multiple components in the electronic device 700 are connected to the I / O interface 705, including an input unit 706, such as a keyboard, a mouse, etc.; an output unit 707, such as various types of displays, speakers, etc.; a storage unit 708, such as a magnetic disk, an optical disk, etc.; and a communication unit 709, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 709 allows the electronic device 700 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0127] The computing unit 701 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the computing unit 701 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 701 performs the various methods and processes described above, such as the surface shortwave radiation product production method or the long short-term memory network training method. For example, in some embodiments, the surface shortwave radiation product production method or the long short-term memory network training method can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 708. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 700 via the ROM 702 and / or the communication unit 709. When the computer program is loaded into the RAM 703 and executed by the computing unit 701, one or more steps of the surface shortwave radiation product production method or the long short-term memory network training method described above can be performed. Alternatively, in other embodiments, the computing unit 701 may be configured to execute the surface shortwave radiation product production method or the long short-term memory network training method in any other appropriate manner (for example, by means of firmware).

[0128] Various embodiments of the systems and techniques described above can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0129] Program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. Such program code can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable surface shortwave radiation product production device or a long-short-term memory network training device, such that when the program code is executed by the processor or controller, the modes / operations specified in the flowcharts and / or block diagrams are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0130] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of machine-readable storage media may include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), optical fibers, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0131] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0132] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0133] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not limited herein.

[0134] The foregoing descriptions of specific exemplary embodiments of the present invention are for purposes of illustration and description. These descriptions are not intended to limit the invention to the precise forms disclosed, and it is apparent that many variations and modifications are possible in light of the foregoing teachings. The exemplary embodiments have been selected and described for the purpose of explaining the specific principles of the invention and their practical application, thereby enabling those skilled in the art to realize and utilize a variety of exemplary embodiments of the invention and various options and modifications. The scope of the invention is intended to be defined by the claims and their equivalents.

Claims

1. A method for producing surface shortwave radiation products, characterized in that: The method comprises: Extracting a time-aligned first shortwave radiation dataset and a second shortwave radiation dataset from first surface shortwave radiation data collected by a satellite, wherein the first shortwave radiation dataset includes at least one first shortwave radiation data, the second shortwave radiation dataset includes at least one second shortwave radiation data, a first temporal resolution of the first shortwave radiation data is greater than a second temporal resolution of the second shortwave radiation data, a first spatial coverage of the first shortwave radiation data is smaller than a second spatial coverage of the second shortwave radiation data, and the first surface shortwave radiation data has a first spatial resolution; Acquire a photovoltaic shortwave radiation dataset collected by a ground-based photovoltaic radiometer and time-aligned with the first shortwave radiation dataset; determining, based on a satellite albedo product and a surface land cover product, an albedo dataset of pixels having a second spatial resolution, wherein the spatial resolution of the satellite albedo product is smaller than a land type spatial resolution of the surface land cover product, and the second spatial resolution is smaller than the land type spatial resolution and greater than the first spatial resolution; Obtaining an input shortwave radiation data sequence having the first temporal resolution and the second spatial resolution based on the first shortwave radiation dataset, the second shortwave radiation dataset, the albedo dataset, the photovoltaic shortwave radiation dataset, and the ground photovoltaic radiometer coordinate vector points includes: resampling the first shortwave radiation dataset and the second shortwave radiation dataset so that the spatial resolutions of the first shortwave radiation dataset and the second shortwave radiation dataset are both the second spatial resolution; registering the resampled second shortwave radiation dataset and the albedo dataset to the resampled first shortwave radiation dataset using a pixel area weighting method to obtain a registered second shortwave radiation dataset, a surface reflectance dataset, and a resampled first shortwave radiation dataset; spatially superimposing the registered second shortwave radiation dataset, the resampled first shortwave radiation dataset, the photovoltaic shortwave radiation dataset, and the surface reflectance dataset to obtain a superimposed dataset; and extracting the input shortwave radiation data sequence having the first temporal resolution from the superimposed dataset using the ground photovoltaic radiometer coordinate vector points by a point value extraction method; Inputting the input shortwave radiation data sequence into a pre-trained long short-term memory network to obtain an initial surface shortwave radiation product having the first temporal resolution, the second spatial resolution and belonging to the second spatial coverage range, and using the long short-term memory network to predict surface shortwave radiation products at future moments and belonging to the second spatial coverage range; Based on the initial surface shortwave radiation product, a target surface shortwave radiation product with a third spatial resolution is obtained, where the third spatial resolution is greater than the second spatial resolution.

2. The method according to claim 1, characterized in that The step of extracting a time-aligned first shortwave radiation dataset and a second shortwave radiation dataset from first surface shortwave radiation data collected by a satellite comprises: Based on the historical time period, the first surface shortwave radiation data is extracted from the minute-level surface shortwave radiation data collected by the satellite; extracting a first satellite dataset and a second satellite dataset from the first surface shortwave radiation data; A dataset with a first time resolution in the first satellite dataset is time-matched with a dataset with a second time resolution in the second satellite dataset to obtain a first shortwave radiation dataset and a second shortwave radiation dataset aligned in minute time.

3. The method according to claim 1, characterized in that Determining the albedo dataset of pixels with a second spatial resolution based on the satellite albedo product and the surface land cover product includes: Obtain satellite albedo products with second spatial resolution and surface land cover products with land type spatial resolution; Spatially superimposing the satellite albedo product and the surface land cover product, and calculating the proportion of each land cover type in each pixel; Based on the proportions, a linear regression model is constructed; The linear regression model is used to spatially assign albedo to different land cover types to recalculate the albedo of each pixel, thereby obtaining an albedo dataset of pixels having the second spatial resolution.

4. The method according to claim 1, wherein Obtaining a target surface shortwave radiation product having a third spatial resolution based on the initial surface shortwave radiation product includes: resampling the second spatial resolution of the initial surface shortwave radiation product to match the data elevation model to obtain a processed surface shortwave radiation product; Calculate the direct radiation component based on the data elevation model, the processed surface shortwave radiation product, and the expression of the direct radiation component in the surface shortwave radiation process; Calculating the scattered radiation component based on the data elevation model, the processed surface shortwave radiation product, and the scattered radiation expression in the surface shortwave radiation process; The direct radiation component and the scattered radiation component are summed to obtain a final surface shortwave radiation product of a third spatial resolution.

5. A long short-term memory network training method, characterized in that: The method comprises: A data training set is obtained, wherein the data training set includes: a photovoltaic shortwave radiation value set, a surface reflectivity value set, a second shortwave radiation value set, and a first shortwave radiation value set aligned at minute time points, wherein a first spatial coverage range of the first shortwave radiation value is smaller than a second spatial coverage range of the second shortwave radiation value, and the first shortwave radiation value, the second shortwave radiation value, the surface reflectivity value set, and the photovoltaic shortwave radiation value set all have a first temporal resolution and a second spatial resolution; obtaining the data training set includes: extracting a first satellite data set and a second satellite data set from minute-level surface shortwave radiation data collected by a satellite, wherein the surface shortwave radiation data has a first spatial resolution; and temporally averaging a data set of the first temporal resolution in the first satellite data set with a data set of the second temporal resolution in the second satellite data set. Matching to obtain a first shortwave radiation dataset and a second shortwave radiation dataset aligned at minute times; obtaining a photovoltaic shortwave radiation dataset collected by a ground-based photovoltaic radiometer that is aligned at minute times with the first shortwave radiation dataset; determining an albedo dataset of pixels at a second spatial resolution based on a satellite albedo product and a surface land cover product, wherein the satellite spatial resolution of the satellite albedo product is smaller than the land type spatial resolution of the surface land cover product, and the second spatial resolution is smaller than the land type spatial resolution; obtaining a photovoltaic shortwave radiation value set, a surface reflectance value set, a second shortwave radiation value set, and a first shortwave radiation value set aligned at minute times based on the first shortwave radiation dataset, the second shortwave radiation dataset, the albedo dataset, the photovoltaic shortwave radiation dataset, and the coordinate vector points of the ground-based photovoltaic radiometer; Based on the data training set, construct a plurality of input sequence and output data pairs; Based on the input sequence and output data pairs, an initial long short-term memory network is trained to obtain a trained long short-term memory network, wherein the long short-term memory network is used to preset a surface shortwave radiation product having the first temporal resolution, the second spatial resolution and belonging to the second spatial coverage range at a future moment.

6. A surface shortwave radiation product production device, characterized in that: The device comprises: an extraction unit configured to extract a time-aligned first shortwave radiation dataset and a second shortwave radiation dataset from first surface shortwave radiation data collected by a satellite, wherein the first shortwave radiation dataset includes at least one first shortwave radiation data, the second shortwave radiation dataset includes at least one second shortwave radiation data, a first temporal resolution of the first shortwave radiation data is greater than a second temporal resolution of the second shortwave radiation data, a first spatial coverage of the first shortwave radiation data is smaller than a second spatial coverage of the second shortwave radiation data, and the first surface shortwave radiation data has a first spatial resolution; a photovoltaic acquisition unit configured to acquire a photovoltaic shortwave radiation dataset collected by a ground photovoltaic radiometer and time-aligned with the first shortwave radiation dataset; a determining unit configured to determine an albedo dataset of pixels having a second spatial resolution based on a satellite albedo product and a surface land cover product, wherein the satellite spatial resolution of the satellite albedo product is smaller than the land type spatial resolution of the surface land cover product, and the second spatial resolution is smaller than the land type spatial resolution and larger than the first spatial resolution; A sequence obtaining unit is configured to obtain an input shortwave radiation data sequence having the first temporal resolution and the second spatial resolution based on the first shortwave radiation dataset, the second shortwave radiation dataset, the albedo dataset, the photovoltaic shortwave radiation dataset, and the ground optical fiber radiometer coordinate vector points; the sequence obtaining unit is further configured to: resample the first shortwave radiation dataset and the second shortwave radiation dataset so that the spatial resolutions of the first shortwave radiation dataset and the second shortwave radiation dataset are both the second spatial resolution; register the resampled second shortwave radiation dataset and the albedo dataset to the resampled first shortwave radiation dataset using a pixel area weighting method to obtain a registered second shortwave radiation dataset, a surface reflectance dataset, and a resampled first shortwave radiation dataset; spatially superimpose the registered second shortwave radiation dataset, the resampled first shortwave radiation dataset, the photovoltaic shortwave radiation dataset, and the surface reflectance dataset to obtain a superimposed dataset; and extract the input shortwave radiation data sequence having the first temporal resolution from the superimposed dataset using the ground photovoltaic radiometer coordinate vector points through a point value extraction method; an input unit configured to input the input shortwave radiation data sequence into a pre-trained long short-term memory network to obtain an initial surface shortwave radiation product having the first temporal resolution, the second spatial resolution and belonging to the second spatial coverage range, wherein the long short-term memory network is used to predict the surface shortwave radiation product at a future time point and belonging to the second spatial coverage range; The product obtaining unit is configured to obtain a target surface shortwave radiation product having a third spatial resolution based on the initial surface shortwave radiation product, where the third spatial resolution is greater than the second spatial resolution.

7. A long short-term memory network training device, characterized in that: The device comprises: The data acquisition unit is configured to acquire a data training set, wherein the data training set includes: a photovoltaic shortwave radiation value set, a surface reflectivity value set, a second shortwave radiation value set, and a first shortwave radiation value set aligned at minute time points, wherein the first spatial coverage range of the first shortwave radiation value is smaller than the second spatial coverage range of the second shortwave radiation value, and the first shortwave radiation value, the second shortwave radiation value, the surface reflectivity value set, and the photovoltaic shortwave radiation value set all have a first temporal resolution and a second spatial resolution; the data acquisition unit is further configured to: extract a first satellite data set and a second satellite data set from minute-level surface shortwave radiation data collected by a satellite, wherein the surface shortwave radiation data has a first spatial resolution; and compare the data set with the first temporal resolution in the first satellite data set and the data set with the second temporal resolution in the second satellite data set. The data sets are time-matched to obtain a first shortwave radiation data set and a second shortwave radiation data set aligned at minute times; a photovoltaic shortwave radiation data set collected by a ground-based photovoltaic radiometer and aligned at minute times with the first shortwave radiation data set is obtained; an albedo data set of pixels at a second spatial resolution is determined based on a satellite albedo product and a surface land cover product, wherein the satellite spatial resolution of the satellite albedo product is smaller than the land type spatial resolution of the surface land cover product, and the second spatial resolution is smaller than the land type spatial resolution; a photovoltaic shortwave radiation value set, a surface reflectance value set, a second shortwave radiation value set, and a first shortwave radiation value set aligned at minute times are obtained based on the first shortwave radiation data set, the second shortwave radiation data set, the albedo data set, the photovoltaic shortwave radiation data set, and the coordinate vector points of the ground-based photovoltaic radiometer; A construction unit, configured to construct an input sequence and an output data pair based on the data training set; The training unit is configured to train an initial long short-term memory network based on the input sequence and the output data pair to obtain a trained long short-term memory network, wherein the long short-term memory network is used to preset a surface shortwave radiation product having the first temporal resolution, the second spatial resolution and belonging to the second spatial coverage range at a future moment.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images