A fusion method for actual evapotranspiration from multiple sources in a watershed considering the eco-hydrological effects of vegetation

Abstract

A method for fusing actual evapotranspiration from multiple sources in a watershed considering the eco-hydrological effects of vegetation includes: 1. collecting target data; 2. setting an appropriate target spatiotemporal resolution, preprocessing the data, and calculating the saturated vapor pressure difference (VPD) based on the air temperature; 3. using sunlight-induced chlorophyll fluorescence data (SIF), saturated vapor pressure difference data (VPD), and surface net radiation data (R) to calculate the evapotranspiration from multiple sources in a watershed. n 4. Calculate evapotranspiration data that explicitly considers vegetation effects; 5. Calculate the variance of multi-source evapotranspiration data using a three-source matching method; 6. Utilize linear data fusion based on the variance of multi-source evapotranspiration data to generate a long series of raster basin evapotranspiration products. This method, through multi-source matching and data fusion techniques, integrates other evapotranspiration products, ensuring that the fused product fully accounts for vegetation effects while retaining the high-resolution advantages of established evapotranspiration products. This provides important and accurate data for basin hydrological and ecological research.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-source data fusion, and in particular to a method for fusing actual evapotranspiration from multiple sources in a watershed taking into account the eco-hydrological effects of vegetation. Background Art

[0002] Evapotranspiration is the primary terrestrial pathway for precipitation to return to the atmosphere during the water cycle. The importance of this water flux in climate regulation has been recognized, and in recent years, its importance in ecological research has been recognized. Obtaining accurate evapotranspiration estimates on a large scale has long been a focus of attention. In recent years, with the advancement of satellite remote sensing technology and inversion algorithms, evapotranspiration products based on satellite remote sensing inversion have become the mainstream data product used in watershed evapotranspiration research. Compared to station-based data, current evapotranspiration products have greater global coverage, making up for the lack of evapotranspiration observation sites. Different satellite products have different accuracies and spatiotemporal resolutions, providing a better reference for understanding the spatiotemporal variations of evapotranspiration on a global scale.

[0003] However, currently available publicly available evapotranspiration data still has numerous flaws due to limitations in satellite sensor accuracy and inversion algorithms. First, for satellites with high spatial coverage, the limitations of inversion algorithms necessitate gridding at low resolution to reduce the impact of noise. Even so, low-resolution data still cannot meet the needs of most current research. Limited by the inherent accuracy of the sensors, further data with lower noise and higher spatial resolution cannot be obtained. Second, current satellite revisit periods and evapotranspiration inversion algorithms cannot effectively reflect the role of vegetation activity in evapotranspiration. As a result, some evapotranspiration products fail to reflect the ecohydrological effects of vegetation or the impact of extreme drought events, resulting in underestimation or overestimation. Therefore, the practical application of evapotranspiration data in ecological and hydrological data remains challenging.

[0004] To more effectively utilize evapotranspiration data for hydrological or ecological studies, researchers have begun to improve remote sensing evapotranspiration products. One major improvement approach is the data fusion of multi-source remote sensing satellite products. Current research focuses on fusing existing evapotranspiration data. However, since publicly available evapotranspiration data still does not fully consider the ecohydrological effects of vegetation, there is still room for improvement.

[0005] Overall, existing research has failed to effectively consider the role of vegetation's ecohydrological effects on evapotranspiration, limiting the feasibility of evapotranspiration data in practical research. Therefore, it is necessary to design a method for fusing actual evapotranspiration from multiple sources in a watershed that considers vegetation's ecohydrological effects to overcome this problem. Summary of the Invention

[0006] To avoid the above problems, a method for fusing multi-source actual evapotranspiration in a watershed that considers the eco-hydrological effects of vegetation is provided. By using multi-source matching and data fusion technology, other evapotranspiration products are integrated to ensure that the vegetation effect is fully considered in the fused product. At the same time, the high-resolution advantages of mature evapotranspiration products are also compatible, providing important and accurate data for watershed hydrological and ecological research.

[0007] The present invention provides a method for fusing actual evapotranspiration from multiple sources in a watershed, taking into account the eco-hydrological effects of vegetation, comprising the following steps:

[0008] Step 1: Collect target data, including the currently available land surface temperature data LST, sunlight-induced chlorophyll fluorescence data SIF, and air temperature data T a , surface net radiation data R n , leaf area index data LAI, MODIS land use type data, MODIS evapotranspiration data, GLEAM evapotranspiration data and FLUXNET2015 flux network flux data;

[0009] Step 2: Set the appropriate target spatiotemporal resolution and perform the calculations on the sunlight-induced chlorophyll fluorescence data SIF and the temperature data T a , 2m dew point temperature data T dew , surface net radiation data R n , pre-process the leaf area index data LAI, and calculate the saturated water vapor pressure difference VPD according to the air temperature;

[0010] Step 3: Use sunlight-induced chlorophyll fluorescence data SIF, saturated water vapor pressure difference data VPD, and surface net radiation data R n The data calculation explicitly considers the actual evapotranspiration data ET of vegetation;

[0011] Step 4: Calculate the variance of the actual evapotranspiration data from multiple sources using the three-source matching method;

[0012] Step 5: Based on the variance of multi-source actual evapotranspiration data, a linear data fusion method is used to obtain a long series of gridded actual basin evapotranspiration products.

[0013] Preferably, in step 1:

[0014] Surface temperature data LST and air temperature data T a The shortwave net radiation data and longwave net radiation data provided by the ERA5-Land reanalysis data are derived by adding them together. The calculation formula is:

[0015] R n =R ns +R nl ;

[0016] Where Rn is the net radiation data, unit: J m -2 ; R ns is the shortwave net radiation, unit: J m -2 ; R nl is the long-wave net radiation data, unit: J m -2 ;

[0017] The sunlight-induced chlorophyll fluorescence data (SIF) uses the CSIF sunlight-induced chlorophyll fluorescence data to reconstruct the product; the leaf area index data (LAI) is calculated using the bidirectional reflectance function-corrected reflectance MCD43C4, with a temporal and spatial resolution of 0.05° on a daily scale; the MODIS land use type data uses the MODIS MCD12C1 land use type data.

[0018] Preferably, step 2 specifically includes:

[0019] 2.1 Based on the basic reflectance data of each site, select appropriate temporal and spatial resolutions to resample the SIF data;

[0020] 2.2 Collected surface temperature data LST and air temperature data T a , dew point temperature data T dew , Normalized Difference Vegetation Index data NDVI, Surface Net Radiation data R n The leaf area index data LAI is preprocessed and clipped, and the temporal and spatial resolution of each data is unified with that of the SIF data; the saturated water vapor pressure difference VPD and relative humidity RH are calculated using the processed air temperature and dew point temperature data. The calculation formula is as follows:

[0021] VPD=e sat ×(1-RH),

[0022]

[0023] Among them, RH can be expressed as:

[0024] That is, VPD is expressed as:

[0025] In the above formula, VPD is the saturated water vapor pressure difference, e sat is the saturated water vapor pressure, e sat，dew is the saturated water vapor pressure at dew point temperature, RH is the relative humidity, T a is the temperature data, T dew is the dew point temperature data.

[0026] Preferably, step 3 specifically includes:

[0027] 3.1 Using SIF and VPD data to calculate plant transpiration T, the calculation formula is as follows:

[0028]

[0029] Where T is plant transpiration, unit is Wm -2 ; SIF is sunlight-induced chlorophyll fluorescence; NDVI is normalized difference vegetation index; γ is psychrometric constant; λ cf is the marginal water use efficiency, which is a constant; O2 is a constant 2.09×104Pa; T a is the temperature, in °C; P a is atmospheric pressure, unit is Pa; C a is the atmospheric carbon dioxide concentration, which is a constant of 426.9 ppm; α and β are calibration parameters, which are calibrated based on the flux station data in the studied basin;

[0030] 3.2 Soil evaporation is calculated using the following formula:

[0031] Where RH is relative humidity; R n is the net radiation of the surface; Δ is the slope of the saturated water vapor pressure difference; k A is the extinction coefficient, which is 0.6; γ is the psychrometric constant; LAI is the leaf area index;

[0032] 3.3 Combining plant transpiration and soil evaporation into actual evapotranspiration: ET = E s +T;

[0033] Where: ET is the actual evapotranspiration; E s is soil evaporation; T is plant transpiration.

[0034] Preferably, step 4 specifically includes:

[0035] 4.1 Based on MODIS, GLEAM, and ET SIF Three different actual evapotranspiration data are used to calculate the variance of different data. There is a linear relationship between the three different evapotranspiration data and the true value of the evapotranspiration data, which is as follows:

[0036]

[0037] Where, ET MODIS ET GLEAM ET SIF represent MODIS, GLEAM, and SIF-based evapotranspiration, respectively; ET SIF For ET;

[0038] α1, α2, and α3 are the cumulative deviation coefficients of the evapotranspiration product relative to the true evapotranspiration value;

[0039] β1, β2, and β3 are the multiplication deviation coefficients of the evapotranspiration product relative to the true evapotranspiration value;

[0040] ε1, ε2, and ε3 are the errors of the evapotranspiration product relative to the true evapotranspiration value;

[0041] 4.2 Calculate the error variance of each product:

[0042]

[0043] Where, are the error variances of different actual evapotranspiration products, σ1, σ2, and σ3 are the errors of different actual evapotranspiration products themselves, and σ True is the theoretical true value variance.

[0044] Preferably, step 5 is specifically: using the fusion principle based on linear least squares method to fuse different data products, the formula is as follows:

[0045] ET 融合 =w1ET MODIS +w2ET GLEAM +w3ET SLF ;

[0046] Where w1, w2, and w3 are MODIS, GLEAM, and ET, respectively. SIF The weight of actual evapotranspiration data in evapotranspiration data fusion, ET 融合 MODIS, GLEAM, ET SIF The fused evapotranspiration data product; w1, w2, and w3 are obtained by the least squares method using ET 融合 The calculation is performed with the goal of minimizing the error variance. The error variance of each evapotranspiration product calculated in step 4 is used to quantify w1, w2, and w3 in the data fusion:

[0047]

[0048] After processing with the above formula, the evapotranspiration fusion data can be finally obtained.

[0049] Compared with existing technologies, the present invention offers the following advantages: The method for fusing multi-source actual evapotranspiration data from a watershed, which considers the eco-hydrological effects of vegetation, is scientifically sound and closely aligned with practical research applications. It fully leverages the advantages of currently available sunlight-induced chlorophyll fluorescence data in accurately reflecting the eco-hydrological effects of vegetation to estimate watershed evapotranspiration. This method also integrates other evapotranspiration products through multi-source matching and data fusion techniques. This ensures that the fused product fully accounts for vegetation effects while also retaining the high-resolution advantages of established evapotranspiration products. Ultimately, this fused evapotranspiration product, which considers the eco-hydrological effects of vegetation, provides valuable and accurate data for watershed hydrological and ecological research. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a flow chart of a method for fusing actual evapotranspiration from multiple sources in a watershed considering the eco-hydrological effects of vegetation according to a preferred embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram of the probability density function of the fused basin actual evapotranspiration data according to a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0053] According to an embodiment of the present invention, a method for fusing actual evapotranspiration products from multiple sources in a watershed that considers the eco-hydrological effects of vegetation is provided. First, publicly available remote sensing sunlight-induced chlorophyll fluorescence observation data, global meteorological reanalysis data, vegetation-related data, and land use type datasets are collected. Then, through the ground basic reflectance station data, the appropriate temporal and spatial resolution of the data sampling is determined by correlation analysis and other methods, and the temporal and spatial resolutions of the collected data are unified. Then, the key variables in the evapotranspiration formula are calculated using the collected remote sensing data. Evapotranspiration is divided into two parts: soil evaporation and plant transpiration. The plant transpiration part is calculated using sunlight-induced chlorophyll fluorescence data in combination with meteorological data to obtain evapotranspiration data calculated using sunlight-induced chlorophyll fluorescence data. The evapotranspiration data obtained by using the three-source matching method are analyzed using MODIS, GLEAM, and evapotranspiration data calculated based on sunlight-induced chlorophyll fluorescence to obtain the error variances of the three different evapotranspiration products. Finally, a fused product of watershed evapotranspiration is obtained through a data fusion method. The specific process is described in detail in [1]. Figure 1 .

[0054] The technical solution of the present invention is further described in detail below through examples and in conjunction with the accompanying drawings. A method for fusing actual evapotranspiration from multiple sources in a watershed, taking into account the eco-hydrological effects of vegetation, comprises the following steps:

[0055] Step 1: Collect target data, including the currently available land surface temperature data LST, sunlight-induced chlorophyll fluorescence data SIF, and air temperature data T a , surface net radiation data R n , leaf area index data LAI, MODIS land use type data, MODIS evapotranspiration data, GLEAM evapotranspiration data and FLUXNET2015 flux network flux data; the specific method is:

[0056] As for the surface temperature data and air temperature data, the shortwave net radiation data and longwave net radiation data provided by the ERA5-Land reanalysis data are added together and derived. The calculation formula is:

[0057] R n =R ns +R nl ;

[0058] In the above formula, R n is the net radiation data (unit: J m -2 ), R ns is the shortwave net radiation (unit: J m -2 ), R nl is the long-wave net radiation data (unit: J m -2 ).

[0059] In terms of sunlight-induced chlorophyll fluorescence data, it is used to characterize the photosynthesis effect in the vegetation eco-hydrological effect. The CSIF sunlight-induced chlorophyll fluorescence data reconstruction product is mainly used; the leaf area index data uses the MODIS leaf area index MCD15A3H, with a temporal and spatial resolution of 500m and 4 days.

[0060] The land use type data uses MODIS MCD12C1 land use type data.

[0061] Step 2: Set the appropriate target spatiotemporal resolution and perform the calculations on the sunlight-induced chlorophyll fluorescence data SIF and the temperature data T a , 2m dew point temperature data T dew , surface net radiation data R n , pre-process the leaf area index data LAI, and calculate the saturated water vapor pressure difference VPD according to the air temperature; the specific method is:

[0062] 2.1 Based on the basic reflectance data of each site, select appropriate temporal and spatial resolutions to resample the SIF data.

[0063] 2.2 Collected surface temperature data LST and air temperature data T a , dew point temperature data T dew , Normalized Difference Vegetation Index data NDVI, Surface Net Radiation data R n , leaf area index data LAI are preprocessed and clipped, and the temporal and spatial resolutions of each data are unified with those of the SIF data.

[0064] The processed air temperature and dew point temperature data are used to calculate the saturated water vapor pressure difference VPD and relative humidity RH. The calculation formula is as follows:

[0065] VPD=e sat ×(1-RH),

[0066]

[0067] Among them, RH can be expressed as:

[0068] Therefore, VPD can be expressed as:

[0069] In the above formula, VPD is the saturated water vapor pressure difference, e sat is the saturated water vapor pressure, e sat，dew is the saturated water vapor pressure at dew point temperature, RH is the relative humidity, T air is the temperature data, T dew is the dew point temperature data.

[0070] Step 3: Use sunlight-induced chlorophyll fluorescence data SIF, saturated water vapor pressure difference data VPD, and surface net radiation data R n The data calculation explicitly considers the actual evapotranspiration data ET of vegetation eco-hydrological effects;

[0071] The actual evapotranspiration is divided into two parts for calculation, namely plant transpiration and soil evaporation.

[0072] 3.1 Plant transpiration is calculated using SIF and VPD data. The calculation formula is as follows:

[0073]

[0074] In the above formula:

[0075] T is plant transpiration, unit is Wm -2 ;

[0076] SIF is sunlight-induced chlorophyll fluorescence;

[0077] NDVI is the normalized difference vegetation index;

[0078] γ is the psychrometric constant;

[0079] λ cf is the marginal water use efficiency, which is a constant;

[0080] O2 is a constant 2.09×104Pa;

[0081] T a is the air temperature (℃);

[0082] P a is the atmospheric pressure, Pa;

[0083] C a is the atmospheric carbon dioxide concentration, using a constant of 426.9 ppm;

[0084] α and β are calibration parameters, which are calibrated based on the flux station data in the studied basin;

[0085] 3.2 Soil evaporation is calculated using the following formula:

[0086]

[0087] Where:

[0088] RH is relative humidity;

[0089] R n is the net radiation from the surface;

[0090] Δ is the slope of the saturated water vapor pressure difference;

[0091] k A is the extinction coefficient, which is 0.6;

[0092] γ is the psychrometric constant;

[0093] LAI is leaf area index;

[0094] 3.3 Combining plant transpiration and soil evaporation into actual evapotranspiration:

[0095] ET=E s +T;

[0096] Where:

[0097] ET is actual evapotranspiration;

[0098] E s for soil evaporation;

[0099] T is plant transpiration;

[0100] This ET is referred to as ET in the following text SIF .

[0101] Step 4: Calculate the variance of the actual evapotranspiration data from multiple sources using the three-source matching method. The specific method is as follows:

[0102] 4.1 Based on MODIS, GLEAM, and ET SIF Three different actual evapotranspiration data are used to calculate the variance of different data. Assuming that there is a linear relationship between the three different evapotranspiration data and the true value of the land evapotranspiration data, the following formula can be obtained:

[0103]

[0104] Where:

[0105] ET MODIS ET GLEAM ET SIF represent the evapotranspiration from MODIS, GLEAM and the SIF mentioned above, respectively; among them, ET SIF For ET;

[0106] α1, α2, and α3 are the cumulative deviation coefficients of the evapotranspiration product relative to the true evapotranspiration value;

[0107] β1, β2, and β3 are the multiplicative deviation coefficients of the evapotranspiration product relative to the true value of evapotranspiration;

[0108] ε1, ε2, and ε3 are the errors of the evapotranspiration product relative to the true evapotranspiration value.

[0109] 4.2 Calculate the error variance of each product according to the above formula:

[0110]

[0111] In the above formula, are the error variances of different evapotranspiration products, σ1, σ2, and σ3 are the errors of different evapotranspiration products themselves, and σ True is the theoretical true value variance.

[0112] Step 5: Based on the variance of multi-source actual evapotranspiration data, a linear data fusion method is used to obtain a long series of grid basin actual evapotranspiration products.

[0113] The fusion principle based on linear least squares is used to fuse different data products. The formula is as follows:

[0114] ET 融合 =w1ETM ODIS +w2ET GLEAM +w3ET SIF ;

[0115] In the above formula, w1, w2, and w3 are MODIS, GLEAM, and ET, respectively. SIFThe weight of the actual evapotranspiration data of the basin in the evapotranspiration data fusion, ET 融合 MODIS, GLEAM, ET SIF The fused evapotranspiration data product.

[0116] In the formula, w1, w2, and w3 are obtained by the least squares method using ET 融合 The calculation is performed with the goal of minimizing the error variance. The error variance of each evapotranspiration product calculated in step 4 is used to quantify w1, w2, and w3 in the data fusion:

[0117]

[0118] After processing with the above formula, the evapotranspiration fusion data can be finally obtained.

[0119] Figure 2 This figure compares the evapotranspiration estimates for the CN-Cng and FI-Hyy evapotranspiration sites using the proposed method with those from the traditional evapotranspiration product PT-JPL. The figure shows that the proposed method's estimates for these sites are generally distributed near the 1:1 line, while the PT-JPL evapotranspiration estimates deviate significantly from the 1:1 line, indicating that the proposed method provides more accurate estimates than traditional remote sensing evapotranspiration estimation methods.

[0120] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A basin multi-source actual evapotranspiration fusion method considering vegetation eco-hydrological effects, characterized by: The steps include: Step 1: Collect target data, including the currently available land surface temperature data LST, sunlight-induced chlorophyll fluorescence data SIF, and air temperature data T a , surface net radiation data R n , leaf area index data LAI, MODIS land use type data, MODIS evapotranspiration data, GLEAM evapotranspiration data and FLUXNET2015 flux network flux data; Step 2: Set the appropriate target spatiotemporal resolution and perform the calculations on the sunlight-induced chlorophyll fluorescence data SIF and the temperature data T a , 2m dew point temperature data T dew , surface net radiation data R n , pre-process the leaf area index data LAI, and calculate the saturated water vapor pressure difference VPD according to the air temperature; Step 3: Use sunlight-induced chlorophyll fluorescence data SIF, saturated water vapor pressure difference data VPD, and surface net radiation data R n The data calculation explicitly considers the actual evapotranspiration data ET of vegetation; Step 4: Calculate the variance of the actual evapotranspiration data from multiple sources using the three-source matching method; Step 5: Based on the variance of multi-source actual evapotranspiration data, a linear data fusion method is used to obtain a long series of grid basin actual evapotranspiration products; Step 4 specifically includes: 4.1 Based on MODIS, GLEAM, and ET SIF Three different actual evapotranspiration data are used to calculate the variance of different data. There is a linear relationship between the three different evapotranspiration data and the true value of the evapotranspiration data, which is as follows: Where, ET MODIS ET GLEAM ET SIF represent MODIS, GLEAM, and SIF-based evapotranspiration, respectively; ET SIF For ET; α1, α2, and α3 are the cumulative deviation coefficients of the evapotranspiration product relative to the true evapotranspiration value; β1, β2, and β3 are the multiplication deviation coefficients of the evapotranspiration product relative to the true evapotranspiration value; ε1, ε2, and ε3 are the errors of the evapotranspiration product relative to the true evapotranspiration value; 4.2 Calculate the error variance of each product: Where, are the error variances of different actual evapotranspiration products, σ1, σ2, and σ3 are the errors of different actual evapotranspiration products themselves, and σ True is the theoretical true value variance; Step 5 is as follows: Use the fusion principle based on linear least squares to fuse different data products. The formula is as follows: AND 融合 =w1ET MODIS +w2ET GLEAM +w3ET SIF ; Where w1, w2, and w3 are MODIS, GLEAM, and ET, respectively. SIF The weight of actual evapotranspiration data in evapotranspiration data fusion, ET 融合 MODIS, GLEAM, ET SIF The fused evapotranspiration data product; w1, w2, and w3 are obtained by the least squares method using ET 融合 The calculation is performed with the goal of minimizing the error variance. The error variance of each evapotranspiration product calculated in step 4 is used to quantify w1, w2, and w3 in the data fusion: After processing with the above formula, the evapotranspiration fusion data can be finally obtained.

2. The method for fusing actual evapotranspiration from multiple sources in a watershed considering the vegetation eco-hydrological effect as claimed in claim 1, characterized in that: In step 1: Surface temperature data LST and air temperature data T a The shortwave net radiation data and longwave net radiation data provided by the ERA5-Land reanalysis data are derived by adding them together. The calculation formula is: R n =R ns +R nl ; Where R n is the net radiation data, unit: J m -2 ; R ns is the shortwave net radiation, unit: J m -2 ; R nl is the long-wave net radiation data, unit: J m -2 ; The sunlight-induced chlorophyll fluorescence data (SIF) uses the CSIF sunlight-induced chlorophyll fluorescence data to reconstruct the product; the leaf area index data (LAI) is calculated using the bidirectional reflectance function-corrected reflectance MCD43C4, with a temporal and spatial resolution of 0.05° on a daily scale; the MODIS land use type data uses the MODIS MCD12C1 land use type data.

3. The method for fusing actual evapotranspiration from multiple sources in a watershed considering the vegetation eco-hydrological effect as claimed in claim 1, characterized in that: Step 2 specifically includes: 2.1 Based on the basic reflectance data of each site, select appropriate temporal and spatial resolutions to resample the SIF data; 2.2 Collected surface temperature data LST and air temperature data T a , dew point temperature data T dew , Normalized Difference Vegetation Index data NDVI, Surface Net Radiation data R n The leaf area index data LAI is preprocessed and clipped, and the temporal and spatial resolution of each data is unified with that of the SIF data; the saturated water vapor pressure difference VPD and relative humidity RH are calculated using the processed air temperature and dew point temperature data. The calculation formula is as follows: VPD=e sat ×(1-RH), Among them, RH can be expressed as: That is, VPD is expressed as: In the above formula, e sat is the saturated water vapor pressure, e sat，dew is the saturated water vapor pressure at the dew point temperature.

4. The method for fusing actual evapotranspiration from multiple sources in a watershed considering the vegetation eco-hydrological effect as claimed in claim 1, characterized in that: Step 3 specifically includes: 3.1 Using SIF and VPD data to calculate plant transpiration T, the calculation formula is as follows: Where T is plant transpiration, unit is Wm -2 ; SIF is sunlight-induced chlorophyll fluorescence; NDVI is normalized difference vegetation index; γ is psychrometric constant; λ cf is the marginal water use efficiency, which is a constant; O2 is a constant 2.09×104Pa; T a is the temperature, in °C; P a is atmospheric pressure, unit is Pa; C a is the atmospheric carbon dioxide concentration, which is a constant of 426.9 ppm; α and β are calibration parameters, which are calibrated based on the flux station data in the studied basin; 3.2 Soil evaporation is calculated using the following formula: Where RH is relative humidity; R n is the net radiation of the surface; Δ is the slope of the saturated water vapor pressure difference; k A is the extinction coefficient, which is 0.6; γ is the psychrometric constant; LAI is the leaf area index; 3.3 Combining plant transpiration and soil evaporation into actual evapotranspiration: ET = E s +T; Where: ET is actual evapotranspiration; E s is soil evaporation; T is plant transpiration.

Images