A method for automatically identifying melt layers using S-band dual-polarization radar

Abstract

The present invention discloses a method for automatically identifying a melting layer using an S-band dual-polarization radar. The method improves the existing MLDA identification method, including adding a radial continuity test on the basis of marking the melting layer on the radar scanning azimuth, increasing the scanning range by 3.3° elevation when the melting layer area is determined, and adjusting the threshold of the total number of distance libraries for distinguishing the occurrence of the melting layer. The method can remove the influence of non-precipitation echoes such as clear sky and ground objects on the MLDA identification of the melting layer to a certain extent, the deviation of the identified melting layer top height is small, and the melting layer bottom height is within a reasonable range; and the adjusted parameters are more suitable for my country's Doppler S-band dual-polarization radar, and the identification efficiency is much higher than the original MLDA.

Description

Technical Field

[0001] The invention relates to the field of atmospheric science and technology, and in particular to a method for automatically identifying a melting layer by using an S-band dual-polarization radar. Background Art

[0002] The melt layer is the area where solid and liquid hydrometeors in the atmosphere transform. The top of the melt layer is usually the lowest altitude where the 0℃ isotherm is located. Above the melt layer are mainly solid hydrometeors, and below are mainly liquid hydrometeors. The melt layer plays an important role in many business applications. For example, the top of the melt layer can be used to classify the phases of hydrometeors. Some hydrometeors, such as dry snow and moderate (light) rain, have roughly the same radar parameter characteristics, and they can be effectively distinguished by the height of the melt layer top. The position of the melt layer can be used to correct the error of radar quantitative estimation of precipitation in the bright band area. The melt layer can also be used in radar data quality control business. Domestic and foreign scholars use the zero-lag correlation coefficient of radar parameters to distinguish precipitation and non-precipitation echoes. In the melt layer area, the zero-lag correlation coefficient characteristics of the two are very similar. By identifying the position of the melt layer, the precipitation echo can be protected from being filtered out.

[0003] As more and more operational radars in my country are upgraded to dual-polarization radars, using dual-polarization radar parameters to identify the melt layer has become an effective way. Compared with sounding data, radar data can provide information on the location of the melt layer with higher temporal and spatial resolution. Giangrande et al. (2008) used the reflectivity factor (abbreviated as Z) of the observation parameter of the WSR-88D dual-polarization radar in the United States to identify the melt layer. H ), differential reflectivity (abbreviated as Z DR ) and the zero-lag correlation coefficient (CC) in the characteristics and position of the melt layer, and designed a melt layer identification method (MLDA), which was used in the U.S. hydrometeor classification (HCA) business; Boodoo et al (2010) used MLDA to test two frontal systems. The results showed that when the frontal system passed, compared with the aircraft reporting data, MLDA could effectively detect the rapid changes in the height of the melt layer.

[0004] Through some test results (see Table 2 for test cases), it was found that there are certain defects in the use of MLDA in my country's Doppler S-band dual-polarization radar to identify the melting layer. The error of the top height of the melting layer identified is too large, the bottom height of the melting layer is too low, and a large number of melting layers in the tests were not identified. Summary of the invention

[0005] In view of the problems that arise when MLDA is used in my country's Doppler S-band dual polarization radar to identify the melting layer, such as large errors in identifying the top of the melting layer, low height of the bottom of the melting layer and low identification efficiency, the present invention provides a method for automatically identifying the melting layer using S-band dual polarization radar.

[0006] In order to solve the technical problem, the present invention adopts a solution: a method for automatically identifying a melt layer using an S-band dual-polarization radar, comprising the following steps:

[0007] S01) Mark the melting layer of the radar scanning azimuth; detect all the distance libraries on any radar scanning azimuth in order from small to large radial distance from the radar. If its CC distribution is in [0.9, 0.97] and there is a Z within 500m above the distance library, H Distributed in [30, 47] dBZ, Z DR If the range is [0.8, 2.5] dB, the distance library is judged to be in the melting layer area and marked, otherwise it is judged to be in the non-melting layer area. In the 4-10° elevation angle, when the height of the distance library is higher than 6 km, the azimuth detection ends;

[0008] S02), marking the melting layer of the radar volume scan; repeating step S01) until all azimuth angles of each elevation layer in the radar volume scan are detected;

[0009] S03), determining the melting layer area; sequentially counting the number of distance bins marked at 10° azimuths before and after each azimuth when the radar scans at an elevation angle of 3-10°; if the total number of marked distance bins exceeds 1500, then the azimuth is determined to be in the melting layer area;

[0010] S04), determine the top and bottom heights of the melting layer; at the azimuth angle where the melting layer is determined to appear, sort the marked distance library of the azimuth angle 10° before and after it from low to high according to the height, and take the 80th percentile of the height sequence as the top height of the melting layer at that azimuth, and the 20th percentile as the bottom height of the melting layer.

[0011] Furthermore, a radial continuity check is added in step S01), and the specific steps are as follows: after all distance bins of one azimuth angle have been detected, the distance bins of the azimuth angle marked as the melting layer are detected again. If the number of distance bins marked as the melting layer within the range of 500 m above and below the melting layer accounts for more than 40% of the total number of distance bins within the detection range, then it is determined that the distance bin belongs to the melting layer.

[0012] Further, in step S03), the scanning range is increased by 3.3° elevation angle, that is, the number of distance bins marked at 10° azimuth angles before and after each azimuth angle is counted at 3.3°, 4.3°, 6.1° and 9.9° elevation angles.

[0013] Furthermore, in step S04), the total thresholds of the distance libraries for judging the presence of melting layers at elevation angles of 3.3°, 4.3°, 6.1° and 9.9° are set to 550, 450, 300 and 200 respectively. If a layer is judged to have a melting layer in more than 10 directions, then a melting layer is judged to exist at this elevation angle, and its corresponding threshold is added to the total threshold.

[0014] Beneficial effects of the present invention: The present invention improves the existing MLDA identification method, and can remove the influence of non-precipitation echoes such as clear sky and ground objects on MLDA identification of the melting layer to a certain extent. The deviation of the identified melting layer top height is small, and the melting layer bottom height is within a reasonable range; and the adjusted parameters are more suitable for my country's Doppler S-band dual-polarization radar, and the identification efficiency is much higher than the original MLDA. There are very few domestic documents or technical documents on the automatic identification of the melting layer using Doppler S-band dual-polarization radar. This optimization method fills the gap in this field in China. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a schematic diagram of the scattered distribution of the 0℃ height of each test sounding and the height of the top of the melting layer identified by MLDA;

[0016] Figure 2 Schematic diagram of the temperature (a) and the thickness of the melting layer (b) corresponding to the height of the bottom of the melting layer identified by MLDA for each test;

[0017] Figure 3 The elevation angle CC(a), Z of 4.3° at Jinan radar station at 23:57 UTC on August 14, 2020 H (b) with Z DR (c) Distribution diagram;

[0018] Figure 4 Schematic diagram of the number of tests where the melting layer characteristics appear at each elevation angle;

[0019] Figure 5 The average vertical profile distribution of CC at 2.4°, 3.3°, 4.3° and 6.0° elevation angles at the Jinan radar station in the 180-210° azimuth at 00:02 UTC on August 5, 2020. The black dotted line is the 0°C height of the sounding, and the blue dotted line is the CC equal to 0.97 contour line;

[0020] Figure 6 This is a schematic diagram of the distribution of the height corresponding to the distance library at each azimuth angle at 23:57 UTC on August 14, 2020, which was judged as the melting layer at an elevation angle of 4.3° by the Jinan radar station. (a) is before the radial continuity test, and (b) is after the radial continuity test.

[0021] Figure 7 This is a schematic diagram of the scattered distribution of the 0℃ height of each test sounding and the height of the top of the melting layer identified by MLDA-R1;

[0022] Figure 8 Schematic diagram of the temperature (a) and the thickness of the melting layer (b) corresponding to the height of the bottom of the melting layer identified by MLDA-R1 for each test;

[0023] Fig. 9 This is a schematic diagram of the scattered distribution of the 0℃ height of each test sounding and the height of the top of the melting layer identified by MLDA-R2;

[0024] Fig.10 Schematic diagram of the temperature (a) and the thickness of the melt layer (b) corresponding to the height of the bottom of the melt layer identified by MLDA-R2 for each test;

[0025] Fig.11 This is a schematic diagram of the scattered distribution of the 0℃ height of each test sounding and the height of the top of the melting layer identified by MLDA-R3;

[0026] Fig.12 Schematic diagram of the temperature (a) and the thickness of the melting layer (b) corresponding to the height of the bottom of the melting layer identified by MLDA-R3 for each test. DETAILED DESCRIPTION

[0027] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0028] Example 1

[0029] In order to check the recognition effect of the existing MLDA on the melting layer and find out the problems and shortcomings of MLDA, this embodiment uses the Doppler S-band dual-polarization radar volume scanning data of Jinan and Qingdao stations to carry out recognition experiments. Since the melting layer usually appears in the stratiform cloud precipitation area, the data of the stratiform cloud precipitation period in Jinan and Qingdao from July to August 2020 are selected. The specific time is shown in Table 1.

[0030] Table 1 The periods of stratiform cloud precipitation in Jinan and Qingdao from July to August 2020

[0031] Site Stratiform precipitation period (UTC) Site Stratiform Cloud Period (UTC) Z9531 July 2nd 07:59:48-July 3rd 01:02:05 Z9532 July 11th 15:57:22-July 12th 06:58:35 Z9531 July 8 22:59:57-9 July 05:59:48 Z9532 July 18th 16:58:05-July 19th 09:59:02 Z9531 July 11 10:00:19-23:00:25 Z9532 July 21, 18:58:39-July 22, 17:02:15 Z9531 July 17 02:03:37-05:02:38 Z9532 July 26 04:01:34-08:58:02 Z9531 July 18th 18:59:58-19th 00:57:50 Z9532 July 31 000035-135903 Z9531 July 21, 22:58:09-July 22, 17:03:08 Z9532 August 1, 16:58:34-August 2, 05:59:56 Z9531 July 25th 14:00:18-26th 03:00:02 Z9532 August 3, 12:58:44-14:58:28 Z9531 August 1 11:01:06-23:59:14 Z9532 August 3, 18:57:54-August 4, 03:59:50 Z9531 August 3, 09:01:53-12:00:21 Z9532 August 5, 12:01:11-16:00:36 Z9531 August 4th 22:01:00-August 5th 10:02:36 Z9532 August 6th 03:57:10-August 7th 13:58:17 Z9531 August 5th 23:01:36-August 6th 16:57:16 Z9532 August 8 23:02:36-9 August 01:59:45 Z9531 August 11, 23:58:17-12, 04:58:32 Z9532 August 11, 18:59:59-12, 02:02:28 Z9531 August 12, 13:02:36-13, 00:57:37 Z9532 August 13th 11:58:09-14th 19:02:43 Z9531 August 14th 18:00:14-15th 09:57:53 Z9532 August 19 10:58:23-16:57:32 Z9531 August 19, 19:57:41-20, 12:01:26 Z9532 August 20, 18:59:56-21, 04:59:09 Z9532 July 2nd 16:00:43-July 3rd 03:59:00 Z9532 August 25, 18:01:37-26, 13:02:40

[0032] The 0°C height of the sounding data was used to verify the top height of the melting layer identified by MLDA. The sounding data was obtained from the MICAPS system of the China Meteorological Administration. The longitudes and latitudes of the Jinan and Qingdao sounding stations were (117.52°E 36.65°N) and (120.33°E36.07°N), respectively, and the longitudes and latitudes of the radar stations were (116.71°E 36.71°N) and (120.23°E 35.99°N), respectively. The distances between the sounding and radar of the two stations were approximately 73km and 13km, which were representative. Since the sounding station data were obtained at 08:00 and 20:00 (Beijing time) every day, in order to objectively test the recognition effect of the method, radar data with a time of no more than 5 minutes from the sounding observation were selected for the test. The specific test cases are shown in Table 2, with a total of 41.

[0033] Table 2 Inspection time of individual cases

[0034]

[0035]

[0036] The indicators for testing the height of the top of the melting layer include the mean absolute error (MAE) and the correlation coefficient (RC). The formulas are shown in (1) and (2).

[0037]

[0038]

[0039] Where ml is the height of the top of the melt layer, o is the 0°C height of the sounding data, n is the number of experiments, Cov is the covariance, and Var is the variance.

[0040] The height of the bottom of the melting layer varies greatly and cannot be obtained through a certain temperature threshold. It depends on the air humidity conditions and the concentration of precipitation particles. The study found that the wet-bulb temperature corresponding to the bottom height of the melting layer is 2°C. According to the conversion table of wet-bulb temperature and temperature, when the relative humidity is 100%, the air temperature is 2°C. As the relative humidity decreases, the corresponding temperature increases. When the relative humidity is 15%, the air temperature is 10°C. Based on this, when the identified bottom height of the melting layer corresponds to a temperature distribution of 2-10°C, the identification is accurate.

[0041] The scattered distribution of the 0℃ height of each test sounding and the height of the top of the melting layer identified by MLDA ( Figure 1 ) It can be seen that some observation tests are significantly higher or lower than the sounding 0℃ height, mainly with negative deviations. Statistics show that the maximum positive error is 0.623km, the maximum negative error is 1.096km, there are 16 tests where MLDA did not identify the melting layer, and the MAE of the top height of the melting layer calculated by the other 25 tests is 0.328km, and the RC is 0.418.

[0042] The sounding temperature corresponding to the height of the melting layer identified by MLDA ( Figure 2 a) It was found that only 6 tests were distributed in the range of 2-10℃, and the temperature of other tests exceeded 10℃, and some tests exceeded 20℃. The temperature was too high, so MLDA identified that the bottom height of the melting layer was too low. Figure 2 b It can be seen that the thickness of the melting layer calculated by MLDA is mainly distributed between 2-4 km. The observed thickness of the melting layer calculated by foreign scholars for a long time is close to 0.5 km. The thickness of the melting layer calculated by Giangrande et al (2008) is 0.2-0.6 km. Therefore, the thickness of the melting layer calculated by MLDA is relatively large.

[0043] From the above results, it can be seen that MLDA can identify the position of the melting layer in most observation tests, but there are certain defects. The error of the top height of the melting layer identified in some observation tests is too large, the bottom height of the melting layer is too low, and the melting layer of a large number of tests is not identified.

[0044] In view of the defects of the existing MLDA identification method, this embodiment analyzes the causes and makes improvements to enhance the recognition effect of the melting layer, and provides a method suitable for automatic identification of the melting layer by Doppler S-band dual-polarization radar in my country, in order to provide more accurate melting layer location information for services such as hydrometeor classification and radar quantitative precipitation estimation.

[0045] In order to solve the technical problem, the technical solution adopted by the present invention is:

[0046] S01) Mark the melting layer of the radar scanning azimuth; detect all the distance libraries on any radar scanning azimuth in the order of radial distance from the radar from small to large. If its CC distribution is in [0.9, 0.97] and there is a Z within 500m above the distance library, H Distributed in [30, 47] dBZ, Z DR If the range is [0.8, 2.5] dB, the distance library is judged to be in the melting layer area and marked, otherwise it is judged to be in the non-melting layer area. In the 4-10° elevation angle, when the height of the distance library is higher than 6 km, the azimuth detection ends;

[0047] S02), marking the melting layer of the radar volume scan; repeating step S01) until all azimuth angles of each elevation layer in the radar volume scan are detected;

[0048] S03), determining the melting layer area; sequentially counting the number of distance bins marked at 10° azimuths before and after each azimuth when the radar scans at an elevation angle of 3-10°; if the total number of marked distance bins exceeds 1500, then the azimuth is determined to be in the melting layer area;

[0049] S04), determine the top and bottom heights of the melting layer; at the azimuth angle where the melting layer is determined to appear, sort the marked distance library of the azimuth angle 10° before and after it from low to high according to the height, and take the 80th percentile of the height sequence as the top height of the melting layer at that azimuth, and the 20th percentile as the bottom height of the melting layer.

[0050] This embodiment adds a radial continuity test in step S01). Figure 3 The CC and Z values ​​of the 4.3° elevation layer at Jinan Radar Station at 23:57 on August 14, 2020 are given. H With Z DR Distribution. Figure 3 As can be seen from a, in the black dashed rectangular area, precipitation echoes and non-precipitation echoes (clear sky echoes and ground object echoes, etc.) are mixed, resulting in the CC distribution of many range libraries in the interval [0.9, 0.97]. H ( Figure 3 b) With Z DR ( Figure 3 c) It also meets the MLDA discrimination conditions, but these distance libraries do not belong to the melting layer area. Since these mixed echoes are not easy to be filtered out through quality control, these distance libraries that are mistakenly identified as melting layers make the bottom height of the melting layer corresponding to the 20th percentile of the height sequence too low, and also affect the accuracy of the top height of the melting layer.

[0051] Depend on Figure 3 a, 3c (solid line area), it can be seen that in the melting layer area, the CC low value area and Z DR The high-value area has strong continuity in the radial direction, while the incorrectly identified distance bins have poor continuity in the radial direction. In order to eliminate these false distance bins, a radial continuity test is added to MLDA. The specific steps are as follows: after all distance bins of an azimuth are tested, the distance bins marked as melting layers in that azimuth are tested again. If the number of distance bins marked as melting layers within the range of 500m above and below it accounts for more than 40% of the total number of distance bins in the detection range, then this distance bin is judged to belong to the melting layer. This improvement is called MLDA-R1.

[0052] The original MLDA uses 4-10° elevation layer data. This is because there are too few distance libraries in the melting layer area for elevation layers above 10°, which will not affect the discrimination threshold. However, for elevations below 4°, the marked distance library will overestimate the melting layer position due to radial broadening. The US WSR-88D dual-polarization radar has 6 elevation angles in this range, namely 4.5°, 5.5°, 6.5°, 7.5°, 8.7° and 10.0°, while my country's Doppler S-band dual-polarization radar has only three elevation angles, namely 4.3°, 6.1° and 9.9°. Figure 4The number of tests with melting layer characteristics at 9 elevation angles was counted, and it can be seen that the elevation angles with melting layer characteristics are mainly 3-6 layers. The elevation angles of 3-4 layers, i.e. 2.4° and 3.3°, are not within the scanning range.

[0053] Average CC vertical profile distribution of 3-6 layers of Jinan radar station at 180-210° elevation angle at 00:02 on August 5, 2020 ( Figure 5 ) It can be found that the 3.3° elevation profile feature is roughly the same as the 4.3° elevation. Due to radial broadening, the height corresponding to the 3.3° elevation CC of 0.97 (the upper intersection point where the profile and the contour line intersect) is slightly greater than the top of the melting layer, while the height corresponding to the 2.4° elevation CC of 0.97 is significantly greater than the top of the melting layer. The 80th percentile of the height sequence can filter out most of the distance libraries with 3.3° elevation angles that are higher than the top of the melting layer, while the distance libraries with 2.4° elevation angles that are higher than the top of the melting layer will not be effectively filtered out, which will cause a serious overestimation of the top of the melting layer. In order to improve the successful recognition rate of the melting layer and avoid large errors in the identified melting layer, the scanning range is changed to 3-10° elevation.

[0054] Specifically, based on MLDA-R1, the number of distance libraries marked at 10° azimuth before and after each azimuth is counted at elevation angles of 3.3°, 4.3°, 6.1° and 9.9°. This improvement is called MLDA-R2.

[0055] Depend on Figure 4 It can be seen that not all elevation angles will show melting layer features, which makes it difficult for the total number of distance bins to reach the threshold (i.e., 1500) for MLDA to distinguish the occurrence of melting layer. Therefore, it is necessary to adjust the total number of distance bin thresholds for identifying melting layer according to the number of radar elevation angle layers where melting layer features appear. After adding radial continuity test and adding 3.3° elevation angle to the scanning range, if a radial is judged to have a melting layer, then the number of distance bins at 3.3°, 4.3°, 6.1°, and 9.9° elevation angles is approximately 28, 22, 16, and 10. Based on this, the total number of distance bin thresholds for distinguishing the occurrence of melting layer is adjusted on the basis of MLDA-R2. The steps are as follows: the total number of distance bin thresholds for distinguishing the occurrence of melting layer at 3.3°, 4.3°, 6.1°, and 9.9° elevation angles are set to 550, 450, 300, and 200, respectively. If a layer is judged to have a melting layer at more than 10 azimuths, then the melting layer is judged to appear at that elevation angle, and its corresponding threshold is accumulated to the total number of thresholds. For example, if a melting layer is detected only at an elevation angle of 3.3°, the threshold of the total distance library is adjusted to 550. Similarly, if a melting layer is detected at 3.3°, 4.3°, 6.1°, and 9.9°, the threshold is adjusted to 1500. This improvement is called MLDA-R3.

[0056] Figure 6The distribution of the distance bin height in all azimuths of the melt layer at 4.3° elevation at Jinan radar station before and after adding radial continuity test at 23:57 on August 14, 2020 is given. Figure 6 a shows that a large number of distance bins in the non-melting layer area are misidentified by MLDA (areas outside the dotted line), while the number of misidentified distance bins is significantly reduced after the radial continuity test is added ( Figure 6 b) The remaining non-melting layer distance library has good radial continuity, so MLDA-R1 cannot effectively filter it out, but the number is small and will not have a significant impact on the identification of the melting layer. In this test, the bottom and top heights of the melting layer calculated by MLDA were 1.196km and 4.847km respectively; the bottom and top heights of the melting layer calculated by MLDA-R1 were 4.263km and 5.072km respectively, while the 0℃ height of the sounding was 5.045km. It is not difficult to find that with the addition of radial continuity test, the deviation between the top height of the melting layer and the 0℃ height of the sounding is smaller, and the thickness of the melting layer is more reasonable.

[0057] The scatter distribution of the top of the melting layer identified by MLDA-R1 and the 0℃ height of the sounding ( Figure 7 ) can be seen, with Figure 1 Compared with MLDA, the scattered points are closer to the diagonal line, and the number of negative errors is significantly reduced. The MAE of the top height of the melting layer identified by MLDA-R1 is 0.246km, and the RC is 0.631. However, only 10 experiments identified the melting layer, and the number of missed experiments increased significantly compared with MLDA. This is because a large number of incorrectly identified distance libraries were removed by using the radial continuity test, and the total number of distance libraries in many experiments did not reach the discrimination threshold (1500).

[0058] Figure 8 a and b respectively show the temperature and thickness distribution of the melting layer corresponding to the bottom height of the melting layer identified by MLDA-R1. The temperature corresponding to the bottom height of the melting layer ( Figure 8 a) Mainly distributed in the range of 2-10℃, the thickness of the melting layer ( Figure 8 b) It is mainly distributed between 0.7 and 1 km. Compared with MLDA, the distribution of melting layer bottom height and melting layer thickness is more reasonable.

[0059] Fig. 9The scatter distribution of the melting layer top height identified by MLDA-R2 and the 0℃ height of the sounding is given for each test. The MAE of the melting layer top height is 0.217km, and the RC is 0.704, which is slightly better than MLDA-R1. Compared with MLDA-R1, the number of missed tests is significantly reduced to 17, which is close to MLDA. There are 24 tests that identify the melting layer, which is much larger than MLDA-R1. This is mainly due to the increase of 3.3° elevation angle, and the number of qualified distance libraries has increased significantly. In most tests, the melting layer height identified by MLDA-R2 is slightly higher than the 0℃ height of the sounding. This is because the newly added 3.3° elevation layer is affected by radial broadening, and the height of some marked distance libraries is slightly higher than the melting layer top.

[0060] The temperature corresponding to the bottom height of the melting layer identified by MLDA-R2 in each test ( Fig.10 a) Distributed at 2-8℃, identify the thickness of the melting layer ( Fig.10 b) It is mainly distributed in the range of 0.5-1 km. It can be seen that the increase of 3.3° elevation angle does not affect the rationality of the height and thickness of the bottom of the melting layer.

[0061] The scatter distribution of the top of the melting layer identified by MLDA-R3 and the 0℃ height of the sounding ( Fig.11 ) It can be seen that only 4 tests were not recognized by MLDA-R3, and the recognition success rate was further improved. The MAE of the melt layer top height was 0.214km, and the RC was 0.678. The adjustment of the threshold of the total number of distance libraries did not cause excessive positive and negative errors in the melt layer top height identified by MLDA-R3.

[0062] From the temperature distribution (12a) and the melting layer thickness distribution (12b) corresponding to the bottom height of the melting layer identified by MLDA-R3, it can be seen that there were two tests in which the temperature corresponding to the bottom of the melting layer was too high and the melting layer thickness was too large, namely, at Jinan Station at 23:59 on August 12 and 23:59 on August 19. In these two tests, the melting layer was not identified by MLDA, MLDA-R1 and MLDA-R2. The temperature corresponding to the bottom height of the melting layer and the melting layer thickness distribution of other tests were relatively reasonable.

[0063] The above description is only the basic principle and preferred embodiments of the present invention. Improvements and substitutions made by those skilled in the art based on the present invention belong to the protection scope of the present invention.

Claims

1. A method for automatically identifying a melt layer using an S-band dual-polarization radar, characterized in that: The following steps are involved: S01) Mark the melting layer of the radar scanning azimuth; detect all the distance libraries on any radar scanning azimuth in order from small to large radial distance from the radar. If the zero-lag correlation coefficient is distributed in [0.9, 0.97], and within 500m above the distance library, there is a reflectivity factor distributed in [30, 47] dBZ and a differential reflectivity distributed in [0.8, 2.5] dB, then the distance library is determined to be in the melting layer area and marked, otherwise it is determined to be in the non-melting layer area. When the height of the distance library is higher than 6km, the azimuth detection ends; S02), marking the melting layer of the radar volume scan; repeating step S01) until all azimuth angles of each elevation layer in the radar volume scan are detected; S03) Determine the melting layer area; in the radar body scan 3-10° elevation angle, count the number of distance bins marked at 10° azimuth angles before and after each azimuth angle in sequence; if the total number of marked distance bins exceeds 1500, then the azimuth is determined to be in the melting layer area; S04) Determine the top and bottom heights of the melting layer; For the azimuth angle where the melting layer is determined to appear, the marked distance library of the 10° azimuth angle before and after it is sorted from low to high according to height, and the 80th percentile of the height sequence is taken as the top height of the melting layer at this azimuth angle, and the 20th percentile is taken as the bottom height of the melting layer.

2. The method for automatically identifying a melt layer using an S-band dual-polarization radar according to claim 1, characterized in that: Step S01) adds radial continuity test, and the specific steps are as follows: after all distance bins of an azimuth angle are detected, the distance bins marked as melting layers in the azimuth angle are detected again. If the number of distance bins marked as melting layers within the range of 500m above and below it accounts for more than 40% of the total number of distance bins in the detection range, then it is determined that this distance bin belongs to the melting layer.

3. The method for automatically identifying a melt layer using an S-band dual-polarization radar according to claim 1, characterized in that: In step S03), an elevation angle of 3.3° is added, that is, the number of distance bins marked with azimuth angles 10° before and after each azimuth angle is counted at elevation angles of 3.3°, 4.3°, 6.1° and 9.9°.

4. The method for automatically identifying a melt layer using an S-band dual-polarization radar according to claim 1, characterized in that: In step S04), the total threshold of the distance library for judging the occurrence of a melting layer is adjusted. Specifically, the total threshold of the distance library for judging the occurrence of a melting layer at elevation angles of 3.3°, 4.3°, 6.1° and 9.9° is set to 550, 450, 300 and 200 respectively. If a layer is judged to have a melting layer at more than 10 azimuths, then a melting layer is judged to have occurred at that elevation angle, and its corresponding threshold is added to the total threshold.

Images