CN114169215A - Surface temperature inversion method coupling remote sensing and regional meteorological model - Google Patents
Surface temperature inversion method coupling remote sensing and regional meteorological model Download PDFInfo
- Publication number
- CN114169215A CN114169215A CN202110904271.XA CN202110904271A CN114169215A CN 114169215 A CN114169215 A CN 114169215A CN 202110904271 A CN202110904271 A CN 202110904271A CN 114169215 A CN114169215 A CN 114169215A
- Authority
- CN
- China
- Prior art keywords
- surface temperature
- image
- modis
- model
- earth surface
- 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
Images
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
-
- 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
Cloud pollution can interfere with the transmission of thermal infrared in the atmosphere, and the potential of a satellite thermal infrared sensor for providing complete space-time coverage for earth surface temperature is limited. In order to solve the problem, the invention discloses a three-step inversion model (TRM), wherein (i) the WRF model is used for simulating the surface temperature which is not polluted by cloud and has the resolution of 1 KM; (ii) fitting the simulated earth surface temperature into a similar MODIS cloudless earth surface temperature by taking the MODIS data as reference and using an optimal machine learning model; (iii) and optimizing by combining a median filtering algorithm, and removing salt and pepper noise points in the fitted image, thereby obtaining the earth surface temperature which has the precision similar to that of the MODIS earth surface temperature and is free from cloud pollution. The constructed model can provide a MODIS cloud-free time sequence earth surface temperature image in a region with frequent cloud coverage, and the defect that the satellite thermal infrared image is polluted by cloud and cannot obtain the complete earth surface temperature is overcome.
Description
Technical Field
The invention relates to the technical field of atmospheric environment and atmospheric remote sensing, in particular to a surface temperature inversion method of a coupling remote sensing and regional meteorological model.
Background
The surface temperature is one of the important indexes for monitoring resource environment and climate change, and plays an important role in surface research. Surface temperature is widely used in various fields including climatology, agriculture and forestry, ecological environment, geology and exploration, etc., and is considered one of the high priority parameters of the international territory and biosphere program. The accurate surface temperature inversion is not only beneficial to evaluating the surface energy, hydrologic balance, thermal inertia and soil humidity, but also beneficial to detecting climate change and mastering the change rule of long-term climate. In recent years, the rapid development of remote sensing technology has made surface temperature inversion by using thermal infrared become the mainstream direction. Due to the complexity of the process of receiving earth surface radiation energy by a satellite and the direct influence of the atmospheric environment, the earth surface temperature reflected by a satellite remote sensing image under a cloudy condition is different from the true value, unfortunately, in the global actual daytime weather, the cloudy weather dominates half of the time, so that the application of an operable earth surface temperature product for space-time analysis is limited, and an effective model needs to be designed to derive the earth surface temperature under the cloudy condition at present.
Many studies have attempted to construct reconstruction methods to recover data lost through cloud contamination, and these methods can be divided into three categories. The first method is based on a surface energy balance method, which needs to assume or know meteorological conditions, and deduces the earth surface temperature under a cloudy condition by inputting the quality of short-wave radiation and other complex regional parameters, so that the parameterization has great uncertainty, and the other disadvantage is that the surface energy balance method cannot be applied at night when the input short-wave radiation cannot dominate the spatial variation of the earth surface temperature; the second method is a gap filling method, which recovers the earth surface temperature by means of space-time information, extracts the space-time information of clear adjacent earth surface temperature pixels from different satellites or other earth surface temperature data on the same day, and deduces the earth surface temperature under a cloud condition by replacing or interpolating, and the method is used on the premise that the earth surface temperature under the cloud is similar to the earth surface temperature under the adjacent geographic area or the adjacent time image under the clear sky, so that only the assumed clear sky earth surface temperature can be deduced, but not the earth surface temperature under the cloudy condition, and a plurality of steps of convolution and iteration are needed to obtain the pixels with large-area cloud pollution, which results in a large accumulated error, so that more information about weather and earth surface characteristics is needed to improve the method, so as to capture the extreme or abnormal earth surface temperature values in the empty area; the third method is a passive microwave measurement method, which can acquire the earth surface temperature in clouds since microwaves can be transmitted through clouds and other atmospheric interference sources, however, estimating the earth surface temperature by microwaves is challenging because the microwave signal varies significantly with the earth surface characteristics, and the technique needs to take into account the spatial and temporal variations in microwave emissivity, and in addition, there are some obstacles in integrating the thermal infrared and passive microwaves to generate all-weather earth surface temperature products, the first problem is that the spatial resolution of the earth surface temperature measured by passive microwaves is much coarser than that of the earth surface temperature measured by thermal infrared, usually on the order of tens of kilometers at the frequency of interest, the second problem is that most passive microwave satellite sensors have bands, resulting in band gaps between two adjacent tracks, passive microwave measurements are therefore often used as a complement to thermal infrared or other ancillary data to reverse the surface temperature.
The WRF is a new generation of mesoscale weather research and forecast model, has perfect physical schemes to describe various complex weather phenomena, can be used as a regional weather simulation model, and has five modularized physical schemes with many options including micro-physics, short wave radiation, long wave radiation, near-ground layers and boundary layers. The method is characterized in that the urban canopy is closely related to the height, density, trend and the like of a building, the variation of urban surface energy can be deeply explored through the combined application of a medium-scale meteorological model and an urban canopy model, the WRF and the urban canopy model can be used for more carefully describing the processes of surface radiation, energy transfer and the like, and the surface temperature under the sunny or cloudy condition can be accurately simulated.
Disclosure of Invention
The invention aims to solve the problems that the transmission of thermal infrared in the atmosphere is interfered by cloud pollution, and the satellite thermal infrared sensor is limited to provide complete space-time coverage potential for the earth surface temperature. In order to solve the problem, a three-step inversion model (TRM) is invented, wherein (i) the WRF model is used for simulating the surface temperature which is not polluted by cloud and has the resolution of 1 KM; (ii) fitting the simulated earth surface temperature into a similar MODIS cloudless earth surface temperature by taking the MODIS data as reference and using an optimal machine learning model; (iii) and optimizing by combining a median filtering algorithm, and removing salt and pepper noise points in the fitted image, thereby obtaining the earth surface temperature which has the precision similar to that of the MODIS earth surface temperature and is free from cloud pollution.
The purpose of the invention is realized by the following technical scheme: a surface temperature inversion method coupling remote sensing and a regional meteorological model comprises the following steps:
(1) and selecting an optimal atmosphere WRF model scheme.
The WRF model is a model for the present study to simulate surface temperature, and is configured as three unidirectional nested domains with spatial resolutions set to 9km, 3km, and 1km, and pixels for the three domains set to 124 × 124, 160 × 160, and 205 × 205, respectively. The WRF is combined with the urban canopy model, so that the surface temperature of the urban area can be simulated more accurately. According to the relevant research, 3 schemes are selected, and 3 × 2 WRF configuration schemes are selected together because static land use data is also taken as an input parameter, and an optimal WRF configuration scheme is selected from the schemes as shown in table 1.
TABLE 1WRF model configuration selection scheme
Errors between the surface temperature simulated by various schemes and the MODIS image need to be compared through the test points, and root mean square errors and correlation coefficients between the surface temperature simulated by various schemes and the MODIS image are used as the quality standard of the judgment schemes. And then simulating the surface temperature free from cloud pollution by using the optimal scheme.
(2) And (3) programming to create a positive sphere used by the WRF model, and converting pixels of the positive sphere into a WGS84 coordinate system so as to enable the pixels to correspond to the pixels of the MODIS image one by one.
(3) And respectively fitting the data by using a linear regression model, a neural network model and a random forest model, selecting an optimal model from the three models, and selecting a good and bad standard which is still the root mean square error and the correlation coefficient. When the MODIS image is less polluted by cloud, establishing a relation between the MODIS image and the WRF simulation image in the earth surface temperature of a clear sky pixel on the same date by using an optimal machine learning model, and applying the model to the simulated earth surface temperature to obtain an earth surface temperature image similar to the MODIS, wherein when the MODIS image is more polluted by cloud, and no enough pixels are used for training the machine learning model, the fitted earth surface temperature image is not highly correlated with the earth surface temperature observed by the MODIS (the correlation coefficient is less than 0.8), the image is considered to be completely polluted by the cloud, and the earth surface temperature image similar to the MODIS is obtained by using the model adjacent in time and the WRF image on the same day, and a specific calculation formula is as follows:
in the formula, LSTcSurface temperature image similar to MODIS, fiFor the time-neighboring training model, N is the number of neighboring images used to estimate the earth's surface temperature, which in this study is 7. R is the correlation coefficient between time-adjacent MODIS image and analog image, T is the number of days between image dates, LSTcWRFFor the current day WRF image, cor is the current day MODIS image and LSTcWRFω is the weight of the image.
(4) The method is characterized in that a two-dimensional sliding window with a certain structure is used, pixels in the window are sorted according to the pixel values, and a monotonously rising (or falling) two-dimensional data sequence is generated. (ii) a
g(x,y)＝med{f(x-k,y-l),(k,l∈W)} (4)
Wherein g (x, y) is the output image of two-dimensional median filtering, f (x, y), g (x, y) are the original image and the processed image respectively, and W is a two-dimensional 3 x 3 window.
Compared with the prior art, the invention has the following advantages and beneficial effects:
the method utilizes WRF and simultaneously combines a city canopy model, and obtains 'real' earth surface temperature through a machine learning model and a median filtering algorithm instead of the prior algorithm which reverses the earth surface temperature under the assumed condition.
Drawings
FIG. 1 is a test study area overview
FIG. 2 is a flow chart of the present invention
FIG. 3 optimal WRF scheme selection test point
FIG. 4 is similar to the MODIS surface temperature acquisition process under the condition of complete cloud pollution or insufficient pixel training.
Fig. 56 WRF configurations corresponding to various land use types and total surface temperature analog numerical root mean square error.
FIG. 6 shows the comparison and correlation between the WRF simulation results and the weather station air temperature.
FIG. 7 is a table showing the surface temperature modeled by scheme 6, which is obtained by linear regression, neural network fitting, and random forest model fitting.
FIG. 8MODIS surface temperature and the correlation coefficient and root mean square error between the surface temperature and the surface temperature fitted by linear regression, neural networks and random forest models.
Fig. 9 shows the MODIS images and TRM simulation images on the corresponding dates and the associated thermodynamic diagrams of the earth surface temperature before and after optimization under the condition of low cloud pollution.
FIG. 10 shows the root mean square error and other rates of change before and after optimization of the surface temperature under conditions of low cloud pollution.
Fig. 11 shows the MODIS image and the TRM simulation image on the corresponding date under the cloud pollution multi-condition, and the associated thermodynamic diagrams of the earth surface temperature before and after optimization.
FIG. 12 root mean square error and other rates of change before and after optimization of surface temperature under various cloud pollution conditions.
Fig. 13 absolute error and correlation coefficient between TRM surface temperature and MODIS surface temperature per soil utilization type.
The TRM model of fig. 14 was inverted to the root mean square error and correlation coefficient for the surface temperature in 3 months of 2020.
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but the present invention is not limited thereto.
The present embodiment will be described in detail with reference to the results of tests conducted in the research area (fig. 1) in beijing as an example, wherein the results are inverted to obtain the cloudless surface temperature by using the present invention. According to the method flowchart shown in fig. 2, the method comprises the steps of:
1: the method comprises the following steps of firstly preprocessing work, including processing of land utilization data and software development of converting an NC format file into a GeoTiff format file. Replacing WRF default geostationary field data with MOD11Q1 land use data; under the Qt platform, format conversion software is developed by using C + + language, a right sphere is created by using a GDAL library, and pixels of the WRF result are converted into a WGS84 coordinate system through ellipsoid conversion.
2: WRF scheme selection and performance, using three-day images which are less polluted by cloud in 3 months and can cover the test points as scheme test data, wherein the three days are respectively 3 months 19 days, 3 months 20 days and 3 months 21 days, and the corresponding time is 2:30(UTC), and fig. 4 shows that the root mean square errors of the simulated values of the earth surface temperatures and the overall earth surface temperatures corresponding to the 6WRF configuration schemes are the closest to the MODIS earth surface temperatures in terms of the overall root mean square errors of the various schemes, and the accuracy is the highest, and the root mean square errors are 1.84, 3.31 and 3.73, respectively, so that the scheme 6 is selected as the optimal scheme. To further verify the reliability of the results of scenario 6, the 19-day, 20-day and 21-day earth surface temperature images were simulated using scenario 6, and the 144 point-to-point comparison results (fig. 5) were obtained using the temperature data provided by the 454110 meteorological site, and the results show that the simulated temperature and the actual temperature have the same trend, and the simulation results have reliability.
3: the selection and performance of the machine learning model are that three machine learning models, namely a linear regression model, a neural network and a random forest are used, three-day 2:30(UTC) earth surface temperature images of 3, 5, 14 and 23 days of 2020 are simulated respectively, the result is shown in fig. 6, visually, the earth surface temperature images fitted by the three machine learning models and the earth surface temperature observed by MODIS have similar spatial distribution patterns, the correlation coefficient between the fitted earth surface temperature images and the earth surface temperature images observed by MODIS on the corresponding date and the root mean square error are calculated to obtain the scatter diagram of fig. 7, the root mean square error of the random forest can be minimum, the correlation is maximum, although some noise points exist, the fitting performance is best, and therefore the random forest is selected as the fitting model.
And 4, generating and verifying a cloudless ground surface temperature image, when the MODIS image is little polluted by clouds, establishing the relation between the ground surface temperatures of the MODIS image and the WRF simulation image in clear sky pixels on the same date by using an optimal machine learning model, and applying the model to the simulated ground surface temperature to obtain the cloudless ground surface temperature image. Because the result image has salt and pepper noise, in order to optimize the surface temperature image, the third step of the TRM model, namely median filtering denoising, is executed. Considering the median filtering optimization effect, the correlation between the images before and after optimization and the MODIS observation images corresponding to the dates is compared, as can be seen from the thermodynamic correlation diagram (fig. 8), the correlation is larger after optimization than before optimization, fig. 9 shows the root mean square error value and the change rate thereof before and after optimization, which indicates that after optimization, the precision of the earth surface temperature is improved, the root mean square error between the earth surface temperature image inverted by the TRM model and the MODIS observation images is about 1.2 ℃, the correlation is larger than 0.9, and when the MODIS images are slightly polluted by clouds, the earth surface temperature can be better inverted.
When the MODIS image is heavily polluted by cloud, the model adjacent in time and the WRF image of the current day are used to obtain the surface temperature image similar to the MODIS, the technical process is shown in fig. 10, the specific calculation formula is (1-3), and the two days of 3, 9 and 18 in 2020 are 2:30(UTC) (fig. 11), the thermodynamic correlation diagram (fig. 11) shows the correlation coefficients of the surface temperature before and after optimization, with the correlation coefficients increasing after optimization, and fig. 12 presents the values of the root mean square error before and after optimization and its rate of change, indicating that the root mean square error decreases overall after optimization. The method shows that the accuracy is improved after the inverted ground surface temperature is optimized under the condition of complete cloud pollution, the root mean square error between the ground surface temperature image inverted by the TRM model and the MODIS observation image is about 1.8 ℃, the correlation is greater than 0.9, and when the MODIS image is completely polluted by the cloud, the TRM can also invert the ground surface temperature with higher accuracy.
5: and (3) performing error analysis on the TRM model, namely comparing the earth surface temperature inverted by the TRM model with the earth surface temperature observed by the corresponding MODIS, wherein the absolute error can more intuitively show the performance of the model, so that the absolute error and a correlation coefficient are used for performing error analysis on the TRM model. Fig. 16 shows absolute errors and correlation coefficients between the MODIS surface temperature and the TRM surface temperature for each land use of 3, 4, 9, 14, 18, and 23 days of 2020, where the surface temperature inverted by the TRM model has a smaller error and a higher correlation with the surface temperature observed by the MODIS, which indicates that the constructed TRM model can invert a higher-precision MODIS-like surface temperature image and can make up for the defects of the cloud images of the MODIS.
6: in order to explain the reliability of our results, the earth surface temperature was inverted for all 3 months in 2020, and from the point line graph (fig. 14) of the correlation coefficient and the root mean square error, it can be seen that the root mean square error was less than 2 ℃, the average value of the overall root mean square error was 1.23 ℃, and the average value of the correlation coefficient was 0.93, so our study results were reliable.
The above examples are preferred embodiments of the present invention, but the present invention is not limited to the above examples, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents thereof, and they are included in the scope of the present invention.
Claims (5)
1. A surface temperature inversion method coupling remote sensing and a regional meteorological model is characterized by comprising the following steps: the method comprises the following steps:
A. simulating the surface temperature without cloud pollution by using an optimal scheme based on an atmosphere WRF (weather research and learning model), wherein the time resolution is 1KM, and the spatial resolution is 1 day;
B. converting the simulated earth surface temperature data to a WGS84 coordinate system to enable the earth surface temperature data to correspond to the MODIS image pixels one by one;
C. fitting the simulated earth surface temperature with a coordinate system into a similar MODIS cloudless earth surface temperature by taking the MODIS data as reference and using an optimal machine learning model;
D. and optimizing by combining a median filtering algorithm, and removing salt and pepper noise points in the fitted image, thereby obtaining the earth surface temperature which has the precision similar to that of the MODIS earth surface temperature and is free from cloud pollution.
2. The method for inverting the earth's surface temperature by coupling remote sensing with the regional meteorological model according to claim 1, wherein the step A specifically comprises:
a1: selecting an optimal atmosphere WRF model scheme;
according to related research, the current commonly used scheme is given, and the optimal scheme is selected from the scheme;
TABLE 1WRF model configuration selection scheme
The root mean square error and the correlation coefficient between the simulated surface temperature of each scheme and the MODIS image are used as the standards for judging the quality of the schemes.
A2: simulating the surface temperature free from cloud pollution by using an optimal scheme;
3. the method for inverting the earth's surface temperature by coupling remote sensing with the regional meteorological model according to claim 1, wherein the step B specifically comprises:
b1: and (3) creating a positive sphere used by the WRF model, and converting the pixel of the positive sphere into a WGS84 coordinate system.
4. The method for inverting the earth's surface temperature by coupling remote sensing with the regional meteorological model according to claim 1, wherein the step C specifically comprises:
c1: respectively fitting the data by using a linear regression model, a neural network model and a random forest model, selecting an optimal model from the three models, and selecting a good and bad standard which is still root mean square error and a correlation coefficient;
c2: when the MODIS image is little polluted by cloud, establishing a relation between the earth surface temperatures of the MODIS image and the WRF simulation image in the clear sky pixels on the same date by using an optimal machine learning model, and applying the model to the simulated earth surface temperature to obtain an earth surface temperature image similar to the MODIS;
c3: when the MODIS image is heavily polluted by cloud and insufficient pixels are used for training a machine learning model, the fitted surface temperature image is not highly correlated with the surface temperature observed by the MODIS (the correlation coefficient is smaller than 0.8), the image is considered to be completely polluted by the cloud, the model adjacent in time and the WRF image of the day are used for obtaining the surface temperature image similar to the MODIS, and the specific calculation formula is as follows:
in the formula, LSTcSurface temperature image similar to MODIS, fiFor training the model in time-neighborhood, N is used to estimate the surface temperatureThe number of adjacent images in this study is 7. R is the correlation coefficient between time-adjacent MODIS image and analog image, T is the number of days between image dates, LSTcWRFFor the current day WRF image, cor is the current day MODIS image and LSTcWRFω is the weight of the image.
5. The method for inverting the earth's surface temperature by coupling remote sensing with the regional meteorological model according to claim 1, wherein the step D specifically comprises:
d1: and (4) optimizing by using a median filtering algorithm, and removing salt and pepper noise points in the fitted image.
The method uses a two-dimensional sliding window with a certain structure to sort the pixels in the window according to the pixel values, and generates a monotonously rising (or falling) two-dimensional data sequence.
g(x,y)＝med{f(x-k,y-l),(k,l∈W)} (4)
Wherein g (x, y) is the output image of two-dimensional median filtering, f (x, y), g (x, y) are the original image and the processed image respectively, and W is a two-dimensional 3 x 3 window.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110904271.XA CN114169215B (en) | 2021-08-06 | 2021-08-06 | Surface temperature inversion method coupling remote sensing and regional meteorological model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110904271.XA CN114169215B (en) | 2021-08-06 | 2021-08-06 | Surface temperature inversion method coupling remote sensing and regional meteorological model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114169215A true CN114169215A (en) | 2022-03-11 |
| CN114169215B CN114169215B (en) | 2022-11-15 |
Family
ID=80476466
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110904271.XA Active CN114169215B (en) | 2021-08-06 | 2021-08-06 | Surface temperature inversion method coupling remote sensing and regional meteorological model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114169215B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115511887A (en) * | 2022-11-22 | 2022-12-23 | 北京慧天卓特科技有限公司 | Drought index prediction method and device based on artificial intelligence fusion |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104765903A (en) * | 2015-01-27 | 2015-07-08 | 中国科学院城市环境研究所 | Comprehensive urban climate environment assessment method integrated with multiple-source remote sensing and climate environment information |
| US20160217230A1 (en) * | 2015-01-23 | 2016-07-28 | Iteris, Inc. | Modeling of crop growth for desired moisture content of bovine feedstuff and determination of harvest windows for corn earlage using field-level diagnosis and forecasting of weather conditions and field observations |
| CN109297605A (en) * | 2018-10-09 | 2019-02-01 | 北京大学 | A kind of Surface Temperature Retrieval method with Thermal Infrared Data infrared based in |
| CN110516816A (en) * | 2019-08-30 | 2019-11-29 | 中国科学院、水利部成都山地灾害与环境研究所 | Round-the-clock surface temperature generation method and device based on machine learning |
-
2021
- 2021-08-06 CN CN202110904271.XA patent/CN114169215B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160217230A1 (en) * | 2015-01-23 | 2016-07-28 | Iteris, Inc. | Modeling of crop growth for desired moisture content of bovine feedstuff and determination of harvest windows for corn earlage using field-level diagnosis and forecasting of weather conditions and field observations |
| CN104765903A (en) * | 2015-01-27 | 2015-07-08 | 中国科学院城市环境研究所 | Comprehensive urban climate environment assessment method integrated with multiple-source remote sensing and climate environment information |
| CN109297605A (en) * | 2018-10-09 | 2019-02-01 | 北京大学 | A kind of Surface Temperature Retrieval method with Thermal Infrared Data infrared based in |
| CN110516816A (en) * | 2019-08-30 | 2019-11-29 | 中国科学院、水利部成都山地灾害与环境研究所 | Round-the-clock surface temperature generation method and device based on machine learning |
Non-Patent Citations (1)
| Title |
|---|
| WEI CHEN等: "The Investigation of Urbanization and Urban Heat Island in Beijing Based on Remote Sensing", 《URBAN PLANNING AND ARCHITECTURE DESIGN FOR SUSTAINABLE DEVELOPMENT》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115511887A (en) * | 2022-11-22 | 2022-12-23 | 北京慧天卓特科技有限公司 | Drought index prediction method and device based on artificial intelligence fusion |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114169215B (en) | 2022-11-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109580003B (en) | Method for estimating near-ground atmospheric temperature by thermal infrared data of stationary meteorological satellite | |
| CN112905560B (en) | Air pollution prediction method based on multi-source time-space big data deep fusion | |
| Nguyen-Xuan et al. | The Vietnam gridded precipitation (VnGP) dataset: construction and validation | |
| CN109974665B (en) | Aerosol remote sensing inversion method and system for short-wave infrared data lack | |
| Fenner et al. | CrowdQC+—a quality-control for crowdsourced air-temperature observations enabling world-wide urban climate applications | |
| CN104636608A (en) | Direct assimilation method of MODIS satellite data | |
| Carouge et al. | What can we learn from European continuous atmospheric CO 2 measurements to quantify regional fluxes–Part 2: Sensitivity of flux accuracy to inverse setup | |
| CN110059745B (en) | Basin rainfall product correction method and system based on satellite-ground fusion | |
| CN111191673B (en) | Ground surface temperature downscaling method and system | |
| CN111982822B (en) | Long-time sequence high-precision vegetation index improvement algorithm | |
| CN114019579A (en) | High-space-time resolution near-surface air temperature reconstruction method, system and equipment | |
| CN107391794B (en) | Typhoon continuous three-dimensional wind field inversion method | |
| Fan et al. | A comparative study of four merging approaches for regional precipitation estimation | |
| CN112926468A (en) | Tidal flat elevation automatic extraction method | |
| CN114169215B (en) | Surface temperature inversion method coupling remote sensing and regional meteorological model | |
| CN113408111B (en) | Atmospheric precipitation inversion method and system, electronic equipment and storage medium | |
| CN117423002B (en) | Small-scale tidal flat image processing method based on DEM | |
| CN115840908B (en) | Method for constructing PM2.5 three-dimensional dynamic monitoring field based on microwave link of LSTM model | |
| CN116609859A (en) | Weather disaster high-resolution regional mode forecasting system and method | |
| CN114417728A (en) | Near-surface air temperature inversion method based on temperature, emissivity and deep learning | |
| Le Toumelin et al. | A two-folds deep learning strategy to correct and downscale winds over mountains | |
| Jiang et al. | Surface global and diffuse solar radiation over China acquired from geostationary Multi-functional Transport Satellite data | |
| CN115993668B (en) | Polynomial correction and neural network-based PWV reconstruction method and system | |
| CN114581043B (en) | Cross-network modular multi-source multi-factor wave-condition analysis system | |
| Dutta et al. | Dynamic annotation and visualisation of the South Esk hydrological sensor web |
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 |