A fuzzy neural network model to forecast the percent cloud coverage and cloud top temperature maps

摘要

Atmospheric processes are highly nonlinear. A small group at the METU inAnkara has been working on a fuzzy data driven generic model of nonlinearprocesses. The model developed is called the Middle East TechnicalUniversity Fuzzy Neural Network Model (METU-FNN-M). The METU-FNN-M consistsof a Fuzzy Inference System (METU-FIS), a data driven Neural Network module(METU-FNN) of one hidden layer and several neurons, and a mapping module,which employs the Bezier Surface Mapping technique. In this paper, thepercent cloud coverage (%CC) and cloud top temperatures (CTT) areforecast one month ahead of time at 96 grid locations. The probableinfluence of cosmic rays and sunspot numbers on cloudiness is considered byusing the METU-FNN-M.