County Clare Flood Forecasting System

Transcription

1 County Clare Flood Forecasting System Paul Moroney BE (Civil), C. Eng. M.I.E.I. M.C.I.Arb. Head of the Environment Section Clare County Council (The views expressed in this paper represent the professional views of the author and not the views of Clare County Council)

2 Flooding in County Clare November 2009 November 2009 notable for high rainfall recorded and severe flooding. Rainfall totals for November were the highest on record. Two-day falls of over 100mm were recorded in parts of the west and southwest on the 18th/19th. 28 day rain totals at Shannon in November 2009 were 270mm

3 Flooding in County Clare November 2009 River Flows on the Fergus in Ennis peaked at 80 cumecs on November 19 th and 24 th. These river flows were the highest on record. These flows lead to significant town centre flooding in Ennis River Flows on the Shannon peaked on November 28 th at 920 cumecs, also the highest on record These flows lead to significant flooding in South East Clare.

4 Flood Forecasting System During the flood events: Immersed in emergency response and management. After the events: developed an interest in the river flows and levels. Excellent national Weather warning systems in place. No national or regional flood forecasting system existed. Could a flood forecasting system be developed from existing data sources To suit County Clare's needs until a better system came online.

5 Existing data sources OPW river gauges, Ballycoree, Inch Corofin over 50 years of records Met Eireann rain gauges, Shannon Airport CSO reports Shannon Foynes Port tides prediction and gauges Ennis Flood Study from 2001 ESB records and documents European Flood Advisory System EFAS

6 Shannon Airport Rainfall records Period from 1958 to 2010 analysed. Reasonably low number of records where cumulative monthly rain amounts exceeds 170mm Monthly rain totals exceeding 200mm generally co-incide with town centre flooding in Ennis Large daily maximum rain amounts at Shannon have not generally correlated with winter flooding to date All historical floods in 50 years in Ennis are in Nov, Dec, Jan or Feb

7 November 2009 Monthly total at Shannon 263mm 277% of normal November rain

8 Shannon Monthly rain totals cumulative monthly rain Monthly totals > 170mm Oct Nov Dec Jan Feb Mar month Source CSO and Met Eireann

9 River Fergus Ballycoree flow records Range of records is from 1958 to 2010 Flows greater than 40 cumecs generally result in flooding in flood plains. Flows greater that 50 cumecs generally result in town centre flooding Highest flows occur between 5 and 7 days after maximum rain Average flows at Ballycoree have increased from 10 cumecs to 11.4 cumecs over the 50 year period

10 OPW Ballycoree Fergus flow November 2009 Max flow at Ballycoree 79.9 cumecs 24 th November flow (cumecs) flow cumecs /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/ /01/2010 Source: OPW date

11 OPW Ballycoree flood comparisons flow (cumecs) Oct 1 Oct 7 Oct 13 Oct 19 Oct 25 Oct 31 Nov 6 Nov 12 Nov 18 Nov 24 Nov 30 Dec 6 Dec 12 Dec Dec 24 date in winter season Dec 30 Jan 5 Jan 11 Jan 17 Jan 23 Jan 29 Feb 4 Feb 10 Feb 16 Feb 22 Feb 28 Mar 6 Mar 12 Mar 18 Mar 24 Mar flood 1995 Flood 1999 Flood 2009 Flood 2008 Flood 2006 flood 1993 Flood 1989 Flood Source: OPW

12 Developing a flood model Uses OPW figures for flow at Ballycoree and Claureen Use the 2001 Ennis main drainage report which includes a flood model. This was required due to complicated flow regime, 2 rivers Fergus and Claureen, groundwater, 2 flood channels Predicts flow through the town and the Fergus Minor and in Lower river near Clarecastle A less complicated river with one flow only requires a gauge and some analysis.

13

14 Flood Model No 1 based on River flow Reports on River Flow through the town under Victoria bridge represents fluvial flood risk Reports on daily rain events at Shannon represents pluvial flood risk References fluvial flow to an EFAS type colour coded warning system Analysed regularly during winter (and Summer) flood season Detailed predictions of levels at the nodes for a given flow is available Reported internally in Clare County Council twice weekly during flood risk periods and used as a basis for issuing flood warnings

15

16 Flood Model No 2 based on River levels Reports on River levels for 4 OPW gauges Ballycoree, Corofin and Victoria on the Fergus Inch bridge on the Claureen Referenced to annual average and maxima levels from OPW records, Amed and Amax References river level to EFAS Peak to peak and response times clearly seen Reported internally in Clare County Council twice weekly during flood risk periods and used as a basis for issuing flood warnings

17

18 Features of the model Uses only pre-existing data Based on flood flows detailed level predictions are available for important nodes in the network. Some of these have gauges in situ. Model is well calibrated Uses readily available software easily transportable home/office Reliant on waterlevel.ie web site. When this site is down manual inspection of gauges required In addition at remote locations if internet is down unable to update model Could easily be developed into an APP

19 Features of the model (contd) Is used as a proxy for flooding county wide Has been extended to other river catchments easily albeit with assumptions on flood models Will need to be updated once CFRAMS are in place after 2015 and may be redundant at that stage Will become redundant when national flood forecasting system is rolled out Heavily reliant on a small group of interested and motivated staff No budget, reliant on old technology and old (ish) people

20 Tidal Coastal Flooding Jan/Feb 2014 Issues regarding tidal and coastal flooding were apparent early in Tidal cycle predicted very high spring tides on Jan 3 rd and Feb 1 st 2014 Awareness of coincidence of Atlantic storms with high tides in the winter season was well known Exact details of interaction between Low pressure surge, high winds, wave action and high seas not known in detail OPW ICPSS surge forecast in use for a number of years ICPSS predicts tidal surge at Shannon Approach in the Atlantic at the mouth of the Shannon Estuary

21 Weather Jan/Feb 2014 Jet Stream over Ireland from mid December 2013 at a much lower latitude than normal Conveyor belt of Low pressure systems feeding Atlantic storms onto the west coast Storms on December 23 rd, 27 th, Jan 3 rd, 6 th and Feb 1 st. Many less significant events in between. Co-incidence of Storm conditions, Low Barometric pressure causing surge, High Seas and wave height and strong onshore winds Tide levels known, could surge and flood risk be predicted?

22 Analysis of Storms Dec 27 th extremely low barometric pressure High tides not at spring tide condition Forecast of wave height, surge and tide carried out Gauges at Rossaveal, Fenit, Doonbeg and Clarecastle Based on forecast surge from ICPSS, wave height, wind and barometric pressure flood risk at Clarecastle estimated Flood warning issued County wide tidal and coastal flooding warning for entire Clare coast.

23 Storm surge at Rossaveal Dec 27th Jan 3rd Jan 6th

24 Storm surge at Clarecastle Dec 27th Jan 3rd Jan 6th

25 Surge Analysis of Storms Date Location Surge est Surge act Highest tide level AOD Malin Head Dec 27 th Jan 3 rd Clarecastle 600mm 900mm Jan 6th Clarecastle 500mm 400mm Feb 1st Jan 6 th Clarecastle 300mm 400mm Feb 1st Clarecastle 300mm 270mm 4.225

26 Storm forecasting Feb 1 st 2014

27 Flood Warnings

28 Flood action Flood warnings issued on CCC web site Flood warnings sent via text alert to over 9000 subscribers Closely examine gauges during events to predict and record Actual surge at Rossaveal and Fenit gives a 1 hour warning for Clarecastle Newspaper and media warnings Local radio warnings Pumps, from 300mm to 100mm, Sandbags 1 tonne and regular

29 Issues for the future 1 Many good systems are in place, but they are not co-ordinated They are not all to Malin head datum. Differing datums between tide predictions and gauges leads to confusion Weather prediction systems are not tailored for flood and surge prediction and they could be. Colour code warnings have different colour codes on different systems, e.g. Met Eireann weather warning and EFAS flood warning and UK EA colour codes

30 Issues for the future 2 In an area like the Shannon Estuary the Local authorities and Port Company should co-ordinate flood and surge prediction A detailed flood surge model should be generated for Shannon, Ennis, Kildysart, Kilrush, Carrigaholt, Kilbaha, Kilkee, Doonbeg, Lahinch Liscannor, Ballyvaughan, etc The ICPSS gives good warnings but has only 2 analysis points on the west coast Shannon approach and Galway bay ICPSS does not report on actual surge just predicted On the west coast and Estuary there is a lack of tidal gauges for reporting and analysis

31 Issues for the future 3 Due to extreme nature of forecasts difficulty with comprehending communicating and being understood Everyone has an opinion on the weather, some of it wrong Prolonged weather events lead to desensitization, complacency and fatigue There is a tendency to put effort into response as opposed to preparedness A credible national flood forecasting, flood warning and flood action system is required

32 Real time video Lahinch Jan 3rd Video from Clare County Council Bottle Banks at the playground Lahinch High tide was at 6.00am First signs of flooding at 3.40am over 2 hours before high tide Prolonged damaging surge effects Extreme Flood Hazard The tide came in for 2 hours before peak and took 2 hours to go out in the dark This flood caused damage to 35 different locations on the Clare coast