CN111737913B - MWHTS clear sky observation bright temperature selection method based on cloud water content inversion - Google Patents
MWHTS clear sky observation bright temperature selection method based on cloud water content inversion Download PDFInfo
- Publication number
- CN111737913B CN111737913B CN202010544507.9A CN202010544507A CN111737913B CN 111737913 B CN111737913 B CN 111737913B CN 202010544507 A CN202010544507 A CN 202010544507A CN 111737913 B CN111737913 B CN 111737913B
- Authority
- CN
- China
- Prior art keywords
- mwhts
- clear sky
- water content
- bright temperature
- data set
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
Abstract
The MWHTS clear sky observation bright temperature selection method based on cloud water content inversion comprises the steps of establishing matching data of MWHTS observation bright temperature and a climatology data set in time and space, calculating corresponding MWHTS simulation bright temperature, and forming an analysis data set and a verification data set; taking cloud water content of 0 in an analysis data set as a strict clear sky threshold, selecting MWHTS (metal wrap through) strict clear sky to observe bright temperature, and adjusting the strict clear sky threshold by taking calculation accuracy of the MWHTS simulated bright temperature under the strict clear sky as a standard to form a corresponding clear sky threshold; based on the analysis data set and the BP neural network, establishing an optimal inversion model of MWHTS observation light Wen Fanyan cloud water content; and using the MWHTS observation light Wen Fanyan cloud water content in the verification data set, and selecting the MWHTS clear sky observation light temperature according to the cloud water content inversion value and the clear sky threshold value. The method can directly utilize the MWHTS observation bright temperature to realize effective selection of the MWHTS clear sky observation bright temperature, and has high accuracy and simple and easy operation.
Description
Technical Field
The invention belongs to the technical field of microwave remote sensing, and particularly relates to an MWHTS clear sky observation bright temperature selection method based on cloud water content inversion.
Background
The microwave humidity temperature detector (MWHTS) is an important load on a wind cloud third-size C star D star and is a microwave radiometer integrating a hygrometer and a thermometer at the first stage in the world, and the MWHTS is provided with eight temperature detection channels (channels 2-9) and five humidity detection channels (channels 11-15) and two window area channels (channels 1 and 10) so as to realize the simultaneous detection of atmospheric temperature, water vapor parameters and surface parameters.
Compared with clear air, radiation transmission of microwaves in cloud and rain air can be influenced by emission and scattering effects generated by cloud and rain, so that nonlinearity of a radiation transmission equation is increased, and calculation accuracy of a radiation transmission model is further reduced. At present, the simulation calculation precision of the satellite-borne microwave radiometer for observing the bright temperature in the clear air is far higher than that in the cloud and rain air, and the calculation precision of the satellite-borne microwave radiometer for simulating the bright temperature has an important influence on the data application of the microwave radiometer. The clear sky observation bright temperature is extracted from the satellite-borne microwave radiometer observation bright temperature, the inversion accuracy of inverting the atmospheric parameters is directly determined, and the method is closely related to the assimilation application effect of the satellite-borne microwave radiometer observation bright temperature. In addition, clear sky observation bright temperature has important theoretical guiding significance for the evaluation of the detection capability of the channel of the satellite-borne microwave radiometer and the development of hardware.
When the forward modeling problem of the satellite-borne microwave radiometer is processed, the setting of the atmospheric relative humidity threshold value and the utilization of the infrared cloud product are common methods for selecting clear sky observation bright temperature from the satellite-borne microwave radiometer observation bright Wen Zhongxuan, but before the clear sky data selection is carried out on satellite-borne microwave radiometer observation data by utilizing the two methods, matching with the atmospheric relative humidity or the infrared cloud product is completed, and the direct application of the data is seriously influenced, so that the two methods are not applicable to inversion of the microwave radiometer observation data. In inversion application of observation bright temperatures of a satellite-borne microwave radiometer, a threshold value can be formed according to a statistical relationship between the observation bright temperatures of a microwave radiometer channel to judge clear sky generally, but the accuracy of clear sky data judgment by the method is low, in addition, the establishment of the threshold value in the method has strong dependence on local atmosphere states, and is not suitable for selecting clear sky observation bright temperatures in a large range.
Disclosure of Invention
In order to solve the technical problems, according to the setting characteristics of a radiometer channel, the MWHTS has high sensitivity to water vapor parameters, the BP neural network can be directly used for inverting the atmospheric cloud water content, the cloud water content is directly related to the judgment of atmospheric clear sky, and further, the MWHTS clear sky bright temperature data can be extracted according to the cloud water content inversion value.
In order to achieve the technical purpose, the adopted technical scheme is as follows: MWHTS clear sky observation bright temperature selection method based on cloud water content inversion comprises the following steps:
step one: establishing matching data of MWHTS observation bright temperature and a climatic data set in time and space, inputting atmospheric parameters in the matching data into a radiation transmission model CRTM, calculating the MWHTS simulation bright temperature, establishing the matching data set, and forming an analysis data set and a verification data set;
step two: taking cloud water content of 0 as a strict clear sky threshold, selecting matching data with cloud water content of 0 as strict clear sky data in an analysis data set, calculating the calculation accuracy of the MWHTS simulated bright temperature, taking the calculation accuracy of the MWHTS simulated bright temperature in the strict clear sky data as a reference, and establishing a corresponding clear sky threshold by increasing the strict clear sky threshold in the analysis data set;
step three: taking MWHTS observation brightness temperature in an analysis data set as input, taking cloud water content in the analysis data set as output, training a BP neural network, and establishing an optimal inversion model of the MWHTS observation brightness Wen Fanyan cloud water content by adjusting the number of neurons of a hidden layer of the BP neural network and taking the mean square error of a cloud water content predicted value as a minimum value standard;
step four: inputting the MWHTS observation bright temperature in the verification data set into an optimal inversion model of the cloud water content of the MWHTS observation bright Wen Fanyan to obtain a cloud water content inversion value, and selecting the MWHTS clear sky observation bright temperature according to the clear sky threshold established in the second step.
The first step specifically comprises: first, the atmospheric parameters in the use climatology dataset are: the atmospheric parameters are matched with MWHTS observation bright temperature according to a matching rule that the longitude and latitude error is smaller than 0.1 DEG and the time error is smaller than 10 minutes, and matching data are obtained; then taking the atmospheric parameters in the matching data as input parameters of a radiation transmission model CRTM, calculating the simulated bright temperature of the MWHTS, and establishing a matching data set of the observed bright temperature of the MWHTS and the simulated bright temperature and the atmospheric parameters of the MWHTS; finally, 80% of the matching data in the matching data set is randomly selected to form an analysis data set, and the remaining 20% of the matching data form a verification data set.
The second step specifically comprises: firstly, taking cloud water content of 0 as a strict clear sky threshold, selecting matching data with cloud water content of 0 as strict clear sky data in an analysis data set formed in the first step, calculating root mean square error between MWHTS observation bright temperature and MWHTS simulated bright temperature in the strict clear sky data, and taking the root mean square error as calculation precision of the MWHTS simulated bright temperature of the strict clear sky data as the strict clear sky precision; then, a strict clear air threshold value is increased by a step length of 0.01mm, the clear air threshold value is used as an adjustment clear air threshold value, matching data smaller than the adjustment clear air threshold value is selected in an analysis data set to serve as adjustment clear air data, root mean square error between MWHTS observed bright temperature and MWHTS simulated bright temperature in the adjustment clear air data is calculated, and the calculation accuracy of the MWHTS simulated bright temperature serving as the adjustment clear air data is recorded as adjustment clear air accuracy; and finally, calculating a difference value between the adjustment clear sky precision and the strict clear sky precision in the MWHTS temperature detection channels 2-9 and the water vapor detection channels 11-15, and if the difference value of any one of the MWHTS temperature detection channels 2-9 and the water vapor detection channels 11-15 is increased to 1K, the adjustment clear sky threshold value at the moment is the clear sky threshold value.
The third step specifically comprises: firstly, taking MWHTS observation bright temperature in an analysis data set as input, taking cloud water content matched with the MWHTS observation bright temperature in the analysis data set as output, and establishing a training sample of the BP neural network; then, selecting BP neural network comprising an input layer, an output layer and a hidden layer, wherein the number of neurons of the hidden layer is increased from 5 to 50 according to step length 1, and training the BP neural network by using training samples to obtain 46 BP neural network models; and finally, comparing the mean square deviations of the 46 BP neural network models to the cloud water content predicted values in the training process, and selecting the BP neural network model corresponding to the minimum mean square deviation value as the optimal inversion model of the MWHTS observation brightness Wen Fanyan cloud water content.
The fourth step specifically comprises: firstly, inputting the MWHTS observation bright temperature in the verification data set into an optimal inversion model of the MWHTS observation bright Wen Fanyan cloud water content established in the step three, and obtaining a cloud water content inversion value; then comparing the clear sky threshold established in the second step with the cloud water content inversion value, and selecting data with the cloud water content inversion value smaller than the clear sky threshold as the clear sky cloud water content inversion value; and finally, according to the one-to-one correspondence between the clear sky cloud water content inversion value and the MWHTS observed bright temperature, acquiring the clear sky bright temperature data of the MWHTS.
The invention has the beneficial effects that: according to the method, a strict clear sky threshold is set according to the cloud water content value of 0, the calculation accuracy of the MWHTS simulated bright temperature under the strict clear sky condition is used as a standard, and the clear sky threshold is established by adjusting the strict clear sky threshold; and directly inverting the cloud water content by utilizing the MWHTS observation bright temperature, comparing the cloud water content inversion value with a clear sky threshold, and selecting the MWHTS observation bright temperature corresponding to the cloud water content inversion value smaller than the clear sky threshold, namely the MWHTS clear sky observation bright temperature. The MWHTS clear sky observation bright temperature selected by the method has higher accuracy and is simple and easy to operate.
Drawings
FIG. 1 is a flow chart of a method for selecting bright temperatures of MWHTS clear sky observations based on cloud water content inversion;
FIG. 2 is a graph of the calculation accuracy of the bright-sky data versus the MWHTS simulated bright temperature under all-weather conditions for example 1;
fig. 3 is a graph showing the difference between the accuracy of clear sky adjustment and the accuracy of clear sky adjustment when the clear sky threshold is 0.11mm in example 1.
Detailed Description
The MWHTS clear sky observation bright temperature selection method based on cloud water content inversion is characterized by comprising the following steps of:
step one: establishing matching data of MWHTS observation bright temperature and a climatic data set in time and space, inputting atmospheric parameters in the matching data into a radiation transmission model CRTM, calculating the MWHTS simulation bright temperature, establishing the matching data set, and forming an analysis data set and a verification data set;
the first step specifically comprises:
first, the atmospheric parameters in the use climatology dataset are: the atmospheric parameters are matched with MWHTS observation bright temperature according to a matching rule that the longitude and latitude error is smaller than 0.1 DEG and the time error is smaller than 10 minutes, and matching data are obtained; then taking the atmospheric parameters in the matching data as input parameters of a radiation transmission model CRTM, calculating the simulated bright temperature of the MWHTS, and establishing a matching data set of the observed bright temperature of the MWHTS and the simulated bright temperature and the atmospheric parameters of the MWHTS; finally, 80% of the matching data in the matching data set is randomly selected to form an analysis data set, and the remaining 20% of the matching data form a verification data set.
Step two: taking cloud water content of 0 as a strict clear sky threshold, selecting matching data with cloud water content of 0 as strict clear sky data in an analysis data set, calculating the calculation accuracy of the MWHTS simulated bright temperature, taking the calculation accuracy of the MWHTS simulated bright temperature in the strict clear sky data as a reference, and establishing a corresponding clear sky threshold by increasing the strict clear sky threshold in the analysis data set;
the second step specifically comprises: firstly, taking cloud water content of 0 as a strict clear sky threshold, selecting matching data with cloud water content of 0 as strict clear sky data in an analysis data set formed in the first step, calculating root mean square error between MWHTS observation bright temperature and MWHTS simulated bright temperature in the strict clear sky data, and taking the root mean square error as calculation precision of the MWHTS simulated bright temperature of the strict clear sky data as the strict clear sky precision; then, a strict clear air threshold value is increased by a step length of 0.01mm, the clear air threshold value is used as an adjustment clear air threshold value, matching data smaller than the adjustment clear air threshold value is selected in an analysis data set to serve as adjustment clear air data, root mean square error between MWHTS observed bright temperature and MWHTS simulated bright temperature in the adjustment clear air data is calculated, and the calculation accuracy of the MWHTS simulated bright temperature serving as the adjustment clear air data is recorded as adjustment clear air accuracy; and finally, calculating a difference value between the adjustment clear sky precision and the strict clear sky precision in the MWHTS temperature detection channels 2-9 and the water vapor detection channels 11-15, and if the difference value of any one of the MWHTS temperature detection channels 2-9 and the water vapor detection channels 11-15 is increased to 1K, the adjustment clear sky threshold value at the moment is the clear sky threshold value.
Step three: taking MWHTS observation brightness temperature in an analysis data set as input, taking cloud water content in the analysis data set as output, training a BP neural network, and establishing an optimal inversion model of the MWHTS observation brightness Wen Fanyan cloud water content by adjusting the number of neurons of a hidden layer of the BP neural network and taking the mean square error of a cloud water content predicted value as a minimum value standard;
the third step specifically comprises: firstly, taking MWHTS observation bright temperature in an analysis data set as input, taking cloud water content matched with the MWHTS observation bright temperature in the analysis data set as output, and establishing a training sample of the BP neural network; then, selecting BP neural network comprising an input layer, an output layer and a hidden layer, wherein the number of neurons of the hidden layer is increased from 5 to 50 according to step length 1, and training the BP neural network by using training samples to obtain 46 BP neural network models; and finally, comparing the mean square deviations of the 46 BP neural network models to the cloud water content predicted values in the training process, and selecting the BP neural network model corresponding to the minimum mean square deviation value as the optimal inversion model of the MWHTS observation brightness Wen Fanyan cloud water content.
Step four: inputting the MWHTS observation bright temperature in the verification data set into an optimal inversion model of the cloud water content of the MWHTS observation bright Wen Fanyan to obtain a cloud water content inversion value, and selecting the MWHTS clear sky observation bright temperature according to the clear sky threshold established in the second step.
The fourth step specifically comprises: firstly, inputting the MWHTS observation bright temperature in the verification data set into an optimal inversion model of the MWHTS observation bright Wen Fanyan cloud water content established in the step three, and obtaining a cloud water content inversion value; then comparing the clear sky threshold established in the second step with the cloud water content inversion value, and selecting data with the cloud water content inversion value smaller than the clear sky threshold as the clear sky cloud water content inversion value; and finally, according to the one-to-one correspondence between the clear sky cloud water content inversion value and the MWHTS observed bright temperature, acquiring the clear sky bright temperature data of the MWHTS.
The invention will be further described with reference to examples and drawings, to which it should be noted that the examples do not limit the scope of the invention as claimed.
Example 1
The climatic data set is selected by using an analysis data set ERA-Interim of an European mid-term weather forecast center (ECMWF), and selecting a D star MWHTS observation bright temperature of the Fengyun No. three to select a clear sky observation bright temperature, wherein the time range is from 9 months in 2018 to 8 months in 2019, and the geographic range is (25-45 DEG N, 160-220 DEG E). And matching the MWHTS observation bright temperature with the temperature profile, the humidity profile, the cloud water profile, the 2m temperature, the 2m humidity, the surface pressure and the 10m wind speed in the ERA interm data set according to a matching rule that the time error is less than 10 minutes and the longitude and latitude error is less than 0.1 degree, inputting the atmospheric parameters in the matched ERA interm data set into a radiation transmission model CRTM, calculating the MWHTS simulation bright temperature, obtaining a matching data set (1060162 groups) of the MWHTS observation bright temperature, the MWHTS simulation bright temperature and the atmospheric parameters, randomly selecting 80% of the matching data in the matching data set to establish an analysis data set (848129 groups), and establishing a verification data set (212033 groups) of the remaining 20% of the matching data.
Taking the cloud water content of 0 as a strict clear sky threshold, selecting the strict clear sky data in the analysis data set to obtain 5830 sets of matching data, and calculating root mean square error between MWHTS observed bright temperature and MWHTS simulated bright temperature in the strict clear sky data to obtain strict clear sky precision; all MWHTS observation brightness and MWHTS simulation brightness temperature direct root mean square errors in the analysis data set are calculated, and all-weather precision can be obtained. The calculation accuracy of the MWHTS simulated bright temperature is shown in fig. 2 under the strict clear sky condition and the all-weather condition, namely, the strict clear sky accuracy and the all-weather accuracy pair. As can be seen from fig. 2, in fifteen channels of MWHTS, the precision of clear sky is higher than the precision of all weather.
Gradually increasing a strict clear sky threshold by 0.01mm step length to form an adjustment clear sky threshold, selecting matching data smaller than the adjustment clear sky threshold from cloud water content in analysis data, calculating root mean square error of MWHTS observed bright temperature and MWHTS simulated bright temperature, and recording as adjustment clear sky precision; when the difference between the adjustment clear sky precision and the strict clear sky precision of any one of the MWHTS temperature detection channels (channels 2-9) and the water vapor detection channels (channels 11-15) is increased to 1K, the adjustment clear sky threshold at the moment is the clear sky threshold. When the clear sky threshold is adjusted to be 0.11mm, the difference between the clear sky precision and the strict clear sky precision is adjusted as shown in fig. 3. Wherein the difference in accuracy in the channel 8 is 1K, thus establishing a clear sky threshold of 0.11mm. According to the clear sky threshold, 306001 sets of clear sky data can be selected in the analysis dataset.
The method comprises the steps of taking a BP neural network as an algorithm model for inverting cloud water content by MWHTS, wherein the structure of the BP neural network is set to be 3 layers (an input layer, an output layer and a hidden layer), taking 848129 groups of MWHTS observation brightness temperatures in an analysis data set as input of the algorithm model, taking corresponding cloud water content as output of the algorithm model, changing the number of neurons of the hidden layer from 5 to 50, and respectively training the neural network. And comparing the obtained 46 trained neural network models, wherein the mean square difference value of the obtained cloud water content predicted value is 0.0091 when the number of hidden layer neurons is 43, and the neural network model when the number of hidden layer neurons is 43 is recorded as the optimal inversion model of the MWHTS observation brightness Wen Fanyan cloud water content.
The bright temperature of the MWHTS observation of 212033 groups in the verification data set is input into an optimal inversion model of the cloud water content of the MWHTS observation bright Wen Fanyan, and 212033 groups of cloud water content inversion values are obtained. And (3) taking the cloud water content in the verification data set as a true value, and obtaining that the average deviation between the true value of the cloud water content and the inversion value of the cloud water content is 0.0022mm. Comparing the clear sky threshold value of 0.11mm with the cloud water content inversion value in the cloud water content inversion value, selecting an MWHTS observation bright temperature corresponding to the cloud water content inversion value smaller than the clear sky threshold value as 76332 groups, and recording the MWHTS observation bright temperature as MWHTS clear sky observation bright temperature selection data; and comparing the clear sky threshold value of 0.11mm with the cloud water content real value in the verification analysis data set, selecting an MWHTS (metal wrap through test) observation bright temperature corresponding to the cloud water content real value smaller than the clear sky threshold value as 79263 groups, and recording the MWHTS observation bright temperature real data as the MWHTS clear sky observation bright temperature real data. The same data in the MWHTS clear sky observation bright temperature selection data and the MWHTS clear sky observation bright temperature real data are 73590 groups, in other words, the MWHTS clear sky bright temperature data selection method based on cloud water content inversion can select 73590 groups of data in the MWHTS clear sky observation bright temperature real data 79263 groups, namely the method can select 92.84% of MWHTS clear sky observation bright temperatures in the MWHTS observation bright temperatures.
Claims (2)
1. The MWHTS clear sky observation bright temperature selection method based on cloud water content inversion is characterized by comprising the following steps of:
step one: establishing matching data of MWHTS observation bright temperature and a climatic data set in time and space, inputting atmospheric parameters in the matching data into a radiation transmission model CRTM, calculating the MWHTS simulation bright temperature, establishing the matching data set, and forming an analysis data set and a verification data set;
the first step specifically comprises the following steps:
first, the atmospheric parameters in the use climatology dataset are: the atmospheric parameters are matched with MWHTS observation bright temperature according to a matching rule that the longitude and latitude error is smaller than 0.1 DEG and the time error is smaller than 10 minutes, and matching data are obtained; then taking the atmospheric parameters in the matching data as input parameters of a radiation transmission model CRTM, calculating the simulated bright temperature of the MWHTS, and establishing a matching data set of the observed bright temperature of the MWHTS and the simulated bright temperature and the atmospheric parameters of the MWHTS; finally, randomly selecting 80% of the matching data in the matching data set to form an analysis data set, and forming a verification data set by the remaining 20% of the matching data;
step two: taking cloud water content of 0 as a strict clear sky threshold, selecting matching data with cloud water content of 0 as strict clear sky data in an analysis data set, calculating the calculation accuracy of the MWHTS simulated bright temperature, taking the calculation accuracy of the MWHTS simulated bright temperature in the strict clear sky data as a reference, and establishing a corresponding clear sky threshold by increasing the strict clear sky threshold in the analysis data set;
the second step specifically comprises the following steps:
firstly, taking cloud water content of 0 as a strict clear sky threshold, selecting matching data with cloud water content of 0 as strict clear sky data in an analysis data set formed in the first step, calculating root mean square error between MWHTS observation bright temperature and MWHTS simulated bright temperature in the strict clear sky data, and taking the root mean square error as calculation precision of the MWHTS simulated bright temperature of the strict clear sky data as the strict clear sky precision; then, a strict clear air threshold value is increased by a step length of 0.01mm, the clear air threshold value is used as an adjustment clear air threshold value, matching data smaller than the adjustment clear air threshold value is selected in an analysis data set to serve as adjustment clear air data, root mean square error between MWHTS observed bright temperature and MWHTS simulated bright temperature in the adjustment clear air data is calculated, and the calculation accuracy of the MWHTS simulated bright temperature serving as the adjustment clear air data is recorded as adjustment clear air accuracy; finally, the clear sky precision and the strict clear sky precision are adjusted in the MWHTS temperature detection channels 2-9 and the water vapor detection channels 11-15 to obtain a difference value, and if the difference value of any one of the MWHTS temperature detection channels 2-9 and the water vapor detection channels 11-15 is increased to 1K, the clear sky threshold is adjusted to be the clear sky threshold at the moment;
step three: taking MWHTS observation brightness temperature in an analysis data set as input, taking cloud water content in the analysis data set as output, training a BP neural network, and establishing an optimal inversion model of the MWHTS observation brightness Wen Fanyan cloud water content by adjusting the number of neurons of a hidden layer of the BP neural network and taking the mean square error of a cloud water content predicted value as a minimum value standard;
the third step specifically comprises the following steps:
firstly, taking MWHTS observation bright temperature in an analysis data set as input, taking cloud water content matched with the MWHTS observation bright temperature in the analysis data set as output, and establishing a training sample of the BP neural network; then, selecting BP neural network comprising an input layer, an output layer and a hidden layer, wherein the number of neurons of the hidden layer is increased from 5 to 50 according to step length 1, and training the BP neural network by using training samples to obtain 46 BP neural network models; finally, comparing the mean square error of the 46 BP neural network models to the cloud water content predicted value in the training process, and selecting the BP neural network model corresponding to the minimum mean square error value as the optimal inversion model of MWHTS observation brightness Wen Fanyan cloud water content;
step four: inputting the MWHTS observation bright temperature in the verification data set into an optimal inversion model of the cloud water content of the MWHTS observation bright Wen Fanyan to obtain a cloud water content inversion value, and selecting the MWHTS clear sky observation bright temperature according to the clear sky threshold established in the second step.
2. The MWHTS clear sky observation bright temperature selection method based on cloud water content inversion of claim 1, characterized by comprising the following steps:
firstly, inputting the MWHTS observation bright temperature in the verification data set into an optimal inversion model of the MWHTS observation bright Wen Fanyan cloud water content established in the step three, and obtaining a cloud water content inversion value; then comparing the clear sky threshold established in the second step with the cloud water content inversion value, and selecting data with the cloud water content inversion value smaller than the clear sky threshold as the clear sky cloud water content inversion value; and finally, according to the one-to-one correspondence between the clear sky cloud water content inversion value and the MWHTS observed bright temperature, acquiring the clear sky bright temperature data of the MWHTS.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010544507.9A CN111737913B (en) | 2020-06-15 | 2020-06-15 | MWHTS clear sky observation bright temperature selection method based on cloud water content inversion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010544507.9A CN111737913B (en) | 2020-06-15 | 2020-06-15 | MWHTS clear sky observation bright temperature selection method based on cloud water content inversion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111737913A CN111737913A (en) | 2020-10-02 |
| CN111737913B true CN111737913B (en) | 2023-08-15 |
Family
ID=72649242
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010544507.9A Active CN111737913B (en) | 2020-06-15 | 2020-06-15 | MWHTS clear sky observation bright temperature selection method based on cloud water content inversion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111737913B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112345151B (en) * | 2020-10-15 | 2022-01-07 | 洛阳师范学院 | Sensitivity test method of MWTS-II to sea surface air pressure based on natural atmosphere |
| CN112329334B (en) * | 2020-10-15 | 2024-01-23 | 洛阳师范学院 | MWHTS and MWTS-II fusion inversion sea surface air pressure method based on simulated bright temperature |
| CN112730313B (en) * | 2020-12-21 | 2023-05-02 | 国家卫星气象中心(国家空间天气监测预警中心) | Multi-frequency terahertz detector channel selection method and device for ice cloud detection |
| CN113311510B (en) * | 2021-05-11 | 2023-07-25 | 洛阳师范学院 | MWHTS (metal wrap through) observation bright temperature classification method based on simulated bright temperature |
| CN113340836A (en) * | 2021-05-18 | 2021-09-03 | 国家卫星气象中心(国家空间天气监测预警中心) | Atmospheric temperature and humidity profile inversion method for high-latitude complex underlying surface |
| CN113822345B (en) * | 2021-09-09 | 2024-03-29 | 南京中科逆熵科技有限公司 | Method and system for inverting cloud water content by fusing MWHTS and MWTS-II |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010032383A (en) * | 2008-07-29 | 2010-02-12 | Fujitsu Ltd | Apparatus and method for forecasting, and program |
| CN103901420A (en) * | 2014-04-18 | 2014-07-02 | 山东科技大学 | Method for dynamic threshold method remote sensing data cloud identification supported by prior surface reflectance |
| CN108051872A (en) * | 2017-12-13 | 2018-05-18 | 湖北省气象服务中心（湖北省专业气象服务台） | Method and apparatus based on steam phase transition process in Ground-Based Microwave Radiometer Retrieval of Cloud |
| CN108508442A (en) * | 2018-03-16 | 2018-09-07 | 哈尔滨工程大学 | A kind of Atmosphere and humidity profiles inversion method based on ground multi-channel microwave radiometer |
| CN109782374A (en) * | 2019-01-15 | 2019-05-21 | 北京华云星地通科技有限公司 | Moisture content by assimilating inverting optimizes the method and device of numerical weather forecast |
| CN110969609A (en) * | 2019-11-29 | 2020-04-07 | 华中科技大学 | Microwave remote sensing bright temperature image cloud detection method |
| CN111060992A (en) * | 2019-12-27 | 2020-04-24 | 国家卫星气象中心(国家空间天气监测预警中心) | Equal-weight conjugate precipitation detection method and system for satellite-borne microwave dioxygen detection channel |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106909722B (en) * | 2017-02-10 | 2019-07-26 | 广西壮族自治区气象减灾研究所 | A kind of accurate inversion method of large area of temperature near the ground |
-
2020
- 2020-06-15 CN CN202010544507.9A patent/CN111737913B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010032383A (en) * | 2008-07-29 | 2010-02-12 | Fujitsu Ltd | Apparatus and method for forecasting, and program |
| CN103901420A (en) * | 2014-04-18 | 2014-07-02 | 山东科技大学 | Method for dynamic threshold method remote sensing data cloud identification supported by prior surface reflectance |
| CN108051872A (en) * | 2017-12-13 | 2018-05-18 | 湖北省气象服务中心（湖北省专业气象服务台） | Method and apparatus based on steam phase transition process in Ground-Based Microwave Radiometer Retrieval of Cloud |
| CN108508442A (en) * | 2018-03-16 | 2018-09-07 | 哈尔滨工程大学 | A kind of Atmosphere and humidity profiles inversion method based on ground multi-channel microwave radiometer |
| CN109782374A (en) * | 2019-01-15 | 2019-05-21 | 北京华云星地通科技有限公司 | Moisture content by assimilating inverting optimizes the method and device of numerical weather forecast |
| CN110969609A (en) * | 2019-11-29 | 2020-04-07 | 华中科技大学 | Microwave remote sensing bright temperature image cloud detection method |
| CN111060992A (en) * | 2019-12-27 | 2020-04-24 | 国家卫星气象中心(国家空间天气监测预警中心) | Equal-weight conjugate precipitation detection method and system for satellite-borne microwave dioxygen detection channel |
Non-Patent Citations (1)
| Title |
|---|
| 基于FY-3C/MWHTS观测资料反演中低纬度海面气压;张子瑾;空间科学学报;第40卷(第3期);364-375 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111737913A (en) | 2020-10-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111737913B (en) | MWHTS clear sky observation bright temperature selection method based on cloud water content inversion | |
| CN111737912B (en) | MWHTS (metal wrap through) simulated bright temperature calculation method based on deep neural network | |
| CN108875254A (en) | A kind of One-Dimensional Variational inversion method of Atmosphere and humidity profiles | |
| CN111060992B (en) | Equal-weight conjugate precipitation detection method and system for satellite-borne microwave dioxygen detection channel | |
| CN110174359B (en) | Aviation hyperspectral image soil heavy metal concentration assessment method based on Gaussian process regression | |
| CN101634711A (en) | Method for estimating temperature of near-surface air from MODIS data | |
| CN110749942B (en) | Clear sky channel detection quality control method suitable for data assimilation of stationary satellite | |
| CN110826693A (en) | Three-dimensional atmospheric temperature profile inversion method and system based on DenseNet convolutional neural network | |
| CN113108918A (en) | Method for inverting air temperature by using thermal infrared remote sensing data of polar-orbit meteorological satellite | |
| CN116449331B (en) | Dust particle number concentration estimation method based on W-band radar and meteorological satellite | |
| CN111737641B (en) | MWHTS channel weight function calculation method based on neural network | |
| CN112560270A (en) | Chemical hazard assimilation system | |
| CN113552080B (en) | Real-time inversion algorithm for ultra-spectrum remote sensing earth surface atmosphere relative humidity | |
| CN112329334B (en) | MWHTS and MWTS-II fusion inversion sea surface air pressure method based on simulated bright temperature | |
| CN113935228A (en) | L-band rough sea surface radiation brightness and temperature simulation method based on machine learning | |
| CN111400658B (en) | Channel selection method for inversion gas profile of hyperspectral thermal infrared data | |
| CN113311509B (en) | Sensitivity test method of MWHTS (metal wrap through) to sea surface air pressure based on neural network | |
| CN110968926A (en) | Method for predicting atmospheric parameters based on improved background error covariance matrix | |
| CN116297288A (en) | Rapid remote sensing inversion method and system for atmospheric methane dry air mixing ratio | |
| CN109543911B (en) | Sunlight radiation prediction method and system | |
| CN114755189B (en) | Feature-optimized self-attention-based hyperspectral satellite LAI inversion method | |
| CN113108949B (en) | Model fusion-based sonde temperature sensor error prediction method | |
| CN114880933A (en) | Atmospheric temperature and humidity profile inversion method and system for non-exploration-site foundation microwave radiometer based on reanalysis data | |
| CN114970663A (en) | Near-shore sea surface temperature inversion method of microwave radiometer based on neural network | |
| CN112254866A (en) | Method for inverting sea surface air pressure by fusion of MWTS-II and MWHTS |
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 |