Benchmarking prediction skill in binary El Niño forecasts

Abstract Reliable El Niño Southern Oscillation (ENSO) prediction at seasonal-to-interannual lead times would be critical for different stakeholders to conduct suitable management. In recent years, new methods combining climate network analysis with claim that they can predict up 1 year in advance by overcoming the spring barrier problem (SPB). Usually this kind of method develops an index representing relationship between nodes related basins, and crossing a certain threshold is taken as warning event next few months. How well performs should measured order estimate any improvements. However, amount recordings available data limited, therefore it difficult validate whether these are truly predictive or their success merely result chance. We propose benchmarking surrogate quantitative forecast validation small sets. apply naïve events based on Index (ONI) time series, where we build data-based scheme using series itself input. assess network-based method, reproduce two forecasts our compare skill all these. Our benchmark shows ONI input does not work moderate times, while least one has above chance about year.