Overshoot cloud top identification and area calculation method and device based on stationary satellite data

Abstract

The invention discloses an overshoot cloud top identification and area calculation method and device based on geostationary satellite data. The method comprises the following steps: acquiring geostationary satellite cloud radiation brightness temperature data and troposphere top temperature data of a target area in a required time period; extracting cloud radiation brightness temperature of water vapor, ozone and infrared channels; after space-time matching, a multi-channel threshold discrimination method is used to identify potential overshoot cloud top pixels; based on the infrared brightness temperature data, utilizing an isotherm analysis method to identify a potential overshoot cloud top area and an anvil cloud area corresponding to each potential overshoot cloud top pixel; identifying an overshoot cloud top region and calculating the area of the overshoot cloud top region by comparing the average brightness temperature difference between the potential overshoot cloud top pixel and the anvil cloud region; the device comprises a data receiving and caching module, a data processing unit and a result output unit, and all the steps in the method are achieved. According to the method, satellite multi-channel brightness temperature data and troposphere top temperature data are combined, interference of cirrus cloud pixels is effectively eliminated, the overshoot cloud top is accurately recognized, and overshoot cloud top area calculation is achieved.

Description

Technical Field

[0001] This invention belongs to the fields of atmospheric science and remote sensing technology, specifically relating to a method and apparatus for identifying and calculating the area of ​​overshoot cloud tops based on geostationary satellite data. Background Technology

[0002] Overpass cloud tops are tower-shaped cloud top structures formed when strong updrafts penetrate the tropopause and extend into the stratosphere. They play an important role in the tropospheric-stratospheric exchange process, directly affecting the content of stratospheric water vapor and greenhouse gases such as ozone, and indirectly modulating the global atmospheric shortwave radiation balance. Their formation is also closely related to extreme weather phenomena such as heavy rain, tornadoes, and typhoons, posing a serious threat to aviation safety.

[0003] Weather radar or polar-orbiting satellites are the most accurate observational means for identifying overrunning cloud tops. Weather radar can obtain the vertical profile of clouds to determine whether they have broken through the tropopause, but its detection range is limited, making it difficult to apply in ocean areas. Furthermore, the equipment construction and maintenance costs are high, and regular calibration is required to ensure reliability. Radars carried by polar-orbiting satellites operate on the same principle as weather radar, but their scanning method limits their temporal resolution, making it difficult to reflect the continuous spatiotemporal evolution of overrunning cloud tops. In contrast, geostationary satellites, due to their high spatiotemporal resolution and wide-area detection capabilities, have become an important means of identifying overrunning cloud tops. Some methods already use geostationary satellite data for overrunning cloud top identification, such as the single-channel infrared texture method proposed by Bedka in 2010. This method identifies overrunning cloud tops by identifying the infrared brightness temperature cold center and calculating the brightness temperature gradient. This method has stable detection capabilities in mid-to-high latitudes, but because it only uses infrared brightness temperature data, it is prone to misidentifying very cold cirrus cloud edges or non-convective cirrus clouds as overrunning cloud tops, resulting in a high false alarm rate, reaching 90%.

[0004] Furthermore, existing identification methods mostly focus on determining the presence of overrun cloud tops, while research on their physical characteristics, such as area, remains relatively limited. The area of ​​an overrun cloud top not only reflects the strength of mid-level updrafts but also has a certain correlation with tornado wind speed. Therefore, developing a method that can accurately identify overrun cloud tops and quantify their area is of great significance for the monitoring and early warning of severe convection. Summary of the Invention

[0005] To address the technical problems existing in the prior art, this invention proposes a method and apparatus for identifying and calculating the area of ​​overshoot cloud tops based on geostationary satellite data, so as to achieve accurate identification and rapid area calculation of overshoot cloud tops.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0007] A method for identifying overshoot cloud tops and calculating their area based on geostationary satellite data includes the following steps:

[0008] S1. Obtain geostationary satellite cloud radiation brightness temperature data and tropopause temperature data for the target area within the required time period;

[0009] S2. Read the acquired geostationary satellite cloud radiation brightness temperature data and extract the cloud radiation brightness temperature from the three channels of water vapor, ozone and infrared.

[0010] S3. Unify the temporal and spatial resolution of geostationary satellite cloud radiative brightness temperature data and tropopause temperature data;

[0011] S4. Based on all the data, identify potential overshoot cloud top pixels using a multi-channel threshold discrimination method;

[0012] S5. Based on the infrared channel brightness temperature data extracted in step S2, the potential overshoot cloud top region and its anvil cloud region of the potential overshoot cloud top pixel are extracted using the isotherm analysis method.

[0013] S6. Subtract the infrared channel brightness temperature of all pixels in the anvil cloud region from the infrared channel brightness temperature of the potential overshoot cloud top pixel. If the difference is greater than or equal to 5K, then the potential overshoot cloud top pixel is determined as the overshoot cloud top, and the corresponding potential overshoot cloud top region is determined as the overshoot cloud top region.

[0014] S7. Based on the identified overshoot cloud top region, calculate the overshoot cloud top area using the Gaussian area formula.

[0015] Furthermore, in step S1, the preset temporal resolution of the geostationary satellite cloud radiative brightness temperature data is [missing information]. Spatial resolution is The temporal resolution of the tropopause temperature data is [missing information]. Spatial resolution is The number of frames acquired for geostationary satellite cloud radiative brightness temperature data is: The number of frames for tropopause temperature data is , , These represent the start and end times for data acquisition.

[0016] Furthermore, in step S3, a spatiotemporal matching algorithm is used to unify the temporal and spatial resolutions of geostationary satellite cloud radiative brightness temperature data and tropopause temperature data. Specific steps include:

[0017] S31. In terms of time, for each frame of geostationary satellite cloud radiation brightness temperature data, select the tropopause temperature data of the frame with the closest time to it for matching.

[0018] S32. In space, linear interpolation is performed on the tropopause temperature data to make its spatial resolution equal to that of the geostationary satellite cloud radiation brightness temperature data.

[0019] Furthermore, the specific steps of step S4 include:

[0020] S41. For each observation pixel in each frame of geostationary satellite cloud radiation brightness temperature data, compare its infrared channel brightness temperature with the corresponding tropopause temperature. If its infrared channel brightness temperature is less than the corresponding tropopause temperature, proceed to the next step; otherwise, determine that the observation pixel is not an overshoot cloud top pixel.

[0021] S42. Calculate the difference between the brightness temperature of the water vapor channel at the observed pixel and the corresponding tropopause temperature. If the difference is less than or equal to 4K, proceed to the next step; otherwise, determine that the observed pixel is not an overshoot cloud top pixel.

[0022] S43. Calculate the difference between the brightness temperature of the ozone channel and the brightness temperature of the water vapor channel at the observed pixel. If the difference is greater than or equal to a preset threshold, the observed pixel is listed as a potential overshoot cloud top pixel; otherwise, the observed pixel is determined not to be an overshoot cloud top pixel. The preset threshold is set according to the latitude of the target area.

[0023] Furthermore, the specific steps of step S5 include:

[0024] S51. For each frame of geostationary satellite cloud radiation brightness temperature data, extract the infrared channel brightness temperature of the current frame;

[0025] S52. For any potential overshoot cloud top pixel identified in the current frame, take its infrared channel brightness temperature as the local minimum value, and draw isotherms around the potential overshoot cloud top pixel using 1K intervals based on the infrared channel brightness temperature of the current frame.

[0026] S53. Starting from the isotherm closest to the potential overshoot cloud top pixel, examine from the inside out, and determine the isotherm that first meets one of the following two conditions as the potential overshoot cloud top isotherm:

[0027] a) The average distance between all pixels on the isotherm and the potential overshoot cloud top pixel is less than or equal to 12km, but the average distance between all pixels on the nearest neighboring isotherm and the potential overshoot cloud top pixel is greater than or equal to 12km.

[0028] b) The difference between the infrared channel brightness temperature of the isotherm and the infrared channel brightness temperature at the potential overshoot cloud top pixel is greater than or equal to 8K;

[0029] S54. Continue the inspection outward from the potential overshoot cloud top isotherm, and determine the isotherms that first meet the following conditions as the outer isotherms:

[0030] c) The average distance between all pixels on the isotherm and the potential overshoot cloud top pixel is greater than or equal to 50 km;

[0031] d) The difference between the infrared channel brightness temperature of the isotherm and the infrared channel brightness temperature at the potential overshoot cloud top isotherm is greater than or equal to 10K;

[0032] S55. The annular region between the potential overshoot cloud top isotherm and the outer isotherm is defined as the anvil-shaped cloud region, and the region within the potential overshoot cloud top isotherm is defined as the potential overshoot cloud top region.

[0033] Furthermore, the specific steps for calculating the area of ​​the overhanging cloud top in step S7 are as follows:

[0034] S71. For any identified overshoot cloud top region, use the universal transverse Mercator projection and dynamically select the projection partition according to the longitude of the overshoot cloud top pixels to project the overshoot cloud top region from the geographic coordinate system to the plane coordinate system.

[0035] S72. In a plane coordinate system, the area of ​​the region passing through the cloud top can be directly calculated using the Gaussian area formula. The specific formula is as follows:

[0036]

[0037] in, For the area of ​​the Chongyunding region, ( ) is the polygon corresponding to the overpass cloud top region. The coordinates of the vertices are sorted in clockwise or counterclockwise order. The total number of polygon vertices; the area of ​​isolines is quickly calculated by cross-multiplying and accumulating the coordinates of isotherm vertices, which is more efficient than the traditional grid method or segmentation method, and the area calculation error is less than 1%.

[0038] This invention also protects a device for identifying and calculating overshoot cloud tops based on geostationary satellite data, comprising:

[0039] The data receiving and caching module is used to receive, preprocess, and transmit geostationary satellite cloud radiation brightness temperature data and tropopause temperature data from external sources.

[0040] The data processing unit is used to schedule and process the geostationary satellite cloud radiation brightness temperature data and tropopause temperature data transmitted by the data receiving and caching module, and generate result data including overshoot cloud top region identification and area calculation.

[0041] The result output unit is used to encapsulate the result data output by the data processing unit and transmit it to the host system.

[0042] Furthermore, the data receiving and caching module includes: a corresponding data interface for near real-time acquisition of geostationary satellite cloud radiation brightness temperature data and tropopause temperature data; a cache control unit for aligning and scheduling data from different sources based on timestamp information, and performing preprocessing operations such as data frame synchronization, integrity verification, and field parsing; a high-speed FIFO cache for writing preprocessed data to achieve time-series decoupling between the data receiving process and the subsequent calculation process; and an onboard DDR memory for persistently storing preprocessed data within a preset time window, providing stable and continuous data input to the data processing unit when the calculation conditions are met.

[0043] Furthermore, the data processing unit is composed of a programmable logic device and an embedded processor. The programmable logic controller is used to implement data control, access the onboard DDR memory, and schedule the calculation process. The embedded processor is used to load the pre-stored calculation program from the onboard non-volatile memory. The computer program is the computer instruction corresponding to the steps of the overshoot cloud top identification and area calculation method. During operation, through the scheduling of the programmable logic device, the programmable logic controller reads the spatiotemporally aligned geostationary satellite cloud radiation brightness temperature data and tropopause temperature data from the onboard DDR memory in chronological order, and completes the overshoot cloud top identification and area calculation according to the set algorithm. The calculation result data is temporarily stored in the internal buffer. After the calculation is completed, an output instruction is sent to the result output unit.

[0044] Furthermore, after receiving the output instruction sent by the data processing unit, the result output unit reads the result data and encapsulates the result data according to the preset data format; the encapsulated result data is sent to the host system in real time through the serial communication interface according to the preset communication protocol; the output result data includes at least a timestamp, overshoot cloud top location information, and the corresponding area calculation result.

[0045] Furthermore, the output data also includes quality control identifiers, which can be subsequently parsed, displayed, or stored by the host software.

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0047] The method provided by this invention extends the original single-infrared channel brightness temperature data by introducing brightness temperature data from channels such as water vapor and ozone. Combined with tropopause temperature data, it eliminates cirrus cloud pixels and non-strong convection pixels, thereby significantly reducing the false alarm rate of overshoot cloud top detection. Furthermore, it uses isotherm analysis with 1K intervals and radius and temperature difference as constraint parameters to accurately extract the overshoot cloud top region and its anvil cloud region and calculate the overshoot cloud top area. Attached Figure Description

[0048] Figure 1 A flowchart of the geostationary satellite overshoot cloud top identification and area calculation method provided by the present invention;

[0049] Figure 2 This is a schematic diagram illustrating the potential overshoot cloud top region and its anvil-shaped cloud region for extracting potential overshoot cloud top pixels using the isotherm analysis method provided by the present invention.

[0050] Figure 3 (a) is the infrared brightness temperature map of the Sunflower-8 satellite provided by the present invention, (b) is the 20dBZ echo height map of the GPM dual-frequency rain measurement radar data that matches (a), and (c) is the vertical reflectivity profile of the overshoot cloud top corresponding to (b).

[0051] Figure 4 This is a graph showing the fitting results between the overshoot cloud top area calculated by the method provided in this invention and the overshoot cloud top area validation set;

[0052] Figure 5 This is a schematic diagram of the overshoot cloud top identification and area calculation device provided by the present invention. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments.

[0054] like Figure 1 As shown, the present invention provides a method for identifying overshoot cloud tops and calculating their area based on geostationary satellite data, comprising the following steps:

[0055] S1. Obtain geostationary satellite cloud radiation brightness temperature data and tropopause temperature data for the target area within the required time period;

[0056] S2. Read the acquired geostationary satellite cloud radiation brightness temperature data and extract the cloud radiation brightness temperature from the three channels of water vapor, ozone and infrared.

[0057] S3. Unify the temporal and spatial resolution of geostationary satellite cloud radiative brightness temperature data and tropopause temperature data;

[0058] S4. Based on all the data, identify potential overshoot cloud top pixels using a multi-channel threshold discrimination method;

[0059] S5. Based on the infrared channel brightness temperature data extracted in step S2, the potential overshoot cloud top region and its anvil cloud region of the potential overshoot cloud top pixel are extracted using the isotherm analysis method.

[0060] S6. Subtract the infrared channel brightness temperature of all pixels in the anvil cloud region from the infrared channel brightness temperature of the potential overshoot cloud top pixel. If the difference is greater than or equal to 5K, then the potential overshoot cloud top pixel is determined as the overshoot cloud top, and the corresponding potential overshoot cloud top region is determined as the overshoot cloud top region.

[0061] S7. Calculate the overshoot cloud top area based on the identified overshoot cloud top region.

[0062] Specifically, in step S1, the geostationary satellite cloud radiation brightness temperature data comes from the Himawari-8 multi-channel spectral imager, and the tropopause temperature data comes from the European Centre for Medium-Range Weather Forecasts. The Himawari-8 multi-channel spectral imager has 16 spectral channels, including 3 visible light channels, 3 near-infrared channels, and 10 infrared channels; its spatial resolution is 2 km, with a minimum of 0.5-1 km, and its temporal resolution is 10 minutes. The tropopause temperature data has a temporal resolution of 1 hour and a spatial resolution of 30 km.

[0063] The geostationary satellite cloud radiation brightness temperature data obtained in step S1 is for the period of 15:40 UTC on August 8, 2019, consisting of one frame; the tropopause temperature data obtained is for the period of 15:00 UTC on August 8, 2019, consisting of one frame; the target area is the South China Sea.

[0064] Specifically, in step S2, the cloud radiation brightness temperature data of the geostationary satellite is read, and the cloud radiation brightness temperature of three channels, namely water vapor, ozone and infrared, is extracted from it; wherein, the water vapor channel is the 8th channel with a wavelength of 6.2μm; the ozone channel is the 12th channel with a wavelength of 9.6μm; and the infrared channel is the 13th channel with a wavelength of 10.4μm.

[0065] Specifically, in step S3, a spatiotemporal matching algorithm is used to unify the temporal and spatial resolutions of geostationary satellite cloud radiation brightness temperature data and tropopause temperature data:

[0066] In terms of time, the current frame of geostationary satellite cloud radiation brightness temperature data has been matched with the tropopause temperature data; in terms of space, linear interpolation of the tropopause temperature data has been performed to make its spatial resolution equal to 2km.

[0067] Specifically, in step S4, based on all the data, a multi-channel threshold discrimination method is used to identify potential overshoot cloud top pixels. The specific steps include:

[0068] S41. For each observed pixel in the current geostationary satellite cloud radiative brightness temperature data, denoted as... ,Compare Infrared channel brightness temperature Corresponding to tropopause temperature The size, if If yes, proceed to the next step; otherwise, determine that the observed pixel is not an overshoot cloud top pixel.

[0069] S42, Calculation Bright temperature of water vapor channel Corresponding to tropopause temperature The difference, if If yes, proceed to the next step; otherwise, determine that the observed pixel is not an overshoot cloud top pixel.

[0070] S43, Calculation Ozone channel brightness temperature Brightness of the water vapor channel The difference, if Then Listed as potential overshoot cloud top pixels; if the target area is located north of 23.5 degrees North latitude... If the target area is located south of 23.5 degrees north latitude, then Since the selected target area is the South China Sea, located south of 23.5 degrees north latitude, then That is, if Then Listed as potential overshoot cloud top pixels, denoted as Otherwise, it will be judged. It's not overshooting the top pixel.

[0071] Specifically, step S5 includes the following steps:

[0072] S51. For each frame of geostationary satellite cloud radiation brightness temperature data, extract the infrared channel brightness temperature of the current frame;

[0073] S52, for any potential overshoot cloud top pixel identified in the current frame. ,by Infrared channel brightness temperature To achieve a local minimum, based on the infrared channel brightness temperature of the current frame, using a 1K interval in... Draw isotherms around the area;

[0074] S53, from the closest The isotherms are examined from the inside out. The isotherm that first meets one of the following two conditions is identified as a potential overshoot cloud top isotherm:

[0075] a) Distance of all pixels on the isotherm The average distance is less than or equal to 12km, but all pixels on its nearest neighbor isotherm are far from it. The average distance is greater than or equal to 12km;

[0076] b) Infrared channel brightness temperature of the isotherm and Infrared channel brightness temperature The difference is greater than or equal to 8K, that is ;

[0077] like Figure 2 As shown, the identified potential overshoot cloud top isotherms are marked.

[0078] S54. Continue the inspection outward from the potential overshoot cloud top isotherm, and determine the isotherms that first meet the following conditions as the outer isotherms:

[0079] c) All pixels on the isotherm line The average distance is greater than or equal to 50km;

[0080] d) Infrared channel brightness temperature of the isotherm Infrared channel brightness temperature at the potential overshoot cloud top isotherm. The difference is greater than or equal to 10K, that is... ;

[0081] like Figure 2 As shown, the identified peripheral isotherms are marked.

[0082] S55. The annular region between the potential overshoot cloud top isotherm and the outer isotherm is defined as the anvil-shaped cloud region, and the region within the potential overshoot cloud top isotherm is defined as the potential overshoot cloud top region.

[0083] Specifically, in step S6, the infrared channel brightness temperature of the potential overshoot cloud top pixel is subtracted from the average infrared channel brightness temperature of all pixels in its anvil cloud region. If the difference is greater than or equal to 5K, the potential overshoot cloud top pixel is determined as the overshoot cloud top, and the corresponding potential overshoot cloud top region is determined as the overshoot cloud top region. Figure 2 The text indicates the identified overshoot cloud top region and anvil cloud region.

[0084] Specifically, step S7, which uses the Gaussian area formula to calculate the area of ​​the overshoot cloud top, involves the following steps:

[0085] S71, for example Figure 2 The cloud-peak region identified in the image is projected using the universal transverse Mercator projection. Based on the longitude of the pixels passing through the cloud-peak, the projection partition is dynamically selected to project the cloud-peak region from the geographic coordinate system to the plane coordinate system. The partition table is shown in Table 1, where the degree zone represents the sequence number of the projection zone. The longitude of the cloud-peak vertex in the current frame is 115.67° East, and the 50N projection partition is selected for projection.

[0086] S72. In a plane coordinate system, the area can be directly calculated using the Gaussian area formula. The specific formula is as follows:

[0087]

[0088] in, For the area of ​​the Chongyunding region, ( ) is the polygon corresponding to the overpass cloud top region. The coordinates of the vertices are sorted in clockwise or counterclockwise order. Given the total number of vertices of the polygon, the area is 71.50. .

[0089] Table 1

[0090] Longitude range (East) Central Meridian Longitude Indexing zone 102°- 108° 105° 48N 108°- 114° 111° 49N 114°- 120° 117° 50N 120°- 126° 123° 51N 126°- 132° 129° 52N 132°- 138° 135° 53N 138°- 144° 141° 54N 144°- 150° 147° 55N 150°- 156° 153° 56N 156°- 162° 159° 57N 162°- 168° 165° 58N

[0091] In one specific embodiment, such as Figure 5 As shown, the present invention provides an overshoot cloud top identification and area calculation device based on geostationary satellite data, which is implemented by a PCIe card with data receiving and processing functions, and includes:

[0092] The data receiving and caching module receives geostationary satellite cloud radiation brightness temperature data and tropopause temperature data through the corresponding data interface; under the scheduling of the cache control unit, it completes the preprocessing operations of data frame synchronization, integrity verification and field parsing; it writes the preprocessed data into the high-speed FIFO cache and further stores it in the onboard DDR memory; and sends a data ready indication signal to the data processing unit when the calculation conditions are met.

[0093] The data processing unit is composed of a programmable logic device (FPGA) and an embedded processor (ARM CPU). After receiving the indication signal, the FPGA retrieves geostationary satellite cloud radiation brightness temperature data and tropospheric temperature data from the onboard DDR memory through the control interface and transmits them to the embedded processor. The embedded processor executes the computer program pre-stored in FLASH to complete the identification and area calculation of the overpass cloud top. The calculation results are temporarily stored in the result buffer inside the embedded processor, and at the same time, an output instruction is sent to the result output module.

[0094] Upon receiving an output command, the result output unit reads the result data to be sent from the result buffer inside the embedded processor and encapsulates the result data according to a pre-set data format. The encapsulated data is then sent to the host system via the RS-232 serial communication interface according to a preset communication protocol. During the transmission process, the result output unit monitors the communication status in real time. If a transmission failure or interface abnormality is detected, it performs a retransmission or status reporting operation. After a successful transmission is completed, the result output unit clears the current result buffer and returns to the idle state, waiting for the next output command from the data processing unit.

[0095] In another specific embodiment, dual-frequency rain-measuring radar observation data and weather radar observation data are used to verify the identification results of the overshoot cloud top region:

[0096] Among them, the dual-frequency rain measurement radar product is the second-generation spaceborne rain measurement radar carried by the Global Rain Measurement Satellite (GPM). It is a dual-frequency rain measurement radar that combines a 13.6GHz Ku-band radar (KuPR) and a 35.5GHz Ka-band radar (KaPR). The data scan width is 245km each time, the detection height is 22km above the ground, the horizontal resolution is about 5km, and the vertical resolution is about 250m.

[0097] The effective horizontal detection range of the weather radar data is 230km, the time resolution is 6 minutes, a total of 9 elevation angle data are used, it can detect up to 20km above the ground, and the spatial resolution is 250m;

[0098] The process of verifying the identification results of the overshoot cloud top region using dual-frequency rain radar observation data and weather radar observation data is as follows:

[0099] Step 1: Construct the overshoot cloud top validation set and the overshoot cloud top area validation set:

[0100] Based on the scanning information of the dual-frequency rain-measuring radar, the profile data corresponding to each planar pixel in the vertical direction is obtained. Based on linear interpolation of the vertical profile, the height corresponding to a pixel when its radar reflectivity in the vertical direction reaches 20 dBz is calculated. Figure 3 As shown in (a), this is the infrared brightness temperature image from the Himawari-8 satellite of a severe thunderstorm that occurred near the equator at 10:40 UTC on April 27, 2019. The corresponding 20 dBz echo height from the GPM dual-frequency rain-measuring radar is as follows: Figure 3 As shown in (b), a vertical section is drawn along section line AB in (b), resulting in the following: Figure 3 The vertical reflectivity profile shown in (c) is shown in the middle.

[0101] All GPM data from May to October each year from 2019 to 2023, between 105° and 165° east longitude and 0° and 40° north latitude, were processed and will be as follows: Figure 3 The longitude, latitude, and time of the pixels with an altitude exceeding 14km, as shown in (b), were recorded and integrated into the overpass cloud top verification set, which contains 2037 verification samples.

[0102] Based on the scanning information of weather radar, the profile data corresponding to each planar pixel in the vertical direction is obtained. Based on the linear interpolation of the vertical profile, the height corresponding to the radar reflectivity of the corresponding pixel in the vertical direction reaches 20dBz is calculated. Finally, the longitude, latitude, time and area of ​​the overshoot cloud top events in central and eastern China from June to July 2019, where the radar reflectivity reached 20dBz at a height of more than 14km, are recorded and integrated into an overshoot cloud top area validation set, which contains 97 validation samples.

[0103] The accuracy of the geostationary satellite identification method was verified using a validation set of overshoot cloud tops generated by a dual-frequency rain-measuring radar, while the accuracy of geostationary satellite area identification was verified using a validation set of overshoot cloud top areas generated by a weather radar. The two datasets overlapped to some extent but were not identical. The reason for creating two validation sets is that the spatial resolution of the dual-frequency rain-measuring radar is relatively low, about 5 km, while the radius of overshoot cloud tops is generally less than 15 km. Therefore, although the dual-frequency rain-measuring radar can obtain the distribution of overshoot cloud tops over a wide area, it is difficult to accurately quantify the area of ​​overshoot cloud tops at a small scale. Although the coverage of the weather radar is limited, its spatial resolution is higher and it can provide more accurate area reference values, which can be used to verify the area of ​​overshoot cloud tops.

[0104] Step 2: Perform spatiotemporal matching between samples in the overshoot cloud top validation set and the overshoot cloud top area validation set and satellite observation images:

[0105] Based on the recording time of each sample in the overshoot cloud top verification set and the overshoot cloud top area verification set, the observation images of the Himawari-8 geostationary satellite are matched within a 10-minute time window before and after the recording time point. The spatial matching rule is as follows: since the infrared brightness temperature at the center of the overshoot cloud top is the lowest point of brightness temperature in a circular area with a radius of 15km centered on its latitude and longitude, the verification samples in the overshoot cloud top verification set and the overshoot cloud top area verification set are matched to the coldest infrared brightness temperature pixel point in a circular area with a radius of 15km centered on its latitude and longitude, and the coldest infrared brightness temperature pixel point is determined as the matching point of the verification sample on the satellite observation image.

[0106] Step 3: Verification, Judgment, and Evaluation:

[0107] For each identified overshoot cloud vertex, if a matching point of any validation sample exists within a 15km radius around its corresponding location, it is considered a successful identification, i.e., a hit; otherwise, it is considered a false alarm. Simultaneously, all validation set samples that were not successfully identified by the method of this invention are counted and recorded as missed alarms. Based on this, the probability of detection (POD) and the false alarm rate (FAR) are calculated, with their optimal values ​​being 1 and 0, respectively.

[0108] Hit rate metric Represented as:

[0109] ,

[0110] False alarm rate metric Represented as:

[0111] ,

[0112] in, This indicates that both the predicted and observed data are positive. This indicates that the predicted data is negative and the observed data is positive. This indicates that the predicted data is positive and the observed data is negative; the overshoot cloud top identification results are evaluated based on the hit rate and false alarm rate to obtain the overshoot cloud top identification capability evaluation of the method of the present invention;

[0113] The calculated area of ​​the successfully identified overshoot cloud apex and its corresponding overshoot cloud top region is matched one-to-one with the weather radar estimated area of ​​the matched verification sample, and the correlation coefficient between the two sets of areas is calculated. Based on the correlation coefficient, the overshoot cloud top area calculated by the method of the present invention and the overshoot cloud top area calculated by the weather radar are fitted to obtain the overshoot cloud top area calculation capability evaluation of the method of the present invention.

[0114] Correlation coefficient Represented as:

[0115] ,

[0116] Where X and Y represent the overshoot cloud top area calculated using the method of this invention and the overshoot cloud top area calculated using weather radar, respectively. Let X be the covariance between X and Y. Let X be the variance. Let Y be the variance.

[0117] like Figure 4 As shown, the overshoot cloud top obtained by the method of the present invention is compared with the overshoot cloud top validation set and the overshoot cloud top area validation set, and the hit rate is 0.887, the false alarm rate is 0.359, and the area correlation coefficient is 0.849, which shows good recognition performance.

[0118] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0119] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for identifying overshoot cloud tops and calculating their area based on geostationary satellite data, characterized in that, Includes the following steps: S1. Obtain geostationary satellite cloud radiation brightness temperature data and tropopause temperature data for the target area within the required time period; S2. Read the acquired geostationary satellite cloud radiation brightness temperature data and extract the cloud radiation brightness temperature from the three channels of water vapor, ozone and infrared. S3. Unify the temporal and spatial resolution of geostationary satellite cloud radiative brightness temperature data and tropopause temperature data; S4. Based on all the data, identify potential overshoot cloud top pixels using a multi-channel threshold discrimination method; S5. Based on the infrared channel brightness temperature data extracted in step S2, the potential overshoot cloud top region and its anvil cloud region of the potential overshoot cloud top pixel are extracted using the isotherm analysis method. S6. Subtract the infrared channel brightness temperature of all pixels in the anvil cloud region from the infrared channel brightness temperature of the potential overshoot cloud top pixel. If the difference is greater than or equal to 5K, then the potential overshoot cloud top pixel is determined as the overshoot cloud top, and the corresponding potential overshoot cloud top region is determined as the overshoot cloud top region. S7. Calculate the overshoot cloud top area based on the identified overshoot cloud top region.

2. The method for identifying overshoot cloud tops and calculating their area based on geostationary satellite data according to claim 1, characterized in that, In step S1, the preset temporal resolution of the geostationary satellite cloud radiative brightness temperature data is 1. Spatial resolution is The temporal resolution of the tropopause temperature data is [missing information]. Spatial resolution is The number of frames acquired for geostationary satellite cloud radiative brightness temperature data is: The number of frames for tropopause temperature data is , , These represent the start and end times for data acquisition, respectively.

3. The method for identifying overshoot cloud tops and calculating their area based on geostationary satellite data according to claim 2, characterized in that, Step S3 employs a spatiotemporal matching algorithm to unify the temporal and spatial resolutions of geostationary satellite cloud radiative brightness temperature data and tropopause temperature data. Specific steps include: S31. In terms of time, for each frame of geostationary satellite cloud radiation brightness temperature data, select the tropopause temperature data of the frame with the closest time to it for matching. S32. In space, linear interpolation is performed on the tropopause temperature data to make its spatial resolution equal to that of the geostationary satellite cloud radiation brightness temperature data.

4. The method for identifying overshoot cloud tops and calculating their area based on geostationary satellite data according to claim 3, characterized in that, The specific steps of step S4 include: S41. For each observation pixel in each frame of geostationary satellite cloud radiation brightness temperature data, compare its infrared channel brightness temperature with the corresponding tropopause temperature. If its infrared channel brightness temperature is less than the corresponding tropopause temperature, proceed to the next step; otherwise, determine that the observation pixel is not an overshoot cloud top pixel. S42. Calculate the difference between the brightness temperature of the water vapor channel at the observed pixel and the corresponding tropopause temperature. If the difference is less than or equal to 4K, proceed to the next step; otherwise, determine that the observed pixel is not an overshoot cloud top pixel. S43. Calculate the difference between the brightness temperature of the ozone channel and the brightness temperature of the water vapor channel at the observed pixel. If the difference is greater than or equal to a preset threshold, the observed pixel is listed as a potential overshoot cloud top pixel; otherwise, the observed pixel is determined not to be an overshoot cloud top pixel. The preset threshold is set according to the latitude of the target area.

5. The method for identifying overshoot cloud tops and calculating their area based on geostationary satellite data according to claim 4, characterized in that, The specific steps of step S5 include: S51. For each frame of geostationary satellite cloud radiation brightness temperature data, extract the infrared channel brightness temperature of the current frame; S52. For any potential overshoot cloud top pixel identified in the current frame, take its infrared channel brightness temperature as the local minimum value, and draw isotherms around the potential overshoot cloud top pixel using 1K intervals based on the infrared channel brightness temperature of the current frame. S53. Starting from the isotherm closest to the potential overshoot cloud top pixel, examine from the inside out, and determine the isotherm that first meets one of the following two conditions as the potential overshoot cloud top isotherm: a) The average distance between all pixels on the isotherm and the potential overshoot cloud top pixel is less than or equal to 12km, but the average distance between all pixels on the nearest neighboring isotherm and the potential overshoot cloud top pixel is greater than or equal to 12km. b) The difference between the infrared channel brightness temperature of the isotherm and the infrared channel brightness temperature at the potential overshoot cloud top pixel is greater than or equal to 8K; S54. Continue the inspection outward from the potential overshoot cloud top isotherm, and determine the isotherms that first meet the following conditions as the outer isotherms: c) The average distance between all pixels on the isotherm and the potential overshoot cloud top pixel is greater than or equal to 50 km; d) The difference between the infrared channel brightness temperature of the isotherm and the infrared channel brightness temperature at the potential overshoot cloud top isotherm is greater than or equal to 10K; S55. The annular region between the potential overshoot cloud top isotherm and the outer isotherm is defined as the anvil-shaped cloud region, and the region within the potential overshoot cloud top isotherm is defined as the potential overshoot cloud top region.

6. The method for identifying overshoot cloud tops and calculating their area based on geostationary satellite data according to claim 5, characterized in that, The specific steps for calculating the area of ​​the overhanging cloud top in step S7 are as follows: S71. For any overshoot cloud top region, a universal transverse Mercator projection is used, and the projection partition is dynamically selected according to the longitude of the overshoot cloud top pixels, so as to project the overshoot cloud top region from the geographic coordinate system to the plane coordinate system. S72. In a plane coordinate system, the area of ​​the region passing through the cloud top can be directly calculated using the Gaussian area formula. The specific formula is as follows: ， in, For the area of ​​the Chongyunding region, ( ) is the polygon corresponding to the overpass cloud top region. The coordinates of the vertices are sorted in clockwise or counterclockwise order. This represents the total number of vertices of the polygon.

7. A device for identifying overshoot cloud tops and calculating area based on geostationary satellite data, characterized in that, The method for identifying overshoot cloud tops and calculating their area based on geostationary satellite data as described in any one of claims 1 to 6 includes: The data receiving and caching module is used to receive, preprocess, and transmit geostationary satellite cloud radiation brightness temperature data and tropopause temperature data from external sources. The data processing unit is used to schedule and calculate the geostationary satellite cloud radiation brightness temperature data and tropopause temperature data transmitted by the data receiving and caching module, and generate result data including overshoot cloud top region identification and area calculation. The result output unit is used to encapsulate the result data output by the data processing unit and transmit it to the host system.

8. The device for identifying and calculating overshoot cloud tops based on geostationary satellite data according to claim 7, characterized in that, The data receiving and caching module includes: a corresponding data interface for near real-time acquisition of geostationary satellite cloud radiation brightness temperature data and tropopause temperature data; a cache control unit for aligning and scheduling data from different sources based on timestamp information, and performing preprocessing operations such as data frame synchronization, integrity verification, and field parsing; a high-speed FIFO cache for writing preprocessed data to achieve time decoupling between the data receiving process and the subsequent calculation process; and an onboard DDR memory for persistently storing preprocessed data within a preset time window, providing stable and continuous data input to the data processing unit when the calculation conditions are met.

9. The device for identifying and calculating overshoot cloud tops based on geostationary satellite data according to claim 8, characterized in that, The data processing unit is composed of a programmable logic device (PLD) and an embedded processor. The PLD is used to implement data control, access the onboard DDR memory, and schedule the calculation process. The embedded processor is used to load the pre-stored calculation program from the onboard non-volatile memory. The computer program is the computer instruction corresponding to the steps of the overshoot cloud top identification and area calculation method. During operation, the PLD is scheduled to read the spatiotemporally aligned geostationary satellite cloud radiation brightness temperature data and tropopause temperature data from the onboard DDR memory in chronological order, and complete the overshoot cloud top identification and area calculation according to the set algorithm. The calculation result data is temporarily stored in an internal buffer. After the calculation is completed, an output instruction is sent to the result output unit.

10. The device for identifying and calculating overshoot cloud tops based on geostationary satellite data according to claim 9, characterized in that, After receiving the output instruction sent by the data processing unit, the result output unit reads the result data and encapsulates the result data according to the preset data format; The packaged result data is sent to the host system in real time via a serial communication interface according to a preset communication protocol. The output data should include at least the timestamp, overshoot cloud top location information, and the corresponding area calculation results.

Images