Seasonal weather forecast by fuzzy logic method Istanbul application

The atmosphere deeply affects our lives with all the events and phenomena that it has. The phrase "deeply affects our life" is perhaps a shallow expression to describe the importance of the atmosphere. Because all living creatures build their living by considering the effects – situtaions of atmospheric conditions. This process that is begining with construction continues with efforts to protect from possible disasters and events, or challenge to be survive after events with minimum damage. Through the centuries of human experiences, the transfer of experienced events, the events we learned with the explanation of observational inferences, and the development of technology, we have tried to examine them with the help of many methods and models in computer. The fact that our atmosphere has a chaotic structure so it show itselfs as greatest obstacle. Especially in long-term forecasts, this hurdle is felt more and more. The seasonal weather forecast is an issue that has become more and more important in recent years. Work on seasonal weather forecasting in many parts of the world is ongoing at the same time. The fact that considering in long term plans and investments, insreasing population growth and urbanization accordingly triggered these studies. The seasonal weather forecast differs from familiar weather forecast becasuse it doesnt contain sudden cahanges and it contains summary statistics for future seasons. By making the seasonal weather forecast successfully, all sectors interested in human comfort will be able to make healthy action plans in advance and take precautionary measures to prevent possible loss of life and property. In this study, Fuzzy Logic modeling is done and it is aimed to compare the results with both measurement data and other models. Why is the Fuzzy Model selected? Atmosphere has a chaotic structure that plays main role in studies lack of certainty in predicting atmospheric events. Formulas or mathematical equations in various assumptions causes the event to diverge from its true state of event. For this reason Sugeno Software is used in the scope of Fuzzy Logic. A model is occured by using temperature, relative humidity and rainfall data of İstanbul Kandilli Observatory station between 1960 and 2016. The model was set up to have two inputs, one output. Inputs are relative humidity and temperature, and output is rainfall data. In the model tht is created using temperature, relative humidity and precipitation data, 5 memberships is set for temperature and relative humidity data for daily data and monthly total data. The numbers of membership groups have been determined as a result of various trials. Taking 5 intervals for each of the temperature and relative humidity parameter according to the result are that the error is low in the training and test phase, and R2 is high in comparison with the measured data and predicted data. Different input data (temperature and relative humidity) each have 5 varied membership. xxii Firstly membership groups determination of daily data sets, membership groups of temperature data; cold (-10 20). The membership groups of the relative humidity data are; Low (x < 40), medium (35 1000). The membership groups of the relative humidity data are; Low (x < 1900), medium (1800