Climatological attribution of wind ramp events and their probabilistic forecast based on self-organizing maps - (PPT)

摘要

In this study, Self-organizing maps (SOM) is applied to analyze and establish the relation-ship between atmospheric synoptic patterns over Japan and wind power generation. SOM is employed on sea level pressure data derived from atmospheric reanalysis over the Tohoku region in Japan, whereby a two-dimensional lattice of classified weather patterns (WPs) is obtained. To compare these results with atmospheric data, long-term wind power generation is reconstructed using the high-resolution surface observation network. Our SOM analysis for WPs extracts seven typical patterns that are linked to frequent occurrences of wind ramp events. The result of this study suggests that detailed classification of synoptic circulation patterns can be a useful tool for first-order approximations of the probability of future wind power generation and its variability. Possibility of application to probabilistic forecasts of wind power generation and ramps based on the obtained SOM is discussed.