A method for generating a near-surface air temperature lapse rate based on MODIS data

Abstract

The application discloses a kind of near-surface air temperature direct reduction rate generation method based on MODIS data, belong to remote sensing near-surface air temperature direct reduction rate generation technical field.The present application directly generates near-surface air temperature data using MODIS data, reduces the uncertainty brought by many auxiliary parameters in the generation process of near-surface air temperature;Using moving window convolution, strictly control the data in moving window, and then generate high temporal resolution, spatially continuous near-surface air temperature direct reduction rate product.The present application reduces the dependence on a variety of auxiliary ground variables and meteorological station measured values in the traditional near-surface air temperature direct reduction rate estimation method;The present application judges whether the field pixel of remote sensing data meets the requirements pixel by pixel, changes the size of moving window according to the judgment result, strictly controls the input data of model, to obtain credible daily spatially continuous SATLR;The near-surface air temperature direct reduction rate produced by the present application has the characteristics of high temporal resolution, spatially continuous, etc.

Description

Technical Field

[0001] This invention belongs to the field of remote sensing near-surface air lapse rate generation technology, specifically involving a method for generating near-surface air lapse rate based on MODIS (Moderate-resolution Imaging Spectroradiometer) data. Background Technology

[0002] Near-surface air temperature lapse rates (SATLR) refer to the rate at which air temperature at a height of 2 meters above the Earth's surface decreases with increasing altitude. SATLR can be used to analyze the spatial distribution characteristics of air temperature in mountainous areas where near-surface air temperature data is scarce, and to simulate the surface environment. It is of great significance for the study of mountain climate change and hydrological processes. In addition, SATLR research can improve the accuracy of regional near-surface air temperature downscaling and provide input data support for the development of high spatial resolution near-surface air temperature datasets. Currently, commonly used SATLR estimation methods are based on simple linear regression using station-measured air temperature and elevation data. A few studies have also used multiple linear regression models to estimate SATLR, that is, to establish a multiple regression model with multiple independent variables such as elevation, longitude, and latitude, and to use the coefficient before the elevation independent variable as the SATLR. However, the above-mentioned SATLR estimation studies are difficult to depict the spatially continuous SATLR over large areas. With the continuous advancement of remote sensing technology, some scholars have applied satellite thermal infrared remote sensing to the acquisition of large-scale near-surface air temperature data, which has made it possible to quantitatively obtain spatially continuous SATLR. However, the following problems still need to be addressed: (1) In most methods for estimating near-surface air temperature based on remote sensing thermal infrared surface temperature data, elevation is used as an input parameter. At the same time, SATLR is also the regression coefficient of the elevation influence factor in the near-surface air temperature estimation model, and SATLR has been indirectly incorporated into the estimation model in the process of estimating near-surface air temperature. Therefore, most near-surface air temperature products estimated based on remote sensing thermal infrared surface temperature cannot be used to estimate SATLR. (2) In the process of estimating near-surface air temperature, it is often necessary to use a variety of auxiliary surface variables and station measured temperature values ​​as auxiliary parameter inputs in order to achieve high-precision near-surface air temperature estimation, which will cause great uncertainty. Summary of the Invention

[0003] This invention provides a method for generating near-surface temperature lapse rate based on MODIS data, which is used to generate near-surface temperature lapse rate products with high temporal resolution and spatial continuity, thereby reducing the uncertainty caused by many auxiliary parameters in the near-surface temperature generation process.

[0004] The technical solution adopted in this invention is as follows:

[0005] A method for generating near-surface air lapse rate based on MODIS data, comprising the following steps:

[0006] Step 1: Extract contour data based on MODIS data of the target region, perform data preprocessing on the extracted contour data to unify the spatiotemporal resolution of the data, and perform spatiotemporal matching.

[0007] The contour data includes: atmospheric temperature contour, atmospheric pressure contour, surface temperature, and surface air pressure;

[0008] Step 2: According to the formula Estimate the near-surface temperature T in the target area a1 ;

[0009] in, This represents the atmospheric pressure closest to the Earth's surface in the atmospheric pressure profile obtained in step 1. express The surface air pressure at the next higher altitude, i.e. The corresponding upper atmospheric pressure, P S This indicates that step 1 involves obtaining the surface air pressure; Indicates correspondence Atmospheric temperature, Indicates correspondence Atmospheric temperature;

[0010] Step 3: Estimate the near-surface temperature T of the target area. a1 Perform averaging parameter processing:

[0011]

[0012] Among them, T a T represents the near-surface air temperature estimated by averaging parameter processing. s This represents the near-surface temperature obtained in step 1;

[0013] Step 4, based on near-surface air temperature T a Linear regression with elevation data was used to estimate the near-surface air lapse rate pixel by pixel using a sliding window method;

[0014] Among them, near-surface temperature T a The linear relationship between T and elevation data is: a = a×Z+b, where Z represents elevation, a and b represent the slope and intercept of the linear equation, and a is used to characterize the estimated near-surface air lapse rate;

[0015] When estimating the near-surface temperature lapse rate pixel by pixel, it is determined whether the pixel values ​​within the sliding window meet the preset test conditions. If not, the sliding window size is adjusted until the pixel values ​​within the sliding window meet the preset test conditions, thus obtaining the optimal sliding window size. Then, the near-surface temperature T within the sliding window is calculated. a Linear regression is performed with elevation Z, and the obtained 'a' is used as an estimate of the near-surface temperature lapse rate of the central pixel within the current sliding window to generate a daily spatially continuous value.

[0016] Furthermore, the specific test conditions in step 4 are as follows: the significance level is greater than or equal to the specified value (the significance level is generally a very small number, such as 0.1, 0.05, etc.), and the elevation difference is greater than or equal to the specified value (the specific value is an empirical value; the elevation difference is the difference between the maximum and minimum values ​​of the pixels within the moving window. It is generally believed that when the elevation difference is greater than 10m, the estimated temperature lapse rate is more reliable), and the number of effective pixels (pixel values ​​meet the value limit, that is, the pixel values ​​are within a reasonable range) within the sliding window is greater than or equal to the specified value.

[0017] Furthermore, in step 4, the preferred initial size of the sliding window is set to 11×11.

[0018] The technical solution provided by this invention brings at least the following beneficial effects:

[0019] (1) This invention reduces the dependence on various auxiliary surface variables and meteorological station measured values ​​in traditional near-surface temperature lapse rate estimation methods;

[0020] (2) This invention determines whether the neighboring pixels of remote sensing data meet the requirements on a pixel-by-pixel basis, and changes the size of the moving window according to the determination result, strictly controlling the input data of the model in order to obtain reliable daily spatial continuity SATLR.

[0021] (3) The near-surface air temperature lapse rate produced by the present invention has the characteristics of high time resolution and spatial continuity. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a schematic diagram of the processing flow of a method for generating near-surface air lapse rate based on MODIS data provided in an embodiment of the present invention;

[0024] Figure 2In this embodiment of the invention, the near-surface temperature verification results are based on the measured temperature at the station. Figure (a) shows the verification results of estimating the near-surface temperature using the atmospheric pressure profile and temperature profile provided by MOD07_L2, and Figure (b) shows the verification results of the near-surface temperature after processing with the average parameter scheme. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0026] This invention provides a method for generating near-surface temperature lapse rate based on MODIS data. The purpose is to provide a method for generating near-surface temperature lapse rate data using MODIS atmospheric products. In this embodiment, near-surface temperature data is directly generated using the MOD07_L2 product (MODIS atmospheric product), reducing the uncertainty caused by numerous auxiliary parameters during the near-surface temperature generation process. Moving window convolution is used to strictly control the data within the moving window, thereby generating a high temporal resolution and spatially continuous near-surface temperature lapse rate product.

[0027] like Figure 1 As shown, the near-surface air lapse rate generation method based on MODIS data provided in this embodiment of the invention can be divided into four stages:

[0028] (1) Use MOD07_L2 atmospheric products to estimate the near-surface temperature of the target area (in this embodiment, the Qinghai-Tibet Plateau is used as the target area).

[0029] MOD07_L2 provides temperature data at 20 altitudes, but these data lack a direct correlation with near-surface temperatures. Based on the characteristic that temperature decreases with increasing altitude, near-surface temperatures are estimated using the atmospheric temperature profile, atmospheric pressure profile, and surface pressure provided by MOD07_L2, using the following formula:

[0030]

[0031] In the formula, T a1 This represents the estimated temperature using MOD07_L2; P L1 This is the atmospheric pressure closest to the Earth's surface among the 20 altitudes of the MOD07_L2 product; yes The corresponding upper atmospheric pressure; P S The surface atmospheric pressure is provided by MOD07_L2; Indicates correspondence Atmospheric temperature, Indicates correspondence Atmospheric temperature.

[0032] (2) Both the MOD07_L2 product and the process of estimating near-surface air temperature involve certain uncertainties. The estimated air temperature needs to be further processed using an averaging parameter scheme. The formula is as follows:

[0033]

[0034] In the formula, T a T represents the temperature estimated by the averaging parameter scheme. a1 This represents the estimated temperature in T, as determined by MOD07_L2. s This indicates the surface temperature data provided by MOD07_L2.

[0035] (3) Linear regression is performed on near-surface air temperature and elevation data of multiple pixels within the moving window to estimate SATLR. The formula is as follows:

[0036] T a = a×Z+b (3)

[0037] In the formula, T a Z is the air temperature, Z is the elevation, and a and b are regression coefficients (a and b correspond to the slope and intercept in the linear equation, respectively), where a is the estimated SATLR.

[0038] Since SATLR is locally accurate, during the pixel-by-pixel estimation of SATLR values, it is determined whether the pixel values ​​within the moving window meet preset verification conditions. If not, the moving window size is adjusted. After determining the optimal moving window size, linear regression is performed on the near-surface air temperature and elevation within the moving window to obtain the SATLR value as the SATLR value of the central pixel within the moving window. Finally, the above process is repeated for each valid pixel to produce a daily spatially continuous SATLR.

[0039] Using the Qinghai-Tibet Plateau as the target area, the specific implementation of the near-surface air lapse rate generation method based on MODIS data provided in this embodiment of the invention is as follows:

[0040] First, select the data, including:

[0041] (1) MODIS atmospheric products (MOD07_L2), the atmospheric temperature profile, atmospheric pressure profile, surface pressure and surface temperature data included in the MOD07_L2 product will be used in this invention. The atmospheric temperature profile is distributed across 20 vertical atmospheric pressure levels. Due to the influence of clouds, MOD07_L2 can only provide atmospheric temperature and atmospheric pressure profiles for clear days;

[0042] (2) Temperature observation data from 86 stations on the Qinghai-Tibet Plateau provided by the China Meteorological Administration (CMA) with a time resolution of daily, were obtained by observation instruments 2m above the ground. The average daily temperature was obtained by averaging the values ​​at four observation times (02:00, 08:00, 14:00 and 20:00 Beijing time).

[0043] (3) The digital elevation model (DEM) acquired and created by the U.S. Space Shuttle Radar Topography Mission (SRTM). The above data is preprocessed, its spatiotemporal resolution is standardized, and spatiotemporal matching is performed. The implementation method can be divided into the following four steps.

[0044] (1) Extraction and transformation of profile data.

[0045] Since the MOD07_L2 product provides data of multiple types across 20 layers, daily atmospheric temperature, atmospheric pressure, surface air pressure, and surface temperature data are extracted as needed. The extracted data undergoes projection transformation, is stitched together, and cropped according to the target area, removing background values ​​to avoid affecting near-surface temperature estimation.

[0046] (2) Estimation of near-surface temperature.

[0047] Using the atmospheric temperature data, atmospheric pressure data, and surface air pressure obtained in step (1), the near-surface air temperature is obtained based on formula (1). The MOD07_L2 atmospheric profile data estimates the instantaneous near-surface air temperature, which includes near-surface air temperatures at multiple times during the day and night. Considering that the daily average near-surface air temperature data provided by CMA is obtained by averaging the instantaneous air temperature data observed at four times during the day, this method first uses the MOD07_L2 product to estimate the instantaneous near-surface air temperature at multiple times during the day (10:30 and 22:30), and then averages them to obtain the daily average near-surface air temperature. The verification results of the near-surface air temperature based on the measured air temperature at the station are as follows: Figure 2 As shown in (a), the estimated near-surface air temperature value and the R value of the measured air temperature at the station are... 2 The value is 0.59, the RMSE is 5.96℃, and the MBE is -1.74℃. It can be seen that the estimated near-surface temperature is not ideal.

[0048] (3) Average parameter scheme for near-surface air temperature.

[0049] This embodiment uses an averaging parameterization scheme to process near-surface air temperature, that is, it averages the estimated near-surface air temperature and the surface temperature pixel by pixel. The verification results of near-surface air temperature based on the station's measured air temperature are as follows: Figure 2 As shown in (b), the processed near-surface air temperature value and the R value of the station's measured air temperature are...2 The accuracy of near-surface air temperature is 0.83, RMSE is 3.90℃, and MBE is 1.52℃. The accuracy of near-surface air temperature is greatly improved after processing with the average parameter scheme, and it can be used to estimate the temperature lapse rate.

[0050] (4) Estimation of the spatial continuous near-surface temperature lapse rate.

[0051] This embodiment uses near-surface air temperature and elevation data from multiple pixels within a moving window to perform linear regression to estimate the spatially continuous SATLR (Formula (3)). Since SATLR is locally accurate, the calculated SATLR is only reliable when the correlation between near-surface air temperature and elevation within the moving window is high and there is a certain elevation difference (i.e., the difference between the maximum and minimum values ​​of pixels within the moving window). In addition, the acquired air temperature data has the problem of too many missing pixels.

[0052] As a preferred processing method, in this embodiment, the initial window size is set to 11×11. If the significance level is less than 0.1, the elevation difference is less than 10m, or the number of effective pixels within the moving window is less than half of the total number of pixels, the window size is further increased until all the above test conditions are met simultaneously. Only then is the SATLR estimated using linear regression on the temperature and elevation within the moving window, and the SATLR at this point is taken as the value of the center pixel of the moving window. If the moving window size reaches 31 and the test conditions are still not met, then the center pixel of the moving window cannot be estimated to obtain a valid SATLR value. That is, when the moving window size exceeds the specified value, it indicates that the SATLR value estimation of the current center pixel has failed, and the pixel is skipped, and the next pixel is processed. The above process is repeated for each effective pixel to obtain the daily spatially continuous SATLR.

[0053] This embodiment is applied to the Tibetan Plateau. Deeper SATLR pixels are mainly concentrated in the northern, southwestern, and central parts of the Tibetan Plateau, while shallower SATLR pixels are mainly concentrated in the southeastern part. These areas have lower altitudes and more complex terrain. SATLR on the Tibetan Plateau is deeper in spring and summer, followed by winter, and shallowest in autumn. Some small areas on the Tibetan Plateau experience temperature inversions year-round. Using high-resolution online maps, these small areas are identified as lakes or basins. These small areas are more prone to deeper temperature inversions in summer, while the areas with inversions become smaller and the inversion values ​​become shallower in spring, autumn, and winter.

[0054] 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.

[0055] The above descriptions are merely some embodiments of the present invention. For those skilled in the art, various modifications and improvements can be made without departing from the inventive concept of the present invention, and all such modifications and improvements fall within the scope of protection of the present invention.

Claims

1. A method for generating near-surface air lapse rate based on MODIS data, characterized in that, Includes the following steps: Step 1: Extract contour data based on MODIS data of the target region, perform data preprocessing on the extracted contour data to unify the spatiotemporal resolution of the data, and perform spatiotemporal matching. The contour data includes: atmospheric temperature contour, atmospheric pressure contour, surface temperature, and surface air pressure; Step 2: According to the formula Estimate near-surface air temperature in the target area ; in, This represents the atmospheric pressure closest to the Earth's surface at multiple altitudes in MODIS data. This represents the atmospheric pressure closest to the Earth's surface in the atmospheric pressure profile obtained in step 1. express The surface air pressure at the next higher altitude, This indicates that step 1 involves obtaining the surface air pressure; Indicates correspondence Atmospheric temperature, Indicates correspondence Atmospheric temperature; Step 3: Estimate the near-surface air temperature of the target area. Perform averaging parameter processing: ； in, This represents the near-surface air temperature estimated using the averaged parameter processing. This represents the near-surface temperature obtained in step 1; Step 4, based on near-surface air temperature Linear regression with elevation data was used to estimate the near-surface air lapse rate pixel by pixel using a sliding window method; Among them, near-surface temperature The linear relationship between the elevation data and the elevation data is as follows: ,in, Indicates elevation. and The slope and intercept of the linear equation are represented, and 'a' is used to characterize the estimated near-surface air lapse rate. When estimating the near-surface air lapse rate pixel by pixel, it is determined whether the pixel values ​​within the sliding window meet the preset test conditions. If not, the size of the sliding window is adjusted until the pixel values ​​within the sliding window meet the preset test conditions, thus obtaining the optimal sliding window size. Then, the near-surface air temperature within the sliding window is calculated. and elevation Linear regression is performed, and the obtained 'a' is used as an estimate of the near-surface temperature lapse rate of the central pixel within the current sliding window to generate a daily spatially continuous near-surface temperature lapse rate.

2. The method as described in claim 1, characterized in that, The specific test conditions in step 4 are: the significance level is greater than or equal to the specified value, the elevation difference is greater than or equal to the specified value, and the number of valid pixels in the sliding window is greater than or equal to the specified value.

3. The method as described in claim 1, characterized in that, The initial size of the sliding window is set to 11 × 11.

4. The method according to any one of claims 1 to 3, characterized in that, The saliency level is specified as 0.1 or 0.05, the elevation difference is specified as 10 meters, and the effective number of pixels is specified as half the total number of pixels in the sliding window.

Images