Application of Multiple Linear Regression Models and Artificial Neural Networks on the Surface Ozone Forecast in the Greater Athens Area, Greece
Application of Multiple Linear Regression Models and Artificial Neural Networks on the Surface Ozone Forecast in the Greater Athens Area, Greece
复制标题
DOI：
10.1155/2012/894714
复制
发表时间：
2012-01-01
影响因子：
2.9
通讯作者：
Paliatsos, A. G.
中科院分区：
文献类型：
--
作者：
Moustris, K. P.;Nastos, P. T.;Paliatsos, A. G.
试图预测未来24小时内，在大雅典地区（GAA）的每日最大表面臭氧浓度。为此，我们应用多元线性回归（MLR）模型对基于人工神经网络（ANN）方法的预测模型。基本气象参数的可用性对于预测臭氧浓度水平至关重要。模型是基于2001年至2005年五年期间GAA（希腊环境、能源和气候变化部网络）内13个监测点记录的气象和空气污染数据。使用适当的统计指标，所构建的模型的性能的评价，清楚地表明，在各个方面，到目前为止的预测模型是人工神经网络模型。这表明，人工神经网络模型可用于向一般人群和主要敏感群体发出警告。
An attempt is made to forecast the daily maximum surface ozone concentration for the next 24 hours, within the greater Athens area (GAA). For this purpose, we applied Multiple Linear Regression (MLR) models against a forecasting model based on Artificial Neural Network (ANN) approach. The availability of basic meteorological parameters is of great importance in order to forecast the ozone's concentration levels. Modelling was based on recorded meteorological and air pollution data from thirteen monitoring sites within the GAA (network of the Hellenic Ministry of the Environment, Energy and Climate Change) over five years from 2001 to 2005. The evaluation of the performance of the constructed models, using appropriate statistical indices, shows clearly that in every aspect, the prognostic model by far is the ANN model. This suggests that the ANN model can be used to issue warnings for the general population and mainly sensitive groups.