Learning Structure from Data and Its Application to Ozone Prediction

In this paper we propose an algorithm for structure learning in predictive expert systems based on a probabilistic network representation. The idea is to have the simplest structure (minimum number of links) with acceptable predictive capability. The algorithm starts by building a tree structure based on measuring mutual information between pairs of variables, and then it adds links as necessary to obtain certain predictive performance. We have applied this method for ozone prediction in México City, where the ozone level is used as a global indicator for the air quality in different parts of the city. It is important to predict the ozone level a day, or at least several hours in advance, to reduce the health hazards and industrial losses that occur when the ozone reaches emergency levels. We obtained as a first approximation a tree-structured dependency model for predicting ozone in one part of the city. We observe that even with only three parameters, its estimations are acceptable.A causal network representation and the structure learning techniques produced some very interesting results for the ozone prediction problem. Firstly, we got some insight into the dependence structure of the phenomena. Secondly, we got an indication of which are the important and not so important variables for ozone forecasting. Taking this into account, the measurement and computational costs for ozone prediction could be reduced. And thirdly, we have obtained satisfactory short term ozone predictions based on a small set of the most important parameters.     

On the correction of the total ozone content over Athens,Greece as deduced from satellite observations

The observations with the Total Ozone Mapping Spectrometer ( TOMS) mounted aboard the Nimbus-7 satellite have previously been used to determine the trends of the total ozone amount over Athens, Greece ( 38° N, 24° E), since 1979, for various months ( Varotsos, C. A., and Cracknell, A. P., 1993, International Journal of Remote Sensing, 14, 2053–2059). The total ozone depletion over the 13-year time period showed a strong seasonal variation of the trend from more than 7 per cent in winter to about 2·5 per cent in summer. However, the TOMS instrument measures the back-scattered ultraviolet radiation in order to determine ozone content and is limited to observations above the cloud level. ln the presence of thick cloud the column ozone content is generally underestimated. This underestimation of the total ozone amount is quantitatively examined, especially in the synoptic cases where ozone-rich air has been transported into the lower troposphere. The influence of this underestimation on the total ozone depletion over Athens, Greece, deduced from TOMS observations, is finally attempted.     

Contour diagram fuzzy model for maximum surface ozone prediction

A contour diagram approach is presented for the identification of surface ozone concentration feature based on a set of rules by considering the meteorological variables such as the solar radiation, wind speed, temperature, humidity and rainfall. A fuzzy rule system approach is used because of the imprecise, insufficient, ambiguous and uncertain data available. The contour diagrams help to identify qualitative ozone concentration variability rules which are more general than conventional statistical or time series analysis. In the methodology, ozone concentration contours are based on a fixed variable as ozone precursor, namely, NO_x and as the third variable one of the meteorological factors. Such contour diagrams for ozone concentration variation are prepared for six months. It is possible to identify the maximum ozone concentration episodes from these diagrams and then to set up the valid rules in the form of IF-THEN logical statements. These rules are obtained from available daily ozone, NO_x and meteorological data as a first approximate reasoning step. In this manner, without mathematical formulations, expert maximum ozone concentration systems are identified. The application of the contour diagram approach is performed for daily ozone concentration measurements on European side of Istanbul city. It is concluded that through approximate reasoning with fuzzy rules, the maximum ozone concentration episodes can be identified and predicted without any mathematical expression.     

Total ozone depletion over Greece as deduced from satellite observations

The reprocessed (version 7) daily total ozone observations made by the Total Ozone Mapping Spectrometer (TOMS) on the Nimbus-7 satellite over Athens (37.6 N, 23.4 E) for the period from November 1978 until April 1993 have been used to investigate total ozone depletion. To estimate the trends in total ozone content a linear fitting to the data has been applied, given that the other components like the quasi-biennial oscillation, the El Nino/Southern Oscillation and the solar cycle have a very small contribution to the total ozone depletion effects over that geographical region. The total ozone depletion over the 15-year period was derived from version 7 shows a strong seasonal variation from more than 6% in winter and early spring to about 1.5% in summer. The total ozone depletion over Greece is found to be about 1% (per decade) less using version 7 than using version 6.     

Ozone dynamics over Greece as derived from satellite and! in situ measurements

A comprehensive analysis of the records of surface ozone available for Athens, Greece ( 38° N, 24° E) for the periods 1901–1940 and 1987–1990 is presented. Both records are analysed to explore the intraseasonal fluctuations and the harmonic components of surface ozone and also compared to other historical surface ozone records. The variation in surface ozone concentration during rainfall is also investigated, using the hourly measurements of the surface ozone concentration obtained by a network of four stations within the Greater Athens area. The results indicate that, during rainfall events which are associated with the passing of a cold front, an important decrease of the surface ozone concentration is observed. Daily measurements of surface ozone and NO_x, from five stations in the Greater Athens Basin overthe period 1986–1990 are also used in order to examine the main features of basin-wide 0₃-HC-NO_x relations. A simple regression model between the surface ozone concentration and the temperature at the 850 hPa level, which was first tested in Los Angeles, gave satisfactory results in reproducing the mean monthly ozone variation in Athens, when coefficients extracted from local data were used in the regression equation. A series of vertical ozone soundings over Athens has been also performed in order to explore the tropospheric ozone variations and to examine further the transport that occurs at the 700hPa level with advection from the north-western sector. The relevant results are discussed. The existing uncertainties concerning the stratosphere-troposphere exchange of ozone which mainly occurs during midlatitude tropopause folding as well as during cut-off low events are also discussed. The examination of the role of the atmospheric circulation in the lower stratosphere in relation to the laminated structure of ozone is also attempted. The data collected during the balloon ascents have been compared with those during the balloon descents. Both profiles are compared with the total ozone measurements derived from the TOMS on the Nimbus-7 satellite and the Dobson spectrophotometer. The data collected for the vertical distribution of ozone and temperature have been compared with the satellite-derived reference models which provide the monthly latitudinal variations of vertical structure of both ozone and temperature. We have also used total ozone measurements obtained with a Dobson spectrophotometer ( No. 118) which has been instituted in Athens from 1989 in order to examine the consistency of data from TOMS with the corresponding Dobson data on a daily basis. Furthermore monthly mean total ozone data were first estimated for the entire period and were then Fourier analysed to obtain the amplitude, phase and percentage contribution to the total variance of the first, second and third harmonics.     

基于模糊神经网络的臭氧净水机智能控制研究

提出了一种基于模糊神经网络的臭氧杀菌净水机的控制方法,首先,分析了影响臭氧净水机臭氧含量的因素,再根据实验数据,利用模糊神经网络建模,得到了外界水温水压与臭氧管上的对应高压脉冲频率的非线性关系,并在此基础上设计出一种以数字信号处理器为控制核心的适应环境变化的智能臭氧净水机,实验表明,这种新型臭氧净水机能够保持稳定的臭氧浓度,有效杀菌.     

A note on the comparison between total ozone from Oslo CTM2 and SBUV satellite data

The results of a comparison between total ozone amounts derived from solar backscatter ultraviolet (SBUV) satellite observations and those calculated from the chemical transport model Oslo CTM2 are presented for the period 2001–2007. Monthly mean total ozone amounts from improved model simulations were used to compute monthly, seasonal and annual zonal means over 10° latitude zones, and compared with respective satellite retrievals over the northern and southern hemispheres. The results show that the improved model simulations slightly underestimate total ozone over the northern hemisphere when compared with the satellites by 1.4% on average, and slightly overestimate total ozone over the southern extra-tropics, middle and high latitudes by 1.6% on average. The mean difference between the model- and satellite-derived total ozone columns from 75°S to 75°N is estimated to be about ?0.3%. A linear regression analysis between the model- and satellite-derived total ozone data shows statistically significant correlations between the two data sets at all latitude zones (about +0.8 in the tropics and more than +0.9 over all other latitudes). The annual cycle of total ozone is shown to be well reproduced by the model at all latitudes.     

Total ozone observations at Athens,Greece by satellite-borne and ground-based instrumentation

A comparison is made of total ozone (TOZ) content observations conducted by the Dobson spectrophotometer No. 118, the SCanning Imaging Absorption SpectroMeter for Atmospheric CHartographY (SCIAMACHY), the Total Ozone Mapping Spectrometer (TOMS) and the Ozone Monitoring Instrument (OMI) over Athens, Greece, during 1991–2008. Spearman's and Wilcoxon's tests were used to determine the measure of the agreement between the ground-based and satellite column ozone data. The correlation coefficient between Dobson and Nimbus-7, ADEOS, Earth Probe, OMI and SCIAMACHY observations was found to be 0.95, 0.96, 0.94, 0.93 and 0.87, respectively, while the correlation coefficient between total ozone observations of SCIAMACHY and Earth Probe-TOMS and OMI is 0.85 and 0.93, respectively. SCIAMACHY overestimates the column ozone with respect to Dobson, Earth Probe-TOMS and OMI by 10, 15 and 3 DU, respectively, while Dobson underestimates the column ozone with respect to Nimbus-7, ADEOS and OMI by 5, 10 and 8 DU. The results obtained confirm that the Athens Dobson station may continue to be considered as a ground-truth total ozone station for the validation of the satellite column ozone observations. In addition, linear regression analysis of the deseasonalized monthly mean column ozone, as derived from Dobson measurements, gives an increase of +0.33 ± 0.07% per year during 1991–2000 and a decrease of –0.33 ± 0.07% per year for the period 2001–2008.     

Assessment and prediction of tropospheric ozone concentration levels using artificial neural networks

This work deals specifically with the use of a neural network for ozone modelling in the lower atmosphere. The development of a neural network model is presented to predict the tropospheric (surface or ground) ozone concentrations as a function of meteorological conditions and various air quality parameters. The development of the model was based on the realization that the prediction of ozone from a theoretical basis (i.e. detailed atmospheric diffusion model) is difficult. In contrast, neural networks are useful for modelling because of their ability to be trained using historical data and because of their capability for modelling highly non-linear relationships. The network was trained using summer meteorological and air quality data when the ozone concentrations are the highest. The data were collected from an urban atmosphere. The site was selected to represent a typical residential area with high traffic influences. Three neural network models were developed. The main emphasis of the first model has been placed on studying the factors that control the ozone concentrations during a 24-hour period (daylight and night hours were included). The second model was developed to study the factors that regulate the ozone concentrations during daylight hours at which higher concentrations of ozone were recorded. The third model was developed to predict daily maximum ozone levels. The predictions of the models were found to be consistent with observations. A partitioning method of the connection weights of the network was used to study the relative percent contribution of each of the input variables. The contribution of meteorology on the ozone concentration variation was found to fall within the range 33.15–40.64%. It was also found that nitrogen oxide, sulfur dioxide, relative humidity, non-methane hydrocarbon and nitrogen dioxide have the most effect on the predicted ozone concentrations. In addition, temperature played an important role while solar radiation had a lower effect than expected. The results of this study indicate that the artificial neural network (ANN) is a promising method for air pollution modelling.     

Characterization of ozone sensors based on WO3 reactively sputtered films: influence of O2 concentration in the sputtering gas, and working temperature

We report on electrical responses of tungsten oxide thin film ozone sensors based on a tungsten trioxide (WO₃)/tin oxide (SiO₂)/Si structure with interdigitated Pt electrodes. The influence of O₂ concentration in the sputtering gas and working temperature of the sensor are investigated. Sensitivity to ozone increases with O₂ content in the sputtering gas. It reaches its highest value for sensors fabricated with 50% O₂. For these sensors, the best ozone sensitivity and shortest response and recovery times are obtained at a working temperature of 523 K. Ozone sensitivity is compared to other ozone sensors.     

Temporal and spatial variability of total ozone column using TOMS satellite observations and comparison with measurements from the Dobson network

This article presents a detailed analysis of seasonal and interannual variability of total ozone content (TOC) at 16 different stations in Africa using Nimbus-7 Total Ozone Mapping Spectrometer (TOMS) data for a period of 14 years (January 1979–December 1992). The analysis provides not only an estimate of the long-term annual and seasonal trends but also statistics of means and variability of ozone on temporal and spatial scales. For example, high negative deviations were observed for stations in Northern Africa in spring (March–May), with as much as –20 DU in Alexandria. A comparison of total ozone column data retrieved from the TOMS satellite with measurements obtained from the Dobson ground-based network is further presented for Cairo, Irene, Nairobi and Springbok. Estimates of the percentage seasonal difference between TOMS satellite and Dobson ground-based measurements reveal that the ground measurements were higher in magnitude at all stations with the exception of Nairobi. To verify the level of correlation between the ground-based and satellite observations, rank correlation coefficients were determined for all stations using daily and monthly observations. The results show that there is good correlation between the compared data sets, with daily coefficient of determination (r ²) values of 0.87, 0.76, 0.58 and 0.87 for Cairo, Irene, Nairobi and Springbok, respectively.     

Extremes in total ozone content over northern India

Using daily station total ozone data from TOMS on Nimbus‐7 (1979–1993) and Earth Probe (1997–2005) satellites of National Aeronautics and Space Administration (NASA)/Goddard Space Flight Center (GSFC) during the period 1979–2005, the characteristic features of extremes in the total ozone content and the frequency of the low/high ozone days have been carried out over the northern parts of India in the winter season. Stations located in the north of 20° N latitude regions of India have been chosen for the study. To determine whether the day is a low, high or normal ozone day, the statistical percentile thresholds are computed based on the daily data during winter months (January and February). It has been observed that the trends in the frequency of low ozone days are increasing and for the high ozone days are decreasing during 1979–1993. Similarly the trends in the highest total ozone reaching during January and February are decreasing. The recent period (1997–2005) shows opposite trends that are not statistically significant or stable during the period. Even the mean total ozone during January and February for the period 1979–1993 show the decreasing trends. Overall the trends in the total ozone extremes and the frequency of low/high ozone days are found to be decreasing over the northern parts of India during the winter season.     

SCIAMACHY/Envisat,OMI/Aura,and ground-based total ozone measurements over Kyiv-Goloseyev station

Validation of satellite ozone measurements is important for data improvement due to instrumental long-term drifts and retrieval algorithm limitations. For satellite data quality estimation, we compare the total ozone content (TOC) derived from the satellite Scanning Imaging Absorption Spectrometer for Atmospheric Chartography (SCIAMACHY)/Envisat and Ozone Monitoring Instrument (OMI)/Aura spectrometer overpass data and the ground-based measurements made with the Dobson spectrophotometer 040 at the GAW station No. 498 Kyiv-Goloseyev. The station was opened for Dobson ozone measurements in 2010. The results for Direct Sun, Zenith Blue, and Zenith Cloud observations are presented separately, in order to assess the influence of weather conditions (clear or cloudy sky) on the difference between satellite and ground-based measurements. Results from the SCIAMACHY–Dobson and OMI–Dobson difference analyses show small relative overestimation of TOC for satellite data. The ground-based Dobson 040 data are of high quality for Direct Sun and Zenith Blue from AD ((305.5 and 325.0 nm) and (317.5 and 339.9 nm)) pair measurements. Seasonal variations of the difference are seen with maximal satellite–Dobson data discrepancy near the winter solstice. Satellite TOC values are systematically higher than Dobson ones at solar zenith angles larger than 70°. This difference could be explained by seasonal non-uniformity in the satellite data.     

Surface and total ozone investigations in the region of Sofia,Bulgaria

Atmospheric ozone behaviour over Sofia has been investigated with remote-sensing and in situ techniques. Surface ozone and boundary layer observations performed in recent years at three city sites have been analysed. It was found that, in the autumn period, at close meteorological conditions, diurnal ozone variations show stable behaviour from year to year during the analysed period. It may be assumed that the boundary layer and ozone precursor concentrations, which are involved in photochemical ozone formation, keep up their state from year to year at the mentioned conditions. These findings may be interesting when surface ozone trends and climate change influence on ozone are investigated. The analysis of the long-term total ozone content (TOC) variations did not find a total ozone trend in the 1997–2008 period.     

Prediction of mean monthly total ozone time series – application of radial basis function network

The primary objective of the present paper is to apply Artificial Neural Network in the form of Radial Basis Function network to predict the mean monthly total ozone concentration over Arosa, Switzerland (46.8° N/9.68° E). The satellite observations of the total ozone content are based on the total ozone observations performed by the ground‐based instrumentation. While analysing the dataset it was found that January, February and March are the months of maximum variability in the mean monthly total ozone over the stated region. Then, these three months were considered as the target months to frame the predictive model. After appropriate training and testing, it was found that Radial Basis Function network is a suitable neural net type for predicting the aforesaid time series. Moreover, this kind of neural net was found most adroit in predicting the mean monthly total ozone in the month of January.     

Spatial prediction of ozone concentration profiles

Ground level ozone is one of the major air pollutants in many urban areas. Ozone formation affects ecosystems and is known to be associated with many adverse health issues in humans. Effective modeling of ozone is a necessary step to develop a system to warn residents of high ozone levels. In the present study we propose a statistical procedure that uses multiscale and functional data analysis to improve the spatial prediction of ozone concentration profiles in the Dallas Fort Worth (DFW) area of Texas. This study uses daily eight-hour ozone concentrations and meteorological predictors during a period between 2003 and 2006 at 14 monitoring sites in the DFW area. Wavelet transformation was used as a means of multiscale data analysis, followed by functional modeling to reduce model complexity. Kriging was then used for spatial prediction. The experimental results with real data demonstrated that the proposed procedures achieved acceptable accuracy of spatial prediction.     

Dynamic neural observers and their application for identification and purification of water by ozone

A dynamic neural network is applied to estimate the state of the “phenol-water-ozone” chemical system. A new method based on dynamic neural observers with sliding mode (signum) term is used to estimate the dynamics of decomposition of phenols by ozone and to identify their kinetic parameters without the use of any process model. Decomposition of phenols and their mixtures by ozone in a semi-batch reactor is regarded as a dynamic process with an uncertain model (“black box”). Only the content of gaseous ozone is measured at the reactor output during ozonization. Variations of this variable are used to construct a total characteristic curve of the ozonization process. A dynamic state observer is used to estimate the phenol ozonization constant at different pH values from 2 to 12. Experimental data on decomposition dynamics are in good agreement with their estimates. Our method is helpful in on-line monitoring of water purification process without the use of special chemical sensors.     

Effect of the out-of-band stray light on the retrieval of the Umkehr Dobson ozone profiles

The technique of producing a vertical ozone profile using ground-based observations of the Umkehr effect is well established and is widely used for ozone trend analysis. Still, observations of the Umkehr effect by Dobson ozone spectrophotometers have a documented problem: simultaneous measurements from collocated instruments produce noticeably different results, with the difference shown to be dependent on solar zenith angle and total ozone. Ozone profiles are thus difficult to compare from station to station. Current investigations into the level and effects of stray light within the instrument suggest that much of the difference can be attributed to out-of-band stray light characteristics of individual Dobson instruments. Resulting changes in retrieved ozone profiles suggest that adjusting measurements for stray light components removes much of the existing bias between Umkehr profiles and other ozone measuring datasets. This new information has implications for both observational methods and completeness of the existing datasets.     

On the total ozone depletion over Greece derived from satellite-flown and ground-based instruments

There is abundant evidence of the depletion of atmospheric ozone concentration, both from ground-based measurements and from satellite data. However, one set of data published in 1989 by Bais et al. relating to Thessaloniki, in northern Greece, reported no depletion but rather said ‘the period March 1982 to March 1988 shows an insignificant change of total ozone, of about +0-02 per cent per year’. The data of Thessaloniki, obtained from the same instrument (Brewer spectrophotometer no. 5) as that used by Bais et al. has now also been published in the Red Book. In this paper we examine TOMS data for Thessaloniki, Athens and Crete for the six-year period from March 1982 to March 1988 and find a common ozone depletion of the order of 30 per cent during this period, for each of these three sites, including Thessaloniki; this is in contradiction to the results of Bais et al. for Thessaloniki. We have analysed the Red Book data for Thessaloniki and found that it shows total ozone depletion consistent with the results obtained from TOMS data, which are themselves consistent with results from other ground-based measurements in mid-latitudes. It is concluded that there are errors both in the data presented by Bais et al. and in their analysis and, in particular, that their statement quoted above is incorrect and should be disregarded.     

Black box modeling and multiobjective optimization of electrochemical ozone production process

In this paper, simultaneous maximization of generated ozone concentration and minimization of electrical energy consumption is investigated in a laboratory-scale electrochemical ozone production system (EOP). Neural network simulation of EOP was carried out for generated ozone concentration prediction by Abbasi et al. (Chem Eng Res Des 92(11):2618–2625, 2014). In this study, neural network models (as black box models) were developed to predict both generated ozone concentration and electrical energy consumption. The models then were used for optimization. Altruistic non-dominated sorting genetic algorithm with jumping gene variant and termination criterion was used for MOO. Generational distance and spread were used in the termination criterion in order to stop algorithm after the right number of generations. Moreover, several optimal solutions from the Pareto-optimal set are chosen and then validated experimentally.