A Framework for an Outlier Pattern Detection in Weather Forecasting?

Data Mining is the process of discovering new patterns from large data sets. Meteorological data mining is a form of data mining which concerned with finding rare patterns inside largely available weather data. To detect rare Weather pattern is difficult challenge because these rare events are characterized by low occurrence and uncertainty. In this paper, we proposed an Adaptive Markov Chain Algorithm Model which uses an open number of states of Markov Chain to accommodate the dynamic temporality of data. The data is collected with the Tropical Atmosphere Ocean (TAO) array which was developed by the international Tropical Ocean Global Atmosphere (TOGA) program. Data Variables including latitude, longitude, zonal wind, meridional wind, humidity, air temperature and sea surface temperature are considered for identifying climate change patterns in this paper. By adding the Markov property as a global restriction, the granular size of the clusters is determined for optimal performance. Our climate change pattern detection algorithm is proven to be of potential use for climatic and meteorological research as well as research focusing on temporal trends in weather and the consequent changes.