The variation of sunlight throughout the year, caused by the tilt of the Earth’s axis and its spherical shape, influences surface heating, temperatures, atmospheric currents, and rainfall distribution. Solar energy, being clean and sustainable, has many applications. Using it efficiently requires understanding its temporal distribution. The daily accumulation of solar irradiation follows a sigmoidal pattern, which can be described by non-linear regression models. These models offer practical and innovative interpretations for renewable energy projects, filling a gap in the scientific literature. This study was conducted to evaluate the suitability of the Logistic, Gompertz, and Von Bertalanffy non-linear regression models to describe the accumulation of daily global solar irradiation (kJ m-²). Data on global irradiation accumulated throughout the day at ground level, during the equinoxes and solstices in Belo Horizonte/MG, were obtained from the Brazilian National Institute of Meteorology (INMET) through the Annual Historical Data for the period 2018 to 2022. The Logistic and Gompertz models met the evaluated statistical assumptions, unlike the Von Bertalanffy model. The Logistic model achieved a better fit for the equinox and winter solstice data, while the Gompertz model was more suitable for the summer solstice data. The asymptotic estimates (α) of the Logistic model were statistically equivalent for the equinoxes but not for the solstices. The β parameter (abscissa of the inflection point) was statistically equivalent for the solstices and the spring equinox.
