Australian Government - Bureau of Meteorology Home | About Us | Contacts | Help | Feedback |

Global | Australia | NSW | Vic. | Qld | WA | SA | Tas. | ACT | NT | Ant. |

Weather & Warnings | Hydrology | Climate | Numerical Prediction | About Services | Learn About Meteorology | Registered User Services |

Cool season storms | Warm season storms | Heavy rain & flooding | Tropical Cyclones | Heatwaves | Bushfires | References | Acknowledgements


Heavy rain and Flooding

Frequency of heavy rainfall
Mechanisms
Flooding
Historical trends
Storm surge


Frequency of heavy rainfall

Perth is not commonly associated with heavy rain and flood events particularly in recent decades. Nevertheless, there is sufficient historical evidence to suggest that the flood risk in Perth is real and serious and the potential impact made worse by the lack of events in recent memory.

Table 1 shows the frequency distribution of daily rainfall by month using composite data from the official Perth site from 1880 to May 2004. The heaviest rainfall occurs in June and July. Of the 78 events when daily rainfall has exceeded 50 mm, half occurred in June and July, and only 15% occurred in the months from September to March. However the two highest rainfall amounts occurred in February (121 mm on 9/2/1992) and January (104 mm on 22/01/2000).

Table 2 shows the peak daily rainfall for the ten highest rainfall events in the Lower West District. This district is approximately centred on Perth, extending for about 150 km in a north-south direction and 50 km inland. While most of these rainfall events occurred from May to July, the top three events occurred in either February or March. The number of sites recording more than 100 mm during the same event is also shown. However, it should be noted that the number of sites recording rainfall has gradually increased over time.

Table 1. Frequency analysis of daily rainfall at Perth, 1880-2003.

Rainfall
Jan
Feb
Mar
April
May
June
July
Aug
Sept
Oct
Nov
Dec
Total
0
3518
3223
3377
2827
2169
1590
1577
1721
1975
2490
2923
3326
30716
0.1 to 5
305
259
427
624
905
966
1023
1095
1088
931
638
419
8680
5 to 10
27
37
58
157
324
398
506
457
370
233
95
59
2721
10 to 15
12
14
36
77
181
258
276
256
143
91
33
19
1396
15 to 20
5
11
15
40
119
186
183
136
93
51
11
13
863
20 to 25
3
7
7
23
73
117
127
85
35
23
15
4
519
25 to 30
0
2
8
12
42
72
75
63
26
9
2
2
313
30 to 35
0
1
3
8
30
65
51
26
10
9
1
0
204
35 to 40
1
0
2
3
26
43
18
14
3
3
2
0
115
40 to 45
1
1
2
2
12
13
18
9
3
1
0
1
63
45 to 50
1
0
0
2
11
22
4
5
2
1
0
1
49
50 to 55
1
1
1
1
7
7
3
3
2
0
0
0
26
55 to 60
0
0
0
1
3
5
5
1
0
1
0
0
16
60 to 65
0
1
0
1
1
2
3
2
0
0
0
0
10
65 to 70
0
0
0
2
2
3
3
0
0
1
0
0
11
70 to 75
0
0
0
0
0
1
0
2
0
0
0
0
3
75 to 80
0
0
1
0
1
2
1
0
0
0
0
0
5
80 to 85
0
0
0
0
0
0
0
0
0
0
0
0
0
85 to 90
0
1
0
0
0
1
1
0
0
0
0
0
3
90 to 95
0
0
0
0
0
0
1
0
0
0
0
0
1
95 to 100
0
0
0
0
0
1
0
0
0
0
0
0
1
100 to 105
1
0
0
0
0
0
0
0
0
0
0
0
1
105 to 110
0
0
0
0
0
0
0
0
0
0
0
0
0
110 to 115
0
0
0
0
0
0
0
0
0
0
0
0
0
115 to 120
0
0
0
0
0
0
0
0
0
0
0
0
0
120 to 125
0
1
0
0
0
0
0
0
0
0
0
0
1
No. of days
3875
3559
3937
3780
3906
3752
3875
3875
3750
3844
3720
3844
45717
No. of Years
125
126
127
126
126
125
125
125
125
124
124
124
125.17
Rainfall >25 mm
5
8
17
32
135
237
183
125
46
25
5
4
822
Rainfall >50 mm
2
4
2
5
14
22
17
8
2
2
0
0
78

 

Table 2. Highest ten daily rainfall events in the Lower West District.

Peak rainfall for the event (mm)

Site of peak rainfall

Date of event

No. other sites recording > 100 mm for the same event

230
Medina
9/02/1992
20
196.3
Bindoon
9/03/1934
17
190.5
Wannamal
16/02/1955
15
164.8
Rockingham
11/06/1945
8
159.2
Serpentine
29/07/1987
22
155.8
Mundaring
30/07/2001
13
154.7
Rottnest Island
21/06/1984
2
145.8
Karnet
26/06/1967
7
138.9
Highgate
10/06/1920
8
137.0
Lupin Valley
2/05/1953
6

Heavy rainfall events are not confined to a twenty-four hour period. Intense rainfall over short durations can result in flash flooding; causing gutters and drains to overflow in five to ten minutes. Perth's sandy soils on the coastal plain and relative lack of topography diminish the impact of heavy rain. However, Perth's infrastructure is also not designed to cope with very heavy rainfall, especially when compared to tropical locations. Every year there are reports of properties becoming inundated due to short-term heavy rainfall. On 30 March 2003, for example, the State Emergency Service responded to over 100 requests for flood-related damage when a thunderstorm caused 29 mm of rain in half an hour at Perth Airport.

To assist in hydrological design procedures the frequency analysis of rainfall data can be indicated by derived Intensity-Frequency-Duration (IFD) design rainfall curves. IFD curves for Perth are shown in Figure 1. These show the rainfall rate for periods ranging from 5 minutes to 72 hours, which correspond to average recurrence intervals (ARI) ranging from 1 year to 100 years. The ARI of 2 years, for example, corresponds with 20 mm of rainfall in one hour, and an ARI of 50 years corresponds with the occurrence of 40 mm of rainfall in one hour.

Intensity Frequency Duration curves for Perth

Figure 1. Intensity Frequency Duration Curves for Perth.

Mechanisms

Heavy rainfall in the Perth region can be caused by a range of meteorological mechanisms, but are usually associated in some way with one of the following:

1. A mid-latitude low and cold front embedded in the westerlies (see winter storms).
The most typical pattern for heavy rain is for a front to interact with a tropical moisture source to produce a band of cloud stretching from the northwest to southeast as shown in Figure 2 (July 2001). A slow-moving low to the south or southwest can result in this rainband causing extensive heavy rain or even a succession of rainbands as several fronts sweep across southern Western Australia. These rainbands have the potential to cause widespread rain although it is common for particularly heavy rain to occur in a strip owing to strong convergence in a northwest to southeast orientation.

2. A tropical low or cyclone off the west coast and moving south (see tropical cyclones ).

3. Thunderstorms (see warm season thunderstorms).

satellite image showing rainband over southwest

Figure 2. Satellite image of a rainband responsible for heavy rain in Perth in July 2001 (Courtesy of the Japan Meteorological Agency).

Flooding

Typically, flooding occurs in the form of:
1. Localised flash flooding as a result of heavy rain over a short period.
2. Inundation of low-lying land after prolonged rain. Improved drainage has reduced the risk of flooding in these areas that were once prone to seasonal flooding.
3. Flooding caused by a rise in river height on the Swan and Canning Rivers. Although uncommon, major river floods have the most significant impact on Perth.
4. Inundation of low-lying coastal areas caused by high tides caused by storm surge.

A heavy rain event does not automatically result in a rise in river levels sufficiently high to flood surrounding areas. While properties may be affected by flash-flooding, more significant flooding on the Swan and Canning Rivers is associated with heavy rainfall over prolonged periods. This is demonstrated in Figure 3 which shows rainfall over a seventy day period for four different rain and flood events in 1926, 1945, 1963 and 2001. The rainfall event in July 2001 recorded the heaviest daily rainfall at 99 mm. However, this rainfall event did not produce a major flood in Perth as just 184 mm fell in the previous seven weeks. Major floods did occur in 1926, 1945 and 1963 as heavy rain over the proceeding period had already raised river levels. In 1945 a total of 842 mm of rainfall fell over a seventy day period compared to just 184 mm prior to the 2001 event.

 

graph showing rainfall for four different rain events

Figure 3. Rainfall over seventy days for the major floods in 1926, 1945 and 1963, and the rainfall event of July 2001 showing the importance of accumulated rainfall. Although Perth registered a daily fall of 99 mm in 2001, relatively dry antecedent conditions prevented significant flooding.

Indeed rainfall can be important as much as twelve months prior to the event. The 1963 flood, for example, influenced the consequences of rain the following year. The Collie flood in 1964 was rated as a one in eighty year event, yet the rainfall which fell was only rated as a one in twenty year event.

The requirement for preceding rain events to saturate the catchment and raise river levels provides an explanation as to why flooding on the Swan River is primarily a winter-time phenomenon. Those uncommon warm-season rainfall events generally do not cause major floods on the Swan River because of the river's capacity to discharge large volumes of water when existing river levels are low. This is not the case for some other towns such as Moora and Greenough that are located on smaller river catchments more susceptible to short-term heavy rain. Rainfall associated with tropical cyclone Elaine, for example, flooded the town of Moora in March 1999.

Very exceptional rainfall events can however cause flooding on large river catchments such as the Swan River. In March 1934, the Swan River reportedly rose 5.8 m in less than eight hours at Guildford in response to heavy rain (134 mm in just four hours at Clackline) associated with a tropical cyclone. In January 1982, 250 mm of rain fell in association with tropical cyclone Bruno, resulting in a flood with a 100 year ARI on the Blackwood River in the Southwest.

Even if summer rain events do not cause a major flood on the Swan River, they can still cause considerable damage and disrupt activities in the city. The impact of one such event is illustrated by the newspaper headlines relating to rainfall on 22 January 2000 (Figure 4).

newspaper headlines, January 2000

Figure 4. The heavy rainfall event of 22 January 2000. While not causing a major flood on the Swan River, the event did considerable damage and disrupted activities in the city.

Historical Trends

Historical rainfall patterns in the southwest have been the subject of intense study in recent times owing to a sustained reduction of annual rainfall in the last forty years. Studies associated with the Indian Ocean Climate Initiative (IOCI) showed that heavy rainfall events decreased both in frequency (the number of events) and intensity (the amount of rainfall in each event) particularly since the 1960s (Haylock and Nicholls, 2000). Using the Manjimup site south of Perth, the ARI for daily rainfall of 50 mm increased from 1.5 years (0.0018 AEP) in the 1930-65 period, to 7.5 years (0.0004 AEP) in the 1966-2001 period (IOCI, 2002). Such changes are less dramatic in Perth, but nevertheless the number of heavy rain events has decreased noticeably. Figure 5 shows the frequency at 15 year intervals of daily rainfall exceeding 50 mm for Perth since 1883. This highlights the sudden decrease in heavy rainfall events in the 1960s.

Frequency of daily rainfall exceeding 50 mm in 15 year intervals 1883 - 2002

Figure 5. Frequency of daily rainfall exceeding 50 mm at Perth in 15 year intervals 1883 -2002.

This drier period agrees with the flood record on the Swan River where the last major floods occurred in 1963 and 1964. Although there have been significant individual rain events in the last forty years they have not caused major floods. The most serious of these was in July 1983, when heavy rain over the Swan-Avon catchment flooded the upper Swan Valley and threatened some properties in Bassendean and Guildford.

Storm Surge

Sustained strong westerly or northwesterly winds can cause a build up of water on the coast that is referred to as storm surge. This enhances the water level above the normal tidal variation. This includes water levels on the Swan River subject to tidal variation so that parts of the river foreshore may be inundated near high tide. A storm surge also exacerbates flooding on the coastal plain by not allowing the river water to escape out to sea. The extent of flooding in Perth is dependent upon:

1. The intensity of the surge itself which relates to the strength and duration of onshore winds.
2. The timing of the peak surge in relation to the predicted high tide.
3. Existing river levels from rainfall in the catchment area.

Most storm surge events are associated with sustained westerly gales caused by intense lows and cold fronts in the cool season (refer cool season storms). During the 23 May 1994 storm, the tidal elevation measured at Fremantle showed a storm surge of 0.98 m. Fortunately the potential for more serious and extensive flooding was not realised as the peak surge occurred near the time of predicted low tide. A strong westerly gale on 20 July 1910 caused damage along the west coast to as far north as Geraldton. The Fremantle North Mole was damaged while on the Swan River all the surrounding low-lying lands and many of the jetties were submerged. More recently, gales associated with a low passing near Busselton on 16 May 2003 caused a storm surge of 0.8 m at Fremantle at close to the time of high tide as shown in Figure 6. The actual tide was 0.5 m above the highest astronomical tide and significant coastal erosion occurred. Figure 7 shows water overflowing onto Riverside Drive near the city centre. Fortunately existing river levels were low owing to dry conditions in the previous months, otherwise flooding would have been significantly worse.

 

Fremantle tide, 16 May 2003

Figure 6. Actual tide above the predicted tide at Fremantle, 15-16 May, 2003, showing the impact of a storm surge during a cool season westerly storm (Data courtesy of the Department of Planning and Infrastructure, Western Australia).

Riverside Drive

Figure 7. The effects of the storm surge on Riverside drive, 16 May 2003 (Courtesy of Lee Evelegh).

Wave action adds to coastal erosion. Intense winter lows to the south generate significant swells. During the May 1994 storm the open-water swell was estimated at eight to nine metres.

In addition to cool season events tropical cyclones can also cause storm surges in Perth. Cyclone Alby in April 1978 generated northerly gales responsible for large waves and a storm surge, which caused substantial coastal erosion along the Lower West coast. At Fremantle the surge was about 0.6 m, causing a high tide of 1.8 m, about 0.5 m above the highest astronomical tide. Very dry conditions and northerly, rather than westerly winds negated the flood potential on the Swan River. Areas further south at Bunbury and Busselton more prone to a northerly surge were flooded in low-lying areas on the coast.


© Copyright Commonwealth of Australia 2008, Bureau of Meteorology (ABN 92 637 533 532)
Please note the Copyright Notice and Disclaimer statements relating to the use of the information on this site and our site Privacy and Accessibility statements. Users of these web pages are deemed to have read and accepted the conditions described in the Copyright, Disclaimer, and Privacy statements. Please also note the Acknowledgement notice relating to the use of information on this site. No unsolicited commercial email.