A T m -f c Calculation method for extreme temperature combinations within a trapezoidal frame

Abstract

The present invention provides a T m ‑f c The calculation method for selecting extreme temperatures within the trapezoidal framework uses the pixel-by-pixel error analysis method PPEA to evaluate the effects of the four extreme temperatures on the final LE and LE v / LE simulation. The pixel-by-pixel error analysis method PPEA specifically includes: (1) m ‑f c The two-dimensional scatter plot is divided into a 100*100 grid; (2) For T in the 2D coordinate system m ‑f c Four corners of the scatter points, select n scatter points for each corner, where n≤10; (3) The total number of extreme temperature combinations is n 4 group; using (actual observation data) observed LE data (LE o ,W / m 2 ) Verify the simulated LE value of each extreme temperature combination, and perform error analysis on each extreme temperature combination by finding the minimum RMSE value as the objective function; 4 The extreme temperature combination corresponding to the minimum RMSE value among the RMSEs is the theoretical extreme temperature.

Description

Technical Field

[0001] The present invention belongs to a method based on a feature space model (T m -f c Trapezoidal frame) remote sensing evapotranspiration calculation field, specifically involving a T m -f c Calculation method for extreme temperatures within a trapezoidal frame. Background Art

[0002] A key step in applying the ladder frame model is to determine T m -f c The absolute dry / wet boundaries of the trapezoid are determined and the theoretical extreme temperatures at its four vertices are located. There are currently three main methods for estimating T m -f c Four theoretical extreme temperatures of the trapezoid: 1) Positioning T m -f c The four actual extreme temperatures (AET) of the trapezoid. This method first plots T m and f c The scatter plot is plotted in a two-dimensional x and y coordinate system, and the upper / lower limits of the scatter distribution are determined by experience or by eye, and the four extreme temperatures are located at T m -f c The scattered points inside the trapezoid closest to the four vertices. In order to reduce the uncertainty caused by experience and visual judgment, some researchers have proposed various automatic boundary location algorithms; 2) Locating T m -f c The four theoretical extreme temperatures (TET) of the trapezoidal shape are directly estimated by the energy balance equation and combined with ground-based meteorological factors without the need for T m -f c The actual boundary of the trapezoid is located; 3) The method combining the actual boundary and the theoretical boundary (ATET). This method first identifies T m -f c The actual boundary of the trapezoid then determines the four actual extreme temperatures, on this basis, the four actual extreme points are coupled to the surface energy balance equation to obtain four theoretical extreme temperatures. The AET method initially relied on visual judgment of T m -f c The four extreme pixel points in the two-dimensional scatter plot.

[0003] The three extreme temperature determination methods, AET, TET and ATET, have been widely used in various types of underlying surfaces and surface types. m -f cThe trapezoidal extreme temperatures are being determined, but the problems in the current research are: 1) Can the extreme temperatures determined by these three methods represent the actual extreme conditions in a specific area well? 2) What are the differences between the extreme temperature estimation and the final LE and LE? v / LE simulations differ from each other? 3) What are the sources of differences among the three methods? 4) How to choose the appropriate method for determining extreme temperatures in a study area with various spatial scales and surface heterogeneity?

[0004] The AET method initially relied on visual judgment of T m -f c Four extreme pixel points in the two-dimensional scatter plot of T. m -f c The trapezoid is divided into several vertical strips, and then the highest / lowest temperature of the upper and lower boundaries of each strip is calculated, and finally the upper and lower boundaries of the trapezoid are located using a linear regression method. The inventor has modified this method by dividing the horizontal strips on the basis of dividing the vertical strips and considering the density distribution of each grid, thereby discarding the grids with lower density distribution. Summary of the invention

[0005] In order to fully evaluate the effects of the four extreme temperatures on the final LE and LE v / LE simulation, the present invention proposes a point-by-point error analysis method (PPEA) to analyze the effect of T m -f c The impact of the possibility of all extreme temperature combinations within the trapezoidal framework on the final flux simulation results.

[0006] To achieve the above purpose, the technical solution of the present invention is as follows:

[0007] A T m -f c The calculation method of extreme temperature combinations in the trapezoidal framework uses the pixel-by-pixel error analysis method PPEA to evaluate the effects of four extreme temperatures on the final LE and LE v / LE simulation, specifically to accurately determine the extreme temperature and eliminate the error caused by determining the extreme temperature on the final LE and LE v / LE simulation effects.

[0008] The specific steps of the pixel-by-pixel error analysis method PPEA are as follows:

[0009] (1) T m -f c The two-dimensional scatter plot is divided into a 100*100 grid;

[0010] (2) For T in the 2D coordinate system m -f c Four corners of the scattered points, select n scattered points for each corner, where n≤10;

[0011] (3) The total number of extreme temperature combinations obtained from step (2) is n 4 group; using LE data observed at flux sites (LE o ,W / m 2 ) The simulated LE value of each extreme temperature combination is verified, and the error analysis of each extreme temperature combination is performed by finding the minimum RMSE value as the objective function;

[0012] (4)n 4 The extreme temperature combination corresponding to the minimum RMSE value among the RMSEs is the theoretical extreme temperature.

[0013] n 4 The RMSE value calculation formula for the extreme temperature combination is as follows:

[0014]

[0015] Among them, LE o is the LE data observed at the flux site (W / m 2 ), N: total number of samples, which should be n here 4 ; LEi: latent heat flux of the i-th sample: RMSE: root mean square error.

[0016] The analog value of LE (LE p ,W / m 2 ) is calculated by PCACA. PCACA refers to Pixel Component Arranging and Comparing Algorithm.

[0017] Advantages of the present invention:

[0018] The present invention divides the horizontal strips again on the basis of dividing the vertical strips. m -f c The two-dimensional scatter plot is divided into several grids. By calculating the RMSE value, the extreme temperature combination is accurately selected, the extreme temperature is accurately determined, and the error caused by determining the extreme temperature is eliminated. v / LE simulation, eliminating the reliance on naked eye judgment of T m -f c The two-dimensional scatter plot shows the extreme uncertainty of the four pixels. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 : Scatter plot of surface temperature and vegetation cover. DETAILED DESCRIPTION

[0021] Example 1

[0022] This embodiment provides a T m -f c The calculation method of extreme temperature combinations in the trapezoidal frame, the calculation process of the PPEA method is as follows:

[0023] 1) T m -f c The two-dimensional scatter plot is divided into a 23*23 grid. The purpose of dividing the grid is to assist in selecting potential extreme pixels; Figure 1 As shown;

[0024] 2) For T in the 2D coordinate system m -f c The four corners of the scatter point, select "n" scatter points for each corner. These 4n scatter points represent T m -f c There are four possible extreme cases in the trapezoid. Figure 1 In the circled area, n is 6;

[0025] 3) The total number of extreme temperature combinations obtained from step 2 is n 4 In order to select the four absolute extreme pixels, the observed LE data (LE o ,W / m 2 ) The simulated LE value of each extreme temperature combination is verified, and the error analysis of each extreme temperature combination is performed by finding the minimum RMSE value as the objective function. First, for each extreme temperature combination, the simulated value of LE is calculated by PCACA (LE p ,W / m 2 ), and calculate RMSE by equation (1), thus obtaining n 4 RMSE value for the extreme temperature combination.

[0026]

[0027] 4)n 4The extreme temperature combination corresponding to the minimum RMSE value among the RMSEs is the theoretical extreme temperature.

[0028] It should be pointed out that in PPEA, if the setting of "n" is large, PPEA requires a lot of computing resources. To a certain extent, the setting of "n" depends on the operator's experience and subjectivity. In order to balance the computing time and the potential error of the model, this study recommends that the maximum setting of "n" should not exceed 10 (when set to 10, 10 4 extreme temperature combinations).

Claims

1. A T m -f c The calculation method for the selection of extreme temperatures within a trapezoidal frame is characterized in that The calculation method uses pixel-by-pixel error analysis (PPEA) to evaluate the effects of four extreme temperature selections on the final LE and LE v / LE simulation, specifically, using accuracy to determine the extreme temperature and eliminating the error caused by determining the extreme temperature on the final LE and LE v / LE simulation; where LE is the latent heat flux, LE v is the latent heat flux on the vegetation surface (W / m 2 ), Tm and fc represent the maximum temperature and vegetation coverage, respectively. The specific steps of the pixel-by-pixel error analysis method PPEA are as follows: (1) T m -f c The two-dimensional scatter plot is divided into a 100*100 grid; (2) For T in the 2D coordinate system m -f c Four corners of the scatter points, select "n" scatter points for each corner; (3) The total number of extreme temperature combinations obtained from step (2) is n 4 group; using LE data observed at flux sites (LE o ,W / m 2 ) The simulated LE value of each extreme temperature combination is verified, and the error analysis of each extreme temperature combination is performed by finding the minimum RMSE value as the objective function; (4)n 4 The extreme temperature combination corresponding to the minimum RMSE value among the RMSEs is the theoretical extreme temperature; 4 The RMSE value calculation formula for the extreme temperature combination is as follows: Among them, LE o is the LE data observed using actual observation data (W / m 2 ), where N is the total number of samples; LEi: latent heat flux of the i-th sample; RMSE: root mean square error; the total number of extreme temperature combinations obtained from step 2 is n 4 In order to select the four absolute extreme pixels, the LE data (LE o ,W / m 2 ) The simulated LE value of each extreme temperature combination is verified, and the error analysis of each extreme temperature combination is performed by finding the minimum RMSE value as the objective function; first, for each extreme temperature combination, the simulated value of LE is calculated by PCACA (LE p ,W / m 2 ), and calculate RMSE by the RMSE value calculation formula to get n 4 The RMSE value of the extreme temperature combination; PCACA refers to the pixel component arranging and comparing algorithm.

2. According to claim 1, m -f c The method for calculating the extreme temperature combinations in a trapezoidal frame is characterized in that: Said n≤10.

Images