Environment pollution monitoring and control is very big problem for the whole world. Taking decision in the environment is becoming more challenging. The aim of this paper is to present the challenges surrounding environmental data sets and to address these in order to develop solutions. Environmental data sets present a number of data management challenges including data collection, integration, quality and data mining. Environment data sets are also very dynamic and this presents additional challenges ranging from data gathering to data integration, particularly as these data sets are normally very large and expanding continuously. Statistical methods are very effective and economical way to analyze small, static data sets but they are not applicable for dynamic, real-time and large data sets. The use of data mining methods to discover hidden knowledge in large datasets therefore presents great potential to improve environmental management decisions. A representative environmental data set from quantitative air quality monitoring instruments has been assessed and will be used to demonstrate some of the issues in applying data mining approaches to poor data quality.