应用气象学报

Diagnosis of a Heavy Rain Event Caused by the Intense Development of Yellow River Cyclone in July, 1998

Liang Feng, Tao Shiyan

2007, 18(5): 577-585.

[FullText HTML] (743) [PDF 3462KB](2076)

Abstract:
A Yellow River cyclone intensifies rapidly during July 5 to 7, 1998. Its center pressure decreases by 12 hPa over a 24 h period and it produces heavy rain with the maximum rainfall exceeding 350 mm in Beijing. A diagnostic study is conducted from a potential vorticity or "PV thinking" perspective using NCAR/NCEP 6 hourly reanalysis data. The results show that the rapid development of the Yellow River Cyclone is related to the coupling between a surface low system and an upper level positive PV anomaly. When the positive PV anomaly near the tropopause advects over a pre existing surface cyclone, the cyclone deepens dramatically. Warm advection at 850 hPa intensifies the development of cyclone. Heavy rain occurs in the rapid intensification stage of the Yellow River Cyclone. This is a synoptic scale precipitation case, which is caused by the convergence between cold air descending from stratosphere and southwest warm and moisture air flow brought by the Monsoon Surge. On the vertical cross section through the region of heavy rain, abrupt jump of tropopause is shown clearly. The tropopause is near 250 hPa on the cold side and rises dramatically to above 100 hPa on the warm side. The extent of the descent of stratospheric air in the storm can be deduced by the tongue of 1 PVU extending from 200 to 600 hPa.While the cyclone intensifies rapidly, there are strong ascending motions, which lead to the deep moisture air from the south of China transporting to the heavy rain region continually by southwest wind. From the low to mid level of trop osphere, the humidity increases. The amount of precipitable water vapor increases 10.2 mm in 24 hours. The atmosphere is baroclinic over heavy rain region. High level jet with wind speed greater than 30 m·s^-1 and low level jet with wind speed greater than 12 m·s^-1 are found at 200 hPa and 850 hPa, respectively. At surface, a tongue of high θ_se value prevails in North China. At 500 hPa, there is a convergent zone of non geostrophic wet Q vector extending from southwest to northeast, which is caused by the large scale force. The convective cloud bands have a good relationship with the convergent center of Q vector. In the convergence zone, a number of MCSs continuously move to North China alone the southwest wind on the northwest side of the Subtropical High and cause amount of rainfall. It is called "train effect". Furthermore, topography influences the location of heavy rain. East wind prevails at surface over Beijing area when rainfall occurs. The west mountain blocks and lifts the east flow and increases the precipitation on upstream side of the mountain. The maximum precipitation centers occur at Changping, Yanqing.

Ensemble Prediction Experiments of Tropical Cyclone Track

Wang Chenxi, Liang Xudong

2007, 18(5): 586-593.

[FullText HTML] (742) [PDF 1659KB](1715)

Abstract:
In order to find the ensemble prediction method that can be used to forecast the tracks of tropical cyclones in the Western North Pacific Ocean, 52 cases tropical cyclone track ensemble prediction experiments are made, the 52 cases are from the 8 tropical cyclones that made landfall at China in 2005.MM5 model is used as the experiment model. The model has a horizontal grid spacing of 45 km with 115×115 points and 23 vertical sigma layers. It is run for 72 h. The model domain center is same as the center of experiment tropical cyclone. There are 12 ensemble members. 12 members consist of 11 perturbation members and a control forecast. Two methods are used to create the perturbation members. One method is breeding of growing modes (BGM), in which two 12-h breeding cycles are carried out. The other method is model physics perturbation (MPP), in which members are created by choosing different physics parameterization schemes. The experiment results show that the ensemble mean of BGM is better than the control forecast and the ensemble mean of MPP is worse than the control forecast. For BGM, the ensemble mean for the tropical cyclones whose initial intensity is smaller than 32.6 m/s is more skillful than the ensemble mean for the tropical cyclones whose initial intensity is larger than 32.6 m/s. For MPP, the ensemble mean for the tropical cyclones whose initial intensity is larger than 32.6 m/s is better than the control forecast. According to the different results of BGM ensemble and MPP ensemble, a new method is used to create the perturbation members. In this new method, when the initial intensity of tropical cyclone is smaller than 32.6 m/s, BGM method is used to create the perturbation members, when the initial intensity of tropical cyclone is larger than 32.6 m/s, both BGM method and MPP method are used to create the perturbation members. The ensemble mean of this method is better than the ensemble mean of BGM or MPP. The spreads of BGM ensemble, MPP ensemble and the new method are all too small.

Impacts of the Cloud Initialization on T213 Forecasting Performances

Guan Chenggong, Chen Qiying, Wang Juan, Tong Hua

2007, 18(5): 594-600.

[FullText HTML] (723) [PDF 1824KB](1320)

Abstract:
Cloud variables such as cloud water, cloud ice and cloud cover are not treated in most operational data objective analyses and initialization schemes. While cloud variations are used as prognostic variables in a forecast model, they are necessary to be defined at initial time. The common method is to set cloud variations to zero at the initial time in the initial field, and some time of several hours are needed by the forecast model in doing spin up at the beginning of the run, which is sure to affect the forecasting capabilities such as precipitation and patterns and so on. The cloud free variation scheme is adopted to realize the adding of cloud variations information to the model initial fields of T213L31, the operational global model, then both a statistic analysis of a 3 month continuous running of test and operation schemes and a case study are carried out. In order to testify that the cloud free variation scheme does not foil the balance between dynamic and thermal variances at the start of model forecast, the stabilities of the model are checked and confirmed firstly by means of analyses on kinetic energy, temperature, and so on by long time integration. A synoptic case study, which happens on June 18—19, 2005, shows that with cloud information including cloud water, cloud ice and cloud cover being added to the initial model field, the spin up phenomenon disappears which always occurs at the starting period of forecasting, and often exists for 12 to 18 hours or longer, before the cloud free scheme being added in the modeling system. And the characteristics of distribution and variation of cloud related variables can be more reasonably depicted by the model than the operational one so that the improvement of the model forecasting performance is led to especially in the short time precipitation prediction. The 3 month continuous parallel experiment in winter and summer respectively are done together with corresponding operational ones. The results based on summer 3 month statistic analyses show that positive contributions to both precipitation and geopotential height pattern are made by adding cloud variations to initial field, that is obvious in the improvement of 24 hours accumulating precipitation's TS from 36 to 108 hours lead time, and the bias of precipitation increases a little at the same time. The 3 month averaged anomaly correlation coefficients and root mean square error at 500　hPa is better than operational one at 1—4 lead days. The results based on winter experiment show that the advantages of threat score over operational one within 60 hours lead time, and the bias of precipitation are identically better than operational one at each lead time, and the 3 month averaged anomaly correlation coefficients and root mean square error at 500 hPa are better than operational one. Positive impacts on operational model forecasting can be made by reasonably initializing cloud variations in the model field. This research is based on the operational global medium range weather forecast model, which confirms that the cloud free variation scheme is reliable and feasible to the operational application in the near future. In the long term, how much contribution will be made to the operation forecasting need to be further evaluated. Cloud variables such as cloud water, cloud ice and cloud cover are not treated in most operational data objective analyses and initialization schemes. While cloud variations are used as prognostic variables in a forecast model, they are necessary to be defined at initial time. The common method is to set cloud variations to zero at the initial time in the initial field, and some time of several hours are needed by the forecast model in doing spin up at the beginning of the run, which is sure to affect the forecasting capabilities such as precipitation and patterns and so on. The cloud free variation scheme is adopted to realize the adding of cloud variations information to the model initial fields of T213L31, the operational global model, then both a statistic analysis of a 3 month continuous running of test and operation schemes and a case study are arried out. In order to testify that the cloud free variation scheme does not foil the balance between dynamic and thermal variances at the start of model forecast, the stabilities of the model are checked and confirmed firstly by means of analyses on kinetic energy, temperature, and so on by long time integration. A synoptic case study, which happens on June 18—19, 2005, shows that with cloud information including cloud water, cloud ice and cloud cover being added to the initial model field, the spin up phenomenon disappears which always occurs at the starting period of forecasting, and often exists for 12 to 18 hours or longer, before the cloud free scheme being added in the modeling system. And the characteristics of distribution and variation of cloud related variables can be more reasonably depicted by the model than the operational one so that the improvement of the model forecasting performance is led to especially in the short time precipitation prediction. The 3 month continuous parallel experiment in winter and summer respectively are done together with corresponding operational ones. The results based on summer 3 month statistic analyses show that positive contributions to both precipitation and geopotential height pattern are made by adding cloud variations to initial field, that is obvious in the improvement of 24 hours accumulating precipitation's TS from 36 to 108 hours lead time, and the bias of precipitation increases a little at the same time. The 3 month averaged anomaly correlation coefficients and root mean square error at 500　hPa is better than operational one at 1—4 lead days. The results based on winter experiment show that the advantages of threat score over operational one within 60 hours lead time, and the bias of precipitation are identically better than operational one at each lead time, and the 3 month averaged anomaly correlation coefficients and root mean square error at 500 hPa are better than operational one. Positive impacts on operational model forecasting can be made by reasonably initializing cloud variations in the model field. This research is based on the operational global medium range weather forecast model, which confirms that the cloud free variation scheme is reliable and feasible to the operational application in the near future. In the long term, how much contribution will be made to the operation forecasting need to be further evaluated.

Spatial and Temporal Structures of Relationship Between Seasonal Mean Temperature and Rainfall in China

Zhou Xiaoxia, Wang Panxing, Duan Mingkeng, Lin Kaiping

2007, 18(5): 601-609.

[FullText HTML] (744) [PDF 2881KB](1314)

Abstract:
The interdecadal and interannual variation components of seasonal mean temperature and rainfall are extracted by applying spectrum analysis on monthly mean temperature and rainfall data of 160 stations during 1955—1998 in China. Evidence shows that the variances of interannual components are much larger than those of interdecadal parts, but it is the opposite while the variances are divided by their freedom degrees respectively, especially for temperature. In winter, most of the significant interdecadal variation areas lie in the northern half of China including Northeast China, North China and northeast part of Xinjiang. While in summer, the significant interdecadal variation areas locate in South China, Central China and Northwest China. As for the interannual variation of seasonal temperature, there are no significant areas both in winter and summer. In contrast with the seasonal temperature, the significant interdecadal and interannual areas of rainfall are fairly small and disconnected, among which two small patches of significant interannual variation in Shandong and Southeast China in winter are noticeable.It is implied in correlation analysis that the negative correlation between seasonal mean temperature and rainfall is remarkable in summer particularly over east of 105°E and south of 35°N of the main land, and it can be inferred that both anomalous hot dry and cool wet summer have more likelihood in most part of this region. For the purpose of verifying this point, observational station data are analyzed. The results present that 30 out of 44 years are either hot dry or cool wet summer in Yangtze Huaihe River Valley and South China, 21 out of 44 years in North China.Singular vector decomposing (SVD) are conducted to reveal the spatial and temporal features of the relationship between the two elements. It is indicated by the SVD modes that the two summer patterns are statistically significant in Yangtze Huaihe River Valley and South China. For years, both interdecadal and interannual time coefficients are negative, the rate of hot dry summer in Yangtze Huaihe River Valley is 6/8, and the rate of cool wet summer in South China is 5/8, while the reverse is true when time coefficients are positive, 6/7 cool wet summer in Yangtze Huaihe River Valley, and 4/7 hot dry summer in South China. The summer patterns occur out of phase in the two regions during the same year which is contributed by the negative correlation of summer mean temperature and rainfall on both interdecadal and interannual time scales.

Decade Variations of Precipitation Event Frequency, Intensity and Duration in the Northeast China

Sun Fenghua, Yang Suying, Ren Guoyu

2007, 18(5): 610-618.

[FullText HTML] (970) [PDF 2543KB](1446)

Abstract:
Against the background of global warming, study on climate extremes has become more important, especially the extreme precipitation events in Northeast China which is one of the most remarkable warming areas in China. Daily rainfall data of 93 weather stations in Northeast China from 1951 to 2002 are used to analyze the temporal and spatial variation of precipitation events, including rainstorm, heavy rain, light rain, extreme dry spell, extreme wetness spell etc. The spatial and temporal characteristics of precipitation events change are studied. The main conclusions are summarized as follows. The number of days both of the total rain events and the light rain decreases. The contribution of light rain to annual precipitation is obviously increasing, the contribution of mediummagnitude rain is slightly decreasing, and the contributions of heavy rain remains unchanged. The annual rain day has a significant decreasing trend, which is mostly due to the decreasing of light rain day. The intensity of annual precipitation shows a significant increasing trend due to the increasing of the intensity of light rain and rainstorm. The events of light rain are more frequent before the middle of 1980s, the events of mediummagnitude rain are more frequent after the middle of 1980s, and the rainstorm events have an obviously positive trend after the middle of 1990s. In the significant warming period of 1991—2000, the total days with rain events have an obvious decreasing trend, but the rainstorm day hasn't an insignificant change, though the intensity of rainstorm is building up in the analyzed period. Since 1980s, the climatic variation range of precipitation (57 mm) has also an obvious increasing trend with the global warming. The value of precipitation variation range (77 mm) in the significant warming period of 1991—2000 reaches the biggest since 1960s, which is about one and a half times of other period of 1960s of 52 mm and 1970s of 41 mm. The long dry spells (there is no rain for 10 days or longer) are with a significant increasing trend. The long wetness spells(there is rain for 6 days or longer)are with a significant decreasing trend. The long dry spells are highly related to drought. Against the background of a little change in the total precipitation amount, the distribution of precipitation has become more asymmetric. The rain events have an obvious trend of extremeness. In a word, the extreme trend of precipitation is a reality during the last half of the 20th century in Northeast China with the remarkable warming. The extremity brings drought and waterlog which is likely to become more severe due to the change trend. Adverse influence on environment, especially agriculture production, will be brought.

The Influence of Drought and Waterlogging Disasters on Crop Yields in Anhui Province

Zhang Aimin, Ma Xiaoqun, Yang Taiming, Sheng Shaoxue, Huang Yong

2007, 18(5): 619-626.

[FullText HTML] (843) [PDF 1178KB](2360)

Abstract:
Anhui Province is located in the north south climate transition zone. Meteorological disasters occur frequently. Particularly, drought, flood and waterlogging are verYharmful to agricultural production. According to the terrain, climate and agriculture characteristics, focusing on key meteorological disasters in drought, flood and waterlogging, and the main crops of wheat and rice, by use of agronomic and disaster information of Anhui Province since the founding of the PRC and weather data since the establishment of weather stations of the whole province and other data, drought and flood disasters identify criteria of climate are established by applying statistical climate analysis, the distribution of drought and flood disasters are analyzed, the impact of drought and flood disasters on wheat and rice is quantitatively analyzed, drought and flood disaster damage assessment models and index and achieve real time monitoring of Anhui drought and flood disaster assessment are established.The drought and flood levels identify criteria of Anhui Province are established and divided into 10 levels each by use of Z index. Crop loss rate (Y_D) shows the difference rate between the actual yield and tendency yield. By analyzing typical drought and flood years of rice(single season rice)and wheat, it is found that drought and flood occurring at different times have different impact on yield. Considering the meteorological conditions and drought and flood extent during the whole growth of wheat and rice from planting to harvest, and taking into account the critical periods of different regions, disaster stricken meteorology index and standard of Z index classification of drought and flood disasters, the degree of drought and flood disasters are classified. Since spring waterlogging is the major meteorological disaster which influences winter wheat production in Anhui Province, therefore evaluating wheat yield loss by the spring waterlogging is studied. The dankness index Q proposed bYhuang Yuhua is modified in the article. New waterlogging index Q_w which takes in to account the precipitation, rainfall days and sunshine time is proposed by which the characteristics of flood and waterlogging's damage can be better reflected. Spring flood and waterlogging are the major disasters which mainly affect winter wheat yield. TheYDo more harm to wheat yield than drought, particularly the flood and waterlogging occurring in April and MaYhave the worst impact on yield which can cause heavy losses up to 40%. The influence on crop yield by flood and waterlogging disaster is not only related to precipitation but also to rainfall duration and distribution. Precipitation and flood and waterlogging index Q_wz is the yield loss quantitative criterion for the analysis of the typical years. Crop loss rate is not only closely related to the crop sensitivity to the disasters, but also to the regional vulnerability. The crop loss rate is different due to the difference of sensitivity and adaptability of crops planted under the same disasters. In order to consider the influence of regional difference on disaster assessment, the trend relative frangibility factor method is proposed. Using this method the assessment error which is induced by the difference of regional disaster resisting ability can be reduced. The winter wheat loss rate from 1961 to 2000 has been revised and tested for the regional disaster vulnerability factor, and some specific cases have been analyzed. After correcting regional vulnerability, fitting errors in average and diversity variations have been remarkably reduced.

Microclimate Inside Sunlight Greenhouse in Semi arid Rain Feed Region in Loess Plateau

Zhao Hong, Zhao Qiang, Yang Qiguo, Deng Zhenyong, Wang Runyuan, Ma Pengli

2007, 18(5): 627-634.

[FullText HTML] (767) [PDF 1856KB](1407)

Abstract:
Monitoring and analysis on microclimate inside sunlight greenhouse are carried out in semi arid rain feed region which lies in Loess Plateau. The results show that during the whole growing period of autumn planting cucumber, the daily mean air temperature maintains between 12—23.5 ℃ with falling fluctuately, which is very in accordance with the cucumber growth demand of high temperature in the prophase and low temperature in the anaphase. The ground temperature also falls wavily in general. Relative humidity, which keeps between 52.4%—93.4%, increases before the initial gathering period, but declines wavily after widespread gathering period. These three factors all change along with the variation of height and horizontal position in the greenhouse. In different growing period of autumn planting cucumber, the trend and fluctuation of daily variation of climate factors are basically similar, but variation degree is different and the time that peak value appears in a day is different too. In the upright direction inside the greenhouse, the air temperature and relative humidity (RH) are distinct at different height (0.5 m, 1.0 m and 1.5 m above the ground respectively). The air temperature increases gradually from 0.5 m to 1.5 m. In other words, air temperature inversion phenomenon occurs. However, RH is the opposite, which is smaller in the upper level than that in the lower level. In this way, there is a microenvironment of low temperature and high humidity at 0.5 m in comparison with 1.5 m, where there is a microenvironment of high temperature and low humidity. The daily variation of air temperature and RH is sharp relatively. In addition, ground temperature and its daily variation are obviously different in different depths (10 cm, 30 cm and 50 cm under the ground respectively). The variation of ground temperature at 10 cm level is most sensitive and its variation degree is the biggest among 10 cm, 30 cm and 50 cm. It has less variation degree on the ground temperature at 30 cm under the ground. In comparison with 10 cm and 30 cm, it has little variation in a day of the ground temperature at 50 cm level, where it is in a state of low temperature all the time. There is a clear lag effect in ground temperature when comparing with air temperature. The appearing time of maximum temperature of 10 cm depth soil is delayed about 5 hours than air temperature, and the delaying degree is more than that of the air temperature outside the greenhouse and the observation data in the weather station. A longer delaying time occurs at the 30 cm depth ground temperature, which reaches more than 7 hours.In horizontal direction inside the greenhouse, the air temperature and ground temperature in the south are higher than that in the north. RH is the reverse, that is to say, RH is lower in the south than that in the north. But these differences are very small, it is regarded approximately that the distribution of temperature and RH in horizontal direction is relatively evenly distributed. It is fit to the planting of vegetation.

Seasonal Physical and Chemical Features Variation of Ambient Aerosol in Lin'an

Zhang Yangmei, Yan Peng, Yang Dongzhen, Wang Shufeng, Tang Jie, Yu Xiangming, Ma Qianli

2007, 18(5): 635-644.

[FullText HTML] (776) [PDF 1767KB](2079)

Abstract:
The seasonally chemical and microphysical characteristics of aerosols in Lin'an regional atmospheric pollution monitoring station is investigated. The aerosol is collected on both Teflon filters and Quartz filters at the same time which are analyzed by different methods for different purpose. The Teflon filters are prepared for analyzing the water soluble ion concentration, while the quartz filters are used for analyzing the organic carbon and element carbon. Seven sampling periods represent different seasons. The details of sampling information are listed below: from March 30, 2002 to April 8, 2002 for spring; from August 14 to 24, in 2002, from July 20 to 30, in 2003, and from August 17 to 29, in 2004 for summer; from November 7 to 23, in 2003, from November 6 to 20, in 2004 for fall, and from January 15 to February 2, 2005 for winter. The ion chromatogram analyzer is used to discuss water soluble ion concentration of aerosol. A sunset carbon analyzer is applied to measure the organic carbon and element carbon in aerosol. Not only the seasonal mass, ion and carbon concentration of ambient aerosol are studied in Lin'an, but also the size distribution characteristic of aerosol components is analyzed in detail. Generally speaking, there are obvious variations among different seasons of mass, ion and carbon concentration. For the whole size range, the mass concentration in spring is the highest with 534 μg/m³, and it is 117.21 μg/m³ in winter which is just lower than that in spring. The lowest concentration occurs in summer with 65.7 μg/m³, and 98.6 μg/m³ in fall. The ions occupy different percentage of mass in different season. In summer, the ions' concentration is 49.4% of mass concentration, and it is 11.3% in spring. At the same time, the total concentration of sulfate, ammonium and nitrate is 75%—83% of the whole ions' concentration. On the other hand, nitrate concentration varies with season greatly. The average concentration of it is only 1.7 μg/m³ in summer. The characteristic of carbon shows that the highest element carbon concentration appears in spring and the lowest appears in summer. For organic carbon, spring's concentration is the highest and winter's concentration is the lowest. Moreover, the characteristic of size distribution is also obvious. The PM₁₁ reaches 90% of the whole mass concentration. PM_2.1 also occupies 53% of the whole mass. Sulfate, ammonium and nitrate are the main ions in fine particles. The characteristic of carbon size distribution shows that the smaller the particles are, the higher the concentration reaches.

Componential Characteristics and Sources Identification of PM_2.5 in Beijing

Xu Jing, Ding Guoan, Yan Peng, Wang Shufeng, Meng Zhaoyang, Zhang Yangmei, Liu Yuche, Zhang Xiaoling, Xu Xiangde

2007, 18(5): 645-654.

[FullText HTML] (3117) [PDF 2010KB](4404)

Abstract:
Statistic analysis is made of the characters of the mass concentrations and chemical compositions of PM_2.5 in Beijing, based on the observations during the period of 2003 to 2004. It is found that the mean concentration of PM_2.5 shows the lowest value in summer, while it reaches the maximum of year in winter and spring. Moreover, the daily average mass concentration of PM_2.5 in summer is 71 μg/m³, which is lower than that in other seasons of about 110 μg/m³. The yearly average mass concentration of PM_2.5 is 100 μg/m³, which is much higher than the U S Air Quality Standard for yearly average mass concentration of PM_2.5 of 15 μg/m³. Then, the relationship of PM_2.5 and PM₁₀ is discussed. The mean ratio of PM_2.5 to PM₁₀ is 0.55 for the whole year, which is close to the values in previous research for other cities including Guangzhou, Wuhan, Chongqing and Lanzhou in China. In addition, the mean ratio of PM_2.5 to PM₁₀ is 0.62 and 0.52 for the heating and non heating season respectively. It shows a slightly high trend in the heating period. Seasonal characteristics of the ratio of PM_2.5 to PM₁₀ is 0.3—0.6 in summer, 0.3—0.8 in spring and autumn, and 0.4—0.9 in winter. The results indicate that diurnal ratio of PM_2.5 to PM₁₀ responds to the meteorological conditions and the anthropogenic activities. Sand dust weather and daily traffic lead to the increase in concentration of coarse particles more rapidly than that of fine particles in atmosphere. As a result, the ratio of PM_2.5 to PM₁₀ decreases. On the other hand, the house heating in winter and the photochemical reaction in summer cause the increase in the ratio of PM_2.5 to PM₁₀ as well. Furthermore, the analysis to meteorological factors reveals that the change of the concentration of PM_2.5 is well related to pressure, relative humidity, and wind speed. In addition to the positive correlation with humidity, the mass concentration of PM_2.5 is negatively correlated with wind and pressure except in summer. Finally, the sources of PM_2.5 are analyzed by using the method of positive matrix factorization. It is found that SO₄^2-, NO₃^- and NH4+ are the primary water soluble ions in PM_2.5 in Beijing. Moreover, five sources of PM_2.5 in Beijing are identified. They are soil dust, coal combustion, traffic, sea salt aerosol and steel production. Compared with the results of previous research in sources identification of aerosol in Beijing, some conclusions are made. Firstly, soil dust and coal combustion have been the primary sources since 1980 s, while the contribution of traffic emission to fine particles has grown gradually from 1983 to 2001. Secondly, the effects of a number of sources including coal combustion, sea salt aerosol, biofuel combustion and second aerosol, on PM_2.5 vary with seasons. Thirdly, the process of transportation affects the composition of PM_2.5 distinctly, and the characteristic element of PM_2.5 from special sources related with the wind direction closely. Besides, the results of sources identification to PM_2.5 in Beijing area are different in different sites and periods. As a result, selecting representative sites and observing period for the study is very important.

The Meso β-Scale Convective System of a Heavy Rain Event on July 10, 2004 in Beijing

He Lifu, Chen Tao, Zhou Qingliang, Li Zechun

2007, 18(5): 655-665.

[FullText HTML] (768) [PDF 3963KB](1891)

Abstract:
An analysis on the meso β-scale convective system of a heavy rain event on July 10, 2004 in Beijing is performed by using the special observational data, including automatic meteorological stations data, radar images and satellite images and NCEP/NCAR reanalysis data on the basis of successful simulation. It is found that the heavy rainfall process is generated by a meso β-convective system which is produced in a large scale warm area. The short wave trough in the mid level of troposphere, the convergence between the southwest air current from west wind trough and the southeast air current from the north part of warm shear line in low troposphere provide a good background condition. The meso β-scale convective system is formed by the mergence of two meso-scale convective clusters, it shows an ellipse shape structure of a horizontal scale of 150 km×100 km and the time scale of about 5 hours. It shows the features of the meso-scale convergence line(or convergence center)in low levels stream fields during its occurrence and development stage, and the strong meso-scale convective clouds echo band and meso-scale convergence line exhibited in radar reflectivity image and in radar velocity fields are often related with the occurrence and the development of the meso β-scale convective system. During the stage of strong development, the meso β-scale convective system shows strong baroclinity perpendicular features and has a similar structure of slantwise updraft current to convective storm. Its occurrence and development are forced by the meso-scale convergence line of low troposphere in strong convective instability condition and a warm tongue below 700 hPa. The convergence between the southern air current and eastern air current and the invading of the cold air in the boundary layer lead to the strengthen of the energy front, which is helpful to induce the generation of the meso β-scale convective system.In addition, the cloud top infrared brightness temperature(TBB)of the meso β-scale convective system that induces the heavy rain on July 10, 2004 in Beijing is at-45 ℃ or so, and the updraft airflow reaches the height of around 300 hPa, which means in this case the convection is only activating in the low level of the troposphere in contrast with the deep convective systems of meso β-scale convective system in the mid and lower reaches of Yangtze River and South China, which symbolize with infrared brightness temperature between-70 and-85 ℃, and the updraft airflow reaches the top of troposphere. Future research is needed on whether this conclusion is characteristic for popular convective heavy rain process in North China.

Wet Q Vector Interpretation Technique with Its Application to Quantitative Precipitation Forecast

Yue Caijun, Shou Yixuan, Shou Shaowen, Zeng Gang, Wang Yongqing

2007, 18(5): 666-675.

[FullText HTML] (747) [PDF 2112KB](1453)

Abstract:
A new kind of wet Q vector interpretation technique is developed for the first time.In the method, the vertical motion ω ₁ can be obtained by solving omega equation whose forcing term is dry ageostrophic Q vector divergence based on iterative method, whereby the wet Q vector divergence is calculated. Then vertical motion ω ₂ can be obtained by solving omega equation whose forcing term is wet Q vector divergence based on iterative method. Finally, precipitable water is calculated on the basis of ω ₂ and vapor, whereby wet Q vector interpretation precipitation is produced. Using a typical Changjiang-Huaihe Meiyu front heavy rainfall, the analy tic results show that the wet Q vector interpretation precipitation field has a certain ability to reflect synchronous actual rain in the context of horizontal distribution characteristic and the extreme intensity, which manifests that wet Q vector interpretation technique is feasible and rational to some extent on the basis of practical application. The technique is applied to Eastern China regional numerical prediction model (which is based on MM5 V3.6, and hereafter termed as MM5) product, whereby wet Q vector interpretation quantitative precipitation fo recast (QPF) field is obtained, which is independent of QPF field output by the MM5 itself in such a way that it has the same spatial and temporal resolutions as the latter.In a Meiyu rainfall process and a landfall typhoon rain process occurring in eastern China during June to August 2004 and combining real rain data, the abilities of the wet Q vector interpretation QPF and MM5 QPF to reflect synchronous surface precipitation are compared and analyzed.The results indicate that the reflecting abilities of the former to fair weather or rain and the rain with intensity over 10 mm/24 h are all superior to that of the latter. Furthermore, the results of forecast statistical verification show that the test scores (TSs) and forecast accuracy of the wet Q vector interpretation fo recast are obviously higher than the counterparts of the MM5 in the context of fair weather or rain, light rain, and the rain with intensity over 10 mm/24 h, on average by 20%, 40% and 60% respectively for TSs and by 6%, 3% 11% respectively for accuracy. Meanw hile, the false-alarm and miss rates of wet Q vector interpretation forecast are evidently lower than those of the MM5, which all manifest sufficiently that the application of the wet Q vector interpretation technique to QPF research is effective. At last, the dependence of the numerical predictionproduct interpretation technique on the performance of the numerical prediction model is discussed, with further modifying directions to the wet Q vector interpretation technique.That is to say, it is necessary to take into consideration the roles of orographic lifting and surface friction, at the same time, the revised wet Q vector (QM) consisting of convective vapor condensational potential heating besides synoptic scale stable vapor condensational potential heating should be considered. Additionally, the wet Q vector interpretation technique introduced in this paper is not limited to apply to Eastern China regional numerical prediction product, it also has the interpretation ability to any model prediction product as long as temperature, wind and specific humidity at conventional layers are included. It has wide application prospects.

Support Vector Data Description in Rainstorm Prediction of the Northwest China

Yan Dongwei, Sun Tianwen, Yang Yan, Fang Jiangang, Liu Zhijing

2007, 18(5): 676-681.

[FullText HTML] (691) [PDF 903KB](1977)

Abstract:
The expert system (ES) has been studied and applied in meteorological field widely. ES depends on know ledge engineers to enter knowledge used in inferring by computer, which is toilsome and error-prone work. As another branch of artificial intelligence (AI), machine learning aims at solving the know ledge obtaining problem automatically and paving a path to remedy the shortcoming of ES. But machine learning still does not work well if it is not tailored to fit characteristics of weather foresting, among which imbalanced class is an important problem deserving study.Although it is usually assumed implicitly by the machine learning research community that the classes are well-balanced, there exist many domains for which one class is represented by a large number of examples while the other is represented by only a few, and there are many applications demanding to classify im portant but rare positive examples (minority). It is a typical example of learning from imbalanced training set to predict such disaster weathers as hail and rainstorm in meteorology. Though they are small probability events, those disastrous weathers will bring about serious destruction. Thus disastrous weathers' prediction has been paid much more attention by meteorologist than normal weather prediction. Normally, the number of examples belonging to normal weather is much more than disaster ones. Aiming at improving the accuracy, trivial classifier that labels every example with majority when faced with imbalanced class distribution will be lead to by traditional machine learning algorithms.By doing so, high accuracy would be obtained.Imbalanced class is a stumbling block stymieing practical attempts to apply machine learning to realistic problem.In order to find algorithms being resistant to imbalanced class distribution, threat score (TS) is used as criterion to evaluate classifiers.As a kernel method, SVM fails to deal with imbalanced class problem too although based on statistical learning theory, and working well in many applications. SVM will incline to the majority class (corresponding to normal weather), and lose very important disaster weather. Support vector data description (SVDD) is another import kernel method originated from SVM. By employing training examples of target set only, one class method is fit for imbalanced class problem. As one class method, SVDD tries to obtain characteristics of target class, and is resistant to class imbalanced problem.The comparative study of SVDD and SVM is conducted to predict rainstorm in Tongchuan City, Shaanxi Province. The experiment shows that SVM is prone to majority class evidently, and brings about many false negative. When normal weather class is select as target, TS of SVDD' is prior to SVM. The result fits the theory analysis on SVDD and SVM.Results show that SVDD is a better choice than such traditional methods as SVM when dealing with imbalanced class problem, better performance could be obtained if the class with more examples is chosen as target class.

Trial Study on Factors Analysis and Prediction of Landslide Hazard Triggered by Extreme Heavy Rainfall

Wei Li, Chen Shuangxi, Bian Xiaogeng

2007, 18(5): 682-689.

[FullText HTML] (787) [PDF 1755KB](1581)

Abstract:
To meet the service needs of predicting and warning landslides triggered by extreme heavy rainfall, based on mechanism of landslide induced by torrential rainstorm and its predicting theory, by monitoring experiment in eight trial grounds, the impacts of rainfall on underg round water table, pore water pressure, soil stress and landslides stability are studied. The influence of vegetation coverage on landslide is discussed. Combining trial data and statistics method, the precipitation values used to predict and warn the landslides are given.The function of precipitation on landslide is a dynamic process. When raining water is injected into landslide mass, water content and volume weight of rock soil mass can be increased, soil is intenerated, rock soil capacity is enhanced, and soil also becomes lubricant while penetrating into bedrock surface under weathering rock mass or layer cutting with water, which diminishes anti-slide force and causes landslide. Surface landslide in torrential rainfall is caused by short time function of violent precipitation. In the case, landslide mass arrives quickly to saturation situation that leads to ground displacement sharply, as water penetrates promptly and soil saturation degree and pore water pressure are increased, slope substance strength is reduced and it ends with landslide which is the surface landslide mechanism.Critical values of precipitation to forecast landslides are investigated, which provide bases to predict and warn landslide disasters. Eight observation sites in sensitive areas in Jiangxi Province are set to monitoring changes of water table, pore water pressure, stress of gliding zone, and landslide mass movement. Based on observational data and historic records, the purpose is to study models of landslides stability and search criteria forpredicting or warning landslide disasters. By monitoring and experiment, it is proved that landslide hazards can be predicted. Critical rainfall value to induce landslide acquired from field trial is in accordance with statistics results. Changes of water table consistent with precipitation are revealed. Water table has negative correlation with landslide stability, but is dependent on precipitation. The fluctuation of water level lags behind the rainfall wave with half to 1 day, the effect of each precipitation process on water table is different from landslide to landslide, some affecting duration lasts 5 days or less, others 15 days.Near the ground surface, water table changes strongly depending on precipitation.Drop speed of underground water varies with water table and rainfall amount. Changing features of pore water pressure in different depths of landslide body are discussed. Pore water pressure is a useful index to represent saturation degree of landslide mass. Research shows that the increase of rainfall and its intensity vary with the decreasing of landslide stability. Pore water pressure declines gradually from March to August till a lowest value and then increases again. Burying depth is in 6 to 7 meters, descending trend becomes weaker, which is related to evaportraspiration. Variations of landslide mass stress having a consistency with landslide movement are studied. Conditions of slope creeping can be well reflected by stress monitoring, indicating the probability of the mass occurring. It is proved that stress-monitoring data have more sensitivity to indicate the creeping of landslide than that of pore water pressure.Effects of vegetation coverage on landslides are derived from the intensity of precipitation. Whether more or less the degrees of vegetation cover and biology amount, or in mountainous areas, landslide disasters would occur, which are determined by the intensity of precipitation. Forest vegetation plays a role in intercepting precipitation, which can relieve water erosion. When rainfall intensity is 50 mm or more, interception amount decreases sharply. In the same geological environment, while precipitation intensity reaches a threat value that can lead to landslide, events of landslide hazard will be sharpened by vegetation weight adding to soil mass.The threat rainfall value to predict landslides in dense forest area is lager than that in sparse zone.

Improvement and Application Test of TREC Algorithm for Convective Storm Nowcast

Chen Mingxuan, Wang Yingchun, Yu Xiaoding

2007, 18(5): 690-701.

[FullText HTML] (834) [PDF 4583KB](1982)

Abstract:
At present, cross-correlation extrapolation is one of the main algorithms for convective storm now cast. Motion vectors of convective storm for every divided equal-sized two-dimensional arrays of radar echo or other data measured at two times several minutes apart by calculating optimal spatial cross-correlation are obtained in the algorithm. The obtained motion vectors are customarily called TREC (tracking radar echoes by correlation) vectors or TREC winds. And then, storm now cast can be achieved by extrapolating radar echoes or other data based on the obtained TREC vectors. The algorithm results involve not only the changing characteristics of magnitude and direction of the motion vectors, but also shape varieties of the whole echoes in the course of their movement. So the result of storm now cast based on the algorithm is assuredly reasonable and significant in meteorology. The basic principle of TREC is introduced firstly. And a number of methods to improve the algorithm result are presented, including noisy vector restriction and clutter contamination removal, discarded or missing motion vector supplement, vector smoothing, and so on. Analysis results of two cases indicate that the tracked motion vectors can be markedly improved after quality control and optimization processes to TREC algorithm. Finally, based on the optimized TREC algorithm and Tianjin radar data, storm nowcast tests and verifications of four intense convective storms that occur in Beijing-Tianjin-Hebei areas during 2004 and 2005 summertime, including two squall line cases, a hailstorm case and a strong thunderstorm case, are described in detail. The results indicate that the improved algorithm is available for convective storm nowcast.The algorithm can automatically produce 30-minute or 60-minute forecast of location and shape of radar echoes or storm characteristics based on the extrapolated vectors. The forecast results are close to what is actually happening. Because the algorithm can automatically produce forecast results in real time mode, it is helpful for convective storm now cast and warning.The forecast results are also clear at a glance, so abilities of forecasters for strong convective storm identification and forecast can be enhanced by the algorithm. An expectation is that the improved algorithm can be used for operational storm now cast in the near future.

Regional Distribution of Surface Heat and Radiation Balance Components over South Ningxia Using Remote Sensing Technique

Guo Jianmao, Lu Weisong, Min Wenbin, Liu Wenquan, Wang Lianxi

2007, 18(5): 702-708.

[FullText HTML] (874) [PDF 1832KB](1814)

Abstract:
The regional land surface heat and radiation balance components are very important and not easy to deal with. To study the components of surface radiation balance and heat balance over inhomogeneous landscape, the utilization of satellite remote sensing is indispensable. In this study, a parameterization method based on Landsat-7 ETM+ data and 20 weather stations data is described to obtain the regional distributions of the components of surface radiation balance and surface heat balance over the south Ningxia area. The south Ningxia area is divided into five surface types:water surface, naked surface, half-naked surface area, grass area and forest area. The regional distributions of the components of surface radiation balance and surface heat balance are calculated and discussed according to each type. Further more, each distribution map and straight-bar figure of the components of surface radiation balance and surface heat balance is given. The results indicate that all the regional distributions are characterized by their terrain nature and the regional distributions are obvious. The figures of the mountains and rivers are very clear, it is because there is a great deal of vegetation over the mountains and rivers edge. It is seen that the derived regional distributions of surface radiation balance and surface heat balance components for the whole mesoscale area are in good accordance with the land surface status. The surface absorbed shortwave radiation is high over Liupan Mountain and Guanmen Mountain, and Yueliang Mountain and the rivers edge is high too, the minimum is in the naked area. In clear day, the surface absorbed shortwave radiation is mainly determined by surface reflectivity. The regional distribution of net radiation is similar to the surface absorbed shortwave distribution. The maximum latent heat flux is at Liupan Mountain which is second by Guanmen Mountain, Yueliang Mountain, rivers edge and other irrigated areas, the low latent heat flux areas are over naked areas.The regional distributions of surface sensible heat flux are opposite to latent heat flux.Soil heat flux is a small quantity in the heat balance function.

The Current Application of Meteorological Code Forms and Impacts of Migration on Table Driven Code Forms

Zhao Fang

2007, 18(5): 709-715.

[FullText HTML] (573) [PDF 1042KB](1570)

Abstract:
Meteorological code is composed of TAC (the Traditional Alphanumeric Code form) and TDCFs (the Table Driven Code Forms). In view of the limitation of TAC and the advantage of TDCFs on data representation, the transition from TAC to TDCFs within the next decade is planned by World Meteorological Organization (WMO).The current applications of meteorological code forms in China are given:TAC is still used mainly in the generation, transmission, processing, storage and application of meteorological observation data, while TDCFs are still not widely used.The main points of the transition plan from TAC to TDCFs of WMO are also summarized. The impacts of the transition to TDCFs, which are mainly to the operational systems such as the current observation system, telecommunication system, data processing and storage system of meteorological services of different levels in China, are analyzed. For the observation system, new coding system has to be deployed to encode the TDCFs observation data. For international and national telecommunication system, the ability of the transmission of TDCFs and the ability to convert between TDCFs and TAC have to be built, the capability of dual transmission of TAC and TDCFs data has to be supported. For data processing and storage system, the functions of the decoding and storage of TDCFs data have to be provided.The primary proposals of the implementation of the transition to TDCFs in China are suggested. The national transition plan should be constituted as quickly as possible; the development of transition core software systems should be arranged; the experimental projects of transition should be conducted nationally and internationally; the national training should be organized.

A 15 L Mixing Cloud Chamber for Testing Ice Nuclei

Yang Shaozhong, Lou Xiaofeng, Huang Geng, Feng Daxiong

2007, 18(5): 716-721.

[FullText HTML] (747) [PDF 1692KB](1415)

Abstract:
Small mixing cloud chamber with liters cubage has been used in the observation of the nature ice nuclei and the detection of the nucleating effectiveness of artificial ice nuclei. Practical usage shows that the testing data are different due to various cubage and methods of supplying fog as well as operation procedure which lead to the difficulty to compare the results. A 15-liter mixing cloud chamber is developed to improve the creditability of data and test the ice nuclei effectiveness of "the 37 model silver iodide shell". The cloud chamber is cooled by a F22 refrigeration system.The lowest temperature within cloud chamber is down to-28℃.In order to keep the temperature stable, a jacket tank filling with glycol-water solution is designed at the cloud chamber periphery. In the glycol-water solution, a 1 kW electric heater is installed to adjust the temperature.Outside the jacket tank, a 10 mm thick cold insulation is made using frothing technology to ensure heat insulation. When the cloud chamber runs, the experiment temperatures are obtained by controlling the cooling compressor and the electric heater. In addition, a device is developed to collect ice crystals with two methods so that to extend the detect range of the ice nuclei concentration. Glass slice is used by one of the methods to collect ice crystals and then the ice crystals are counted by a microscope, sugar plate is used by another method to collect ice crystals and then they are counted by naked eye. The two methods can be used alternatively according to the needs in experiment. The former is often used if the ice nuclei effectiveness is high (cloud seeding agents) and the latter can be used at low concentration (natural ice nuclei). Another important improvement is that the super-cooled fog can be supplied to ensure the nucleation full of the ice nuclei in cloud chamber. At first fog with normal temperatures is generated by an ultrasonic atomizer and passes through a low temperature narrow access so as to arrive at supercooled degree. And then it enters the chamber after the temperature of super-cooled fog equals or is under the mid temperature of cloud chamber. Thus the effect of instantaneous high super-saturation is avoided in this technology and the temperature is not disturbed. In order to improve the ice nuclei effectiveness of the cloud seeding agents carried by shell, several composite formulations are tested with the cloud chamber. The results show that the cloud chamber has a better stability and reproducibility than other small mixing cloud chambers. The better performance should be ascribed to the design of supplying super-cooled fog and the improvement of ice crystals collecting method.

A Dynamic Cluster Model Based on Projection Pursuit with Its Application to Climate Zoning

Wang Shunjiu, Li Yueqing

2007, 18(5): 722-726.

[FullText HTML] (598) [PDF 798KB](1519)

Abstract:
Climate zoning analysis is a typical multifactor problem.The difficulty frequently encountered in climate zoning analysis is that there are so many factors and the complex interrelationship among them cannot be analyzed according to only one factor, all the effect factors associated with climate zoning must be taken into consideration. Aiming at the problem mentioned above, a dynamic cluster model based on projection pursuit principle (DCPP), in which dynamic cluster is combined with projection pursuit principle, is developed, and it is used in climate zoning successfully for the first time. Firstly, multifactor cluster problem can be converted into singlefactor (projected characteristic value) cluster problem according to linear projection principle. Secondly, a new projection index based on dynamic cluster rule is constructed in the dynamic cluster model based on projection pursuit principle, which is the clustering basis for the projected characteristic value. Thirdly, genetic algorithm (GA) is applied to optimize the dynamic cluster model based on projection pursuit principle, and the steps of genetic algorithm are introduced in detail. Finally, a case study on climate zoning is used to test the effect of the dynamic cluster model based on projection pursuit principle.The results show that the dynamic cluster model based on projection pursuit principle for climate zoning is reasonable and effective. On the other hand, based on the dynamic cluster model based on projection pursuit principle, the cluster results can be obtained directly according to the characteristic of data set. Since there is no parameter calibration in the dynamic cluster model based on projection pursuit principle, the results are more objective and less subjective.The dynamic cluster model based on projection pursuit principle is a new method and powerful tool in climate zoning. A new approach to the problem of complicated multifactor cluster analysis is provided by the dynamic cluster model based on projection pursuit principle.

Application of Dynamic Extended Forecast Products to Monthly Precipitation Forecast in Guangxi

He Hui, Jin Long, Qin Zhinian, Chen Jian

2007, 18(5): 727-731.

[FullText HTML] (692) [PDF 917KB](1169)

Abstract:
Based on the monthly mean 500 hPa geopotential height fields from NCEP/NCAR reanalysis data during 1958 to 2005 and the products of dynamic extended forecast from China National Climate Center during 2003 to 2005, monthly precipitation in Guangxi is predicted.The eigenvectors with typical spatial distribution patterns for the predicting key areas and time coefficients reflecting their variation trends can be developed to by way of making empirical orthogonal function (EOF) with the 500 hPa geopotential height anomalies over every predicting key areas. The monthly precipitation in Guangxi is predicted by using analog deviation to find out historical samples similar to predictors in the prediction year. Prediction models are tested by independent samples and results show that the models with predictors from products of dynamic extended forecast are superior in prediction ability to those with predictors from higher correlation areas of former 500 hPa geopotential height.

Spatial Interpolation Methods of Daily Precipitation

Gao Ge, Gong Lebing, Zhao Shanshan, Zhang Qiang

2007, 18(5): 732-736.

[FullText HTML] (314) [PDF 2134KB](2917)

Abstract:

High-resolution precipitation field is very useful in the study of precipitation-induced geological hazards. The quality of the precipitation field varies if different interpolation methods are used. In this study inverse distance weighting and ordinary Kriging methods are applied to interpolate daily precipitation of stations over 26°—34°N, 103°—115°E into 1 km×1 km grid.The interpolation results are then compared to indicate which method better represents the spatial pattern and intensity of precipitation. Results show that the two methods perform similarly well by their correlation coefficients of 0.83 and 0.82 between observed and interpolated precipitation in the cross validation test. When daily precipitation is equal to or greater than 10 mm, both methods are less efficient and the correlation is down to 0.66 and 0.67. The correlation coefficient, as an indicator of the interpolation quality, has a clear seasonal trend with a maximum in spring and a minimum in summer. From analysis to the accuracy of interpolation for heavier precipitation, both methods show unsatisfied results and Inverse distance weighting method with more error rate. The accuracy decrease from 80% for precipitation exceeding 10 mm to 65% for precipitation exceeding 25 mm, particular in rainstorm with above 50 mm precipitation, the accuracy is only 50%. The observed precipitation is systematically underestimated by both methods and the interpolation quality gradually decreases for heavy precipitation. Generally, ordinary Kriging is better than inverse distance weighting. But the running time is a disadvantage when the ordinary Kriging method is applied into operational system, inverse distance weighting method may be substituted for it.

Vol.18, NO.5, 2007