Predicting
the future?
Comparative Forecasting and a Test for Persistence in the El Nino
Southern Oscillation and new ideas
Daniel S. Zachary
John Hopkins University
noon 27 May Conference Center
Abstract:
We present a statistical analysis on El Nino. Analysis is done for two
single-indicator forecasting methods for the El Nino Southern
Oscillation based on oscillation persistence. We use the Southern
Oscillation Index (SOI) to produce short term 5 month forecasts and a
Bayesian approach to explore SOI persistence, with results compared to
a benchmarking Taylor Series expansion. We find signal persistence is
important when forecasting more than a few months and the models
presented may provide a relatively simple approach to environmental
risk forecasting in situations where the underlying phenomenon exhibits
substantial persistence. We update the analysis with new results for
2014. We supplement the discussion with some new ideas and initial
results that are expected to improve the analysis and strengthen the
statistical forecasting method.