Method and system for improving the accuracy of remote sensing evapotranspiration

Abstract

The present application relates to a kind of method and system for improving the precision of remote sensing evapotranspiration, comprising the following steps: according to two years of LAI data measured in field, combining sentinel image to retrieve crop growth period vegetation red light edge index CI red‑edge , monitor the two-stage precision retrieval LAI model parameters of crop are calibrated and verified;Based on the two-stage precision retrieval LAI model established, using Landsat8 remote sensing image and sentinel image data as the driving data of typical remote sensing dual-source evapotranspiration model, the surface temperature obtained by remote sensing retrieval is divided into components, and canopy temperature and soil temperature are obtained;Based on the results of LAI precision retrieval and component temperature division, the typical remote sensing dual-source evapotranspiration model is used to estimate crop evapotranspiration and its components in irrigation area.The present application can be widely applied in the field of water-saving irrigation technology.

Description

Technical Field

[0001] The present invention relates to a method and system for improving the accuracy of remote sensing evapotranspiration based on two-stage precise inversion of leaf area index, which can provide technical support for rapid and accurate estimation of regional remote sensing evapotranspiration (ET) and belongs to the technical field of water-saving irrigation in irrigation areas. Background Art

[0002] With agriculture accounting for 76% of my country's total water use in 2022, research on efficient water-saving irrigation is crucial for promoting a water-saving society. Accurate assessment of crop water consumption in irrigated areas is a key component of this process.

[0003] The remote sensing dual-source evapotranspiration model can simultaneously estimate crop evapotranspiration (ET), soil evaporation (E) and vegetation transpiration (T). Canopy temperature (Tc) and soil temperature (Ts) are the key parameters driving the dual-source ET model. However, the land surface temperature (LST) of the irrigated area obtained by remote sensing image inversion represents a mixed result of canopy temperature and soil temperature. Therefore, how to accurately divide LST into Tc and Ts is a key step in estimating evapotranspiration using the dual-source ET model. According to different division principles, it can be divided into two typical methods: PT iteration method and trapezoidal space method. The representative models are Two-Source Energy Balance (TSEB) and Hybrid Dual-Source Scheme and Trapezoid Framework-Based Evapotranspiration (HTEM) model. The former updates the initial parameter α by continuous iteration. PT The final solution is energy balance, while the latter establishes a trapezoidal space between vegetation index and surface temperature. By solving the four key end-member temperatures, the temperature partitioning (Tc and Ts) and ET estimation are achieved. Therefore, the accuracy of the temperature partitioning affects the degree of improvement in remote sensing evapotranspiration estimation.

[0004] As a key input factor of the remote sensing dual-source ET model, the crop leaf area index (LAI) affects the calculation of vegetation cover and the distribution of available energy, thereby affecting the temperature division and ET estimation. Therefore, the precise remote sensing inversion of LAI is of great significance for improving the accuracy of evapotranspiration. One of the widely used methods is to accurately invert LAI based on the linear relationship between vegetation index and LAI. Among various vegetation indices, the vegetation red edge index (CI red-edge ) is often used to establish the correlation equation between NDVI and LAI to estimate LAI because it overcomes the saturation phenomenon of NDVI. red-edgeSince LAI may have a lag in time series changes, simple linear models may ignore the impact of lag on LAI estimation, which in turn affects the temperature partitioning and ET estimation accuracy. Therefore, it is necessary to consider these factors and propose a reasonable method to determine the precise equation for remote sensing inversion of LAI to improve the accuracy of temperature partitioning and ultimately improve the accuracy of ET estimation. Summary of the Invention

[0005] In view of the above problems, the purpose of the present invention is to provide a method and system for improving the accuracy of remote sensing evapotranspiration based on two-stage precise inversion of leaf area index. This method uses the field measured crop LAI and the vegetation red edge index (CI red-edge ) information, a two-stage LAI precision inversion model was established, and a typical dual-source ET model (TSEB and HTEM) was used as an example to achieve accurate division of remote sensing surface temperature components, thereby improving the estimation accuracy of evapotranspiration.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a method for improving the accuracy of remote sensing evapotranspiration, comprising the following steps:

[0008] Based on two years of field-measured LAI data, combined with Sentinel images, the vegetation red edge index CI during the crop growth period was inverted. red-edge , calibrate and validate the parameters of the two-stage precision inversion LAI model for monitoring crops;

[0009] Based on the established two-stage precise LAI inversion model, Landsat8 remote sensing imagery and Sentinel imagery data were used as driving data for a typical remote sensing dual-source evapotranspiration model. The surface temperature obtained by remote sensing inversion was divided into component temperatures to obtain canopy temperature and soil temperature.

[0010] Based on the LAI precise estimation and temperature component classification results, the evapotranspiration of crops in irrigated areas and its components were estimated using a typical remote sensing dual-source evapotranspiration model.

[0011] Furthermore, the LAI data measured in the field for two years was combined with the sentinel image to invert the vegetation red light edge index CI during the crop growth period. red-edge , calibrate and validate the parameters of the two-stage precision inversion LAI model for crop monitoring, including:

[0012] Select typical crop plots in the irrigation area as monitoring points, manually collect LAI data of the monitored crops regularly for two consecutive years, and record the corresponding collection dates;

[0013] Using the GEE platform, we analyzed the available sentinel images during the two-year crop growth period in the irrigation area according to the judgment principle of cloud cover <20%, and used band operation inversion to obtain the vegetation red edge index CI during the crop growth period. red-edge ;

[0014] CI based on the first year's measured LAI data and remote sensing inversion red-edge The data were combined with the corresponding crop day sequence, and the measured values ​​were fitted using the modified Logistic model to obtain the LAI and CI of the monitored crops during their entire growth period. red-edge Dynamic curve;

[0015] Based on the LAI and CI of the monitored crops during their entire growth period red-edge Dynamic curve, determine LAI and CI red-edge There is a lag between the two, and based on this, the whole growth period of the monitored crops is divided into a rapid growth period and a gradual withering period;

[0016] The LAI values ​​measured during the first year's rapid growth and gradual withering periods and the CI obtained by remote sensing were used. red-edge , respectively calibrate the parameters of the two-stage precise inversion LAI model for monitored crops;

[0017] CI of monitored crops obtained using remote sensing in the second year red-edge The data were combined with the day sequence number and the modified Logisic model was used to fit the CI corresponding to the LAI measurement date. red-edge value;

[0018] The CI obtained by fitting red-edge The values ​​were substituted into the calibrated two-stage precise inversion LAI model to estimate the LAI of the monitored crops, and the accuracy of the estimated LAI values ​​was compared with the measured values ​​in the second year to verify the parameters of the established two-stage precise LAI estimation model.

[0019] Furthermore, when selecting monitored crops in the irrigation area, corresponding crop types are selected according to different photosynthesis pathways.

[0020] Furthermore, based on the established two-stage precise inversion LAI model, Landsat8 remote sensing imagery and Sentinel imagery data are used as driving data for a typical remote sensing dual-source evapotranspiration model. The surface temperature obtained by remote sensing inversion is divided into components to obtain canopy temperature and soil temperature, including:

[0021] On the GEE platform, we analyzed available Landsat 8 remote sensing images based on the screening principle of cloud cover <20%. Surface temperature, normalized difference vegetation index, albedo, and solar zenith angle were obtained through remote sensing inversion. These data served as the basic driving dataset for a typical remote sensing dual-source evapotranspiration model.

[0022] Adjust and correct key parameters in the HTEM model;

[0023] According to the two-stage precise inversion LAI model after calibration and verification, the CI obtained by remote sensing is used red-edge Estimate the LAI of the irrigated area and input the LAI of the irrigated area and the basic driving dataset into a typical remote sensing dual-source evapotranspiration model, dividing the surface temperature into canopy temperature and soil temperature;

[0024] The CTMS-Online crop canopy temperature and environmental factor measurement system is deployed in the crop monitoring plots in the irrigation area to monitor the surface temperature in real time, and the actual canopy temperature and soil temperature are obtained based on the precision identification method;

[0025] The canopy temperature and soil temperature estimated by a typical remote sensing dual-source evapotranspiration model were compared with the measured soil temperature and canopy temperature to determine the accuracy of the component temperature division.

[0026] Furthermore, the adjustment and correction of key parameters in the HTEM model include:

[0027] Based on the measured plant height, the logistic model is used to simulate the plant height curve of the crop during the entire growth period and determine the maximum plant height h of the crop. c_max ; Among them, the simulation formula is:

[0028] h=E / (1+F×exp(-G×DOY)

[0029] Where h is the plant height of the crop, E, F and G are empirical coefficients obtained by fitting the measured plant height of the crop using Origin;

[0030] The PT formula is combined with the energy balance equation to obtain the canopy temperature during crop potential transpiration as T c_min , the formula is:

[0031]

[0032] H c =R nc -LE c

[0033]

[0034] Where, LE c is the canopy latent heat flux, R nc is the canopy net radiation, α PT Take it as 1.3, Δ is the slope of the saturated water vapor pressure and temperature curve, γ is the hygrometer constant, H c is the sensible heat flux, is the aerodynamic impedance of heat transfer between the canopy and the reference height, ρ is the air density, C pis the specific heat of air.

[0035] Furthermore, the aforementioned method, based on the component temperature division results, uses a typical remote sensing dual-source evapotranspiration model to estimate evapotranspiration and its components in the irrigated area, including:

[0036] Using a typical remote sensing dual-source evapotranspiration model combined with remote sensing images to obtain instantaneous evapotranspiration of monitored crops, soil evaporation, and vegetation transpiration;

[0037] The instantaneous evapotranspiration of the monitored crops was converted into daily values ​​using the evaporation ratio method;

[0038] Cubic spline interpolation method is used to interpolate daily crop evapotranspiration, soil evaporation and vegetation transpiration;

[0039] Based on daily evapotranspiration data, monthly and annual crop evapotranspiration and its components can be obtained;

[0040] Combined with remote sensing inversion, crop planting distribution is obtained, and evapotranspiration and its component distribution maps of monitored crops at different time scales are extracted.

[0041] Furthermore, the method of combining remote sensing inversion to obtain crop planting distribution includes: in the early stage of crop growth, using the global satellite positioning system to conduct a planting structure survey of the irrigation area, recording the latitude and longitude of each point and the crops planted, combining the sentinel data obtained from the GEE platform, and using the random forest method to classify the planting structure to obtain a planting distribution map of the monitored crops.

[0042] In a second aspect, the present invention provides a system for improving the accuracy of remote sensing evapotranspiration, comprising:

[0043] The precision inversion module is used to invert the vegetation red edge index CI during the crop growth period based on two years of field measured LAI data and combined with sentinel images. red-edge , calibrate and validate the parameters of the two-stage precision inversion LAI model for monitoring crops;

[0044] The component temperature partitioning module is used to perform component temperature partitioning on the surface temperature obtained by remote sensing inversion based on the established two-stage precise inversion LAI model. It uses Landsat8 remote sensing imagery and Sentinel imagery data as driving data for a typical remote sensing dual-source evapotranspiration model to obtain canopy temperature and soil temperature.

[0045] The evapotranspiration precision estimation module is used to estimate the evapotranspiration of irrigated crops and its components based on the LAI precision estimation and component temperature division results using a typical remote sensing dual-source evapotranspiration model.

[0046] In a third aspect, the present invention provides a computer-readable storage medium storing one or more programs, wherein the one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any method.

[0047] In a fourth aspect, the present invention provides a computing device comprising: one or more processors and a memory, wherein the memory stores one or more programs and is configured to be executed by the one or more processors, and the one or more programs include instructions for executing any method.

[0048] The present invention has the following advantages due to the adoption of the above technical solution:

[0049] 1. By considering LAI and CI red-edge In order to improve the LAI estimation accuracy, a LAI precision inversion model was established in two stages.

[0050] 2. By establishing different forms of two-stage LAI models, the different growth characteristics of C3 and C4 crops are effectively considered, and the universality of the LAI precision inversion model is enhanced;

[0051] 3. The two-stage precise LAI inversion effectively improves the accuracy of surface temperature component division from remote sensing inversion;

[0052] 4. The two-stage precise inversion of the driving parameters of the typical dual-source ET model for LAI crops effectively improved the estimation accuracy of evapotranspiration and its components in the irrigation area, providing theoretical support for the precise estimation of water consumption in the irrigation area and irrigation decision-making management. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. Throughout the drawings, the same reference numerals are used to denote the same components. In the drawings:

[0054] Figure 1 Flowchart of the method for improving remote sensing evapotranspiration accuracy based on two-stage precise LAI inversion provided by the present invention;

[0055] Figure 2 The two-stage precise inversion LAI model parameter calibration flow chart provided by the present invention;

[0056] Figure 3 A flow chart for the classification of remotely sensed surface temperature components provided by the present invention;

[0057] Figure 4 A flow chart for remote sensing evapotranspiration estimation provided by the present invention;

[0058] Figure 5 Schematic diagram of the system for improving the accuracy of remote sensing evapotranspiration based on two-stage precise inversion of LAI provided by the present invention. DETAILED DESCRIPTION

[0059] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field are within the scope of protection of the present invention.

[0060] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0061] In some embodiments of the present invention, a method for improving the accuracy of remote sensing evapotranspiration based on two-stage precise inversion of leaf area index is provided. This method is based on the different variation characteristics of the leaf area index of typical crops sunflower (C3) and corn (C4), utilizes the variation relationship between the leaf area index and the vegetation red light edge index, considers the hysteresis between the two, constructs a two-stage leaf area index precise inversion model, and explores the improvement of the accuracy of component temperature partitioning (Tc and Ts) by precise estimation of the leaf area index. The method then evaluates the degree of improvement in the accuracy of remote sensing evapotranspiration estimation based on the improvement of the precision of the leaf area index and component temperature partitioning, providing a theoretical basis and technical support for the study of crop water consumption in irrigated areas and agricultural production decision-making and management.

[0062] Correspondingly, in other embodiments of the present invention, a system, device and medium are provided for improving the accuracy of remote sensing evapotranspiration based on two-stage precise inversion of leaf area index.

[0063] Example 1

[0064] like Figure 1 As shown, this embodiment provides a method for improving the accuracy of remote sensing evapotranspiration based on two-stage precise inversion of leaf area index, which includes the following steps:

[0065] 1) Based on two years of field-measured LAI data, combined with Sentinel images, the vegetation red edge index (CI) during the crop growth period was inverted. red-edge , calibrate and validate the parameters of the two-stage precision inversion LAI model for monitoring crops;

[0066] 2) Based on the established two-stage precise LAI inversion model, Landsat8 remote sensing imagery and Sentinel imagery data were used as driving data for a typical remote sensing dual-source evapotranspiration model (TSEB and HTEM). The land surface temperature (LST) obtained by remote sensing inversion was divided into components to obtain canopy temperature (Tc) and soil temperature (Ts).

[0067] 3) Based on the LAI precise estimation and temperature classification results, the typical remote sensing dual-source evapotranspiration model (TSEB and HTEM) was used to estimate ET and its components (E and T) in the irrigation area.

[0068] Preferably, Figure 2 As shown, the above step 1) specifically includes:

[0069] 1.1) Select typical crop plots in the irrigation area as monitoring sites, manually collect LAI data of the monitored crops regularly for two consecutive years, and record the corresponding collection dates.

[0070] 1.2) Using the Google Earth Engine (GEE) platform, we analyzed available sentinel images during the two-year crop growth period in the irrigation area, using the cloud cover <20% judgment principle. We then used band calculations to invert the vegetation red edge index (CI) during the crop growth period. red-edge , the calculation formula is:

[0071]

[0072] Where R 783 and R 705 They represent the reflectance at wavelengths of 783 nm and 705 nm in the Sentinel image, respectively.

[0073] 1.3) CI based on the first year's measured LAI data and remote sensing inversion red-edge The data were combined with the corresponding crop day sequence, and the measured values ​​were fitted using the modified Logistic model to obtain the LAI and CI of the monitored crops during their entire growth period. red-edge Dynamic curve.

[0074] Among them, the calculation formula of the modified Logistic model is:

[0075] y=A / (1+exp(B×DOY 2 +C×DOY+D))

[0076] Where A, B, C and D are empirical parameters that can be obtained by fitting the Origin model based on measured data; DOY is the crop day number; y is LAI or CI red-edge .

[0077] 1.4) Based on the LAI and CI of the monitored crops during their entire growth period obtained in step 1.3) red-edge Dynamic curve, determine LAI and CI red-edge There is a lag between the two, and based on this, the entire growth period of the monitored crops is divided into a rapid growth period and a gradual withering period.

[0078] 1.5) Using the measured LAI values ​​and remote sensing CI values ​​during the first year of rapid growth and gradual withering red-edge , and calibrate the parameters of the two-stage precise inversion LAI model for monitoring crops.

[0079] 1.6) Using CI of monitored crops obtained by remote sensing in the second year red-edge The data were combined with the day sequence number and the modified Logisic model was used to fit the CI corresponding to the LAI measurement date. red-edge value.

[0080] 1.7) The CI obtained by fitting red-edge The values ​​were substituted into the two-stage precise inversion LAI model calibrated in step 1.5) to estimate the LAI of the monitored crops, and the LAI estimates were compared with the measured values ​​in the second year to verify the parameters of the established two-stage precise LAI estimation model.

[0081] In particular, if there are sufficient monitoring points, it is also possible to use only one year's worth of data, divided into two groups, for modeling and validation. In this example, the first year's data was used for modeling, and the second year's data was used for model validation. This is primarily because: in actual field trials, the number of monitoring points is often limited due to regional, labor-intensive, and time constraints. If the data were divided into two groups, the amount of data would often be very small. The LAI model involved estimates LAI throughout the crop's growth period, so data from that entire growth period must be used for modeling. Verification using the second year's data demonstrates the model's interannual applicability.

[0082] Preferably, in step 1.1), when selecting the monitored crops, different crop types can be selected according to different photosynthesis pathways. In this embodiment, sunflower and corn are selected as representatives of C3 and C4 crops, respectively.

[0083] Preferably, Figure 3 As shown, the above step 2) specifically includes:

[0084] 2.1) Based on the GEE platform, available Landsat 8 remote sensing images were analyzed according to the cloud cover screening principle of <20%. Land surface temperature (LST), normalized difference vegetation index (NDVI), albedo, and solar zenith angle (SZA) were obtained by remote sensing inversion. These data serve as the basic driving datasets for a typical remote sensing dual-source evapotranspiration model (TSEB and HTEM).

[0085] 2.2) Adjust and correct the key parameters in the HTEM model, specifically including two parameters:

[0086] ①h c_max :h c_max In the original HTEM model, it is taken as a constant value of 1m, which represents T c_max The plant height at the point (the plant height of the crop when it is completely covered by vegetation under drought conditions) is calculated. In this embodiment, based on the measured plant height, the maximum plant height of the crop is used as h after the plant height curve of the crop during the entire growth period is simulated using the Logistic model. c_max .

[0087] h=E / (1+F×exp(-G×DOY)

[0088] Where h is the plant height (m), E, ​​F, and G are empirical coefficients obtained by fitting the measured plant height using Origin.

[0089] ②T c_min :T c_min It represents the temperature of the cold boundary of the vegetation index-surface temperature trapezoidal space in the HTEM model. In the original HTEM model, T c_min is taken as the mean temperature of the study area. In this embodiment, the PT formula is combined with the energy balance equation to obtain the canopy temperature at the time of crop potential transpiration as T c_min The relevant formula is:

[0090]

[0091] H c =R nc -LE c

[0092]

[0093] Where, LE c is the canopy latent heat flux (W / m 2 ), R nc is the canopy net radiation (W / m 2 ), α PT is taken as 1.3, Δ is the slope of the saturated water vapor pressure and temperature curve (kPa / ℃), γ is the hygrometer constant (kPa / ℃), H c is the sensible heat flux (W / m 2 ), is the aerodynamic impedance of heat transfer between the canopy and the reference height (m / s), ρ is the air density (kg / m 3 ), C p is the specific heat of air (J / kg / K).

[0094] 2.3) Based on the two-stage precise inversion LAI model after calibration and verification, the CI obtained by remote sensing is used red-edge Estimate the LAI of the irrigated area and input the LAI of the irrigated area and the basic driving dataset in step 2.1) into a typical remote sensing dual-source evapotranspiration model (TSEB and HTEM) to divide the LST into Ts and Tc.

[0095] 2.4) Deploy the CTMS-Online crop canopy temperature and environmental factor measurement system in the irrigated crop fields to monitor surface temperature in real time, and obtain the actual canopy temperature and soil temperature based on the precision identification method.

[0096] 2.5) Compare the temperature components (Tc and Ts) estimated using a typical remote sensing dual-source evapotranspiration model with the measured soil and canopy temperatures to determine the accuracy of the temperature component classification.

[0097] Preferably, Figure 4 As shown, the above step 3) specifically includes:

[0098] 3.1) Based on the improved accuracy of LAI estimation and temperature component classification, a typical remote sensing dual-source evapotranspiration model (TSEB and HTEM) is combined with remote sensing imagery to obtain instantaneous ET, E, and T of monitored crops.

[0099] 3.2) Using the evaporation ratio method, the instantaneous ET of the monitored crops is converted into daily values. The evaporation ratio method formula is:

[0100]

[0101] Where, the subscripts i and d represent the instantaneous and daily values ​​of the variables, respectively; EF stands for the evaporation ratio, which is the ratio of evaporation to available energy; λ stands for the latent heat of vaporization (MJ / kg); R n is the net radiation (W / m 2 ), G is the soil heat flux (W / m 2 ).

[0102] 3.3) Use cubic spline interpolation to obtain daily ET, E, and T.

[0103] 3.4) Based on daily evapotranspiration data, monthly and annual crop evapotranspiration and its components can be obtained.

[0104] 3.5) Combine remote sensing inversion to obtain crop planting distribution and extract evapotranspiration and its component distribution maps at different time scales for monitoring crops.

[0105] Preferably, in step 3.5), obtaining crop planting distribution by combining remote sensing inversion includes:

[0106] In the early stage of crop growth (i.e., the period after crop seedlings emerge, generally from the end of June to the beginning of July), the global satellite positioning system is used to conduct a planting structure survey of the irrigation area, recording the latitude and longitude of each point and the crops planted. Combined with the sentinel data obtained from the GEE platform, the random forest method is used to classify the planting structure and obtain the planting distribution map of the monitored crops.

[0107] Example 2

[0108] like Figure 1 As shown, this embodiment takes sunflower (C3 photosynthesis) and corn (C4 photosynthesis), typical crops in the Yongji irrigation area of ​​a certain city, as examples to provide a method for improving the accuracy of remote sensing evapotranspiration based on two-stage precise LAI inversion, including the following steps:

[0109] 1) If Figure 2 As shown, based on the measured LAI data in 2021 and 2022, combined with the available sentinel imagery data during the crop growth period in the Yongji irrigation area, the two-stage LAI model parameters for sunflower (C3) and corn (C4) were calibrated and verified.

[0110] Specifically:

[0111] 1.1) Sunflower and corn monitoring points were evenly distributed in typical plots in the irrigation area. Leaf area of ​​the monitored crops was manually collected regularly during the growth period to calculate LAI, and the corresponding collection time was recorded.

[0112] 1.2) Using Google Earth Engine (GEE) platform, we analyzed the available sentinel images of the irrigation area according to the judgment principle of cloud cover <20%, and calculated the vegetation red edge index (CI) of the corresponding date using band operation. red-edge ), the calculation formula is:

[0113]

[0114] Where R 783 and R 705 They represent the reflectance at wavelengths of 783 nm and 705 nm in the Sentinel image, respectively.

[0115] 1.3) CI based on measured LAI and remote sensing monitoring red-edge Combined with the crop day number (DOY), the modified logistic model was used to obtain the LAI and CI of sunflower and corn during their entire growth period. red-edge Dynamic curve, the calculation formula of the modified Logistic model is:

[0116] y=A / (1+exp(B×DOY 2 +C×DOY+D))

[0117] Where A, B, C and D are empirical parameters that can be obtained by fitting measured data using the Origin model; y is LAI or CI red-edge .

[0118] 1.4) Based on LAI and CI red-edge The full growth period change curve divides the growth period of the crop to be tested into a rapid growth period and a gradual withering period based on the lag in time changes between the two.

[0119] 1.5) Using the measured values ​​of sunflower and corn LAI in 2021 and CI obtained by remote sensing red-edge , respectively calibrate the two-stage LAI model parameters of the tested crops. Since the LAI of sunflower (C3) and corn (C4) have different variation characteristics, that is, the peak duration of corn LAI in the middle growth period is longer than that of sunflower, different model expressions are established for the two crops, as follows:

[0120] sunflower:

[0121] corn:

[0122] Where LAI up LAI is the leaf area index during the rapid growth period. down is the leaf area index during the gradual wilting period, a1, b1, a2, b2, a3, b3, a4, b4 and c are empirical coefficients, and the CI obtained by using the measured LAI and remote sensing is red-edge Fitting obtained.

[0123] 1.6) Extract the CI of remote sensing monitoring at the monitoring point in 2022 red-edge The CI corresponding to the LAI measurement date was obtained by fitting the modified logistic model. red-edge The values ​​were substituted into the calibrated two-stage LAI model to estimate the LAI values ​​of corn and sunflower in 2022, and compared with the measured values ​​to verify the reliability of the two-stage LAI model parameters.

[0124] 2) If Figure 3 As shown in the figure, using Landsat8 remote sensing images and Sentinel images, combined with the two-stage LAI model calibrated and verified in step 1) and the typical remote sensing dual-source ET model (TSEB and HTEM), the remotely sensed land surface temperature (LST) is divided into components, specifically:

[0125] 2.1) In 2022, during the early stages of crop growth, a global positioning system (GPS) survey will be conducted on the Yongji Irrigation Area's planting structure, recording the latitude and longitude of sample points and crop types.

[0126] 2.2) After the collected sample point information is converted into vector data, it is uploaded to the GEE platform. Combined with the available time series sentinel data, the random forest method is used to classify the planting structure and obtain the distribution map of the main crops in the study area.

[0127] 2.3) Based on the screening principle of cloud cover <20%, available Landsat 8 remote sensing images were downloaded from the GEE platform. LST, normalized difference vegetation index (NDVI), albedo, and solar zenith angle (SZA) were obtained by remote sensing inversion as the basic input datasets for the dual-source ET model (TSEB and HTEM model).

[0128] 2.4) Adjust and correct the key parameters in the HTEM model, specifically including two parameters:

[0129] ①h c_max :h c_max In the original HTEM model, it is taken as a fixed value of 1m, which represents the plant height of crops when the crops are fully covered by vegetation under drought conditions. In this embodiment, based on the measured plant height, the plant height curve of crops during the entire growth period is simulated using the Logistic model to obtain h c_max . Sunflower and corn h c_max The values ​​are taken as 1.9m and 3.0m respectively. The Logistic model formula is:

[0130] h=E / (1+F×exp(-G×DOY)

[0131] Where h is the plant height (m), E, ​​F, and G are empirical coefficients obtained by fitting the measured plant height using Origin.

[0132] ②T c_min :T c_min It represents the temperature of the cold boundary of the vegetation index-surface temperature trapezoidal space in the HTEM model. In the original HTEM model, T c_min is taken as the mean temperature of the study area. In this embodiment, the PT formula is combined with the energy balance equation to obtain the canopy temperature at the time of crop potential transpiration as T c_min The relevant formula is:

[0133]

[0134] H c =R nc -LE c

[0135]

[0136] Where, LE c is the canopy latent heat flux (W / m2 ), R nc is the canopy net radiation (W / m 2 ), α PT is taken as 1.3, Δ is the slope of the saturated water vapor pressure and temperature curve (kPa / ℃), γ is the hygrometer constant (kPa / ℃), H c is the sensible heat flux (W / m 2 ), is the aerodynamic impedance of heat transfer between the canopy and the reference height (m / s), ρ is the air density (kg / m 3 ), C p is the specific heat of air (J / kg / K).

[0137] 2.5) Based on the two-stage LAI model that has been calibrated and verified, the CI obtained by remote sensing is used red-edge The LAI of the irrigated area was estimated and used as the driving data for the dual-source ET model (TSEB and HTEM model). The LST was divided into component temperatures to obtain Ts and Tc for sunflower and corn, respectively.

[0138] 2.6) Deploy a CTMS-Online crop canopy temperature and environmental factor measurement system in typical crop plots in the irrigation area to monitor surface temperature in real time, and obtain actual canopy and soil temperatures using precision identification methods.

[0139] 2.7) Compare the temperature components estimated by the dual-source model with the measured soil and canopy temperatures to determine the accuracy of the temperature component division.

[0140] 3) If Figure 4 As shown in Figure 2, based on the improved accuracy of LAI and component temperature, a typical dual-source ET model is used to estimate the ET and its components of typical C3 (sunflower) and C4 (corn) crops in irrigated areas. Specifically:

[0141] 3.1) Based on the improved accuracy of LAI estimation and component temperature partitioning, a dual-source ET model (TSEB and HTEM) combined with remote sensing imagery is used to obtain instantaneous crop ET, E, and T.

[0142] 3.2) Using the evaporation ratio method, the instantaneous ET and its components are converted into daily scale values. The evaporation ratio method formula is:

[0143]

[0144] Where, the subscripts i and d represent the instantaneous value and daily value of the variable, respectively. EF stands for evaporation ratio, which is the ratio of evapotranspiration to available energy. R n is the net radiation (W / m 2 ), G is the soil heat flux (W / m 2 ), λ represents the latent heat of vaporization, and its unit is MJ / kg.

[0145] 3.3) Use cubic spline interpolation to obtain daily scale ET, E, and T.

[0146] 3.4) Based on daily evapotranspiration data, crop evapotranspiration on a monthly and entire growing season scale can be obtained.

[0147] 3.5) Based on the distribution of crop planting structure, the evapotranspiration of corn and sunflower can be extracted.

[0148] Example 3

[0149] The above-mentioned Example 1 provides a method for improving the accuracy of remote sensing evapotranspiration based on a two-stage precise inversion of leaf area index. Correspondingly, this embodiment provides a system for improving the accuracy of remote sensing evapotranspiration based on a two-stage precise inversion of leaf area index. The system provided in this embodiment can implement the method for improving the accuracy of remote sensing evapotranspiration based on a two-stage precise inversion of leaf area index in Example 1. The system can be implemented through software, hardware, or a combination of software and hardware. For example, the system can include integrated or separate functional modules or functional units to perform the corresponding steps in each method of Example 1. Because the system of this embodiment is substantially similar to the method embodiment, the description of the process in this embodiment is relatively simple. For relevant details, please refer to the partial description of Example 1. The system embodiment provided in this embodiment is merely illustrative.

[0150] like Figure 5 As shown, this embodiment provides a system for improving the accuracy of remote sensing evapotranspiration based on two-stage precise inversion of leaf area index corresponding to this embodiment 1, which includes:

[0151] The precision inversion module is used to invert the vegetation red edge index CI during the crop growth period based on two years of field measured LAI data and combined with sentinel images. red-edge , calibrate and validate the parameters of the two-stage precision inversion LAI model for monitoring crops;

[0152] The component temperature partitioning module is used to partition the land surface temperature (LST) obtained by remote sensing into components based on the established two-stage precise inversion LAI model. It uses Landsat8 remote sensing imagery and Sentinel imagery data as driving data for a typical remote sensing dual-source evapotranspiration model (TSEB and HTEM). This module obtains canopy temperature (Tc) and soil temperature (Ts).

[0153] The evapotranspiration accuracy estimation module is used to estimate ET and its components (E and T) in irrigation areas based on the precise inversion of LAI and component temperature division results using typical remote sensing dual-source evapotranspiration models (TSEB and HTEM).

[0154] Example 4

[0155] This embodiment provides a processing device corresponding to the method for improving the accuracy of remote sensing evapotranspiration based on two-stage precise inversion of leaf area index provided in this embodiment 1. The processing device can be a processing device used for a client, such as a mobile phone, laptop computer, tablet computer, desktop computer, etc., to execute the method of embodiment 1.

[0156] The processing device includes a processor, a memory, a communication interface, and a bus. The processor, memory, and communication interface are connected via the bus to facilitate communication between them. The memory stores a computer program executable on the processor. When the processor executes the computer program, it executes the method for improving the accuracy of remote sensing evapotranspiration based on two-stage precise inversion of leaf area index, as provided in Example 1.

[0157] Preferably, the memory may be a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory.

[0158] Preferably, the processor may be a central processing unit (CPU), a digital signal processor (DSP), or other general-purpose processors of various types, which are not limited here.

[0159] Example 5

[0160] The method for improving the accuracy of remote sensing evapotranspiration based on two-stage precise inversion of leaf area index in this embodiment 1 can be specifically implemented as a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing the method for improving the accuracy of remote sensing evapotranspiration based on two-stage precise inversion of leaf area index in this embodiment 1.

[0161] Computer readable storage media can be tangible devices that hold and store instructions used by instruction execution devices. Computer readable storage media can be, for example, but not limited to, electronic storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any combination thereof.

[0162] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0163] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0164] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0165] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0166] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A method for improving the accuracy of remote sensing evapotranspiration, characterized in that: The following steps are involved: Based on two years of field-measured LAI data, combined with Sentinel images, the vegetation red edge index CI during the crop growth period was inverted. red-edge , calibrate and validate the parameters of the two-stage precision inversion LAI model for crop monitoring, including: Select typical crop plots in the irrigation area as monitoring points, manually collect LAI data of the monitored crops regularly for two consecutive years, and record the corresponding collection dates; Using the GEE platform, we analyzed the available sentinel images during the two-year crop growth period in the irrigation area according to the judgment principle of cloud cover <20%, and used band operation inversion to obtain the vegetation red edge index CI during the crop growth period. red-edge ; CI based on the first year's measured LAI data and remote sensing inversion red-edge The data were combined with the corresponding crop day sequence, and the measured values ​​were fitted using the modified Logistic model to obtain the LAI and CI of the monitored crops during their entire growth period. red-edge Dynamic curve; Based on the LAI and CI of the monitored crops during their entire growth period red-edge Dynamic curve, determine LAI and CI red-edge There is a lag between the two, and based on this, the whole growth period of the monitored crops is divided into a rapid growth period and a gradual withering period; The LAI values ​​measured during the first year's rapid growth and gradual withering periods and the CI obtained by remote sensing were used. red-edge , respectively calibrate the parameters of the two-stage precise inversion LAI model for monitored crops; CI of monitored crops obtained using remote sensing in the second year red-edge The data were combined with the day sequence number and the modified Logisic model was used to fit the CI corresponding to the LAI measurement date. red-edge value; The CI obtained by fitting red-edge The values ​​were substituted into the calibrated two-stage precise inversion LAI model to estimate the LAI of the monitored crops, and the LAI estimates were compared with the measured values ​​in the second year to verify the parameters of the established two-stage precise LAI estimation model. Based on the established two-stage precise LAI inversion model, Landsat8 remote sensing imagery and Sentinel imagery data were used as driving data for a typical remote sensing dual-source evapotranspiration model. The surface temperature obtained by remote sensing inversion was divided into component temperatures to obtain canopy temperature and soil temperature. Based on the LAI precise estimation and temperature component classification results, the evapotranspiration of crops in irrigated areas and its components were estimated using a typical remote sensing dual-source evapotranspiration model.

2. The method for improving the accuracy of remote sensing evapotranspiration according to claim 1, characterized in that: When selecting monitored crops in the irrigation area, corresponding crop types are selected according to different photosynthesis pathways.

3. The method for improving the accuracy of remote sensing evapotranspiration according to claim 1, characterized in that: Based on the established two-stage precise inversion LAI model, Landsat8 remote sensing imagery and Sentinel imagery data are used as driving data for a typical remote sensing dual-source evapotranspiration model. The surface temperature obtained by remote sensing inversion is divided into components to obtain canopy temperature and soil temperature, including: On the GEE platform, we analyzed available Landsat 8 remote sensing images based on the screening principle of cloud cover <20%. Surface temperature, normalized difference vegetation index, albedo, and solar zenith angle were obtained through remote sensing inversion. These data served as the basic driving dataset for a typical remote sensing dual-source evapotranspiration model. Adjust and correct key parameters in the HTEM model; According to the two-stage precise inversion LAI model after calibration and verification, the CI obtained by remote sensing is used red-edge Estimate the LAI of the irrigated area and input the LAI of the irrigated area and the basic driving dataset into a typical remote sensing dual-source evapotranspiration model, dividing the surface temperature into canopy temperature and soil temperature; The CTMS-Online crop canopy temperature and environmental factor measurement system is deployed in the crop monitoring plots in the irrigation area to monitor the surface temperature in real time, and the actual canopy temperature and soil temperature are obtained based on the precision identification method; The canopy temperature and soil temperature estimated by a typical remote sensing dual-source evapotranspiration model were compared with the measured soil temperature and canopy temperature to determine the accuracy of the component temperature division.

4. The method for improving the accuracy of remote sensing evapotranspiration according to claim 3, characterized in that: The adjustment and correction of key parameters in the HTEM model include: Based on the measured plant height, the logistic model is used to simulate the plant height curve of the crop during the entire growth period and determine the maximum plant height h of the crop. c_max ; Among them, the simulation formula is: h=E / (1+F×exp(-G×DOY)) Where h is the plant height of the crop, E, F and G are empirical coefficients obtained by fitting the measured plant height of the crop using Origin; The PT formula is combined with the energy balance equation to obtain the canopy temperature during crop potential transpiration as T c_min , the formula is: H c =R nc -THE c Where, LE c is the canopy latent heat flux, R nc is the canopy net radiation, α PT Take it as 1.3, Δ is the slope of the saturated water vapor pressure and temperature curve, γ is the hygrometer constant, H c is the sensible heat flux, is the aerodynamic impedance of heat transfer between the canopy and the reference height, ρ is the air density, C p is the specific heat of air.

5. The method for improving the accuracy of remote sensing evapotranspiration according to claim 1, characterized in that: Based on the component classification results, a typical remote sensing dual-source evapotranspiration model was used to estimate crop evapotranspiration and its components in the irrigated area, including: Using a typical remote sensing dual-source evapotranspiration model combined with remote sensing images to obtain instantaneous crop evapotranspiration, soil evaporation, and vegetation transpiration for monitored crops; The instantaneous evapotranspiration of monitored crops was converted into daily values ​​using the evaporation ratio method; Cubic spline interpolation method is used to interpolate daily crop evapotranspiration, soil evaporation and vegetation transpiration; Based on daily evapotranspiration data, monthly and annual crop evapotranspiration and its components can be obtained; Combined with remote sensing inversion, crop planting distribution is obtained, and evapotranspiration and its component distribution maps of monitored crops at different time scales are extracted.

6. The method for improving the accuracy of remote sensing evapotranspiration according to claim 5, characterized in that: The method of combining remote sensing inversion to obtain crop planting distribution includes: in the early stage of crop growth, using the global satellite positioning system to conduct a planting structure survey of the irrigation area, recording the latitude and longitude of each point and the crops planted, combining the sentinel data obtained from the GEE platform, using the random forest method to classify the planting structure, and obtaining a planting distribution map of the monitored crops.

7. A system for improving the accuracy of remote sensing evapotranspiration, characterized in that: include: The precision inversion module is used to invert the vegetation red edge index CI during the crop growth period based on two years of field measured LAI data and combined with sentinel images. red-edge , calibrate and validate the parameters of the two-stage precision inversion LAI model for crop monitoring, including: Select typical crop plots in the irrigation area as monitoring points, manually collect LAI data of the monitored crops regularly for two consecutive years, and record the corresponding collection dates; Using the GEE platform, we analyzed the available sentinel images during the two-year crop growth period in the irrigation area according to the judgment principle of cloud cover <20%, and used band operation inversion to obtain the vegetation red edge index CI during the crop growth period. red-edge ; CI based on the first year's measured LAI data and remote sensing inversion red-edge The data were combined with the corresponding crop day sequence, and the measured values ​​were fitted using the modified Logistic model to obtain the LAI and CI of the monitored crops during their entire growth period. red-edge Dynamic curve; Based on the LAI and CI of the monitored crops during their entire growth period red-edge Dynamic curve, determine LAI and CI red-edge There is a lag between the two, and based on this, the whole growth period of the monitored crops is divided into a rapid growth period and a gradual withering period; The LAI values ​​measured during the first year's rapid growth and gradual withering periods and the CI obtained by remote sensing were used. red-edge , respectively calibrate the parameters of the two-stage precise inversion LAI model for monitored crops; CI of monitored crops obtained using remote sensing in the second year red-edge The data were combined with the day sequence number and the modified Logisic model was used to fit the CI corresponding to the LAI measurement date. red-edge value; The CI obtained by fitting red-edge The values ​​were substituted into the calibrated two-stage precise inversion LAI model to estimate the LAI of the monitored crops, and the LAI estimates were compared with the measured values ​​in the second year to verify the parameters of the established two-stage precise LAI estimation model. The component temperature partitioning module is used to perform component temperature partitioning on the surface temperature obtained by remote sensing inversion based on the established two-stage precise inversion LAI model. It uses Landsat8 remote sensing imagery and Sentinel imagery data as driving data for a typical remote sensing dual-source evapotranspiration model to obtain canopy temperature and soil temperature. The evapotranspiration accuracy estimation module is used to estimate the evapotranspiration of irrigated crops and its components based on the LAI precision inversion and component temperature division results using a typical remote sensing dual-source evapotranspiration model.

8. A computer-readable storage medium storing one or more programs, characterized in that: The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform any one of the methods of claims 1 to 6 .

9. A computing device, characterized in that include: One or more processors and a memory, wherein the memory stores one or more programs and is configured to be executed by the one or more processors, wherein the one or more programs include instructions for executing any one of the methods according to claims 1 to 6.

Images