Comparative study on the performance of ConvLSTM and ConvGRU in classification problems-taking early warning of short-duration heavy rainfall as an example

卷积长短期记忆单元ConvLSTM和卷积门控循环单元ConvGRU是两种广泛应用的深度学习单元,通过将循环机制与卷积运算相结合,常常用于时空序列的预测.为了明确上述两种模型的收敛速度和分类能力,需要使用相同的模型架构对相同的分类问题进行预测.本研究将北京短时强降水区级预警问题看作深度学习中的二分类问题,使用京津冀雷达网的组合反射率数据和北京区域内的自动气象站降雨数据进行深度学习模型的训练和评估.结果表明,ConvGRU的收敛速度比ConvLSTM快约25%.ConvLSTM和ConvGRU的预警性能随地区,时间,降雨强度的变化趋势相似,但大部分ConvLSTM的得分较高,少数情况下ConvGRU的得分较高.

Contrast in Secondary Organic Aerosols between the Present Day and the Preindustrial Period:The Importance of Nontraditional Sources and the Changed Atmospheric Oxidation Capability

Quantifying differences in secondary organic aerosols(SOAs)between the preindustrial period and the present day is crucial to assess climate forcing and environmental effects resulting from anthropogenic activities.The lack of vegetation information for the preindustrial period and the uncertainties in describing SOA formation are two leading factors preventing simulation of SOA.This study calculated the online emissions of biogenic volatile organic compounds(VOCs)in the Aerosol and Atmospheric Chemistry Model of the Institute of Atmospheric Physics(IAP-AACM)by coupling the Model of Emissions of Gases and Aerosols from Nature(MEGAN),where the input vegetation parameters were simulated by the IAP Dynamic Global Vegetation Model(IAP-DGVM).The volatility basis set(VBS)approach was adopted to simulate SOA formation from the nontraditional pathways,i.e.,the oxidation of intermediate VOCs and aging of primary organic aerosol.Although biogenic SOAs(BSOAs)were dominant in SOAs globally in the preindustrial period,the contribution of nontraditional anthropogenic SOAs(ASOAs)to the total SOAs was up to 35.7%.In the present day,the contribution of ASOAs was 2.8 times larger than that in the preindustrial period.The contribution of nontraditional sources of SOAs to SOA was as high as 53.1%.The influence of increased anthropogenic emissions in the present day on BSOA concentrations was greater than that of increased biogenic emission changes.The response of BSOA concentrations to anthropogenic emission changes in the present day was more sensitive than that in the preindustrial period.The nontraditional sources and the atmospheric oxidation capability greatly affect the global SOA change.

Comparison of Adaptive Simulation Observation Experiments of the Heavy Rainfall in South China and Sichuan Basin

This study examines the effectiveness of adaptive observation experiments using the ensemble transformation sensitivity(ETS) method to improve precipitation forecasts during heavy rainfall events in South China and the Sichuan Basin. High-resolution numerical models are employed to simulate adaptive observations. By identifying the sensitive areas of key weather system positions 42 hours before heavy rainfall events, the adaptive observations improve the prediction of jet streams, strong winds, and shear lines, which are essential for accurate heavy rainfall forecasting. This improvement is reflected in both the precipitation structure and location accuracy within the verification region. In South China, targeted observations enhance rainfall predictions by improving water vapor transport. In the Sichuan Basin, adaptive observations refine water vapor transport and adjust vortex dynamics. This research highlights the importance of accurately predicting shear lines and jet streams for forecasting heavy rainfall in these areas. Overall, this study found that adaptive observation enhances the precipitation forecast skills of the structure and location for heavy rainfall in South China and the Sichuan Basin, emphasizing their potential utility in operational numerical weather prediction.

Scientific Advances and Weather Services of the China Meteorological Administration’s National Forecasting Systems during the Beijing 2022 Winter Olympics

Since the Beijing 2022 Winter Olympics was the first Winter Olympics in history held in continental winter monsoon climate conditions across complex terrain areas,there is a deficiency of relevant research,operational techniques,and experience.This made providing meteorological services for this event particularly challenging.The China Meteorological Administration(CMA)Earth System Modeling and Prediction Centre,achieved breakthroughs in research on short-and medium-term deterministic and ensemble numerical predictions.Several key technologies crucial for precise winter weather services during the Winter Olympics were developed.A comprehensive framework,known as the Operational System for High-Precision Weather Forecasting for the Winter Olympics,was established.Some of these advancements represent the highest level of capabilities currently available in China.The meteorological service provided to the Beijing 2022 Games also exceeded previous Winter Olympic Games in both variety and quality.This included achievements such as the“100-meter level,minute level”downscaled spatiotemporal resolution and forecasts spanning 1 to 15 days.Around 30 new technologies and over 60 kinds of products that align with the requirements of the Winter Olympics Organizing Committee were developed,and many of these techniques have since been integrated into the CMA’s operational national forecasting systems.These accomplishments were facilitated by a dedicated weather forecasting and research initiative,in conjunction with the preexisting real-time operational forecasting systems of the CMA.This program represents one of the five subprograms of the WMO’s high-impact weather forecasting demonstration project(SMART2022),and continues to play an important role in their Regional Association(RA)II Research Development Project(Hangzhou RDP).Therefore,the research accomplishments and meteorological service experiences from this program will be carried forward into forthcoming highimpact weather forecasting activities.This article provides an overview and assessment of this program and the operational national forecasting systems.

Multimodel Ensemble Forecast of Global Horizontal Irradiance at PV Power Stations Based on Dynamic Variable Weight

In the present study,multimodel ensemble forecast experiments of the global horizontal irradiance(GHI)were conducted using the dynamic variable weight technique.The study was based on the forecasts of four numerical models,namely,the China Meteorological Administration Wind Energy and Solar Energy Prediction System,the Mesoscale Weather Numerical Prediction System of China Meteorological Administration,the China Meteorological Administration Regional Mesoscale Numerical Prediction System-Guangdong,and the Weather Research and Forecasting Model-Solar,and observational data from four photovoltaic(PV)power stations in Yangjiang City,Guangdong Province.The results show that compared with those of the monthly optimal numerical model forecasts,the dynamic variable weight-based ensemble forecasts exhibited 0.97%-15.96%smaller values of the mean absolute error and 3.31%-18.40%lower values of the root mean square error(RMSE).However,the increase in the correlation coefficient was not obvious.Specifically,the multimodel ensemble mainly improved the performance of GHI forecasts below 700 W m^(-2),particularly below 400 W m^(-2),with RMSE reductions as high as 7.56%-28.28%.In contrast,the RMSE increased at GHI levels above 700 W m^(-2).As for the key period of PV power station output(02:00-07:00),the accuracy of GHI forecasts could be improved by the multimodel ensemble:the multimodel ensemble could effectively decrease the daily maximum absolute error(AE max)of GHI forecasts.Moreover,with increasing forecasting difficulty under cloudy conditions,the multimodel ensemble,which yields data closer to the actual observations,could simulate GHI fluctuations more accurately.

Contrasting Characteristics of the Surface Energy Balance between the Urban and Rural Areas of Beijing

A direct comparison of urban and rural surface energy balances, as well as a variety of other variables including incoming shortwave/longwave radiation and aerosol optical depth, is conducted for the Beijing metropolitan area. The results indicate that, overall, the urban area receives a smaller amount of incoming shortwave radiation but a larger amount of incoming longwave radiation. However, comparisons in the aerosol optical depth and cloud fraction at the two locations suggest that neither aerosol optical depth nor cloud fraction alone can explain the difference in the incoming shortwave radiation. The urban–rural differences in the incoming longwave radiation are unlikely to be caused by the presence of more abundant greenhouse gases over the urban area, as suggested by some previous studies, given that water vapor is the most dominant greenhouse gas and precipitable water is found to be less in urban areas. The higher incoming longwave radiation observed over the urban area is mostly likely due to the higher temperatures of the ambient air. The urban area is also found to always produce higher sensible heat fluxes and lower latent heat fluxes in the growing season. Furthermore, the urban area is associated with a larger amount of available energy（the sum of sensible and latent heat fluxes） than the rural area, except in May and October when evapotranspiration in the rural area significantly exceeds that in the urban area. This study provides observational evidence of urban–rural contrasts in relevant energy-balance components that plausibly arise from urban–rural differences in atmospheric and land-surface conditions.

Impact of Urbanization on Low-Temperature Precipitation in Beijing during 1960–2008

Daily precipitation and temperature records at 13 stations for the period 1960-2008 were analyzed to identify climatic change and possible effects of urbanization on low-temperature precipitation [LTP, precipitation of ≥ 0.1 mm d^-1 occurring under a daily minimum temperature （Tmin） of ≤ 0℃] in the greater Beijing region （B JR）, where a rapid process of urbaniza tion has taken place over the last few decades. The paper provides a climatological overview of LTP in B JR. LTP contributes 61.7% to the total amount of precipitation in B JR in the cold season （November-March）. There is a slight increasing trend [1.22 mm （10 yr）^-1] in the amount of total precipitation for the cold season during 1960-2008. In contrast, the amount of LTP decreases by 0.6 mm （10 yr）^-1. The warming rate of Train in B JR is 0.66℃ （10 yr）^-1. Correspondingly, the frequency of LTP decreases with increasing Tmin by -0.67 times per ℃. The seasonal frequency and amount of LTP in southeast B JR （mostly urban sites） are 17%-20% less than those in the northwestern （rural and montane sites）. The intensity of LTP for the urban sites and northeastern B JR exhibited significant enhancing trends [0.18 and 0.15 mm d^- 1 （10 yr）^- 1, respectively]. The frequency of slight LTP （〈0.2 mm d^-1） significantly decreased throughout B JR [by about -15.74% （10 yr）^-1 in the urban area and northeast B JR], while the contribution of the two heaviest LTP events to total LTP amount significantly increased by 3.2% （10 yr） ^-1.

Impact of Urban 3D Morphology on Particulate Matter 2.5(PM2.5) Concentrations:Case Study of Beijing, China

Urban particulate matter 2.5(PM2.5)pollution and public health are closely related,and concerns regarding PM2.5 are widespread.Of the underlying factors,the urban morphology is the most manageable.Therefore,investigations of the impact of urban three-dimensional(3D)morphology on PM2.5 concentration have important scientific significance.In this paper,39 PM2.5 monitoring sites of Beijing in China were selected with PM2.5 automatic monitoring data that were collected in 2013.This data set was used to analyze the impacts of the meteorological condition and public transportation on PM2.5 concentrations.Based on the elimination of the meteorological conditions and public transportation factors,the relationships between urban 3D morphology and PM2.5 concentrations are highlighted.Ten urban 3D morphology indices were established to explore the spatial-temporal correlations between the indices and PM2.5 concentrations and analyze the impact of urban 3D morphology on the PM2.5 concentrations.Results demonstrated that road length density(RLD),road area density(RAD),construction area density(CAD),construction height density(CHD),construction volume density(CVD),construction otherness(CO),and vegetation area density(VAD)have positive impacts on the PM2.5 concentrations,whereas water area density(WAD),water fragmentation(WF),and vegetation fragmentation(VF)(except for the 500 m buffer)have negative impacts on the PM2.5 concentrations.Moreover,the correlations between the morphology indices and PM2.5 concentrations varied with the buffer scale.The findings could lay a foundation for the high-precision spatial-temporal modelling of PM2.5 concentrations and the scientific planning of urban 3D spaces by authorities responsible for controlling PM2.5 concentrations.

A Microscale Model for Air Pollutant Dispersion Simulation in Urban Areas： Presentation of the Model and Performance over a Single Building

A microscale air pollutant dispersion model system is developed for emergency response purposes. The model includes a diagnostic wind field model to simulate the wind field and a random-walk air pollutant dispersion model to simulate the pollutant concentration through consideration of the influence of urban buildings. Numerical experiments are designed to evaluate the model＇s performance, using CEDVAL （Compilation of Experimental Data for Validation of Microscale Disper- sion Models） wind tunnel experiment data, including wind fields and air pollutant dispersion around a single building. The results show that the wind model can reproduce the vortexes triggered by urban buildings and the dispersion model simulates the pollutant concentration around buildings well. Typically, the simulation errors come from the determination of the key zones around a building or building cluster. This model has the potential for multiple applications; for example, the prediction of air pollutant dispersion and the evaluation of environmental impacts in emergency situations; urban planning scenarios; and the assessment of microscale air quality in urban areas.

Urban Boundary-Layer Stability and Turbulent Exchange during Consecutive Episodes of Particle Air Pollution in Beijing,China

Based on measurements at the Beijing 325-m Meteorological Tower,this study reports an analysis of atmospheric stability conditions and turbulent exchange during consecutive episodes of particle air pollution in Beijing(China),primarily due to haze and dust events(15–30 April 2012).Of particular interest were relevant vertical variations within the lower urban boundary layer(UBL).First,the haze and dust events were characterized by different atmospheric conditions,as quite low wind speed and high humidity are typically observed during haze events.In addition,for the description of stability conditions,the bulk Richardson number(RiB) was calculated for three different height intervals: 8–47,47–140,and 140–280 m.The values of RiB indicated an apparent increase in the occurrence frequency of stably-stratified air layers in the upper height interval—for the 140–280-m height interval,positive values of RiB occurred for about 85% of the time.The downward turbulent exchange of sensible heat was observed at 280 m for the full diurnal cycle,which,by contrast,was rarely seen at 140 m during daytime.These results reinforce the importance of implementing high-resolution UBL profile observations and addressing issues related to stably-stratified flows.

Large Eddy Simulation and Study of the Urban Boundary Layer

Based on a pseudo-spectral large eddy simulation (LES) model, an LES model with an anisotropy turbulent kinetic energy (TKE) closure model and an explicit multi-stage third-order Runge-Kutta scheme is established. The modeling and analysis show that the LES model can simulate the planetary boundary layer (PBL) with a uniform underlying surface under various stratifications very well. Then, similar to the description of a forest canopy, the drag term on momentum and the production term of TKE by subgrid city buildings are introduced into the LES equations to account for the area-averaged effect of the subgrid urban canopy elements and to simulate the meteorological fields of the urban boundary layer (UBL). Numerical experiments and comparison analysis show that: (1) the result from the LES of the UBL with a proposed formula for the drag coefficient is consistent and comparable with that from wind tunnel experiments and an urban subdomain scale model; (2) due to the effect of urban buildings, the wind velocity near the canopy is decreased, turbulence is intensified, TKE, variance, and momentum flux are increased, the momentum and heat flux at the top of the PBL are increased, and the development of the PBL is quickened; (3) the height of the roughness sublayer (RS) of the actual city buildings is the maximum building height (1.5-3 times the mean building height), and a constant flux layer (CFL) exists in the lower part of the UBL.

Assimilation of the FY-4A AGRI Clear-Sky Radiance Data in a Regional Numerical Model and Its Impact on the Forecast of the“21·7”Henan Extremely Persistent Heavy Rainfall

Assimilation of the Advanced Geostationary Radiance Imager(AGRI)clear-sky radiance in a regional model is performed.The forecasting effectiveness of the assimilation of two water vapor(WV)channels with conventional observations for the“21·7”Henan extremely heavy rainfall is analyzed and compared with a baseline test that assimilates only conventional observations in this study.The results show that the 24-h cumulative precipitation forecast by the assimilation experiment with the addition of the AGRI exceeds 500 mm,compared to a maximum value of 532.6 mm measured by the national meteorological stations,and that the location of the maximum precipitation is consistent with the observations.The results for the short periods of intense precipitation processes are that the simulation of the location and intensity of the 3-h cumulative precipitation is also relatively accurate.The analysis increment shows that the main difference between the two sets of assimilation experiments is over the ocean due to the additional ocean observations provided by FY-4A,which compensates for the lack of ocean observations.The assimilation of satellite data adjusts the vertical and horizontal wind fields over the ocean by adjusting the atmospheric temperature and humidity,which ultimately results in a narrower and stronger WV transport path to the center of heavy precipitation in Zhengzhou in the lower troposphere.Conversely,the WV convergence and upward motion in the control experiment are more dispersed;therefore,the precipitation centers are also correspondingly more dispersed.

Evaluation of the Added Value of Probabilistic Nowcasting Ensemble Forecasts on Regional Ensemble Forecasts

Ensemble forecasting systems have become an important tool for estimating the uncertainties in initial conditions and model formulations and they are receiving increased attention from various applications.The Regional Ensemble Prediction System(REPS),which has operated at the Beijing Meteorological Service(BMS)since 2017,allows for probabilistic forecasts.However,it still suffers from systematic deficiencies during the first couple of forecast hours.This paper presents an integrated probabilistic nowcasting ensemble prediction system(NEPS)that is constructed by applying a mixed dynamicintegrated method.It essentially combines the uncertainty information(i.e.,ensemble variance)provided by the REPS with the nowcasting method provided by the rapid-refresh deterministic nowcasting prediction system(NPS)that has operated at the Beijing Meteorological Service(BMS)since 2019.The NEPS provides hourly updated analyses and probabilistic forecasts in the nowcasting and short range(0-6 h)with a spatial grid spacing of 500 m.It covers the three meteorological parameters:temperature,wind,and precipitation.The outcome of an evaluation experiment over the deterministic and probabilistic forecasts indicates that the NEPS outperforms the REPS and NPS in terms of surface weather variables.Analysis of two cases demonstrates the superior reliability of the NEPS and suggests that the NEPS gives more details about the spatial intensity and distribution of the meteorological parameters.

Analysis on Discipline Construction and Development Strategy of National-level Professional Research Institutes

Discipline construction is development core of a professional research institute. The paper describes function,guiding ideology,and construction of discipline construction,based on giving a national-level professional research institute directly under the Ministry of Science and Technology,as an example. Furthermore,according to current situation of discipline construction in a professional research institute,the paper discusses development ideas of discipline construction. The final goal of the paper is to provide basis and reference for reinforcing discipline construction of professional research institutes.

Cooling effect of urban parks and their relationship with urban heat islands

It is claimed that open spaces in cities, such as parks, have an urban cooling effect. However, the relationship between urban parks and adjacent districts is still not explicit. In order to clarify the interaction between urban parks and their urban surroundings, this paper takes the Temple of Heaven Park （THP） as an example of a park station and focuses on analyzing the differences with a nearby urban station.THP is located in the center of Beijing, and the nearest urban station is Tian An Men. It is interesting that the cooling effect of THP reaches a peak and remains stable when its city background urban heat island （UHI） varies within a given range, but becomes unstable when the UHI goes beyond the range. This is called an enhanced cooling effect in this paper. As a result, the UHi intensities （UHIIs） are calculated in order to comprehend the role of the park cooling effect in the urban heating characteristics of Beijing. By comparison with five other park-district pairs, this paper attempts to identify the causes of the enhanced cooling effect. It is found that six park-district pairs consistently demonstrate a persistently stronger cooling rate during the night, and that the water coverage might be a key factor in enhancing the park cooling effect. Based on further investigation of the influence of surrounding UHIs on the park cooling effect, it is found that the UHII differences in park-district pairs show quasi-linear changes within a given range as the UHli of the surrounding district increases.

Sensitivity Assessment and Optimization Strategy of Climatic Environment in Central Urban Area of Beijing

As one of the methods of urban environmental assessment,sensitivity assessment of urban climatic environment can accurately reflect the problem areas of current urban climate and built environment through visual means,and help cities to carry out sustainable transformation and renewal in the stock era to optimize the living environment.Taking the central urban area of Beijing as the research object,the sensitivity of climatic environment was evaluated by meteorological data and built environment element data.The results showed that the distribution of sensitive units showed a centrifugal radiation pattern of"high in the center and low in the periphery".Block units with high climate sensitivity were mostly concentrated in Xicheng District and Dongcheng District,accounting for 93%of the total number of units.Further research showed that complex built environment and lack of elements to regulate climatic environment were the main reasons for the poor local climatic environment in the region with high climate sensitivity.Finally,a comprehensive evaluation of the remolding capacity,renewal potential and future development intensity of different grades of climate sensitive units was given.The results will provide scientific reference for urban renewal and reconstruction under climatic and environmental changes,so as to promote urban sustainable development and improvement of living environment.

Impact of Local Climate Zones on the Urban Heat and Dry Islands in Beijing:Spatial Heterogeneity and Relative Contributions

Based on the building height and density data on a 100-m resolution,hourly 2-m temperature and humidity data at83 automatic weather stations,and gridded local climate zone(LCZ)data on a 120-m resolution in urban Beijing in2020,this study first employs the semivariogram combined with building parameters to calculate spatial correlations and has identified an LCZ grid resolution of 500 m suitable for best usage of the available observation data.Then,how the spatially heterogeneous LCZs affect and contribute to the canopy urban heat island intensity(UHII)and urban dry island intensity(UDII)are quantitatively investigated.It is found that UHII is high in winter and low in summer with a unimodal diurnal variation while UDI is low in winter but high in summer with a bimodal diurnal variation.The LCZ with compact mid-rise(open high-rise)buildings exhibits the highest UHII(UDII),followed by the compact high-rise(compact low-rise),while the LCZ of scattered trees presents both the lowest UHII and the lowest UDII.The most significant difference in the UHII(UDII)among the nine LCZ types in the urban area of Beijing is2.62℃(1.1 g kg^(-1)).Area-weighted averaging analysis reveals that the open mid-rise LCZ is the most significant contributor to the UHII(UDII),immediately followed by compact mid-rise(open low-rise),with the least contribution from bare rock or paved(scattered trees).The results also indicate that beyond the intrinsic physical properties of the LCZs of a city,their area proportions cannot be overlooked in evaluating their impact on the UHI and UDI.These quantitatively findings could help urban planners to create a livable urban climate and environment by adjusting the relevant land use.

A Deep Learning Approach for Forecasting Thunderstorm Gusts in the Beijing–Tianjin–Hebei Region

Thunderstorm gusts are a common form of severe convective weather in the warm season in North China,and it is of great importance to correctly forecast them.At present,the forecasting of thunderstorm gusts is mainly based on traditional subjective methods,which fails to achieve high-resolution and high-frequency gridded forecasts based on multiple observation sources.In this paper,we propose a deep learning method called Thunderstorm Gusts TransU-net(TGTransUnet)to forecast thunderstorm gusts in North China based on multi-source gridded product data from the Institute of Urban Meteorology(IUM)with a lead time of 1 to 6 h.To determine the specific range of thunderstorm gusts,we combine three meteorological variables:radar reflectivity factor,lightning location,and 1-h maximum instantaneous wind speed from automatic weather stations(AWSs),and obtain a reasonable ground truth of thunderstorm gusts.Then,we transform the forecasting problem into an image-to-image problem in deep learning under the TG-TransUnet architecture,which is based on convolutional neural networks and a transformer.The analysis and forecast data of the enriched multi-source gridded comprehensive forecasting system for the period 2021–23 are then used as training,validation,and testing datasets.Finally,the performance of TG-TransUnet is compared with other methods.The results show that TG-TransUnet has the best prediction results at 1–6 h.The IUM is currently using this model to support the forecasting of thunderstorm gusts in North China.

Quantify Urbanization-Induced Precipitation and Runoff Anomalies over the Qinhuai River Basin of China: Sensitivity Experiments with WRF-Hydro

Urbanization-related precipitation and surface runoff changes have been widely investigated,but few studies have directly quantified these changes and their link to urbanization in the hydrological cycle.A two-way dynamically coupled atmospheric–hydrological modeling system,Weather Research and Forecasting(WRF)-Hydro,has been applied in this study to perform the quantification.The offline WRF-Hydro was first calibrated and validated for several flooding events against gauge observed streamflow data,with the Nash–Sutcliffe efficiency reaching 0.9.Compared to the WRF model,WRF-Hydro resolves more detailed rainfall pattern features and reproduces the gauge rainfall with a correlation coefficient of 0.8.Then,the impact of urbanization on hydrometeorological processes was investigated with coupled WRF-Hydro sensitivity simulations over the Qinhuai River basin of China during 2 June–31 July 2015.The results indicate that urbanization enhances regional precipitation,resulting in an indirect increase in surface runoff,overland flow,and streamflow by 16.7,93.5,and 111.2 mm,respectively;however,the impervious area results in higher surface runoff,overland flow,and streamflow.Moreover,changes in main hydrometeorological processes further impact the atmospheric–terrestrial water budget,resulting in a decrease in terrestrial water storage and an increase(a decrease)in precipitable water storage in the middle(lower)parts of the lower troposphere.These changes are likely associated with the warmer urban environment than rural areas.Increased water vapor and strengthened convective conditions in the middle part of the lower troposphere due to urban warming are advantageous to the formation of precipitation in urban areas,which in turn increases surface runoff,thereby facilitating the water cycle and altering the atmospheric–terrestrial water budget.

Quantitative relationship between visibility and mass concentration of PM2.5 in Beijing

The pollution of particulate matter less than 2.5μm （PM2.5） is a serious environmental problem in Beijing. The annual average concentration of PM2.5 in 2001 from seasonal monitor results was more than 6 times that of the U,S, national ambient air quality standards proposed by U.S. EPA. The major contributors to mass of PM2.5 were organics, crustal elements and sulfate. The chemical composition of PM2.5 varied largely with season, but was similar at different monitor stations in the same season. The fine particles （PM2.5） cause atmospheric visibility deterioration through light extinction, The mass concentrations of PM2.5 were anti-correlated to the visibility, the best fits between atmospheric visibility and the mass concentrations of PM2.5 were somehow different： power in spring, exponential in summer, logarithmic in autumn, power or exponential in winter. As in each season the meteorological parameters such as air temperature and relative humidity change from day to day, probably the reason of above correlations between PM2.5 and visibility obtained at different seasons come from the differences in chemical compositions of PM2.5.