CN116189796A - Machine learning-based satellite-borne short wave infrared CO 2 Column concentration estimation method - Google Patents
Machine learning-based satellite-borne short wave infrared CO 2 Column concentration estimation method Download PDFInfo
- Publication number
- CN116189796A CN116189796A CN202211594763.4A CN202211594763A CN116189796A CN 116189796 A CN116189796 A CN 116189796A CN 202211594763 A CN202211594763 A CN 202211594763A CN 116189796 A CN116189796 A CN 116189796A
- Authority
- CN
- China
- Prior art keywords
- data
- column concentration
- satellite
- machine learning
- value
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000010801 machine learning Methods 0.000 title claims abstract description 17
- 238000007637 random forest analysis Methods 0.000 claims abstract description 18
- 238000012216 screening Methods 0.000 claims abstract description 8
- 239000000443 aerosol Substances 0.000 claims abstract description 6
- 238000002310 reflectometry Methods 0.000 claims abstract description 5
- 238000012876 topography Methods 0.000 claims abstract description 5
- 238000010219 correlation analysis Methods 0.000 claims abstract description 4
- 238000003066 decision tree Methods 0.000 claims description 8
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 claims description 6
- 230000002159 abnormal effect Effects 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 5
- 238000012952 Resampling Methods 0.000 claims description 4
- 229910002092 carbon dioxide Inorganic materials 0.000 claims description 3
- 239000001569 carbon dioxide Substances 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 230000010354 integration Effects 0.000 claims description 3
- 238000010206 sensitivity analysis Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 5
- 238000010521 absorption reaction Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000003203 everyday effect Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000005855 radiation Effects 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Chemical compound O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 229910052799 carbon Inorganic materials 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000005431 greenhouse gas Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 238000010792 warming Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/30—Prediction of properties of chemical compounds, compositions or mixtures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- 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
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Computational Biology (AREA)
- Medical Informatics (AREA)
- Crystallography & Structural Chemistry (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a machine learning-based satellite-borne short wave infrared CO 2 A method of estimating column concentration comprising: s1, extracting OCO-2 satellite wave band data to obtain 9 weakCO 2 Band and 6O 2 Band data; s2, 9 weakCO 2 Band and 6O 2 Performing feature screening on the wave band data, the NDVI normalized vegetation index, the SR surface reflectivity data, the DEM elevation topography data, the ERA5 atmosphere data, the AOD aerosol data and the TCCON station observation data, and reserving the first 31 features of screening according to importance; s3, performing correlation analysis on the first 31 screened features through a heat map to find out the correlation with CO 2 Features of stronger correlation and weaker features of column concentration; s4, will be combined with CO 2 Combining the characteristic with strong correlation with weak characteristic, inputting to five regression models with integrated learning, and predicting accuracy by using the different modelsComparative analysis, determining coefficient R of extreme random forest regression model 2 Highest, least error, best prediction effect.
Description
Technical Field
The invention relates to the technical field of atmospheric satellite remote sensing prediction, in particular to a satellite-borne short wave infrared CO based on machine learning 2 Column concentration estimation method.
Background
CO 2 Is the main greenhouse gas in the atmosphere and has very important influence on global climate change. Since the industrial age, CO 2 The concentration has increased to about 30% and there is a continuing trend. So CO is monitored 2 Has important significance for the research of global climate warming; thus accurately grasping CO in the atmosphere 2 The content and the change thereof can provide support for climate prediction and environmental decision.
Conventional ground atmosphere CO 2 Although the detection method has the advantages of high precision and high reliability, the detection method is single-point measurement and lacks the capability of detecting the region and the global large scale in real time, so the satellite observation CO is developed 2 Is imperative. Short wave infrared band near ground CO 2 And the system is more sensitive and is more suitable for monitoring dynamic changes of ground carbon source sinks.
Currently international shortwave infrared CO 2 The observation data adopts an all-physical inversion algorithm, the whole optical path needs to be simulated, and the radiation transmission equation is complex in calculation and time-consuming. Because aerosol, water vapor and earth surface reflectivity have complex influence on the short-wave infrared radiation process, the existing physical inversion model needs more input parameters and has uncertainty.
Disclosure of Invention
The invention aims at solving the problem that the method is used for solving the problem of CO by using decision trees, XGBoost, common random forests, extreme random forests and gradient lifting regression models 2 Estimating column concentration, comparing and analyzing to find out the model with highest estimation accuracy, and performing remote sensing on the atmospheric CO of the satellite 2 The method has the advantages of high prediction accuracy and strong interpretability, and greatly improves the prediction efficiency.
In order to achieve the above purpose, the technical scheme of the application is as follows: machine learning-based satellite-borne short wave infrared CO 2 A method of estimating column concentration comprising:
s1, acquiring OCO-2 satellite wave band data, and extracting the OCO-2 satellite wave band data through sensitivity analysis of atmospheric carbon dioxide inversion parameters to obtain 9 weak_CO 2 Band and 6O 2 Band data;
s2, 9 weakCO 2 Band and 6O 2 Band data, NDVI normalizationThe vegetation index, SR surface reflectivity data, DEM elevation topography data, ERA5 atmospheric data, AOD aerosol data and TCCON station observation data are subjected to feature screening, and the first 31 screened features are reserved according to importance;
s3, performing correlation analysis on the first 31 screened features through a heat map to find out the correlation with CO 2 Features of stronger correlation and weaker features of column concentration;
s4, will be combined with CO 2 Combining the characteristic with strong column concentration correlation and the characteristic with weak column concentration correlation as an input characteristic data set, and then adopting decision trees, XGBoost, common random forests, extreme random forests and gradient lifting regression models to carry out CO respectively 2 Average column concentration is estimated by determining coefficient R estimated by different regression models 2 Performing contrast analysis on the accuracy of prediction in the error allowable range, namely, the root mean square error RMSE, the average absolute error MAE, the average relative error MRE and the average relative error MRE, finding out the model with the highest prediction accuracy as an extreme random forest regression model, and using the extreme random forest regression model to carry out CO 2 The average concentration of the columns was predicted.
Further, the OCO-2 satellite wave band data comprises longitude lon, dimension lat, zenith angle and azimuth angle of the sun, zenith angle and azimuth angle of the satellite; the ERA5 atmospheric data includes temperature, humidity, pressure, U/V component of wind, rainfall, boundary layer height (blh), cloud floor height (cbh), cloud cover (tcc), total rainfall (tp), vertical velocity of wind.
Further, before extracting the OCO-2 satellite band data, determining an extraction range by adopting a resampling mode, namely drawing grids according to the longitude and latitude range of a target area, setting the resolution after sampling to be 0.5 degrees multiplied by 0.5 degrees, and obtaining the Euclidean distance between the central point of each grid and the central point of each pixel corresponding to the original image by the longitude and latitude of each grid, wherein:
in lon k Longitude, lat for fixed site k Being fixed-siteLatitude, lon i 、lat i The longitude and latitude of the grid respectively.
Further, for 9 weakCO 2 Band and 6O 2 The abnormal value in the band data is processed as follows:
wherein sigma is the standard deviation of the data of the same day, namely, all abnormal values except + -3 sigma are removed, and the average value of the data of each wave band measured for each site for a plurality of times every day is obtained.
Further, the decision tree uses the base index to divide the attributes, assuming that the k-th sample in the current sample set X occupies a proportion p k (k=1, 2,3, …, y), then the base value is:
gini (X) indicates the possibility of random sampling of inconsistencies between two different types of tags; the genie purity refers to the probability that the sample is selected multiplied by the probability of error. The smaller Gini (X), the higher the purity of sample set X; when all samples in a node are of one class, the keni is not pure 0.
Assuming that the discrete attribute a has v possible values, if a is used to classify the sample set X, v branch nodes are generated and X is recorded v The v branch node comprises all samples which take values on the attribute a in the sample set X; the base index of attribute a is defined as:
the Gini index Gini (X, a) represents the uncertainty of the sample set X after a=a segmentation; the larger the base index, the greater the uncertainty of the sample.
Further, in XGBoost, K trees are assumed in total, and F tableTree model, predicted value
Expressed as:
in which x is i As an input example, a feature vector representing an i-th data point; k is the number of CART trees; f (f) k To represent the kth CART tree;
the corresponding objective function L is:
wherein, l is a loss function and represents the error between the predicted value and the true value; y is i Is a true value; omega is a regularization function that prevents model overfitting.
Further, in the common random forest, a model h (X, theta) is established for the characteristic parameter set X of the data set i ) I=1, 2, …, k, randomly selecting m features, so that each leaf node selects the feature of the maximum information gain for splitting; wherein the information gain is expressed as:
wherein i is a regression value, p i Represents the probability of occurrence of the corresponding value, w is the number of dividing nodes,
the weight value of the leaf node is divided for the mth.
Further, in extreme random forests, generalized errors of individual learners are assumedThe difference is E i The generalized error weight of the learner is:
assume that the divergence value of the individual learner is A i The learner's weighted divergence value is:
the generalization error after integration is expressed as:
w in i And T is the total number of decision trees with different structures for the weight.
Furthermore, the new learner obtained by each iteration of gradient lifting is fitted to the residual error of the previous learner, and finally, predictions of all trees are added, so that a prediction task is completed; the residual error obtaining mode is as follows:
r ni ＝y i -f n-1 (x i )
wherein y is i For the measured value of the ith sample, f n-1 (x i ) The predicted value of the former round of learner; fitting the residual remembering to obtain a fitted residual model h n (x) Updating the regression tree:
f n (x)＝f n-1 (x)+h n (x)
further, the decision coefficient R 2 The obtaining modes of the root mean square error RMSE, the average absolute error MAE and the average relative error MRE are as follows:
wherein: n is the number of samples; f (f) i Is a predicted value; y is i Is a true value;
is the average value.
By adopting the technical scheme, the invention can obtain the following technical effects: the method uses different integrated learning methods to realize the CO of the atmosphere through satellite, vegetation, ground topography, atmosphere, aerosol data and the like 2 The prediction of the average concentration of the column is of practical significance. Compared with the traditional physical inversion method, the method has the advantages of sufficient considered characteristics, easy modification, easy explanation and simple operation, and greatly improves the prediction efficiency. Can better predict CO 2 Column concentration, so that the integrated learning model predicts the CO of different satellites 2 The column concentration result can be more accurate, and data support is provided for decisions of environmental protection departments.
Drawings
FIG. 1 is an OCO-2 satellite observation spectrum, observation samples from Tsukuba site, (36.05N DEG, 140.12E DEG), 1 month and 1 day 2019; wherein (a) is CO 2 (b) is the O2-A absorption band;
FIG. 2 is a bar graph of the first 31 feature importance;
FIG. 3 is CO 2 Correlation graphs among all influence factors in column average concentration satellite inversion;
FIG. 4 is a training result diagram of five predictive models;
FIG. 5 is a graph of CO prediction for five prediction model test sets 2 Differences between column concentration and true valuesA value map;
FIG. 6 is a graph showing the influence of the prediction performance of the extreme random forest regression model on the parameters of the model;
FIG. 7 is a machine learning based satellite borne short wave infrared CO 2 A flow chart of a column concentration estimation method.
Detailed Description
The embodiment of the invention is implemented on the premise of the technical scheme of the invention, and a detailed implementation mode and a specific operation process are provided, but the protection scope of the invention is not limited to the following embodiment.
Example 1
The embodiment provides a machine learning-based satellite-borne short wave infrared CO 2 A method of estimating column concentration comprising:
s1, acquiring OCO-2 satellite wave band data, and extracting the OCO-2 satellite wave band data through sensitivity analysis of atmospheric carbon dioxide inversion parameters; due to CO 2 The weak absorption band is greatly affected by water vapor, so that 9 weak_CO are obtained by selecting corresponding absorption channels at the strong absorption band (1.61 μm) 2 Band and 6O 2 Band data;
the OCO-2 satellite wave band data comprises longitude lon, latitude lat, zenith angle and azimuth angle of the sun, and zenith angle and azimuth angle of the satellite.
It should be noted that: before OCO-2 satellite wave band data are extracted, a resampling mode is adopted to determine an extraction range, namely grids are drawn according to the longitude and latitude range of a target area, the resolution after sampling is set to be 0.5 degrees multiplied by 0.5 degrees, and the Euclidean distance between the center point of each grid and the center point of each pixel corresponding to an original image is obtained through the longitude and latitude of each grid:
in lon k Longitude, lat for fixed site k Is the latitude and lon of the fixed site i 、lat i The longitude and latitude of the grid respectively.
PreferablyFor 9 weakCO 2 Band and 6O 2 The abnormal value in the band data is processed as follows:
wherein sigma is the standard deviation of the data of the same day, namely, all abnormal values except + -3 sigma are removed, and the average value of the data of each wave band measured for each site for a plurality of times every day is obtained.
S2, 9 weakCO 2 Band and 6O 2 Performing feature screening on the wave band data, the NDVI normalized vegetation index, the SR surface reflectivity data, the DEM elevation topography data, the ERA5 atmosphere data, the AOD aerosol data and the TCCON station observation data, and reserving the first 31 features of screening according to importance;
specifically, feature screening is performed by the resampling method.
S3, performing correlation analysis on the first 31 screened features through a heat map to find out the correlation with CO 2 Features of stronger correlation and weaker features of column concentration;
s4, will be combined with CO 2 Combining the characteristic with strong correlation with weak characteristic, inputting to five regression models of integrated learning, and respectively outputting predicted CO 2 Column concentration;
specifically, in the extreme random forest, it is assumed that the generalization error of the individual learner is E i The generalized error weight of the learner is:
assume that the divergence value of the individual learner is A i The learner's weighted divergence value is:
the generalization error after integration can be expressed as:
w in i And T is the total number of decision trees with different structures for the weight.
The accuracy of prediction by using the different models is determined by comparing and analyzing the determination coefficient R of the extremely random forest regression model 2 The method has the advantages of highest prediction effect, minimum error and best prediction effect, and is obviously superior to the prediction results of other models; the four evaluation index related data are shown in the following table:
table 1 four evaluation index related data
The embodiments of the present invention are preferred embodiments and are not intended to be limiting in any way. The technical features or combinations of technical features described in the embodiments of the present invention should not be regarded as isolated, and they may be combined with each other to achieve a better technical effect. Additional implementations are also included within the scope of the preferred embodiments of the present invention and should be understood by those skilled in the art to which the inventive embodiments pertain.
Claims (10)
1. Machine learning-based satellite-borne short wave infrared CO 2 A method for estimating column concentration, comprising:
s1, acquiring OCO-2 satellite wave band data, and extracting the OCO-2 satellite wave band data through sensitivity analysis of atmospheric carbon dioxide inversion parameters to obtain 9 weak_CO 2 Band and 6O 2 Band data;
s2, 9 weakCO 2 Band and 6O 2 Performing feature screening on the wave band data, the NDVI normalized vegetation index, the SR surface reflectivity data, the DEM elevation topography data, the ERA5 atmosphere data, the AOD aerosol data and the TCCON station observation data, and reserving the first 31 features of screening according to importance;
s3, performing correlation analysis on the first 31 screened features through a heat map to find out the correlation with CO 2 Features of stronger correlation and weaker features of column concentration;
s4, will be combined with CO 2 Combining the characteristic with strong column concentration correlation and the characteristic with weak column concentration correlation as an input characteristic data set, and then adopting decision trees, XGBoost, common random forests, extreme random forests and gradient lifting regression models to carry out CO respectively 2 Average column concentration is estimated by determining coefficient R estimated by different regression models 2 Performing contrast analysis on the accuracy of prediction in the error allowable range, namely, the root mean square error RMSE, the average absolute error MAE, the average relative error MRE and the average relative error MRE, finding out the model with the highest prediction accuracy as an extreme random forest regression model, and using the extreme random forest regression model to carry out CO 2 The average concentration of the columns was predicted.
2. Machine learning-based satellite-borne short wave infrared CO according to claim 1 2 The column concentration estimation method is characterized in that the OCO-2 satellite wave band data comprise longitude lon, latitude lat, zenith angle and azimuth angle of the sun, and zenith angle and azimuth angle of the satellite; the ERA5 atmospheric data includes temperature, humidity, pressure, U/V component of wind, rainfall, boundary layer height, cloud base height, cloud cover, total rainfall, vertical velocity of wind.
3. Machine learning-based satellite-borne short wave infrared CO according to claim 1 2 The column concentration estimation method is characterized in that an extraction range is determined by adopting a resampling mode before extracting OCO-2 satellite wave band data, namely, grids are drawn according to the longitude and latitude range of a target area, the resolution after sampling is set to be 0.5 degrees multiplied by 0.5 degrees, and the Euclidean distance between the center point of each grid and the center point of each pixel corresponding to an original image is obtained through the longitude and latitude of each grid:
in lon k Longitude, lat for fixed site k Is the latitude and lon of the fixed site i 、lat i The longitude and latitude of the grid respectively.
4. Machine learning-based satellite-borne short wave infrared CO according to claim 1 2 The column concentration estimation method is characterized in that for 9 weakCO 2 Band and 6O 2 The abnormal value in the band data is processed as follows:
wherein sigma is the standard deviation of the data of the same day, namely all abnormal values except + -3 sigma are removed.
5. Machine learning-based satellite-borne short wave infrared CO according to claim 1 2 The column concentration estimation method is characterized in that a decision tree uses a base index to divide attributes, and the proportion of a kth sample in a current sample set X is assumed to be p k (k=1, 2,3, …, y), then the base value is:
gini (X) indicates the possibility of random sampling of inconsistencies between two different types of tags;
assuming that the discrete attribute a has v possible values, if a is used to classify the sample set X, v branch nodes are generated and X is recorded v The v branch node comprises all samples which take values on the attribute a in the sample set X; the base index of attribute a is defined as:
the Gini index Gini (X, a) represents the uncertainty of the sample set X after a=a segmentation; the larger the base index, the greater the uncertainty of the sample.
6. Machine learning-based satellite-borne short wave infrared CO according to claim 1 2 A column concentration estimation method is characterized in that K trees are assumed in XGBoost, F represents a tree model, and then a predicted value is obtained
Expressed as:
in which x is i As an input example, a feature vector representing an i-th data point; k is the number of CART trees; f (f) k To represent the kth CART tree;
the corresponding objective function L is:
wherein, l is a loss function and represents the error between the predicted value and the true value; y is i Is a true value; omega is a regularization function that prevents model overfitting.
7. Machine learning-based satellite-borne short wave infrared CO according to claim 1 2 The column concentration estimation method is characterized in that in a common random forest, a model h (X, theta) is established for a characteristic parameter set X of a data set i ) I=1, 2, …, k, randomly selecting m features, so that each leaf node selects the feature of the maximum information gain for splitting; wherein the information gain is expressed as:
wherein i is a regression value, p i Represents the probability of occurrence of the corresponding value, w is the number of dividing nodes,
the weight value of the leaf node is divided for the mth.
8. Machine learning-based satellite-borne short wave infrared CO according to claim 1 2 The column concentration estimation method is characterized in that in an extremely random forest, the generalization error of an individual learner is assumed to be E i The generalized error weight of the learner is:
assume that the divergence value of the individual learner is A i The learner's weighted divergence value is:
the generalization error after integration is expressed as:
w in i And T is the total number of decision trees with different structures for the weight.
9. Machine learning-based satellite-borne short wave infrared CO according to claim 1 2 The column concentration estimating method is characterized in that each iteration of gradient lifting obtains a new learner which is fit for the residual error of the former learner, and finally all the learners are fittedThe predictions of the tree are added, thereby completing the prediction task; the residual error obtaining mode is as follows:
r ni ＝y i -f n-1 (x i )
wherein y is i For the measured value of the ith sample, f n-1 (x i ) The predicted value of the former round of learner; fitting the residual remembering to obtain a fitted residual model h n (x) Updating the regression tree:
f n (x)＝f n-1 (x)+h n (x)。
10. machine learning-based satellite-borne short wave infrared CO according to claim 1 2 The column concentration estimation method is characterized in that the determination coefficient R 2 The obtaining modes of the root mean square error RMSE, the average absolute error MAE and the average relative error MRE are as follows:
wherein: n is the number of samples; f (f) i Is a predicted value; y is i Is a true value;
is the average value. />
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211594763.4A CN116189796A (en) | 2022-12-13 | 2022-12-13 | Machine learning-based satellite-borne short wave infrared CO 2 Column concentration estimation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211594763.4A CN116189796A (en) | 2022-12-13 | 2022-12-13 | Machine learning-based satellite-borne short wave infrared CO 2 Column concentration estimation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116189796A true CN116189796A (en) | 2023-05-30 |
Family
ID=86451347
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211594763.4A Pending CN116189796A (en) | 2022-12-13 | 2022-12-13 | Machine learning-based satellite-borne short wave infrared CO 2 Column concentration estimation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116189796A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117455066A (en) * | 2023-11-13 | 2024-01-26 | 哈尔滨航天恒星数据系统科技有限公司 | Corn planting accurate fertilizer distribution method based on multi-strategy optimization random forest, electronic equipment and storage medium |
-
2022
- 2022-12-13 CN CN202211594763.4A patent/CN116189796A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117455066A (en) * | 2023-11-13 | 2024-01-26 | 哈尔滨航天恒星数据系统科技有限公司 | Corn planting accurate fertilizer distribution method based on multi-strategy optimization random forest, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113919448B (en) | Method for analyzing influence factors of carbon dioxide concentration prediction at any time-space position | |
| CN111401602B (en) | Assimilation method for satellite and ground rainfall measurement values based on neural network | |
| JP4719893B2 (en) | Control device, control method, and program thereof | |
| CN112052627A (en) | Method, device, medium and equipment for estimating near-surface ozone space distribution | |
| Kumar et al. | A machine learning approach to estimating the error in satellite sea surface temperature retrievals | |
| CN107122606A (en) | The Trends of Sea Level Changes computational methods and device counted based on satellite altitude | |
| Li et al. | A neural network-based approach for the detection of heavy precipitation using GNSS observations and surface meteorological data | |
| CN116189796A (en) | Machine learning-based satellite-borne short wave infrared CO 2 Column concentration estimation method | |
| Liu et al. | Hyperspectral infrared sounder cloud detection using deep neural network model | |
| Ziv et al. | Flash floods prediction using precipitable water vapor derived from GPS tropospheric path delays over the eastern mediterranean | |
| CN116609858A (en) | Tropical cyclone atmospheric waveguide prediction method and system based on interpretability XGBoost model | |
| Pegion et al. | Understanding predictability of daily southeast US precipitation using explainable machine learning | |
| Zhang et al. | Improved forest signal detection for space-borne photon-counting lidar using automatic machine learning | |
| Wang et al. | Relative importance of radar variables for nowcasting heavy rainfall: A machine learning approach | |
| CN112330079B (en) | Earth surface information prediction method and device | |
| Khouider et al. | A novel method for interpolating daily station rainfall data using a stochastic lattice model | |
| Amaireh et al. | Atmospheric humidity estimation from wind profiler radar using a cascaded machine learning approach | |
| CN113987782A (en) | Near-ground ultraviolet irradiance satellite missing data completion method based on cloud information | |
| Jayaminda et al. | Rating Performances of Global Climate Models in Capturing Monsoon Rainfall Patterns in Sri Lanka | |
| Oliver et al. | Estimating extremes from global ocean and climate models: A Bayesian hierarchical model approach | |
| Li et al. | Deep learning for ground-level PM 2.5 prediction from satellite remote sensing data | |
| CN112730300B (en) | GIIRS observation area atmosphere profile set sampling method based on shannon entropy | |
| Karpatne et al. | Earth science applications of sensor data | |
| CN114998719A (en) | Forest fire prediction method based on deep learning and multi-source remote sensing data | |
| CN116449460B (en) | Regional month precipitation prediction method and system based on convolution UNet and transfer learning |
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 |