Using Analog Ensembles with Alternative Metrics for Hindcasting with Multistations

[摘要] This study concerns making weather predictions for a location where no data is available, usingmeteorological datasets from nearby stations. The hindcast with multiple stations is performed withdifferent variants of the Analog Ensemble (AnEn) method. In addition to the traditional Monachemetric used to identify analogs in datasets from one or two stations, several new metrics are explored, namely cosine similarity, normalization, and k-means clustering. These were analyzed andbenchmarked to find the ones that bring improvements. The best results were obtained with thek-means metric, yielding between 3% and 30% of lower quadratic error when compared againstthe Monache metric. Also, by making the predictors to include two stations, the performance ofthe hindcast improved, decreasing the error up to 16%, depending on the correlation between thepredictor stations.

[学科分类] 环境工程

[关键词] Analog Ensembles;Metrics;Hindcasting;Time series;Meteorological data