Reasearchers from the HPCWE project presented a poster at the virtual EERA JP Wind event.
A wind resource assessment for the Rune experiment area at the west coast of Denmark. The study experiments data fusion in both space and time domains using high-fidelity measurements by lidar, but are expensive and scarce, particularly for offshore sites. On the other hand, numerical simulations using, mesoscale models, for example the Weather Research and Forecasting (WRF) Model, generate temporally and spatially continuous data with relatively low fidelity. We first perform a temporal fusion using multi-fidelity Gaussian process regression to combine the intermittent and short measurement with the continuous and long simulation data at an onshore location. Following up, spatial fusion using neural networks is conducted to project the data from the onshore location to an offshore one. The result showed that the proposed technique has an accurate offshore wind resource assessment of about 7% margin error.