ARIMA models in the rain attenuation prediction in a Mexican tropical area

Rain attenuation is a dominant impairment in frequencies greater than 10 GHz, raindrops absorbs electromagnetic energy, turning it into thermal energy, furthermore dispersing the energy outside the original direction; rain effect over propagation depends on rain rate, its distribution and the raindrops size. Water vapor, fog, clouds and gasses influence even attenuation over the trajectory, rain fading is the most important and then of essential interest in microwave communications. Statistical analysis over specific regions, where rainfall predictions are necessary, becomes a must in microwave link design. Rain prediction is made through phenomenological analysis, like those founded in literature, to establish models to be used in attenuation analysis, and analyzing long term rainfall over the studied region to forecasting. This paper uses rain data in Mexico tropical area, collected by Comision Nacional del Agua (CNA) and the Institute Mexicano de Tecnologia del Agua, from Villahermosa Tabasco region, to establish a rain prediction model using ARIMA models (auto-regressive integrated moving average) and the MINITAB software. The model is used to predict rainfall in Villahermosa and near sites to determine its effectiveness.