In-cloud particle velocity inversion method based on EarthCard 1B

Abstract

The invention discloses an in-cloud particle velocity inversion method based on EarthCard 1B, and aims to solve the problems of insufficient error suppression, poor scene adaptability, disjunction of processing flow and single verification system in the prior art when the in-cloud particle velocity is inversed by EC-CPR data of an EarthCard satellite. A full-link collaborative optimization process of data preprocessing-mask generation-velocity inversion-result output is constructed, invalid value replacement in a preprocessing stage, velocity calibration and a subsequent inversion step are in deep linkage, and mask rules are uniformly applied to Doppler velocity and settling velocity calculation. High-precision inversion of the in-cloud particle velocity in a full-cloud scene is realized, and reliable data support is provided for cloud microphysical process analysis and climate model parameterization optimization.

Description

A Cloud-Based Particle Velocity Inversion Method Based on EarthCare 1B Technical Field

[0001] This invention belongs to the field of cloud microphysics technology, specifically involving the design of a cloud particle velocity inversion method based on EarthCare 1B. Background Technology

[0002] EarthCare is a next-generation meteorological observation satellite jointly developed by the Japan Aerospace Exploration Agency (JAXA) and the European Space Agency (ESA). Its onboard Cloud Profiler Radar (EC-CPR) is the world's first spaceborne W-band (94GHz) Doppler radar, possessing the unique capability to detect the Doppler velocities of particles within clouds. This enables routine observation of vertical cloud motion on a global scale, providing crucial data support for cloud microphysics analysis and climate model optimization. With its low-Earth orbit altitude of 400km and a large-aperture 2.5m antenna, the radar boasts a minimum detectable reflectivity of -36dBZ, superior to the previous generation CloudSat satellite's -29dBZ. This allows for more efficient capture of thin cirrus clouds and low-level small water droplet cloud systems, and its observational data has become a core data source for cloud particle velocity inversion.

[0003] The existing JAXA standard algorithm and basic data processing technology and the inversion method and performance evaluation technology based on observation data have laid the foundation for EarthCare cloud particle velocity inversion. However, there are still significant defects in practical applications, which are difficult to meet the business needs of high precision and wide adaptability. The specific shortcomings are as follows: (1) The error suppression effect is limited and the velocity inversion accuracy is insufficient; (2) There is a lack of universal particle velocity separation model; (3) The data preprocessing and inversion process are disconnected and lack collaborative optimization; (4) The verification system is single and the credibility of the inversion results is not guaranteed. Summary of the Invention

[0004] The purpose of this invention is to address the problems of insufficient error suppression, poor scene adaptability, disconnected processing flow, and single verification system in existing technologies for inverting cloud particle velocities using EarthCare satellite EC-CPR data. This invention proposes a cloud particle velocity inversion method based on EarthCare 1B, which achieves high-precision inversion of cloud particle velocities in all cloud scenarios, providing reliable data support for cloud microphysical process analysis and climate model parameterization optimization.

[0005] The technical solution of the present invention is: a cloud particle velocity inversion method based on EarthCare 1B, comprising the following steps: S1, acquiring EarthCare 1B data.

[0006] S2. Preprocess the EarthCare 1B data to obtain preprocessed data.

[0007] S3. Calculate the radar reflectivity based on the preprocessed data.

[0008] S4. Generate Doppler velocity mask and settlement velocity mask based on radar reflectivity.

[0009] S5. Obtain the smoothed Doppler velocity and settlement velocity along the track by inversion based on the Doppler velocity mask and settlement velocity mask.

[0010] S6. Save the smoothed Doppler velocity and settlement velocity along the track as a NetCDF file.

[0011] Furthermore, the EarthCare 1B data in step S1 includes geographic and geometric information and scientific observation data.

[0012] Geographic and geometric information includes the timestamp, latitude and longitude, actual altitude of each distance cell, surface elevation, and cell number of the surface location for each radar profile.

[0013] Scientific observation data includes radar reflectivity factor in linear units, raw Doppler velocity, and radar wavelength.

[0014] Furthermore, the preprocessing in step S2 includes invalid value processing and Doppler velocity calibration.

[0015] Invalid value handling involves identifying and replacing special padding values.

[0016] The formula for Doppler velocity calibration is: in This represents the true radial Doppler velocity. Represents the original Doppler velocity. Indicates the calibration scaling factor. This indicates the calibration offset.

[0017] Furthermore, the formula for calculating radar reflectivity in step S3 is as follows: in Indicates radar reflectivity. Represents the reflectivity factor. Indicates the radar wavelength. This represents the normalized radar backscattering cross section. Represents the dielectric constant. Indicates radar receiving power. Indicates radar transmit power. Represents the radar calibration coefficient. This indicates the slant range between the radar and the target.

[0018] Further, step S4 includes the following sub-steps: S41, obtaining the range cell number corresponding to the ground surface in each radar profile.

[0019] S42. Taking the surface warehouse as the center, extend 5 distance warehouses upwards and downwards, and mark all warehouses within this range as the surface clutter area.

[0020] S43. Assign the value True to the surface clutter region to indicate that a mask is needed, and assign the value False to the remaining regions to indicate that the mask is valid, thus obtaining the surface clutter mask. .

[0021] S44. Based on surface clutter masking and radar reflectivity Generate synthesis mask : in Represents logical OR, Indicates a null mask. This represents an infinity mask.

[0022] S45, In the composite mask Based on this, the Doppler velocity mask is obtained. and settling velocity mask .

[0023] Furthermore, the method for inverting the smoothed Doppler velocity along the track in step S5 is as follows: A1. For each central profile i, calculate the distance between all radar profiles within the window and the central profile i. : in Indicates the section number.

[0024] A2. Based on distance Calculate initial weights : in This represents the weight decay factor.

[0025] A3. Initial weights Normalization is performed to obtain the first weight. .

[0026] A4. Obtain effective Doppler velocity data for each vertical height level. And through the first weight For effective Doppler velocity data A weighted average is performed to obtain the smoothed Doppler velocity.

[0027] A5. Using Doppler velocity masking The same masking rules are used to encapsulate the smoothed Doppler velocity into a mask array, resulting in the smoothed Doppler velocity along the track.

[0028] Furthermore, the inversion method for settlement velocity in step S5 is as follows: B1, setting a height threshold. , transition area width The atmosphere with an altitude less than 4000m is classified as the lower atmosphere, the atmosphere with an altitude between 4000m and 6000m is classified as the transitional atmosphere, and the atmosphere with an altitude greater than 6000m is classified as the upper atmosphere.

[0029] B2. Masking the settlement velocity The settling velocity of the region is assigned a value of 0, and a settling velocity mask that does not belong to the region is calculated based on the atmospheric region division results. The subsidence rate of the area : in Represents the reflectivity factor. Indicates the second weight. Indicates atmospheric altitude. The lower value represents the value calculated based on the radar reflectivity factor. This represents the upper value calculated based on the radar reflectivity factor.

[0030] Further, step S6 specifically involves: taking time, latitude, longitude, altitude, radar reflectivity, smoothed Doppler velocity along the track, settlement velocity, and surface elevation as variables, adding unit and long name attributes to each variable, and adding title, source, creation time, and original file name as global attributes, and saving it as a NetCDF file.

[0031] The beneficial effects of the present invention are: (1) The present invention accurately calibrates the Doppler velocity by extracting the native scaling factor and offset of EarthCare 1B data, and combines the surface clutter mask and the radar reflectivity multi-threshold comprehensive mask to efficiently filter surface interference, invalid values ​​and abnormal data, thereby reducing system errors and noise interference from the source and greatly improving the purity of the inversion data.

[0032] (2) The present invention innovatively designs a layered settling velocity calculation model, adopts differentiated calculation logic for the lower and upper atmosphere and the transition region, and achieves smooth connection of high and low altitude velocities through linear transition weights, perfectly adapts to the motion physical characteristics of particles in clouds at different altitudes, and achieves accurate inversion of the entire cloud type and the entire altitude range.

[0033] (3) The present invention constructs a full-link collaborative optimization process of “data preprocessing-mask generation-velocity inversion-result output”. The invalid value replacement and velocity calibration in the preprocessing stage are deeply linked with the subsequent inversion steps. The mask rules are uniformly applied to the calculation of Doppler velocity and settlement velocity, avoiding the accuracy loss caused by process disconnection and ensuring the data consistency of each link.

[0034] (4) The present invention adopts the distance attenuation weighted integral algorithm. Through dynamic normalization weight allocation, it retains the spatial distribution characteristics of effective data while weakening the influence of isolated outliers, which significantly improves the smoothness and continuity of velocity data along the track direction. The settlement velocity calculation is based on the physical correlation model of radar reflectivity factor, which further ensures the physical rationality and numerical accuracy of the inversion results. Attached Figure Description

[0035] Figure 1 shows a flowchart of a cloud particle velocity inversion method based on EarthCare 1B provided by an embodiment of the present invention. Detailed Implementation

[0036] Exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be understood that the embodiments shown and described in the drawings are merely exemplary and are intended to illustrate the principles and spirit of the invention, and are not intended to limit the scope of the invention.

[0037] This invention provides a cloud particle velocity inversion method based on EarthCare 1B, as shown in Figure 1, including the following steps S1~S6: S1, Obtain EarthCare 1B data.

[0038] In this embodiment of the invention, EarthCare 1B data refers to EarthCare CPR 1B level HDF data (.h5 file), which mainly includes geographic and geometric information and scientific observation data.

[0039] The geographic and geometric information includes the timestamp (profileTime, based on TAI) for each radar profile, latitude and longitude, actual height (binHeight) corresponding to each distance bin, surface elevation (in km), and surface bin number (surfaceBinNumber).

[0040] Scientific observation data includes the radar reflectivity factor (in mm²) in linear units. 6 / m 3 ), raw Doppler velocity (doppler Velocity, integer) and radar wavelength (rayHeaderLambda).

[0041] S2. Preprocess the EarthCare 1B data to obtain preprocessed data.

[0042] In this embodiment of the invention, the preprocessing includes invalid value processing and Doppler velocity calibration.

[0043] Invalid value handling involves identifying and replacing special padding values. For example, the maximum value in binHeight (9.96921e+36) is set to 0; missing or invalid flags in other variables (such as NaN, Inf, and dBZ values ​​outside the reasonable range) are excluded by masking.

[0044] The original Doppler velocity (dopplerVelocity) is integer compressed data, and linear decompression is required by combining its scale_factor and add_offset attributes. in This represents the true radial Doppler velocity. This represents the raw Doppler velocity (uncalibrated). This indicates the calibration scaling factor, with a default value of 2. This indicates the calibration offset, with a default value of 0.

[0045] S3. Calculate the radar reflectivity based on the preprocessed data.

[0046] In this embodiment of the invention, the formula for calculating radar reflectivity is: in Indicates radar reflectivity. Represents the reflectivity factor. Indicates the radar wavelength (in meters). Normalized radar backscattering cross section (in meters) 2 / m 3 ), Represents the dielectric constant. This indicates the radar receive power, with the noise floor removed. Indicates radar transmit power. Represents radar calibration coefficients (unit: m) -3 ), This indicates the slant range (in meters) between the radar and the target. Radar reflectivity. Reflectivity factor The 10-fold logarithm converts the linear-scale reflectivity factor to a logarithmic scale, which helps to highlight the differences in weak signals.

[0047] S4. Generate Doppler velocity mask and settlement velocity mask based on radar reflectivity.

[0048] Step S4 includes the following sub-steps S41~S45: S41, Read the ScienceData / Data / surfaceBinNumber field of the HDF file to obtain the range bin number corresponding to the surface in each radar profile.

[0049] S42. With the surface cell as the center, extend five range cells upwards and downwards (corresponding to a vertical range of approximately 1.2 km, since the CPR radar range cell resolution is fixed), and mark all cells within this range as surface clutter areas.

[0050] S43. Assign the value True to the surface clutter region to indicate that a mask is needed, and assign the value False to the remaining regions to indicate that the mask is valid, thus obtaining the surface clutter mask. : in Indicates the index along the track direction. (To avoid array out-of-bounds errors) (Python slicing: left-closed, right-open; adding 6 ensures the inclusion of the surface warehouse plus 5). This represents the total number of height layers, ensuring it does not exceed the array boundaries.

[0051] S44. Based on surface clutter masking and radar reflectivity Generate synthesis mask : in Represents logical OR, Indicates a null mask. This represents an infinity mask.

[0052] In this embodiment of the invention, radar reflectivity is preserved by setting a reflectivity threshold. Data in the range of [-30, 100] needs to be excluded: (1) Weak noise: dBZ≤-30 (signal strength close to the noise floor, no cloud information).

[0053] (2) Abnormal values: dBZ>100 (exceeds the reasonable range of natural cloud reflectance, which may be due to instrument malfunction) or dBZ=10 (code-specific abnormal value marker).

[0054] (3) Invalid values: mask np.isnan(dBZ) (null value) and np.isinf(dBZ) (infinite value).

[0055] S45, In the composite mask Based on this, the Doppler velocity mask is obtained. and settlement velocity mask .

[0056] in, , For radar reflectivity The mask (to ensure that Doppler velocity data is preserved only in areas with cloud signals).

[0057] Settlement velocity mask With Doppler velocity mask Consistent with the logic, this ensures that the settlement velocity Vr calculation is only for the effective cloud area.

[0058] S5. Based on the Doppler velocity mask and settlement velocity mask, the smoothed Doppler velocity and settlement velocity along the track are obtained by inversion.

[0059] In this embodiment of the invention, the method for inverting the smoothed Doppler velocity along the track is as follows: A1. For each central profile i, calculate the distance between all radar profiles within the window and the central profile i. : in Indicates the section number.

[0060] A2. Based on distance Calculate initial weights : in This represents the weight decay factor.

[0061] In this embodiment of the invention, a distance attenuation weight is adopted (the weight of the central sample is the largest, and it gradually decreases towards both sides) to avoid the problem of uneven weight of edge samples in traditional moving average. The window size is fixed at 20 (covering 20 radar profiles along the track direction to balance the smoothing effect and spatial resolution), and the attenuation factor is fixed at 1.8 (to control the weight attenuation rate; the larger the attenuation factor, the smaller the weight of the edge samples).

[0062] A3. Initial weights Normalization is performed to obtain the first weight. Ensure the total weights are 1: A4. Obtain effective Doppler velocity data for each vertical height level. : in Indicates the non-masked profile number. This represents a masked two-dimensional array that stores calibrated Doppler velocity data. The index representing the height level.

[0063] And through the first weight For effective Doppler velocity data After performing a weighted average, the smoothed Doppler velocity is obtained: A5. Using Doppler velocity masking The same masking rules encapsulate the smoothed Doppler velocity into a mask array, resulting in the smoothed Doppler velocity along the track, ensuring data validity.

[0064] In this embodiment of the invention, the method for inverting the settlement velocity is as follows: B1, setting a height threshold. , transition area width The atmosphere with an altitude less than 4000m is classified as the lower atmosphere, the atmosphere with an altitude between 4000m and 6000m is classified as the transitional atmosphere, and the atmosphere with an altitude greater than 6000m is classified as the upper atmosphere.

[0065] B2. Masking the settlement velocity The settling velocity of the region is assigned a value of 0 to avoid interference from invalid data, and a settling velocity mask that does not belong to the region is calculated based on the atmospheric region division results. The subsidence rate of the area : in Represents the reflectivity factor. Indicates the second weight. Indicates atmospheric altitude. The value below is calculated based on the radar reflectivity factor and is numerically equal to... , This represents the upper value calculated based on the radar reflectivity factor, which is numerically equal to... The deposition of particles in the lower atmosphere is greatly affected by air resistance, and the reflectivity factor... The effect on velocity exhibits a high-power relationship; the deposition of particles in the upper atmosphere is closer to free fall, and the reflectivity factor... The effect on velocity exhibits a low-power relationship, making it the second weighting factor in the transition layer atmosphere. The speed value is linearly increased from 0 to 1 to ensure that there are no sudden changes in the speed value.

[0066] S6. Save the smoothed Doppler velocity and settlement velocity along the track as a NetCDF file.

[0067] In this embodiment of the invention, time, latitude, longitude, altitude, radar reflectivity, smoothed Doppler velocity along the track, subsidence velocity, and surface elevation are used as variables. Each variable is given a unit and long name attribute, and a title, source, creation time, and original file name are added as global attributes. The data is then saved as a NetCDF file.

[0068] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A cloud-based particle velocity inversion method based on EarthCare 1B, characterized in that, Includes the following steps: S1. Acquire EarthCare 1B data; S2. Preprocess the EarthCare 1B data to obtain preprocessed data; S3. Calculate the radar reflectivity based on the preprocessed data; S4. Generate Doppler velocity mask and settlement velocity mask based on radar reflectivity; S5. Invert the smoothed Doppler velocity and settlement velocity along the track based on the Doppler velocity mask and settlement velocity mask; S6. Save the smoothed Doppler velocity and settlement velocity along the track as a NetCDF file.

2. The cloud particle velocity inversion method based on EarthCare 1B according to claim 1, characterized in that, The EarthCare 1B data in step S1 includes geographic and geometric information and scientific observation data; the geographic and geometric information includes the timestamp, latitude and longitude, actual height corresponding to each distance cell, surface elevation, and cell number; the scientific observation data includes radar reflectivity factor in linear units, raw Doppler velocity, and radar wavelength.

3. The cloud particle velocity inversion method based on EarthCare 1B according to claim 1, characterized in that, The preprocessing in step S2 includes invalid value processing and Doppler velocity calibration; the invalid value processing involves identifying and replacing special filler values; the formula for Doppler velocity calibration is: in This represents the true radial Doppler velocity. Represents the original Doppler velocity. Indicates the calibration scaling factor. This indicates the calibration offset.

4. The cloud particle velocity inversion method based on EarthCare 1B according to claim 1, characterized in that, The formula for calculating radar reflectivity in step S3 is as follows: in Indicates radar reflectivity. Represents the reflectivity factor. Indicates the radar wavelength. This represents the normalized radar backscattering cross section. Represents the dielectric constant. Indicates radar receiving power. Indicates radar transmit power. Represents the radar calibration coefficient. This indicates the slant range between the radar and the target.

5. The cloud particle velocity inversion method based on EarthCare 1B according to claim 1, characterized in that, Step S4 includes the following sub-steps: S41, obtaining the range cell number corresponding to the ground surface in each radar profile; S42, expanding upwards and downwards by 5 range cells from the center of the surface cell, and marking all cells within this range as surface clutter areas; S43, assigning the surface clutter areas a value of True to indicate that a mask is needed, and assigning the remaining areas a value of False to indicate that they are valid, thus obtaining the surface clutter mask. ； S44. Based on surface clutter masking and radar reflectivity Generate synthesis mask ： in Represents logical OR, Indicates a null mask. S45, representing the infinite value mask; in the synthesized mask Based on this, the Doppler velocity mask is obtained. and settlement velocity mask 。 6. The cloud particle velocity inversion method based on EarthCare 1B according to claim 1, characterized in that, The method for inverting the smoothed Doppler velocity along the track in step S5 is as follows: A1. For each central profile i, calculate the distance between all radar profiles within the window and the central profile i. ： in Indicates the section number; A2, based on distance Calculate initial weights ： in This represents the weight decay factor. A3, for the initial weights... Normalization is performed to obtain the first weight. ； A4. Obtain effective Doppler velocity data for each vertical height level. And through the first weight For effective Doppler velocity data A weighted average is performed to obtain the smoothed Doppler velocity; A5. Using Doppler velocity masking The same masking rule encapsulates the smoothed Doppler velocity into a mask array, resulting in the smoothed Doppler velocity along the track.

7. The cloud particle velocity inversion method based on EarthCare 1B according to claim 1, characterized in that, The method for inverting the settlement velocity in step S5 is as follows: B1, setting a height threshold. , transition area width The atmosphere at altitudes less than 4000m is classified as the lower atmosphere, the atmosphere at altitudes between 4000m and 6000m is classified as the transition layer, and the atmosphere at altitudes greater than 6000m is classified as the upper atmosphere; B2, the settling velocity mask. The settling velocity of the region is assigned a value of 0, and a settling velocity mask that does not belong to the region is calculated based on the atmospheric region division results. The subsidence rate of the area ： in Represents the reflectivity factor. Indicates the second weight. Indicates atmospheric altitude. The lower value represents the value calculated based on the radar reflectivity factor. This represents the upper value calculated based on the radar reflectivity factor.

8. The cloud particle velocity inversion method based on EarthCare 1B according to claim 1, characterized in that, Step S6 specifically involves taking time, latitude, longitude, altitude, radar reflectivity, smoothed Doppler velocity along the track, subsidence velocity, and surface elevation as variables, adding unit and long name attributes to each variable, and adding title, source, creation time, and original file name as global attributes, and saving it as a NetCDF file.

Images