Search documents

Browse topics

Document details

Short Term Forecasting of Solar Irradiance by Combining Satellite Data and Numerical Weather Predictions
E. Lorenz, J. Kühnert, D. Heinemann
PV - a Major Electricity Source
Subtopic: PV in the Electricity Mix
Event: 27th European Photovoltaic Solar Energy Conference and Exhibition
Session: 6DO.12.1
4401 - 4405
ISBN: 3-936338-28-0
Paper DOI: 10.4229/27thEUPVSEC2012-6DO.12.1
0,00 EUR
Document(s): paper


With the constantly increasing contribution of PV to the electricity mix, reliable predictions of the expected PV power production are getting more and more important as a basis for management and operation strategies. Here, we present a new approach for irradiance forecasting to improve the operational PV power prediction system of University of Oldenburg and Meteocontrol GmbH by integrating additional data sources, including satellite images and different numerical weather prediction systems. For short term forecast horizons up to four hours ahead, irradiance forecasts based on cloud motion vector fields from satellite images perform significantly better than forecasts of numerical weather prediction systems, demonstrated in an evaluation with 290 measurement stations in Germany. The investigation of an optimized combination of the different data sources in dependence on the forecast horizon is a second focus of the paper and the benefit of this approach is shown.