Atmospheric pressure spatial downscaling method based on atmospheric static equation

Abstract

The application discloses a pressure space downscaling method based on an atmospheric static equation, and comprises the following steps: obtaining a third polar region TPMFD and preprocessing; converting daily average temperature data, daily average pressure data and specific humidity data of the third polar region TPMFD after preprocessing into relative humidity data, and substituting the relative humidity data and an elevation difference between low resolution and target resolution into a pressure-height formula, i.e., a relationship between pressure and height, to perform downscaling based on the pressure-height formula, so as to obtain 500m resolution daily average pressure data and realize spatial downscaling of pressure; and comparing pressure data of the third polar region TPMFD and pressure data of the obtained high resolution image with station pressure data. The application realizes obtaining high resolution pressure data through low resolution data and an elevation difference between high and low resolution pixels, has good applicability in pressure downscaling, and expands a physical spatial downscaling method for studying pressure.

Description

Technical Field

[0001] This invention pertains to pressure spatial downscaling methods, specifically a pressure spatial downscaling method based on atmospheric hydrostatic equations. Background Technology

[0002] Spatial scale is a core issue in remote sensing science, and different research fields have different requirements for the spatial scale of remote sensing. Due to the limitations of remote sensors, the spatial resolution of remote sensing products is mostly large, which cannot meet the needs of precision. Therefore, spatial downscaling is needed to improve the spatial resolution of remote sensing data.

[0003] Currently, there are various spatial downscaling methods. These mainly include multi-parameter fusion methods, spatiotemporal fusion methods, remote sensing-model fusion methods, and implicit fusion methods for super-resolution reconstruction. Different downscaling methods have different advantages and disadvantages and are suitable for different spatial downscaling scenarios.

[0004] Given the advantages of physical downscaling methods in multi-parameter fusion, such as theoretical rigor, relatively stable models, and physical interpretability, downscaling through a defined physical model has become a popular direction for spatial downscaling in recent years.

[0005] Spatial downscaling using physical downscaling methods mainly focuses on soil moisture, with less research on air pressure. However, air pressure is greatly affected by temperature and humidity, making it more advantageous for physical downscaling. Physical downscaling methods for air pressure are also more stable, so physical downscaling methods for air pressure are considered. Summary of the Invention

[0006] Purpose of the invention: In order to overcome the shortcomings of the existing technology, the purpose of this invention is to provide a simple and stable method for atmospheric pressure space downscaling based on atmospheric hydrostatic equations.

[0007] Technical solution: The present invention provides a method for spatial downscaling of atmospheric pressure based on atmospheric hydrostatic equations, comprising the following steps:

[0008] Step 1: Obtain the long-term high-resolution surface meteorological element driving dataset (TPMFD) of the target area's Third Pole region and perform preprocessing;

[0009] Step 2: Using the atmospheric hydrostatic equation and Dalton's law of partial pressures, the relationship between air pressure and altitude is derived by introducing geopotential height. The daily average temperature data, daily average air pressure data, and relative humidity data converted from specific humidity data of the preprocessed high-resolution surface meteorological element driven dataset (TPMFD) of the Third Pole region are substituted with the elevation difference between the low resolution and target resolution into the relationship between air pressure and altitude, i.e., the pressure-height formula. Downscaling is then performed based on the pressure-height formula to obtain daily average air pressure data at a resolution of 500m, thus achieving spatial downscaling of air pressure.

[0010] Step 3: Compare the air pressure data from the long-term high-resolution surface meteorological element driven dataset (TPMFD) of the Third Pole region and the air pressure data from the resulting high-resolution images with the station air pressure data.

[0011] Furthermore, in step one, preprocessing includes image cropping and resampling.

[0012] Furthermore, image cropping involves cropping the long-term high-resolution surface meteorological element-driven dataset (TPMFD) of the Third Pole region to the same extent.

[0013] Furthermore, resampling involves bilinear interpolation resampling of the air pressure data, temperature data, and specific humidity data to transform the air pressure data, temperature data, and specific humidity data in the TPMFD dataset into the target resolution.

[0014] Furthermore, in step two, the relationship between air pressure and altitude is as follows:

[0015]

[0016]

[0017] in, For elevation difference, The interlayer average virtual temperature, , The air pressure at two altitudes, The interlayer mean absolute temperature. The average relative humidity is... Indicates temperature The corresponding saturated vapor pressure, The average air pressure is (hPa).

[0018] Furthermore, in step two, the scaling down based on the pressure height formula is as follows:

[0019]

[0020] in, Indicates high-resolution air pressure. This indicates low-resolution air pressure. This represents the elevation difference between high-resolution pixels and low-resolution pixels. For low-resolution average absolute temperature, This represents the low-resolution average relative humidity. Indicates temperature The corresponding saturated vapor pressure.

[0021] Furthermore, low-resolution mean absolute temperature With low-resolution average Celsius temperature The relationship is: .

[0022] Furthermore, saturated water vapor pressure The saturated vapor pressure is calculated using the Magnus formula based on the horizontal plane: .

[0023] Furthermore, in step two, the formula for converting the specific humidity data of TPMFD into relative humidity data is as follows:

[0024]

[0025] in, It refers to relative humidity. It's air pressure. It is more wet. It's the temperature.

[0026] Furthermore, in step three, the formula for calculating the coefficient of determination obtained through comparison is as follows:

[0027]

[0028] Where xᵢ represents the i-th observation (obs) and yᵢ represents the i-th predicted value (pre). This represents the mean of the observed values. This represents the mean of the predicted values.

[0029] Beneficial effects: Compared with the prior art, the present invention has the following significant features:

[0030] 1. The problem of spatial downscaling of air pressure was solved by using the pressure height formula, and high-resolution air pressure data was obtained by using low-resolution data and the elevation difference between high and low resolution pixels. It has good applicability in air pressure downscaling and expands the physical spatial downscaling method for studying air pressure.

[0031] 2. This method for pressure downscaling only requires low-resolution pressure data, temperature data, relative humidity data, and the difference between the low-resolution elevation and the target resolution elevation; the required data is relatively simple. Furthermore, the key steps involve mathematical calculations, making the operation straightforward.

[0032] 3. The air pressure data from the long-term high-resolution surface meteorological element driven dataset (TPMFD) of the Third Pole region and the air pressure data from the derived high-resolution images were compared with the station air pressure data. It can be seen that the accuracy of the two products is comparable, so the results of this method are relatively stable. Attached Figure Description

[0033] Figure 1 This is a research technology roadmap provided by the present invention;

[0034] Figure 2 These are air pressure comparison images provided by this invention on January 1, 2009, where a is the original TPMFD air pressure image and b is the downscaled air pressure image.

[0035] Figure 3 The present invention provides scatter plots between TPMFD and station data, and scatter plots between downscaled results and station data. Specifically, a is a scatter plot between TPMFD and station data at Namtso station in 2009; b is a scatter plot between downscaled results and station data at Namtso station in 2009; c is a scatter plot between TPMFD and station data at Mount Everest station in 2009; d is a scatter plot between downscaled results and station data at Mount Everest station in 2009; e is a scatter plot between TPMFD and station data at high-altitude stations in southeastern Tibet in 2009; and f is a scatter plot between downscaled results and station data at high-altitude stations in southeastern Tibet in 2009. Detailed Implementation

[0036] Unless otherwise specified, all materials and reagents used in the following embodiments are commercially available. Experimental methods not specifically described in the embodiments are generally performed under standard conditions or as recommended by the manufacturer.

[0037] Table 1. Research data provided in this embodiment.

[0038]

[0039] Based on the atmospheric hydrostatic equation and Dalton's law of partial pressures, and by simultaneously introducing geopotential height, the relationship between air pressure and altitude is derived—that is, the pressure-height formula—for spatial downscaling of air pressure. For example... Figure 1 Specifically, it includes the following steps:

[0040] Step 1: Data Preparation. This invention focuses on the Third Pole region; it uses the TPMFD dataset and other similar datasets as primary data sources, and data from the Yarlung Tsangpo River stations as validation data. The research data includes... Figure 2 .

[0041] Step 2, Data Preprocessing. The TPMFD dataset is acquired and preprocessed. The preprocessing primarily involves image cropping and resampling. Resampling involves bilinear interpolation resampling of the air pressure, temperature, and specific humidity data to convert the air pressure, temperature, and specific humidity data in the TPMFD dataset (3km) to the target resolution (500m).

[0042] Step 3: Obtain data and verify accuracy.

[0043] Using the atmospheric hydrostatic equations—specifically (in For the density of the gas parcel, It is the acceleration due to gravity. For air pressure, (Height) and Dalton's law of partial pressures—specifically, in any ideal gas mixture within a container, if the components do not chemically react with each other, each gas, when uniformly distributed throughout the container, produces the same pressure as it would produce if it occupied the entire container alone. Simultaneously, the relationship between gas pressure and altitude is derived using the potential height, i.e., the pressure-height formula, which is:

[0044]

[0045]

[0046] in The elevation difference is in meters (m). The interlayer average virtual temperature (K) , The air pressure (hPa) at the two altitudes. The interlayer mean absolute temperature (K) is given. The average relative humidity is (%). Indicates temperature The corresponding saturated water vapor pressure (hPa). The average air pressure is (hPa).

[0047] (1) Considering that the low-resolution elevation is derived from the high-resolution elevation, and there is an elevation difference between the low-resolution elevation and the high-resolution elevation within its pixel range, downscaling based on the pressure-height formula is performed using this elevation difference and the low-resolution air pressure. During formula conversion, considering that the low-resolution air pressure is derived from the high-resolution air pressure, the average air pressure in the pressure-height formula is approximated by the low-resolution air pressure; the mathematical expression is:

[0048]

[0049] in, This indicates high-resolution atmospheric pressure (hPa). This indicates low-resolution atmospheric pressure (hPa). This represents the elevation difference (m) between high-resolution pixels and low-resolution pixels. This represents the low-resolution average absolute temperature (K). Low-resolution average relative humidity (%) Indicates temperature The corresponding saturated vapor pressure (hPa).

[0050] Low-resolution mean absolute temperature (K) and low-resolution average Celsius temperature The relationship (°C):

[0051] Saturated vapor pressure (hPa) is calculated using the Magnus saturated vapor pressure formula based on a horizontal plane:

[0052]

[0053] Since the humidity data in the TPMFD dataset is specific humidity data, it is necessary to convert the specific humidity data into relative humidity data.

[0054]

[0055] in It is air pressure (Pa). It is the specific moisture content (kg). It refers to the air temperature (°C).

[0056] By substituting the air pressure data, temperature data, and relative humidity data converted from specific humidity data into the TPMFD dataset (3km) after bilinear interpolation resampling, as well as the elevation difference between the low resolution (3km) and the target resolution (500m), into the expression, air pressure data at a resolution of 500m is obtained, thus achieving spatial downscaling of air pressure.

[0057] (2) Accuracy verification: Accuracy verification was performed using air pressure data from three stations in the Yarlung Tsangpo River station data: Namtso, Mount Everest, and Southeast Tibet. The air pressure data from the TPMFD and the resulting high-resolution images were compared with the station air pressure data, and the coefficient of determination R was used. 2 They are almost identical. The formula for calculating the coefficient of determination is:

[0058]

[0059] in Let represent the i-th observation (obs). This represents the i-th predicted value (pre). This represents the mean of the observed values. This represents the mean of the predicted values.

[0060] like Figure 3 The coefficients of determination (CCRs) for the air pressure data from the Namtso station and the air pressure data from the derived high-resolution images compared to the station's air pressure data were 0.115 and 0.198, respectively, showing a slight improvement in accuracy. For the Mount Everest station, the CCRs for the air pressure data from the TPMFD and the derived high-resolution images compared to the station's air pressure data were 0.486 and 0.458, respectively, showing a slight decrease in accuracy. For the southeastern Tibetan high-altitude stations, the CCRs for the air pressure data from the TPMFD and the derived high-resolution images compared to the station's air pressure data were 0.303 and 0.365, respectively, showing a slight improvement in accuracy. Overall, the accuracy of the derived data is comparable to that of TPMFD, indicating that the results of this method are relatively stable.

Claims

1. A pressure spatial downscaling method based on atmospheric hydrostatic equations, characterized in that, Includes the following steps: Step 1: Obtain a long-term high-resolution surface meteorological element-driven dataset of the target area's Third Pole region and perform preprocessing. Step 2: Using the atmospheric hydrostatic equation and Dalton's law of partial pressures, the relationship between air pressure and altitude is derived by introducing geopotential height. The daily average temperature data, daily average air pressure data, and relative humidity data converted from specific humidity data of the preprocessed high-resolution surface meteorological element-driven dataset of the Third Pole region are substituted with the elevation difference between the low resolution and target resolution into the relationship between air pressure and altitude, i.e., the pressure-height formula. Downscaling is then performed based on the pressure-height formula to obtain daily average air pressure data at a resolution of 500m, thus achieving spatial downscaling of air pressure. Step 3: Compare the air pressure data from the long-term high-resolution surface meteorological element-driven dataset of the Third Pole region and the air pressure data from the resulting high-resolution images with the station air pressure data.

2. The pressure spatial downscaling method based on atmospheric hydrostatic equations according to claim 1, characterized in that: In step one, preprocessing includes image cropping and resampling.

3. The pressure spatial downscaling method based on atmospheric hydrostatic equations according to claim 2, characterized in that: The image cropping involves cropping a long-term, high-resolution surface meteorological element-driven dataset of the Third Pole region to the same extent.

4. The pressure spatial downscaling method based on atmospheric hydrostatic equations according to claim 2, characterized in that: The resampling involves bilinear interpolation resampling of the air pressure data, temperature data, and specific humidity data to convert the air pressure data, temperature data, and specific humidity data in the TPMFD dataset into the target resolution.

5. The pressure spatial downscaling method based on atmospheric hydrostatic equations according to claim 1, characterized in that: In step two, the relationship between air pressure and altitude is as follows: in, For elevation difference, The interlayer average virtual temperature, At a certain altitude, The air pressure at another altitude, The interlayer mean absolute temperature. The average relative humidity is... Indicates temperature The corresponding saturated vapor pressure, The average air pressure is (hPa).

6. The pressure spatial downscaling method based on atmospheric hydrostatic equations according to claim 1, characterized in that: In step two, the scaling down based on the pressure height formula is as follows: in, Indicates high-resolution air pressure. This indicates low-resolution air pressure. This represents the elevation difference between high-resolution pixels and low-resolution pixels. For low-resolution average absolute temperature, This represents the low-resolution average relative humidity. Indicates temperature The corresponding saturated vapor pressure.

7. The pressure spatial downscaling method based on atmospheric hydrostatic equations according to claim 6, characterized in that: The low-resolution mean absolute temperature With low-resolution average Celsius temperature The relationship is: .

8. The pressure spatial downscaling method based on atmospheric hydrostatic equations according to claim 7, characterized in that: The saturated water vapor pressure The saturated vapor pressure is calculated using the Magnus formula based on the horizontal plane: .

9. The pressure spatial downscaling method based on atmospheric hydrostatic equations according to claim 1, characterized in that: In step two, the formula for converting the specific humidity data of TPMFD into relative humidity data is as follows: in, It refers to relative humidity. It's air pressure. It is more wet. It's the temperature.

10. A pressure spatial downscaling method based on atmospheric hydrostatic equations according to claim 1, characterized in that: In step three, the formula for calculating the determination coefficient obtained through comparison is as follows: in Let represent the i-th observation (obs). This represents the i-th predicted value (pre). This represents the mean of the observed values. This represents the mean of the predicted values.

Images