Combining Trend-Based Loss with Neural Network for Air Quality Forecasting in Internet of Things

Internet of Things(IoT)is a network that connects things in a special union.It embeds a physical entity through an intelligent perception system to obtain information about the component at any time.It connects various objects.IoT has the ability of information transmission,information perception,and information processing.The air quality forecasting has always been an urgent problem,which affects people’s quality of life seriously.So far,many air quality prediction algorithms have been proposed,which can be mainly classified into two categories.One is regression-based prediction,the other is deep learning-based prediction.Regression-based prediction is aimed to make use of the classical regression algorithm and the various supervised meteorological characteristics to regress themeteorological value.Deep learning methods usually use convolutional neural networks(CNN)or recurrent neural networks(RNN)to predict the meteorological value.As an excellent feature extractor,CNN has achieved good performance in many scenes.In the same way,as an efficient network for orderly data processing,RNN has also achieved good results.However,few or none of the above methods can meet the current accuracy requirements on prediction.Moreover,there is no way to pay attention to the trend monitoring of air quality data.For the sake of accurate results,this paper proposes a novel predicted-trend-based loss function(PTB),which is used to replace the loss function in RNN.At the same time,the trend of change and the predicted value are constrained to obtain more accurate prediction results of PM_(2.5).In addition,this paper extends the model scenario to the prediction of the whole existing training data features.All the data on the next day of the model is mixed labels,which effectively realizes the prediction of all features.The experiments show that the loss function proposed in this paper is effective.

Research on Decision Support System (DSS) of Atmospheric Environment Management in Anhui Province Based on Air Quality Forecasting

With the atmospheric stereoscopic monitoring, air quality forecasting and decision of environment management as the main line, and comprehensive management system as the guidance, five platforms including infrastructure, technological support, monitoring and early monitoring, decision support and information services were established. These platforms have 15 subsystems, including stereoscopic monitoring network, visual business consultation, high-performance computing environment, comprehensive management of atmospheric data, emission inventories of pollu-tion sources, evaluation tools of atmospheric models, monitoring and management of air pollution, forecasting and early warning of air quality, diag-nostic analysis of atmospheric environment, tracking of air pollution sources, emergency management of air pollution, conformity management of air quality, comprehensive display of information, releasing of information to external networks, and releasing of information by mobile networks. The decision support system （DSS） of atmospheric environment management could realize an integration business system of 11 air quality forecast - heavy pollution weather warning - diagnosis of pollution causes （dynamic analysis of pollution sources） -air quality conformity planning （air pollu-tion emergency management） -evaluation of forecasting and warning results （evaluation pf management measures） -air quality forecasting＂ and provide the technical support for the prevention and control of atmosphere pollution in Anhui province.

Effect of Meteorological Data Assimilation on Regional Air Quality Forecasts over the Korean Peninsula

The Weather Research and Forecasting model coupled with Chemistry(WRF-Chem),a type of online coupled chemistry-meteorology model(CCMM),considers the interaction between air quality and meteorology to improve air quality forecasting.Meteorological data assimilation(DA)can be used to reduce uncertainty in meteorological field,which is one factor causing prediction uncertainty in the CCMM.In this study,WRF-Chem and three-dimensional variational DA were used to examine the impact of meteorological DA on air quality and meteorological forecasts over the Korean Peninsula.The nesting model domains were configured over East Asia(outer domain)and the Korean Peninsula(inner domain).Three experiments were conducted by using different DA domains to determine the optimal model domain for the meteorological DA.When the meteorological DA was performed in the outer domain or both the outer and inner domains,the root-mean-square error(RMSE),bias of the predicted particulate matter(PM)concentrations,and the RMSE of predicted meteorological variables against the observations were smaller than those in the experiment where the meteorological DA was performed only in the inner domain.This indicates that the improvement of the synoptic meteorological fields by DA in the outer domain enhanced the meteorological initial and boundary conditions for the inner domain,subsequently improving air quality and meteorological predictions.Compared to the experiment without meteorological DA,the RMSE and bias of the meteorological and PM variables were smaller in the experiments with DA.The effect of meteorological DA on the improvement of PM predictions lasted for approximately 58-66 h,depending on the case.Therefore,the uncertainty reduction in the meteorological initial condition by the meteorological DA contributed to a reduction of the forecast errors of both meteorology and air quality.

An air quality forecasting system in Beijing-Application to the study of dust storm events in China in May 2008

An air pollution forecast system,ARIA Regional,was implemented in 2007–2008 at the Beijing Municipality Environmental Monitoring Center,providing daily forecast of main pollutant concentrations.The chemistry-transport model CHIMERE was coupled with the dust emission model MB95 for restituting dust storm events in springtime so as to improve forecast results.Dust storm events were sporadic but could be extremely intense and then control air quality indexes close to the source areas but also far in the Beijing area.A dust episode having occurred at the end of May 2008 was analyzed in this article,and its impact of particulate matter on the Chinese air pollution index （API） was evaluated.Following our estimation,about 23 Tg of dust were emitted from source areas in Mongolia and in the Inner Mongolia of China,transporting towards southeast.This episode of dust storm influenced a large part of North China and East China,and also South Korea.The model result was then evaluated using satellite observations and in situ data.The simulated daily concentrations of total suspended particulate at 6：00 UTC had a similar spatial pattern with respect to OMI satellite aerosol index.Temporal evolution of dust plume was evaluated by comparing dust aerosol optical depth （AOD） calculated from the simulations with AOD derived from MODIS satellite products.Finally,the comparison of reported Chinese API in Beijing with API calculated from the simulation including dust emissions had showed the significant improvement of the model results taking into accountmineral dust correctly.

Study of Combination Forecasting in Airline Traffic Turnover

The basic theory and method of the combination forecasting are introduced. Based on the actual data in an airline, the case study was presented. In the case study, two basic forecasting models are set up, which are the time-regression plus seasonal factor model and the logarithm additive Winters model. And two combination models are established with the basic models, which are the optimal combination model and the regressive combination model. The results of the study are guidable to the practice.

Large inter annual variation in air quality during the annual festival ‘Diwali' in an Indian megacity

A network of air quality and weather monitoring stations was established under the System of Air Quality Forecasting and Research（SAFAR） project in Delhi. We report observations of ozone（O_3）, nitrogen oxides（NO_x）, carbon monoxide（CO） and particulate matter（PM_2.5and PM_（10）） before, during and after the Diwali in two consecutive years, i.e., November 2010 and October 2011. The Diwali days are characterised by large firework displays throughout India. The observations show that the background concentrations of particulate matter are between 5 and 10 times the permissible limits in Europe and the United States. During the Diwali-2010, the highest observed PM_（10） and PM_2.5mass concentration is as high as2070 μg/m^3 and 1620 μg/m3, respectively（24 hr mean）, which was about 20 and 27 times to National Ambient Air Quality Standards（NAAQS）. For Diwali-2011, the increase in PM_（10） and PM_2.5mass concentrations was much less with their peaks of 600 and of 390 μg/m^3 respectively, as compared to the background concentrations. Contrary to previous reports,firework display was not found to strongly influence the NO_x, and O_3 mixing ratios, with the increase within the observed variability in the background. CO mixing ratios showed an increase. We show that the large difference in 2010 and 2011 pollutant concentrations is controlled by weather parameters.

Study on the assimilation of the sulphate reaction rates based on WRF-Chem/DART

Sulphate(SO_(4)^(2-))is a main component of PM_(2.5)in China.The chemical formation mechanisms of sulphate are complex,and many air quality models have been used to analyse these mechanisms.To improve the accuracy of Weather Research Forecast-Chemistry(WRF-Chem)on sulphate,an alternative method is proposed in this paper.Moreover,data assimilation is performed to adjust the chemical reaction rates of sulphate.Based on the original reactions,a new sulphate parameterisation scheme,which includes two hypothetical reactions and six undetermined parameters,was added.Based on the WRF-Chem/DART(Data Assistance Research Testbed)system,the near-ground concentrations of SO_(4)^(2-),SO_(2),NO_(2),O_(3)and particulate matter are assimilated to adjust the six parameters.After adjusting the parameters,the new scheme can effectively solve the underestimation of SO_(4)^(2-)and overestimation of SO_(2).The simulation of SO_(4)^(2-)improved as the mean bias changed from-13.1μg m^(-3)to 3.5μg m^(-3)while SO_(2)improved from 17.0μg m^(-3)to 6.3μg m^(-3).The temporal and spatial variation characteristics predicted by the new scheme are consistent with the theoretical research results,indicating that the complex mechanism of sulphate formation could be replaced by the temporal and spatial variation characteristics predicted by the new scheme and that the parameters can be adjusted by data assimilation.Furthermore,the reaction rates of the SO_(4)^(2-)parameterisation scheme of the WRF-Chem model are improved in this study,and a new method for improving the accuracy of the air quality model is provided.

GAT-EGRU:A Deep Learning Prediction Model for PM2.5 Coupled with Empirical Modal Decomposition Algorithm

With the rapid development of the economy and industry and the improvement of pollution monitoring,how to accurately predict PM2.5 has become an issue of concern to the government and society.In the field of PM2.5 pollution forecasting,a series of results have emerged so far.However,in the existing research field of PM2.5 prediction,most studies tend to predict short-term temporal series.Existing studies tend to ignore the temporal and spatial characteristics of PM2.5 transport,which leads to its poor performance in long-term prediction.In this paper,by optimizing previous PM2.5 deep learning prediction models,we propose a model GAT-EGRU.First,we add a spatial modular Graph Attention Network(GAT)and couple an Empirical Modal Decomposition algorithm(EMD),considering the temporal and spatial properties of PM2.5.Then,we use Gated Recurrent Unit(GRU)to filter spatio-temporal features for iterative rolling PM2.5 prediction.The experimental results show that the GAT-EGRU model has more advantages in predicting PM2.5 concentrations,especially for long time steps.This proves that the GAT-EGRU model outperforms other models for PM2.5 forecasting.After that,we verify the effectiveness of each module by distillation experiments.The experimental results show that each model module has an essential role in the final PM2.5 prediction results.The new model improves the ability to predict PM2.5 after a long time accurately and can be used as a practical tool for predicting PM2.5 concentrations.