IrMaSet: Intelligent Weather Forecaster System for Hyper-Local Renewable Energies

Weather forecasting plays a vital role in estimating energy generation from variable renewable energy sources (VRES). Current weather forecast methods for estimating energy from VRES are typically at a scale of a few kilometers, which is not the fine spatial resolution needed for distributed energy planning. In this paper, we propose IrMaSet (an intelligent weather forecaster system) that integrates weather data obtained from existing massive sensing infrastructure and processes data using state-of-the-art communication and computational technologies. IrMaSet generates hyper-local real-time weather and forecast information (HyReF) and enables accurate estimation of the amount of electricity to be generated by VRES, resulting in the optimal management of microgrid energy systems. We present challenges and possible solutions for deploying IrMaSet and demonstrate the feasibility of IrMaSet in saving energy, and reducing costs and carbon emissions, using meteorological data including wind and solar radiation from eleven official monitoring stations (OMS) in greater Helsinki, Finland.