Table of Content

28 February 2021, Volume 37 Issue 1 Previous Issue    Next Issue

Articles

Select

Analysis of an air pollution event due to dust and anthropogenic emission in Fenwei Plain Xing-xing GAO,Hai-lin GUI,Nan WANG,Li ZHANG,Jian-peng WANG 2021, 37 (1):  1-8.  doi: 10.3969/j.issn.1673-503X.2021.01.001 Abstract ( 262 )   HTML ( 133 )   PDF (2586KB) ( 230 )   Save Using the data retrieved from the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite, the Ozone Monitoring Instrument (OMI), and the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO), ground environmental monitoring and meteorological observational data, and a backward trajectory model (Hybrid Single-Particle Lagrangian Integrated Trajectory Model, HYSPLIT), we analyzed the air pollution progress caused by dust and anthropogenic emission in Fenwei Plain from November 23 to December 6, 2018.The results showed that the severest air pollution occurs from November 26 to December 3, with the heaviest haze and dust pollution on December 1 and the heaviest dust pollution during the night on November 26 and the night on December 2.Heavy pollution lasts for 66 h, 42 h, and 37 h in Xi'an, Linfen, and Luoyang, respectively, and the total duration of the pollution process is 336 h, of which the relative humidity is more than 50% in 199 h, but not for the dust episodes.The haze episode in Xi'an is predominately caused by air pollutants accumulated locally and transported from the southwest and northeast.Haze in Linfen is mainly due to local accumulation of pollutants, and haze in Luoyang mainly resulted from pollutants transported from the northeast.Dust particles are mainly transported from the southwest to Xi'an and from the northwest to Luoyang, respectively, and Linfen is influenced less by dust.During the haze episode, the spatial distribution of high values of aerosol optical depth (AOD) is greatly affected by topography, and the values of absorbing aerosol index (AAI) are low.Air pollutants are mainly concentrated under 1.5 km above the ground, with polluted dust predominately trapped at the lower altitudes.During dust episodes, both AOD and AAI are high, exceeding 2.0 and 4.5, respectively.Pollutant concentration is least in the planetary boundary layer (PBL), and dust is the major pollutant at the lower altitudes.Whereas in the haze-dust mixed episode, pollutant concentration in the PBL is relatively higher, and the major pollutants at the lower altitudes include desert dust and polluted dust.The analysis from the HYSPLIT model shows that dust particles in the former dust episode are generated from the Xinjiang Uygur Autonomous Region and transported longer and higher, and traveled faster than those of the second haze-dust mixed episode that come from the west of Inner Mongolia Autonomous Region, causing severer pollution in Fenwei Plain. Figures and Tables | References | Related Articles | Metrics

Select

Impact of a south branch trough in winter on heavy pollution process in southern Sichuan urban agglomeration Li CHEN,Shi-gong WANG,Gui-cai NING,Jin FAN,Jun-biao LIAN 2021, 37 (1):  9-15.  doi: 10.3969/j.issn.1673-503X.2021.01.002 Abstract ( 199 )   HTML ( 16 )   PDF (2557KB) ( 98 )   Save We analyzed the pollution characteristics and meteorological causes of a heavy air pollution event in the southern Sichuan urban agglomeration (including Yibin, Zigong, Luzhou, Neijiang, and Leshan) from December 26, 2016, to January 11, 2017, based on the observational data of air pollutants' concentration and meteorological parameters and using methods of statistical analysis and weather diagnosis.The results indicated that the southern branch trough is the major weather system that contributes to the formation of this heavy pollution process.It strengthens and guides the dry, warm southwesterly flows in front of the trough, forming a local low-level isothermal or inversion layer and suppressing the dispersion of local air pollutants.The convergence of wind at low levels, the hygroscopic growth effect due to high relative humidity (without precipitation), and low-level isothermal temperature or temperature inversion, contributes to the peak concentration of air pollutants (AQI=286) in southern Sichuan agglomeration.As the south branch trough weakens, temperature inversion and low-level wind convergence disappear.The dispersion of air pollutants enhances, and air pollutants' concentration decreases.Finally, the wet removal effect of precipitation reduces pollutants' concentration quickly, ending this air pollution event.This study provides new important reference for understanding the meteorological causes of heavy pollution in the southern Sichuan urban agglomeration and its prevention and control, which is different from the causes for air pollution in the northwestern Sichuan Basin. Figures and Tables | References | Related Articles | Metrics

Select

Vertical distribution of atmospheric particulate concentration over the surface layer and its relationship with meteorological factors in Shenyang Li-guang LI,Xiao-lan LI,Zi-qi ZHAO,Hong-bo WANG,Li-du SHEN,Yang-feng WANG,Ning-wei LIU,Yan-jun MA 2021, 37 (1):  16-25.  doi: 10.3969/j.issn.1673-503X.2021.01.003 Abstract ( 284 )   HTML ( 17 )   PDF (3167KB) ( 612 )   Save Based on the observed PM2.5, PM10, TSP concentrations and meteorological factors including wind speed, air temperature, relative humidity at 1.5 m, 15 m, and 90 m heights from October 1, 2018, to September 30, 2019, in Shenyang, the vertical variation characteristics of atmospheric particulate concentration (APC) at different heights and their relationships with meteorological factors were analyzed.The results indicated that concentrations of different atmospheric particulates experience seasonal variations at different heights.The autumn and winter APCs are higher than those in spring and summer.The values of winter PM2.5 concentration at 1.5 m, 15 m, and 90 m heights are in turn 54.98±12.67 μg·m-3, 63.77±15.1 μg·m-3, 39.27±5.62 μg·m-3, i.e., the value at 15 m is greater than those at 1.5 m and 90 m, the corresponding values in autumn, spring and summer are in order of 1.5 m, 15 m, 90 m, the value at 1.5 m≈that at 15 m while the value at 15 m>that at 90 m, i.e.value (15 m) > value (90 m) > value (1.5 m), respectively.The diurnal variations of PM2.5, PM10, TSP concentrations are in the significant bimodal patterns in autumn and winter and unimodal patterns in spring at the three vertical heights, while the variations are in a unimodal pattern at 15 m height and have no obvious patterns at the other heights in summer.The variations of monthly mean APC show a significant difference between winter and summer seasons.Also, the concentrations of different sizes of atmospheric particulates vary with heights.In winter half-year, the monthly mean PM2.5, PM10, TSP concentrations at 1.5 m height and PM2.5 concentration at 90 m height indicate the change characteristics of increase-decrease-increase-decrease, while the monthly mean PM2.5, PM10, TSP concentrations at 15 m height and PM10, TSP concentrations at 90 m height tend to increase first and then decrease.The maximum values of APC in winter and summer half years occurred in January at 15 m height and June at three heights, respectively.The APCs at the two lower heights are higher than those at 90 m height in the winter half-year.In the summer half-year, the APCs among three heights have little difference and are much less than those in the winter half-year.The daily mean PM2.5, PM10, TSP concentrations are negatively correlated with wind speed and air temperature, and the correlation coefficients increase with the increase of the height and decrease with increase of the particle size.The relationships between daily mean PM2.5, PM10, TSP concentrations, and relative humidity are relatively complex, and the correlation coefficients vary with increase of the vertical height and particular size. Figures and Tables | References | Related Articles | Metrics

Select

Analysis of ensemble forecast error for spring soaking precipitation in Liaoning province Rui-wen YANG,Xin SUN,Zheng-hua TAN,Chen-he ZHANG,Ya-xin YU 2021, 37 (1):  26-32.  doi: 10.3969/j.issn.1673-503X.2021.01.004 Abstract ( 213 )   HTML ( 16 )   PDF (1049KB) ( 90 )   Save Based on the 24 h precipitation ensemble forecast products from three global numerical forecast centers (CMA, ECMWF, and NCEP) and surface precipitation observations in the Liaoning province from April 1 to June 30 from 2016 to 2018 and using Threaten Score, Bias, Talagrand and Brier Score methods, the forecast errors for spring soaking precipitation of the above-mentioned products in corresponding time were compared and analyzed.The results show that the dispersions of the ensemble forecast systems of the three centers are all small and their Talagrand diagrams all have a U-shaped distribution.Specifically, each ensemble forecast system overestimates the small magnitude of precipitation and has a high false alarm, and is incapable in forecasting the large magnitude of precipitation and underestimates the extreme value, which is easy to produce forecast deviation.In addition, by comparing the deterministic and probabilistic test results, it is found that TS value is higher and B values are closer to 1 for ECMWF relative to the other two forecast centers, implying that ECMWF has less false alarms than the other centers for spring soaking precipitation forecast in Liaoning province.Besides, the comparisons on the BS value and its decomposition score value also show that ECMWF is better than the other two centers.In summary, with the spring precipitation forecasts closest to the observations in Liaoning province and the best test results, ECMWF is available for forecast service work in the future as the main reference. Figures and Tables | References | Related Articles | Metrics

Select

Evaluation and projection of temperature change in Northeast China Xue AO,Chun-yu ZHAO,Yan CUI,Xiao-yu ZHOU,Qing-fei ZHAI,Li-du SHEN,Tao WANG 2021, 37 (1):  33-42.  doi: 10.3969/j.issn.1673-503X.2021.01.005 Abstract ( 338 )   HTML ( 12 )   PDF (1552KB) ( 214 )   Save Based on the RegCM4 (Regional Climate Model 4) regional climate model, CMIP5 (Coupled Model Intercomparison Project Phase 5) global climate model dataset and temperature observation data from 162 meteorological stations in Northeast China, the abilities of RegCM4 and CMIP5 to simulate temperature in Northeast China were evaluated using the deviation and correlation analyses.The future temperature changes under three emission scenarios of RCP2.6, RCP4.5, and RCP8.5 in Northeast China were estimated as well.The results show that both the regional and global models can better represent the spatiotemporal changes in air temperature.The simulation effects in winter and summer are better than those in autumn and spring.In terms of regional-scale information, the simulated values are smaller than the observed ones.The simulation results of the RegCM4 are significantly better than those of the CMIP5, and the simulated deviation cold is improved.In the future, the annual and seasonal temperatures in Northeast China will increase.The warming magnitude in the RCP8.5 scenario is most significant, followed by the RCP4.5, and RCP2.6 is the smallest.The warming effects in winter and autumn are larger than those in summer.Compared with the CMIP5 model, the RegCM4 model has a larger temperature increasing amplitude and more obvious interannual oscillation characteristics.In terms of space, the distribution patterns of short-, medium-, and late-of-term warming predicted by the regional and global models are consistent, showing a zonal distribution reducing from the north to the south, with the minimum amplitude of warming in Liaoning province, the high-value area locating in the Greater Khingan Mountains area of Heilongjiang province.Although the warming effect in the north is more obvious than that in the south, the temperature distribution in the northeast region will continue to be higher in the south than that in the north after warming. Figures and Tables | References | Related Articles | Metrics

Select

Comparative analysis of the mean temperature trend before and after homogenization of the mean temperature data in He'nan province Xing-jie JI,Ya-lei DING,Feng-xiu LI,Xuan ZUO 2021, 37 (1):  43-52.  doi: 10.3969/j.issn.1673-503X.2021.01.006 Abstract ( 393 )   HTML ( 12 )   PDF (3827KB) ( 159 )   Save Based on the data of daily mean temperature before and after homogenization from 111 meteorological stations during 1961-2017 in He'nan province, the statistical analysis method was used to analyze the changes of annual, seasonal and monthly mean temperature and their spatial distribution before and after homogenization in order to clarify the effects of unnatural factors on the trend change of mean temperature in He'nan province.The results show that from 1961 to 2017, the annual mean temperature in He'nan province increases significantly before and after homogenization.The increasing rate after homogenization is 0.21 ℃ per decade which is 0.02 ℃ higher than that before homogenization.Among the four seasons, the increasing rate of seasonal mean temperature in winter before and after homogenization both are the highest, which are 0.36 ℃ per decade and 0.38 ℃ per decade, respectively.Among the twelve months, the increasing rates of monthly mean temperature in February before and after homogenization both are the highest, the values are 0.49 ℃ per decade and 0.51 ℃ per decade, respectively.Among 111 meteorological stations, 41.4% of the total sites are not affected by non-natural factors, while the rest 58.6% are greatly affected by non-natural factors.For the annual mean temperature, after homogenization, 96.4% of the total stations increase significantly which are averagely 6.3% bigger than that i.e.90.1% before homogenization.Among the rest 58.6% of the total stations, the climatic tendencies of 43.3% and 15.3%are averagely underestimated and overestimated, respectively.For the seasonal mean temperature, 99.1% of the total stations significantly increase in both winter and spring after homogenization, while the percentages are 98.2% and 92.8% before homogenization, respectively.Among the twelve months, the months whose mean temperatures increase significantly in more than 50% of the 111 stations are February, March, April, October and December before homogenization, and January is added after homogenization.The effects of non-natural factors on the mean temperature result in the underestimation of the warming rate in He'nan province. Figures and Tables | References | Related Articles | Metrics

Select

Spatio-temporal distribution characteristics of dry lightning in the Greater Khingan Mountains of Inner Mongolia from 2014 to 2018 Xue-bin QU,Yan-ping WANG,Shu-xiang YANG,Hai-qing SONG,Yue-ji ZHAO,Xiao-hua ZOU 2021, 37 (1):  53-58.  doi: 10.3969/j.issn.1673-503X.2021.01.007 Abstract ( 274 )   HTML ( 16 )   PDF (2959KB) ( 112 )   Save Based on the data of lightning and the CLDAS (CMA Land Data Assimilation System) daily precipitation monitored at the lightning locators in the Greater Khingan Mountains of Inner Mongolia Autonomous Region from 2014 to 2018, the precipitation threshold during the dry lightning period was determined and the data of dry lightning monitoring was screened to analyze the spatio-temporal distribution characteristics of dry lightning in the Greater Khingan Mountains, to provide a scientific basis for the prevention of lightning fire in the Greater Khingan Mountains.The results show that the precipitation error between the CLDAS and meteorological stations in the Greater Khingan Mountains is small and has high applicability, which can meet the demand for dry lightning analysis.The annual average lightning frequency in the Greater Khingan Mountains is 36013.4 times, mainly with the negative ground flash.The annual average lightning days are 110.4 d, of which 42.3% of the amount of daily precipitation is less than 5 mm when the lightning occurs, which should be used as the precipitation threshold when dry lightning occurs in this region.According to the screening of this threshold, the annual average dry lightning occurring in the Greater Khingan Mountains is 1, 5229.8 times.The annual average dry lightning occurs mostly in July.The absolute intensity of the dry lightning current in May and September is stronger.Among the Forestry Services in the region, the intensity of the lightning current is generally strong at these regions including the eastern part of Yong'an Mountain, the south of Mangui, the eastern part of Along Mountain, the northern part of the Khanma Nature Reserve, the eastern part of the Ganhe River, the western part of the Ali River, the northern part of the Jiwen, the southeast of the Poplar, the southern part of Bahrain, the South Wood, and the eastern part of Chao'er, with the annual average dry lightning reaching 0.4 times/km2.These regions belong to the high-risk areas for dry lightning.The daily monitoring of dry lightning and the lightning fire prevention work in the above areas should be strengthened. Figures and Tables | References | Related Articles | Metrics

Select

Spatiotemporal distribution characteristics of thunderstorm tracks in Zhejiang province from 2007 to 2017 Xue-dong CUI,Wei-bin ZHANG 2021, 37 (1):  59-66.  doi: 10.3969/j.issn.1673-503X.2021.01.008 Abstract ( 170 )   HTML ( 17 )   PDF (2829KB) ( 108 )   Save Based on the nearly 11 a (2007-2017) cloud-to-ground lightning data monitored by the ADTD (Advanced TOA and Direction system) 2D lightning location system in Zhejiang province, the identification of thunderstorms and tracking their paths were realized using the density maximum fast search clustering and Kalman filtering algorithms.The spatial and temporal distribution characteristics of the process of large-scale thunderstorms were discussed.The results show that the above method can recognize and track all kinds of thunderstorms in Zhejiang province.261 tracks of thunderstorm processes are screened out, which show considerable annual and interannual variations.There is a good corresponding relationship between the number of the annual tracks and the frequency of annual lightning.The number of monthly tracks demonstrates a bimodal distribution, with peak values appearing in early spring and summer.The thunderstorms in spring generally move longer than those in other seasons.In the spatial distribution, the direction of 88.51% of thunderstorm tracks is from west to east, and more thunderstorms move toward the north.There is a difference in the monthly distribution of the number of thunderstorm tracks in each moving direction.The movement direction of thunderstorms in spring is more concentrated than those in summer.The dominant movement direction of summer thunderstorms is different.The moving speed of thunderstorms is about 50 km per hour.The thunderstorms moving from west to east and southwest to northeast are faster than in other directions.In terms of spatial distribution, there are two main mobile tracks.One is the area along the Jinqu basin, Shaoxing and Ningbo, the second is the area from Huzhou Jiaxing Plain to the north of Tianmu Mountain.From the perspective of topographic features, the tracks mainly occur in hilly areas, with relatively few plains and mountains, and the higher the terrain, the fewer paths. Figures and Tables | References | Related Articles | Metrics

Select

Error analysis of detection data of microwave radiometer and wind profiler radar under different weather conditions Shu-cheng CHEN,Xiao-bo LI,Ming CUI,Yan WANG 2021, 37 (1):  67-72.  doi: 10.3969/j.issn.1673-503X.2021.01.009 Abstract ( 222 )   HTML ( 10 )   PDF (979KB) ( 145 )   Save In order to explore the accuracy and availability of the detection data of microwave radiometer and wind profiler radar, the error characteristics of temperature and humidity detected by microwave radiometer and wind profiler radar under different weather conditions were analyzed using the GPS sounding data obtained during the National Sports Games in Tianjin.The results show that under sunny, cloudy, and rainy days, the microwave radiometer retrieves the low-altitude temperature profile well, while the inversion error of the upper-air temperature profile is large.Under cloudy days, the correlation between the inverted whole-layer temperature profile and the measured values from radiosonde is optimal.Under three kinds of weather conditions, the error of relative humidity profile inverted by the microwave radiometer is large, and the correlation with sounding data is poor.Under sunny and cloudy conditions, the error of wind direction and speed detection by wind profile radar is small, and the wind detection effect of wind profile radar on rainy days is poor.Under sunny and cloudy conditions above 1750 m, and rainy days above 3000 m, the wind speed data detected by the wind profile radar have good correlations with sounding data, while those detected at low altitude have poor correlations with sounding data.Below 3500 m, the correlation between the wind direction detected by the wind profile radar and sounding data is poor under three weather conditions, while above 3500 m, the correlation is good, and the value fluctuates between 0.6-1.0. Figures and Tables | References | Related Articles | Metrics

Select

Comparison and verification of 10-m wind field based on ECMWF fine grid and observations in Dalian Jing-tai YANG,Yu-xiu SUI,Jian XIAO,Li-na WANG,Hui-lin CHANG,Xiao-dong WU,Yan WANG 2021, 37 (1):  73-81.  doi: 10.3969/j.issn.1673-503X.2021.01.010 Abstract ( 511 )   HTML ( 29 )   PDF (1101KB) ( 199 )   Save The 10-m wind data of European Center for Medium-Range Weather (ECMWF)fine mesh grid from January of 2016 to December of 2018 were evaluated using various observations from 8 national meteorological observation stations in Dalian.The results indicated that the 10 m wind forecast of ECMWF fine mesh grid is the closest to the maximum wind speed, and has the best correlation with the maximum instantaneous wind speed.The whole forecast results of ECMWF 10 m wind speed for eight stations in Dalian are relatively larger on average.From the error statistical analysis between ECMWF 10 m wind forecast data and its corresponding maximum wind speed, the results showed that forecast is closest to the observation when the wind is at level 3, larger than observations when the wind is below level 3, and smaller than that when the wind is above level 3 by the classification of the observation data from the mean error (ME) of wind speed.The wind speed errors of different directions are also obvious, but smaller than those of different wind scales.Mean absolute error (MAE) is the smallest when the wind is between levels 2 and 3.There is no significant difference in ME during all the forecast times, which is mainly between 0.1-0.3 m·s-1.But MAE increases slowly with the extension of forecast time.The wind speed errors had obvious diurnal variations, which show the characteristics of small in the day and large at night, the smallest in the afternoon, and the largest after midnight.They also vary at different stations, wind levels, and wind directions. Figures and Tables | References | Related Articles | Metrics

Select

Study on the closure of cloud condensation nuclei based on ion pairing method at Huangshan Mountain Xin QIN,Jin-guang ZHANG,Ze-feng ZHANG,Qing MIAO,Hong-qiang LI,Shu-hui ZHAO,Si-hang QUAN,Jia-li MA 2021, 37 (1):  82-90.  doi: 10.3969/j.issn.1673-503X.2021.01.011 Abstract ( 172 )   HTML ( 7 )   PDF (1750KB) ( 69 )   Save To conduct a closed study on cloud condensation nuclei (CCN) at Guangmingding, Huangshan Mountain, physical and chemical properties and CCN number concentration of atmospheric aerosols were observed from June 30 to July 28, 2014.The temporal characteristics of chemical composition, spectral distribution, and CCN number concentration were analyzed.The CCN number concentration calculated by ion-pairing method based on κ-Köhler theory was compared with the observations.The results indicated that the calculated and measured CCN concentrations show close correlation and high accuracy.The results of low supersaturation CCN closure are better than those of the high supersaturation, the CCN number concentration is underestimated when the supersaturation is low and overestimated when that is high, which shows that the chemical composition data of aerosol is very important for the prediction of the CCN number concentration, and the use of this method can lead to the closure of CCN.Considering the influence of 40% Water Soluble Organic Carbon (WSOC) on the hygroscopicity of aerosol particles, CCN closure results are better at the lower supersaturation, but the effect is not significant, especially when the fitting results are relatively low, there is no effect under poor high supersaturation.Therefore, the water-soluble inorganic components in aerosol particles have an important effect on the activation of CCN, while the WSOC, which contains more content, complex chemical components, and uncertain hygroscopicity, has a limited effect on the activation of CCN, which is consistent with some studies that the influence of aerosol hygroscopicity is more important than that of organic components with complex characteristics. Figures and Tables | References | Related Articles | Metrics

Select

Effect of climate change on energy consumption in office building in Northeast China Yan CUI,Xue AO,Jing-fu CAO,Xiao-yu ZHOU,Ming-cai LI,Chun-yu ZHAO,Ming-yan LIU,Rong LIN 2021, 37 (1):  91-99.  doi: 10.3969/j.issn.1673-503X.2021.01.012 Abstract ( 235 )   HTML ( 14 )   PDF (1348KB) ( 98 )   Save Based on the meteorological data including air temperature, relative humidity, etc, and energy consumption data simulated with Transient System (TRNSYS) from 1961 to 2019 in three typical cities (Harbin, Changchun, and Shenyang) in Northeast China, we analyzed the impact of climate change on design meteorological parameters and energy consumption in office buildings.The results showed that the values of three office-building outdoor design parameters (i.e., outdoor design temperature for heating, and outdoor design temperature for winter air-conditioning, and outdoor design temperature for summer air-conditioning) increase in Northeast China during the past 30 years (1991-2019) in comparison with years of 1961-1990.The outdoor design temperature for summer air-conditioning has a lower increase than the rest two parameters.The outdoor design temperature for heating in the three typical cities increases by 2.1 ℃, 1.7 ℃, and 0.2 ℃, respectively.During 1961-2017, the heating energy consumption of office buildings in the three cities shows a decreasing trend in winter and an increasing trend in summer, and the annual cumulative energy consumption exhibits a declining trend.The decreasing rate of annual total energy consumption in Harbin and Changchun is higher than in Shenyang, at a rate of 5.02 MJ·m-2/10 a, 6.15 MJ·m-2/10 a, and 1.99 MJ·m-2/10 a, respectively.Air temperature is a major meteorological factor that affects the energy consumption in an office building in Northeast China, accounting for 95%, 96%, and 93% of the variation in winter heating energy consumption and 72%, 71%, and 72% of the variation in summer cooling energy consumption at three cities, respectively.With a 1 ℃ increase in temperature in the three cities, winter heating energy consumption will decrease 20.6 MJ·m-2, 21 MJ·m-2, and 18.9 MJ·m-2, summer cooling energy consumption will increase 15.1 MJ·m-2, 16.1 MJ·m-2, and 18.8 MJ·m-2, and the annual cumulative energy consumption will decrease 5.5 MJ·m-2, 4.9 MJ·m-2, and 0.1 MJ·m-2, respectively. Figures and Tables | References | Related Articles | Metrics

Bulletins

Select

Characteristic analysis of red rainstorm warning signals from 2015 to 2019 in Liaoning province Jing LIU,Chuan-lei CHEN,Jun YAN,Ying WANG,Kui-zhi CAI,Chuan REN,Mei HAN,Wei DONG 2021, 37 (1):  100-105.  doi: 10.3969/j.issn.1673-503X.2021.01.013 Abstract ( 334 )   HTML ( 15 )   PDF (1750KB) ( 89 )   Save Using the information of the red rainstorm warning signals and the minute-level precipitation observation data of 1605 automatic stations in Liaoning province from 2015 to 2019, the inter-annual variation characteristics and spatio-temporal distribution of the red rainstorm warning signals and the short-term heavy rainstorm are counted.Furthermore, the centralized area and time of the red rainstorm warning signals are analyzed.The results showed that the number of red rainstorm warning signals from 2015 to 2017 has increased year by year, with a maximum of 147 in 2017.The advance time of red rainstorm warning signals has little decreased for the improvement of accuracy, with the minimum value of 19 minutes in 2018.The number of warning signals issued in 2019 has 59 less than in 2018, the advance time increases 29 minutes.The frequency of the red rainstorm warning signals is high in the southeast and low in the central areas.Statistics also show that the red rainstorm warning signals are mostly issued at night.In more than 50% of the cases, the time when the accumulated rainfall meets the issuing requirements of the red rainstorm warning signal is 90 minutes after the time with the maximum rainfall intensity.So, the occurrence time of the maximum rainfall intensity is an important indication for the issuing of the red rainstorm warning signal.To improve the surface effect on disaster prevention and mitigation, when considering the issuing of red rainstorm warning signals, the occurring time of the maximum rainfall intensity, the issuing time, the high-incidence area and the geography should be considered carefully. Figures and Tables | References | Related Articles | Metrics

Select

Identification of the fog image features on the Yangtze River waterways in Chongqing based on machine learning Yuan-mou WANG,Jia-qi LI,Shi-ji CHEN,Jia-ping TANG,Bai-cheng XIA,Shi-gang HAN 2021, 37 (1):  106-112.  doi: 10.3969/j.issn.1673-503X.2021.01.014 Abstract ( 216 )   HTML ( 16 )   PDF (1052KB) ( 204 )   Save Based on the image data of fog events taken by radar stations on the Yangtze River waterways in Chongqing, images without fog events and with five types of fog events were trained using algorithms including K-nearest neighbor, support vector machine, back propagation neural network, and random forest.According to the training results, a fog image identification model was built, and the identification accuracy was tested.The results show that machine learning can effectively identify fog images, and random forest performances better than the other three algorithms.The model has a recognition accuracy of 100% for non-fog recognition, over 90% for light fog and strong dense fog events, and over 70% for fog, heavy fog, and dense fog events. Figures and Tables | References | Related Articles | Metrics