Abstract:Large-scale water vapor transportation is an important condition for persistent heavy rainfall.The water vapor transportation and its sources are important factors influencing summer precipitation in China, and they have been highly concerned by the domestic and international meteorological circles.In the summer of 1998, there was a persistent heavy rainfall in the Yangtze River Basin, resulting in a basin-wide large flood, which caused heavy losses to the people's lives and property.This paper has divided the heavy rainfall process over the Yangtze River Basin from 12 June to 27 August in 1998 into four precipitation stages.The variation features of water vapor transportation, main water vapor sources and their contributions in the different stages have been discussed by using the NCEP reanalysis data and the daily precipitation data of 754 weather stations in China, as well as the airflow trajectory model based on Lagrangian method (HYSPLIT_V4.9), which combined with the trajectory clustering method and the air parcel tracing method.The results suggest that the water vapor of the heavy rainfall mainly comes from three regions, namely the Indian Ocean, the Bay of Bengal-South China Sea and the Pacific Ocean.The water vapor of the Indian Ocean flows into the Yangtze River Basin through Southern India Peninsula, the Bay of Bengal and northern Indo-China Peninsula, while the water vapor of the Bay of Bengal-South China Sea region passes Indo-China Peninsula and the north part of South China Sea into the Yangtze River Basin.The water vapor of the Pacific region flows into the Yangtze River Basin along with the southeast airflow on the edge of subtropical anticyclone.Every precipitation stage has its own water vapor trajectory and sources.The transportation of water vapor originating from the Bay of Bengal-South China Sea is most important and accounts for 35% of the total vapor transportation during the first stage, while the water vapor from the India Ocean and the Pacific Ocean accounting for only 21% and 19%, respectively.During the second stage, the Indian Ocean, the Bay of Bengal-South China Sea and the Pacific Ocean are three major contributors, accounting for 32%, 28% and 31% respectively.During the third stage, the water vapor is mainly provided by the Indian Ocean and the Bay of Bengal-South China Sea, accounting for 33% and 41%, respectively, and the water vapor contribution rate from the Pacific Ocean is only 5%.During the fourth stage, the Bay of Bengal-South China Sea is dominant, the same as during the first stage, and accounts for 40% of total moisture transportation, while the water vapor from the India Ocean and the Pacific Ocean only accounting for 23% and 22%, respectively.The adjustment of atmospheric circulation during the heavy rainfall process leads to the differences of water vapor source and its contribution rate.The enhancement of cross-equatorial jet stream in Somalia during the second and third stages causes its water vapor contribution rate significantly higher than that during the first and fourth stages.The intensity change of the South China Sea monsoon results in the differences of water vapor contribution from the Bay of Bengal-South China Sea during the different stages.

Abstract:Dozens of East Asian summer monsoon(EASM)indices have been proposed to represent the main features of anomalous summer monsoon activity based on different physical consideration.In this study, sixteen representative EASM indices were calculated based on the NCEP/NCAR reanalysis data (1961—2010) and detailed comparison has been performed to detect the differences among the EASM indices.We also explored the basic physical features of various EASM indices by comparing their differences in the related anomalous patterns of atmospheric general circulation and summer rainfall.The possible linkages of EASM indices with the predominant modes of EASM were also discussed.
It is found that sixteen EASM indices can be generally classified into two categories (Category I and Category II).Category I mainly exhibits the decadal variation of EASM while Category II is featured by strong interannual variability of EASM.Further analysis shows that:1) Category I mainly reflects a consistent spatial variation of EASM and is closely linked to the first leading mode of EASM acquired by the extended empirical orthogonal function (EEOF) analysis of 850 hPa wind.Positive(negative)index of category shows intensified (weakened) southerly in the whole East Asian area, which is accompanied with anomalously intensified cyclonic (anti-cyclonic) circulation around the Lake Baikal and slightly northward (southward)western Pacific subtropical high (WPSH).There are also corresponding anomalous 500 hPa geopotential height fields exhibiting a "+-+"("-+-") anomalous zonal pattern in the mid-high latitudes, originating from the western Europe, across the West Siberia plain, and extending to northeast Asia.Meanwhile, there is significantly increased (reduced) rainfall in North and Northeast China but reduced (increased) rainfall in the Yangtze River and its southern area, which consists of a typical dipole of monsoon rainfall anomalies.Category I shows better description of summer precipitation anomalies in North and Northeast China.2)Category II of EASMI generally represents an out-of-phase variation of the summer wind between southern and northern China.Positive (negative) index corresponds to weaker (stronger) summer monsoon in southeast China but stronger (weaker) summer monsoon in northeast China, which is accompanied with anomalously intensified cyclonic (anti-cyclonic) circulation over the Northwest Pacific and northward (southward) shifting of WPSH.Correspondingly, 500 hPa geopotential height anomalies exhibit a "-+-"("+-+") meridional tripole pattern in East China, from the western North Pacific (WNP), across Japan to the Okhotsk Sea, which is similar to the typical East Asia-Pacific (EAP) teleconnection pattern.The summer precipitation decreases (increases) in the middle and lower reaches of the Yangtze River, while it increases (decreases) over the east of Northeast China and the southeast coast of China.Category II shows better description of summer precipitation over the Yangtze River basin.

Abstract:Based on the daily precipitation data from 753 stations in China during 1979 and 2010, the spatial scale(23—30°N, 110—120°E) and time range(from pentad 12 to 27) of the Spring Persistent Rains(SPR) are redefined, and the spatial and temporal distribution of SPR are analyzed by EOF(empirical orthogonal function) method.Three main modes, namely, the region consistent type, south-north reverse type and east-west reverse type, are obtained.On this basis, by using the NCEP/NCAR reanalysis dataset, we analyzed the circulation differences between SPR drought and flood years.The results show that in drought years, the southwester over the southeast area of Tibet Plateau strengthens and the western Pacific subtropical high enhances and extends westward, which are conducive to water vapor transport from the south of the plateau and of the subtropical high to the south area of the Yangtze River, and further induce the enhancement of ascending motion and precipitation over the south area of the Yangtze River.Further analysis shows that the heat source intensity over Tibet Plateau in SPR flood years is obviously stronger than that in SPR drought years, which leads to the enhancement of the flow around the southwestern side of the plateau.Therefore, it is conducive to the precipitation in the south area of the Yangtze River.But in drought years, the situation is just the opposite.In addition, by comparing the zonal land-sea thermal differences between East Asia continent and the Western Pacific, it is found that the inversion between cold and heat source occurs at about the 11th pentad in flood years.However, the atmospheric source reverses at the 16th pentad in drought years.The intensity of heat source during SPR over the south area of the Yangtze River in flood years is also stronger than that in drought years, which further explains that the inversion and enhancement of the zonal land-sea thermal contrast play an important role in producing the SPR, and it is also an important indicator in determining the monsoon precipitation characteristics of the SPR.

Abstract:This paper analyzes three long duration heavy precipitation events in Liaoning Province using NCEP 1°×1° grid reanalysis data, FY-2E infrared 0.1°×0.1° TBB data and automatic weather station data.The relationship between TBB and precipitation and the characteristics and maintenance mechanism of mesoscale environmental field are studied.The conceptual forecasting model of long duration heavy precipitation in Liaoning is initially established.The results demonstrate that: When the pattern, location and intensity of subtropical high are conducive to heavy precipitation occurring in Liaoning, a favorable environmental background for the occurrence and maintenance of heavy precipitation is provided by the long-time interaction between the moist warm air transferred by low-level jet on the west side of subtropical high and the arid cold air on upper levels.There must be a continuous and stable water vapor transport before the occurrence of heavy rainfall, and there is no significant change in the intensity of water vapor transport and the thickness of wet layer in the process of heavy rainfall.The time length of strong vertical movement has good correspondence with the duration of heavy rainfall, and it is generally weak unstable stratification or neutral stratification on upper levels during heavy precipitation.The heavy precipitation can occur not only in the inner or edge of the cool cloud area of strong convective cloud but also in the warm cloud area where TBB value is small.TBB value is not necessarily related to precipitation intensity, but the rapid decrease of TBB value indicates imminent heavy rainfall.It is necessary to pay attention to water vapor transport, convergence and upward motion below 900 hPa in order to forecast location and intensity of precipitation in the warm cloud with low center of mass.These conclusions are helpful to deepen the understanding of the cause of long duration heavy precipitation in Liaoning area and provide clues for forecasting.

Abstract:By analyzing the easterly waves that influence Southern China Coast during 1999—2013, the easterly waves can be classified into three types:south path easterly waves, westward easterly waves and offshore easterly.The Pacific ocean high is a key system which controls the typical paths and positions of the three types of easterly waves especially through its motions towards west and north.The composite analysis based on different types of easterly waves shows that the strong vorticity center, which mostly locates under 850 hPa level, denotes the easterly wave center.As for the structures of the easterly waves, there are a strong convergence area in its lower part and a vertical rising area over the whole level on its back side.The trough axis tilts westwards with its height for the south path easterly wave, but it is almost vertical for the westward and offshore easterly waves.The composite analysis also shows that the easterly wave tends to move towards the warmer SST area.The numerical simulation results confirm that by warming up the SST warm center in front of the easterly wave trough, it will cause minus geo-potential height at low level of the easterly wave trough area and minus center of geo-potential height at mid level of its back side.It shows that the warming up of SST would increase the intensity of the easterly wave by sensitive and latent heats.It will also increase the intensity of the easterly wave precipitation, push the rainfall area northward and add more cyclonic component of wind fields in lower part of the easterly wave.The added component will support stronger convergence at low level and maintain the rising vertical velocity over the whole level.The analysis of a typical westward easterly wave shows that the helicity is positively correlated with the intensity of the easterly wave.Potential temperature zonal difference shows that the easterly wave has a structure of warm top and cool bottom.When the vorticity of the easterly wave increases, the perturbation kinetic energy transforms to the layer perturbation potential energy;while the easterly wave weakens, the latter transforms to the former.

Abstract:Based on the TIGGE datasets including the European Centre for Medium-Range Weather Forecasts(ECMWF), the U.S.National Centers for Environmental Prediction(NCEP), the United Kingdom Met Office(UKMO) and its multi-center ensemble systems, Bayesian Model Averaging(BMA) probabilistic forecasts of winter surface air temperature over East Asia are established.Anomaly correlation coefficient(ACC) and root mean square(RMSE) are used for the evaluation of the BMA deterministic forecasts.Furthermore, Brier score(BS), Ranked probability score(RPS), BSS and RPSS are applied to evaluate the performance of BMA probabilistic forecasts.The results show that the BMA forecast distributions are considerably better calibrated than the raw ensemble forecasts, and BMA forecasts of ECMWF, NCEP and UKMO EPSs provide better deterministic forecasts than the individual model forecasts.The BMA models for multi-center EPSs outperform those for single-center EPS for lead times of 240-360 h, and the optimal length of the training period is about 35 days.In addition, BMA provides a more reasonable probability distribution, which depicts the quantitative uncertainty of the forecasts.The uncertainty on the land(higher latitude) is larger than that on the sea(lower latitude).

Abstract:Observed sea surface temperature anomaly(SSTA) in the equatorial eastern Pacific exhibits an interesting evolution characteristic in La Niña life cycle, which is characterized by a weaker decay after its peak and a re-intensification of cold SSTA in the second year.Based on the output data in 19 CMIP5 models, the special evolution features of La Niña events are investigated.The simulation capability of CMIP5 multi-models for La Niña life cycle is evaluated.Evaluation results show that only a few good models can capture the slow decaying and re-intensification processes in La Niña life cycle, while La Niña continues decaying to the neutral state in the rest of pool models.The physical mechanisms that caused the distinctive evolution features are carried out by an oceanic subsurface heat budget analysis.The result shows that the major differences between good models and pool models lie in wind, which induces anomalous zonal advection and mean meridional advection associated with subtropical cell.The first possible mechanism that caused the re-intensification process of La Niña is the wind-forced equatorial wave dynamics.As both the mean upwelling and the mean zonal temperature gradient along the equator reach a maximum in boreal fall, northern autumn is the season of the strongest coupled ENSO instability when the cold tongue is the strongest.With re-intensification of easterly wind anomalies in northern fall, the thermocline shoals again continue towards the end of the year.Thus, it is found that a positive feedback of SST-convection-wind, which makes La Niña re-develop, can be established in the good models, while such a positive feedback does not exist in the poor models.The other possible mechanism is the meridional mean ocean advection process related to the ocean mean subtropical cell.The good models can simulate the intensity of climatological mean subtropical cell well, thus the off-equatorial anomalous cold subsurface water is likely to be advected equatorward by the mean current, and La Niña intensifies again through the meridional mean ocean advection.In contrast, the pool models are weak in simulating the intensity of mean subtropical cell, therefore the equatorial negative ocean temperature anomaly tends to recover to the neutral state in the end of the second year.

Abstract:Data quality is a basic assurance for meteorological researches and data applications.In this paper, considering local climate in Shanghai, a localized quality control flow is formulated from 59 AWSs(Automatic Weather Stations) hourly temperature data during the period of 2006—2013.It includes station format parameter check, climate extreme value check, time consistency check and space consistency check, especially by using the methods of dynamic threshold in climate extreme value check and different distance standards for different areas in spatial consistency check.After quality control, the annual mean data missing rates of this data set are all below 10%, which means the AWS data in Shanghai bears good integrality and confidence.Thus a set of high-quality and high-resolution temperature data set is obtained.Compared with the manual observation data, it is found that the average annual and seasonal mean temperatures in Shanghai are almost the same, which proves that the quality of this data set is reliable.But the spatial difference is more obvious, which just indicates that the high resolution temperature data is more representative and effective in the researches about fine spatial distribution features in urban thermal environment.Based on standardized temperature, the spatial distribution of urban heat island in Shanghai is analyzed with this data set.The results indicate that the AWS data can reflect many fine features of the distribution of urban heat island.The urban heat island center has expanded from the city center to periphery and the southwest area, especially presents a multi-centered feature.In addition to the main center of heat island at the city center, there are two sub-centers in the north of Minhang district and the south of Songjiang district respectively, which are associated with rapid urbanization.Two regional construction projects, namely, "Songjiang New Town" in 2009 and "Big Hongqiao Section" in 2010, have greatly accelerated the process of urbanization.They have not only changed the underlying surface status, but also made a large amount of anthropogenic heat caused by lots of people moving into the sub-centers of the city.Also the heat islands are located in the southeast area in autumn and winter and in the northwest area in spring and summer, which is affected by local sea-land wind and seasonal transition of atmospheric circulation.The above detailed characteristics are not obvious or can not be reflected from manual data.Therefore, the AWS data with quality control is better applicable than manual data in the research of fine structures of urban heat island.

Abstract:By analyzing the aerosol number concentration and the size distribution observed in July 2014 at Mt.Huang, the characteristics of aerosol distribution at Mt.Huang in summer were analyzed.On this basis, log-normal fitting were performed on aerosol size distribution.The results show that the average aerosol number concentration at Mt.Huang is about 3 518.27 cm^-3.The aerosol number concentration is mainly concentrated in Aitken mode.The daily variation of aerosol average concentration showed a bimodal distribution, and the peak concentration was accompanied by the increase of small particles.The aerosol number concentration was negatively correlated with relative humidity and wind speed, and high aerosol number concentration usually occurred under weak southeastern wind condition.The aerosol number concentration in the accumulation mode was obviously influenced by wind direction.With the effect of different air masses, the difference of aerosol number distribution mainly existed in the particle size of less than 100 nm and between 500 and 1 000 nm.Under the southwest air mass with high humidity, there appeared the highest aerosol number concentration and the narrowest size distribution in Aitken mode.However, under the southeast air mass with high temperature and low humidity, there appeared the lowest aerosol number concentration and the widest size distribution.Aerosol number concentration and spectral width corresponding to northern air mass were in medium level.The size distribution characteristics of aerosol ranging from 500 to 1 000 nm were contrary to those in Aitken mode.The aerosol size distribution and volume distribution under different backgrounds can all be fitted to a log-normal distribution by using the three modes of Aitken mode, accumulation mode 1 and accumulation mode 2, but the aerosol number concentration spectral parameters under different air masses were quite different.

Abstract:To further explore tropical cyclones'(TCs) law and climate characteristics of occurrence, development and disappearance, this study applies Finite-Mixed-Model(FMM) which is based on clustering algorithms to cluster landfall tracks of TCs over China during summer and autumn(from June to November).The data of typhoon best-tracks and circulation fields are from the Shanghai Typhoon Institute(STI) of China Meteorological Administration(CMA) and 6-hourly reanalysis data of wind and other fields from NCEP/NCAR for the period 1951—2012 during summer and autumn.The three clusters are analyzed for comparison in terms of TC's active season, frequency and circulation.
The TC tracks landing China are divided into three categories by using FMM algorithm, which corresponds with the traditional classification result.Cluster-1 approximates the recurve-landing type, Cluster-2 approximates to the westward-landing type, and Cluster-3 approximates to the northwest-landing type.The findings are summarized as follows.In summer, Cluster-1 and Cluster-2 appear more frequently than Cluster-3, however, Cluster-3 has the highest frequency of occurrence in autumn.The generate location of Cluster-1 is the most northward, and the intensity is relatively strong.The track of Cluster-1's development shows northward trend, affecting the most extensive region, such as the Yangtze River Delta, the Pearl River Delta, Beijing, Tianjin and other economically developed areas, and even affect the Korean Peninsula and the Japanese Islands.The generate location of Cluster-2 is the most westward, its lifetime is the shortest and its intensity is the weakest, but it owns the highest frequency of occurrence.This type mainly affects China's Guangdong, Guangxi and Fujian.The generate location of Cluster-3 is the most southward, and its lifetime is the most enduring, but more inactive than Cluster-1 and Cluster-2.With respect to the track, this type slightly move to the northwest direction.Cluster-3 mainly affects China's Fujian, Zhejiang.
Through the analysis of the circulation situation, results are shown as follows.The west-extending ridge point of the subtropical ridge line's location of Cluster-1 is the most northward, and the southeast airflow turns to the southwest airflow around the cyclone, which leads to the minimum steering flow velocity.When generating, Cluster-1 has the lowest convergence degree, zonal wind vertical shear and water vapor transportation.The water vapor supply during the decaying process is also the weakest among the three classes, but still has a complete cyclonic structure.The west-extending ridge point of the subtropical ridge line's location of Cluster-2 locates the most westward, with the southeastward easterly airflow dominated by the cyclone.When disappearing, the convergence over the lower level is the weakest.The location of Cluster-3's west-extending ridge point is the most eastward and southward, and the southeast airflow turns to the southwest airflow around the cyclone, which leads to the maximum steering flow velocity.This cluster owns the strongest vertical zonal wind shear when occurring.Both in occurring and decaying stage, water vapor supply of Cluster-3 is the strongest among the three clusters.
The above work and conclusions not only enrich the research of TC tracks, but also provide reference for climate diagnosis and new ideas for scientific research in the field of data mining and meteorology.

Abstract:How to obtain the atmospheric gravity wave parameters is one of the most important topics in atmospheric dynamic theory.By taking a thunder storm process on 6—8 August 2013 in northwest Shanxi province as an example in the paper and by using the commonly used second-order polynomial fitting, third-order polynomial fitting, fourth-order polynomial fitting and Bandpass filtering methods to obtain disturbance fields(denoted as SeOp, ThOp, FoOp and BpFp, respectively) from sounding data with high resolution, the paper conducted comparative analysis on the atmospheric gravity wave parameters obtained from those disturbance fields by using different methods.The results showed that different disturbance fields(second-order curve fitting, third-order curve fitting, fourth-order curve fitting and Bandpass filtering) exerted certain influence on the value and trend of the gravity wave parameters to different degrees.They exerted obvious effect on the value and trend of wave period, horizontal and vertical wavelength while posed little influence on the propagation direction, group velocity and phase velocity.By considering the dynamic characterisitcs of the thunder storm process during 6—8 Ausugst 2013, the gravity wave parameters obtained from SeOp and BpFp showed the highest reliability while the gravity wave parameter obtained from FoOp showed the lowest reliability.

Abstract:Since ozone hole in the polar region has been found, ozone depletion aroused extensive concern all over the world.The ozone valley over the Tibetan Plateau(OVTP) and over North America(OVNA) are seasonal ozone lows out of the polar region.They occurred both in middle latitudes and in summer half year.In order to compare their vertical structure and formation mechanism, we calculated ozone zonal deviation rate to show vertical structure of ozone valley over the Tibetan Plateau(OVTP) and North America, using MLS(Microwave Limb Sounder) data from 2005 to 2013, where ozone zonal deviation rate is ratio of ozone zonal deviation to ozone zonal mean, and ozone zonal deviation equals to ozone concentration minus ozone zonal mean.The ozone zonal deviation rate indicates that there are two centers of OVTP and one center of OVNA.One center of OVTP is located in the upper troposphere and lower stratosphere(UTLS) with peak ozone zonal deviation rate of -0.3, the other center is located in the upper stratosphere with peak ozone zonal deviation rate of -0.01.The center of OVNA is also located in the UTLS region with peak value of -0.18.The integration of zonal deviation shows that the UTLS center of OVTP with -15 DU is the strongest, the UTLS center of OVNA with -5 DU is in the middle and the upper center of OVTP with -1 DU is the weakest.Therefore, the vertical structure of ozone valley is clear.Then we analyzed the dynamic and chemical mechanism of the ozone valley.Summer circulation calculated from ERA-interim reanalysis data and summer ozone flux divergence calculated from ERA-interim data show that anticyclones are located over Tibetan Plateau and North America in the UTLS region.The anticyclone domain over the Tibetan Plateau is larger than that over North America.Moreover, corresponding ozone flux divergence over the Tibetan Plateau(1×10^-12 kg·kg^-1·s^-1) is more robust than that over North America(5×10^-13 kg·kg^-1·s^-1), which is the main reason why UTLS center of OVTP is stronger than that of OVNA.UTLS center of OVTP, UTLS center and the only center of OVNA are mainly caused by dynamic processes.However, the MLS data implies chemical reaction may play a role in the upper center of OVTP.Zonal deviation of chlorine monoxide and hydrogen chloride are negative while zonal deviation of nitrogen dioxide is positive at 10 hPa over the Tibetan Plateau, which means chlorine monoxide and hydrogen chloride concentration are higher over the Tibetan Plateau than that in the same latitudes.Besides, nitrogen dioxide concentration are lower over the Tibetan Plateau than that in the same latitudes.Higher chlorine monoxide and hydrogen chloride concentration imply that chlorine catalytic reactions make ozone loss stronger over the Tibetan Plateau.The lower nitrogen dioxide concentrations may slow the reaction which deactivates reactive chlorine into nonreactive chlorine and may strengthen the chlorine catalytic reactions.Consequently, chemical processes may have an effect on the upper center of OVTP.To sum up, there are two centers of OVTP and only one center of OVNA.UTLS center of OVTP with -15 DU is the strongest, the UTLS center of OVNA with -5 DU is in the middle and the center of OVTP in the upper stratosphere with -1 DU is the weakest.UTLS center of OVTP, UTLS center and the only center of OVNA are mainly caused by dynamic processes.However, chemical processes may have an effect on the upper center of OVTP.

Abstract:Over the past 10 years, due to the Arctic Amplification of global warming in the North Pole region, there have been situations of accelerated temperature rise and drastic decrease in sea ice area in the North Pole;nevertheless, the air temperature of continents in the north hemisphere continue to drop.Particularly in recent years, there have been abnormally cold winters in large areas.During the winter season in China, the cold air activity that causes temperature drop, including cold waves that can cause dramatic temperature drops throughout the entire country, often travels southward together with the cold high pressure on the ground.Therefore, winter temperature anomaly in China could be related to the anticyclone activities on the Eurasian continent.In this study, on the basis of Lagrange’s theories, the NCEP/NCAR reanalysis data are used to trace anticyclone activities on the Eurasian continent from 1979—2012, and the strength index of the winter anticyclone on the Eurasian continent is defined.Based on this, we then analyze the features of anticyclone on the Eurasian continent for the past 30 years, as well as its relationship to winter temperature in China.The results show the following:
1)The index of anticyclone intensity exhibits obvious inter-annual and inter-decades fluctuations in the winter during 1979—2012, and the ACI exhibits a pronounced intensification over the past 10 years.
2)The winter temperature anomaly in China is negatively correlated to the strength of anticyclone activities on the Eurasian continent, i.e.when the anticyclone activities are strong, most part of China are colder, and vice visa.The negative anomalies of winter temperature increases in China in the past years were mostly related to the strong phase of anticyclone activities on the Eurasian continent.
3)Anticyclones with short lifetime occur most frequently, and have a more apparent impact on the winter temperature in China in larger regions.Anticyclones with long lifetimes occur less frequently, and mainly affect the temperatures of the Liaoning and central Inner Mongolia regions.Anticyclones with extremely long lifetimes account for 10%.
4)The Mongolian plateaus are the key region for winter anticyclone activities on the Eurasian continent, with 40% of anticyclones passing this region annually.They are highly mobile, with the Scandinavia peninsula, Mediterranean, Caspian Sea, Black Sea, Aral Sea and central Siberia areas as their main origins.

Abstract:Based on the daily 500 hPa geopotential height data between June and August, 2007—2012, the historical reanalysis grid data of NCEP global 2.5°×2.5°and the daily precipitation data of 158 meteorological stations in north of Zhejiang province, the relationships between local precipitation and large-scale precipitation in different atmospheric circulations are studied in this paper.The BP neural network combined with 4 forecasting objects and corresponding predictor variables in different circulations are employed to design 4 downscaling function models to approximate the precipitation data.The 4 models are used to simulate and forecast the daily precipitation data of 158 meteorological stations in north of Zhejiang province, and the results show that the BP neural network model with 2 hidden layers has good simulation accuracy.Through Jenkinson atmospheric circulation to classify the precipitation into SE(SE type), NW(NW type), C(C type) and SW(SW type), NW type and C type generally outperform the SW type and SE type in simulation of the extreme precipitation.Compared with the area of Ningbo and Zhoushan, other areas of north Zhejiang reflect the greater error value from 4 atmospheric circulations.The prediction accuracy of the downscaling model is the best of three types of rainstorm forecast after categorizing rainfall into different levels.