Abstract:
keep_len="250">The typical circulation configuration, fine structure of mesoscale system and thermodynamic development mechanism of affecting system associated with the extreme heavy rainfall of Henan from 19 July to 21 July in 2021 are analyzed with data which contain minutely automatic weather station observations, FY-4A satellite high-resolution measurements, Doppler radar products and the fifth-generation European Center for Medium-Range Weather Forecasts (ECMWF) atmospheric reanalysis (ERA5). Results show that the extreme heavy rainfall occurs under a weaken saddle field which is between the subtropical high and the continental high. The dominant systems are the weak low-pressure system at 500 hPa and easterly shear line at low level. A long-term maintained mesoscale convective cloud plays great roles in the extreme heavy rainfall of Henan, which features a nearly circular structure with a horizontal scale of about 300 km. The long-term maintenance of this cloud is related to the merging of its inside multiple β-mesoscale convective systems. Meanwhile, the confluence of three easterly flows in the east of warm zone in the southeast of the periphery keep triggering new convective cells which are constantly integrated into the mesoscale convective cloud. These two processes lead to the train effect, which is crucial to the extreme heavy rainfall in Henan. The rainfall intensity at Zhengzhou Station is 201.9 mm·h^-1, and breaks the hourly rainfall record in inland regions. It is mainly caused by quasi-stationary β-mesoscale convective systems, which show bow echo in Doppler weather radar. The vertical structure of the convective system, which has the strong echo centroid below 5 km in Doppler weather radar, reflects the extremely efficient precipitation. During the extreme rainfall from 1600 BT to 1700 BT on 20 Jul 2021, the minutely continuous precipitation is stable at 3-4.7 mm, and the 5-minute rainfall maximum could reach 21 mm. The strong convection is triggered by the dynamic convergence of the boundary layer wind, leading to the pseudo equivalent temperature (θ_se) front above the heavy rainfall area maintaining in a neutral stratification of nearly barotropic structure for a long time. Meanwhile, the low-level convergence collocates with the high-level divergence, which benefits an intense ascend through the tropopause. The high-level divergent flow forms a sinking branch of the secondary circulation near the Northwest Pacific subtropical high. The significant positive vorticity advection near the cyclonic circulation at 500 hPa, the continuous warm advection transport by easterly flow at 925 hPa, the frontogenesis of low-level deformation field and the abnormally strong jet water vapor transport from the coast of East China are the thermodynamic mechanisms of the development and maintenance of the extreme heavy rainfall in Henan. The typical circulation configuration, fine structure of mesoscale system and thermodynamic development mechanism of affecting system associated with the extreme heavy rainfall of Henan from 19 July to 21 July in 2021 are analyzed with data which contain minutely automatic weather station observations, FY-4A satellite high-resolution measurements, Doppler radar products and the fifth-generation European Center for Medium-Range Weather Forecasts (ECMWF) atmospheric reanalysis (ERA5). Results show that the extreme heavy rainfall occurs under a weaken saddle field which is between the subtropical high and the continental high. The dominant systems are the weak low-pressure system at 500 hPa and easterly shear line at low level. A long-term maintained mesoscale convective cloud plays great roles in the extreme heavy rainfall of Henan, which features a nearly circular structure with a horizontal scale of about 300 km. The long-term maintenance of this cloud is related to the merging of its inside multiple β-mesoscale convective systems. Meanwhile, the confluence of three easterly flows in the east of warm zone in the southeast of the periphery keep triggering new convective cells which are constantly integrated into the mesoscale convective cloud. These two processes lead to the train effect, which is crucial to the extreme heavy rainfall in Henan. The rainfall intensity at Zhengzhou Station is 201.9 mm·h^-1, and breaks the hourly rainfall record in inland regions. It is mainly caused by quasi-stationary β-mesoscale convective systems, which show bow echo in Doppler weather radar. The vertical structure of the convective system, which has the strong echo centroid below 5 km in Doppler weather radar, reflects the extremely efficient precipitation. During the extreme rainfall from 1600 BT to 1700 BT on 20 Jul 2021, the minutely continuous precipitation is stable at 3-4.7 mm, and the 5-minute rainfall maximum could reach 21 mm. The strong convection is triggered by the dynamic convergence of the boundary layer wind, leading to the pseudo equivalent temperature (θ_se) front above the heavy rainfall area maintaining in a neutral stratification of nearly barotropic structure for a long time. Meanwhile, the low-level convergence collocates with the high-level divergence, which benefits an intense ascend through the tropopause. The high-level divergent flow forms a sinking branch of the secondary circulation near the Northwest Pacific subtropical high. The significant positive vorticity advection near the cyclonic circulation at 500 hPa, the continuous warm advection transport by easterly flow at 925 hPa, the frontogenesis of low-level deformation field and the abnormally strong jet water vapor transport from the coast of East China are the thermodynamic mechanisms of the development and maintenance of the extreme heavy rainfall in Henan.

Abstract:
keep_len="250">Phased array radar is faster in scanning speed than mechanical scanning radar, but due to the influence of antenna structure and attenuation, phased array radar will produce larger system error and random error. The mainstream fuzzy logic hydrometeor classification method has little difference in the weight coefficients of each parameter, and the calculated composite values are often close, so the hydrometeor classification results are easily affected by data errors. The detection data of the X-band dual polarization phased array radar at Qiuyutan, Shenzhen, from March to September in 2020 are compared with the S-band dual polarization radar at the same location. The points close to the elevation, azimuth and radial distance of the two radars are obtained to establish matching datasets to calculate the errors of X-band dual-polarization phased array radar. Quantitative analysis of the causes for the introduction of errors through certain restriction conditions reveals that the calibration error and random error of the reflectance factor Z_H and the differential reflectance Z_DR are relatively large. The error range of Z_H is -0.5-4.5 dB, and the error of Z_DR is -0.7-0.2 dB. After the preliminary correction of calibration error and random error, it is found that there are still some errors in data, which make the hydrometeor classification result of fuzzy logic method unreliable, so the decision tree hydrometeor classification method with the basic structure of binary tree is established according to the characteristic range of radar parameters of different hydrometeors and the height of the melting layer. In order to verify the practical application effects of the above methods, the error sensitivity of hydrometeor classification results and the rationality of hydrometeor spatial distribution are evaluated respectively. Typical examples are selected to further evaluate the rationality of the decision tree hydrometeor classification method by comparing the parameters and hydrometeor classification results of X-band dual polarization phased array radar and S-band dual polarization radar. The evaluation results show that the stability of the decision tree hydrometeor classification method is higher than that of the fuzzy logic method, and the hydrometeor distribution in the convective cloud is more reasonable, which can give full play to the advantage of X-band dual polarization phased array radar in studying the phase evolution of particles in the cloud. Phased array radar is faster in scanning speed than mechanical scanning radar, but due to the influence of antenna structure and attenuation, phased array radar will produce larger system error and random error. The mainstream fuzzy logic hydrometeor classification method has little difference in the weight coefficients of each parameter, and the calculated composite values are often close, so the hydrometeor classification results are easily affected by data errors. The detection data of the X-band dual polarization phased array radar at Qiuyutan, Shenzhen, from March to September in 2020 are compared with the S-band dual polarization radar at the same location. The points close to the elevation, azimuth and radial distance of the two radars are obtained to establish matching datasets to calculate the errors of X-band dual-polarization phased array radar. Quantitative analysis of the causes for the introduction of errors through certain restriction conditions reveals that the calibration error and random error of the reflectance factor Z_H and the differential reflectance Z_DR are relatively large. The error range of Z_H is -0.5-4.5 dB, and the error of Z_DR is -0.7-0.2 dB. After the preliminary correction of calibration error and random error, it is found that there are still some errors in data, which make the hydrometeor classification result of fuzzy logic method unreliable, so the decision tree hydrometeor classification method with the basic structure of binary tree is established according to the characteristic range of radar parameters of different hydrometeors and the height of the melting layer. In order to verify the practical application effects of the above methods, the error sensitivity of hydrometeor classification results and the rationality of hydrometeor spatial distribution are evaluated respectively. Typical examples are selected to further evaluate the rationality of the decision tree hydrometeor classification method by comparing the parameters and hydrometeor classification results of X-band dual polarization phased array radar and S-band dual polarization radar. The evaluation results show that the stability of the decision tree hydrometeor classification method is higher than that of the fuzzy logic method, and the hydrometeor distribution in the convective cloud is more reasonable, which can give full play to the advantage of X-band dual polarization phased array radar in studying the phase evolution of particles in the cloud.

Abstract:
keep_len="250">Local torrential rain and short-term heavy rainfall of small spatial-temporal scale are caused by northward-moving Typhoon Lekima (1909) and Typhoon Bavi (2008) in Qingdao area,with the maximum hourly rainfall of 60.3 mm·h^-1 and 130.1 mm·h^-1,respectively, while the prediction performance of numerical weather prediction model is very poor. Using NCEP FNL analysis data, raindrop spectrum and polarimetric radar data, the microphysics characteristics of the heavy rainfall are analyzed. The rainfall mainly occurs in a narrow belt region extending northwestward from the coastal mountainous area. The warm and humid air is transported by the southeast wind strengthens the instability. Convective cells are constantly triggered by topography or boundary layer front, and then move northwestward and form linear multicell storms under strong wind condition, or merges into local strong storms when the wind is weak. Both can cause local heavy rainfall. The mass weighted average diameter (D_m) and logarithmic normalized intercept (lgN_w) are 1.89 mm and 3.86,respectively, which are between tropical marine-time and continental convective precipitation, indicating a larger mean diameter and lower number concentration compared to the typhoon rainfall in East China and South China. The μ-Λ slope is also significantly different, indicating the dominant microphysical processes are different. With the increase of rainfall intensity, the proportion of small particles below 1 mm decreases significantly, and the proportion of medium-large particles increases, indicating significant collision-coalescence process. Particles with 1-4 mm diameters contribute more than 90% to short-term heavy rainfall. When hourly rainfall is more than 50 mm·h^-1, the proportion of small particles increases and particles with 2-3 mm diameter changes little, indicating that breakup and collision-coalescence process reaches equilibrium. Aggregate process and dry snow is dominant above -20℃ level and grapuel produced by riming process is dominant between -10℃ and 0℃ level. With the decrease of height, the values of Z_H, Z_DR and K_DP increase, and raindrops change from light rain to heavy rain particles. At the same time, the liquid water content is significantly greater than ice water content, indicating that the collision-coalescence and accretion process play a critical role in the formation of heavy rainfall. Riming process also plays an important role in extreme heavy rainfall, during which its height can reach near -20℃ layer. The positive feedback of latent heat release leads to the strengthening of convective activity, resulting in more graupel particles and greater ice water content. The melting of graupel directly increases the rainfall. On the other hand, it produces big droplets, which enhance the warm-rain processes and leads to the increase of rainfall intensity. Local torrential rain and short-term heavy rainfall of small spatial-temporal scale are caused by northward-moving Typhoon Lekima (1909) and Typhoon Bavi (2008) in Qingdao area,with the maximum hourly rainfall of 60.3 mm·h^-1 and 130.1 mm·h^-1,respectively, while the prediction performance of numerical weather prediction model is very poor. Using NCEP FNL analysis data, raindrop spectrum and polarimetric radar data, the microphysics characteristics of the heavy rainfall are analyzed. The rainfall mainly occurs in a narrow belt region extending northwestward from the coastal mountainous area. The warm and humid air is transported by the southeast wind strengthens the instability. Convective cells are constantly triggered by topography or boundary layer front, and then move northwestward and form linear multicell storms under strong wind condition, or merges into local strong storms when the wind is weak. Both can cause local heavy rainfall. The mass weighted average diameter (D_m) and logarithmic normalized intercept (lgN_w) are 1.89 mm and 3.86,respectively, which are between tropical marine-time and continental convective precipitation, indicating a larger mean diameter and lower number concentration compared to the typhoon rainfall in East China and South China. The μ-Λ slope is also significantly different, indicating the dominant microphysical processes are different. With the increase of rainfall intensity, the proportion of small particles below 1 mm decreases significantly, and the proportion of medium-large particles increases, indicating significant collision-coalescence process. Particles with 1-4 mm diameters contribute more than 90% to short-term heavy rainfall. When hourly rainfall is more than 50 mm·h^-1, the proportion of small particles increases and particles with 2-3 mm diameter changes little, indicating that breakup and collision-coalescence process reaches equilibrium. Aggregate process and dry snow is dominant above -20℃ level and grapuel produced by riming process is dominant between -10℃ and 0℃ level. With the decrease of height, the values of Z_H, Z_DR and K_DP increase, and raindrops change from light rain to heavy rain particles. At the same time, the liquid water content is significantly greater than ice water content, indicating that the collision-coalescence and accretion process play a critical role in the formation of heavy rainfall. Riming process also plays an important role in extreme heavy rainfall, during which its height can reach near -20℃ layer. The positive feedback of latent heat release leads to the strengthening of convective activity, resulting in more graupel particles and greater ice water content. The melting of graupel directly increases the rainfall. On the other hand, it produces big droplets, which enhance the warm-rain processes and leads to the increase of rainfall intensity.

Abstract:
keep_len="250">Previous studies show that two typhoons with similar landing area and similar moving tracks may have significant differences in precipitation intensity, which are caused by different structure and characteristics of the cloud systems. Based on FY-2H, Aqua, CALIPSO(Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation) and GPM(Global Precipitation Measurement) satellite data, the horizontal and vertical structural characteristics of the cloud system, 3-dimensional structure and characteristics of Typhoon Lekima(1909) and Typhoon Yagi(1814), which landed along the Wenling coast of Zhejiang Province, are discussed. The visibility of typhoon eye area and the helicity of typhoon cloud system in TBB images are important indicators of typhoon development. The precipitation near the typhoon center is the largest, and the spatial distribution of typhoon precipitation is asymmetrical. For the typhoons with similar paths, strong typhoon induces circular strong precipitation center, while weak typhoon is along with belt-type strong precipitation center. In the mature stage of typhoon development, the maximum cloud top height is near the typhoon eye. When the cloud top height on the north side of the spiral rain belt area is lower than that on the south side, the typhoon develops strongly. When the cloud top height on the south side of the spiral rain belt area is lower than that the the north side, the typhoon is relatively weak. Before typhoon landing, the proportion of single layer cloud is higher when the typhoon is stronger, and the atmosphere is optically thicker. Typhoon clouds are mainly deep convective clouds and cirrus clouds consisting of non-directional ice. The height of cloud base and thickness in spiral rainband are related to the development of typhoon. Before typhoon landing, the area of high and low brightness temperature under the same channel, the precipitation type of typhoon, the length and number of convective columns in the 3-dimensional precipitation structure, and the precipitation rate in vertical direction can all indicate the development of typhoon. Regardless the typhoon strength, the total amount of ice water particles is roughly the same, and the difference in intensity is reflected in the areas of the high and low brightness temperature under the same channel. The spiral rain belt of a strong typhoon is dominated by stratiform precipitation, while a weak typhoon is dominated by convective precipitation. The number and length of convective columns of a strong typhoon are far greater than those of a weak typhoon. Previous studies show that two typhoons with similar landing area and similar moving tracks may have significant differences in precipitation intensity, which are caused by different structure and characteristics of the cloud systems. Based on FY-2H, Aqua, CALIPSO(Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation) and GPM(Global Precipitation Measurement) satellite data, the horizontal and vertical structural characteristics of the cloud system, 3-dimensional structure and characteristics of Typhoon Lekima(1909) and Typhoon Yagi(1814), which landed along the Wenling coast of Zhejiang Province, are discussed. The visibility of typhoon eye area and the helicity of typhoon cloud system in TBB images are important indicators of typhoon development. The precipitation near the typhoon center is the largest, and the spatial distribution of typhoon precipitation is asymmetrical. For the typhoons with similar paths, strong typhoon induces circular strong precipitation center, while weak typhoon is along with belt-type strong precipitation center. In the mature stage of typhoon development, the maximum cloud top height is near the typhoon eye. When the cloud top height on the north side of the spiral rain belt area is lower than that on the south side, the typhoon develops strongly. When the cloud top height on the south side of the spiral rain belt area is lower than that the the north side, the typhoon is relatively weak. Before typhoon landing, the proportion of single layer cloud is higher when the typhoon is stronger, and the atmosphere is optically thicker. Typhoon clouds are mainly deep convective clouds and cirrus clouds consisting of non-directional ice. The height of cloud base and thickness in spiral rainband are related to the development of typhoon. Before typhoon landing, the area of high and low brightness temperature under the same channel, the precipitation type of typhoon, the length and number of convective columns in the 3-dimensional precipitation structure, and the precipitation rate in vertical direction can all indicate the development of typhoon. Regardless the typhoon strength, the total amount of ice water particles is roughly the same, and the difference in intensity is reflected in the areas of the high and low brightness temperature under the same channel. The spiral rain belt of a strong typhoon is dominated by stratiform precipitation, while a weak typhoon is dominated by convective precipitation. The number and length of convective columns of a strong typhoon are far greater than those of a weak typhoon.

Abstract:
keep_len="250">In order to reconstruct the historical typhoon wind field in the Northwest Pacific Ocean and calculate the maximum wind speed in 50 years of the Northwest Pacific Ocean, Yan Meng wind field model is used to simulate the wind field. There are 3 important parameters for wind field simulation in Yan Meng wind field model: The radius of maximum wind, pressure distribution constant B, and roughness z₀. Therefore, it is necessary to test and reasonably optimize the value of the three parameters by measured data of the buoy stations during the typhoon in the Northwest Pacific Ocean.First, based on the JTWC (Joint Typhoon Warning Center) dataset, the relationship between the radius of maximum wind and its impact factors is discussed and four combinations scheme of calculating the radius of maximum wind are proposed, and then the best combination scheme is selected through the measured data. Second, the values of B and z₀ are estimated with the observed wind speed of buoy stations during different typhoons. Finally, the simulation effect of the typhoon wind field at sea is evaluated with 19 typhoon processes, and the applicability of the model and estimation scheme of three parameters are verified.The results show that it is more reasonable to find the radius of maximum wind by combination scheme of V_max (the maximum wind speed of typhoon center) and the latitude of typhoon. In the parameter value test, the wind speed simulation effect of sea surface (buoy stations) is better given z₀ being 0.005 m and B being 1.0, according to the absolute deviation between the simulated and the measured maximum wind speed at 10 buoy stations during 6 typhoons. Except for the parameter test, 19 other typhoon processes landing in northern Fujian and Zhejiang, heading north to the East China Sea, moving west to the South China Sea, and crossing Taiwan Island into the Taiwan Strait are selected to test the simulation effect, which illustrates that when the V_max published by Central Meteorological Observatory is below 40 m·s^-1, the simulated V_max is close to the published V_max if B is equal to 1.0 and z₀ is equal to 0.005 m, and the simulated wind speed in the non-maximum wind speed region is well fitted with the observed wind speed of the buoy stations. In addition, When V_max published is greater than or equal to 40 m·s^-1, the simulated V_max is close to the published V_max if B is equal to 1.4 and z₀ is equal to 0.005 m, and the simulated wind speed of non-maximum wind speed region is more reasonable when B is equal to 1.0 and z₀ is equal to 0.005 m. In order to reconstruct the historical typhoon wind field in the Northwest Pacific Ocean and calculate the maximum wind speed in 50 years of the Northwest Pacific Ocean, Yan Meng wind field model is used to simulate the wind field. There are 3 important parameters for wind field simulation in Yan Meng wind field model: The radius of maximum wind, pressure distribution constant B, and roughness z₀. Therefore, it is necessary to test and reasonably optimize the value of the three parameters by measured data of the buoy stations during the typhoon in the Northwest Pacific Ocean.First, based on the JTWC (Joint Typhoon Warning Center) dataset, the relationship between the radius of maximum wind and its impact factors is discussed and four combinations scheme of calculating the radius of maximum wind are proposed, and then the best combination scheme is selected through the measured data. Second, the values of B and z₀ are estimated with the observed wind speed of buoy stations during different typhoons. Finally, the simulation effect of the typhoon wind field at sea is evaluated with 19 typhoon processes, and the applicability of the model and estimation scheme of three parameters are verified.The results show that it is more reasonable to find the radius of maximum wind by combination scheme of V_max (the maximum wind speed of typhoon center) and the latitude of typhoon. In the parameter value test, the wind speed simulation effect of sea surface (buoy stations) is better given z₀ being 0.005 m and B being 1.0, according to the absolute deviation between the simulated and the measured maximum wind speed at 10 buoy stations during 6 typhoons. Except for the parameter test, 19 other typhoon processes landing in northern Fujian and Zhejiang, heading north to the East China Sea, moving west to the South China Sea, and crossing Taiwan Island into the Taiwan Strait are selected to test the simulation effect, which illustrates that when the V_max published by Central Meteorological Observatory is below 40 m·s^-1, the simulated V_max is close to the published V_max if B is equal to 1.0 and z₀ is equal to 0.005 m, and the simulated wind speed in the non-maximum wind speed region is well fitted with the observed wind speed of the buoy stations. In addition, When V_max published is greater than or equal to 40 m·s^-1, the simulated V_max is close to the published V_max if B is equal to 1.4 and z₀ is equal to 0.005 m, and the simulated wind speed of non-maximum wind speed region is more reasonable when B is equal to 1.0 and z₀ is equal to 0.005 m.

Abstract:
keep_len="250">Tropical cyclone is one of the major weather disasters affecting coastal areas, which can produce high winds and heavy rains, posing serious threats to the safety of people’s lives and property in coastal areas. China is in the Northwest Pacific, which is affected more frequently by tropical cyclones than any other area in the world. Therefore, it is of great significance to strengthen the research on tropical cyclone in the Northwest Pacific. In recent years, observations and studies have proved that lightning activity often occurs in tropical cyclone, which is closely related to the convective evolution and intensity variation of tropical cyclone. Based on data of lightning imaging sensor (LIS) carried on the TRMM (Tropical Rainfall Measuring Mission) satellite during 1998-2014, the characteristics of lightning properties (including duration, extended distance, channel area and optical radiant energy) of tropical cyclone in the Northwest Pacific are studied by establishing the lightning dataset of tropical cyclone in the region. The results show that all attributes of tropical cyclone lightning present lognormal distribution characteristics, and the distribution of peak values of these attributes is consistent with that of the Northwest Pacific thunderstorm system, but different from that of East Asia land thunderstorm system. The maximum of tropical cyclone lightning tends to occur over the ocean at tropical depression intensity levels. The maximum proportion of lightning in the outer rain belt is the lowest, while the maximum proportion of the duration and optical radiant energy of lightning in the inner core is the highest. The tropical cyclone lightning duration of different intensity levels has no significant difference, but the mean value of lightning spatial scale and optical radiation energy of tropical storm are lower than those of tropical depression and typhoon. For different areas of tropical cyclone, the maximum value of duration and optical radiant energy of core lightning decrease with the increase of distance between lightning and tropical cyclone center. In terms of maritime-continental contrasts, tropical cyclone lightning occurs over the ocean with larger spatial scale and stronger optical radiant energy than that over the land, while lightning duration is roughly the same. After tropical cyclone landing, the spatial scale of lightning decreases and the optical radiant energy of lightning weakens. Compared with non-tropical cyclone lightning, tropical cyclone lightning has shorter extension distance, narrower channel area and weaker optical radiant energy, but the average duration of lightning is longer. Tropical cyclone is one of the major weather disasters affecting coastal areas, which can produce high winds and heavy rains, posing serious threats to the safety of people’s lives and property in coastal areas. China is in the Northwest Pacific, which is affected more frequently by tropical cyclones than any other area in the world. Therefore, it is of great significance to strengthen the research on tropical cyclone in the Northwest Pacific. In recent years, observations and studies have proved that lightning activity often occurs in tropical cyclone, which is closely related to the convective evolution and intensity variation of tropical cyclone. Based on data of lightning imaging sensor (LIS) carried on the TRMM (Tropical Rainfall Measuring Mission) satellite during 1998-2014, the characteristics of lightning properties (including duration, extended distance, channel area and optical radiant energy) of tropical cyclone in the Northwest Pacific are studied by establishing the lightning dataset of tropical cyclone in the region. The results show that all attributes of tropical cyclone lightning present lognormal distribution characteristics, and the distribution of peak values of these attributes is consistent with that of the Northwest Pacific thunderstorm system, but different from that of East Asia land thunderstorm system. The maximum of tropical cyclone lightning tends to occur over the ocean at tropical depression intensity levels. The maximum proportion of lightning in the outer rain belt is the lowest, while the maximum proportion of the duration and optical radiant energy of lightning in the inner core is the highest. The tropical cyclone lightning duration of different intensity levels has no significant difference, but the mean value of lightning spatial scale and optical radiation energy of tropical storm are lower than those of tropical depression and typhoon. For different areas of tropical cyclone, the maximum value of duration and optical radiant energy of core lightning decrease with the increase of distance between lightning and tropical cyclone center. In terms of maritime-continental contrasts, tropical cyclone lightning occurs over the ocean with larger spatial scale and stronger optical radiant energy than that over the land, while lightning duration is roughly the same. After tropical cyclone landing, the spatial scale of lightning decreases and the optical radiant energy of lightning weakens. Compared with non-tropical cyclone lightning, tropical cyclone lightning has shorter extension distance, narrower channel area and weaker optical radiant energy, but the average duration of lightning is longer.

Abstract:
keep_len="250">The multiple upward leader’s attachment process on buildings is an important topic in lightning physics research, but the research on its physical mechanism is still insufficient. An improved 3D high-resolution multiple upward leader’s stochastic method is used to simulate the development and attachment process of downward negative cloud-to-ground lightning in the near-ground area. The model allows upward leaders to be initiated on both tall and low buildings. The attachment process is analyzed when the multiple upward leaders are initiated and connected from both high and low buildings. The results show that low buildings have very small probability of initiating upward leaders directly and being struck, while high buildings have a clear influence on the initiation of upward leader of low buildings. Once low buildings initiate upward leaders, they are more likely to be stricken. The height difference between buildings is the main factor affecting the lightning attachment process. When the height difference between buildings is not large, the shielding effect of high buildings on the low buildings is not obvious, and the relative position of the downward leader channel will affect whether the low buildings can first initiate the upward leader. With the increase of the height difference between buildings, it is difficult for low buildings to preferentially initiate the upward leader. Only when the downward leader channel is obviously closer to low buildings, low buildings can initiate the upward leader and have a certain probability to connect with the downward leader to form a return stroke. When the height difference between buildings is large enough to a certain extent, the space form of the downward leader has little influence on the lightning attachment process. The low buildings will not be struck without initiating an upward leader. After the upward leader is initiated, it will have a certain inhibitory effect on the electric field intensity of the surrounding top angle. This inhibitory effect is related to the number of upward leaders and the horizontal distance between the top angle of the initiating upward leader and other top angles. The inhibitory effect is positively correlated with the amount of upward leaders while negatively correlated with the horizontal distance. The multiple upward leader’s attachment process on buildings is an important topic in lightning physics research, but the research on its physical mechanism is still insufficient. An improved 3D high-resolution multiple upward leader’s stochastic method is used to simulate the development and attachment process of downward negative cloud-to-ground lightning in the near-ground area. The model allows upward leaders to be initiated on both tall and low buildings. The attachment process is analyzed when the multiple upward leaders are initiated and connected from both high and low buildings. The results show that low buildings have very small probability of initiating upward leaders directly and being struck, while high buildings have a clear influence on the initiation of upward leader of low buildings. Once low buildings initiate upward leaders, they are more likely to be stricken. The height difference between buildings is the main factor affecting the lightning attachment process. When the height difference between buildings is not large, the shielding effect of high buildings on the low buildings is not obvious, and the relative position of the downward leader channel will affect whether the low buildings can first initiate the upward leader. With the increase of the height difference between buildings, it is difficult for low buildings to preferentially initiate the upward leader. Only when the downward leader channel is obviously closer to low buildings, low buildings can initiate the upward leader and have a certain probability to connect with the downward leader to form a return stroke. When the height difference between buildings is large enough to a certain extent, the space form of the downward leader has little influence on the lightning attachment process. The low buildings will not be struck without initiating an upward leader. After the upward leader is initiated, it will have a certain inhibitory effect on the electric field intensity of the surrounding top angle. This inhibitory effect is related to the number of upward leaders and the horizontal distance between the top angle of the initiating upward leader and other top angles. The inhibitory effect is positively correlated with the amount of upward leaders while negatively correlated with the horizontal distance.

Abstract:
keep_len="250">The viticulture area around the Bohai Bay is the largest grape producing area in China. Waterlogging disaster is a major agricultural meteorological disaster in China, which seriously threatens grape production. Waterlogging indexes are utilized on field crops widely, but most of them can only be evaluated after the end of the growing season, which lacks the timeliness of monitoring and evaluating the process of waterlogging disasters. Taking the main grape producing areas in the Bohai Rim of China as the research object, the waterlogging grade index is constructed based on the daily meteorological data, grape growth stage data and grape waterlogging historical disaster data from 303 meteorological stations in the study area from 1980 to 2019. In the process of index construction, the influence of previous water surplus and deficit status on the current waterlogging process is fully considered, and the climate adaptability of crops in a certain place is considered. The daily waterlogging index of grapes is constructed by referring to the relative humidity index method of crops. Taking historical disaster inversion and disaster process analysis as the main line, Lilliefors test of normal distribution and t-distribution interval estimation method are used to construct the grape waterlogging disaster grade index system suitable for the main grape producing areas around the Bohai Bay, starting from the duration and intensity of waterlogging disaster. Based on the classification index of grape waterlogging disasters constructed above, the frequency of waterlogging disasters at each site in the Bohai Rim from 1980 to 2019 is counted, and the probability of disasters at each site is obtained by using information diffusion theory. Considering the probability and intensity of waterlogging disasters of each grade, the risk index of grape waterlogging in each station in the region is calculated. The results show that the occurrence range of waterlogging disaster in the same growth period of grape decreases with the increase of the disaster level, while the occurrence range of severe waterlogging disaster in different growth periods gradually increases with the advancement of development process. The risk of grape waterlogging is relatively low during the period of bud-shoot growth and flowering and fruit-setting, while the high-risk period of grape waterlogging is the period of fruit expansion and coloring and maturity. The high-risk areas of grape waterlogging disaster are mainly located in the southeast of Shandong Province, the southeast of Liaoning Province and the northeast of Hebei Province. The viticulture area around the Bohai Bay is the largest grape producing area in China. Waterlogging disaster is a major agricultural meteorological disaster in China, which seriously threatens grape production. Waterlogging indexes are utilized on field crops widely, but most of them can only be evaluated after the end of the growing season, which lacks the timeliness of monitoring and evaluating the process of waterlogging disasters. Taking the main grape producing areas in the Bohai Rim of China as the research object, the waterlogging grade index is constructed based on the daily meteorological data, grape growth stage data and grape waterlogging historical disaster data from 303 meteorological stations in the study area from 1980 to 2019. In the process of index construction, the influence of previous water surplus and deficit status on the current waterlogging process is fully considered, and the climate adaptability of crops in a certain place is considered. The daily waterlogging index of grapes is constructed by referring to the relative humidity index method of crops. Taking historical disaster inversion and disaster process analysis as the main line, Lilliefors test of normal distribution and t-distribution interval estimation method are used to construct the grape waterlogging disaster grade index system suitable for the main grape producing areas around the Bohai Bay, starting from the duration and intensity of waterlogging disaster. Based on the classification index of grape waterlogging disasters constructed above, the frequency of waterlogging disasters at each site in the Bohai Rim from 1980 to 2019 is counted, and the probability of disasters at each site is obtained by using information diffusion theory. Considering the probability and intensity of waterlogging disasters of each grade, the risk index of grape waterlogging in each station in the region is calculated. The results show that the occurrence range of waterlogging disaster in the same growth period of grape decreases with the increase of the disaster level, while the occurrence range of severe waterlogging disaster in different growth periods gradually increases with the advancement of development process. The risk of grape waterlogging is relatively low during the period of bud-shoot growth and flowering and fruit-setting, while the high-risk period of grape waterlogging is the period of fruit expansion and coloring and maturity. The high-risk areas of grape waterlogging disaster are mainly located in the southeast of Shandong Province, the southeast of Liaoning Province and the northeast of Hebei Province.

Abstract:
keep_len="250">Drought has brought great potential threat to agriculture, ecology, economy, society and available water resources of China, while accurate drought prediction can help risk management and development of early warning system, and it can reduce the destructive impact of drought. Among many prediction methods, data-driven model is a suitable tool with small data demand and fast development speed. With the development of machine learning, especially deep neural network (DNN), deep learning method shows great ability in drought prediction, and is reported to outperform traditional time series model (e.g. integrated moving average autoregressive model, ARIMA). However, its use needs to be widely estimated, further developed and adjusted for geoscience analysis. The standardized precipitation evapotranspiration index (SPEI) is reported to meet the needs of agricultural drought monitoring and early warning under the background of climate warming. SPEI at 1-, 3- and 6-month time scale are selected as the quantitative description of agricultural drought, and DNN model driven by meteorological and circulation variable is presented to explore the ability of SPEI prediction at the lead time of 1-3 months. The traditional long short-term memory neural network (TLSTM) has been used in drought prediction, which is limited by the quality of prediction factors and noise. Therefore, an improved TLSTM model (ILSTM) is proposed. With highlight of large-scale climate characteristics, a convolution neural network (CNN) module is combined with the ILSMT model. This newly-developed model (CLSTM) can extract circulation information that contributes to the regional drought change, as well as other outputs of prediction model. Evaluation of the drought prediction capabilities in different models is based on the Pearson correlation coefficient, the root mean square error, and the mean absolute error. Results indicate that overall prediction ability of DNN models outperforms the ARIMA model. And the comparative evaluation results among DNN models show that the architecture of the model has an important impact on the prediction performance. The ILSTM model can extract comprehensive information that contributes to future drought change by nonlinear coding of input variables through the full connected layer. When the correlation coefficient can be raised by 0.04-0.25, the root mean square error can be reduced by 0.07-0.32 and the mean absolute error can be reduced by 0.06-0.27 at the validation stage with different lead time comparing with the TLSTM model. Taking advantage of the circulation information as extra inputs to the ILSTM model, the CLSTM model outperform the ILSTM model, when the correlation coefficient can be raised by 0.03-0.44, the root mean square error can be reduced by 0.09-0.33 and the mean absolute error can be reduced by 0.05-0.26. Both results show that deep learning method has great ability in short-term regional climatic drought prediction. Drought has brought great potential threat to agriculture, ecology, economy, society and available water resources of China, while accurate drought prediction can help risk management and development of early warning system, and it can reduce the destructive impact of drought. Among many prediction methods, data-driven model is a suitable tool with small data demand and fast development speed. With the development of machine learning, especially deep neural network (DNN), deep learning method shows great ability in drought prediction, and is reported to outperform traditional time series model (e.g. integrated moving average autoregressive model, ARIMA). However, its use needs to be widely estimated, further developed and adjusted for geoscience analysis. The standardized precipitation evapotranspiration index (SPEI) is reported to meet the needs of agricultural drought monitoring and early warning under the background of climate warming. SPEI at 1-, 3- and 6-month time scale are selected as the quantitative description of agricultural drought, and DNN model driven by meteorological and circulation variable is presented to explore the ability of SPEI prediction at the lead time of 1-3 months. The traditional long short-term memory neural network (TLSTM) has been used in drought prediction, which is limited by the quality of prediction factors and noise. Therefore, an improved TLSTM model (ILSTM) is proposed. With highlight of large-scale climate characteristics, a convolution neural network (CNN) module is combined with the ILSMT model. This newly-developed model (CLSTM) can extract circulation information that contributes to the regional drought change, as well as other outputs of prediction model. Evaluation of the drought prediction capabilities in different models is based on the Pearson correlation coefficient, the root mean square error, and the mean absolute error. Results indicate that overall prediction ability of DNN models outperforms the ARIMA model. And the comparative evaluation results among DNN models show that the architecture of the model has an important impact on the prediction performance. The ILSTM model can extract comprehensive information that contributes to future drought change by nonlinear coding of input variables through the full connected layer. When the correlation coefficient can be raised by 0.04-0.25, the root mean square error can be reduced by 0.07-0.32 and the mean absolute error can be reduced by 0.06-0.27 at the validation stage with different lead time comparing with the TLSTM model. Taking advantage of the circulation information as extra inputs to the ILSTM model, the CLSTM model outperform the ILSTM model, when the correlation coefficient can be raised by 0.03-0.44, the root mean square error can be reduced by 0.09-0.33 and the mean absolute error can be reduced by 0.05-0.26. Both results show that deep learning method has great ability in short-term regional climatic drought prediction.

Abstract:
keep_len="250">The upper reaches of the Yangtze River is the hydropower resources and flood control focus for the whole river. Summer is an important period for flood diversion operation and hydropower development. Therefore, the relationships between summer runoff, precipitation and surface air temperature are analyzed, and the precursory physical climate signals for the runoff in the upper reaches of the Yangtze River are analyzed. By optimal subset regression and some other statistical methods, an annual increment prediction model with multi climatic factors for the runoff is built. The results show that the runoff directly depends on total precipitation in the basin, and they both show a slow downward trend in the past 40 years with a prominent quasi biennial oscillation. Their temporal correlation coefficient (TCC) is 0.81, exceeding the significant level of 0.001. By contrast, the average temperature of the watershed shows a significant upward trend, while influents less on the amount of runoff. On interannual time scale, the decisive role of precipitation on runoff is more prominent, while the influence of average temperature further weakens. Based on physical mechanism analysis,8 key climate preceding signals of runoff are selected. They are the Bay of Bengal monsoon and Australian High in winter, Indonesia Australia meridional wind shear, meridional position of the northern hemisphere polar vortex, Ural Mountain circulation, plateau monsoon and the temperature at high altitude basin in spring, and autumn sea level pressure dipole of the Indian Ocean. The prediction model for summer runoff built on these factors is tested by TCC, sign consistent rate (SCR), root mean square error (RMSE), absolute relative error (AE) and some other techniques. By the indication of test, fitting rate of the model is 0.81 during its modeling period from 1981 to 2015. In addition, SCR between the simulated and observed value is 77.1%, which is 100.0% for the abnormal years, and the RMSE is 0.57. After inversion calculation, TCC of the simulated with observed runoff is 0.66, exceeding the significant level of 0.001, and the average AE is 14.5%. In the post-test from 2016 to 2020, SCR and RMSE of the model are 80.0% and 0.99, respectively. The average AE of predicted runoff is 19.3%. Overall, the prediction accuracy of this model for summer runoff and its interannual variation characteristics of the upper reaches of the Yangtze River is more than 80%. Compared with the existing prediction models, prediction skills of this model are significantly improved, indicating a potential applicability. The upper reaches of the Yangtze River is the hydropower resources and flood control focus for the whole river. Summer is an important period for flood diversion operation and hydropower development. Therefore, the relationships between summer runoff, precipitation and surface air temperature are analyzed, and the precursory physical climate signals for the runoff in the upper reaches of the Yangtze River are analyzed. By optimal subset regression and some other statistical methods, an annual increment prediction model with multi climatic factors for the runoff is built. The results show that the runoff directly depends on total precipitation in the basin, and they both show a slow downward trend in the past 40 years with a prominent quasi biennial oscillation. Their temporal correlation coefficient (TCC) is 0.81, exceeding the significant level of 0.001. By contrast, the average temperature of the watershed shows a significant upward trend, while influents less on the amount of runoff. On interannual time scale, the decisive role of precipitation on runoff is more prominent, while the influence of average temperature further weakens. Based on physical mechanism analysis,8 key climate preceding signals of runoff are selected. They are the Bay of Bengal monsoon and Australian High in winter, Indonesia Australia meridional wind shear, meridional position of the northern hemisphere polar vortex, Ural Mountain circulation, plateau monsoon and the temperature at high altitude basin in spring, and autumn sea level pressure dipole of the Indian Ocean. The prediction model for summer runoff built on these factors is tested by TCC, sign consistent rate (SCR), root mean square error (RMSE), absolute relative error (AE) and some other techniques. By the indication of test, fitting rate of the model is 0.81 during its modeling period from 1981 to 2015. In addition, SCR between the simulated and observed value is 77.1%, which is 100.0% for the abnormal years, and the RMSE is 0.57. After inversion calculation, TCC of the simulated with observed runoff is 0.66, exceeding the significant level of 0.001, and the average AE is 14.5%. In the post-test from 2016 to 2020, SCR and RMSE of the model are 80.0% and 0.99, respectively. The average AE of predicted runoff is 19.3%. Overall, the prediction accuracy of this model for summer runoff and its interannual variation characteristics of the upper reaches of the Yangtze River is more than 80%. Compared with the existing prediction models, prediction skills of this model are significantly improved, indicating a potential applicability.