CN105938517A - Method of estimating water use efficiency of temperate forest - Google Patents

Abstract

The invention discloses a method for estimating the water availability rate of temperate forests, which comprises the following steps: (1) obtaining satellite remote sensing data EVI of enhanced vegetation index from several flux stations respectively; Satellite remote sensing data Ts of surface temperature; (3) Obtain WUE observations from the daily flux data of flux stations; (4) Substitute the above data sets of EVI, Ts and WUE observations into the following temperate forest water availability Model calculation formula WUE=a ₀ +a ₁ EVI+a ₂ EVI·T _s ; (5) The undetermined coefficients a ₀ , a ₁ and a ₂ in the model calculation formula are estimated by the model calibration method, and the calculation formula is obtained Estimates of water availability in temperate forests. The invention adopts the easily obtained remote sensing observation data for statistical modeling, and can effectively estimate the water availability rate of the temperate forest through the model probabilistic method, which is significantly correlated with the actual observation value, thereby providing support for forest management and other applications.

Description

A method for estimating water availability in temperate forests

technical field

The invention relates to a method for estimating the water availability rate of a temperate forest.

Background technique

Forest water availability (Water Use Efficiency, WUE) is an important biophysical property of the ecosystem. It represents the amount of carbon fixed by vegetation per unit amount of water consumed. According to the definition, estimating WUE needs to obtain the gross primary product (GPP) and evapotranspiration (Evapotranspiration, ET) of the terrestrial ecosystem at the same time, and the ratio of the two is WUE, as shown in the following formula:

WUE=GPP/ET.

Although people can easily observe the above two types of fluxes at the observation site, the acquisition of GPP and ET in a large area depends on the estimation of the model. These models require more input parameters, which may lead to the accumulation of estimation errors. Therefore, people also use statistical models to estimate WUE indirectly. The calculation formula of this statistical model is as follows:

WUE＝a ₀ θ _F +a ₁ (1-e- ^0.6LAI ),

Among them, a ₀ and a ₁ are undetermined coefficients of the model; θ _F is soil field water holding capacity; LAI is vegetation leaf area index. The environmental variables θ _F and LAI in the model can theoretically be obtained through ground measurements and remote sensing observations, and the statistical relationship between environmental variables and WUE can be established to calculate WUE. However, the soil field water holding capacity θ _F in this model is difficult to obtain in a large scale and with high time efficiency, which limits the application of this model.

Contents of the invention

The purpose of the present invention is to solve the technical problem that the direct acquisition method of the current forest water availability rate has large errors, and the indirect estimation method of the statistical model is difficult to obtain the soil field water holding capacity θ _F in a large range and with high time efficiency.

In order to achieve the above object of the invention, the present invention provides a method for estimating the water availability of temperate forests, comprising the steps of:

(1) Obtain satellite remote sensing data EVI of Enhanced Vegetation Index from several flux stations;

(2) Obtain satellite remote sensing data Ts of daytime surface temperature from several flux stations;

(3) Obtain the daily flux GPP and LE from several flux stations respectively, and convert LE into ET, where GPP is the total productivity of the ecosystem, LE is the latent heat flux, and ET is the evapotranspiration, through WUE=GPP/ET Find the observed value of WUE;

(4) Substitute the EVI, Ts and WUE data sets obtained above into the following model calculation formula of temperate forest water availability:

WUE＝a ₀ ＋a ₁ EVI+a ₂ EVI·T _s ,

In the formula, WUE is the estimated value of water availability in temperate forest, a ₀ , a ₁ and a ₂ are undetermined coefficients, T _s is the multi-year average daytime surface temperature of each flux station; EVI is the enhancement at the flux station vegetation index;

(5) Estimate the undetermined coefficients a ₀ , a ₁ and a ₂ in the calculation formula of the model by the model calibration method, so as to obtain the calculation formula without undetermined coefficients of the temperate forest water availability rate, by combining different EVI, Substituting the Ts value into the calculation formula without undetermined coefficients gives the estimated value WUE of temperate forest water availability.

Further, the satellite remote sensing data EVI described in step (1) is also preprocessed to remove cloud noise, so as to obtain standard EVI time series data every 16 days.

Further, the preprocessing is Savitzky-Golay filtering.

Further, the satellite remote sensing data Ts described in step (2) is also preprocessed to obtain the 8-day average daytime surface temperature, and its annual average is calculated to obtain the annual average daytime surface temperature.

Further, the preprocessing is to eliminate invalid values and values whose daytime surface temperature is less than 5°C.

Further, the undetermined coefficients a ₀ =-0.205, a ₁ =246.505, and a ₂ =-0825 estimated by the model calibration method in step (4), thus obtaining the undetermined value of the temperate forest water availability The formula for calculating the coefficient is:

WUE=-0.205+246.505·EVI-0.825·EVI·T _s .

Further, the observed value WUE of the temperate forest water availability is obtained by the following method:

Obtain the 16-day average total ecosystem productivity GPP and the 16-day average latent heat flux LE on each of the flux sites respectively, and convert the latent heat flux value LE into evapotranspiration ET, by calculating the formula WUE=GPP/ ET yields observations of water availability for the temperate forests,

Among them, ET=LE/λ, λ is the energy consumed by liquid water per unit mass, λ≈2.454MJ/kg.

Further, the satellite remote sensing data EVI of the enhanced vegetation index described in the step (1) and the satellite remote sensing data Ts of the daytime surface temperature described in the step (2) are also processed as follows:

Through the rainfall data obtained from several flux stations, the data of the day when the rainfall is greater than 0 and the data of the two days after the rainfall are eliminated.

Further, described step (5) also includes after:

Compare the observed value described in step (3) with the estimated value described in step (5) to assess the accuracy of the estimation by the model calculation formula of temperate forest water availability described in step (4)

Compared with prior art, the beneficial effect of the present invention is:

1. Using easily obtained satellite remote sensing data and ground flux observation data for statistical modeling, the water availability rate of temperate forests can be estimated more effectively through the model calibration method, which is significantly related to the actual observation value, thus providing a basis for forest management and other applications provide support.

2. After the remote sensing data is preprocessed, it is put into the model formula for estimation, which improves the usability and estimation accuracy of the remote sensing data.

Description of drawings

Fig. 1 is a flowchart of an embodiment of the present invention;

Fig. 2 is a flowchart of another embodiment of the present invention;

Fig. 3 is a flowchart of another embodiment of the present invention;

Figure 4 is a geographical distribution map of flux sites;

Figure 5 is a graph of regression analysis results of estimated values and observed values;

Figure 6 is a time series diagram of WUE from 2004 to 2005 on station US-MMS;

Figure 7 is the time series diagram of WUE in 2005-2006 at the station US-WCr;

Figure 8 is the time series diagram of WUE in 2002-2003 on the site US-UMB;

Figure 9 is a time series diagram of WUE from 2004 to 2005 at station US-Bar.

detailed description

The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

Example 1:

As shown in Figure 1, the method for estimating temperate forest moisture availability of the present invention, comprises the steps:

S100: Obtain satellite remote sensing data EVI of Enhanced Vegetation Index from several flux stations respectively;

S110: Obtain satellite remote sensing data Ts of daytime surface temperature from several flux stations;

S120: Obtain the daily fluxes GPP and LE from several flux sites, and convert LE into ET, where GPP is the total productivity of the ecosystem, LE is the latent heat flux, and ET is the evapotranspiration, calculated by WUE=GPP/ET Get the observed value of WUE;

S130: Substituting the data set of Ts in step S100, EVI in S110 and WUE in S120 into the model calculation formula of water availability in temperate zone forest as follows:

WUE=a ₀ +a ₁ EVI+a ₂ EVI T _s ,

In the formula, WUE is the water availability efficiency of the temperate forest, a ₀ , a ₁ and a ₂ are undetermined coefficients, T _s is the annual average daytime surface temperature of each station; EVI is the enhanced vegetation index on the station;

S140: Estimate the undetermined coefficients a ₀ , a ₁ and a ₂ in the calculation formula of the model by the model calibration method, so as to obtain the calculation formula of the temperate forest water availability rate without undetermined coefficients, by combining different EVI, Ts Substituting the values into the formula without undetermined coefficients to obtain the estimated value WUE of temperate forest water availability.

Among them, satellite remote sensing data EVI refers to MODIS (Moderate Resolution Imaging Spectrometer) enhanced vegetation index EVI data MOD13Q1; satellite remote sensing data Ts refers to MODIS daytime surface temperature data MOD11A2.

Example 2:

As shown in Figure 2, the method for estimating temperate forest moisture availability of the present invention, comprises the steps:

S200: Obtain satellite remote sensing data EVI of Enhanced Vegetation Index from several flux stations, and remove the cloud noise in it through a preprocessing algorithm to obtain EVI time series data that meets the standard every 16 days; the preferred preprocessing algorithm is Savitzky– Golay filter;

S210: Obtain the satellite remote sensing data Ts of the daytime surface temperature from several flux stations, and obtain the 8-day average daytime surface temperature through a preprocessing algorithm, and calculate its annual average, so as to obtain the annual average daytime surface temperature; preferred The preprocessing algorithm is to eliminate invalid values and values whose daytime surface temperature is less than 5°C, invalid values such as blank or 0 values;

S220: Obtain the daily fluxes GPP and LE from several flux sites, and convert LE into ET, where GPP is the total productivity of the ecosystem, LE is the latent heat flux, and ET is the evapotranspiration. Calculate by WUE=GPP/ET Obtain the observed value of WUE, and through the rainfall data obtained from several flux stations, eliminate the observed value data of WUE on the day when the rainfall is greater than 0 and the observed value data of WUE two days after the rainfall;

S230: Substituting the data set obtained in Ts in step S200, EVI in S210, and WUE in S220 into the model calculation formula of water availability in temperate forests as follows:

WUE=a ₀ +a ₁ EVI+a ₂ EVI T _s ,

In the formula, WUE is the water availability efficiency of the temperate forest, a ₀ , a ₁ and a ₂ are undetermined coefficients, T _s is the annual average daytime surface temperature of each station; EVI is the enhanced vegetation index on the station;

S240: Estimate the undetermined coefficients a ₀ , a ₁ and a ₂ in the above model calculation formula through the model calibration method, and the estimated results are a ₀ = -0.205, a ₁ = 246.505, a ₂ = -0.825, thus obtaining the temperate forest The estimated value of water availability is WUE=-0.205+246.505·EVI-0.825·EVI·T _s .

Among them, satellite remote sensing data EVI refers to MODIS (Moderate Resolution Imaging Spectrometer) enhanced vegetation index EVI data MOD13Q1; satellite remote sensing data Ts refers to MODIS daytime surface temperature data MOD11A2.

Example 3:

As shown in Figure 3, the method for estimating temperate forest moisture availability of the present invention, comprises the steps:

S300: Obtain satellite remote sensing data EVI of Enhanced Vegetation Index from several flux stations, and remove the cloud noise in it through a preprocessing algorithm to obtain EVI time series data that meets the standard every 16 days; the preferred preprocessing algorithm is Savitzky– Golay filter;

S310: Obtain the satellite remote sensing data Ts of the daytime surface temperature from several flux stations, and obtain the 8-day average daytime surface temperature through a preprocessing algorithm, and calculate its annual average, so as to obtain the annual average daytime surface temperature; preferred The preprocessing algorithm is to eliminate invalid values and values whose daytime surface temperature is less than 5°C, invalid values such as blank or 0 values;

S320: Obtain the daily fluxes GPP and LE from several flux stations, and convert LE into ET, where GPP is the total productivity of the ecosystem, LE is the latent heat flux, and ET is the evapotranspiration, calculated by WUE=GPP/ET Obtain the observed value of WUE, and through the rainfall data obtained from several flux stations, eliminate the observed value data of WUE on the day when the rainfall is greater than 0 and the observed value data of WUE two days after the rainfall; The 16-day average total ecosystem productivity GPP (unit gC m ^-2 d ^-1 ) and latent heat flux LE (unit MJ m ^-2 d ^-1 ) at the site, and convert the value of latent heat flux LE into evapotranspiration ET, Observational value WUE (unit: gC kg ^-1 H ₂ O) of temperate forest water availability is obtained by calculating the formula WUE=GPP/ET; the original data is the average value every half hour, in order to be consistent with the time scale of remote sensing data ( That is, 16 days), the original data is synthesized into a 16-day average value, and the data on the day when the rainfall is greater than 0 and the data on the day after the rainfall are removed, only the arithmetic mean of the 16 days needs to be calculated for the original observation value, namely:

$\mathrm{<span\; class="notranslate">\; G</span>}\mathrm{<span\; class="notranslate">\; P</span>}\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; GPP</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; ;</span>$

$\mathrm{<span\; class="notranslate">\; L</span>}\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; IE</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; ;</span>$

Among them, GPP ₁ and LE ₁ represent the half-hour average total ecosystem productivity and latent heat flux respectively, N represents the number of original observations within 16 days, ET=LE/λ, λ is the energy consumed per unit mass of liquid water, λ≈2.454MJ/kg;

S330: Substituting the data set obtained in Ts in step S300, EVI in S310 and WUE in S320 into the model calculation formula of water availability in temperate zone forest as follows:

WUE=a ₀ +a ₁ EVI+a ₂ EVI T _s ,

In the formula, WUE is the water availability rate of temperate forest, a ₀ , a ₁ and a ₂ are undetermined coefficients;

S340: Estimate the undetermined coefficients a ₀ , a ₁ and a ₂ in the above model calculation formula through the model calibration method, and the estimated results are a ₀ = -0.205, a ₁ = 246.505, a ₂ = -0.825, thus obtaining the temperate forest The estimated value of water availability is WUE＝-0.205+246.505·EVI-0.825·EVI·T _s ;

S350: Compare the observed value of WUE in step S320 with the estimated value in step S340, so as to evaluate the accuracy of the estimation by the model calculation formula of the temperate forest water availability in step S330.

Among them, satellite remote sensing data EVI refers to MODIS (Moderate Resolution Imaging Spectrometer) enhanced vegetation index EVI data MOD13Q1; satellite remote sensing data Ts refers to MODIS daytime surface temperature data MOD11A2.

In the following, the estimation effect of the model of the present invention is evaluated by the data obtained from specific flux sites.

Figure 4 is the geographical distribution of flux stations, in which the observation data of the stations marked by solid circles are all used for model verification, part of the observation data of stations marked by solid triangles, such as 50%, are used for model calibration, and the remaining observation data are used for model verify.

The experiment uses statistical methods to evaluate the effect of the model. The main indicators used are: correlation coefficient, determination coefficient and root mean square error. The value of the correlation coefficient is between -1 and 1, indicating the degree of correlation between two variables. Negative values represent negative correlations, and positive values represent positive correlations. The closer the absolute value is to 1, the greater the degree of correlation between the two variables. The value of the coefficient of determination is also between 0 and 1, indicating the degree of interpretation of the independent variable on the dependent variable in the model, and the closer the value is to 1, the higher the degree of interpretation. The root mean square error reflects the degree of deviation between the predicted value of the model and the observed value, and the smaller the value, the higher the accuracy of the result.

The flux data used in the experiment comes from the La Thuile data set, which is a data set composed of more than 250 flux site data around the world. The observation data set includes carbon dioxide flux, water heat flux, air temperature and air humidity, etc. Earth-like surface biophysical property data. The data used in the experiment belongs to the sub-dataset that can be downloaded and used for free in the dataset.

In the experiment, the undetermined coefficients a ₀ , a ₁ and a ₂ of the model were first calibrated using the flux tower data. The flux tower data used for model calibration are shown in Table 1 below:

Table 1

Among them, the names of flux towers represent observation sites in different forest areas (as shown in Figure 4). A station chronology represents the observation data of a station for one year. Part of the data in flux towers US-MMS, US-WCr and US-UMB is used for model calibration, and the rest is used for model validation (as shown in Table 1). The model calibration includes the annual data of 11 stations in total, and the calibration results are a ₀ =-0.205, a ₁ =246.505, a ₂ =-0.825. The coefficient of determination of the model calibration population reached 0.5. Then the model equation is as follows:

WUE=-0.205+246.505·EVI-0.825·EVI·T _s .

The flux tower data used for model validation are shown in Table 2 below:

Table 2

A total of 8 site-years of data were included for model validation. Regression analysis of WUE model estimation results and observation results is shown in Figure 5, where the abscissa is the WUE observed at the flux site, and the ordinate is the model-estimated WUE. The correlation coefficient between the model estimation results and the observation results is 0.64 ,Significant correlation. The slope of the regression line is 0.985, the intercept is 0.156, close to the 1:1 line, and the root mean square error is 1.64, which is close to the accuracy of the existing estimation results.

The comparison of the predicted and observed seasonal variations of WUE at different flux stations is shown in the time series plots in Figures 6-9. In these four panels, open circles represent observed values of WUE, and solid circles represent estimated values of WUE. As can be seen from the WUE time series diagrams of different flux sites, although there are differences in the estimation accuracy of the WUE of different flux sites, their time variation trends are consistent with the observed values, and there are obvious seasonal changes, which can prove that the present invention The estimation model is feasible.

In addition to the above-mentioned embodiments, the present invention can also have other implementations, and all technical solutions formed by equivalent replacement or transformation fall within the protection scope of the present invention.

Claims (9)

1. A method for estimating temperate zone forest water availability, is characterized in that, comprises the steps: (1) Obtain satellite remote sensing data EVI of Enhanced Vegetation Index from several flux stations; (2) Obtain satellite remote sensing data Ts of daytime surface temperature from several flux stations; (3) Obtain the daily flux GPP and LE from several flux stations respectively, and convert LE into ET, where GPP is the total productivity of the ecosystem, LE is the latent heat flux, and ET is the evapotranspiration, through WUE=GPP/ET Find the observed value of WUE; (4) Substitute the EVI, Ts and WUE data sets obtained above into the following model calculation formula of temperate forest water availability: WUE=a ₀ +a ₁ EVI+a ₂ EVI T _s , In the formula, WUE is the estimated value of water availability of temperate forest, a ₀ , a ₁ and a ₂ are undetermined coefficients, T _s is the multi-year average daytime surface temperature of each flux station; EVI is the enhancement at the flux station vegetation index; (5) Estimate the undetermined coefficients a ₀ , a ₁ and a ₂ in the calculation formula of the model by the model calibration method, so as to obtain the calculation formula without undetermined coefficients of the temperate forest water availability rate, by combining different EVI, Substituting the Ts value into the calculation formula without undetermined coefficients gives the estimated value WUE of temperate forest water availability. 2. The method according to claim 1, characterized in that, the satellite remote sensing data EVI of step (1) is also carried out to remove the preprocessing of cloud noise, so as to obtain the EVI time series data of every 16 days up to the standard. 3. The method according to claim 2, wherein the preprocessing is Savitzky-Golay filtering. 4. the method for claim 1, is characterized in that, also carry out preprocessing to described satellite remote sensing data Ts of step (2), to obtain 8 days average daytime surface temperature, and calculate its annual average value, thereby obtain Annual mean daytime surface temperature. 5 . The method according to claim 4 , wherein the preprocessing is to eliminate invalid values and values whose daytime surface temperature is less than 5° C. 6. The method according to claim 1, characterized in that, the undetermined coefficients a ₀ =-0.205, a ₁ =246.505, a ₂ =-0.825 estimated by the model calibration method in step (4), Therefore, the formula for calculating the undetermined coefficient of water availability in temperate forests is as follows: WUE=-0.205+246.505·EVI-0.825·EVI·T _s . 7. The method according to claim 1, characterized in that, the observed value WUE of the temperate forest water availability is obtained by the following methods: Obtain the 16-day average total ecosystem productivity GPP and the 16-day average latent heat flux LE on each of the flux sites respectively, and convert the latent heat flux value LE into evapotranspiration ET, by calculating the formula WUE=GPP/ ET yields observations of water availability for the temperate forests, Among them, ET=LE/λ, λ is the energy consumed by liquid water per unit mass, λ≈2.454MJ/kg. 8. the method for claim 1 is characterized in that, also to the satellite remote sensing data EVI of enhanced vegetation index described in step (1) and the satellite remote sensing data Ts of daytime surface temperature described in step (2) as follows deal with: Through the rainfall data obtained from several flux stations, the data of the day when the rainfall is greater than 0 and the data of the two days after the rainfall are eliminated. 9. method as claimed in claim 1 is characterized in that, described step (5) also comprises afterwards: Compare the observed value described in step (3) with the estimated value described in step (5) to evaluate the accuracy of the estimation by the model calculation formula of temperate forest water availability described in step (4).