Using fuzzy set theory to address the uncertainty of susceptibility to drought

This paper presents the technical aspects of a new methodology for assessing the susceptibility of society to drought. The methodology consists of a combination of inference modelling and fuzzy logic applications. Four steps are followed: (1) model input variables are selected—these variables reflect the main factors influencing susceptibility in a social group, population or region, (2) fuzzification—the uncertainties of the input variables are made explicit by representing them as ‘fuzzy membership functions’, (3) inference modelling—the input variables are used to construct a model made up of linguistic rules, and (4) defuzzification—results from the model in linguistic form are translated into numerical form, also through the use of fuzzy membership functions. The disadvantages and advantages of this methodology became apparent when it was applied to the assessment of susceptibility from three disciplinary perspectives: Disadvantages include the difficulty in validating results and the subjectivity involved with specifying fuzzy membership functions and the rules of the inference model. Advantages of the methodology are its transparency, because all model assumptions have to be made explicit in the form of inference rules; its flexibility, in that informal and expert knowledge can be incorporated through ‘fuzzy membership functions’ and through the rules in the inference model; and its versatility, since numerical data can be converted to linguistic statements and vice versa through the procedures of ‘fuzzification’ and ‘defuzzification’.     

Critical loads for Finnish lakes: comparison of three steady-state models

Three simple steady-state water-chemistry models are used to calculate critical loads of sulfur for lakes in Finland. Because of the high concentrations of organic matter in Finnish lakes, the influence of organic anions on the calculation of critical loads has been given special attention. The first two methods are well known ion-balance methods which have been used in many previous lake-acidification studies. The third method, developed for this study, includes the numerical solution of equilibrium equations for organic anions, inorganic carbon species and inorganic monomeric aluminum. The original pH and aluminum concentration of the lakes are estimated with this model, and a method to estimate the original acid neutralizing capacity (ANC) by simulating a Gran-titration is also tested on the lake data. Uncertainty in the predictions is estimated by varying the most critical model parameters.     

Yield and grain quality of spring wheat (Triticum aestivum L., cv. Drabant) exposed to different concentrations of ozone in open-top chambers

Spring wheat (Triticum aestivum L., cv. Drabant) was exposed to different concentrations of ozone in open-top chambers for two growing seasons, 1987 and 1988, at a site located in south-west Sweden. The chambers were placed in a field of commercially grown spring wheat. The treatments were charcoal-filtered air (CF), non-filtered air (NF) and non-filtered air plus extra ozone (NF(+)). In 1988, one additional ozone concentration (NF(++)) was used. Grain yield was affected by the ozone concentration of the air. Air filtration resulted in an increase in grain yield of about 7% in both years, compared to NF. The addition of ozone (NF(+), NF(++)) reduced grain yield and increased the content of crude protein of the grain in both years. Filtration of the air had no significant effect on the content of crude protein, compared to NF. The results showed a strong positive chamber effect on grain yield in the cold and wet summer of 1987. In 1988, there was no net chamber effect on grain yield. The relative differences between the CF, NF and NF(+) treatments with respect to grain yield were of the same magnitude in the two years, despite the very different weather conditions.     

Analysis of pollutant levels in central Hong Kong applying neural network method with particle swarm optimization

Air pollution has emerged as an imminent issue in modernsociety. Prediction of pollutant levels is an importantresearch topic in atmospheric environment today. For fulfillingsuch prediction, the use of neural network (NN), and inparticular the multi-layer perceptrons, has presented to be acost-effective technique superior to traditional statisticalmethods. But their training, usually with back-propagation (BP)algorithm or other gradient algorithms, is often with certaindrawbacks, such as: 1) very slow convergence, and 2) easilygetting stuck in a local minimum. In this paper, a newlydeveloped method, particle swarm optimization (PSO) model, isadopted to train perceptrons, to predict pollutant levels, andas a result, a PSO-based neural network approach is presented. The approach is demonstrated to be feasible and effective bypredicting some real air-quality problems.     

Reduction of SO2 emissions by ammonia gas during unstaged combustion

The reduction of SO₂ by the addition of ammonia gas has been studied in a 2 m high fluidized bed combustor having a 30 cm static bed height and a freeboard height of 170 cm. Ammonia gas was injected at 52 cm above the distributor where the temperature is ca. 700° C by an uncooled stainless steel tube injector. Experiments were carried out to investigate the effects of amminia gas injection on sulphur dioxide emissions at unstaged conditions of: (i) excess air level, (ii) NH₃:SO₂ molar ratio, (iii) fluidizing velocity and (iv) bed height.A maximum reduction of 75% in SO₂ emissions was found at 40% excess air, at an NH₃:SO₂ molar ratio of 5.4. The onset of SO₂ reduction occurred at an NH₃:SO₂ ratio of 1.5 However, the most effective ratio was found to be between 3 and 5. Fluidizing velocity and bed height were also found to have significant influence on SO₂ reduction.It is difficult to determine how the SO₂ reduction varied with operating conditions. When ammonia is added in the main combustor zone, the temperature is much higher than that required for the occurrence of sulphur dioxide-ammonia and sulphur trioxide-ammonia reactions. However, this paper points out the significance of ammonia addition in the reduction of sulphur dioxide.     

Using improved neural network model to analyze RSP,NOx and NO2 levels in urban air in Mong Kok,Hong Kong

As the health impact of air pollutants existing in ambient addresses much attention in recent years, forecasting of airpollutant parameters becomes an important and popular topic inenvironmental science. Airborne pollution is a serious, and willbe a major problem in Hong Kong within the next few years. InHong Kong, Respirable Suspended Particulate (RSP) and NitrogenOxides NO_x and NO₂ are major air pollutants due to thedominant diesel fuel usage by public transportation and heavyvehicles. Hence, the investigation and prediction of the influence and the tendency of these pollutants are ofsignificance to public and the city image. The multi-layerperceptron (MLP) neural network is regarded as a reliable andcost-effective method to achieve such tasks. The works presentedhere involve developing an improved neural network model, whichcombines the principal component analysis (PCA) technique and theradial basis function (RBF) network, and forecasting thepollutant levels and tendencies based in the recorded data. Inthe study, the PCA is firstly used to reduce and orthogonalizethe original input variables (data), these treated variables arethen used as new input vectors in RBF neural network modelestablished for forecasting the pollutant tendencies. Comparingwith the general neural network models, the proposed modelpossesses simpler network architecture, faster training speed,and more satisfactory predicting performance. This improvedmodel is evaluated by using hourly time series of RSP, NO_x and NO₂ concentrations collected at Mong Kok Roadside Gaseous Monitory Station in Hong Kong during the year 2000. By comparing the predicted RSP, NO_x and NO₂ concentrationswith the actual data of these pollutants recorded at the monitorystation, the effectiveness of the proposed model has been proven.Therefore, in authors' opinion, the model presented in the paper is a potential tool in forecasting air quality parameters and hasadvantages over the traditional neural network methods.