The abundant aspects of big data and it's technology are increasing due to new methods of fetching data and diverse needs. Meteorological data is also the source of big data in terms of volume, variety, veracity and velocity, and it includes structured, unstructured and hybrid forms. This paper aims to apply Hadoop architecture and MapReduce algorithm into meteorological big data. It also describes the difficult situations of dealing complex meteorological data by present computer systems and how we replace it with distributed computing. Automatic text classification is used to divide the text into predefined classes, which help individuals utilize information with an effective method. This paper gives an idea about how to implement meteorological text classification through big data approach. Traditional text classification requires a high cost of domain knowledge, but it has low classification efficiency, which can't satisfy the requirement of large documents processing. Distributed parallel method is well developed to solve the problem of a large number of documents classification. Finally, this paper describes big data applications of meteorological big data combined with agriculture, transportation and power grid industries, and illustrates how we analyze tourism industry by our meteorological tourism big data platform.