Accurate localized short term weather prediction for renewables planning

Abstract

Short-term prediction of meteorological variables is important for many applications. For example, many `smart grid' planning and control scenarios rely on accurate short term prediction of renewable energy generation, which in turn requires accurate forecasts of wind-speed, cloud-cover, and other such variables. Accurate short-term weather forecasting therefore enables smooth integration of renewables into future intelligent power systems. Weather forecasting at a specific location is currently achieved by numerical weather prediction (NWP), or by statistical models built from local time series data, or by a hybrid of these two methods broadly known as `downscaling'. We introduce a new data-intensive approach to localized short-term weather prediction that relies on harvesting multiple freely available observations and forecasts pertaining to the wider geographic region. Our hypothesis is that NWP-based forecast resources, despite the benefit of a dynamical physics-based model, tend to be only sparsely informed by observation-based inputs at a local level, while statistical downscaling models, though locally well-informed, invariably miss the opportunity to include rich additional data sources concerning the wider local region. By harvesting the data stream of multiple forecasts and observations from the wider local region we expect to achieve better accuracy than available otherwise. We describe the approach and demonstrate results for three locations, focusing on the 1hr-24hrs ahead forecasting of variables crucial for renewables forecasting. This work is part of the ORIGIN EU FP7 project (www.origin-concept.eu) and the weather forecasting approach, used in ORIGIN as input for both demand and renewables prediction, began live operation (initially for three European locations) in October 2014.