Abstract:To improve the forecasting accuracy of high-temperature weather,this study conducted a statistical analysis of the characteristics of high-temperature events in Meizhou over the past 60 years,based on daily maximum temperature data from seven national meteorological stations in Meizhou from 1964 to 2023 and ERA5 reanalysis data from the European Centre for Medium-Range Weather Forecasts(ECM-WF).High-temperature events exceeding 37℃and 39℃were classified.The results indicate the follow-ing:(1)Over the past 60 years,high-temperature weather in Meizhou exhibited distinct interdecadal vari-ation characteristics.The number of days with temperatures≥35℃showed an overall increasing trend,with both the number of high-temperature days and the annual extreme maximum temperature demonstrating a significant upward trend.High-temperature events were predominantly short-duration,with prolonged events being relatively rare.Among them,one-day events were the most frequent,occurring 230 times,while the longest continuous high-temperature period reached 21 days.(2)The extreme maximum temper-atures across Meizhou over the past 60 years ranged between 39.3℃and 39.9℃.The maximum center was located in the Meixian-Dapu area,with an average annual high-temperature days of 38.6 days.In contrast,the southern regions of Wuhua and Fengshun experienced fewer high-temperature days,with an average of only 21.1 days per year.(3)Based on the 500 hPa circulation patterns,240 anomalous high-temperature events were classified into three types:subtropical high circulation type,peripheral typhoon type,and combined subtropical high circulation and peripheral typhoon type.Among these,the subtropical high circulation type and the combined type were equally prevalent.The stable control of the 588 dagpm line of the subtropical high over Guangdong Province,with its strong subsidence airflow,easily induced high-temperature heatwaves in Meizhou.The ridge line of the subtropical high was most frequently located near 30°N.When tropical cyclones were 500-1500 km away from Meizhou,temperatures exceeding 37℃were likely to occur.Under appropriate distance conditions,the occurrence of anomalous high temperatures was not significantly correlated with the intensity of tropical cyclones.

Abstract:To enhance flood disaster preparedness in the Danshui River Basin,the HBV(Hydrologiska Byråns Vattenbalansavdelning)hydrological model was employed to simulate runoff using hydrological and meteorological data.The study established relationships among rainfall,discharge,and water level,and in-vestigated critical rainfall thresholds.After calibration and verification,the model performance indices—Nash-Sutcliffe Efficiency Coefficient(NSE)and Water Balance(WB)—both exceeded 0.6,indicating that the HBV model effectively captured the overall trend and peak flood processes of the Danshui River run-off.Based on the water level-discharge relationship,the rainfall-discharge-water level correlation was derived using the discharge outputs from the HBV model.By incorporating antecedent water level data,the water level-rainfall relationship under different initial conditions was determined.The three critical water levels at the Danshui Hydrological Station were then applied to this relationship to establish the corresponding rainfall-antecedent water level thresholds for each critical water level.Finally,the derived thresholds were validated against four historical flood events,with two predictions successful and two unsuccessful.Further analysis revealed that the two failed predictions were primarily due to significant errors in flow simulation,at-tributed mainly to inaccuracies in the empirical parameters related to the runoff coefficient and surface runoff regression coefficient.

Abstract:This study aims to analyze the patterns and characteristics of flood-triggering rainstorms in Qingxin District,Qingyuan City.Based on Micaps data,the CROSS Guangdong Provincial Climate Reanaly-sis Statistical Data Operational Service System,and NCEP/NCAR reanalysis data,a statistical analysis was conducted on 22 flood records corresponding to rainstorm processes from 2000 to 2024 provided by the Flood Control,Drought Relief,and Wind Prevention Office of Qingxin District.The results indicate that rainstorm processes can be classified into six types:typhoon type,frontal type,warm-sector type,southwest mon-soon+shear line type,upper-level trough+low vortex type,and low-pressure trough+easterly dis-turbance type.These types differ significantly in terms of occurrence season,influencing systems,rainfall intensity,duration,and cumulative rainfall.

Abstract:To improve the prediction and early warning capabilities of rainfall-induced geological dis-asters,this study utilized geological disaster data from Longchuan County from 2010 to 2020,along with hourly and daily meteorological data from the same period.Rainfall-induced geological disasters were iden-tified using effective rainfall,and their rainfall impact periods were classified to determine rainfall thresholds for different rainfall types.The research shows that among rainfall-induced geological disasters in Long-chuan County,60.6%were triggered by short-term rainfall,while long-duration rainfall accounted for 39.4%.Among these,short-duration rainfall-induced geological disasters made up 20.5%of the short-term rainfall types.Geological disasters are likely to be triggered when the cumulative rainfall of short-duration heavy rainfall over 12 hours reaches 70 mm,the effective cumulative rainfall 1-3 days prior ex-ceeds 65 mm,or the effective cumulative rainfall over 4-10 days exceeds 46 mm.

Abstract:Using hourly ground station data,ERA5 reanalysis data,and other sources,this study in-vestigates warm-sector rainstorms in northern Fujian from April to June between 2012 and 2020,summari-zing their weather background and climatic characteristics.The results are shown as follows.(1)Precipitati-on on the northwestern side of the Wuyi Mountains is significantly higher than on the southeastern side,with the windward slope near Huanggang Mountain and the bell-mouthed terrain formed by Dajue Mountain and Mount Wuyi receiving the most rainfall.(2)The peak rainfall of warm-sector rainstorms in northern Fujian occurs from midnight to morning,while the lowest values appear in the afternoon.(3)These rainstorms can be classified into four types:(a)low-level southerly winds transitioning to cold shear passage,(b)warm shear lifting northward under southerly wind control,(c)southerly wind control with upper-level wave dis-turbances,and(d)low-level southerly winds influenced by the subtropical high.(4)Key synoptic-scale conditions for warm-sector rainstorms include a high-level westerly jet,500 hPa fluctuations,a low-lev-el southwest jet,wind speed convergence,a boundary layer jet,and terrain effects.The combined influence of topographic uplift and atmospheric circulation creates a strong low-level convergence center,which,cou-pled with upper-level divergence,provides favorable upward motion for rainstorm development.

Abstract:To further improve the forecasting of short-term heavy rainfall,this study utilizes hourly precipitation data from national and regional meteorological stations in northern Guangdong from 2014 to 2023.Through comparative analysis,the influence of different station distribution densities on the spatiotem-poral distribution of extreme short-term heavy rainfall is investigated.The results indicate that regional sta-tions better reflect the extremity,dispersion,and locality of extreme hourly heavy rainfall compared to na-tional stations.The 99.8th percentile value is identified as the most suitable threshold for extreme short-term heavy rainfall in this region.The extreme values of hourly heavy rainfall in northern Guangdong are pri-marily concentrated between 50-80 mm.Among the stations with thresholds above 50 mm at the 99.8th percentile,the majority are located in Qingcheng,Fogang,and the eastern part of Yingde in Qingyuan;Wu-jiang and Qujiang in southern Shaoguan;the eastern and southern parts of Heyuan;and the northern part of Meizhou.In terms of monthly variations,the frequency exhibits a bimodal distribution,consistent across both national and regional stations,with peaks occurring in June and August.Regarding intensity,regional stations show higher hourly rainfall intensity and extreme values compared to national stations.The diurnal variations in both the frequency and intensity of extreme hourly heavy rainfall display a typical bimodal pat-tern,with peaks occurring in the evening and early morning.

Abstract:To carry out professional meteorological services more accurately,this study focuses on the Dengta Basin in Heyuan City,Guangdong Province,utilizing minimum bare-ground temperature data from 1974 to 2023 and minimum air temperature data from 1965 to 2023 from the national basic meteorological stations in Lianping and Heping.The research analyzes the climatic variation characteristics of frost days and low-temperature probability in the region.The findings reveal that the annual average number of frost days in the Dengta Basin shows a significant decreasing trend,particularly after 1996,with a notable reduction in frost days,delayed peak frost occurrence,and a shortened frost period.From December to February,crops are susceptible to freezing and chilling injuries.The study recommends that farmers consider the probability of low temperatures when selecting crops,especially for cold-sensitive tropical and subtropical varieties.They should opt for cultivars that mature and can be harvested during months with lower low-temperature probability or implement appropriate overwintering measures.

Abstract:To enhance the capability of specialized meteorological services for marine operations,wind speed data from the Beishan automatic weather station,offshore wind tower station,and buoy station along the Maoming coast from 2019 to 2020 were selected.A comparative analysis was conducted on the character-istics of destructive winds at these coastal stations and the differences in gust factors under various weather systems.The results indicate the following.(1)There is a significant variation in the number of windy days among the stations.The buoy station recorded twice as many windy days as the offshore wind tower,while the wind tower's count was similar to that of the Beishan station.(2)October had the highest number of windy days across all stations,followed by November,with December ranking third.Similarly,October also had the most strong wind days,but August and September ranked second and third,respectively.(3)Both tropical cyclones and cold air are the primary causes of strong winds at the Maoming buoy station,while se-vere convective weather is the least likely to induce strong winds.In contrast,strong winds at the Beishan station and offshore wind tower are mainly triggered by tropical cyclones.(4)The highest gust factor was ob-served at the offshore wind tower,followed by the Beishan station,with the buoy station recording the low-est.This is closely related to the surface roughness of the underlying terrain—the rougher the surface,the larger the gust factor,and the smoother the surface,the smaller the gust factor.

Abstract:To provide a reference for drought and flood forecasting during the flood season,this study recalculates the Standardized Precipitation Evapotranspiration Index(SPEI)using data from 86 national me-teorological stations in Guangdong and NCEP reanalysis data from 1973 to 2022.Methods such as Empirical Orthogonal Function(EOF)decomposition and Mann-Kendall test were employed to classify drought and flood patterns during Guangdong's flood season and analyze their causes.The results indicate the following.1)The spatial distribution patterns of drought and flood during Guangdong's flood season can be categorized into three types:province-wide consistency,east-west pattern,and north-south pattern.2)Over the past 50 years,Guangdong's flood season has shown a consistent drying trend,particularly during the 1989-1993 period,where the drying trend was pronounced.3)The main rainfall areas during the flood season are Yangjiang,Jiangmen,and Shanwei;eastern Guangdong is more sensitive to drought conditions,while northern Guangdong is more sensitive to flood conditions.4)The province-wide consistency pattern may be influenced by the intensity of the southwest monsoon,the east-west pattern may be related to the subtropi-cal high and westerly trough,and the north-south pattern may be affected by low-pressure systems in the South China Sea,such as typhoons or monsoon troughs.

Abstract:To study the convective storm structure and evolution characteristics of westerlies tornadoes,this paper utilizes surface and upper-air meteorological observation data,as well as Doppler weather radar observations,to conduct a detailed observational analysis and comparative study of tornado events in the Pearl River Delta under the background of westerlies weather.The results show that(1)Tornadoes occur in environments with strong low-level(0～1 km)vertical wind shear,characterized by high convective availa-ble potential energy,a high K-index,and a clear"dry above,moist below"unstable stratification.(2)From the perspective of local meteorological element evolution,westerlies tornadoes have weaker precipitati-on intensity but are more localized;tornadoes occur on the side of negative temperature zones produced by precipitation,where the cold pool formed by the precipitation interacts with the warm and moist air at the surface,facilitating local small-scale cyclonic convergence and causing tornadoes to form and develop.(3)Tornadoes are produced from supercells,which exhibit clear hook echo features,deep mesocyclone charac-teristics,and inflow notches/bounded weak echo regions.(4)Numerical simulations reveal that the interac-tion between the triple point convergence boundary and supercell storms provides favorable conditions for the formation of tornado storms.

Abstract:To improve the forecasting capability of consecutive cold air processes,this study analyzes the differences between two consecutive cold air events affecting Guangzhou in late February 2024 using ERA5 reanalysis data,surface automatic weather station observations,and vertical sounding data.The re-sults show that the first cold air mass moved southward relatively slowly,while the second one advanced fas-ter.Additionally,the difference in minimum temperatures between southern and northern Guangzhou was larger during the first event but smaller during the second.A comparison between numerical model forecasts and observations revealed that,for the first and second cold air processes,the underlying surface conditions were initially warmer and cooler,respectively.The model predicted a faster southward movement of cold air than observed in the first event but slower in the second.Furthermore,the daytime maximum temperature forecasts were lower and higher than the actual observations for the first and second events,respectively.This study provides valuable insights into the mechanisms of consecutive cold air processes and offers impor-tant references for improving the accuracy of temperature forecasts in subtropical coastal cities.

Abstract:To understand lightning characteristics and implement scientific lightning protection,this study analyzes the cloud-to-ground(CG)lightning features during two rainstorm events in 2015 using data from a lightning location system,combined with synoptic weather conditions and observations from 40 region-al automatic weather stations.The results show that during the squall line passage,the active periods of lightning in both rainstorm processes occurred in the latter half of the events,dominated by negative CG fla-shes,which accounted for over 70％of the total CG flashes.Variations in the frequency of positive and neg-ative CG flashes exhibited certain differences at different stages.The number of CG flashes changed sharply,with the trends of both positive and negative CG flashes aligning with the overall trend,though negative CG flashes exhibited more abrupt variations.The spatial distribution of CG lightning frequency was uneven in both rainstorm processes,with significantly higher activity in the western and central-southern regions com-pared to the northeastern mountainous areas.The areas with heavy precipitation largely coincided with those of high CG lightning frequency.The findings of this study provide valuable insights into the relationship be-tween lightning activity and convective weather systems,which can enhance lightning risk assessment and early warning capabilities for extreme weather events.

Abstract:To understand the accuracy of machine learning in wave prediction,this study utilizes hourly observational data from the Maoming buoy station in Guangdong from February 2013 to June 2024.The SHAP(SHapley Additive exPlanations)machine learning method is applied to analyze the specific influ-ences of meteorological factors such as wind speed,air temperature,and air pressure on significant wave height(SWH)in the coastal waters of Guangdong.Additionally,the performance characteristics of ten ma-chine learning models in predicting SWH are evaluated.The results indicate the following.Wind speed is the most critical factor in SWH prediction,followed by air temperature and month.In two experimental sce-narios-single-factor(wind speed)and multi-factor(wind speed,month,air temperature)model fitting-support vector regression(SVR)consistently outperforms other models.Particularly in multi-factor fitting,SVR achieves a root mean square error(RMSE)of approximately 0.31,a correlation coefficient(CORR)of 0.87,and a coefficient of determination(R2)of 0.74,demonstrating excellent predictive performance.Light gradient boosting machine regression(LGBM)exhibits comparable stability to SVR in prediction.

Abstract:To investigate the environmental background of pre-typhoon severe convective weather,this study analyzes radar echo characteristics,circulation patterns,and selected physical parameters of pre-ty-phoon severe convection events in the Guangdong-Hong Kong-Macao Greater Bay Area from 2007 to 2022,using MICAPS upper-air and surface observations,radar reflectivity,typhoon tracks,and NCEP FNL reanal-ysis data.The results show that pre-typhoon severe convection occurs most frequently from July to August,particularly from noon to evening.These events mainly develop at the edge or within the circulation of the subtropical high,with low-level southerly flow or northerly flow from the typhoon periphery,and the surface is often located near a trough or in a quasi-uniform pressure field.Nearly 30％of pre-typhoon squall lines have a maximum 40 dBZ echo length of 200-250 km.Over 90％of cases exhibit maximum echo intensities between 55-65 dBZ,with 78％of non-squall line events falling within this range.Radar echoes most fre-quently move northwestward,with pre-typhoon squall lines averaging a speed of 32 km·h-1and a mean li-fespan of about 5.1 hours.Considering seasonal variations,t850-500,t925-700,and the Most Unstable Lifted Index(MULI)show certain predictive indications for pre-typhoon severe convection events from September to October.This study enhances the understanding of pre-typhoon severe convection mechanisms in the Greater Bay Area,providing valuable insights for early warning systems and improving forecasting accuracy of high-impact weather events under typhoon influence.

Abstract:To enhance the understanding of extreme heavy rainfall associated with typhoon remnant vor-tices,this study analyzes the extreme heavy rainfall event induced by the remnant vortex of Typhoon"Haikui"in Guangdong,using precipitation data from national and regional automatic weather stations and ERA5 reanalysis data.The results show that:(1)The most intense rainfall occurred from the night of the 7th to the daytime of the 8th.Multiple stations in the Pearl River Delta region recorded exceptionally extreme precipitation,with Donghu Street in Shenzhen and Hecheng Street in Foshan breaking local historical records for 12-hour and 24-hour rainfall.(2)The stable maintenance of synoptic-scale systems,including the conti-nental high to the north of the remnant vortex and a deep cold vortex to the northeast,provided favorable background conditions for the extreme rainfall.A southeasterly low-level jet-formed by the convergence of easterly return flow and southerly monsoon winds-delivered highly favorable moisture,thermal,and dynamic conditions.(3)The anomalies in dynamic and moisture conditions were remarkable,with the maximum low-level dynamic convergence intensity and moisture flux divergence reaching over 5σ in standardized anoma-lies.This study provides critical insights into the mechanisms of extreme rainfall associated with typhoon remnant vortices,aiding in improved forecasting and disaster mitigation for such high-impact weather events.

Abstract:To investigate the microphysical processes of a hailstorm event in Shunde District,Foshan,on April 21,2023,observational data from a DSG5 laser disdrometer were processed to analyze the particle distribution characteristics during the hailstorm.Instantaneous rainfall intensity,radar reflectivity factor,and other microphysical parameters were calculated.The results indicate the following.The precipitation parti-cles during this hailstorm were primarily concentrated in the size range of 0.3-3.0 mm and the velocity range of 2-7 m/s.During the hailfall,the particle spectrum evolved from a unimodal to a bimodal structure,with a rapid increase in rainfall intensity.Both the particle size and velocity spectra exhibited significant broade-ning,along with an increase in particle number concentration.The entire particle spectrum shifted upward,and the disdrometer sampled precipitation particles larger than 7 mm.A sudden surge in radar reflectivity factor was observed during the hailfall,indicating that intense convective motions generated larger precipita-tion particles.The velocity-diameter distribution of this rainfall event mostly appeared above the empirical curve,likely due to the combined effects of hailstones and strong downdrafts.

Abstract:To better conduct research on cold air and low-temperature processes and assess their im-pacts,an operational and objective identification method for cold air-low temperature processes is proposed based on their synchronous characteristics.By applying this method,the cold air-low temperature processes identified between 2020 and 2022 were compared with historical records.The results show that 87％of the processes had start and end time differences within 1 day,with 47％being completely consistent.In typical cold air-low temperature processes in Guangzhou,the longest-lasting processes and those with the lowest ex-treme minimum temperatures mainly occurred in December and January,while the processes with the largest temperature drops primarily occurred between February and April.This study provides a robust and opera-tional framework for identifying cold air-low temperature processes,which is crucial for improving climate risk assessment,disaster prevention,and public health preparedness in urban areas.

Abstract:To improve the accuracy of forecasting extreme heavy rainfall associated with typhoons,this study utilizes 3D retrieved wind fields from Guangdong S-band radars and Foshan X-band radars to analyze the relationship among low-level jets,structural changes in the remnant vortex of Typhoon"Haikui,"and rainfall distribution.The results show the following.(1)The development of the jet on the southeastern side of the remnant vortex is closely related to heavy rainfall.This jet first appears at low levels and gradually ex-tends upward,approaching the vortex center,with heavy rainfall areas shifting from the right side of the vor-tex toward its center as the jet strengthens.(2)The low-level jet causes the remnant vortex center to tilt ver-tically,leading to the development of a new circulation center.The new and original centers merge as they move closer,causing the vortex to slow down and meander west of the Pearl River Estuary,resulting in pro-longed heavy rainfall over the Pearl River Delta.(3)The intensification and northward movement of the jet over the northern South China Sea lead to a renewed strengthening of rainfall in the Pearl River Delta,with heavy precipitation occurring in areas where different airflows converge pairwise.This study provides critical insights into the dynamic mechanisms linking low-level jets and remnant vortex evolution with extreme rain-fall distribution,offering valuable references for improving typhoon-related heavy rainfall forecasting in coast-al metropolitan regions.

Abstract:To enhance the understanding of the occurrence patterns of short-term heavy rainfall in Fos-han City and improve the forecasting and warning capabilities for such rainfall and its derived disasters,hourly observational data from 194 automatic weather stations in Foshan from 2015 to 2024 were analyzed to investigate the variation characteristics of short-term heavy rainfall at different intensity levels and time peri-ods.The results indicate the following.(1)The annual trends in the frequency of short-term heavy rainfall vary inconsistently across different intensity levels,with greater interannual fluctuations as intensity increa-ses.Heavy rainfall primarily occurs between 13:00-20:00 from April to September.(2)Its frequency is generally higher between 14:00-20:00 than at other times,with the smallest interannual fluctuations during this period.The frequency shows an increasing trend from 02:00-08:00 and 14:00-20:00 but a decrea-sing trend from 20:00-02:00 and 08:00-14:00.(3)The spatial distribution of short-term heavy rainfall frequency varies significantly across different intensity levels and time periods,with an inverse relationship between frequency distribution and the coefficient of variation.(4)In operational forecasting,particular at-tention should be paid to short-term heavy rainfall occurring between 14:00-20:00 during the rainy season in the eastern part of Nanhai District and the southern part of Shunde District.This study provides critical in-sights into the spatiotemporal patterns of short-term heavy rainfall,offering a scientific basis for improved dis-aster prevention and urban resilience in Foshan City and similar subtropical regions.