Hydrology and Water Resources Symposium 1988 ANU, Canberra, 1-3 February 1988 The Sydney Floods of 1986: Warnings, Damages, Policy and the Future J.W. HANDMER Post Doctural Fellow, Australian National University D.I. SMITH Senior Fellow, Australian National University T.L. LUSTIG Environmental Management SUMMARY Following the ~ugust1986 flood, detai led surveys of f lood-prone properties i n the Georges River, Prospect and Toongabbie Creek catchments of western S.ydney were undertaken to: estimate tangihle and intangible flood damage; estimate the damages averted by the actions of those flooded; and examine the flood warning system. The damage data and analysis of the flood warning svstem w i l l he presented together .with the implications and recommendations for urban floodplain management policy. I During August 5 and 6 1986, S,ydney experienced i t s worst flooding for decades. Record or near record rainfalls were recorded at stations in western and southern Sydney and in parts of the Georges River catchment. These intense storms were not forecast by the Bureau of Meteorology. Rut, b,y late afternoon of August 5, the Georges River was rising and man.y metropolitan areas were starting to flood from urban creeks and storm-water surcharge. The River peaked at Liverpool early in the morning of August 6 reaching a height of 4.5m - a 'major' flood. Much o f F a i r f i e l d was also flooded by the Prospect Creek system and Toongabbie i n Parramatta was inundated b.y Toongabbie Creek. All areas were subject to flooding by overflow or back-up from storm and septic sewers. In all areas flooding occurred with little or no warning. A total of 2248 residential properties were flooded, 1000 of them having water over their floors, and 390 commercial or industrial premises of which 215 had water over their floors. In addition, a l l forms of transport were severely disrupted and i n f r a s t r u c t u r e was damaged. Immediately after the flood, Environmental Management in a s s o c i a t i on with the Centre for Resource and Environmental Studies (CRES), of the Australian National University was commissioned .by the NSW Public Works Department and rlepartment of Water Resources to: assess the social and economic losses fran the flood; examine the functioning of the flood warning and emergency system; and assess the consequences of larger floods. This paper presents a selection of the key results. Full detai 1s are contained i n the Project Report (Smith et al, 1987). / 2 METHODOLOGY I nata were obtained through field and questionnaire surveys and through access t o the relevant records r of the Department of Youth and Community Services (YACS) and GI0 insurance claims for flooding. The following surveys were undertaken: a social survey of 40 households which had experienced overfloor flooding; a media survey which included nearly all Sydney's radio and TV stations; interviews with officia 1s from the warning and emergency management system, affected counci 1s and other overnment authorities; loss adjusters surveys of i h and 23 business enterprises t o develop stage damage curves f o r ANUFLOOD, 38 large enterprises mostly not flooded were also surveyed; and t h e ANUFLOOO external hui l di ng survey of all properties within the 1% floodplain. ANlJFLOOn is a commercial1,y available computer package designed t o assess t a n g i hle f lood damages to buildings and contents and the benefits of certain flood mitigation notions (Smith et al, 1986). Indirect losses were assumed t o he 15% of direct losses for the residential sector and 55% from the busi ness sector. Intangi h le damages are d i vided into .disruption, i 11-health and deaths with results for disruption and i 11-health presented i n terms of 'lost time' (Handmer e t al, 1996). In this study inputs and therefore results for the 1986 flood are i n terms of actual rather than p o t e n t i a l damage. 3 THE WARNIW ANn EMERGENCY Y4N4GEblENT SYYTFY 1 3.1 River forecasts I In NSW the Sureau of Yeteorology provides quantitative forecasts of flood hei qhts for qost of the state's major rivers including the Georqes River. In arldition, it issues warnings of weather likely to lead to flooding for the whole state. It is important to note that only major rivers, such as the Georges, are covered by formal flood warninq systems. Those parts of 5ydney drained hy urhan creeks and stormwater channels are covered hy weather forecasts only. Thus, the flood~lains of Toongahbie and Prospect Creeks, representing 60% of the properties at risk fran flooding in the stulv area, are not covered hy a flood warning system. The Sureau aims to provide 6-17 hours warning for I the Georges River at Liverpool. To do this it relies on 5 manual rain gallges which are read h . ~ volunteers who telephone i n their ohservations at Specified intervals during a f lnod. llsing these data (and predicted rainfall) a computerised river- model estiqates flood heights for 1-i verpool. Ilnfortunate ly, the gauge netuirk performed ~oorly. Only one of the 5 readers responded as required, two stations fai led to report at a1 1 and the other two experienced intermittent telephone f a i lures. nespite the absence of important data the Rureau was ahle t o produce flood forecasts for the Georges River at Liverpool. Yowever, the warnings, issued ever,y 3 hours, consistently underestimated the rate of river rise tending to predict river heights some 3-6 hours later than actually occurred. The first warning was issued at about 3.30pm August 5, b.y which time some areas were already heginning t o