CN117633441A - Sea surface temperature data filling and correcting method and device and electronic equipment - Google Patents
Sea surface temperature data filling and correcting method and device and electronic equipment Download PDFInfo
- Publication number
- CN117633441A CN117633441A CN202410096143.0A CN202410096143A CN117633441A CN 117633441 A CN117633441 A CN 117633441A CN 202410096143 A CN202410096143 A CN 202410096143A CN 117633441 A CN117633441 A CN 117633441A
- Authority
- CN
- China
- Prior art keywords
- surface temperature
- sea surface
- temperature data
- data
- error variance
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 78
- 238000012545 processing Methods 0.000 claims abstract description 15
- 239000011159 matrix material Substances 0.000 claims description 49
- 230000002159 abnormal effect Effects 0.000 claims description 35
- 238000012549 training Methods 0.000 claims description 33
- 230000008878 coupling Effects 0.000 claims description 28
- 238000010168 coupling process Methods 0.000 claims description 28
- 238000005859 coupling reaction Methods 0.000 claims description 28
- 230000004927 fusion Effects 0.000 claims description 22
- 230000015654 memory Effects 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 16
- 238000004422 calculation algorithm Methods 0.000 claims description 15
- 230000008569 process Effects 0.000 claims description 15
- 238000011176 pooling Methods 0.000 claims description 10
- 238000012937 correction Methods 0.000 claims description 9
- 238000003860 storage Methods 0.000 claims description 8
- 230000004913 activation Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 6
- 238000011065 in-situ storage Methods 0.000 claims 1
- 230000000875 corresponding effect Effects 0.000 description 21
- 238000010586 diagram Methods 0.000 description 8
- 230000002596 correlated effect Effects 0.000 description 5
- 238000013135 deep learning Methods 0.000 description 4
- 230000007246 mechanism Effects 0.000 description 4
- 238000005070 sampling Methods 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Radiation Pyrometers (AREA)
Abstract
The invention provides a filling and correcting method, a device and electronic equipment of sea surface temperature data, and belongs to the technical field of data processing.
Description
Technical Field
The present invention relates to the field of data processing technologies, and in particular, to a method and an apparatus for filling and correcting sea surface temperature data, and an electronic device.
Background
At present, sea surface temperature data (sea surface temperature data acquired by a mode of placing equipment at each point of a sea surface site) obtained based on site observation is few, satellite remote sensing data are affected by cloud, and the like, the satellite remote sensing sea surface temperature data are sparse and uneven in distribution, and the sea surface temperature product produced by a traditional model only depends on the satellite remote sensing sea surface temperature data, so that the sea surface temperature data observed in site are not incorporated into a data product production process, and the potential application of the sea surface temperature data is greatly limited.
The traditional method is used for solving the problem of low space-time coverage rate of satellite remote sensing sea surface temperature data by constructing a fitting model to predict a missing value, but the transient and random properties of the sea surface temperature data are easy to smooth and filter in the model operation process, and the traditional model algorithm needs complex prior information, so that the time dimension characteristics and the space dimension characteristics of the satellite remote sensing sea surface temperature data cannot be simultaneously extracted and correlated, the method is limited when correcting and filling the satellite remote sensing sea surface temperature data, the space-time coverage rate of model results is low, the precision and the resolution are low, and the real physical process that the sea surface temperature data cannot be reflected by the sea surface temperature data produced by depending on the traditional model is dependent.
In summary, the traditional sea surface temperature data correction and filling method has the technical problems of low space-time coverage rate and low precision of the corrected and filled sea surface temperature data.
Disclosure of Invention
In view of the above, the invention aims to provide a filling and correcting method, a device and electronic equipment for sea surface temperature data, so as to solve the technical problems of low space-time coverage rate and low precision of the sea surface temperature data obtained by the traditional sea surface temperature data correcting and filling method and the filled sea surface temperature data.
In a first aspect, an embodiment of the present invention provides a method for filling and correcting sea surface temperature data, including:
performing gridding pretreatment on the sea surface temperature data observed on site to obtain gridding sea surface temperature data observed on site;
carrying out weighted fusion on satellite remote sensing sea surface temperature data and the on-site observed grid sea surface temperature data to obtain fused sea surface temperature data;
performing mathematical operation on the fused sea surface temperature data to obtain a mathematical operation data matrix, wherein the data in the mathematical operation data matrix comprises: inverse of error variance, scaled sea surface temperature anomaly, normalized longitude and latitude data, and normalized time data;
Inputting the mathematical operation data matrix into a convolution self-encoder of a double-layer coupling structure for iterative training, and outputting a plurality of logarithms of expected error variance reciprocal values and sea surface temperature anomalies scaled by the plurality of reciprocal values of expected error variance values in the iterative training process;
calculating a target sea surface temperature anomaly value from a plurality of logarithms of the expected error variance reciprocals and a plurality of sea surface temperature anomalies scaled by the expected error variance reciprocals;
and correcting an arithmetic average value corresponding to the fused sea surface temperature data by adopting the target sea surface temperature abnormal value to obtain corrected and filled sea surface temperature data.
Further, performing gridding pretreatment on the sea surface temperature data observed in the field, wherein the gridding pretreatment comprises the following steps:
constructing an empty grid with the same size as the grid of the satellite remote sensing sea surface temperature data;
adding the field observed sea surface temperature data into the empty grid, and calculating the field observed sea surface temperature data of the adjacent grid of the grid where the field observed sea surface temperature data is positioned by adopting an interpolation algorithm;
and placing the calculated on-site observed sea surface temperature data adjacent to the grid into the corresponding adjacent grid to obtain the on-site observed meshed sea surface temperature data.
Further, the weighted fusion of satellite remote sensing sea surface temperature data and the on-site observed grid sea surface temperature data comprises the following steps:
acquiring a preset weight;
and carrying out weighted fusion on the satellite remote sensing sea surface temperature data and the on-site observed grid sea surface temperature data according to the preset weight to obtain the fused sea surface temperature data.
Further, performing a mathematical operation on the fused sea surface temperature data, including:
calculating an error variance of the fused sea surface temperature data, and determining the inverse of the error variance according to the error variance;
calculating an arithmetic mean of the fused sea surface temperature data;
calculating the abnormal value of the sea surface temperature according to the fused sea surface temperature data and the arithmetic average value;
scaling the sea surface temperature abnormal value corresponding to the inverse of the error variance to obtain the scaled sea surface temperature abnormal value;
carrying out standardization processing on the longitude and latitude data to obtain the standardized longitude and latitude data;
and carrying out standardization processing on the time data to obtain the standardized time data, and further obtaining a mathematical operation data matrix composed of the inverse of the error variance, the scaled sea surface temperature abnormal value, the standardized longitude and latitude data and the standardized time data.
Further, the mathematical operation data matrix is input into a convolution self-encoder with a double-layer coupling structure to perform iterative training, and the method comprises the following steps:
inputting the mathematical operation data matrix into a first convolution self-encoder to obtain a first output result;
and inputting the first output result and the mathematical operation data matrix into a second convolution self-encoder, and outputting sea surface temperature anomalies which are obtained by scaling the logarithm of the inverse of the expected error variance and the inverse of the expected error variance.
Further, each of the convolutional self-encoders of the double-layer coupling structure includes: an encoder and a decoder;
the encoder includes: a plurality of coding layers, each of the coding layers comprising: a convolution layer, an attention module, an activation function and a pooling layer;
the decoder includes: the same number of decoding layers as the encoding layers, each decoding layer comprising: an upsampling layer, an attention module, and a convolution layer;
a summing skip connection is employed between the encoding layer and the decoding layer.
Further, calculating a target sea surface temperature anomaly value from a plurality of logarithms of the inverse of the expected error variance and a plurality of inverse-scaled sea surface temperature anomalies of the expected error variance, comprising:
Calculating the average value of the logarithms of the expected error variance reciprocal to obtain the logarithm of the target expected error variance reciprocal;
calculating the average value of the inverse scaled sea surface temperature anomalies of the plurality of expected error variances to obtain the inverse scaled sea surface temperature anomalies of the target expected error variances;
calculating according to abnormal value of sea surface temperatureCalculating the target sea surface temperature anomaly value, wherein +_>Representing the abnormal value of the target sea surface temperature, < >>Sea surface temperature anomalies, which represent reciprocal scaling of the target expected error variance,/>Representing error variance->，，/>，/>Representing the logarithm of the inverse of the target expected error variance.
In a second aspect, an embodiment of the present invention further provides a device for filling and correcting sea surface temperature data, including:
the gridding pretreatment unit is used for carrying out gridding pretreatment on the sea surface temperature data observed on site to obtain gridding sea surface temperature data observed on site;
the weighted fusion unit is used for carrying out weighted fusion on satellite remote sensing sea surface temperature data and the on-site observed grid sea surface temperature data to obtain fused sea surface temperature data;
the mathematical operation unit is used for performing mathematical operation on the fused sea surface temperature data to obtain a mathematical operation data matrix, wherein the data in the mathematical operation data matrix comprises: inverse of error variance, scaled sea surface temperature anomaly, normalized longitude and latitude data, and normalized time data;
The iterative training unit is used for inputting the mathematical operation data matrix into a convolution self-encoder of a double-layer coupling structure to carry out iterative training and outputting a plurality of logarithms of expected error variance reciprocal pairs and sea surface temperature anomalies of reciprocal scaling of the expected error variance pairs in the iterative training process;
a calculation unit for calculating a target sea surface temperature anomaly value from a plurality of logarithms of the inverse of the expected error variance and a plurality of inverse-scaled sea surface temperature anomalies of the expected error variance;
and the correction unit is used for correcting the arithmetic average value corresponding to the fused sea surface temperature data by adopting the target sea surface temperature abnormal value to obtain corrected and filled sea surface temperature data.
In a third aspect, an embodiment of the present invention further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the steps of the method according to any one of the first aspects when the processor executes the computer program.
In a fourth aspect, embodiments of the present invention also provide a computer-readable storage medium storing machine-executable instructions which, when invoked and executed by a processor, cause the processor to perform the method of any one of the first aspects.
In an embodiment of the present invention, a method for filling and correcting sea surface temperature data is provided, including: performing gridding pretreatment on the sea surface temperature data observed on site to obtain gridding sea surface temperature data observed on site; carrying out weighted fusion on satellite remote sensing sea surface temperature data and on-site observed grid sea surface temperature data to obtain fused sea surface temperature data; performing mathematical operation on the fused sea surface temperature data to obtain a mathematical operation data matrix, wherein the data in the mathematical operation data matrix comprises: inverse of error variance, scaled sea surface temperature anomaly, normalized longitude and latitude data, and normalized time data; inputting a mathematical operation data matrix into a convolution self-encoder of a double-layer coupling structure for iterative training, and outputting a plurality of logarithms of expected error variance reciprocal values and sea surface temperature anomalies scaled by the reciprocal values of the expected error variance values in the iterative training process; calculating a target sea surface temperature anomaly value from the logarithm of the plurality of expected error variance reciprocals and the sea surface temperature anomaly scaled by the plurality of expected error variance reciprocals; and correcting an arithmetic average value corresponding to the fused sea surface temperature data by adopting the target sea surface temperature abnormal value to obtain corrected and filled sea surface temperature data. According to the method for filling and correcting the sea surface temperature data, the sea surface temperature data observed on site and the satellite remote sensing sea surface temperature data are fused together and are filled and corrected through a deep learning algorithm (namely, a convolution self-encoder of a double-layer coupling structure), the satellite remote sensing sea surface temperature data are utilized, the sea surface temperature data observed on site are utilized, potential application of the sea surface temperature data is promoted, in addition, the convolution self-encoder of the double-layer coupling structure can capture instantaneous and small-scale characteristic information of the fused sea surface temperature data, further model result precision is improved, time dimension characteristics and space dimension characteristics (comprising standardized longitude and latitude data and standardized time data in a mathematical operation data matrix) of the fused sea surface temperature data can be simultaneously extracted and correlated, and finally, space-time coverage rate and precision of the corrected and filled sea surface temperature data are improved, and the technical problems of low space-time coverage rate and precision of the corrected and filled sea surface temperature data obtained by the traditional sea surface temperature data filling method are relieved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for filling and correcting sea surface temperature data according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a convolutional self-encoder according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a filling and correcting device for sea surface temperature data according to an embodiment of the present invention;
fig. 4 is a schematic diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The technical solutions of the present invention will be clearly and completely described in connection with the embodiments, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The space-time coverage rate and the precision of the corrected and filled sea surface temperature data obtained by the traditional sea surface temperature data correction and filling method are low.
Based on the above, in the filling and correcting method of the sea surface temperature data, the sea surface temperature data observed on site and the satellite remote sensing sea surface temperature data are fused together and are filled and corrected through a deep learning algorithm (namely, a convolution self-encoder of a double-layer coupling structure), the satellite remote sensing sea surface temperature data are utilized, the sea surface temperature data observed on site are utilized, potential application of the sea surface temperature data is promoted, in addition, the convolution self-encoder of the double-layer coupling structure can capture instantaneous and small-scale characteristic information of the fused sea surface temperature data, further, model result precision is improved, time dimension characteristics and space dimension characteristics (including standardized longitude and latitude data and standardized time data in a mathematical operation data matrix) of the fused sea surface temperature data can be simultaneously extracted and correlated, and finally, space-time coverage rate and precision of the corrected and filled sea surface temperature data are improved.
For the convenience of understanding the present embodiment, a method for filling and correcting sea surface temperature data disclosed in the present embodiment will be described in detail.
Embodiment one:
according to an embodiment of the present invention, there is provided an embodiment of a method of padding and modifying sea surface temperature data, it being noted that the steps illustrated in the flow chart of the drawings may be performed in a computer system such as a set of computer executable instructions, and that although a logical sequence is illustrated in the flow chart, in some cases the steps illustrated or described may be performed in a different order than that illustrated herein.
Fig. 1 is a flowchart of a method for filling and correcting sea surface temperature data according to an embodiment of the present invention, as shown in fig. 1, the method includes the following steps:
step S102, performing gridding pretreatment on the field-observed sea surface temperature data to obtain the field-observed gridded sea surface temperature data;
in the embodiment of the invention, the sea surface temperature data observed on site and the satellite remote sensing sea surface temperature data can be obtained, and after the two data are obtained, the sea surface temperature data observed on site is subjected to gridding pretreatment, and the process of gridding pretreatment is described in detail hereinafter and is not described in detail.
The satellite remote sensing sea surface temperature data may be sea surface temperature data observed by an infrared radiometer, and the satellite remote sensing sea surface temperature data is not particularly limited in the embodiment of the invention.
Step S104, carrying out weighted fusion on satellite remote sensing sea surface temperature data and on-site observed grid sea surface temperature data to obtain fused sea surface temperature data;
step S106, carrying out mathematical operation on the fused sea surface temperature data to obtain a mathematical operation data matrix, wherein the data in the mathematical operation data matrix comprises: inverse of error variance, scaled sea surface temperature anomaly, normalized longitude and latitude data, and normalized time data;
step S108, inputting a mathematical operation data matrix into a convolution self-encoder of a double-layer coupling structure for iterative training, and outputting a plurality of logarithms of expected error variance reciprocal pairs and sea surface temperature anomalies of reciprocal scaling of the expected error variances in the iterative training process;
specifically, assuming that the training is iterated 1000 times, the sea surface temperature abnormality which outputs a logarithm of the inverse of the expected error variance and a reciprocal scaling of the expected error variance every 100 times is set, so that a plurality of logarithms of the inverse of the expected error variance and a plurality of reciprocal scaling of the expected error variance can be obtained.
Step S110, calculating a target sea surface temperature anomaly value according to the logarithm of the inverse of the plurality of expected error variances and the sea surface temperature anomaly scaled by the inverse of the plurality of expected error variances;
And S112, correcting an arithmetic average value corresponding to the fused sea surface temperature data by adopting the target sea surface temperature abnormal value to obtain corrected and filled sea surface temperature data.
Specifically, the abnormal value of the target sea surface temperature is added with the arithmetic average value to obtain corrected and filled sea surface temperature data.
According to the comparison of related experiments, the corrected and filled sea surface temperature data obtained by the method greatly improves the space-time coverage rate of the sea surface temperature data.
In an embodiment of the present invention, a method for filling and correcting sea surface temperature data is provided, including: performing gridding pretreatment on the sea surface temperature data observed on site to obtain gridding sea surface temperature data observed on site; carrying out weighted fusion on satellite remote sensing sea surface temperature data and on-site observed grid sea surface temperature data to obtain fused sea surface temperature data; performing mathematical operation on the fused sea surface temperature data to obtain a mathematical operation data matrix, wherein the data in the mathematical operation data matrix comprises: inverse of error variance, scaled sea surface temperature anomaly, normalized longitude and latitude data, and normalized time data; inputting a mathematical operation data matrix into a convolution self-encoder of a double-layer coupling structure for iterative training, and outputting a plurality of logarithms of expected error variance reciprocal values and sea surface temperature anomalies scaled by the reciprocal values of the expected error variance values in the iterative training process; calculating a target sea surface temperature anomaly value from the logarithm of the plurality of expected error variance reciprocals and the sea surface temperature anomaly scaled by the plurality of expected error variance reciprocals; and correcting an arithmetic average value corresponding to the fused sea surface temperature data by adopting the target sea surface temperature abnormal value to obtain corrected and filled sea surface temperature data. According to the method for filling and correcting the sea surface temperature data, the sea surface temperature data observed on site and the satellite remote sensing sea surface temperature data are fused together and are filled and corrected through a deep learning algorithm (namely, a convolution self-encoder of a double-layer coupling structure), the satellite remote sensing sea surface temperature data are utilized, the sea surface temperature data observed on site are utilized, potential application of the sea surface temperature data is promoted, in addition, the convolution self-encoder of the double-layer coupling structure can capture instantaneous and small-scale characteristic information of the fused sea surface temperature data, further model result precision is improved, time dimension characteristics and space dimension characteristics (comprising standardized longitude and latitude data and standardized time data in a mathematical operation data matrix) of the fused sea surface temperature data can be simultaneously extracted and correlated, and finally, space-time coverage rate and precision of the corrected and filled sea surface temperature data are improved, and the technical problems of low space-time coverage rate and precision of the corrected and filled sea surface temperature data obtained by the traditional sea surface temperature data filling method are relieved.
The above-mentioned contents briefly introduce the filling and correction method of sea surface temperature data of the present invention, and the detailed description of the contents is provided below.
In an alternative embodiment of the invention, the on-site observed sea surface temperature data is subjected to gridding pretreatment, and the method specifically comprises the following steps:
(1) Constructing an empty grid with the same size as the grid of the satellite remote sensing sea surface temperature data;
specifically, each line of the grid represents a latitude and longitude.
(2) Adding field-observed sea surface temperature data into the empty grid, and calculating the field-observed sea surface temperature data of the grid adjacent to the grid where the field-observed sea surface temperature data is located by adopting an interpolation algorithm;
specifically, when the on-site observed sea surface temperature data is added, the on-site observed sea surface temperature data is placed in the grid according to longitude and latitude information in the on-site observed sea surface temperature data, and meanwhile, an interpolation algorithm is adopted to calculate on-site observed sea surface temperature data of a grid adjacent to the grid where the on-site observed sea surface temperature data is located.
When the sea surface temperature data observed in the field is on the grid point, the sea surface temperature data is used as the grid point; when the on-site observed sea surface temperature data are located in the grids, calculating on-site observed sea surface temperature data of the adjacent four grids by using an interpolation algorithm; when the field observed sea surface temperature data is positioned at the grid boundary, calculating the field observed sea surface temperature data of two adjacent grids by using an interpolation algorithm.
(3) And placing the calculated on-site observed sea surface temperature data adjacent to the grid into the corresponding adjacent grid to obtain on-site observed meshed sea surface temperature data.
The purpose of the above-mentioned gridding is to perform weighted fusion with satellite remote sensing sea surface temperature data, and because satellite remote sensing sea surface temperature data is gridding, it is also necessary to gridde sea surface temperature data observed on site, and meanwhile, difference is also performed during gridding, so as to increase data volume.
In an alternative embodiment of the present invention, the satellite remote sensing sea surface temperature data and the on-site observed grid sea surface temperature data are subjected to weighted fusion, and the method specifically includes the following steps:
(1) Acquiring a preset weight;
specifically, the preset weight refers to a weight corresponding to satellite remote sensing sea surface temperature data and a weight corresponding to on-site observed grid sea surface temperature data.
(2) And carrying out weighted fusion on the satellite remote sensing sea surface temperature data and the on-site observed grid sea surface temperature data according to preset weights to obtain fused sea surface temperature data.
In an alternative embodiment of the invention, the mathematical operation is performed on the fused sea surface temperature data, and specifically comprises the following steps:
(1) Calculating error variance of the fused sea surface temperature data, and determining error variance reciprocal according to the error variance;
specifically, the root mean square error of the fused sea surface temperature data is calculated, and the variance of the root mean square error (i.e., the error variance) is calculated.
The error variance reciprocal is specifically the reciprocal of the error variance of the fused sea surface temperature data of T-1 day, T day and T+1 day.
(2) Calculating an arithmetic mean of the fused sea surface temperature data;
(3) Calculating a sea surface temperature anomaly value according to the fused sea surface temperature data and the arithmetic average value;
specifically, sea surface temperature outlier = fused sea surface temperature data-arithmetic mean.
(4) Scaling the corresponding sea surface temperature abnormal value by adopting the reciprocal of the error variance to obtain a scaled sea surface temperature abnormal value;
how to exemplify, the inverse of the error variance of the fused sea surface temperature data obtained in the T-1 day, T day and t+1 day is scaled according to the inverse of the error variance of the fused sea surface temperature data in the T-1 day, T day and t+1 day, so as to obtain the scaled sea surface temperature anomaly.
(5) Carrying out standardization processing on the longitude and latitude data to obtain standardized longitude and latitude data;
(6) And (3) carrying out standardization processing on the time data to obtain standardized time data, and further obtaining a mathematical operation data matrix consisting of the inverse of the error variance, the scaled sea surface temperature abnormal value, the standardized longitude and latitude data and the standardized time data.
The mathematical operation data matrix is specifically 10×m×n×h, where m and n represent rows and columns of the fused sea surface temperature data, and H is the length of the whole small batch fused sea surface temperature data.
In an alternative embodiment of the present invention, the mathematical operation data matrix is input into a convolution self-encoder with a double-layer coupling structure for iterative training, which specifically comprises the following steps:
(1) Inputting the mathematical operation data matrix into a first convolution self-encoder to obtain a first output result;
(2) The first output result and the mathematical operation data matrix are input into a second convolution self-encoder, and the sea surface temperature anomaly which obtains the logarithm of the expected error variance reciprocal and the reciprocal scaling of the expected error variance is output.
Specifically, each of the convolutional self-encoders of the double-layer coupling structure includes: an encoder and a decoder;
Referring to fig. 2, the encoder includes: a plurality of coding layers, each coding layer comprising: a convolution layer, an attention module, an activation function (not shown in the figure), and a pooling layer;
the decoder includes: decoding layers of the same number as the number of encoding layers, each decoding layer including: an upsampling layer, an attention module, and a convolution layer;
a summation skip connection (i.e., an arrow connection between each encoding layer and decoding layer in fig. 2) is employed between the encoding layer and decoding layer.
In an embodiment of the present invention, a two-layer convolutional self-encoder neural network is constructed, the convolutional self-encoder of each layer is composed of an encoder and a decoder, the encoder comprises a plurality of encoding layers, and each encoding layer comprises: a convolution layer (where the convolution kernel is a filter), a attention module, an activation function, and a pooling layer. The decoder restores the output result of the encoder to original data, the decoder comprising: the decoding layer having the same number as the encoding layer includes: the system comprises an up-sampling layer, an attention module and a convolution layer, wherein an interpolation algorithm is added in the up-sampling layer, the interpolation algorithm in the up-sampling layer adopts nearest neighbor interpolation or bilinear interpolation, and the convolution layer is added after the up-sampling layer.
In order to adapt to the data in the model operation process, a convolution layer is used to replace a full-connection layer (the last coding layer in the coder is the full-connection layer originally), and the invention changes the convolution layer. The whole model adopts a convolution self-encoder with a double-layer coupling structure, so that the depth of the model is increased, and the training precision of the model is improved.
In addition, the model provides that the summation skip connection is used in the coding-decoding structure of each layer, so that the characteristic data quantity of model output is reduced, and the weight and deviation of the convolution layer are more relevant to the output of the neural network, so that the proposal of the solution of the summation skip connection ensures that the characteristic quantity of model output is the same, and the problem of model gradient disappearance is reduced. The sum skip connection formula is as follows:
wherein,results showing the convolution layer, attention module and pooling layer in FIG. 2, +.>Representing the results of the upsampling layer, attention module and convolution layer in fig. 2,/for the example>Representing the summation result.
The large-scale space-time characteristics of the sea surface temperature data are easy to capture in the training process, but key instantaneous and small-scale characteristics are subjected to model training smoothing treatment, and the instantaneous and small-scale characteristic information of the sea surface temperature data can be further captured by skipping connection, so that the accuracy of a model result is improved.
In addition, to prevent the gradient from dropping too fast during model training, a leak RELU is used, with the following formula:
the loss function of the first convolution self-encoder adopts a Gaussian negative log likelihood function, and the formula is as follows:
indicating that the reconstructed sea surface temperature is abnormal, < >>Representing the corresponding error variance +. >Representing the number of significant values.
After the output result of the first convolution self-encoder is obtained, the mathematical operation data matrix is connected with the output result of the first convolution self-encoder to serve as the input of the second convolution self-encoder, and the loss function is constructed through the output of the second convolution self-encoder and the output of the first convolution self-encoder, and the intermediate output result is incorporated into the construction function to improve the model training precision. The loss function calculation formula is as follows:
wherein,、/>representing the reconstruction produced by the second convolution self-encoder and its expected error variance, respectively.
The deep learning model optimizer selects Adam, the initial learning rate is 0.002, and the learning rate attenuation is set as follows:
wherein,representing the initial learning rate, < >>Exponential decay indicative of control learning rate, +.>Representing the number of training rounds.
Meanwhile, on the basis of the convolution self-encoder of each layer, an attention mechanism module is introduced, and the model adopts a CBAM attention mechanism algorithm, and the formula is expressed as follows:
wherein,representing an input profile, < >>Representing global average pooling operations,/->Representing global maximization,/-pooling>And->Weight matrix representing fully connected layer, +.>Representing an activation function->Representing the channel attention vector, " >Representing a spatial attention vector, ">Representing convolution operations +.>Representing the weighted output, obtaining +.>By spatial attention mechanisms maximum pooling and average pooling to give +.>，/>Representing a multi-layer perceptron.
And the attention mechanism module is used for extracting the internal relation between the time dimension characteristic and the space dimension characteristic of the sea surface temperature data, so that the accuracy of a model reconstruction result is improved, and the real physical characteristic of the sea surface temperature data is further reflected.
In an alternative embodiment of the present invention, the calculation of the target sea surface temperature anomaly value from the logarithm of the plurality of inverse desired error variances and the inverse scaled sea surface temperature anomaly of the plurality of inverse desired error variances specifically comprises the steps of:
(1) Calculating the average value of the logarithms of the expected error variance reciprocal to obtain the logarithm of the target expected error variance reciprocal;
(2) Calculating the average value of the inverse scaled sea surface temperature anomalies of the plurality of expected error variances to obtain the inverse scaled sea surface temperature anomalies of the target expected error variances;
(3) Calculating according to abnormal value of sea surface temperatureCalculating an abnormal value of the target sea surface temperature, wherein +_ >Indicating the abnormal value of the target sea surface temperature, +.>Sea surface temperature anomalies representing reciprocal scaling of target expected error variance, +.>Representing error variance->，/>，，/>Representing the logarithm of the inverse of the target expected error variance.
Since the model has strong randomness of weight and deviation, the method calculates the minimum and maximum functions (i.eMax and min in the formula) to stabilize the network model and improve the robustness of the model.
The original sea surface temperature data product and the sea surface temperature data product obtained by the method are respectively compared with field observation data to obtain root mean square error, the root mean square error is reduced to 0.2750 ℃ from 0.3948 ℃, and the linear correlation coefficient is improved to 0.9941 from 0.988, so that the model result and the observation data have high consistency, the effectiveness and the accuracy of the method are verified.
The invention provides a filling and correcting method of sea surface temperature data, in the method, a model realizes reconstruction and filling of sea surface temperature data observed by an infrared radiometer by utilizing the relation between sea surface temperature data observed by the infrared radiometer and sea surface temperature data observed on site, and space-time coverage rate and precision of the existing sea surface temperature data product are improved.
Embodiment two:
the embodiment of the invention also provides a device for filling and correcting the sea surface temperature data, which is mainly used for executing the method for filling and correcting the sea surface temperature data provided in the first embodiment of the invention, and the device for filling and correcting the sea surface temperature data provided in the first embodiment of the invention is specifically introduced below.
Fig. 3 is a schematic diagram of a device for filling and correcting sea surface temperature data according to an embodiment of the present invention, as shown in fig. 3, the device mainly includes: the device comprises a gridding preprocessing unit 10, a weighted fusion unit 20, a mathematical operation unit 30, an iterative training unit 40, a calculation unit 50 and a correction unit 60, wherein:
the gridding pretreatment unit is used for carrying out gridding pretreatment on the sea surface temperature data observed on site to obtain gridding sea surface temperature data observed on site;
the weighted fusion unit is used for carrying out weighted fusion on the satellite remote sensing sea surface temperature data and the on-site observed grid sea surface temperature data to obtain fused sea surface temperature data;
the mathematical operation unit is used for carrying out mathematical operation on the fused sea surface temperature data to obtain a mathematical operation data matrix, wherein the data in the mathematical operation data matrix comprises: inverse of error variance, scaled sea surface temperature anomaly, normalized longitude and latitude data, and normalized time data;
The iterative training unit is used for inputting the mathematical operation data matrix into a convolution self-encoder of a double-layer coupling structure to carry out iterative training and outputting a plurality of logarithms of expected error variance reciprocal pairs and sea surface temperature anomalies of reciprocal scaling of the expected error variance pairs in the iterative training process;
a calculation unit for calculating a target sea surface temperature anomaly value from the logarithm of the plurality of expected error variance reciprocals and the sea surface temperature anomaly scaled by the reciprocals of the plurality of expected error variances;
and the correction unit is used for correcting the arithmetic average value corresponding to the fused sea surface temperature data by adopting the target sea surface temperature abnormal value to obtain corrected and filled sea surface temperature data.
In an embodiment of the present invention, a filling and correcting device for sea surface temperature data is provided, including: performing gridding pretreatment on the sea surface temperature data observed on site to obtain gridding sea surface temperature data observed on site; carrying out weighted fusion on satellite remote sensing sea surface temperature data and on-site observed grid sea surface temperature data to obtain fused sea surface temperature data; performing mathematical operation on the fused sea surface temperature data to obtain a mathematical operation data matrix, wherein the data in the mathematical operation data matrix comprises: inverse of error variance, scaled sea surface temperature anomaly, normalized longitude and latitude data, and normalized time data; inputting a mathematical operation data matrix into a convolution self-encoder of a double-layer coupling structure for iterative training, and outputting a plurality of logarithms of expected error variance reciprocal values and sea surface temperature anomalies scaled by the reciprocal values of the expected error variance values in the iterative training process; calculating a target sea surface temperature anomaly value from the logarithm of the plurality of expected error variance reciprocals and the sea surface temperature anomaly scaled by the plurality of expected error variance reciprocals; and correcting an arithmetic average value corresponding to the fused sea surface temperature data by adopting the target sea surface temperature abnormal value to obtain corrected and filled sea surface temperature data. According to the sea surface temperature data filling and correcting device, field-observed sea surface temperature data and satellite remote sensing sea surface temperature data are fused together and are filled and corrected through a deep learning algorithm (namely, a convolution self-encoder of a double-layer coupling structure), the satellite remote sensing sea surface temperature data are utilized, the field-observed sea surface temperature data are utilized, potential application of the sea surface temperature data is promoted, in addition, the convolution self-encoder of the double-layer coupling structure can capture instantaneous and small-scale characteristic information of the fused sea surface temperature data, further model result precision is improved, time dimension characteristics and space dimension characteristics (comprising standardized longitude and latitude data and standardized time data in a mathematical operation data matrix) of the fused sea surface temperature data can be simultaneously extracted and correlated, and finally, space-time coverage rate and precision of corrected and filled sea surface temperature data are improved, and the technical problems of low space-time coverage rate and precision of the corrected and filled sea surface temperature data obtained by a traditional sea surface temperature data correcting and filling method are solved.
Optionally, the meshing preprocessing unit is further configured to: constructing an empty grid with the same size as the grid of the satellite remote sensing sea surface temperature data; adding field-observed sea surface temperature data into the empty grid, and calculating the field-observed sea surface temperature data of the grid adjacent to the grid where the field-observed sea surface temperature data is located by adopting an interpolation algorithm; and placing the calculated on-site observed sea surface temperature data adjacent to the grid into the corresponding adjacent grid to obtain on-site observed meshed sea surface temperature data.
Optionally, the weighted fusion unit is further configured to: acquiring a preset weight; and carrying out weighted fusion on the satellite remote sensing sea surface temperature data and the on-site observed grid sea surface temperature data according to preset weights to obtain fused sea surface temperature data.
Optionally, the mathematical operation unit is further configured to: calculating error variance of the fused sea surface temperature data, and determining error variance reciprocal according to the error variance; calculating an arithmetic mean of the fused sea surface temperature data; calculating a sea surface temperature anomaly value according to the fused sea surface temperature data and the arithmetic average value; scaling the corresponding sea surface temperature abnormal value by adopting the reciprocal of the error variance to obtain a scaled sea surface temperature abnormal value; carrying out standardization processing on the longitude and latitude data to obtain standardized longitude and latitude data; and (3) carrying out standardization processing on the time data to obtain standardized time data, and further obtaining a mathematical operation data matrix consisting of the inverse of the error variance, the scaled sea surface temperature abnormal value, the standardized longitude and latitude data and the standardized time data.
Optionally, the iterative training unit is further configured to: inputting the mathematical operation data matrix into a first convolution self-encoder to obtain a first output result; the first output result and the mathematical operation data matrix are input into a second convolution self-encoder, and the sea surface temperature anomaly which obtains the logarithm of the expected error variance reciprocal and the reciprocal scaling of the expected error variance is output.
Optionally, each of the convolutional self-encoders of the double-layer coupling structure comprises: an encoder and a decoder; the encoder includes: a plurality of coding layers, each coding layer comprising: a convolution layer, an attention module, an activation function and a pooling layer; the decoder includes: decoding layers of the same number as the number of encoding layers, each decoding layer including: an upsampling layer, an attention module, and a convolution layer; a summation skip connection is employed between the encoding and decoding layers.
Optionally, the computing unit is further configured to: calculating the average value of the logarithms of the expected error variance reciprocal to obtain the target expected error variance reciprocalLogarithm; calculating the average value of the inverse scaled sea surface temperature anomalies of the plurality of expected error variances to obtain the inverse scaled sea surface temperature anomalies of the target expected error variances; calculating according to abnormal value of sea surface temperature Calculating an abnormal value of the target sea surface temperature, wherein +_>Indicating the abnormal value of the target sea surface temperature,sea surface temperature anomalies representing reciprocal scaling of target expected error variance, +.>The variance of the error is represented as,，/>，/>，/>representing the logarithm of the inverse of the target expected error variance.
The device provided by the embodiment of the present invention has the same implementation principle and technical effects as those of the foregoing method embodiment, and for the sake of brevity, reference may be made to the corresponding content in the foregoing method embodiment where the device embodiment is not mentioned.
As shown in fig. 4, an electronic device 600 provided in an embodiment of the present application includes: the sea surface temperature data filling and correcting method comprises a processor 601, a memory 602 and a bus, wherein the memory 602 stores machine-readable instructions executable by the processor 601, when the electronic device is running, the processor 601 and the memory 602 are communicated through the bus, and the processor 601 executes the machine-readable instructions to execute the steps of the sea surface temperature data filling and correcting method.
Specifically, the memory 602 and the processor 601 can be general-purpose memories and processors, and are not limited herein, and the above-mentioned filling and correction method of sea surface temperature data can be performed when the processor 601 runs a computer program stored in the memory 602.
The processor 601 may be an integrated circuit chip with signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in the processor 601 or instructions in the form of software. The processor 601 may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU for short), a network processor (Network Processor, NP for short), etc.; but may also be a digital signal processor (Digital Signal Processing, DSP for short), application specific integrated circuit (Application Specific Integrated Circuit, ASIC for short), off-the-shelf programmable gate array (Field-Programmable Gate Array, FPGA for short), or other programmable logic device, discrete gate or transistor logic device, discrete hardware components. The disclosed methods, steps, and logic blocks in the embodiments of the present application may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the embodiments of the present application may be embodied directly in hardware, in a decoded processor, or in a combination of hardware and software modules in a decoded processor. The software modules may be located in a random access memory, flash memory, read only memory, programmable read only memory, or electrically erasable programmable memory, registers, etc. as well known in the art. The storage medium is located in a memory 602, and the processor 601 reads information in the memory 602 and performs the steps of the above method in combination with its hardware.
Corresponding to the above sea surface temperature data filling and correcting method, the embodiment of the application also provides a computer readable storage medium, wherein the computer readable storage medium stores machine executable instructions, and the computer executable instructions, when being called and executed by a processor, cause the processor to execute the steps of the sea surface temperature data filling and correcting method.
The filling and correcting device for the sea surface temperature data provided by the embodiment of the application can be specific hardware on equipment or software or firmware installed on the equipment and the like. The device provided in the embodiments of the present application has the same implementation principle and technical effects as those of the foregoing method embodiments, and for a brief description, reference may be made to corresponding matters in the foregoing method embodiments where the device embodiment section is not mentioned. It will be clear to those skilled in the art that, for convenience and brevity, the specific operation of the system, apparatus and unit described above may refer to the corresponding process in the above method embodiment, which is not described in detail herein.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. The above-described apparatus embodiments are merely illustrative, for example, the division of the units is merely a logical function division, and there may be other manners of division in actual implementation, and for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some communication interface, device or unit indirect coupling or communication connection, which may be in electrical, mechanical or other form.
As another example, the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments provided in the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing an electronic device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the filling and correcting method of sea surface temperature data according to the various embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
It should be noted that: like reference numerals and letters in the following figures denote like items, and thus once an item is defined in one figure, no further definition or explanation of it is required in the following figures, and furthermore, the terms "first," "second," "third," etc. are used merely to distinguish one description from another and are not to be construed as indicating or implying relative importance.
Finally, it should be noted that: the foregoing examples are merely specific embodiments of the present application, and are not intended to limit the scope of the present application, but the present application is not limited thereto, and those skilled in the art will appreciate that while the foregoing examples are described in detail, the present application is not limited thereto. Any person skilled in the art may modify or easily conceive of the technical solution described in the foregoing embodiments, or make equivalent substitutions for some of the technical features within the technical scope of the disclosure of the present application; such modifications, changes or substitutions do not depart from the spirit of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application. Are intended to be encompassed within the scope of this application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A filling and correcting method for sea surface temperature data is characterized by comprising the following steps:
performing gridding pretreatment on the sea surface temperature data observed on site to obtain gridding sea surface temperature data observed on site;
carrying out weighted fusion on satellite remote sensing sea surface temperature data and the on-site observed grid sea surface temperature data to obtain fused sea surface temperature data;
performing mathematical operation on the fused sea surface temperature data to obtain a mathematical operation data matrix, wherein the data in the mathematical operation data matrix comprises: inverse of error variance, scaled sea surface temperature anomaly, normalized longitude and latitude data, and normalized time data;
inputting the mathematical operation data matrix into a convolution self-encoder of a double-layer coupling structure for iterative training, and outputting a plurality of logarithms of expected error variance reciprocal values and sea surface temperature anomalies scaled by the plurality of reciprocal values of expected error variance values in the iterative training process;
calculating a target sea surface temperature anomaly value from a plurality of logarithms of the expected error variance reciprocals and a plurality of sea surface temperature anomalies scaled by the expected error variance reciprocals;
And correcting an arithmetic average value corresponding to the fused sea surface temperature data by adopting the target sea surface temperature abnormal value to obtain corrected and filled sea surface temperature data.
2. The method of claim 1, wherein the step of gridding the sea surface temperature data observed in situ comprises:
constructing an empty grid with the same size as the grid of the satellite remote sensing sea surface temperature data;
adding the field observed sea surface temperature data into the empty grid, and calculating the field observed sea surface temperature data of the adjacent grid of the grid where the field observed sea surface temperature data is positioned by adopting an interpolation algorithm;
and placing the calculated on-site observed sea surface temperature data adjacent to the grid into the corresponding adjacent grid to obtain the on-site observed meshed sea surface temperature data.
3. The method of claim 1, wherein weighted fusion of satellite remote sensing sea surface temperature data with the site observed meshed sea surface temperature data comprises:
acquiring a preset weight;
and carrying out weighted fusion on the satellite remote sensing sea surface temperature data and the on-site observed grid sea surface temperature data according to the preset weight to obtain the fused sea surface temperature data.
4. The method of claim 1, wherein mathematically operating on the fused sea surface temperature data comprises:
calculating an error variance of the fused sea surface temperature data, and determining the inverse of the error variance according to the error variance;
calculating an arithmetic mean of the fused sea surface temperature data;
calculating the abnormal value of the sea surface temperature according to the fused sea surface temperature data and the arithmetic average value;
scaling the sea surface temperature abnormal value corresponding to the inverse of the error variance to obtain the scaled sea surface temperature abnormal value;
carrying out standardization processing on the longitude and latitude data to obtain the standardized longitude and latitude data;
and carrying out standardization processing on the time data to obtain the standardized time data, and further obtaining a mathematical operation data matrix composed of the inverse of the error variance, the scaled sea surface temperature abnormal value, the standardized longitude and latitude data and the standardized time data.
5. The method of claim 1, wherein inputting the matrix of mathematical operation data into a convolutional self-encoder of a double-layer coupling structure for iterative training comprises:
Inputting the mathematical operation data matrix into a first convolution self-encoder to obtain a first output result;
and inputting the first output result and the mathematical operation data matrix into a second convolution self-encoder, and outputting sea surface temperature anomalies which are obtained by scaling the logarithm of the inverse of the expected error variance and the inverse of the expected error variance.
6. The method of claim 1, wherein each of the convolutional self encoders of the double-layer coupling structure comprises: an encoder and a decoder;
the encoder includes: a plurality of coding layers, each of the coding layers comprising: a convolution layer, an attention module, an activation function and a pooling layer;
the decoder includes: the same number of decoding layers as the encoding layers, each decoding layer comprising: an upsampling layer, an attention module, and a convolution layer;
a summing skip connection is employed between the encoding layer and the decoding layer.
7. The method of claim 1, wherein calculating a target sea surface temperature anomaly value from a plurality of logarithms of the desired error variance reciprocals and a plurality of sea surface temperature anomalies scaled by the desired error variances, comprises:
Calculating the average value of the logarithms of the expected error variance reciprocal to obtain the logarithm of the target expected error variance reciprocal;
calculating the average value of the inverse scaled sea surface temperature anomalies of the plurality of expected error variances to obtain the inverse scaled sea surface temperature anomalies of the target expected error variances;
calculating according to abnormal value of sea surface temperatureCalculating the target sea surface temperature anomaly value, wherein +_>Representing the abnormal value of the target sea surface temperature, < >>Sea surface temperature anomalies, which represent reciprocal scaling of the target expected error variance,/>Representing error variance->，/>，，/>Representing the logarithm of the inverse of the target expected error variance.
8. A filling and correction device for sea surface temperature data, comprising:
the gridding pretreatment unit is used for carrying out gridding pretreatment on the sea surface temperature data observed on site to obtain gridding sea surface temperature data observed on site;
the weighted fusion unit is used for carrying out weighted fusion on satellite remote sensing sea surface temperature data and the on-site observed grid sea surface temperature data to obtain fused sea surface temperature data;
the mathematical operation unit is used for performing mathematical operation on the fused sea surface temperature data to obtain a mathematical operation data matrix, wherein the data in the mathematical operation data matrix comprises: inverse of error variance, scaled sea surface temperature anomaly, normalized longitude and latitude data, and normalized time data;
The iterative training unit is used for inputting the mathematical operation data matrix into a convolution self-encoder of a double-layer coupling structure to carry out iterative training and outputting a plurality of logarithms of expected error variance reciprocal pairs and sea surface temperature anomalies of reciprocal scaling of the expected error variance pairs in the iterative training process;
a calculation unit for calculating a target sea surface temperature anomaly value from a plurality of logarithms of the inverse of the expected error variance and a plurality of inverse-scaled sea surface temperature anomalies of the expected error variance;
and the correction unit is used for correcting the arithmetic average value corresponding to the fused sea surface temperature data by adopting the target sea surface temperature abnormal value to obtain corrected and filled sea surface temperature data.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method of any of the preceding claims 1 to 7 when the computer program is executed.
10. A computer readable storage medium storing machine executable instructions which, when invoked and executed by a processor, cause the processor to perform the method of any one of the preceding claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410096143.0A CN117633441B (en) | 2024-01-24 | 2024-01-24 | Sea surface temperature data filling and correcting method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410096143.0A CN117633441B (en) | 2024-01-24 | 2024-01-24 | Sea surface temperature data filling and correcting method and device and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117633441A true CN117633441A (en) | 2024-03-01 |
| CN117633441B CN117633441B (en) | 2024-05-10 |
Family
ID=90021887
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410096143.0A Active CN117633441B (en) | 2024-01-24 | 2024-01-24 | Sea surface temperature data filling and correcting method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117633441B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118097455A (en) * | 2024-04-29 | 2024-05-28 | 自然资源部第一海洋研究所 | Sea surface temperature data reconstruction method based on deep learning algorithm |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109668635A (en) * | 2019-01-16 | 2019-04-23 | 中国人民解放军61741部队 | Sea surface temperature fusion method and system |
| CN113051795A (en) * | 2021-03-15 | 2021-06-29 | 哈尔滨工程大学 | Three-dimensional temperature-salinity field analysis and prediction method for offshore platform guarantee |
| CN114721264A (en) * | 2022-03-21 | 2022-07-08 | 浙江工业大学 | Industrial information physical system attack detection method based on two-stage self-encoder |
| CN114779368A (en) * | 2022-03-22 | 2022-07-22 | 同济大学 | ENSO prediction method, device, equipment and medium based on multivariable sea air coupler |
| CN115393540A (en) * | 2022-08-30 | 2022-11-25 | 无锡九方科技有限公司 | Intelligent fusion method and system of three-dimensional marine environment field based on deep learning |
-
2024
- 2024-01-24 CN CN202410096143.0A patent/CN117633441B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109668635A (en) * | 2019-01-16 | 2019-04-23 | 中国人民解放军61741部队 | Sea surface temperature fusion method and system |
| CN113051795A (en) * | 2021-03-15 | 2021-06-29 | 哈尔滨工程大学 | Three-dimensional temperature-salinity field analysis and prediction method for offshore platform guarantee |
| US20220326211A1 (en) * | 2021-03-15 | 2022-10-13 | Harbin Engineering University | Marine Transportation Platform Guarantee-Oriented Analysis and Prediction Method for Three-Dimensional Temperature and Salinity Field |
| CN114721264A (en) * | 2022-03-21 | 2022-07-08 | 浙江工业大学 | Industrial information physical system attack detection method based on two-stage self-encoder |
| CN114779368A (en) * | 2022-03-22 | 2022-07-22 | 同济大学 | ENSO prediction method, device, equipment and medium based on multivariable sea air coupler |
| CN115393540A (en) * | 2022-08-30 | 2022-11-25 | 无锡九方科技有限公司 | Intelligent fusion method and system of three-dimensional marine environment field based on deep learning |
Non-Patent Citations (3)
| Title |
|---|
| LI YUHENG 等: "Reconstruction of sea surface temperature data from sea satellite observation based on convolutional automatic encoder", 《EARTH AND SPACE: FROM INFRARED TO TERAHERTZ (ESIT 2022)》, 31 January 2023 (2023-01-31), pages 2 - 3 * |
| 奚萌;宋清涛;李文君;邹斌;林明森;: "西北太平洋海表温度融合产品交叉比对分析", 海洋学报, no. 12, 15 December 2017 (2017-12-15) * |
| 肖贤俊;何娜;张祖强;刘怀明;王东晓;: "卫星遥感海表温度资料和高度计资料的变分同化", 热带海洋学报, no. 03, 15 May 2011 (2011-05-15) * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118097455A (en) * | 2024-04-29 | 2024-05-28 | 自然资源部第一海洋研究所 | Sea surface temperature data reconstruction method based on deep learning algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117633441B (en) | 2024-05-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112380098B (en) | Timing sequence abnormality detection method and device, computer equipment and storage medium | |
| CN111880998B (en) | Service system anomaly detection method and device, computer equipment and storage medium | |
| CN117633441B (en) | Sea surface temperature data filling and correcting method and device and electronic equipment | |
| CN107832837B (en) | Convolutional neural network compression method and decompression method based on compressed sensing principle | |
| US10147057B2 (en) | Systems and methods for flood zone modeling | |
| CN111818346A (en) | Image encoding method and apparatus, image decoding method and apparatus | |
| CN110892419B (en) | Stop code tolerant image compression neural network | |
| CN112991472B (en) | Image compressed sensing reconstruction method based on residual error dense threshold network | |
| Unnikrishnan et al. | Data-driven multi-time-step ahead daily rainfall forecasting using singular spectrum analysis-based data pre-processing | |
| US20160285476A1 (en) | Method for encoding and decoding of data based on binary reed-solomon codes | |
| CN114186583B (en) | Method and system for recovering abnormal signal of corrosion detection of tank wall of oil storage tank | |
| Cazals et al. | Reconstructing 3d compact sets | |
| CN113011501A (en) | Method and device for predicting typhoon water level based on graph convolution neural network | |
| Chen et al. | Inexact alternating direction method based on proximity projection operator for image inpainting in wavelet domain | |
| CN103401650A (en) | Blind identification method for (n, 1 and m) convolutional code with error codes | |
| CN116562120A (en) | RVE-based turbine engine system health condition assessment method and RVE-based turbine engine system health condition assessment device | |
| CN115840774A (en) | Multi-element time sequence abnormity detection method and device, computer equipment and storage medium | |
| Li et al. | An investigation on image compression using the trigonometric Bézier curve with a shape parameter | |
| Gamito et al. | Ray casting implicit fractal surfaces with reduced affine arithmetic | |
| CN117633712B (en) | Sea level height data fusion method, device and equipment based on multi-source data | |
| Gou et al. | Reduced iteration image reconstruction of incomplete projection CT using regularization strategy through Lp norm dictionary learning | |
| CN111565314A (en) | Image compression method, coding and decoding network training method and device and electronic equipment | |
| CN114548367B (en) | Reconstruction method and device of multimodal data based on countermeasure network | |
| Cardoso et al. | Spatio-Temporal Forecasting With Gridded Remote Sensing Data Using Feed-Backward Decoding | |
| CN117033352A (en) | Data restoration method and device, terminal equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |