CN115511887A - Drought index prediction method and device based on artificial intelligence fusion - Google Patents

Drought index prediction method and device based on artificial intelligence fusion Download PDF

Info

Publication number
CN115511887A
CN115511887A CN202211464979.9A CN202211464979A CN115511887A CN 115511887 A CN115511887 A CN 115511887A CN 202211464979 A CN202211464979 A CN 202211464979A CN 115511887 A CN115511887 A CN 115511887A
Authority
CN
China
Prior art keywords
pixel
quasi
image
target area
thermal infrared
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
Application number
CN202211464979.9A
Other languages
Chinese (zh)
Other versions
CN115511887B (en
Inventor
方莉
唐剑
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Huitian Zhuote Technology Co ltd
Original Assignee
Beijing Huitian Zhuote Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Huitian Zhuote Technology Co ltd filed Critical Beijing Huitian Zhuote Technology Co ltd
Priority to CN202211464979.9A priority Critical patent/CN115511887B/en
Publication of CN115511887A publication Critical patent/CN115511887A/en
Application granted granted Critical
Publication of CN115511887B publication Critical patent/CN115511887B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J5/00—Radiation pyrometry, e.g. infrared or optical thermometry
    • G01J5/48—Thermography; Techniques using wholly visual means
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10048—Infrared image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; 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

Abstract

The invention discloses a drought index prediction method and a drought index prediction device based on artificial intelligence fusion, wherein a specific implementation mode of the method comprises the following steps: firstly, acquiring a surface thermal infrared image corresponding to a target area to be detected at daily observation time; secondly, aiming at any pixel of a target area to be detected: acquiring a surface thermal infrared observation value corresponding to the pixel from the surface thermal infrared image; carrying out cloud interference detection on the surface thermal infrared observation value; determining the quasi-surface temperature corresponding to the pixel based on the detection result; then, determining a quasi-surface temperature image corresponding to the target area to be detected based on the quasi-surface temperature corresponding to each pixel; and finally, performing prediction processing on the earth surface temperature image to generate a drought index image corresponding to the daily observation time of the target area to be detected. Therefore, the drought index of the target area to be detected is predicted in all weather, and the spatial coverage and accuracy of the drought index image prediction corresponding to the target area to be detected are improved.

Description

Drought index prediction method and device based on artificial intelligence fusion
Technical Field
The invention belongs to the technical field of image processing, and particularly relates to a drought index prediction method and device based on artificial intelligence fusion.
Background
Acquiring a surface thermal infrared image of a target area by using a remote sensing thermal infrared sensor carried by a static orbit satellite; the collected surface thermal infrared image not only has high time resolution and high spatial resolution; and compared with a visible wave band in a thermal infrared wave band in the surface thermal infrared image, the signal that the vegetation in the target area reduces transpiration due to water shortage can be detected 7-14 days in advance. However, the thermal infrared band of the surface thermal infrared image is seriously disturbed by clouds and is limited to drought index inversion under the condition of no clouds.
Because the microwave band is long and can penetrate through a cloud layer to obtain earth surface information, an earth surface microwave image of a target area is acquired by adopting a remote sensing microwave sensor carried by a polar orbit satellite, and the earth surface microwave image is an optimal choice for filling up cloud interference conditions in earth surface thermal infrared images. However, remote sensing microwave sensors also have drawbacks, such as: firstly, the lower intensity of the microwave signal causes the spatial resolution of the microwave sensor to be lower than that of the thermal infrared sensor; secondly, the microwave sensor carried on the polar orbit satellite only has twice transit observation every day, and compared with a static orbit thermal infrared sensor which can image once in 5 minutes to 1 hour, the time resolution of the microwave sensor is poorer. The land pattern analysis system can also simulate a land pattern estimation image of the target area; firstly, the land pattern analysis system has high time resolution and high matching degree with a thermal infrared band; secondly, the land pattern analysis system is an all-weather coverage sample with clear sky observation and cloud observation. However, the land pattern analysis system also has drawbacks, such as: because the land pattern analysis system is limited by model-driven parameters, the spatial resolution of large-scale land pattern simulation is generally lower than that of remote sensing products. Therefore, the image observation system or the analysis system based on the single wave band and the physical model has advantages and disadvantages.
Therefore, it is urgently needed to provide a drought index prediction method with high resolution, high precision and all-day coverage so as to overcome the defects of the prior art.
Disclosure of Invention
The invention provides a drought index prediction method and a drought index prediction device based on artificial intelligence fusion; the method can monitor the all-weather drought index of the target area to be detected, and improves the accuracy of the drought index prediction of the target area to be detected.
In order to achieve the above object, according to a first aspect of embodiments of the present application, there is provided a drought index prediction method based on artificial intelligence fusion, the method including: a drought index prediction method based on artificial intelligence fusion is characterized by comprising the following steps: acquiring a surface thermal infrared image corresponding to the daily observation time of a target area to be detected; aiming at any pixel of the target area to be detected: acquiring a surface thermal infrared observation value corresponding to the pixel from the surface thermal infrared image; performing cloud interference detection on the surface thermal infrared observation value; if the detection result represents that the surface thermal infrared observation value is free of cloud interference, taking the surface thermal infrared observation value as a quasi surface temperature; if the detection result represents that the surface thermal infrared observation value is cloud interference, respectively acquiring a surface microwave observation value and a land mode estimation value corresponding to the pixel in the daily observation time; predicting the surface microwave observed value and the land pattern estimated value to obtain a quasi-surface temperature corresponding to the pixel; determining a quasi-surface temperature image corresponding to the target area to be detected based on the quasi-surface temperature corresponding to each pixel; and carrying out prediction processing on the quasi-surface temperature image to generate a drought index image corresponding to the daily observation time of the target area to be detected.
Optionally, the performing prediction processing on the surface microwave observed value and the land pattern estimated value to obtain a quasi-surface temperature corresponding to the pixel includes: acquiring a quasi-surface microwave image, a quasi-land mode estimation image and a quasi-surface thermal infrared image without cloud interference, which correspond to the daily observation time of a target area; aiming at any pixel of the target area: respectively acquiring a quasi-surface microwave observation value, a quasi-land mode estimation value and a quasi-surface thermal infrared observation value corresponding to the pixel from the quasi-surface microwave image, the quasi-land mode estimation image and the quasi-surface thermal infrared image; the quasi-surface microwave observation value and the quasi-land surface mode estimation value are jointly used as a first training sample, and the quasi-land surface mode estimation value and the quasi-surface thermal infrared observation value are jointly used as a second training sample; performing model training on the first training sample data and the second training sample data based on a decision tree algorithm to obtain a double-decision tree model; and predicting the earth surface microwave observed value and the land surface mode estimated value by using the double decision tree model to obtain the quasi-earth surface temperature corresponding to the pixel.
Optionally, the obtaining of the quasi-surface thermal infrared image without cloud interference corresponding to the target area at the daily observation time includes: acquiring a surface thermal infrared image and a cloud image corresponding to the daily observation time of a target area; for any pixel of the target area: respectively acquiring a surface thermal infrared observation value and a cloud template observation value corresponding to the pixel from the surface thermal infrared image and the cloud image; if the observed value of the cloud template corresponding to the pixel meets a first preset condition, taking the surface thermal infrared observed value corresponding to the pixel as an effective value; if the cloud template observed value corresponding to the pixel meets a second preset condition, taking the surface thermal infrared observed value corresponding to the pixel as an invalid value; and obtaining a quasi-surface thermal infrared image without cloud interference corresponding to the target area based on the surface thermal infrared observation value corresponding to each pixel in the target area.
Optionally, the obtaining of the quasi-surface microwave image corresponding to the target area at the daily observation time includes: acquiring a surface microwave image corresponding to the daily observation time of a target area; and carrying out scale reduction processing on the earth surface microwave image based on the earth surface leaf area index corresponding to the preset spatial resolution to obtain a quasi-earth surface microwave image.
Optionally, the obtaining of the quasi-land mode estimation image corresponding to the daily observation time of the target area includes: acquiring a land mode estimation image corresponding to a daily observation time of a target area; and carrying out downscaling processing on the land mode estimation image based on the land surface leaf area index corresponding to the preset spatial resolution to obtain a quasi-land mode estimation image.
Optionally, the acquiring of the surface thermal infrared image of the target area corresponding to the daily observation time includes: determining the pixel longitude and latitude of a disc corresponding to the static orbit satellite based on the platform height of the static orbit satellite, the platform attitude parameter and the corresponding sub-satellite point longitude and latitude; establishing a mapping relation between the longitude and latitude of all pixels of the disc and the row and column numbers of the image of the disc corresponding to the stationary orbit satellite; acquiring a disc image of a target region acquired by the stationary orbit satellite at the daily observation time; for any pixel in the target area: determining the longitude and latitude of a target corresponding to the pixel; based on the mapping relation, selecting a sampling value corresponding to a pixel closest to the target longitude and latitude distance from the disc image as an earth surface thermal infrared observation value of the pixel; and acquiring a surface thermal infrared image acquired by the geostationary orbit satellite aiming at the target area at the daily observation time based on the surface thermal infrared observation value corresponding to each pixel of the target area.
Optionally, the predicting the quasi-surface temperature image to generate a drought index image corresponding to the daily observation time of the target area to be detected includes: acquiring quasi-surface temperature corresponding to each pixel in the target area to be detected from the quasi-surface temperature image; for any pixel of the target region: carrying out prediction processing on the quasi-surface temperature corresponding to the pixel to generate the daily actual evapotranspiration corresponding to the pixel; determining the daily potential evapotranspiration amount corresponding to the pixel based on the soil net radiant energy and vegetation net radiant energy corresponding to the pixel at the observation time; determining a daily drought index corresponding to the pixel based on the daily potential evapotranspiration amount and the daily actual evapotranspiration amount; and determining a daily drought index image corresponding to the daily observation time of the target area to be detected based on the daily drought index corresponding to each pixel of the target area to be detected.
To achieve the above object, there is also provided, according to a second aspect of the embodiments of the present application, an artificial intelligence fusion-based drought index prediction apparatus, including: the acquisition module is used for acquiring a surface thermal infrared image corresponding to the daily observation time of a target area to be detected; the first determining module is used for aiming at any pixel of the target area to be detected: acquiring a surface thermal infrared observation value corresponding to the pixel from the surface thermal infrared image; performing cloud interference detection on the surface thermal infrared observation value; if the detection result indicates that the surface thermal infrared observation value is free of cloud interference, taking the surface thermal infrared observation value as a quasi surface temperature; if the detection result represents that the surface thermal infrared observation value is cloud interference, respectively acquiring a surface microwave observation value and a land mode estimation value corresponding to the pixel in the daily observation time; predicting the surface microwave observed value and the land pattern estimated value to obtain a quasi-surface temperature corresponding to the pixel; the second determination module is used for determining a quasi-surface temperature image corresponding to the target area to be detected based on the quasi-surface temperature corresponding to each pixel; and the prediction module is used for performing prediction processing on the quasi-surface temperature image to generate a drought index image corresponding to the daily observation time of the target area to be detected.
To achieve the above object, according to a third aspect of the embodiments of the present application, there is also provided an electronic apparatus, including: one or more processors; memory for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to carry out the method according to the first aspect.
To achieve the above object, according to a fourth aspect of embodiments of the present application, there is also provided a computer readable medium having a computer program stored thereon, which when executed by a processor, implements the method according to the first aspect.
Compared with the prior art, the embodiment of the invention provides a drought index prediction method and a drought index prediction device based on artificial intelligence fusion, wherein the method comprises the following steps: firstly, acquiring a surface thermal infrared image corresponding to a target area to be detected at daily observation time; secondly, aiming at any pixel of the target area to be detected: acquiring a surface thermal infrared observation value corresponding to the pixel from the surface thermal infrared image; carrying out cloud interference detection on the surface thermal infrared observation value; if the detection result indicates that the surface thermal infrared observation value is free of cloud interference, taking the surface thermal infrared observation value as a quasi surface temperature; if the detection result indicates that the surface thermal infrared observation value is cloud interference, respectively obtaining a surface microwave observation value and a land mode estimation value corresponding to the pixel in daily observation time; predicting the surface microwave observation value and the land pattern estimation to obtain a quasi-surface temperature corresponding to the pixel; then, determining a quasi-surface temperature image corresponding to the target area to be detected based on the quasi-surface temperature corresponding to each pixel; and finally, carrying out prediction processing on the quasi-surface temperature image to generate a drought index image corresponding to the daily observation time of the target area to be detected. Therefore, the advantages of the surface thermal infrared observation value, the surface microwave observation value and the land mode estimation value are combined, the artificial intelligence fusion algorithm is utilized to predict the drought index of the target area to be detected in all weather, and the spatial coverage and the accuracy of the drought index image prediction corresponding to the target area to be detected are improved.
Drawings
Some specific embodiments of the invention will be described in detail hereinafter, by way of illustration and not limitation, with reference to the accompanying drawings. The same reference numbers in the drawings identify the same or similar elements or components. Those skilled in the art will appreciate that the drawings are not necessarily drawn to scale. In the drawings:
fig. 1 is a schematic flow chart of a drought index prediction method based on artificial intelligence fusion according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart illustrating a process of determining a quasi-surface temperature corresponding to a pixel based on a model according to an embodiment of the present invention;
FIG. 3 is a schematic flow chart illustrating a process of generating a daily drought index image corresponding to a target area to be detected according to an embodiment of the present invention;
FIG. 4 is a schematic representation of a subsurface thermal infrared image and a quasi-subsurface thermal infrared image in accordance with an embodiment of the present invention; wherein fig. 4a represents a surface thermal infrared image and fig. 4b represents a quasi-surface thermal infrared image;
FIG. 5 is a diagram illustrating a comparison analysis of advantages and disadvantages of different data sources according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of a drought index prediction device based on artificial intelligence fusion according to an embodiment of the present invention.
Detailed Description
In order to make the objects, features and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a schematic flow diagram of a drought index prediction method based on artificial intelligence fusion according to an embodiment of the present invention.
A drought index prediction method based on artificial intelligence fusion at least comprises the following steps:
s101, acquiring a surface thermal infrared image corresponding to a target area to be detected at daily observation time;
s102, aiming at any pixel of a target area to be detected: acquiring a surface thermal infrared observation value corresponding to the pixel from the surface thermal infrared image; carrying out cloud interference detection on the surface thermal infrared observation value; if the detection result represents that the surface thermal infrared observation value is free of cloud interference, taking the surface thermal infrared observation value as a quasi surface temperature; if the detection result represents that the earth surface thermal infrared observation value is cloud interference, respectively acquiring an earth surface microwave observation value and a land surface mode estimation value corresponding to the pixel; predicting the surface microwave observed value and the land mode estimated value to obtain the quasi-surface temperature corresponding to the pixel;
s103, determining a quasi-surface temperature image corresponding to the target area to be detected based on the quasi-surface temperature corresponding to each pixel;
and S104, performing prediction processing on the surface temperature image to generate a drought index image corresponding to the daily observation time of the target area to be detected.
In S101, a surface thermal infrared image corresponding to a target area to be measured is acquired from a thermal infrared sensor mounted on a stationary orbit satellite.
In S102 and S103, the target area to be measured includes a plurality of pixels.
Aiming at any pixel: acquiring a surface thermal infrared observation value corresponding to the pixel from the surface thermal infrared image; detecting whether the earth surface thermal infrared observation value is larger than a preset threshold value or not; if not, determining that the earth surface thermal infrared observation value is free of cloud interference, and taking the earth surface thermal infrared observation value free of cloud interference as a quasi-earth surface temperature; if so, determining that the earth surface thermal infrared observation value is cloud interference; then, acquiring a ground surface microwave image corresponding to the daily observation time of the target area to be detected from a microwave sensor carried by the polar orbit satellite; acquiring a land pattern estimation image corresponding to a daily observation time of a target area to be detected from a land pattern analysis system; extracting a surface microwave observation value corresponding to the pixel from the surface microwave image; extracting a land mode estimation value corresponding to the pixel from the land mode estimation image; inputting the earth surface microwave observed value and the land surface mode estimated value into a double decision tree model, performing prediction processing, and outputting a quasi-earth surface temperature corresponding to the pixel; because each pixel has a corresponding quasi-surface temperature, a quasi-surface temperature image corresponding to the target area can be obtained.
And in S104, inputting the quasi-surface temperature image into a surface-atmosphere energy exchange model, performing prediction processing, and outputting a daily drought index image corresponding to the daily observation time of the target area to be detected.
The method and the device combine respective advantages of the surface thermal infrared observation value, the surface microwave observation value and the land mode estimation value, based on the artificial intelligence fusion algorithm, the daily drought index image corresponding to the target area to be detected is predicted, the spatial coverage and accuracy of the daily drought index image prediction corresponding to the target area to be detected are improved, and the problems that in the prior art, due to the fact that a single-waveband physical model or an analysis system is adopted to predict the drought index of the target area to be detected, the prediction effective range is small and the accuracy is low are solved.
Fig. 5 is a schematic diagram illustrating analysis results of different data sources according to an embodiment of the invention.
Collecting a ground surface thermal infrared image corresponding to a target area to be measured by adopting a remote sensing thermal infrared sensor, and collecting a ground surface microwave image corresponding to the target area to be measured by adopting a remote sensing microwave sensor; collecting a land mode estimation image corresponding to a target area to be detected by using a land mode analysis system; the method is adopted to acquire the quasi-surface temperature image corresponding to the target area to be measured.
As can be seen from fig. 5, the surface thermal infrared image is characterized by: high image precision, high time resolution, high space resolution and no cloud interference pixels. The surface microwave image has the characteristics that: low image accuracy, low temporal resolution, medium spatial resolution, cloud-interference pixels. The land mode estimation image is characterized in that: medium image accuracy, high temporal resolution, low spatial resolution, cloud-interference pixels. The quasi-surface temperature image has the characteristics that: high image precision, high time resolution, high spatial resolution, and cloud-interference pixels.
Therefore, compared with the earth surface temperature image of the target area to be detected acquired by a single wave band and a physical model, the earth surface temperature image corresponding to a plurality of wave bands is combined, the precision, the time resolution and the spatial resolution of the quasi earth surface temperature image corresponding to the target area to be detected are improved, cloud interference pixels in the target area to be detected can be accurately observed, the spatial coverage and the accuracy of observation of the target area to be detected are improved, and the accuracy of daily drought index image prediction of the target area to be detected is improved.
Fig. 2 is a schematic flow chart illustrating a process of determining a quasi-surface temperature corresponding to a pixel based on a model according to an embodiment of the present invention.
S201, acquiring a quasi-surface microwave image, a quasi-land mode estimation image and a quasi-surface thermal infrared image without cloud interference, which correspond to a target area at daily observation time;
s202, aiming at any pixel of the target area: respectively acquiring a quasi-surface microwave observation value, a quasi-land mode estimation value and a quasi-surface thermal infrared observation value corresponding to the pixel from the quasi-surface microwave image, the quasi-land mode estimation image and the quasi-surface thermal infrared image; taking the quasi-surface microwave observed value and the quasi-land mode estimated value as a first training sample together, and taking the quasi-land mode estimated value and the quasi-surface thermal infrared observed value as a second training sample together;
s203, model training is carried out on the first training sample data and the second training sample data based on a decision tree algorithm to obtain a double-decision tree model;
and S204, performing prediction processing on the earth surface microwave observed value and the land surface mode estimated value by using a double decision tree model to obtain the quasi-earth surface temperature corresponding to the pixel.
Here, the quasi-surface microwave image, the quasi-land pattern estimation image, and the quasi-surface thermal infrared image without the cloud interference are all images having a spatial resolution of 2 km.
The first training sample data is a data set formed by a plurality of first training samples; the second training sample data is a data set formed of several second training samples. For example: the target area comprises a plurality of image elements, and each image element is provided with a corresponding first training sample and a corresponding second training sample; the target area thus has a plurality of first training samples and a plurality of second training samples. When the same day has two day observation times, the number of the first training samples and the second training samples is doubled on the basis of the above.
And performing model training by using a decision tree algorithm aiming at the first training sample data and the second training sample data to obtain a double-decision tree model. When the earth surface thermal infrared observation value corresponding to the pixel of the target area to be detected is in the condition of cloud interference, an error exists when the earth surface thermal infrared observation value corresponding to the pixel is adopted to carry out drought index estimation; therefore, a double-decision tree model is adopted to predict the earth surface microwave observed value and the land surface mode estimated value corresponding to the pixel, and the quasi-earth surface temperature corresponding to the pixel is output. Therefore, the quasi-surface temperature corresponding to the pixel is determined based on the surface microwave observed value and the land pattern estimated value, the influence of cloud interference on the pixel observed value is reduced, and the accuracy of obtaining the quasi-surface temperature corresponding to the pixel is improved.
It should be noted that, when the dual decision tree model is trained, the training sample data set includes not only the non-cloud interference pixel but also the cloud interference pixel.
In addition, the implementation utilizes the land mode estimated value as a bridge to establish a double decision tree model of the surface thermal infrared observed value and the surface microwave observed value, so that all-weather observation is realized on the target area to be detected, the observation result has high spatial resolution, high temporal resolution and high precision, and the accuracy of the drought index image prediction corresponding to the target area to be detected is improved.
FIG. 4 is a schematic representation of a subsurface thermal infrared image and a quasi-subsurface thermal infrared image in accordance with an embodiment of the present invention; fig. 4a shows a surface thermal infrared image, and fig. 4b shows a quasi-surface thermal infrared image.
In another preferred embodiment of this embodiment, the obtaining of the quasi-surface thermal infrared image without cloud interference corresponding to the daily observation time of the target area at least includes the following steps:
s1, acquiring a surface thermal infrared image and a cloud image of a target area corresponding to daily observation time;
s2, aiming at any pixel of the target area: respectively acquiring a surface thermal infrared observation value and a cloud template observation value corresponding to the pixel from the surface thermal infrared image and the cloud image;
s3, if the observed value of the cloud template corresponding to the pixel meets a first preset condition, taking the surface thermal infrared observed value corresponding to the pixel as an effective value; if the cloud template observed value corresponding to the pixel meets a second preset condition, taking the surface thermal infrared observed value corresponding to the pixel as an invalid value;
and S4, obtaining a quasi-surface thermal infrared image without cloud interference corresponding to the target area based on the surface thermal infrared observation value corresponding to each pixel in the target area.
In S1, exemplarily, based on the platform height of the geostationary orbit satellite, the platform attitude parameter, and the corresponding infrasatellite longitude and latitude, determining the pixel longitude and latitude of the disc corresponding to the geostationary orbit satellite; establishing a mapping relation between the longitude and latitude of all pixels of the disc and the image row number and column number of the disc corresponding to the stationary orbit satellite; acquiring a disc image of a static orbit satellite in a target area acquired at daily observation time; aiming at any pixel in the target area: determining the longitude and latitude of a target corresponding to the pixel; based on the mapping relation, selecting a sampling value corresponding to a pixel closest to the target longitude and latitude from the disc image as a surface thermal infrared observation value of the pixel; and acquiring a surface thermal infrared image acquired by the stationary orbit satellite aiming at the target area at daily observation time based on the surface thermal infrared observation value corresponding to each pixel of the target area.
For example: taking FY-4A as an example for explanation, based on a geometric positioning algorithm of latitude and longitude of a disc of a stationary orbit, according to the height of an FY-4A satellite platform, attitude parameters of the satellite platform and corresponding latitude and longitude of an infrastar point, the latitude and longitude of a pixel of the FY-4A disc are calculated, a mapping relation between the latitude and longitude of the pixel of the disc and an image row number and column number of the disc is established, and a latitude and longitude lookup table is generated.
Resampling by using a sensor on FY-4A according to the range of the target area and the spatial resolution (res) to obtain a disc image; and then traversing each pixel of the target area one by one, wherein a target longitude and latitude calculation formula corresponding to the (i, j) th pixel is shown as a formula (1):
Figure DEST_PATH_IMAGE001
formula (1);
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE002
a start longitude representing a target area is shown,
Figure DEST_PATH_IMAGE003
representing a starting latitude of the target area;
Figure DEST_PATH_IMAGE004
indicates the end longitude of the target area and,
Figure DEST_PATH_IMAGE005
representing the terminal latitude of the target area;
Figure DEST_PATH_IMAGE006
representing the spatial resolution of the target region.
The target longitude and latitude of the (i, j) th pixel of the target area
Figure DEST_PATH_IMAGE007
And comparing with the longitude and latitude lookup table of the disc, and selecting a sampling value corresponding to the pixel closest to the target longitude and latitude from the disc image as the surface infrared observation value of the pixel.
Because the target area corresponds to a plurality of pixels, the earth surface thermal infrared image collected by the FY-4A aiming at the target area at the daily observation time can be determined based on the earth surface infrared observation value corresponding to each pixel.
Therefore, the accuracy of acquiring the earth surface thermal infrared image corresponding to the target area is improved by establishing the longitude and latitude lookup table of the disc corresponding to the first stationary orbit satellite and determining the earth surface infrared image of the target area according to the longitude and latitude lookup table after the disc image is acquired.
Acquiring a disc cloud coverage image of a static orbit satellite in a target area acquired at daily observation time; aiming at any pixel in the target area: determining the longitude and latitude of a target corresponding to the pixel; based on the mapping relation, selecting a sampling value corresponding to a pixel closest to the target longitude and latitude distance from the disc cloud coverage image as an earth surface thermal infrared cloud observation value of the pixel; and acquiring a cloud image acquired by the geostationary orbit satellite aiming at the target area at daily observation time based on the earth surface thermal infrared cloud observation value corresponding to each pixel of the target area.
In S2 to S4, the target area has a plurality of pixels; for any pixel: if the cloud template observed value corresponding to the pixel is not more than 1 (for example, the cloud template observed value is 0 or the cloud template observed value is 1), taking the earth surface thermal infrared observed value corresponding to the pixel as an effective value; and if the cloud template observed value corresponding to the pixel is greater than 1 (for example, the cloud template observed value is 2 or the cloud template observed value is 3), taking the surface thermal infrared observed value corresponding to the pixel as an invalid value. Therefore, cloud interference can be effectively removed from the earth surface thermal infrared observation value corresponding to the target area, and the quasi-earth surface thermal infrared image without cloud interference corresponding to the target area can be accurately acquired.
In another preferred embodiment of this embodiment, the acquiring a quasi-surface microwave image corresponding to a daily observation time of a target area at least includes the following steps:
s1, acquiring a ground surface microwave image corresponding to a target area at daily observation time;
and S2, carrying out scale reduction processing on the earth surface microwave image based on the earth surface leaf area index corresponding to the preset spatial resolution to obtain a quasi-earth surface microwave image.
Specifically, a surface microwave image corresponding to a target area is obtained from a microwave sensor carried on a polar orbit satellite; acquiring all pixels of a target area, and aiming at any pixel: carrying out downscaling processing on the surface microwave observed value corresponding to the pixel to obtain a quasi-surface microwave observed value; and determining a quasi-surface microwave image corresponding to the target area based on the quasi-surface microwave observation value corresponding to each pixel of the target area.
For example: and (3) carrying out scale reduction processing on the earth surface microwave observed value corresponding to the image element according to the following formula (2):
Figure DEST_PATH_IMAGE008
Figure DEST_PATH_IMAGE009
formula (2);
wherein, the first and the second end of the pipe are connected with each other,WM C representing a surface microwave observation value corresponding to the pixel under the condition that the spatial resolution is 9km;WM H representing a quasi-surface microwave observation value corresponding to the pixel under the condition that the preset spatial resolution is 2km,LAIrepresenting the surface leaf area index corresponding to the spatial resolution of 2 km;
Figure DEST_PATH_IMAGE010
representing pixel weights for indicating weights based on the surface leaf area index; i represents the corresponding pixel number under the condition that the spatial resolution is 9 km.
It should be noted that, in this embodiment, the algorithm used in the downscaling process is implemented by using an LAI product of a medium-resolution imaging spectrometer MODIS (mode-resolution imaging spectrometer).
Therefore, the ground surface microwave image is subjected to downscaling processing, so that the quasi-ground surface microwave image subjected to downscaling processing and the quasi-ground surface thermal infrared image have the same spatial resolution, consistency of model input data can be ensured, and accuracy of dual-decision tree model training is improved.
In another preferred implementation manner of this embodiment, the obtaining a quasi-land mode estimated image of the target area corresponding to the daily observation time includes:
s1, obtaining a land mode estimation image corresponding to a daily observation time of a target area;
and S2, carrying out downscaling processing on the land mode estimated image based on the land surface leaf area index corresponding to the preset spatial resolution to obtain a quasi-land mode estimated image.
The preset spatial resolution is used to indicate a preset value of spatial resolution of 2 km.
For example, the Climate Forecast Reanalysis System (CFSR) has characteristics of global coverage, spatial resolution of 25km, and temporal resolution of 3 h. Collecting a land pattern estimation image corresponding to a target area by using a CFSR as a land pattern analysis system; then all pixels of the target area are obtained, and aiming at any pixel: carrying out downscaling processing on the land mode estimated value corresponding to the image element to obtain a quasi-land mode estimated value; and determining a quasi-land mode estimated image corresponding to the target area based on the quasi-land mode estimated value corresponding to each pixel of the target area.
Here, the method of performing downscaling processing on the land pattern estimation value is similar to the method of performing downscaling processing on the surface microwave observation value, and repeated description is omitted here.
Therefore, the land mode estimated image is subjected to dimensionality reduction processing, so that the quasi-land mode estimated image subjected to dimensionality reduction and the quasi-surface thermal infrared image are kept consistent in spatial resolution, and the accuracy of the double-decision tree model training is improved.
Fig. 3 is a schematic flow chart illustrating a process of generating a daily drought index image corresponding to a target area to be detected according to an embodiment of the present invention.
In a preferred further implementation manner of this embodiment, generating a daily drought index image corresponding to a target area to be detected at least includes the following steps:
s301, acquiring the quasi-surface temperature corresponding to each pixel in the target area to be detected from the quasi-surface temperature image;
s302, aiming at any pixel of the target area: carrying out prediction processing on the quasi-surface temperature corresponding to the pixel to generate the daily actual evapotranspiration corresponding to the pixel; determining the daily potential evapotranspiration amount corresponding to the pixel based on the soil net radiant energy and vegetation net radiant energy corresponding to the pixel in the observation time; determining a daily drought index corresponding to the pixel element based on the daily potential evapotranspiration and the daily actual evapotranspiration;
s303, determining a daily drought index image corresponding to the daily observation time of the target area to be detected based on the daily drought index corresponding to each pixel of the target area to be detected.
Specifically, the quasi-surface temperature corresponding to the pixel is input into a surface-atmosphere energy exchange model for prediction processing, and the daily actual evapotranspiration corresponding to the pixel is output; and then, determining the daily potential evapotranspiration amount corresponding to the pixel based on the soil net radiant energy and the vegetation net radiant energy corresponding to the daily observation time of the pixel. For example: the daily potential evapotranspiration of the pixel corresponding to the observation time is also divided into two parts of soil and vegetation, and is specifically obtained by a Priestley-Taylor approximate formula, as shown in the following formulas (3) to (5):
Figure DEST_PATH_IMAGE011
formula (3);
Figure DEST_PATH_IMAGE012
formula (4);
Figure DEST_PATH_IMAGE013
formula (5);
wherein the content of the first and second substances,
Figure 958721DEST_PATH_IMAGE014
and
Figure DEST_PATH_IMAGE015
the daily potential evapotranspiration of vegetation and soil respectively, gamma is the measurement constant of 0.067 kPa/degree centigrade, and S is the difference between the saturated vapor pressure and the temperature curve. Tau is a canopy transfer factor and is covered by vegetation
Figure 344703DEST_PATH_IMAGE016
And the zenith angle of the sun
Figure DEST_PATH_IMAGE017
And (6) determining.
Figure DEST_PATH_IMAGE018
Is a constant of 1.3, however
Figure DEST_PATH_IMAGE019
Is determined by tau, if tau is less than or equal to 0.5, then
Figure 858861DEST_PATH_IMAGE019
= 1; if τ is greater than 0.5, then
Figure DEST_PATH_IMAGE020
。
And finally, obtaining the daily potential evapotranspiration amount corresponding to the pixel as follows:
Figure DEST_PATH_IMAGE021
。
determining a daily drought index corresponding to the pixel element based on the daily potential evapotranspiration and the daily actual evapotranspiration, wherein the daily drought index comprises the following steps: determining the daily evapotranspiration proportion corresponding to the pixel based on the daily potential evapotranspiration and the daily actual evapotranspiration; acquiring a meteorological reference of a target area; and (4) making a ratio of the daily evapotranspiration proportion of the pixel to a meteorological reference to obtain a daily drought index corresponding to the pixel.
For example: the daily evapotranspiration ratio was calculated by the following formula (6):
Figure DEST_PATH_IMAGE022
formula (6).
Parallel preset observation time periods are daily
Figure DEST_PATH_IMAGE023
Thereby obtaining the meteorological reference of the image element. The meteorological reference is within a preset observation time period f PET The specific calculation formulas are shown in the following formulas (7) and (8):
Figure DEST_PATH_IMAGE024
formula (7);
Figure DEST_PATH_IMAGE025
formula (8).
Figure DEST_PATH_IMAGE026
Comparing with weather reference, and calculating the daily drought index according to the following formula (9)
Figure DEST_PATH_IMAGE027
：
Figure DEST_PATH_IMAGE028
And (9).
It should be noted that, the surface leaf area index parameter LAI is the LAI product (product number mcd15a2 h.006) of the medium-resolution imaging spectrometer MODIS (simulation-resolution imaging spectrometer) adopted in the present invention. The Climate forecasting System analyzes data (CFSR), extracts long-wave/short-wave uplink radiation and downlink radiation, atmospheric temperature profile, atmospheric pressure profile and relative humidity profile from the CFSR analysis data, and is used for calculating boundary layer energy and PET parameters.
Therefore, the model can accurately acquire the daily drought index corresponding to the pixel in the target area to be detected, so as to effectively acquire the daily drought index image corresponding to the target area to be detected,
in a further preferred embodiment of this embodiment, a drought index prediction method based on artificial intelligence fusion further includes:
s1, aiming at any one daily drought index in the drought index image: judging whether the daily drought index meets a preset condition or not;
s2, if the daily drought index meets a first preset condition, marking a pixel corresponding to the daily drought index for indicating a first color with drought;
and S3, if the daily drought index meets a second preset condition, marking the pixel corresponding to the daily drought index for indicating a second color without drought.
For example, the drought index ranges from-3.5 to 3.5, the (-3.5, 0) interval indicates that the region corresponding to the pixel element is dry, and the (0, 3.5) interval indicates that the region corresponding to the pixel element is abundant in water and no drought occurs.
Therefore, whether the regions corresponding to the pixels are dry or not is judged through the drought index, and the judgment results are distinguished through different colors, so that the output of the drought index image is more visual, the user identification is facilitated, and the user experience is improved.
It should be understood that, in the various embodiments of the present invention, the sequence numbers of the above-mentioned processes do not mean the execution sequence, and the execution sequence of each process should be determined by its function and the inherent logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.
The schematic flow chart of the drought index prediction of the target area to be detected in the application embodiment of the invention is shown.
The method of the present embodiment will be described in detail with reference to specific applications, and the specific procedures are as follows.
The method comprises the following steps: and acquiring a surface thermal infrared image and a cloud image corresponding to the daily observation time of the target area. For any pixel of the target area: respectively acquiring a surface thermal infrared observation value and a cloud template observation value corresponding to the pixel from the surface thermal infrared image and the cloud image; if the observed value of the cloud template corresponding to the pixel is 0 or 1, taking the earth surface thermal infrared observed value corresponding to the pixel as an effective value; if the cloud template observation value corresponding to the pixel is 2 or 3, taking the earth surface thermal infrared observation value corresponding to the pixel as an invalid value; and obtaining a quasi-surface thermal infrared image without cloud interference corresponding to the target area based on the surface thermal infrared observation value corresponding to each pixel in the target area.
Acquiring a surface microwave image corresponding to the daily observation time of a target area; the spatial resolution corresponding to the surface microwave image is 9km; and then, carrying out downscaling processing on the earth surface microwave image based on the earth surface leaf area index corresponding to the 2km spatial resolution to obtain a quasi-earth surface microwave image with the spatial resolution of 2 km.
Acquiring a land mode estimation image corresponding to a daily observation time of a target area; the spatial resolution corresponding to the land mode estimation image is 25km; and performing downscaling processing on the land mode estimation image based on the land surface leaf area index corresponding to the spatial resolution of 2km to obtain a quasi-land mode estimation image with the spatial resolution of 2 km.
Step two: aiming at any pixel of the target area: respectively acquiring a quasi-surface microwave observation value, a quasi-land mode estimation value and a quasi-surface thermal infrared observation value corresponding to the pixel from the quasi-surface microwave image, the quasi-land mode estimation image and the quasi-surface thermal infrared image; and taking the quasi-surface microwave observed value and the quasi-land mode estimated value as a first training sample together, and taking the quasi-land mode estimated value and the quasi-surface thermal infrared observed value as a second training sample together. And performing model training on the first training sample data and the second training sample data based on a decision tree algorithm to obtain a double-decision tree model.
Step three: and acquiring a surface thermal infrared image acquired by the stationary orbit satellite aiming at the target area to be detected at daily observation time. Aiming at any pixel of a target area to be detected: acquiring a surface thermal infrared observation value corresponding to the pixel from the surface thermal infrared image; carrying out cloud interference detection on the surface thermal infrared observation value; if the detection result represents that the surface thermal infrared observation value is free of cloud interference, taking the surface thermal infrared observation value as a quasi surface temperature; if the detection result represents that the earth surface thermal infrared observation value is cloud interference, respectively obtaining an earth surface microwave observation value and a land surface mode estimation value corresponding to the daily observation time of the pixel; inputting the surface microwave observed value and the land pattern estimated value into a double decision tree model; outputting the quasi-surface temperature corresponding to the pixel; and determining a quasi-surface temperature image corresponding to the target area to be detected based on the quasi-surface temperature corresponding to each pixel.
Step four: acquiring the quasi-surface temperature corresponding to each pixel in the target area to be detected from the quasi-surface temperature image; for any pixel of the target area: carrying out prediction processing on the quasi-surface temperature corresponding to the pixel by using a surface-atmosphere energy exchange model to generate the daily actual evapotranspiration corresponding to the pixel; determining the daily potential evapotranspiration amount corresponding to the pixel based on the soil net radiant energy and vegetation net radiant energy corresponding to the pixel in the daily observation time; determining a daily drought index corresponding to the pixel based on the daily potential evapotranspiration amount and the daily actual evapotranspiration amount;
step five: and determining a daily drought index image corresponding to the daily observation time of the target area to be detected based on the daily drought index corresponding to each pixel of the target area to be detected.
The implementation benefit is based on the advantages and disadvantages of the drought products of the single wave band and the physical model, and the drought monitoring and forecasting products with high resolution, high precision and all-weather coverage are generated by utilizing the artificial intelligence fusion algorithm and combining the advantages of the remote sensing thermal infrared, the remote sensing microwave and the land mode data
In addition, the artificial intelligence decision tree model of the thermal infrared and microwave products established by taking the land mode product as the bridge has the advantages of complete coverage, high representativeness and better precision. The fused product has all-weather coverage and has the advantages of high spatial resolution, high temporal resolution and high precision.
Fig. 6 is a schematic structural diagram of an artificial intelligence fusion-based drought index prediction apparatus according to an embodiment of the present invention. An artificial intelligence fusion-based drought index prediction device at least comprises: the acquisition module 601 is used for acquiring a surface thermal infrared image corresponding to a target area to be detected at daily observation time; a first determining module 602, configured to, for any pixel of the target area to be measured: acquiring a surface thermal infrared observation value corresponding to the pixel from the surface thermal infrared image; carrying out cloud interference detection on the surface thermal infrared observation value; if the detection result represents that the surface thermal infrared observation value is free of cloud interference, taking the surface thermal infrared observation value as a quasi surface temperature; if the detection result represents that the surface thermal infrared observation value is cloud interference, respectively acquiring a surface microwave observation value and a land mode estimation value corresponding to the pixel in the daily observation time; predicting the surface microwave observed value and the land pattern estimated value to obtain a quasi-surface temperature corresponding to the pixel; a second determining module 603, configured to determine, based on the quasi-surface temperature corresponding to each pixel, a quasi-surface temperature image corresponding to the target region to be detected; and the prediction module 604 is configured to perform prediction processing on the quasi-surface temperature image to generate a drought index image corresponding to the daily observation time of the target area to be detected.
In a preferred embodiment, the apparatus further comprises: the second determining module includes: the acquisition unit is used for acquiring a quasi-surface microwave image, a quasi-land mode estimation image and a quasi-surface thermal infrared image without cloud interference, which correspond to the daily observation time of a target area; a sample unit, configured to, for any pixel of the target region: respectively acquiring a quasi-surface microwave observation value, a quasi-land mode estimation value and a quasi-surface thermal infrared observation value corresponding to the pixel from the quasi-surface microwave image, the quasi-land mode estimation image and the quasi-surface thermal infrared image; using the quasi-surface microwave observation value and the quasi-land mode estimation value as a first training sample together, and using the quasi-land mode estimation value and the quasi-surface thermal infrared observation value as a second training sample together; the training unit is used for carrying out model training on the first training sample data and the second training sample data based on a decision tree algorithm to obtain a double-decision tree model; and the prediction unit is used for performing prediction processing on the earth surface microwave observed value and the land surface mode estimated value by using the double decision tree model to obtain the quasi-earth surface temperature corresponding to the pixel.
In a preferred embodiment, the acquisition unit comprises: the system comprises a first acquisition subunit, a second acquisition subunit and a third acquisition subunit, wherein the first acquisition subunit is used for acquiring a surface thermal infrared image and a cloud image corresponding to a target area at daily observation time; a second obtaining subunit, configured to, for any pixel of the target region: respectively acquiring a surface thermal infrared observation value and a cloud template observation value corresponding to the pixel from the surface thermal infrared image and the cloud image; the first determining subunit is used for taking the earth surface thermal infrared observation value corresponding to the pixel as an effective value if the cloud template observation value corresponding to the pixel meets a first preset condition; if the cloud template observed value corresponding to the pixel meets a second preset condition, taking the surface thermal infrared observed value corresponding to the pixel as an invalid value; and the second determining subunit is used for obtaining a quasi-surface thermal infrared image without cloud interference corresponding to the target area based on the surface thermal infrared observation value corresponding to each pixel in the target area.
In a preferred embodiment, the obtaining unit further comprises: the third acquisition subunit is used for acquiring a ground surface microwave image corresponding to the daily observation time of the target area; and the first preprocessing unit is used for carrying out downscaling processing on the earth surface microwave image based on the earth surface leaf area index corresponding to the preset spatial resolution to obtain a quasi-earth surface microwave image.
In a preferred embodiment, the acquiring unit further comprises: the fourth acquisition subunit is used for acquiring a land mode estimation image corresponding to the daily observation time of the target area; and the second preprocessing unit is used for carrying out downscaling processing on the land mode estimated image based on the land surface leaf area index corresponding to the preset spatial resolution to obtain a quasi-land mode estimated image.
In a preferred embodiment, the first acquisition subunit comprises: the device comprises a first determining unit, a second determining unit and a control unit, wherein the first determining unit is used for determining the pixel longitude and latitude of a disc corresponding to the geostationary orbit satellite based on the platform height of the geostationary orbit satellite, the platform attitude parameter and the corresponding infrasatellite point longitude and latitude; the establishing unit is used for establishing a mapping relation between the longitude and latitude of all pixels of the disc and the image row-column number of the disc corresponding to the stationary orbit satellite; the first acquisition unit is used for acquiring a disc image of a target area acquired by the stationary orbit satellite at the daily observation time; a second determining unit, configured to, for any pixel in the target region: determining the longitude and latitude of a target corresponding to the pixel; based on the mapping relation, selecting a sampling value corresponding to a pixel closest to the target longitude and latitude distance from the disc image as an earth surface thermal infrared observation value of the pixel; and the second acquisition unit is used for acquiring a surface thermal infrared image acquired by the geostationary orbit satellite aiming at the target area at the daily observation time based on the surface thermal infrared observation value corresponding to each pixel of the target area.
In a preferred embodiment, the prediction module comprises: the acquisition unit is used for acquiring the quasi-surface temperature corresponding to each pixel in the target area to be detected from the quasi-surface temperature image; a first determining unit, configured to, for any pixel of the target area: carrying out prediction processing on the quasi-surface temperature corresponding to the pixel to generate the daily actual evapotranspiration corresponding to the pixel; determining the daily potential evapotranspiration amount corresponding to the pixel based on the soil net radiant energy and vegetation net radiant energy corresponding to the pixel at the observation time; determining a daily drought index corresponding to the pixel element based on the daily potential evapotranspiration and the daily actual evapotranspiration; and the second determining unit is used for determining a daily drought index image corresponding to the daily observation time of the target area to be detected based on the daily drought index corresponding to each pixel of the target area to be detected.
The device can execute the drought index prediction method based on artificial intelligence fusion, and has the corresponding functional modules and beneficial effects of executing the drought index prediction method based on artificial intelligence fusion. For technical details that are not described in detail in this embodiment, reference may be made to the artificial intelligence fusion-based drought index prediction method provided in the embodiment of the present invention.
The present invention also provides an electronic device comprising: a processor; a memory for storing the processor-executable instructions; the processor is used for reading the executable instructions from the memory and executing the instructions to realize the drought index prediction method based on artificial intelligence fusion.
In addition to the above-described methods and apparatus, embodiments of the present application may also be a computer program product comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps in the methods according to the various embodiments of the present application described in the "exemplary methods" section of this specification, above.
The computer program product may be written with program code for performing the operations of embodiments of the present application in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server.
Furthermore, embodiments of the present application may also be a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, cause the processor to perform steps in a method according to embodiments of the present application described in the "exemplary methods" section above of this specification.
The computer-readable storage medium may take any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may include, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The foregoing describes the general principles of the present application in conjunction with specific embodiments, however, it is noted that the advantages, effects, etc. mentioned in the present application are merely examples and are not limiting, and they should not be considered essential to the various embodiments of the present application. Furthermore, the foregoing disclosure of specific details is for the purpose of illustration and description and is not intended to be limiting, since the foregoing disclosure is not intended to be exhaustive or to limit the disclosure to the precise details disclosed.
The block diagrams of devices, apparatuses, devices, systems referred to in this application are only used as illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. These devices, apparatuses, devices, systems may be connected, arranged, configured in any manner, as will be appreciated by one skilled in the art. Words such as "including," "comprising," "having," and the like are open-ended words that mean "including, but not limited to," and are used interchangeably herein. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".
It should also be noted that in the devices, apparatuses, and methods of the present application, the components or steps may be decomposed and/or recombined. These decompositions and/or recombinations are to be considered as equivalents of the present application.
The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
The foregoing description has been presented for purposes of illustration and description. Furthermore, the description is not intended to limit embodiments of the application to the form disclosed herein. While a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, alterations, additions and sub-combinations thereof.
In the description of the specification, reference to the description of "one embodiment," "some embodiments," "an example," "a specific example," or "some examples" or the like means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality" means two or more unless specifically defined otherwise.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.

Claims (10)

1. A drought index prediction method based on artificial intelligence fusion is characterized by comprising the following steps:
acquiring a surface thermal infrared image corresponding to the daily observation time of a target area to be detected;
aiming at any pixel of the target area to be detected: acquiring a surface thermal infrared observation value corresponding to the pixel from the surface thermal infrared image; carrying out cloud interference detection on the surface thermal infrared observation value; if the detection result indicates that the surface thermal infrared observation value is free of cloud interference, taking the surface thermal infrared observation value as a quasi surface temperature; if the detection result represents that the surface thermal infrared observation value is cloud interference, respectively acquiring a surface microwave observation value and a land mode estimation value corresponding to the pixel in the daily observation time; predicting the surface microwave observed value and the land pattern estimated value to obtain a quasi-surface temperature corresponding to the pixel;
determining a quasi-surface temperature image corresponding to the target area to be detected based on the quasi-surface temperature corresponding to each pixel;
and performing prediction processing on the quasi-surface temperature image to generate a drought index image corresponding to the daily observation time of the target area to be detected.
2. The method according to claim 1, wherein the predicting the surface microwave observed value and the land pattern estimated value to obtain the quasi-surface temperature corresponding to the pixel comprises:
acquiring a quasi-surface microwave image, a quasi-land mode estimation image and a quasi-surface thermal infrared image without cloud interference, which correspond to the daily observation time of a target area;
aiming at any pixel of the target area: respectively acquiring a quasi-surface microwave observation value, a quasi-land mode estimation value and a quasi-surface thermal infrared observation value corresponding to the pixel from the quasi-surface microwave image, the quasi-land mode estimation image and the quasi-surface thermal infrared image; the quasi-surface microwave observation value and the quasi-land surface mode estimation value are jointly used as a first training sample, and the quasi-land surface mode estimation value and the quasi-surface thermal infrared observation value are jointly used as a second training sample;
performing model training on the first training sample data and the second training sample data based on a decision tree algorithm to obtain a double-decision tree model;
and predicting the earth surface microwave observed value and the land surface mode estimated value by using the double decision tree model to obtain the quasi-earth surface temperature corresponding to the pixel.
3. The method according to claim 2, wherein the obtaining of the quasi-surface thermal infrared image without cloud interference of the target area at the daily observation time comprises:
acquiring a surface thermal infrared image and a cloud image corresponding to the daily observation time of a target area;
for any pixel of the target area: respectively acquiring a surface thermal infrared observation value and a cloud template observation value corresponding to the pixel from the surface thermal infrared image and the cloud image;
if the observed value of the cloud template corresponding to the pixel meets a first preset condition, taking the surface thermal infrared observed value corresponding to the pixel as an effective value; if the cloud template observed value corresponding to the pixel meets a second preset condition, taking the earth surface thermal infrared observed value corresponding to the pixel as an invalid value;
and obtaining a quasi-surface thermal infrared image without cloud interference corresponding to the target area based on the surface thermal infrared observation value corresponding to each pixel in the target area.
4. The method of claim 2, wherein the obtaining of the quasi-surface microwave image of the target region corresponding to the daily observation time comprises:
acquiring a surface microwave image corresponding to the daily observation time of a target area;
and carrying out scale reduction processing on the earth surface microwave image based on the earth surface leaf area index corresponding to the preset spatial resolution to obtain a quasi-earth surface microwave image.
5. The method of claim 2, wherein the obtaining of the estimated quasi-terrestrial mode image of the target area at the daily observation time comprises:
acquiring a land mode estimation image corresponding to a daily observation time of a target area;
and carrying out downscaling processing on the land mode estimation image based on the land surface leaf area index corresponding to the preset spatial resolution to obtain a quasi-land mode estimation image.
6. The method of claim 3, wherein the obtaining of the surface thermal infrared image of the target area corresponding to the daily observation time comprises:
determining the pixel longitude and latitude of a disc corresponding to the static orbit satellite based on the platform height of the static orbit satellite, the platform attitude parameter and the corresponding sub-satellite point longitude and latitude;
establishing a mapping relation between the longitude and latitude of all pixels of the disc and the image row number and column number of the disc corresponding to the stationary orbit satellite;
acquiring a disc image of a target area acquired by the stationary orbit satellite at the daily observation time;
for any pixel in the target region: determining the longitude and latitude of a target corresponding to the pixel; based on the mapping relation, selecting a sampling value corresponding to a pixel closest to the target longitude and latitude distance from the disc image as an earth surface thermal infrared observation value of the pixel;
and acquiring a surface thermal infrared image acquired by the geostationary orbit satellite aiming at the target area at the daily observation time based on the surface thermal infrared observation value corresponding to each pixel of the target area.
7. The method according to claim 1, wherein the predicting the quasi-surface temperature image to generate a drought index image corresponding to the daily observation time of the target area to be detected comprises:
acquiring quasi-surface temperature corresponding to each pixel in the target area to be detected from the quasi-surface temperature image;
for any pixel of the target region: carrying out prediction processing on the quasi-surface temperature corresponding to the pixel to generate the daily actual evapotranspiration corresponding to the pixel; determining the daily potential evapotranspiration amount corresponding to the pixel based on the soil net radiant energy and vegetation net radiant energy corresponding to the pixel at the observation time; determining a daily drought index corresponding to the pixel element based on the daily potential evapotranspiration and the daily actual evapotranspiration;
and determining a daily drought index image corresponding to the daily observation time of the target area to be detected based on the daily drought index corresponding to each pixel of the target area to be detected.
8. A drought index prediction device based on artificial intelligence fusion is characterized by comprising:
the acquisition module is used for acquiring a surface thermal infrared image corresponding to the daily observation time of a target area to be detected;
the first determining module is used for aiming at any pixel of the target area to be detected: acquiring a surface thermal infrared observation value corresponding to the pixel from the surface thermal infrared image; carrying out cloud interference detection on the surface thermal infrared observation value; if the detection result represents that the surface thermal infrared observation value is free of cloud interference, taking the surface thermal infrared observation value as a quasi surface temperature; if the detection result represents that the surface thermal infrared observation value is cloud interference, respectively acquiring a surface microwave observation value and a land mode estimation value corresponding to the pixel in the daily observation time; predicting the surface microwave observed value and the land pattern estimated value to obtain a quasi-surface temperature corresponding to the pixel;
the second determination module is used for determining a quasi-surface temperature image corresponding to the target area to be detected based on the quasi-surface temperature corresponding to each pixel;
and the prediction module is used for performing prediction processing on the quasi-surface temperature image to generate a drought index image corresponding to the daily observation time of the target area to be detected.
9. An electronic device, comprising:
one or more processors;
a memory for storing one or more programs,
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of any of claims 1-7.
10. A computer-readable medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-7.
CN202211464979.9A 2022-11-22 2022-11-22 Drought index prediction method and device based on artificial intelligence fusion Active CN115511887B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211464979.9A CN115511887B (en) 2022-11-22 2022-11-22 Drought index prediction method and device based on artificial intelligence fusion

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211464979.9A CN115511887B (en) 2022-11-22 2022-11-22 Drought index prediction method and device based on artificial intelligence fusion

Publications (2)

Publication Number Publication Date
CN115511887A true CN115511887A (en) 2022-12-23
CN115511887B CN115511887B (en) 2023-04-07

Family

ID=84514226

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211464979.9A Active CN115511887B (en) 2022-11-22 2022-11-22 Drought index prediction method and device based on artificial intelligence fusion

Country Status (1)

Country Link
CN (1) CN115511887B (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160203245A1 (en) * 2015-01-14 2016-07-14 State Grid Corporation Of China Method for simulating wind field of extreme arid region based on wrf
CN110319938A (en) * 2019-06-26 2019-10-11 西安空间无线电技术研究所 A kind of high spatial resolution surface temperature generation method
CN111191673A (en) * 2019-11-29 2020-05-22 广州地理研究所 Ground surface temperature downscaling method and system
CN114169215A (en) * 2021-08-06 2022-03-11 中国矿业大学(北京) Surface temperature inversion method coupling remote sensing and regional meteorological model

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160203245A1 (en) * 2015-01-14 2016-07-14 State Grid Corporation Of China Method for simulating wind field of extreme arid region based on wrf
CN110319938A (en) * 2019-06-26 2019-10-11 西安空间无线电技术研究所 A kind of high spatial resolution surface temperature generation method
CN111191673A (en) * 2019-11-29 2020-05-22 广州地理研究所 Ground surface temperature downscaling method and system
CN114169215A (en) * 2021-08-06 2022-03-11 中国矿业大学(北京) Surface temperature inversion method coupling remote sensing and regional meteorological model

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
刘蓉等: "结合可见光近红外和微波遥感估算藏北高原日蒸散发量", 《红外与毫米波学报》 *

Also Published As

Publication number Publication date
CN115511887B (en) 2023-04-07

Similar Documents

Publication Publication Date Title
Yang et al. Comparison of three dual‐source remote sensing evapotranspiration models during the MUSOEXE‐12 campaign: Revisit of model physics
US20210208307A1 (en) Training a machine learning algorithm and predicting a value for a weather data variable, especially at a field or sub-field level
Kolassa et al. Soil moisture retrieval from multi‐instrument observations: Information content analysis and retrieval methodology
Anjum et al. Quantification of spatial temporal variability of snow cover and hydro-climatic variables based on multi-source remote sensing data in the Swat watershed, Hindukush Mountains, Pakistan
Zhu et al. Reconstruction of land surface temperature under cloudy conditions from Landsat 8 data using annual temperature cycle model
Liu et al. Estimation of surface and near-surface air temperatures in arid northwest china using landsat satellite images
Brogniez et al. Validation of upper-tropospheric humidity from SAPHIR on board Megha-Tropiques using tropical soundings
de Carvalho Alves et al. Geostatistical improvements of evapotranspiration spatial information using satellite land surface and weather stations data
Eswar et al. Comparison of three remote sensing based models for the estimation of latent heat flux over India
Ramírez-Cuesta et al. Evaluating the impact of adjusting surface temperature derived from Landsat 7 ETM+ in crop evapotranspiration assessment using high-resolution airborne data
Pan et al. Satellite retrieval of surface evapotranspiration with nonparametric approach: Accuracy assessment over a semiarid region
Fan et al. A comparative study of four merging approaches for regional precipitation estimation
Lee et al. A comparative study on generating simulated Landsat NDVI images using data fusion and regression method—the case of the Korean Peninsula
Liu et al. Simultaneous retrieval of land surface temperature and emissivity from the FengYun-4A advanced geosynchronous radiation imager
Zhao et al. Triangle Space-Based Surface Soil Moisture Estimation by the Synergistic Use of $ In\Situ $ Measurements and Optical/Thermal Infrared Remote Sensing: An Alternative to Conventional Validations
Lou et al. An effective method for canopy chlorophyll content estimation of marsh vegetation based on multiscale remote sensing data
Rawat et al. Parameterization of the modified water cloud model (MWCM) using normalized difference vegetation index (NDVI) for winter wheat crop: a case study from Punjab, India
Olsen et al. Evaluating EO-based canopy water stress from seasonally detrended NDVI and SIWSI with modeled evapotranspiration in the Senegal River Basin
Ding et al. Reconstruction of hourly all-weather land surface temperature by integrating reanalysis data and thermal infrared data from geostationary satellites (RTG)
Zhang et al. Improved model for estimating the biomass of Populus euphratica forest using the integration of spectral and textural features from the Chinese high-resolution remote sensing satellite GaoFen-1
Zhao et al. Cloud identification and properties retrieval of the Fengyun-4A satellite using a ResUnet model
CN115578650B (en) Method and device for predicting drought index based on surface temperature image
Gao et al. An improved algorithm for retrieving land surface emissivity and temperature from MSG-2/SEVIRI data
CN107576399A (en) Towards bright the temperature Forecasting Methodology and system of MODIS forest fire detections
CN115511887B (en) Drought index prediction method and device based on artificial intelligence fusion

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