Nowadays the number of devices connected to internet which offer the possibility to collect data is increasing. The interconnectivity of these new sensors favors the creation of sustainable cities, in which the optimization of resources is based on the collected data. These sensors are also a big source of information for forecasting future values. In this work we present an Edge Computing approach for the analysis and forecasting of rainfall data that is later validated on the CITI Laboratory Youpi Platform. To this end, we built a container image with the necessary tools and libraries to use the time series prediction models SARIMA and Prophet on ARMv7 architectures. A Raspberry Pi 3 node was chosen to evaluate performance on an Edge Computing device. Colombia was chosen due to its tropical location and its variant geography which present a wide range of historical rainfall data. The data we used to train our models consisted on the historical mesures from sensors deployed in Colombia by the "Instituto de Hidrología, Meteorología y es-tudios Ambientales de Colombia-IDEAM". In first place, we selected Bucaramanga to study its data sensors and to define the well-suited parameters for SARIMA and Prophet trend models. The comparison between them presented a high degree of similarity, offering a good prediction of dry and wet seasons. Thereafter, the SARIMA and Prophet model of Bucaramanga were used to observe its adaptability to the cities of Bogotá and Medellín, getting a successful outcome at seasonal predictions. After this estimation of the SARIMA parameters and its analysis offline, a container image was created to simplify and speed up the models implementation in the devices for predicting the two years monthly rainfall for Bucaramanga, Bogotá and Medellín. The container is available for armv7 architectures, that is usually used for IoT nodes on the Youpi platform. The proposed model allows to create a network of sensors, with distributed analysis capacity, that improve the prevention of flood or drought emergencies in Smart Cities on Colombia, helping to manage resources for agriculture or prevent catastrophes.