Recent developments in climate science: implications for Australian flood guidelines

Abstract

A fundamental assumption in flood estimation is that climate data is stationary, such that while the weather at any point in time will vary randomly, the underlying climate statistics (including both averages and extremes for rainfall and stream flow)will remain constant irrespective of the period of record. This paper outlines a number of recent developments in our understanding of natural climate variability and human induced climate change which challenge this assumption, and which will require new approaches to flood estimation. Recent research on both natural climate variability and human induced climate change suggest that failure to take non-stationarity into account in flood assessments will heighten the possibility of a systematic underestimation in future flood risk across much of Australia. For example, a recently discovered naturally occurring climate mode known as the Inter-decadal Pacific Oscillation (IPO) has been shown to significantly affect estimates of design rainfall intensity. This mode varies at a frequency of 15-35 years, and has been biased towards the drier phase for much of the instrumental record suggesting a possible downward bias in flood risk. Humaninduced climate change will also have an increasing impact on various aspects of design flood estimation, due to projected increases in design rainfall intensity, changes to antecedent catchment conditions, and increases in tail water levels due to sea level rise and storm surge. This paper recommends that a new framework for future flood assessments be developed that accounts for both the observed non-stationarity of the historic data, as well as the implications of future climate change. Specifically, this paper recommends that recently developed methodologies need to be brought into more widespread practice in order to reconcile our understanding of natural climate variability with human-induced climate change projections. The development of future flood guidelines that address these issues, and in particular the release of the forthcoming version of Australian Rainfall and Runoff, will be essential in the transition to a more non-stationarity paradigm in flood hydrology.