Evidence for Using Lagged Climate Indices to Forecast Australian Seasonal Rainfall

Abstract

Lagged oceanic and atmospheric climate indices are potentially useful predictors of seasonal rainfall totals. A rigorous Bayesian joint probability modeling approach is applied to find the cross-validation predictive densities of gridded Australian seasonal rainfall totals using lagged climate indices as predictors over the period of 1950–2009. The evidence supporting the use of each climate index as a predictor of seasonal rainfall is quantified by the pseudo-Bayes factor based on cross-validation predictive densities. The evidence strongly supports the use of climate indices from the Pacific region with weaker, but positive, evidence for the use of climate indices from the Indian region and the extratropical region. The spatial structure and seasonal variation of the evidence for each climate index is mapped and compared. Spatially, the strongest supporting evidence is found for forecasting in northern and eastern Australia. Seasonally, the strongest evidence is found from August–October to November–January and the weakest evidence is found from March–May to May–July. In some regions and seasons, there is little evidence supporting the use of climate indices for forecasting seasonal rainfall. Climate indices derived from sea surface temperature anomalies in the Pacific region show stronger persistence in the relationship with Australian seasonal rainfall totals than climate indices derived from sea surface temperature anomalies in the Indian region. Climate indices derived from atmospheric variables are also strongly supported, provided they represent the large-scale circulation. Many climate indices are found to show similar supporting evidence for forecasting Australian seasonal rainfall, leading to the prospect of combining climate indices in multiple predictor models and/or model averaging.

Describing rainfall in northern Australia using multiple climate indices

“…Ashok et al, 2007;Kamruzzaman et al, 2013;Schepen et al, 2012). A climate index is a numerical value which provides a measure of the change away from the mean in an oscillatory climate system (National Center for Atmospheric Research, 2012).…”

“…For example, Australian rainfall is well correlated with the Southern Oscillation Index (SOI), which is in turn associated with ENSO (e.g. Murphy and Ribbe, 2004;Risbey et al, 2009;Schepen et al, 2012).…”

“…Despite this there has been a paucity of research undertaken on the climatic influences of rainfall in northern Australia and savannas; however, there have been a number of continental-scale analyses (e.g. Nicholls, 1989;Risbey et al, 2009;Schepen et al, 2012).…”

Assessment of methods for seasonal streamflow forecasting in the Upper Indus Basin of Pakistan

“…The BJP offers state-of-the-art capabilities for developing seasonal forecast models that optimally utilise information available on antecedent catchment conditions, large-scale climate forcing (through climate indices) and flow forecast scenarios from 5 hydrological models Robertson and Wang, 2012;Schepen et al, 2012;Wang and Robertson, 2011). The BJP models simulate predictor-predictand relationships using conditional multivariate normal distributions, with predictor and predictand data transformed to normal using either a log-sinh or Yeo-Johnson (Yeo and Johnson, 2000) transformation.…”

“…Statistical approaches relate antecedent catchment conditions and/or climate indices to streamflow using techniques such as multiple linear regression (Maurer and Lettenmaier, 2003). Statistical approaches require predictor-predictand records of sufficient 20 length to determine robust relationships, stationarity in the relationships, and rigorous cross-validation to avoid over-fitting or an inflated skill assessment Schepen et al, 2012). Dynamical approaches use hydrological models initialised with observed inputs up to the beginning of the forecast season (to account for antecedent conditions), that can be driven either by historical or climate modelled precipitation and temperature forecasts (Yuan et al, 2015;Zheng et al, 2013).…”