Coupling sunlight-induced chlorophyll fluorescence with physiological constraints to estimate global evapotranspiration and a system thereof

Abstract

The application provides a global evapotranspiration estimation method and system coupled with sunlight-induced chlorophyll fluorescence physiological constraints. The method comprehensively utilizes ERA5-Land meteorological reanalysis data, leaf area index, global rainfall data, land cover classification data and globally reconstructed CSIF data, and introduces FLUXNET2015 ground station observation to correct and match the site ET and meteorological data. At the site scale, a semi-empirical relationship model between each component of ET, meteorological factors and SIF is constructed, and the least square method is used for parameter calibration to obtain a set of site parameters. Further, according to the land cover type, the parameters are classified and combined with the environmental variables to establish a multiple regression model, realizing the climate-driven expression of the parameters. At the global scale, the parameters are spatialized according to the environmental variables and land cover zoning to construct a unified global semi-empirical model, thereby realizing the continuous estimation of global ET. The application improves the physical consistency and accuracy of global ET estimation by introducing SIF into the evapotranspiration physiological constraint process, and has strong popularization and application value.

Description

Technical Field

[0001] This invention relates to the interdisciplinary field of ecological remote sensing and hydrology, and in particular to a method and system for estimating global evapotranspiration coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence. Background Technology

[0002] Terrestrial evapotranspiration (ET) is a central link in the global water, carbon, and energy cycles. As the primary pathway for water to return from the land surface to the atmosphere, ET consumes more than half of the solar radiation energy from the land surface and re-transports approximately 60% of global precipitation back into the atmosphere. Accurate quantification of ET and its three core components—vegetation transpiration (T...)—is crucial. r Soil evaporation (E) s ) and canopy intercepted water (E) i This is crucial for assessing the evolution of the hydrological cycle, drought monitoring, and ecosystem dynamics under climate change.

[0003] Solar-induced chlorophyll fluorescence (SIF) is a direct probe for monitoring photosynthesis in terrestrial ecosystems. SIF has a real physiological correlation with actual photosynthetic rates and light use efficiency in vegetation, and can more sensitively capture physiological fluctuations and transient changes in photosynthetic activity under environmental stress. Global SIF data can be obtained using satellites such as GOSAT, OCO-2, and Tropomi. However, satellite-observed SIF often suffers from low spatial resolution and spatiotemporal discontinuities. To overcome these shortcomings, a series of long-term, high-resolution (0.05°) spatiotemporally continuous SIF products, such as CSIF, GOSIF, and RTSIF, have been developed.

[0004] In recent years, the successful application of SIF in ecosystem carbon cycle monitoring has prompted its further expansion into the field of evapotranspiration estimation. Studies have shown a significant correlation between SIF and ET, and using SIF as a constraint on canopy conductance can more effectively simulate vegetation transpiration and optimize the assessment of transpiration proportions compared to traditional models.

[0005] Currently, there is still controversy regarding the effectiveness of global ET estimation, and existing methods generally have the following problems: (1) T rThe estimation of stomatal conductance is highly uncertain; (2) Existing models often use simplified semi-empirical models when dealing with soil evaporation, which leads to inaccurate estimation of the proportions of each component of ET. Overall, existing studies have failed to fully reflect the actual ET information on the land surface. Summary of the Invention

[0006] This invention provides a method and system for estimating global evapotranspiration coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence, in order to address the shortcomings of existing SIF-driven evapotranspiration models in distinguishing components (such as neglecting canopy retention E). i This addresses the shortcomings of parameter spatialization schemes in lacking dynamic climate-driven expression, enabling refined and coordinated estimation of vegetation transpiration, soil evaporation, and canopy interception.

[0007] In a first aspect, the present invention provides a method for estimating global evapotranspiration coupled with sunlight-induced chlorophyll fluorescence physiological constraints, comprising:

[0008] Collect and preprocess ERA5-Land meteorological reanalysis data, leaf area index data, global rainfall data, global land cover classification data, and global reconstructed CSIF daylight-induced chlorophyll fluorescence data;

[0009] FLUXNET2015 ground station observation data is acquired, and the ET observation data and meteorological data of the ground stations are time-averaged and corrected to obtain corrected ground station ET observation data and corrected ground station meteorological data. The globally reconstructed CSIF daylight-induced chlorophyll fluorescence data is matched to each station according to the station's latitude and longitude coordinates to obtain the station-matched SIF observation data.

[0010] A semi-empirical formula is constructed between the corrected ET observation data of the ground station, the corrected meteorological data of the ground station, and the SIF observation data matched by the station. The semi-empirical formula is then calibrated at various ground stations. The least squares algorithm is used to obtain the estimated parameter set of the semi-empirical formula at various ground stations.

[0011] Based on the global land cover classification data, the ground stations are classified, and the station environmental variables are statistically extracted from the corrected meteorological data of the ground stations. The station environmental variables are used to perform multiple linear regression on the estimated parameter set under each land cover classification to obtain the parameter regression model.

[0012] The global land surface is classified based on the global land cover classification data. Global environmental variables are statistically extracted from the ERA5-Land meteorological reanalysis data. The global environmental variables are substituted into the parameter regression model under each land cover classification to obtain the global parameter distribution.

[0013] A global semi-empirical formula is constructed using the global parameter distribution. The global ET data are obtained by substituting the ERA5-Land meteorological reanalysis data, the leaf area index data, the global precipitation data, and the global reconstructed CSIF daylight-induced chlorophyll fluorescence data into the global semi-empirical formula.

[0014] According to the present invention, a global evapotranspiration estimation method coupled with physiological constraints of sunlight-induced chlorophyll fluorescence includes the acquisition and preprocessing of ERA5-Land meteorological reanalysis data, leaf area index data, global precipitation data, global land cover classification data, and globally reconstructed CSIF sunlight-induced chlorophyll fluorescence data, comprising:

[0015] For the hourly ERA5-Land meteorological reanalysis data, temperature variables, dew point temperature variables, net radiation variables, and wind speed variables are extracted at a preset number of days time resolution. The spatial resolution of the temperature variables, dew point temperature variables, net radiation variables, and wind speed variables is unified to a preset number of degrees through spatial interpolation.

[0016] For the leaf area index data with a spatial resolution of 500m, the spatial resolution is unified to a preset degree through spatial resampling;

[0017] For the global rainfall data on a three-hour timescale, the raw data is labeled according to whether rainfall occurred. If rainfall occurred during that time period, it is labeled as a rainfall event. The average rainfall intensity of the rainfall events in that month is calculated on a monthly basis.

[0018] For the global land cover classification data at the annual scale, a mapping from code data to actual land cover classification is constructed according to the IGBP classification standard.

[0019] According to a global evapotranspiration estimation method coupled with sunlight-induced chlorophyll fluorescence physiological constraints provided by the present invention, FLUXNET2015 ground station observation data is acquired, and the ET observation data and meteorological data of the ground stations are time-averaged and corrected to obtain corrected ground station ET observation data and corrected ground station meteorological data, including:

[0020] For the hourly-scale FLUXNET2015 ground station observation data, temperature, relative humidity, saturated vapor pressure difference, and net radiation variables at a preset number of days of time resolution are extracted. Quality control bands are used to perform quality control on these variables. If the proportion of unusable data at the preset number of days exceeds 50%, that data is deleted.

[0021] For the global reconstructed CSIF daylight-induced chlorophyll fluorescence data, the original data is global raster data. Based on the latitude and longitude coordinates of the FLUXNET2015 stations, the raster position of the stations is located, and the CSIF data is matched with the stations.

[0022] According to the present invention, a global evapotranspiration estimation method coupled with sunlight-induced chlorophyll fluorescence physiological constraints is proposed. A semi-empirical formula is constructed between the corrected ground station ET observation data, the corrected ground station meteorological data, and the station-matched SIF observation data. The semi-empirical formula is then calibrated at various ground stations. A least squares algorithm is used to obtain the estimated parameter set of the semi-empirical formula at each ground station, including:

[0023] A semi-empirical formula is constructed to connect the corrected ET observation data from the ground stations, the corrected meteorological data from the ground stations, and the SIF observation data matched to the stations. First, ET is divided into T... r E s E i Model the three components separately;

[0024] T r The model is constructed based on leaf water and carbon exchange processes. The model conceptualizes canopy gross primary productivity (GPP) as a canopy path-averaged Fick diffusion process.

[0025]

[0026] In the formula (mol m) -2 s -1 () is the CO2 conductance of the canopy stomata. (ppm) and (ppm) represent the CO2 concentration in the atmosphere and between plant cells, respectively. This represents the CO2 concentration gradient inside and outside plant leaves, driving CO2 exchange between the leaves and the atmosphere. Similarly, T r Described using Dalton's Law

[0027]

[0028] In the formula (mol m) -2 s -1 () is the canopy stomatal conductivity for water vapor; when water vapor and CO2 share the same pathway, VPD stands for vapor pressure difference (kPa). It is atmospheric pressure (100 kPa). This represents the equivalent mole fraction gradient of water vapor inside and outside plant leaves, driving water vapor exchange between leaves and the atmosphere. Combined with the above formula, canopy conductance can be eliminated.

[0029]

[0030] In the formula, VPD, and If there are direct observations or the value is fixed as a constant, then SIF is used as a direct indicator of vegetation physiological state to constrain GPP, and a linear relationship is used to describe the SIF-GPP relationship.

[0031]

[0032] The parameters to be calibrated in the formula (μmol m) -2 s -1 / mW m -2 nm -1 sr -1 )and (μmol m) -2 s -1 ) describes SIF(mW m -2 nm -1 sr -1 ) and GPP (μmol m -2 s -1 The relationship between ) The formula is as follows:

[0033]

[0034] In the formula (mol mol) -1 The marginal water use efficiency is the reciprocal of the marginal water use efficiency, and its value depends on the IGBP type. See Table 1 for specific values. (ppm) is the CO2 compensation point in a leaf without mitochondrial respiration, and can be calculated using the standard equation.

[0035]

[0036] In the formula (°C) represents air temperature. (Pa) is the partial pressure of oxygen inside the leaf, set as a constant of 2.09 × 10⁻⁶. 4 Pa. Combining the above formula, and... Units from mol m -2 s -1 Transform into W m -2 ,get expression,

[0037]

[0038] E s The model is constructed as follows:

[0039]

[0040] In the formula, RH (no unit) is the relative humidity of the air. It is the slope of the saturated water vapor pressure-temperature curve. It is the humidity constant, 'a' is the parameter to be calibrated, and R... ns (W m) -2 The net radiation reaching the soil is determined by... Calculate; G(W m) -2 ( ) represents soil heat flux, taken as... , It is the proportion of green vegetation in the environment.

[0041]

[0042] In the formula Take 0.8, Set to 0;

[0043] E i The model is constructed as follows:

[0044]

[0045] In the formula for , This is the extinction coefficient, and its value depends on the IGBP type. See Table 1 for details. (mm hr) -1 () is the average rainfall rate during the rainfall period. (mm hr) -1 The average evaporation rate during the rainfall period is calculated using the Daltonian evaporation equation. (mm) represents the total rainfall of the rainfall event. (mm) is the canopy saturation precipitation threshold, calculated by the following formula:

[0046]

[0047] In the formula (mm) represents the rainfall storage capacity of a specific vegetation type; values ​​are shown in Table 1. (m) 2 m -2 () is the area index of dead leaves, branches, and stems of vegetation, and together with LAI, it represents the vegetation area index. The calculation is performed iteratively by the following formula:

[0048]

[0049] In the formula and Representing the nth and (n-1)th times respectively , and Representing the nth and (n-1)th times respectively , express The minimum value that can be achieved, This represents the residual rate. and The value of is related to the IGBP type, see Table 1;

[0050] Table 1 Parameter Values

[0051] IGBP ENF 250 0.45 0.85 1 0.25 EBF 750 0.59 0.85 1 0.15 DNF 250 0.45 0.85 1 0.25 DBF 1250 0.59 0.85 1 0.15 MF 1000 0.59 0.85 1 0.2 CSH 700 0.56 0.85 1 0.55 OSH 700 0.56 0.85 1 0.55 WSA 1200 0.5 0.85 1 0.12 SAV 1200 0.5 0.85 1 0.09 GRA 700 0.5 0.85 1 0.06 WET 560 0.56 0.85 1 0.06 CRO 800 0.62 0.9 0.1 0.06 URB N / A N / A N / A N / A N / A CVM 750 0.56 0.9 0.5 0.06 SNO N / A N / A N / A N / A N / A BSV 700 0.56 0 0 0 WAT N / A N / A N / A N / A N / A

[0052] ET three components T r E s E i Adding these together yields a semi-empirical formula for calculating ET. The corrected ground station ET observation data, the corrected ground station meteorological data, and the station-matched SIF observation data are input into the semi-empirical formula for calculating ET, and the least squares method is used to determine the formula... (T) r Model), β (T) r (model) and a (E) s The model uses three parameters to obtain the set of estimated parameters for the semi-empirical formula at various ground stations.

[0053] According to the present invention, a global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence is provided. Based on the global land cover classification data, the surface stations are classified, and station environmental variables are statistically derived from the corrected meteorological data of the surface stations. The station environmental variables are then used to perform multiple linear regression on the estimated parameter set under each land cover classification to obtain a parameter regression model, including:

[0054] First, the estimated parameter set is classified according to IGBP type using the global land cover classification data. Then, temperature (Ta), precipitation (P), net radiation (Rn), saturated vapor pressure difference (VPD), relative humidity (RH), chlorophyll fluorescence (SIF), potential evapotranspiration (PET), and normalized latitude location (|LAT| / 90) are selected as environmental variables from the ground station observation data.

[0055] Except for latitude, the multi-year average (Mean) and variability (Range) of the other environmental factors are extracted to represent the average climate conditions and seasonal characteristics, respectively. The environmental variables are used to perform multiple linear regression on the estimated parameter set to obtain the parametric regression model.

[0056] According to the present invention, a global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence is provided. Based on the global land cover classification data, the global land surface is classified; global environmental variables are statistically derived from the ERA5-Land meteorological reanalysis data; and the global environmental variables are substituted into the parameter regression model under each land cover classification to obtain the global parameter distribution, including:

[0057] First, the global land area is classified according to the IGBP type using the global land cover classification data. The environmental variables are statistically extracted from the ERA5-Land meteorological reanalysis data and substituted into the multiple regression parameter model based on land cover classification to calculate the parameters of each grid cell globally, thus obtaining the global parameter distribution.

[0058] According to the global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence provided by the present invention, a global semi-empirical formula is constructed using the global parameter distribution. The global semi-empirical formula is then substituted with the ERA5-Land meteorological reanalysis data, the leaf area index data, the global precipitation data, and the global reconstructed CSIF sunlight-induced chlorophyll fluorescence data to obtain global ET data, including:

[0059] After obtaining the global parameter distribution, a global semi-empirical formula can be constructed for each grid cell globally based on the structure of the semi-empirical formula. The global semi-empirical formula is then substituted with the ERA5-Land meteorological reanalysis data, the leaf area index data, the global rainfall data, and the global reconstructed CSIF daylight-induced chlorophyll fluorescence data to calculate the global ET data and its distribution.

[0060] Secondly, the present invention also provides a global evapotranspiration estimation system coupled with sunlight-induced chlorophyll fluorescence physiological constraints, comprising:

[0061] The data acquisition module is used to collect and preprocess ERA5-Land meteorological reanalysis data, leaf area index data, global rainfall data, global land cover classification data, and global reconstructed CSIF daylight-induced chlorophyll fluorescence data.

[0062] The acquisition module is used to acquire FLUXNET2015 ground station observation data, perform time averaging and correction on the ground station ET observation data and meteorological data to obtain corrected ground station ET observation data and corrected ground station meteorological data, and match the global reconstructed CSIF daylight-induced chlorophyll fluorescence data to each station according to the station's latitude and longitude coordinates to obtain the station-matched SIF observation data.

[0063] The calibration module is used to construct a semi-empirical formula between the corrected ground station ET observation data, the corrected ground station meteorological data, and the station-matched SIF observation data, and to calibrate the semi-empirical formula at various ground stations. The least squares algorithm is used to obtain the estimated parameter set of the semi-empirical formula at various ground stations.

[0064] The regression module is used to classify the ground stations based on the global land cover classification data, to extract station environmental variables from the corrected meteorological data of the ground stations, and to perform multiple linear regression on the estimated parameter set under each land cover classification using the station environmental variables to obtain a parameter regression model.

[0065] The distribution module is used to classify the global land surface based on the global land cover classification data, to statistically extract global environmental variables from the ERA5-Land meteorological reanalysis data, and to substitute the global environmental variables into the parameter regression model under each land cover classification to obtain the global parameter distribution.

[0066] The calculation module is used to construct a global semi-empirical formula using the global parameter distribution. The global semi-empirical formula is substituted with the ERA5-Land meteorological reanalysis data, the leaf area index data, the global rainfall data, and the global reconstructed CSIF daylight-induced chlorophyll fluorescence data to obtain global ET data.

[0067] Thirdly, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence as described above.

[0068] The global evapotranspiration estimation method and system coupled with sunlight-induced chlorophyll fluorescence physiological constraints provided by this invention fully utilizes satellite remote sensing observation data and ground station observation data to construct an accurate evapotranspiration estimation model based on SIF, which can better reflect the actual ET process on the Earth's surface. By using a reasonable method to generalize the station-scale model parameters, a global parameter distribution is generated, and accurate ET estimation is successfully achieved on a global scale. It has the characteristics of being scientific, reasonable and close to engineering practice. Attached Figure Description

[0069] To more clearly illustrate the technical solutions in this invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without any creative effort.

[0070] Figure 1 This is one of the flowcharts of the global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence provided by the present invention;

[0071] Figure 2 This is the second flowchart of the global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence provided by the present invention;

[0072] Figure 3 This is a graph showing the fitting effect of the model parameters for multiple linear regression provided by this invention;

[0073] Figure 4 This is a schematic diagram of the probability density function for reconstructing the multi-year average of ET globally, provided by this invention.

[0074] Figure 5 This is a schematic diagram of the global evapotranspiration estimation system coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence provided by the present invention;

[0075] Figure 6 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0076] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, embodiments of this invention. Other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are all within the protection scope of this invention.

[0077] To address the problems of high uncertainty in vegetation transpiration parameters and inaccurate estimation of the proportions of various evapotranspiration components in existing technologies, this invention proposes a global evapotranspiration estimation method and system coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence. By introducing sunlight-induced chlorophyll fluorescence data acquired by satellite remote sensing, a quantitative relationship is established between the photosynthetic physiological state of vegetation and the water consumption process, thereby achieving high-precision estimation of global evapotranspiration.

[0078] Figure 1This is one of the flowcharts illustrating the global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence provided by this invention, such as... Figure 1 As shown, it includes:

[0079] Step 100: Collect and preprocess ERA5-Land meteorological reanalysis data, leaf area index data, global rainfall data, global land cover classification data, and global reconstructed CSIF daylight-induced chlorophyll fluorescence data;

[0080] Step 200: Obtain FLUXNET2015 ground station observation data, perform time averaging and correction on the ground station ET observation data and meteorological data to obtain corrected ground station ET observation data and corrected ground station meteorological data, and match the global reconstructed CSIF daylight-induced chlorophyll fluorescence data to each station according to the station's latitude and longitude coordinates to obtain the station-matched SIF observation data.

[0081] Step 300: Construct a semi-empirical formula between the corrected ground station ET observation data, the corrected ground station meteorological data, and the station-matched SIF observation data, and calibrate the semi-empirical formula at various ground stations. Use the least squares algorithm to obtain the estimated parameter set of the semi-empirical formula at various ground stations.

[0082] Step 400: Classify the ground stations based on the global land cover classification data, extract the station environmental variables from the corrected meteorological data of the ground stations, and use the station environmental variables to perform multiple linear regression on the estimated parameter set under each land cover classification to obtain the parameter regression model;

[0083] Step 500: Classify the global land surface based on the global land cover classification data, statistically extract global environmental variables from the ERA5-Land meteorological reanalysis data, and substitute the global environmental variables into the parameter regression model under each land cover classification to obtain the global parameter distribution;

[0084] Step 600: Construct a global semi-empirical formula using the global parameter distribution. Substitute the ERA5-Land meteorological reanalysis data, the leaf area index data, the global precipitation data, and the global reconstructed CSIF daylight-induced chlorophyll fluorescence data into the global semi-empirical formula to obtain global ET data.

[0085] Specifically, such as Figure 2As shown, firstly, preprocessed ERA5-Land meteorological reanalysis data, leaf area index data, global precipitation data, global land cover classification data, and global reconstructed CSIF day-induced chlorophyll fluorescence data were collected; then, FLUXNET2015 ground station observation data were acquired and averaged, and CSIF data were matched according to station location; the parameters of the semi-empirical formula were calibrated at the FLUXNET station scale; multiple linear regression of the parameters was performed using station environmental variables, and then the parameters were extrapolated using global environmental variables; the preprocessed global meteorological data were substituted into the global raster to calculate the global ET data and the quantities of each component of global ET.

[0086] The global evapotranspiration estimation method and system coupled with sunlight-induced chlorophyll fluorescence physiological constraints provided by this invention fully utilizes satellite remote sensing observation data and ground station observation data to construct an accurate evapotranspiration estimation model based on SIF, which can better reflect the actual ET process on the Earth's surface. By using a reasonable method to generalize the station-scale model parameters, a global parameter distribution is generated, and accurate ET estimation is successfully achieved on a global scale. It has the characteristics of being scientific, reasonable and close to engineering practice.

[0087] Based on the above embodiments, step 100 includes:

[0088] Collect and process 2m temperature data (T) from ERA5-Land meteorological reanalysis data. air ), 2m dew point temperature data (T) dew ), 10mU component wind speed data (Ws) U ), 10mV component wind speed data (Ws) V ), Net solar radiation data (Rn) sw ), net surface thermal radiation data (Rn) lw ); Leaf area index data (LAI), global rainfall data (P), global land cover classification data (IGBP), and global reconstructed CSIF data on light-induced chlorophyll fluorescence (CSIF).

[0089] Specifically, based on ERA5-Land hourly meteorological reanalysis data, 2m temperature data (T) with a time resolution of 4 days were extracted. air ), 2m dew point temperature data (T) dew ), 10mU component wind speed data (Ws) U ), 10mV component wind speed data (Ws) V ), Net solar radiation data (Rn) sw ), net surface thermal radiation data (Rn) lw Meanwhile, the spatial resolution of the above data is unified to 0.05° through spatial interpolation.

[0090] Using these data, global relative humidity (RH), saturated vapor pressure difference (VPD), and net radiation (R) were calculated. n The calculation formula is as follows: (The data includes wind speed (Ws) and wind speed (Ws)).

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097] In the formula The saturated vapor pressure, This represents the saturated vapor pressure. Through recalculation of the original ERA5 data, global-scale RH, VPD, and R were obtained. n Ws data.

[0098] T of the world air RH, VPD, R n The Ws, LAI, and CSIF data are uniformly maintained at a spatiotemporal resolution of 4 days and 0.05° for subsequent input into the model for calculation; the global precipitation data (P) at a 3-hour resolution is marked as raw data according to whether rainfall occurred. If there is rainfall during that period, it is marked as a rainfall event, and the average rainfall intensity of the rainfall event in that month is calculated on a monthly basis; the global IGBP data retains the original spatiotemporal resolution of 0.05° for 1 year.

[0099] Based on the above embodiments, step 200 includes:

[0100] FLUXNET2015 ground station observation data were acquired, and the ET observation data and meteorological data of the ground stations were time-averaged and corrected to obtain corrected ground station ET observation data and corrected ground station meteorological data. The globally reconstructed CSIF solar-induced chlorophyll fluorescence data were then matched to each station according to the station's latitude and longitude coordinates to obtain the station-matched SIF observation data.

[0101] Specifically, for FLUXNET 2015 hourly resolution site data, VPD, RH, and R are filtered. n P, T air Meteorological data, such as E, will be used for station-scale parameter calibration, with the station's ET observation data serving as the target value. Because E iThe calculation formula does not contain parameters, and the influence of this component can be easily eliminated by screening drought data, which facilitates the calculation of T. r and E s The parameters in the formula are calibrated. Here, the station data is filtered in a certain way. For station data every 4 days, if the period of rainfall is greater than 10%, the data for those 4 days is deleted. The filtered dry data is then averaged over a 4-day time window.

[0102] For the global reconstructed CSIF daylight-induced chlorophyll fluorescence data, matching is performed according to the latitude and longitude of the stations. The global CSIF data is 0.05° raster data, and each raster corresponds to two boundaries in longitude and latitude. If the latitude and longitude of a certain station are within these boundaries, the CSIF data of the raster is considered to be the data of that station.

[0103] Based on the above embodiments, step 300 includes:

[0104] Evapotranspiration is decomposed into vegetation transpiration (T). r ), soil evaporation (E) s ) and canopy retention evaporation (E i The model consists of three components, each of which is modeled separately. The dry data of the stated sites and the matched CSIF data of the sites are input into the model, and the formula parameters are obtained using the least squares method for calibration.

[0105] Specifically, vegetation transpiration (T r The model is constructed based on the physiological processes of water and carbon exchange in leaves. First, canopy total primary productivity (GPP) is represented as the diffusion process of CO2 through the stomatal pathway in leaves; simultaneously, vegetation transpiration is represented as the diffusion process of water vapor through the same pathway. By utilizing the assumption that CO2 and water vapor share a stomatal pathway, the stomatal conductance parameter, which is difficult to observe directly, is eliminated. Soil evaporation (E...) s The model employs a semi-empirical expression based on energy constraints, considering net radiation reaching the soil surface, soil heat flux, and air humidity conditions, and sets parameters to be calibrated to comprehensively reflect soil and underlying surface characteristics. Canopy evaporation interception (E0) i The model is driven by rainfall events and explicitly considers factors such as rainfall intensity, canopy water storage capacity, vegetation area index and extinction coefficient. It uses a recursive approach to describe the dynamic changes in canopy intercepted water.

[0106] The site VPD, RH, R n P, T air Meteorological data and the matched station CSIF data drive the model. Using the station's ET observation data as the target value, the least squares method is used to determine the three parameters α(T) of the model for each station. r Model), β (T) r (model) and a (E)s (Model). The basic calibration process of the least squares method is as follows:

[0107] (1) Establishing the model structure: given the observation data Establish a parameterized model.

[0108]

[0109] (2) Construction error and objective function: Define residual The objective function is to minimize the sum of squared residuals.

[0110]

[0111] (3) Find the optimal parameters: Write the model in matrix form.

[0112]

[0113] By analyzing the objective function with respect to Taking the derivative and setting it to zero, we obtain the normal equation.

[0114]

[0115] Solving this normal equation will yield the optimal parameters.

[0116] Based on the above embodiments, step 400 includes:

[0117] Based on global land cover classification data, station parameters were categorized according to IGBP type, and variables such as temperature, precipitation, net radiation, vapor pressure difference, relative humidity, chlorophyll fluorescence, and normalized latitude were selected as environmental factors. Except for latitude, multi-year averages and variability were extracted for the remaining environmental factors to characterize climate background and seasonal characteristics. Multiple linear regression was then performed on the model parameters under each land cover type to establish a parametric regression model.

[0118] Specifically, from the site VPD, RH, R n P, T air From long-term data such as meteorological data and the matched station CSIF data, the multi-year average (mean) and the annual range average (range) are calculated. The average is used to reflect the overall situation, and the range is used to measure seasonality.

[0119] The parameters of the multiple linear regression model are determined using the above environmental variables.

[0120]

[0121] In the formula This represents the parameter of the i-th station under this IGBP type; This is the environment variable matrix for this site; This is the regression coefficient matrix for this IGBP type; This represents the prediction error. The key to the solution lies in determining the environmental variable matrix. Includes environmental variable types and regression coefficient matrix To determine the optimal number and combination of environmental variables, all possible combinations were explored. To prevent overfitting and reduce computational complexity, the number of variables was set between 1 and 4. Leave-one-out cross-validation (LOOCV) was used to evaluate the effect of each combination, removing one site at a time as an independent validation set and using the remaining sites to build the model. This process ensured that each site was simulated "without participating in the modeling." Finally, the optimal number and combination of variables were determined by predicting the results.

[0122] The effect of each parameter combination was measured using the Akaike Information Criterion (AIC).

[0123]

[0124] AIC (Adaptive Inference Capacity) can comprehensively quantify the model's predictive performance and complexity. A lower value indicates that the model maintains a high fit while possessing reasonable complexity. k represents the number of model parameters. The optimal combination of parameters for each IGBP (Integrated Geometric Parameters) results in the following multivariate linear regression model performance: Figure 3 As shown.

[0125] Based on the above embodiments, step 500 includes:

[0126] From the global T air RH, VPD, R n Environmental variables are statistically derived from the global rainfall data, including Ws, LAI, CSIF, and the aforementioned data. The global land surface is classified based on the aforementioned global land cover classification data. For each IGBP type, the environmental variables are substituted into the aforementioned parameter multiple linear regression analysis model to obtain the global parameter distribution.

[0127] Based on the above embodiments, step 600 includes:

[0128] A global semi-empirical formula is constructed using the global parameter distribution, and the global T... air RH, VPD, R n By substituting the data from Ws, LAI, and CSIF into the formula for calculation, global ET data with a spatial resolution of 0.05° and a temporal resolution of 4 days can be obtained. For example... Figure 4 The figure shown is a schematic diagram of the probability density function of the global ET mean.

[0129] The global evapotranspiration estimation system coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence provided by this invention is described below. The global evapotranspiration estimation system coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence described below can be referred to in correspondence with the global evapotranspiration estimation system coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence described above.

[0130] Figure 5 This is a schematic diagram of the structure of the global evapotranspiration estimation system coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence provided in an embodiment of the present invention, as shown below. Figure 5 As shown, it includes: a data acquisition module 51, an acquisition module 52, a calibration module 53, a regression module 54, a distribution module 55, and a calculation module 56, wherein:

[0131] The acquisition module 51 is used to acquire and preprocess ERA5-Land meteorological reanalysis data, leaf area index data, global precipitation data, global land cover classification data, and globally reconstructed CSIF daylight-induced chlorophyll fluorescence data. The acquisition module 52 is used to acquire FLUXNET2015 ground station observation data, perform time averaging and correction on the ground station ET observation data and meteorological data to obtain corrected ground station ET observation data and corrected ground station meteorological data, and match the globally reconstructed CSIF daylight-induced chlorophyll fluorescence data to each station according to the station's latitude and longitude coordinates to obtain the station-matched SIF observation data. The calibration module 53 is used to construct a semi-empirical formula between the corrected ground station ET observation data, the corrected ground station meteorological data, and the station-matched SIF observation data, and calibrate the semi-empirical formula at each ground station using the least squares algorithm to obtain the semi-empirical formula in each... The system comprises: a set of estimated parameters for various ground stations; a regression module 54 for classifying the various ground stations based on the global land cover classification data, statistically analyzing station environmental variables from the corrected meteorological data of the ground stations, and performing multiple linear regression on the set of estimated parameters under each land cover classification using the station environmental variables to obtain a parameter regression model; a distribution module 55 for classifying the global land surface based on the global land cover classification data, statistically analyzing global environmental variables from the ERA5-Land meteorological reanalysis data, and substituting the global environmental variables into the parameter regression model under each land cover classification to obtain a global parameter distribution; and a calculation module 56 for constructing a global semi-empirical formula using the global parameter distribution, substituting the ERA5-Land meteorological reanalysis data, the leaf area index data, the global rainfall data, and the global reconstructed CSIF daylight-induced chlorophyll fluorescence data into the global semi-empirical formula to obtain global ET data.

[0132] Figure 6 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 5 As shown, the electronic device may include: a processor 610, a communication interface 620, a memory 630, and a communication bus 640, wherein the processor 610, the communication interface 620, and the memory 630 communicate with each other through the communication bus 640. Processor 610 can call logic instructions in memory 630 to execute a global evapotranspiration estimation method coupled with physiological constraints of sunlight-induced chlorophyll fluorescence (SIF). This method includes: acquiring and preprocessing ERA5-Land meteorological reanalysis data, leaf area index data, global precipitation data, global land cover classification data, and globally reconstructed CSIF sunlight-induced chlorophyll fluorescence data; acquiring FLUXNET2015 ground station observation data, performing time averaging and correction on the ground station ET observation data and meteorological data to obtain corrected ground station ET observation data and corrected ground station meteorological data, and matching the globally reconstructed CSIF sunlight-induced chlorophyll fluorescence data to each station according to the station's latitude and longitude coordinates to obtain station-matched SIF observation data; constructing a semi-empirical formula between the corrected ground station ET observation data, the corrected ground station meteorological data, and the station-matched SIF observation data to be solved, and applying the semi-empirical formula to various ground stations. The stations are calibrated using a least squares algorithm to obtain the estimated parameter set of the semi-empirical formula for each surface station. Based on the global land cover classification data, each surface station is classified, and station environmental variables are statistically derived from the corrected meteorological data. These station environmental variables are then used to perform multiple linear regression on the estimated parameter set under each land cover classification to obtain a parameter regression model. Based on the global land cover classification data, the global land surface is classified, and global environmental variables are statistically derived from the ERA5-Land meteorological reanalysis data. These global environmental variables are then substituted into the parameter regression model under each land cover classification to obtain the global parameter distribution. Using the global parameter distribution, a global semi-empirical formula is constructed. The ERA5-Land meteorological reanalysis data, the leaf area index data, global precipitation data, and the global reconstructed CSIF daylight-induced chlorophyll fluorescence data are substituted into the global semi-empirical formula to obtain global ET data.

[0133] Furthermore, the logical instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0134] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0135] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0136] 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 them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for estimating global evapotranspiration coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence, characterized in that, include: Collect and preprocess ERA5-Land meteorological reanalysis data, leaf area index data, global rainfall data, global land cover classification data, and global reconstructed CSIF daylight-induced chlorophyll fluorescence data; FLUXNET2015 ground station observation data is acquired, and the ET observation data and meteorological data of the ground stations are time-averaged and corrected to obtain corrected ground station ET observation data and corrected ground station meteorological data. The globally reconstructed CSIF daylight-induced chlorophyll fluorescence data is matched to each station according to the station's latitude and longitude coordinates to obtain the station-matched SIF observation data. A semi-empirical formula is constructed between the corrected ET observation data of the ground station, the corrected meteorological data of the ground station, and the SIF observation data matched by the station. The semi-empirical formula is then calibrated at various ground stations. The least squares algorithm is used to obtain the estimated parameter set of the semi-empirical formula at various ground stations. Based on the global land cover classification data, the ground stations are classified, and the station environmental variables are statistically extracted from the corrected meteorological data of the ground stations. The station environmental variables are used to perform multiple linear regression on the estimated parameter set under each land cover classification to obtain the parameter regression model. The global land surface is classified based on the global land cover classification data. Global environmental variables are statistically extracted from the ERA5-Land meteorological reanalysis data. The global environmental variables are substituted into the parameter regression model under each land cover classification to obtain the global parameter distribution. A global semi-empirical formula is constructed using the global parameter distribution. The global ET data are obtained by substituting the ERA5-Land meteorological reanalysis data, the leaf area index data, the global precipitation data, and the global reconstructed CSIF daylight-induced chlorophyll fluorescence data into the global semi-empirical formula.

2. The global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence as described in claim 1, characterized in that, Collected and preprocessed ERA5-Land meteorological reanalysis data, leaf area index data, global precipitation data, global land cover classification data, and global reconstructed CSIF daylight-induced chlorophyll fluorescence data, including: For the hourly ERA5-Land meteorological reanalysis data, temperature variables, dew point temperature variables, net radiation variables, and wind speed variables are extracted at a preset number of days time resolution. The spatial resolution of the temperature variables, dew point temperature variables, net radiation variables, and wind speed variables is unified to a preset number of degrees through spatial interpolation. For the leaf area index data with a spatial resolution of 500m, the spatial resolution is unified to a preset degree through spatial resampling; For the global rainfall data on a three-hour timescale, the raw data is labeled according to whether rainfall occurred. If rainfall occurred during that time period, it is labeled as a rainfall event. The average rainfall intensity of the rainfall events in that month is calculated on a monthly basis. For the global land cover classification data at the annual scale, a mapping from code data to actual land cover classification is constructed according to the IGBP classification standard.

3. The global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence as described in claim 1, characterized in that, FLUXNET2015 ground station observation data were acquired. Time averaging and correction were performed on the ground station ET observation data and meteorological data to obtain corrected ground station ET observation data and corrected ground station meteorological data. The globally reconstructed CSIF day-induced chlorophyll fluorescence data were then matched to each station according to their latitude and longitude coordinates to obtain the station-matched SIF observation data, including: For the hourly-scale FLUXNET2015 ground station observation data, temperature, relative humidity, saturated vapor pressure difference, and net radiation variables at a preset number of days of time resolution are extracted. Quality control bands are used to perform quality control on these variables. If the proportion of unusable data at the preset number of days exceeds 50%, that data is deleted. For the global reconstructed CSIF daylight-induced chlorophyll fluorescence data, the original data is global raster data. Based on the latitude and longitude coordinates of the FLUXNET2015 stations, the raster position of the stations is located, and the CSIF data is matched with the stations.

4. The global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence as described in claim 1, characterized in that, A semi-empirical formula is constructed between the corrected ET observation data of the ground stations, the corrected meteorological data of the ground stations, and the SIF observation data matched to the stations. The semi-empirical formula is then calibrated at various ground stations. A least squares algorithm is used to obtain the estimated parameter set of the semi-empirical formula at each ground station, including: A semi-empirical formula is constructed to connect the corrected ET observation data from the ground stations, the corrected meteorological data from the ground stations, and the SIF observation data matched to the stations. First, ET is divided into T... r E s E i The three components are modeled separately, and the calculation formulas for each component are as follows; ET three components T r E s E i The summation yields a semi-empirical formula for calculating ET. The corrected ground station ET observation data, the corrected ground station meteorological data, and the station-matched SIF observation data are input into the semi-empirical formula for calculating ET. The least squares method is used to determine the three parameters in the formula, thus obtaining the estimated parameter set of the semi-empirical formula for each ground station.

5. The global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence as described in claim 1, characterized in that, Based on the global land cover classification data, the various ground stations are classified. Station environmental variables are statistically extracted from the corrected meteorological data of the ground stations. Multiple linear regression is then performed on the estimated parameter set using these station environmental variables under each land cover classification to obtain a parameter regression model, including: First, the estimated parameter set is classified according to IGBP type using the global land cover classification data. Then, temperature (Ta), precipitation (P), net radiation (Rn), saturated vapor pressure difference (VPD), relative humidity (RH), chlorophyll fluorescence (SIF), potential evapotranspiration (PET), and normalized latitude (|LAT| / 90) are selected as environmental variables from the ground station observation data. The estimated parameter set is then subjected to multiple linear regression using the environmental variables to obtain the parameter regression model.

6. The global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence as described in claim 1, characterized in that, Based on the global land cover classification data, the global land surface is classified. Global environmental variables are statistically derived from the ERA5-Land meteorological reanalysis data. These global environmental variables are then substituted into the parametric regression model for each land cover classification to obtain the global parameter distribution, including: First, the global land area is classified according to the IGBP type using the global land cover classification data. The environmental variables are statistically extracted from the ERA5-Land meteorological reanalysis data and substituted into the multiple regression parameter model based on land cover classification to calculate the parameters of each grid cell globally, thus obtaining the global parameter distribution.

7. The global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence as described in claim 1, characterized in that, A global semi-empirical formula was constructed using the global parameter distribution. The ERA5-Land meteorological reanalysis data, the leaf area index data, the global precipitation data, and the global reconstructed CSIF daylight-induced chlorophyll fluorescence data were substituted into the global semi-empirical formula to obtain global ET data, including: After obtaining the global parameter distribution, a global semi-empirical formula can be constructed for each grid cell globally based on the structure of the semi-empirical formula. The global semi-empirical formula is then substituted with the ERA5-Land meteorological reanalysis data, the leaf area index data, the global rainfall data, and the global reconstructed CSIF daylight-induced chlorophyll fluorescence data to calculate the global ET data and its distribution.

8. A global evapotranspiration estimation system coupled with sunlight-induced chlorophyll fluorescence physiological constraints, characterized in that, include: The data acquisition module is used to collect and preprocess ERA5-Land meteorological reanalysis data, leaf area index data, global rainfall data, global land cover classification data, and global reconstructed CSIF daylight-induced chlorophyll fluorescence data. The acquisition module is used to acquire FLUXNET2015 ground station observation data, perform time averaging and correction on the ground station ET observation data and meteorological data to obtain corrected ground station ET observation data and corrected ground station meteorological data, and match the global reconstructed CSIF daylight-induced chlorophyll fluorescence data to each station according to the station's latitude and longitude coordinates to obtain the station-matched SIF observation data. The calibration module is used to construct a semi-empirical formula between the corrected ground station ET observation data, the corrected ground station meteorological data, and the station-matched SIF observation data, and to calibrate the semi-empirical formula at various ground stations. The least squares algorithm is used to obtain the estimated parameter set of the semi-empirical formula at various ground stations. The regression module is used to classify the ground stations based on the global land cover classification data, to extract station environmental variables from the corrected meteorological data of the ground stations, and to perform multiple linear regression on the estimated parameter set under each land cover classification using the station environmental variables to obtain a parameter regression model. The distribution module is used to classify the global land surface based on the global land cover classification data, to statistically extract global environmental variables from the ERA5-Land meteorological reanalysis data, and to substitute the global environmental variables into the parameter regression model under each land cover classification to obtain the global parameter distribution. The calculation module is used to construct a global semi-empirical formula using the global parameter distribution. The global semi-empirical formula is substituted with the ERA5-Land meteorological reanalysis data, the leaf area index data, the global rainfall data, and the global reconstructed CSIF daylight-induced chlorophyll fluorescence data to obtain global ET data.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the global evapotranspiration estimation method coupled with the physiological constraints of sunlight-induced chlorophyll fluorescence as described in any one of claims 1 to 8.

Images