A method for extracting evapotranspiration based on high temporal and spatial resolution remote sensing data
By performing spectral screening and analysis in high-spatial-temporal remote sensing data, combined with Fourier transform and feature filtering, the problem of inaccurate evaporation extraction in the prior art is solved, and higher accuracy and accuracy are achieved.
Patent Information
- Application Number
- CN202410622365.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-20
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-05-20
AI Technical Summary
The prior art is difficult to effectively extract evaporation through high-spatial-temporal remote sensing data, especially in the process of spectral screening and analysis, and there is a lack of accurate methods to obtain the evaporation difference value.
By setting the acquisition area, capturing high-quality remote sensing data, performing spectral screening and traversing cycles, obtaining matching spectral coefficients, using the extraction analysis index and variable cardinality for interval decomposition, combining Fourier transform and feature filtering, spectral characteristic coefficients are generated, and the spectral characteristic coefficients are generated, and the The virtual extraction space is compared and the evaporation difference value is calculated.
The accuracy and accuracy of evaporation extraction are improved, the ability to observe the evaporation changes in the acquisition area is enhanced, and the accuracy of classification matching and extraction accuracy are improved.
Smart Images

Figure CN118447259B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of remote sensing technology, in particular to an evapotranspiration extraction method based on high temporal and spatial resolution remote sensing data. Background Art
[0002] High temporal and spatial resolution remote sensing data refers to the image data of the earth's surface acquired multiple times in a short period of time by high-resolution remote sensing satellites. This kind of data can provide detailed surface information and change trends, and is of great value in monitoring changes in the earth's surface environment, resource surveys, disaster warning and response, urban planning and other fields.
[0003] Evapotranspiration refers to the amount of water that vegetation absorbs from the soil during its growth and releases into the atmosphere after transpiration in the plant body. Extracting evapotranspiration from the collection area through high-temporal and spatial resolution remote sensing data can help researchers better understand vegetation growth, water resource utilization, and the health of the ecosystem. This is of great significance for environmental protection, agricultural production, water resource management, and other aspects; therefore, intelligent extraction of evapotranspiration through high-temporal and spatial resolution remote sensing data has important theoretical and practical significance.
[0004] The problem we need to solve is how to use remote sensing technology to perform spectral screening on the collected high-quality remote sensing data and traverse the loop to obtain matching spectral coefficients; set the variation base and extraction analysis index to summarize and extract the matching spectral coefficients to obtain spectral characteristic coefficients; construct an extraction prediction space to perform virtual transformation on the spectral characteristic coefficients to obtain extracted spectral coefficients; set a comparison period to perform virtual extraction to obtain the extraction comparison difference, and obtain the evaporation difference value based on the obtained extraction comparison difference, capture time and comparison period. To this end, a method for extracting evaporation based on high temporal and spatial resolution remote sensing data is provided. Summary of the invention
[0005] The purpose of the present invention can be achieved through the following technical solutions:
[0006] A method for extracting evapotranspiration based on high temporal and spatial resolution remote sensing data comprises the following steps:
[0007] Step S1: Set a collection area, build a capture terminal according to the set collection area, capture the comprehensive data of the collection area through the capture terminal, and obtain high-quality remote sensing data and capture time;
[0008] Step S2: setting a screening band to perform spectral screening on high-quality remote sensing data to obtain high-quality spectral data, setting a capture window to traverse and cycle the high-quality spectral data to obtain matching spectral coefficients, setting an extraction analysis index and a variable base, performing interval decomposition on the extraction analysis index through the variable base, and obtaining a decomposition interval coefficient;
[0009] Step S3: performing amplitude-frequency acquisition on the obtained decomposition interval coefficients to obtain the coefficient frequency shift, summarizing and extracting the matching spectrum coefficients according to the obtained coefficient frequency shift to obtain the spectrum feature coefficients, and constructing the extraction prediction space according to the acquisition area;
[0010] Step S4: Perform Fourier transform on the spectral characteristic coefficients by extracting the prediction space to obtain the frequency domain characteristic coefficients, perform characteristic filtering on the frequency domain characteristic coefficients by setting the target filter end to obtain the extracted spectral coefficients, generate a spectral coefficient map according to the obtained extracted spectral coefficients, set the comparison period to perform virtual extraction on the obtained extraction prediction space to obtain the virtual spectral characteristic coefficients, perform characteristic filtering and virtual comparison on the obtained virtual spectral characteristic coefficients to obtain the extracted comparison difference, and obtain the evaporation difference value according to the obtained extraction comparison difference, capture time and comparison period.
[0011] The process of obtaining matching spectral coefficients includes:
[0012] Set the statistical period according to the obtained capture time;
[0013] Set the screening band to perform spectral screening on the acquired high-quality remote sensing data to obtain high-quality spectral data;
[0014] Matching the obtained high-quality spectral data with the corresponding acquisition area to obtain a matching area and matching spectral data, and associating the matching spectral data with the corresponding matching area;
[0015] A capture window is set according to the obtained matching spectrum data, and the length of the capture window is marked to obtain the window length;
[0016] According to the obtained window length, the interception window is short-cut to obtain a short window sequence;
[0017] The matching spectrum data is traversed and cycled according to the obtained short window sequence to obtain the matching spectrum coefficients.
[0018] The process of obtaining the decomposition interval coefficient includes:
[0019] Set the extraction analysis index and variation base according to the obtained capture time and matching spectrum coefficient;
[0020] Setting the variation interval according to the obtained matching spectrum coefficient;
[0021] Obtain the number of changes according to the obtained change interval and change base;
[0022] The extraction analysis index is decomposed into intervals according to the obtained number of changes to obtain the decomposition interval coefficient.
[0023] The process of obtaining the spectral characteristic coefficients includes:
[0024] Obtain a total value of coefficients according to the obtained decomposition interval coefficients, perform boundary interception on the obtained total value of coefficients, and obtain a center value of the coefficients;
[0025] Performing complex conversion on the obtained decomposition interval coefficient to obtain a complex interval coefficient;
[0026] Performing phase marking on the obtained complex interval coefficients to obtain the coefficient phase;
[0027] Obtaining a phase mean value according to the obtained coefficient phase, and obtaining a coefficient frequency shift amount according to the phase mean value and the coefficient center value;
[0028] The matching spectrum coefficients are summarized and extracted according to the obtained coefficient frequency shift to obtain the spectrum characteristic coefficients.
[0029] The process of constructing the extraction prediction space based on the acquisition area includes:
[0030] Construct the extraction prediction space and generate a virtual geographic subspace according to the set collection area;
[0031] The obtained virtual geographic subspace is uploaded to the extraction prediction space, and the corresponding capture end is marked in the virtual geographic subspace, which is recorded as the virtual matching end;
[0032] The obtained spectral feature coefficients are uploaded to the extraction prediction space and associated with the corresponding virtual matching end.
[0033] The process of generating a spectral coefficient map based on the obtained extracted spectral coefficients includes:
[0034] Performing Fourier transform on the obtained spectral characteristic coefficients to obtain frequency domain characteristic coefficients;
[0035] Set the retention frequency range according to the set filtering band and matching sampling frequency;
[0036] Setting the target filter end according to the set reserved frequency interval;
[0037] Performing characteristic filtering on the frequency domain characteristic coefficients through the set target filter end to obtain the extracted spectral coefficients;
[0038] A spectral coefficient map is generated based on the obtained extracted spectral coefficients.
[0039] The process of obtaining the virtual spectrum characteristic coefficients includes:
[0040] Set the comparison period according to the set statistical period;
[0041] Setting a change adjustment instruction, uploading the obtained change adjustment instruction to the virtual geographic subspace and executing the change adjustment instruction;
[0042] The virtual matching end is used to capture the comprehensive data of the virtual geographic subspace to obtain virtual high-quality remote sensing data, and the obtained virtual high-quality remote sensing data is spectrally screened and matched to obtain virtual matching data;
[0043] The virtual matching data is traversed and cycled according to the obtained short window sequence to obtain virtual matching spectrum coefficients, and the virtual matching spectrum coefficients are summarized and extracted according to the obtained coefficient frequency shift to obtain virtual spectrum characteristic coefficients.
[0044] The process of obtaining the evapotranspiration difference value based on the obtained extraction comparison difference, capture time and comparison period includes:
[0045] Perform Fourier transformation on the obtained virtual spectrum characteristic coefficients and perform characteristic filtering through the target filtering end to obtain virtual extraction spectrum coefficients;
[0046] Generate a sample optical system diagram according to the obtained virtual extracted spectral coefficients;
[0047] Virtually comparing the obtained sample optical system diagram with the spectral coefficient diagram associated with the virtual matching end to obtain a comprehensive optical system diagram;
[0048] The obtained comprehensive light system diagram is subjected to evaporation extraction to obtain the extraction comparison difference;
[0049] The evapotranspiration difference value was obtained based on the obtained extraction comparison difference, capture time and comparison period.
[0050] Compared with the prior art, the beneficial effects of the present invention are as follows: spectral screening is performed on the collected high-quality remote sensing data to obtain high-quality spectral data, and the extraction analysis index and the variation base are set to summarize and extract the matching spectral coefficients to obtain spectral characteristic coefficients; the collected remote sensing data are processed into a spectral characteristic coefficient form with a unified form, so that the spectral characteristic coefficients can be Fourier transformed and characteristic filtered in the constructed extraction prediction space to obtain the extracted spectral coefficients, which can be further extracted and compared with the virtual spectral characteristic coefficients obtained by virtual extraction of the virtual geographic subspace, thereby improving the accuracy of classification matching;
[0051] A comprehensive light system diagram is generated in the extraction prediction space, and evapotranspiration extraction is performed on the obtained comprehensive light system diagram to obtain the extraction comparison difference, and the evapotranspiration difference value is obtained according to the obtained extraction comparison difference, capture time and comparison period; by virtually adjusting the collection area in the virtual space, it is convenient to observe the change of evapotranspiration in the collection area and increase the extraction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0053] Figure 1 It is a schematic diagram of the present invention. DETAILED DESCRIPTION
[0054] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0055] like Figure 1 As shown, a method for extracting evapotranspiration based on high temporal and spatial resolution remote sensing data includes the following steps:
[0056] Step S1: Set a collection area, build a capture terminal according to the set collection area, capture the comprehensive data of the collection area through the capture terminal, and obtain high-quality remote sensing data and capture time;
[0057] Step S2: setting a screening band to perform spectral screening on high-quality remote sensing data to obtain high-quality spectral data, setting a capture window to traverse and cycle the high-quality spectral data to obtain matching spectral coefficients, setting an extraction analysis index and a variable base, performing interval decomposition on the extraction analysis index through the variable base, and obtaining a decomposition interval coefficient;
[0058] Step S3: performing amplitude-frequency acquisition on the obtained decomposition interval coefficients to obtain the coefficient frequency shift, summarizing and extracting the matching spectrum coefficients according to the obtained coefficient frequency shift to obtain the spectrum feature coefficients, and constructing the extraction prediction space according to the acquisition area;
[0059] Step S4: Perform Fourier transform on the spectral characteristic coefficients by extracting the prediction space to obtain the frequency domain characteristic coefficients, perform characteristic filtering on the frequency domain characteristic coefficients by setting the target filter end to obtain the extracted spectral coefficients, generate a spectral coefficient map according to the obtained extracted spectral coefficients, set the comparison period to perform virtual extraction on the obtained extraction prediction space to obtain the virtual spectral characteristic coefficients, perform characteristic filtering and virtual comparison on the obtained virtual spectral characteristic coefficients to obtain the extracted comparison difference, and obtain the evaporation difference value according to the obtained extraction comparison difference, capture time and comparison period.
[0060] It should be further explained that, in the specific implementation process, the process of collecting high-quality remote sensing data includes:
[0061] Set the collection area, build a capture terminal according to the set collection area, capture the comprehensive data of the collection area through the capture terminal, and obtain high-quality remote sensing data;
[0062] The time of capturing high-quality remote sensing data is recorded as the capture time;
[0063] It should be further explained that, in the specific implementation process, the collection area includes all types of surfaces such as farmland, grassland, forest, lakes, rivers, etc.; the high-quality remote sensing data represents high temporal and spatial resolution remote sensing data of the surface, including surface cover type, vegetation conditions, soil moisture, and surface temperature.
[0064] Set the statistical period according to the obtained capture time;
[0065] Setting a screening band, wherein the screening band is used to intercept a target spectral band in high-quality remote sensing data;
[0066] Performing spectral screening on the obtained high-quality remote sensing data according to the set screening band to obtain high-quality spectral data, wherein the high-quality spectral data includes a surface evaporation spectrum and a vegetation transpiration spectrum, and associating the obtained high-quality spectral data with a corresponding capture time;
[0067] It should be further explained that, in the specific implementation process, the spectral screening is performed by setting a screening band to screen the high-quality remote sensing data, that is, different substances show corresponding spectral responses in different bands, and high-quality spectral data corresponding to the screening band is obtained;
[0068] In particular, the evapotranspiration is composed of surface evaporation and vegetation transpiration. Therefore, it is necessary to screen the surface evaporation spectrum and vegetation transpiration spectrum in high-quality remote sensing data; for example, for the surface evaporation spectrum, the screening bands are blue, green, red and near-infrared bands, reflecting the surface radiation balance, water vapor and carbon dioxide concentration; for the vegetation transpiration spectrum, the screening bands are green, red and near-infrared bands, reflecting the growth status, water absorption and evaporation characteristics of vegetation;
[0069] Matching the obtained high-quality spectral data with the corresponding acquisition area, marking the successfully matched acquisition area as the matching area, marking the successfully matched high-quality spectral data as the matching spectral data, and associating the matching spectral data with the corresponding matching area;
[0070] Performing frequency marking on the obtained matching spectrum data to obtain matching spectrum sampling points and matching sampling frequencies;
[0071] A capture window is set according to the obtained matching spectrum data, and the length of the capture window is marked to obtain the window length;
[0072] Further, the interception window is a function sequence determined according to the matching spectrum range; the matching spectrum range represents the range from the lowest effective frequency to the highest effective frequency, and the length of the interception window is related to the matching spectrum range, the larger the matching spectrum range, the larger the length of the interception window, and the smaller the matching spectrum range, the smaller the length of the interception window;
[0073] According to the obtained window length, the interception window is short-cut to obtain a short window sequence;
[0074] Furthermore, the short truncation means dividing the window length by e to obtain a shorter short window sequence;
[0075] According to the obtained short window sequence, the matching spectrum data is traversed and cycled to obtain the matching spectrum coefficients;
[0076] Further, the traversal loop represents multiplying the short window sequence with the matching spectrum sampling points corresponding to the matching spectrum data to obtain matching spectrum coefficients;
[0077] Set the extraction analysis index according to the obtained capture time and matching spectrum coefficient;
[0078] Performing time-frequency transformation on the obtained extraction analysis index to obtain a time extraction axis and a frequency extraction axis;
[0079] Furthermore, the time-frequency transformation is to decompose the extraction analysis index in the time dimension and the frequency dimension to obtain the time extraction axis and the frequency extraction axis;
[0080] Setting a variation base, wherein the variation base is used to control the parameters of the expansion and contraction and translation of the extraction analysis index on the time extraction axis or the frequency extraction axis;
[0081] The variation interval is set according to the obtained matching spectrum coefficient, and the obtained variation interval is marked as Δj;
[0082] Furthermore, the variation interval represents the interval between two adjacent variation bases;
[0083] The number of changes is obtained according to the obtained change interval and change base, and the obtained number of changes is marked as g, where: L represents the length of the matching spectrum coefficient, a represents the number of change intervals, B represents the change base, and β represents the change factor;
[0084] Perform interval decomposition on the extraction analysis index according to the obtained number of changes to obtain a decomposition interval coefficient;
[0085] Furthermore, the interval decomposition is to equally decompose the extraction analysis index according to the number of changes to obtain decomposition interval coefficients of equal length.
[0086] The obtained decomposition interval coefficients are subjected to amplitude and frequency acquisition to obtain the coefficient frequency shift amount;
[0087] It should be further explained that, in the specific implementation process, the amplitude-frequency acquisition process includes:
[0088] Obtain a total coefficient value according to the obtained decomposition interval coefficients, wherein the total coefficient value is the sum of the absolute values of each decomposition spectrum coefficient;
[0089] The total value of the obtained coefficients is intercepted by boundaries to obtain the center value of the coefficients;
[0090] The boundary interception means dividing the total value of the coefficients obtained by the number of decomposition interval coefficients to obtain the center value of the coefficients, and marking the obtained center value of the coefficients as z;
[0091] Performing complex conversion on the obtained decomposition interval coefficient to obtain a complex interval coefficient;
[0092] Furthermore, the complex number conversion means converting the decomposed interval coefficient into a complex number form, the complex number interval coefficient includes a complex real part and a complex imaginary part, the obtained complex real part is marked as k, and the obtained complex imaginary part is marked as m;
[0093] Performing phase marking on the obtained complex interval coefficients to obtain the coefficient phase;
[0094] The phase mark indicates that the coefficient phase is obtained according to the obtained complex real part and the complex imaginary part, and the obtained coefficient phase is marked as x, wherein,
[0095] Obtain a phase sum according to the obtained coefficient phase, and obtain a phase mean according to the obtained phase sum;
[0096] Furthermore, the phase sum is the sum of the coefficient phases of the decomposition interval coefficients, and is divided by the total number of the decomposition interval coefficients to obtain a phase mean;
[0097] Obtain the coefficient frequency shift amount according to the obtained phase mean value and the coefficient center value;
[0098] The obtained coefficient frequency shift amount is marked as PX, where PX = (z*α1)*(x*α2), α1 and α2 represent combined coefficients, and 0<α1<1, 0<α2<1;
[0099] According to the obtained coefficient frequency shift amount, the matching spectrum coefficients are summarized and extracted to obtain the spectrum characteristic coefficients;
[0100] It should be further explained that, in the specific implementation process, the inductive extraction means performing a convolution operation on the coefficient frequency shift amount and the matching spectrum sampling point corresponding to the matching spectrum coefficient until all the matching spectrum sampling points of the matching spectrum coefficient complete the convolution operation, and sum them up to obtain the spectrum characteristic coefficient;
[0101] Furthermore, according to the surface evaporation spectrum and the vegetation transpiration spectrum included in the high-quality spectral data, after processing the high-quality spectral data, the spectral characteristic coefficients obtained include the surface evaporation characteristic coefficient and the vegetation transpiration characteristic coefficient;
[0102] Construct the extraction prediction space and generate a virtual geographic subspace according to the set collection area;
[0103] Furthermore, the virtual geographic subspace is generated according to the actual acquisition area, including the geographic features of the acquisition area and the generated regional data, and one acquisition area corresponds to one virtual geographic subspace;
[0104] The obtained virtual geographic subspace is uploaded to the extraction prediction space, and the corresponding capture end is marked in the virtual geographic subspace, which is recorded as the virtual matching end;
[0105] In particular, the locations of the virtual geographic subspace uploaded into the extraction prediction space correspond to the same locations of the acquisition area in reality;
[0106] The obtained spectral feature coefficients are uploaded to the extraction prediction space and associated with the corresponding virtual matching end.
[0107] Performing Fourier transform on the obtained spectral characteristic coefficients to obtain frequency domain characteristic coefficients;
[0108] Setting a reserved frequency interval according to the set screening band and matching sampling frequency, wherein the reserved frequency interval includes a low cutoff frequency and a high cutoff frequency;
[0109] Further, the reserved frequency interval is the lowest frequency and the highest frequency allowed to pass, and the lowest frequency allowed to pass is recorded as the low cutoff frequency, and the highest frequency allowed to pass is recorded as the high cutoff frequency;
[0110] Setting the target filter end according to the set reserved frequency interval;
[0111] It should be further explained that, in the specific implementation process, the target filter end is a filter that performs characteristic filtering on the frequency domain characteristic coefficients, and the set target filter includes filter parameters, such as the order of the filter, passband and stopband fluctuations; further, the frequency response corresponding to the target filter end is designed according to the filter parameters included in the target filter end;
[0112] Performing characteristic filtering on the frequency domain characteristic coefficients through the set target filter end to obtain the extracted spectral coefficients;
[0113] Furthermore, the characteristic filtering is to pass the obtained frequency domain characteristic coefficient through the target filtering end, multiply the obtained frequency response by the frequency domain characteristic coefficient, and obtain the extracted spectrum coefficient;
[0114] generating a spectral coefficient map according to the obtained extracted spectral coefficients;
[0115] Further, the spectral coefficient map is expressed in the form of a spectrum map, and according to the surface evaporation characteristic coefficient and the vegetation transpiration characteristic coefficient included in the spectral characteristic coefficient, the extracted spectral coefficient obtained also includes the evaporation spectral coefficient and the transpiration spectral coefficient; in particular, the spectral coefficient map generated according to the extracted spectral coefficient also includes the evaporation coefficient map and the transpiration coefficient map;
[0116] Setting a comparison period according to the set statistical period, wherein the comparison period is another statistical period that differs from the statistical period by a certain capture time;
[0117] Performing virtual extraction on the obtained virtual geographic subspace based on the comparison period to obtain an extraction comparison difference;
[0118] It should be further explained that, in the specific implementation process, the virtual extraction process includes:
[0119] Setting a change adjustment instruction, uploading the obtained change adjustment instruction to the virtual geographic subspace and executing the change adjustment instruction;
[0120] Further, the change adjustment instruction indicates changing the surface water content and the vegetation water content by adjusting the virtual geographic subspace so as to extract the evapotranspiration of the collection area;
[0121] The virtual matching end is used to capture the comprehensive data of the virtual geographic subspace to obtain virtual high-quality remote sensing data;
[0122] Perform spectral screening and matching on the obtained virtual high-quality remote sensing data to obtain virtual matching data;
[0123] The virtual matching data is traversed and cycled according to the obtained short window sequence to obtain the virtual matching spectrum coefficient;
[0124] The virtual matching spectrum coefficients are summarized and extracted according to the obtained coefficient frequency shift amount to obtain the virtual spectrum characteristic coefficients;
[0125] Perform Fourier transformation on the obtained virtual spectrum characteristic coefficients and perform characteristic filtering through the target filtering end to obtain virtual extraction spectrum coefficients;
[0126] Generate a sample optical system diagram according to the obtained virtual extracted spectral coefficients;
[0127] Furthermore, the sample optical system diagram is also presented in the form of a spectrum diagram, and the capture time of corresponding positions in the comparison period is the same;
[0128] Virtually comparing the obtained sample optical system diagram with the spectral coefficient diagram associated with the virtual matching end to obtain a comprehensive optical system diagram;
[0129] Mark the capture time in the comprehensive optical system diagram to obtain a virtual node, and mark the obtained virtual node as t, where t=1, 2, 3, ..., u1, and u1 is a positive integer;
[0130] Furthermore, the virtual comparison is based on the spectral coefficient map, and the sample optical system map is mapped to the spectral coefficient map to obtain the comprehensive optical system map; in particular, the mapping order is performed according to the time order of the capture time, and the curve mapped by the sample optical system map in the comprehensive optical system map is marked as the sample spectral line, and the curve of the spectral coefficient map in the comprehensive optical system map is marked as the initial spectral line, and the initial spectral line includes the evaporation initial spectral line and the transpiration initial spectral line;
[0131] The obtained comprehensive light system diagram is subjected to evaporation extraction to obtain the extraction comparison difference;
[0132] Further, the evaporation extraction process includes:
[0133] The sample spectrum line in the comprehensive light system diagram is marked as y(t), and the initial light spectrum in the comprehensive light system diagram is marked as s(t);
[0134] Integrate the sample spectrum line and the initial light spectrum in the comprehensive light system diagram to obtain the extracted comparison difference, and mark the obtained extracted comparison difference as Q(t), where Q(t)=∫∣y(t)-s(t)∣dt;
[0135] The evapotranspiration difference value is obtained according to the obtained extraction comparison difference, capture time and comparison period;
[0136] The obtained evapotranspiration difference value is marked as ΔH, where F 比 Indicates the last capture time of the comparison cycle, F 统 Indicates the first capture time of the statistical period;
[0137] In particular, γ1 represents the equalization factor of the initial spectral line of evaporation, γ2 represents the equalization factor of the initial spectral line of transpiration; V1 大 Indicates the daily average wind speed, obtained based on the comprehensive data of the collection area, P1 大represents the saturated vapor pressure, P1 represents the vapor pressure of the current virtual node, both of which are obtained based on the comprehensive data of the collection area; W represents the number of days for plant growth.
[0138] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.
Claims
1. A method for extracting evapotranspiration based on high temporal and spatial resolution remote sensing data, characterized in that: The following steps are involved: Step S1: Set a collection area, build a capture terminal according to the set collection area, capture the comprehensive data of the collection area through the capture terminal, and obtain high-quality remote sensing data and capture time; Step S2: setting a screening band to perform spectral screening on high-quality remote sensing data to obtain high-quality spectral data, setting a capture window to traverse and cycle the high-quality spectral data to obtain matching spectral coefficients, setting an extraction analysis index and a variable base, performing interval decomposition on the extraction analysis index through the variable base, and obtaining a decomposition interval coefficient; Step S3: performing amplitude-frequency acquisition on the obtained decomposition interval coefficients to obtain the coefficient frequency shift, summarizing and extracting the matching spectrum coefficients according to the obtained coefficient frequency shift to obtain the spectrum feature coefficients, and constructing the extraction prediction space according to the acquisition area; Step S4: performing Fourier transform on the spectral characteristic coefficients by extracting the prediction space to obtain the frequency domain characteristic coefficients, performing characteristic filtering on the frequency domain characteristic coefficients by setting the target filter end to obtain the extracted spectral coefficients, generating a spectral coefficient map according to the obtained extracted spectral coefficients, setting a comparison period to perform virtual extraction on the obtained extraction prediction space to obtain virtual spectral characteristic coefficients, performing characteristic filtering and virtual comparison on the obtained virtual spectral characteristic coefficients to obtain extraction comparison difference, and obtaining the evaporation difference value according to the obtained extraction comparison difference, capture time and comparison period; The process of obtaining matching spectral coefficients includes: Set the statistical period according to the obtained capture time; Set the screening band to perform spectral screening on the acquired high-quality remote sensing data to obtain high-quality spectral data; Matching the obtained high-quality spectral data with the corresponding acquisition area to obtain a matching area and matching spectral data, and associating the matching spectral data with the corresponding matching area; A capture window is set according to the obtained matching spectrum data, and the length of the capture window is marked to obtain the window length; According to the obtained window length, the interception window is short-cut to obtain a short window sequence; According to the obtained short window sequence, the matching spectrum data is traversed and cycled to obtain the matching spectrum coefficients; The process of obtaining the decomposition interval coefficient includes: Set the extraction analysis index and variation base according to the obtained capture time and matching spectrum coefficient; Setting the variation interval according to the obtained matching spectrum coefficient; Obtain the number of changes according to the obtained change interval and change base; Perform interval decomposition on the extraction analysis index according to the obtained number of changes to obtain a decomposition interval coefficient; The process of obtaining the spectral characteristic coefficients includes: Obtain a total value of coefficients according to the obtained decomposition interval coefficients, perform boundary interception on the obtained total value of coefficients, and obtain a center value of the coefficients; Performing complex conversion on the obtained decomposition interval coefficient to obtain a complex interval coefficient; Performing phase marking on the obtained complex interval coefficients to obtain the coefficient phase; Obtaining a phase mean value according to the obtained coefficient phase, and obtaining a coefficient frequency shift amount according to the phase mean value and the coefficient center value; According to the obtained coefficient frequency shift amount, the matching spectrum coefficients are summarized and extracted to obtain the spectrum characteristic coefficients; The process of constructing the extraction prediction space based on the acquisition area includes: Construct the extraction prediction space and generate a virtual geographic subspace according to the set collection area; The obtained virtual geographic subspace is uploaded to the extraction prediction space, and the corresponding capture end is marked in the virtual geographic subspace, which is recorded as the virtual matching end; The obtained spectral characteristic coefficients are uploaded to the extraction prediction space and associated with the corresponding virtual matching end; The process of generating a spectral coefficient map based on the obtained extracted spectral coefficients includes: Performing Fourier transform on the obtained spectral characteristic coefficients to obtain frequency domain characteristic coefficients; Set the retention frequency range according to the set filtering band and matching sampling frequency; Setting the target filter end according to the set reserved frequency interval; Performing characteristic filtering on the frequency domain characteristic coefficients through the set target filter end to obtain the extracted spectral coefficients; generating a spectral coefficient map according to the obtained extracted spectral coefficients; The process of obtaining the virtual spectrum characteristic coefficients includes: Set the comparison period according to the set statistical period; Setting a change adjustment instruction, uploading the obtained change adjustment instruction to the virtual geographic subspace and executing the change adjustment instruction; The virtual matching end is used to capture the comprehensive data of the virtual geographic subspace to obtain virtual high-quality remote sensing data, and the obtained virtual high-quality remote sensing data is spectrally screened and matched to obtain virtual matching data; The virtual matching data is traversed and cycled according to the obtained short window sequence to obtain virtual matching spectrum coefficients, and the virtual matching spectrum coefficients are summarized and extracted according to the obtained coefficient frequency shift amount to obtain virtual spectrum characteristic coefficients; The process of obtaining the evapotranspiration difference value based on the obtained extraction comparison difference, capture time and comparison period includes: Perform Fourier transformation on the obtained virtual spectrum characteristic coefficients and perform characteristic filtering through the target filter end to obtain virtual extraction spectrum coefficients; Generate a sample optical system diagram according to the obtained virtual extracted spectral coefficients; Virtually comparing the obtained sample optical system diagram with the spectral coefficient diagram associated with the virtual matching end to obtain a comprehensive optical system diagram; The obtained comprehensive light system diagram is subjected to evaporation extraction to obtain the extraction comparison difference; The evapotranspiration difference value was obtained based on the obtained extraction comparison difference, capture time and comparison period.
Citation Information
Patent Citations
Grassland above-ground biomass retrievalinversion method based on high-time resolution and high-spatial resolution multispectral remote sensing data
CN108152212A
Remote sensing image enhancement method
CN111583140A
