A cloud height identification method integrating cloud radar and laser wind radar

Abstract

This invention relates to a cloud height identification method integrating cloud radar and laser wind radar, belonging to the field of radio. The identification method includes: S1, obtaining the cloud base height and cloud top height by inverting the reflectivity product from the cloud radar in top-view mode; S2, obtaining the cloud base height and cloud top height by inverting the signal-to-noise ratio product from the laser wind radar in DBS mode; S3, performing time matching on the output products of the cloud radar and the laser wind radar, and then fusing the results obtained from the cloud radar and the laser wind radar to obtain the final cloud height identification result. This invention complements the characteristics of the two devices. The laser wind radar can compensate for the detection limitations of millimeter-wave radar in thin clouds and distinguish between cloud layers and aerosol layers, while the cloud radar can supplement the inability of laser wind radar to penetrate thick cloud layers, achieving accurate detection of cloud height across the entire cloud area.

Description

Technical Field

[0001] The present invention relates to the field of radio, and particularly to a method for identifying cloud height by fusing cloud radar and lidar wind profiler. Background Art

[0002] Clouds are the core elements of meteorological observations, and their characteristics such as height directly affect weather forecasting and climate research; accurate cloud height identification is crucial for improving the forecasting accuracy. The millimeter-wave cloud radar emits millimeter-wave electromagnetic waves, and inversely calculates the cloud structure by means of the scattering characteristics of hydrometeors in the cloud. It has a significant response to larger particles and strong penetration ability. However, due to the weak scattering of millimeter waves by tiny particles, it is difficult to detect thin clouds and is prone to missing or misjudging their heights. The lidar wind profiler emits near-infrared laser, utilizes the scattering effect of cloud particles, and inversely calculates the cloud top height based on the time difference of the laser round trip. It is sensitive to tiny cloud particles and has a vertical resolution of the meter level. However, due to its extremely short wavelength and the energy being consumed by the upper cloud particles, it cannot penetrate thick clouds or high-concentration hydrometeors and it is difficult to obtain the entire cloud structure. Therefore, it is difficult for a single millimeter-wave cloud radar or lidar wind profiler to achieve accurate detection of the entire cloud height range. Summary of the Invention

[0003] The purpose of the present invention is to overcome the shortcomings of the prior art, and provides a method for identifying cloud height by fusing cloud radar and lidar wind profiler, which solves the deficiencies existing in the prior art.

[0004] The purpose of the present invention is realized through the following technical solutions: A method for identifying cloud height by fusing cloud radar and lidar wind profiler, the identification method includes:

[0005] S1. Invert the cloud bottom height and cloud top height through the reflectivity product in the top mode of the cloud radar;

[0006] S2. Use the signal-to-noise ratio product in the DBS mode of the lidar wind profiler to invert the cloud bottom height and cloud top height; [[ID=​​​​​​​​​​​

[0011] S103. If there is only one candidate cloud layer, the height of the cloud base and cloud top of the candidate cloud layer shall be used as the cloud base and cloud top height. If there are multiple candidate cloud layers, the cloud base and cloud top height of the two thickest candidate cloud layers shall be selected as the cloud height product of the cloud radar.

[0012] S2 specifically includes the following:

[0013] S201. Analyze a single base data file of the laser wind measuring radar in DBS mode to obtain the signal-to-noise ratio profile, calculate the logarithmic form of the signal-to-noise ratio product, and record the minimum value on the profile as MinSNR.

[0014] S202. Calculate the gradient magnitude of each distance library on a single signal-to-noise ratio profile, and calculate the mean and standard deviation of the gradient profile. Then, calculate the dynamic threshold from the mean and standard deviation as the basis for distinguishing between clouds and aerosols. Five thresholds are obtained for each signal-to-noise ratio profile.

[0015] S203. Based on the thresholds calculated in S202, the single signal-to-noise ratio gradient profile is judged, and the candidate cloud bottom and middle, and aerosol layer bottom and middle are calculated.

[0016] S204. Based on the bottom and middle parts of the candidate cloud layer and aerosol layer obtained in S203, calculate the top of the candidate cloud layer and aerosol layer.

[0017] S205. If there are multiple candidate aerosol layers, the bottom and top heights of the candidate aerosol layers are determined, and adjacent multi-layer aerosols are merged. If the distance between the bottom of the higher aerosol layer and the top of the lower aerosol layer is less than or equal to two distances, the two aerosol layers are merged. The bottom position of the lower aerosol layer is taken as the bottom of the new aerosol layer, and the top position of the higher aerosol layer is taken as the top of the new aerosol layer.

[0018] S206. If candidate aerosol layers exist, remove aerosols below the threshold and classify aerosol layers above the threshold as clouds. Calculate the mean signal-to-noise ratio (SNR) aversnr and maximum signal-to-noise ratio (SNR) maxsnr from bottom to top for each candidate aerosol layer, and calculate the difference x1 = aversnr - MinSNR and the difference x2 = maxsnr - MinSNR. If x1 is less than the threshold n1 and x2 is less than the threshold m1, remove the aerosol layer from the candidate aerosol layers. If x1 is greater than the threshold n2 and x2 is greater than the threshold m2, and the height of the aerosol layer from bottom to top is greater than height1, classify the candidate aerosol layer as a candidate cloud.

[0019] S207. If a cloud layer converted from an aerosol layer exists in S206, repeat step S204 to find the top of the candidate cloud layer again.

[0020] S208. If both candidate cloud layers and candidate aerosol layers exist simultaneously, the bottom and top distance information of the candidate cloud layers and aerosol layers is used to determine the distance information, and the cloud that meets the conditions is fused with the aerosols within the range of the three distance databases in its vicinity.

[0021] S209. If there are multiple candidate cloud layers, the distances from the bottom and top of the candidate cloud layers to the database are used to determine the distances and merge adjacent multi-layer clouds.

[0022] S210. If candidate cloud layers exist, further judgment is made on them to obtain the Cloud High product.

[0023] The calculation expressions for the five thresholds in S202 include thresh i = averD + r i ×varpower, where averD is the mean of the signal-to-noise ratio profile gradient, varpower is the standard deviation of the signal-to-noise ratio profile gradient, and r i Used to dynamically adjust the threshold size; resulting in five thresholds: threshLow1, threshLow2, threshMid1, threshMid2, and threshTop.

[0024] threshLow1 and threshLow2 are used to identify the bottom of clouds or aerosol layers located below and above 3600m, respectively. threshMid1 and threshMid2 are used to identify the middle of clouds or aerosol layers located below and above 3600m, respectively. threshTop is used to identify the top of clouds or aerosol layers.

[0025] S203 specifically includes the following:

[0026] Choose a threshold from threshLow1 and threshLow2 as thresh1, and choose a threshold from threshMid1 and threshMid2 as thresh2, where thresh1 > thresh2;

[0027] A profile segment satisfying the following conditions is identified: within a continuous range of several distances [bin1, bin2], the signal-to-noise ratio gradient at bin1 is greater than the threshold thresh1, and the signal-to-noise ratio gradient at other altitudes is greater than the threshold thresh2. If the number of distances in the profile segment is greater than 2, then bin1 and bin2 in the profile segment are marked as candidate cloud bottom and middle, respectively. If the number of distances in the profile segment is equal to 2, then bin1 and bin2 in the profile segment are marked as candidate aerosol layer bottom and middle, respectively. bin1>0 and bin2<=binNum.

[0028] S204 specifically includes the following:

[0029] Using threshTop as the threshold thresh3 for discrimination, the traversal starts from the middle position of the candidate cloud layer and aerosol layer respectively;

[0030] A profile segment that meets the following conditions is identified: within a continuous range of distances [bin2, bin3], if the signal-to-noise ratio gradient is greater than the threshold thresh3, or if the signal-to-noise ratio is less than or equal to the signal-to-noise ratio at the bottom bin1 position of the candidate cloud or aerosol layer, then bin3 is taken as the top of the candidate cloud or aerosol layer. Different identified cloud and aerosol layers may have overlapping parts, and bin3 <= binNum.

[0031] The specific aerosols within the three distance ranges of the cloud that meet the fusion criteria in S208 include:

[0032] If the distance between the bottom of the higher aerosol layer and the top of the lower cloud layer is less than or equal to three distance units, or the distance between the bottom of the higher cloud layer and the top of the lower aerosol layer is less than or equal to three distance units, then the aerosol layer and the cloud layer are merged, with the lower cloud layer or the bottom of the aerosol layer being taken as the new cloud layer bottom, and the higher aerosol layer or the top of the cloud layer being taken as the new cloud layer top.

[0033] The fusion of adjacent multi-layer clouds in S209 specifically includes:

[0034] If there are two candidate cloud layers with overlapping parts from bottom to top, merge them. The bottom and top of the new cloud layer are respectively taken from the minimum distance library and the maximum distance library covered by the bottom to top of the two candidate cloud layers.

[0035] If there are two candidate cloud layers whose bottoms and tops do not overlap, and the distance between the bottom of the higher cloud layer and the top of the lower cloud layer is less than or equal to three distance units, then the two cloud layers are merged, with the bottom of the lower cloud layer as the new bottom and the top of the higher cloud layer as the new top.

[0036] S210 specifically includes the following:

[0037] Calculate the mean and maximum signal-to-noise ratio (SNR) of all candidate cloud layers from bottom to top, respectively, and calculate the difference x3 = aversnr - MinSNR and the difference x4 = maxsnr - MinSNR.

[0038] If x3 is less than threshold n3 and x4 is less than threshold m3, then the candidate cloud layer is removed from the candidate cloud layer list. If x3 is greater than threshold n4 and x4 is greater than threshold m4, and the height from the bottom to the top of the candidate cloud layer is greater than height2, then the candidate cloud layer is retained.

[0039] If there is only one candidate cloud layer, then the height of the cloud base and cloud top of that candidate cloud layer shall be taken as the cloud base and cloud top heights.

[0040] If there are multiple candidate cloud layers, the cloud base and cloud top heights of the two candidate cloud layers with the largest values ​​of y=w3×6x3+w4×x4 are selected as the product output of the laser wind measuring radar, where w3 and w4 are weights.

[0041] S3 specifically includes the following:

[0042] Cloud radar and laser wind radar were simultaneously observed at the same location. The basic data time resolution of the cloud radar was 5 seconds, and that of the laser wind radar was 1 minute. The cloud height product along each of the R radial lines of the cloud radar was calculated to obtain the average cloud height product over 1 minute. The cloud base products of the cloud radar and the laser wind radar were denoted as follows: , Genting products are respectively , ;

[0043] Let the height of the merged cloud base be The height of the cloud top is The cloud height products of the two radars within the same minute are calculated;

[0044] If neither the cloud radar nor the laser wind radar detects clouds, then record that there are no clouds at this moment. and ;

[0045] If the cloud radar detects clouds but the laser wind radar does not, then the cloud base and cloud top heights are respectively... and If the laser wind radar detects clouds but the cloud radar does not, then the cloud base and cloud top heights are respectively... and ;

[0046] If both cloud radar and laser wind radar detect clouds, then set the height difference threshold to [value]. When both cloud radar and laser wind radar detect one or two layers of clouds, the system searches for the cloud layer with the smallest difference in cloud height detected by the two radars. If the difference in cloud base or cloud top height is less than [missing information], then [missing information]. If the cloud base height is the average of the cloud base heights of the two radars, and the cloud top height is the average of the cloud top heights of the two radars, then the cloud base height is the minimum of the cloud base heights of the two radars, and the cloud top height is the maximum of the cloud base heights of the two radars. When the final output result is two layers of clouds, if there is a case where the cloud top height of the lower layer is greater than the cloud base height of the upper layer, then the cloud height product output of the two radars is further judged. If the algorithm of the single radar product runs correctly, then the lower cloud top is replaced with the minimum of the cloud top heights output by the two radars.

[0047] This invention has the following advantages: a cloud height identification method that integrates cloud radar and laser wind radar complements the characteristics of the two devices. The laser wind radar can compensate for the detection defects of millimeter-wave radar in thin clouds and distinguish between cloud layers and aerosol layers. The cloud radar can supplement the shortcomings of laser wind radar in penetrating thick cloud layers, thus achieving accurate detection of cloud height across the entire area. Attached Figure Description

[0048] Figure 1 This is a schematic diagram of the process of the present invention;

[0049] Figure 2 A schematic diagram of obtaining cloud height products from cloud radar reflectivity factor;

[0050] Figure 3 This is a schematic diagram of the signal-to-noise ratio profile of a laser wind-measuring radar.

[0051] Figure 4 This is a schematic diagram of the signal-to-noise ratio gradient profile;

[0052] Figure 5 This is a schematic diagram illustrating the signal-to-noise ratio profiles of the cloud layer and the aerosol layer from the bottom to the middle.

[0053] Figure 6 This is a schematic diagram illustrating the signal-to-noise ratio gradient profile between the bottom and middle of the cloud and aerosol layers.

[0054] Figure 7 This is a schematic diagram illustrating the signal-to-noise ratio profiles from the middle to the top of the cloud and aerosol layers.

[0055] Figure 8 This is a schematic diagram illustrating the signal-to-noise ratio gradient profile from the middle to the top of the cloud and aerosol layers.

[0056] Figure 9 A schematic diagram of the signal-to-noise ratio profile after fusing adjacent aerosol layers;

[0057] Figure 10 Schematic diagram of the signal-to-noise ratio profile after removing the aerosol layer or classifying the aerosol layer as a cloud layer;

[0058] Figure 11 Schematic diagram of the signal-to-noise ratio profile after re-identifying the cloud top;

[0059] Figure 12 Schematic diagram of obtaining cloud height products from the signal-to-noise ratio profile of a lidar wind profiler;

[0060] Figure 13 Schematic diagram of the comparison before and after the fusion of cloud height products from a cloud radar and a lidar wind profiler. Detailed implementation manners

[0061] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Usually, the components of the embodiments of the present application described and illustrated herein can be arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of the present application provided below with reference to the accompanying drawings is not intended to limit the protection scope of the claimed present application, but merely represents the selected embodiments of the present application. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative efforts fall within the protection scope of the present application. The present invention will be further described below with reference to the accompanying drawings.

[0062] As Figure 1 shown, the present invention specifically relates to a cloud height identification method for the fusion of a cloud radar and a lidar wind profiler. Among them, the cloud radar has a distance resolution of 30 m and a time resolution of 5 seconds, and the lidar wind profiler has a distance resolution of 60 m and a time resolution of 1 minute, and specifically includes the following content:

[0063] Step 1: Cloud height identification by the cloud radar;

[0064] Using the reflectivity product in the top mode of the cloud radar, the cloud base height and cloud top height are retrieved, and specifically include the following content:

[0065] (1) Analyze a single basic data file of the cloud radar to obtain the reflectivity (Z) profile data.

[0066] (2) Set the threshold threshold1, and let the distance from a certain point on the profile to the cloud radar be H. If 0 < H < 3.5 km, then the height range that satisfies threshold1 < Z < 51 is recorded as the cloud layer; otherwise, the height range where Z > -50 dBz is recorded as the cloud layer.

[0067] (3) To facilitate subsequent cloud height fusion calculations, only the height information of a maximum of two cloud layers will be output. If only one candidate cloud layer exists, the heights of its cloud base and cloud top will be used as the cloud base and cloud top heights. If multiple candidate cloud layers exist, the cloud base and cloud top heights of the two thickest candidate cloud layers will be selected as the cloud height product of the cloud radar. The cloud radar cloud height product effect is as follows: Figure 2 As shown in the figure, (a) is a schematic diagram of the cloud radar reflectivity factor, and (b) is a schematic diagram of the cloud radar cloud height product.

[0068] Step 2: Identification of cloud height using laser wind radar;

[0069] The cloud base height and cloud top height are obtained by using the signal-to-noise ratio product of the laser wind radar in DBS mode; the specific method is as follows:

[0070] (1) Analyze a single base data file of the lidar wind measurement radar in DBS mode to obtain the signal-to-noise ratio (SNR) profile. Calculate the logarithmic SNR product, recording the minimum value (MinSNR) and maximum range number (binNum) on the profile. Considering the impact of the lidar's geometric overlap factor on data accuracy near the ground, data below 500m will be discarded during calculation. The single radial SNR profile is shown below. Figure 3 As shown.

[0071] (2) Calculate the gradient magnitude of each distance on a single signal-to-noise ratio profile, and calculate the mean and standard deviation of the gradient profile. Then, calculate the dynamic threshold from the mean and standard deviation as the basis for distinguishing between clouds and aerosols. Each signal-to-noise ratio profile will yield five thresholds, the expressions of which are as follows:

[0072] thresh i = averD + r i ×varpower;

[0073] Where averD is the mean of the signal-to-noise ratio profile gradient, varpower is the standard deviation of the signal-to-noise ratio profile gradient, and r i ∈(-2,1), used to dynamically adjust the threshold value.

[0074] like Figure 4 As shown, dSNR represents the signal-to-noise ratio gradient profile. threshLow1 and threshLow2 are used to identify the bottom of clouds or aerosol layers below and above 3600m, respectively. threshMid1 and threshMid2 are used to identify the middle of clouds or aerosol layers below and above 3600m, respectively. threshTop is used to identify the top of clouds or aerosol layers.

[0075] (3) Based on the thresholds calculated in the previous step (2), a single signal-to-noise ratio gradient profile is discriminated, and candidate cloud bottom and middle, and aerosol layer bottom and middle are calculated. According to the current altitude, a suitable threshold is selected from threshLow1 and threshLow2 as thresh1, and a suitable threshold is selected from threshMid1 and threshMid2 as thresh2 (thresh1>thresh2). A profile that meets the conditions is obtained: within a continuous range of several distances [bin1,bin2] (bin1>0 and bin2<=binNum), the signal-to-noise ratio gradient at bin1 is greater than the threshold thresh1, and the signal-to-noise ratio gradient at other altitudes is greater than the threshold thresh2. If the distance to the cloud within a given profile segment is greater than 2, then bin1 and bin2 within that segment are marked as candidate cloud base and middle regions, respectively; if the distance to the cloud within a given profile segment is equal to 2, then bin1 and bin2 within that segment are marked as candidate aerosol layer base and middle regions, respectively. The results of the discrimination are as follows: Figure 5 and Figure 6 As shown, the bottom and middle parts of the cloud layer and aerosol layer are marked with solid and hollow triangles, respectively, and the area between the bottom and middle parts is marked with solid and hollow dots, respectively.

[0076] (4) If the bottom and middle parts of the candidate cloud and aerosol layers are obtained from the previous step (3), then the top of the candidate cloud and aerosol layers is calculated. Using threshTop as the discrimination threshold thresh3, the process starts from the middle position of the candidate cloud and aerosol layers. A profile that meets the following conditions is determined: within several consecutive distance ranges [bin2, bin3] (bin3 <= binNum), the signal-to-noise ratio gradient is greater than the threshold thresh3, or the signal-to-noise ratio is less than or equal to the signal-to-noise ratio at the bottom bin1 position of the candidate cloud or aerosol layer. Then bin3 is considered the top of the candidate cloud or aerosol layer. Different identified cloud and aerosol layers may have overlapping parts. The discrimination results are as follows: Figure 7 and Figure 8 As shown, the bottom and top of the cloud layer and aerosol layer are marked with solid and hollow triangles, respectively, and the area between the bottom and top is marked with solid and hollow dots, respectively.

[0077] (5) If multiple candidate aerosol layers exist, the bottom and top heights of the candidate aerosol layers are determined, and adjacent multilayer aerosols are merged. If the distance between the bottom of the higher aerosol layer and the top of the lower aerosol layer is less than or equal to two distances, the two aerosol layers are merged, with the bottom of the lower aerosol layer becoming the bottom of the new aerosol layer, and the top of the higher aerosol layer becoming the top of the new aerosol layer. After aerosol layer fusion... Figure 9 As shown, the bottom and top of the fused aerosol layer are marked with hollow triangles, and the area between the bottom and top is marked with hollow dots.

[0078] (6) If candidate aerosol layers exist, remove aerosols below the threshold and classify aerosol layers above the threshold as clouds. Calculate the mean signal-to-noise ratio (SNR) *aversnr* and the maximum signal-to-noise ratio (SNR) *maxsnr* for each candidate aerosol layer from bottom to top, and calculate the difference *x1* = *aversnr* - *MinSNR* and the difference *x2* = *maxsnr* - *MinSNR*. If *x1* is less than the threshold *n1* and *x2* is less than the threshold *m1*, then the aerosol layer is removed from the candidate aerosol layers; if *x1* is greater than the threshold *n2*, *x2* is greater than the threshold *m2*, and the height of the aerosol layer from bottom to top is greater than *height1*, then the candidate aerosol layer is classified as a candidate cloud. The process after an aerosol layer is removed or included in a cloud layer is as follows: Figure 10 As shown. In this example, the candidate aerosol layer was classified as a cloud, and the bottom and top were marked with solid triangles, while the area between the bottom and top was marked with solid dots.

[0079] (7) Figure 11 As shown, if there is a cloud layer converted from an aerosol layer in the previous step (6), then repeat step (4) to find the top of the candidate cloud layer again.

[0080] (8) If both candidate cloud layers and candidate aerosol layers exist simultaneously, the bottom and top distance information of the candidate cloud layers and aerosol layers is used to determine the distances to the database. Clouds that meet the criteria are then merged with aerosols within a distance database of three adjacent distance databases. If the distance between the bottom of the higher aerosol layer (or cloud layer) and the top of the lower cloud layer (or aerosol layer) is less than or equal to three distance database lengths (i.e., the distance between the aerosol layer and the cloud layer is less than or equal to three distance database lengths), then the aerosol layer and cloud layer are merged, with the bottom of the lower layer becoming the new cloud bottom and the top of the higher layer becoming the new cloud top. In this example, there are no cloud layers or aerosol layers that need to be merged.

[0081] (9) If multiple candidate cloud layers exist, the bottom and top distance database information of the candidate cloud layers is used to determine the fusion of adjacent multi-layer clouds. If two candidate cloud layers have overlapping bottom-to-top portions, they are merged, and the bottom and top of the new cloud layer are taken from the minimum and maximum distance databases covered by the bottom-to-top portions of the two candidate cloud layers, respectively. If two candidate cloud layers do not have overlapping bottom-to-top portions, and the distance between the bottom of the higher cloud layer and the top of the lower cloud layer is less than or equal to three distance database lengths, the two cloud layers are merged, with the bottom position of the lower cloud layer as the new cloud layer bottom and the top position of the higher cloud layer as the new cloud layer top. In this example, there are no multi-layer clouds that need to be merged.

[0082] (10) If candidate cloud layers exist, further judgment is made to obtain the cloud height product. The mean and maximum signal-to-noise ratio (SNR) of all candidate cloud layers from bottom to top are calculated respectively, and the difference x3 = aversnr - MinSNR and the difference x4 = maxsnr - MinSNR are calculated. If x3 is less than the threshold n3 and x4 is less than the threshold m3, the candidate cloud layer is removed from the candidate cloud layer list; if x3 is greater than the threshold n4 and x4 is greater than the threshold m4, and the height from bottom to top of the candidate cloud layer is greater than height2, the candidate cloud layer is retained. To facilitate the subsequent fusion calculation of cloud layer height, the height information of at most two cloud layers is output. If there is only one candidate cloud layer, the height of the cloud bottom position and the cloud top position of the candidate cloud layer are used as the cloud bottom and cloud top heights; if there are multiple candidate cloud layers, the cloud bottom and cloud top heights of the two candidate cloud layers with the largest values ​​of y = w3 × x3 + w4 × x4 (w3 and w4 are weights) are selected as the product output of the laser wind radar. Laser wind measurement radar cloud height product effect as follows Figure 12 As shown.

[0083] Step 3: Integration of cloud radar and laser wind measurement radar cloud products;

[0084] Before fusing the cloud height products, time matching of the output products from the two radars is required. The cloud radar and the laser wind radar conduct simultaneous observations at the same location. The cloud radar's baseline data time resolution is 5 seconds, and the laser wind radar's baseline data time resolution is 1 minute. The cloud height product is calculated for every 12 radial lines from the cloud radar to obtain the 1-minute average cloud height product. Let the cloud base products from the cloud radar and the laser wind radar be denoted as follows: , Genting products are respectively , ( ).

[0085] Let the height of the merged cloud base be The height of the cloud top is Calculate the cloud height products of the two radars within the same minute:

[0086] (1) If neither the cloud radar nor the laser wind radar detects clouds, then record that there are no clouds at this moment. and .

[0087] (2) If the cloud radar (or laser wind radar) detects clouds, but the laser wind radar (or cloud radar) does not, then the cloud base and cloud top heights are respectively and (or and ).

[0088] (3) If both cloud radar and laser wind radar detect clouds, set a height difference threshold. When both cloud radar and laser wind radar detect one or two layers of clouds, the system searches for the cloud layer with the smallest difference in cloud height detected by the two radars. If the difference in cloud base or cloud top height is less than [missing information], then [missing information]. If the cloud base height is the average of the cloud base heights of the two radars, then the cloud top height is the average of the cloud top heights of the two radars; otherwise, the cloud base height is the minimum of the cloud base heights of the two radars, and the cloud top height is the maximum of the cloud base heights of the two radars.

[0089] When the final output result is two layers of clouds, if there is a case where the height of the lower cloud top is greater than the height of the upper cloud bottom, the cloud height output of the two radars needs to be further judged. If the algorithm of the single radar product runs correctly, the lower cloud top is replaced with the minimum value of the cloud top height output by the two radars.

[0090] like Figure 2 and Figure 12 The image shows a comparison of the reflectivity factor and cloud height product of the cloud radar before fusion, and the signal-to-noise ratio and cloud height product of the laser wind measurement radar. Figure 12 In the diagram, (a) shows the signal-to-noise ratio (SNR) of a laser wind-measuring radar, and (b) shows a schematic diagram of the cloud height product of a laser wind-measuring radar. Figure 13 As shown in the figure, (a) is a schematic diagram of the cloud height product of the cloud radar, (b) is a schematic diagram of the cloud height product of the laser wind radar, and (c) is a schematic diagram of the cloud height product after the fusion of the two radars.

[0091] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and improvements, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A cloud height identification method integrating cloud radar and laser wind radar, characterized in that: The identification method includes: S1. The cloud base height and cloud top height are obtained by inverting the reflectivity product of the cloud radar top mode; S2. Using the signal-to-noise ratio product in the DBS mode of the laser wind radar, the cloud base height and cloud top height are obtained by inversion. S3. Time matching is performed on the output products of cloud radar and laser wind radar, and then the results obtained from cloud radar and laser wind radar are fused to obtain the final cloud height recognition result. S2 specifically includes the following: S201. Analyze a single base data file of the laser wind measurement radar in DBS mode to obtain the signal-to-noise ratio profile, calculate the logarithmic form of the signal-to-noise ratio product, and record the minimum value of the profile as MinSNR and the maximum range library number as binNum. S202. Calculate the gradient magnitude of each distance library on a single signal-to-noise ratio profile, and calculate the mean and standard deviation of the gradient profile. Then, calculate the dynamic threshold from the mean and standard deviation as the basis for distinguishing between clouds and aerosols. Five thresholds are obtained for each signal-to-noise ratio profile. S203. Based on the thresholds calculated in S202, the single signal-to-noise ratio gradient profile is judged, and the candidate cloud bottom and middle, and aerosol layer bottom and middle are calculated. S204. Based on the bottom and middle parts of the candidate cloud layer and aerosol layer obtained in S203, calculate the top of the candidate cloud layer and aerosol layer. S205. If there are multiple candidate aerosol layers, the bottom and top heights of the candidate aerosol layers are determined, and adjacent multi-layer aerosols are merged. If the distance between the bottom of the higher aerosol layer and the top of the lower aerosol layer is less than or equal to two distances, the two aerosol layers are merged. The bottom position of the lower aerosol layer is taken as the bottom of the new aerosol layer, and the top position of the higher aerosol layer is taken as the top of the new aerosol layer. S206. If candidate aerosol layers exist, remove aerosols below the threshold and classify aerosol layers above the threshold as clouds. Calculate the mean signal-to-noise ratio (SNR) aversnr and maximum signal-to-noise ratio (SNR) maxsnr from bottom to top for each candidate aerosol layer, and calculate the difference x1 = aversnr - MinSNR and the difference x2 = maxsnr - MinSNR. If x1 is less than the threshold n1 and x2 is less than the threshold m1, remove the aerosol layer from the candidate aerosol layers. If x1 is greater than the threshold n2 and x2 is greater than the threshold m2, and the height of the aerosol layer from bottom to top is greater than height1, classify the candidate aerosol layer as a candidate cloud. S207. If a cloud layer converted from an aerosol layer exists in S206, repeat step S204 to find the top of the candidate cloud layer again. S208. If both candidate cloud layers and candidate aerosol layers exist simultaneously, the bottom and top distance information of the candidate cloud layers and aerosol layers is used to determine the distance information, and the cloud that meets the conditions is fused with the aerosols within the range of the three distance databases in its vicinity. S209. If there are multiple candidate cloud layers, the distances from the bottom and top of the candidate cloud layers to the database are used to determine the distances and merge adjacent multi-layer clouds. S210. If there is a candidate cloud layer, further judgment is performed on it to obtain the cloud height product.

2. The cloud height identification method fused with cloud radar and laser wind radar according to claim 1, characterized in that: The specific content of S1 is as follows: S101. Analyze a single basic data file of the cloud radar to obtain the reflectivity Z profile data; S102. Set the threshold threshold1. Let the distance from a certain point on the profile to the cloud radar be H. If there is 0 < H < 3.5 km, then the height range that satisfies threshold1 < Z < 51 is recorded as the cloud layer. Otherwise, the height range where Z > -50 dBz is recorded as the cloud layer; S103. If there is only one candidate cloud layer, the heights where the cloud bottom position and cloud top position of the candidate cloud layer are located are used as the cloud bottom and cloud top heights. If there are multiple candidate cloud layers, the cloud bottom and cloud top heights of the two thickest candidate cloud layers are selected as the cloud height product of the cloud radar.

3. The cloud height identification method fused with cloud radar and laser wind radar according to claim 1, characterized in that: The calculation expressions for the five thresholds in S202 include thresh i = averD + r i ×varpower, where averD is the mean of the signal-to-noise ratio profile gradient, varpower is the standard deviation of the signal-to-noise ratio profile gradient, and r i Used to dynamically adjust the threshold size; resulting in five thresholds: threshLow1, threshLow2, threshMid1, threshMid2, and threshTop. threshLow1 and threshLow2 are respectively used to distinguish the bottom of the cloud layer or aerosol layer below and above 3600 m, threshMid1 and threshMid2 are respectively used to distinguish the middle of the cloud layer or aerosol layer below and above 3600 m, and threshTop is used to distinguish the top of the cloud layer or aerosol layer.

4. The cloud height identification method fused with cloud radar and laser wind radar according to claim 3, characterized in that: The specific content of S203 is as follows: Select the threshold from threshLow1 and threshLow2 as thresh1, select the threshold from threshMid1 and threshMid2 as thresh2, and thresh > thresh2; Judge to obtain a section of profile that meets the conditions: within a continuous range of several distance bins [bin1, bin2], the signal-to-noise ratio gradient at bin1 is greater than the threshold thresh1, and the signal-to-noise ratio gradients at the remaining heights are all greater than the threshold thresh2. If the number of distance bins within this section of the profile is greater than 2, then bin1 and bin2 in this section of the profile are respectively marked as the candidate cloud bottom and middle. If the number of distance bins within this section of the profile is equal to 2, then bin1 and bin2 in this section of the profile are respectively marked as the candidate aerosol layer bottom and middle, where bin > 0 and bin2 <= binNum.

5. The cloud height identification method fused with cloud radar and laser wind radar according to claim 4, characterized in that: The specific content of S204 is as follows: Use threshTop as the discrimination threshold thresh3 and traverse from the middle positions of the candidate cloud layer and aerosol layer respectively; Judge to obtain a section of profile that meets the conditions: within a continuous range of several distance bins [bin2, bin,], the signal-to-noise ratio gradients are all greater than the threshold thresh, or the signal-to-noise ratio is less than or equal to the signal-to-noise ratio at the corresponding bottom bin1 position of the candidate cloud layer or aerosol layer. Then bin3 is used as the candidate cloud layer or aerosol layer top. There may be overlapping parts between the identified different cloud layers and aerosol layers, and bin3 <= binNum.

6. The cloud height identification method fused with cloud radar and laser wind radar according to claim 1, characterized in that: The fusion of the cloud that meets the conditions and the aerosol within the three adjacent distance bins in S208 specifically includes: If the distance between the bottom of the higher aerosol layer and the top of the lower cloud layer is less than or equal to three distance units, or the distance between the bottom of the higher cloud layer and the top of the lower aerosol layer is less than or equal to three distance units, then the aerosol layer and the cloud layer are merged, with the lower cloud layer or the bottom of the aerosol layer being taken as the new cloud layer bottom, and the higher aerosol layer or the top of the cloud layer being taken as the new cloud layer top.

7. The cloud height identification method fused with cloud radar and laser wind radar according to claim 1, characterized in that: The fusion of adjacent multi-layer clouds in S209 specifically includes: If there are two candidate cloud layers with overlapping parts from bottom to top, merge them. The bottom and top of the new cloud layer are respectively taken from the minimum distance library and the maximum distance library covered by the bottom to top of the two candidate cloud layers. If there are two candidate cloud layers whose bottoms and tops do not overlap, and the distance between the bottom of the higher cloud layer and the top of the lower cloud layer is less than or equal to three distance units, then the two cloud layers are merged, with the bottom of the lower cloud layer as the new bottom and the top of the higher cloud layer as the new top.

8. The cloud height identification method fused with cloud radar and laser wind radar according to claim 1, characterized in that: S210 specifically includes the following: Calculate the mean and maximum signal-to-noise ratio (SNR) of all candidate cloud layers from bottom to top, respectively, and calculate the difference x3 = aversnr - MinSNR and the difference x4 = maxsnr - MinSNR. If x3 is less than threshold n3 and x4 is less than threshold m3, then the candidate cloud layer is removed from the candidate cloud layer list. If x3 is greater than threshold n4 and x4 is greater than threshold m4, and the height from the bottom to the top of the candidate cloud layer is greater than height2, then the candidate cloud layer is retained. If there is only one candidate cloud layer, then the height of the cloud base and cloud top of that candidate cloud layer shall be taken as the cloud base and cloud top heights. If there are multiple candidate cloud layers, the cloud base and cloud top heights of the two candidate cloud layers with the largest values ​​of y=w3×6x3+w4×x4 are selected as the product output of the laser wind measuring radar, where w3 and w4 are weights.

9. The cloud height identification method fused with cloud radar and laser wind radar according to claim 1, characterized in that: S3 specifically includes the following: Cloud radar and laser wind radar were simultaneously observed at the same location. The basic data time resolution of the cloud radar was 5 seconds, and that of the laser wind radar was 1 minute. The cloud height product along each of the R radial lines of the cloud radar was calculated to obtain the average cloud height product over 1 minute. The cloud base products of the cloud radar and the laser wind radar were denoted as follows: , Genting products are respectively , ; Let the height of the merged cloud base be The height of the cloud top is The cloud height products of the two radars within the same minute are calculated; If neither the cloud radar nor the laser wind radar detects clouds, then record that there are no clouds at this moment. and ; If the cloud radar detects clouds but the laser wind radar does not, then the cloud base and cloud top heights are respectively... and If the laser wind radar detects clouds but the cloud radar does not, then the cloud base and cloud top heights are respectively... and ; If both cloud radar and laser wind radar detect clouds, then set the height difference threshold to [value]. When both cloud radar and laser wind radar detect one or two layers of clouds, the system searches for the cloud layer with the smallest difference in cloud height detected by the two radars. If the difference in cloud base or cloud top height is less than [missing information], then [missing information]. If the cloud base height is the average of the cloud base heights of the two radars, and the cloud top height is the average of the cloud top heights of the two radars, then the cloud base height is the minimum of the cloud base heights of the two radars, and the cloud top height is the maximum of the cloud base heights of the two radars. When the final output result is two layers of clouds, if there is a case where the cloud top height of the lower layer is greater than the cloud base height of the upper layer, then the cloud height product output of the two radars is further judged. If the algorithm of the single radar product runs correctly, then the lower cloud top is replaced with the minimum of the cloud top heights output by the two radars.

Images