Weather Prediction Based on Big Data Using Hadoop Map Reduce Technique

Abstract: The assists of big data collects large volume of data, it is great computational challenge for the big data Hadoop which uses map reduce to maintain and process this data and also helps to extract useful information in an efficient manner . Keeping these things in mind there is need for designing system architecture that helps to predict weather forecast for future. It helps people to take decision in advance for their any outdoor events. In our proposed architecture there are 3 main units, such as: Data acquisition unit, Data processing unit, Data analysis and decision unit. DAU collects data from the satellite and sends to the different base station and finally stores in the national climatic data centre (NCDC).DPU it plays key role in the architecture, it takes the data from the NCDC and HDFS takes the data from this NCDC and store in it and later processed by MapReduce framework and produce required output .DADU it performs the compilation based on the results produced by the DPU and makes decision to produce required results. Hence in our proposed architecture we are focusing on the offline data that is stored in the NCDC to predict the weather analysis by Hadoop map reduce framework the output of result consists of, minimum temperature, maximum temperature, number of hot days and cold days and also predict future weather forecast, which brings the great significance of our work.

Keywords: Big data, HDFS, Hadoop, YARN, MAPREDUCE.