A simple and accurate mathematical model for estimating maximum and minimum daily environmental temperatures in a year

Environmental forecasting temperature models have high application value for building design, environmental impact assessment, and human wellness, as they apply to the development of human comfort devices. The current models used in this field are complex and involve many parameters, resulting in a considerable amount of data and a complicated modeling process. This research proposes two models: First, a type-Gaussian model with four parameters calculated to fit the environmental temperature. The proposed model employs the maximum and minimum temperatures (TMax, TMin) registered in a year for previously selected city or town and the relative time (day number of the year, 1-365). Second: A cosenoidal model to predict the continuous temperature in the daynight cycle of the selected day. This model requires TMax and TMin previously estimated via the Gaussian model and the hour in the day-night cycle at which the lowest temperature occurs. The Gaussian model is validated with data from different cities, achieving a high degree of confidence with a CV(RMSE) less than 2.0% in the cases analyzed. This model was compared to Artificial Neural Networks and other mathematical models previously published to confirm its functionality. The proposed models can be used in agriculture, civil engineering, food industry, power plants, biology, heat transfer, transportation, etc.