A Decade of Flood Inundation Assessment Under Climate Change in the Auckland Region

Stormwater Conference

Increased rainfall intensity and variability due to climate change are projected to increase the risk of flooding and operation of water infrastructure, such as stormwater collection and flood defence systems. Rainfall is a principal input in rainfall-runoff and floodplain delineation studies. This paper chronicles the efforts from 2000 to 2009 to downscale global climate models to local scale rainfall intensities and assess the expected flood inundation under climate change. Existing climate design storms are based on depth-duration-frequency relationships that are documented in the Auckland Regional Council’s Technical Publication 108 (TP108). TP108 represents adaptation of the Soil Conservation Service rainfall-runoff method to the Auckland region. It uses a 24-hour nested alternating block centred probabilistic (Chicago) design storm. Twenty-four hour design storms were derived from the climate change rainfall scenarios. Two global climate downscaling exercises were undertaken. In 2000, the global climate models based on the Intergovernmental Panel on Climate Change (IPCC2) were downscaled and combined with extreme value theory to generate depth-duration-frequency statistics for the Albert Park raingauge in Auckland City, the gauge with the longest period of record in the Auckland region. The projected depth-duration-frequencies in 2050 varied from a 29.8% increase at 10- minutes to 16.8% increase at 24-hours as compared to the TP108 1% annual exceedance probability (AEP) storm event. In 2004, global climate models again were downscaled for the Auckland region using IPCC3 scenarios. The analysis included extreme value theory and consideration of atmospheric moisture content. The projected depth-durations varied from existing climate by 16.8% for the 1% AEP storm event for the most probable scenario in 2090. The most probable scenario for rainfall and sea level rise were simulated in a number of catchments in the Auckland region using DHI Water and Environment software (MOUSE, MIKE-11, MIKE-21 and MIKE-FLOOD): (1) three highly urbanised, small residential and industrial catchments in Auckland City, (2) a medium sized residential catchment that is developing with generally larger individual property sizes in North Shore City, and (3) a larger, mixed land use (predominantly rural) in the Papakura Stream catchment that bisects Papakaura District Council and Manukau City Council. The studies show mixed response in predicted flood inundation under climate change. Predicted flood levels remained largely the same in two of the Auckland City study areas, principally due to the relatively steep nature of the catchments, lack of storage and quick rainfall-runoff response. In the Papakura Stream, simulations of updated climate change intensities for 2090, predict an expansion of the expected floodplains at lower extreme rainfall events (10% and 20% AEP) yet similar floodplains at higher AEP rainfall events (1% AEP). However, simulations of Lucas Creek and Onehunga predict a marked increase in the floodplains in the flatter portions as compared to the existing floodplains. Overall, the findings debunk the general understanding that future climate rainfall translates into more extensive flood extents in every case. The implications for future climate rainfall is that it must be assessed carefully and systematically, and considered together with land use, drainage infrastructure, and the provision of the level of service desired (e.g., provision of primary stormwater network and protection from extreme events). Sea level rise was considered but is not reported herein.

Conference Papers Resource - Conference Papers Stormwater

Session 2 3 T. van Kalken.pdf

pdf
3 MB
06 Jul 2016