Report R08/23 ISBN Kilmore Street PO Box 345 Christchurch Phone (03) Fax (03)

Transcription

1

2

3 Report R08/23 ISBN Kilmore Street PO Box 345 Christchurch Phone (03) Fax (03) Church Street PO Box 550 Timaru Phone (03) Fax (03) Website: Customer Services Phone Environment Canterbury Technical Report

4 EXECUTIVE SUMMARY This study of the Ashley River floodplain uses a combined one and two dimensional hydraulic computer model to estimate flood extent and depths on the Ashley River floodplain. Modelling has indicated the current capacity of the Ashley River stopbanked system is approximately equivalent to the 2% AEP (50 year return period) flood event. Breakouts however, could occur in more frequent events. Stopbank breaches and outflows onto the floodplain could potentially occur anywhere along the stopbanks. Breakout scenarios onto the floodplain have been modelled for the 1%, 0.5% and 0.2% AEP (i.e. 100, 200 and 500 year return period) events at the most likely breakout locations. The modelling indicates significant flooding to large areas of land between the Ashley and Waimakariri Rivers. Kaiapoi and adjacent areas are predicted to be flooded to depths over 1 metres in the 0.2% AEP (500 year return period) event. The impact of future urbanisation in Kaiapoi has also been modelled. It is hoped the floodplain maps and associated depths will assist land use planning within the area and provide information on minimum floor levels for new dwellings located on the floodplain, where appropriate. Environment Canterbury Technical Report 1

5 January 1953 Flood Looking across Ashley River at flooding between Waikuku Beach and Woodend Beach. Floodwaters extended through to the Pines and Kairaki. Photograph by L Ernle Clark for North Canterbury Catchment Board. Environment Canterbury Technical Report

6 Table of Contents 1. Introduction Background Ashley Catchment and Geomorphology Historical Flooding Ashley River control Scheme Flood Investigations Risk Assessment Flood Hydrology Climate Change Hydraulic Model Description Model Simulations and Results Discussion Glossary...43 Appendices Appendix 1: Appendix 2: Appendix 3: Appendix 4: Appendix 5: Appendix 6: Appendix 7: Appendix 8: Floodplain Geomorphology Ashley River Flood Levels Floodplain Roughness Ashley River and breakout flow hydrographs Floodplain Roughness Sensitivity Analysis Kaiapoi Area Breakout Flow Sensitivity Analysis Hazard Category Definition Flood Probability List of Photos Photo 1: Damage to the Ashley River railway bridge in March 1902 flood Photo 2: Surface flooding in High Street, Rangiora February Photo 3: Kaiapoi River (Williams Street) in flood February Photo 4: Kaiapoi February 1945 flood... 9 Photo 5: Rangiora Loburn traffic bridge at the height of the flood on 25 January Donated by Mrs C Tyler to Rangiora and Districts Early Records Society Photo 6: The Coldstream area, showing the major overflows, and extent of flooding between Rangiora Golf Course and the sea. Photo by L Ernle Clark for North Canterbury Catchment Board Environment Canterbury Technical Report 3

7 List of Figures Figure 1: Ashley, Waimakariri, and Kaiapoi River Stopbanks Figure 2: Potential failure zones and historical breakouts, Ashley River Figure 3: Ashley River Floodplain Topography Figure 4: Potential Flooding in 1% AEP Breakout A Figure 5: Potential Flooding in 1% AEP Breakout B Figure 6: Potential Flooding in 1% AEP Breakout E/F Figure 7: Potential Flooding in 1% AEP Combined Breakout Figure 8: Potential Flooding in 0.5% AEP Breakouts A & B Figure 9: Potential Flooding in 0.5% AEP Breakouts A & C Figure 10: Potential Flooding in 0.5% AEP Breakouts E/F Figure 11: Potential Flooding in 0.5% AEP Combined Breakout Flooding Figure 12: Potential Flooding in 0.2% AEP Breakout Figure 13: Potential Flooding Kaiapoi area in 0.5% AEP Event Figure 14: Impact of Filling Kaiapoi Growh areas 0.5% AEP Event Figure 15: Potential Flooding Kaiapoi area in 0.2% AEP Event Figure 16: Impact of Filling Kaiapoi Growth area 0.2% AEP Event Figure 17: Flood Hazard Map 0.5% AEP Event Figure 18: Flood Hazard Map 0.2% AEP Event Environment Canterbury Technical Report

8 1. INTRODUCTION The Ashley River floodplain covers approximately 190 km 2 and includes two large urban areas, Rangiora and Kaiapoi. A previous investigation on flooding of the Ashley River Floodplain was undertaken in Ashley River Flood Plain Management Plan Technical Investigation Report 95(6), This investigation highlighted potential breakout risk and predicted overland flow paths and ponding areas, using a one dimensional hydraulic computer model, Mike 11. In 2003 the Waimakariri District Council and the Canterbury Regional Council adopted a Waimakariri District Flood Hazard Management Strategy. Mitigation measures in the strategy included maintenance of the existing stopbank system and land use management measures, such as appropriate development in flood prone areas and appropriate minimum floor levels for new dwellings. The investigations described in this report have been undertaken with the aid of very detailed topographic data, obtained from a LiDAR (Aerial Laser) survey and a more physically realistic two dimensional hydraulic computer model of the floodplain (using the Mike 21 software). This more accurate (c.f. previous investigation) should enable better informed decisions to be made regarding land use management of the floodplain and will also enable a more accurate assessment of appropriate minimum floor levels for new dwellings to be made. The hydraulic computer model has also been used to estimate the effects of various development scenarios (i.e. filling) on the floodplain in the Kaiapoi area. The assessment of the probability of Ashley River breakouts (outflows) and locations is described in a separate report Ashley Floodplain Hazard Risks Assessment, Report No. U08/1, January BACKGROUND 2.1 ASHLEY CATCHMENT AND GEOMORPHOLOGY The Ashley River is a relatively steep braided river and together with its major tributary, the Okuku, drains a catchment of approximately 1200 km 2. The catchment is relatively steep with the highest point in the Puketeraki Range at over 1900m. The unconfined floodplain commences on the south bank of the Ashley River, immediately downstream of the Okuku River confluence. Numerous breakouts and change of course here over thousands of years have resulted in alluvial deposition leading to the formation of the adjacent alluvial fans, extending to Kaiapoi and the Waimakariri River. On the north side of the Ashley, however, the river is confined by a prominent 5-10 metre high terrace. This terrace extends downstream from the Okuku River as far as Ashley Township. Downstream of the Ashley Township the river is confined by stopbanks but the floodwaters can break out onto the Ashley Fan. Environment Canterbury Technical Report 5

9 A full description of the Ashley Floodplain Geomorphology can be found in the earlier Technical Investigation Report on the Ashley, Report 95(6). A map of the Ashley Floodplain geomorphology is shown in Appendix HISTORICAL FLOODING Floods in the Ashley generally result from north easterly or south easterly weather systems and orographic rainfall in the hills and mountains headwaters. Heavy rain from decaying tropical depressions can also penetrate the catchment from the north east. Records from early European settlers and newspapers indicate that flooding from the Ashley River has always been a problem. The New Zealand Rivers Commission reporting on the Ashley in 1921, stated that evidence showed that in times of very high flood the water has escaped the Ashley River and run into the low country lying to the north and west of Kaiapoi. The flooding of the adjacent lands is caused by the fact that along some portions of the river the natural banks are lower than the grade of a high flood, and furthermore by the fact that the river, in common with most other Canterbury rivers, is running on a fan, and once the floodwaters get over the immediate river bank they tend to follow old channels that lead away from the main river and do not return lower down, as is the case with valley rivers. Further on, in discussing the flood hazard to which Kaiapoi has been subjected, the Rivers Commission also make important observations with regard to the Ashley, vis á vis the Waimakariri, when they state that in addition to these floods (of 1868 and 1987, emanating from the Waimakariri) Kaiapoi has been inundated by local floods from the Eyre and Cust Rivers, and also by flood overflows from the Ashley River. As a rule floods in the Eyre, Cust and Ashley Rivers do not synchronise with those of the Waimakariri, but this happened in 1868 and Numerous floods were reported in the latter half of the 19 th century, but the largest appears to be the February 1868 event. The Ashley River broke its banks flooding Rangiora and floodwaters extended all the way to Kaiapoi. Two children were drowned near Rangiora and Kaiapoi suffered severely with water up to 5 ft 6 inches (1.7m) in some streets, and almost every shop and house was invaded with consequent heavy losses the flooding of Kaiapoi was due to, in the main, to the overflows from the Cust, Eyre and Ashley Rivers, and flooding from the Waimakariri River. Further significant floods were recorded in the early 20 th century, particularly March 1902, June 1905 and May The 1923 flood was reported as the most disastrous since 1868 A large amount of damage was done to Kaiapoi and Rangiora was similarly affected Between Kaiapoi and Southbrook the country is a sea of water, in some places up to the tops of the fences the main road from Flaxton to Southbrook was transformed into a river of water from fence to fence. Kaiapoi was again flooded as a result of a breakout from the Ashley above Rangiora. Rangiora suffered some local (surface) flooding but Kaiapoi was inundated. Two hundred families were evacuated from their homes and the Revell family marked a new flood level at their home fifteen inches above the 1868 mark. At Waikuku the wool works were flooded to a depth of 8 ft (2.4m). Environment Canterbury Technical Report

10 Photo 1: Damage to the Ashley River railway bridge in March 1902 flood. Following this devastating flood, the original Ashley River Flood Protection Scheme was constructed in the 1930 s. However the partially completed stopbanks were breached in at least three places by a large flood in February The most serious flood damage was at Waikuku and up to 9ft (2.7m) at the Waikuku Wool Works. After this flood the height of the stopbanks was increased and belts of willows were planted as an additional defence. However despite completion of the original scheme in 1938, the flood of March 1941 resulted in the Ashley River breaking its banks in four or five places. Again this caused major inundation of farm land, stock losses, flooded houses and occupants being rescued by boat. Mr Wylie had to be rescued by boat, from a haystack and on the south bank at the same time an elderly woman, Miss M Leggatt was rescued by boat from the upper story of her home. Both had been isolated for two and a half hours. The stopbanks were again breached in 1945, 1951 and These events all caused widespread flooding and major flood damages, and evacuation of people from their homes. In the 1953 flood event, Waikuku and Woodend Beach were flooded and floodwaters flowed through the Pines and Kairaki. A Woodend farmer spent 12 hours perched 14 feet up on a pine tree, until rescued by the Lyttelton Harbour Board. Floodwaters from the Ashley also caused the Cam to overflow its banks and threatened the low-lying Camside area of Kaiapoi. There have been a number of reasonably large river flows since The largest being December 1993, approximately 1900 m 3 /s. Although the stopbanks have been close to breaching/overtopping a number of times, there have been no breakouts onto the floodplain since Environment Canterbury Technical Report 7

11 In addition to flooding from the Ashley River, significant flooding has occurred from local rainfall events and from waterways such as the Cam and the Cust. Photo 2: Surface flooding in High Street, Rangiora February 1945 Environment Canterbury Technical Report

12 Photo 3: Kaiapoi River (Williams Street) in flood February 1945 Photo 4: Kaiapoi February 1945 flood Environment Canterbury Technical Report 9

13 Photo 5: Rangiora Loburn traffic bridge at the height of the flood on 25 January Donated by Mrs C Tyler to Rangiora and Districts Early Records Society. Photo 6: The Coldstream area, showing the major overflows, and extent of flooding between Rangiora Golf Course and the sea. Photo by L Ernle Clark for North Canterbury Catchment Board Environment Canterbury Technical Report

14 2.3 ASHLEY RIVER CONTROL SCHEME The original stopbank scheme built in the 1930 s was designed to contain 2000 m3/s (with no freeboard). The present scheme was designed and constructed by the North Canterbury Catchment Board in 1976 to contain 2400 m3/s which at the time was estimated to be the 1% AEP (Annual Exceedance Probability) or 100 year return period. Based on the current hydrology, 2400 m3/s is estimated to be between the 5% AEP (20 year return period) and the 2% AEP (50 year return period). The adopted design stopbank heights allowed for 600mm freeboard above the design flood level, or at the existing bank heights, where higher. Generally, the existing bank heights were higher. It should be noted however, that for much of its length the river and stopbanks system has a capacity of approximately 3000 m3/s (plus freeboard). This is estimated to be approximately equivalent to a 2% AEP (50 year return period) flow. The Ashley River Control Scheme includes 33.7 km of stopbanks, groynes, tree planting and rock protection. (Note: the stopbanks are continuous on the north banks from the mouth to the main trunk railway at Rangiora, where it meets a natural terrace. On the south bank the stopbanks are continuous from the mouth to opposite the Okuku River confluence see Figure 1). Additionally annual maintenance is undertaken to control vegetation and shingle, including controlling commercial shingle extraction. Figure 1: Ashley, Waimakariri, and Kaiapoi River Stopbanks Environment Canterbury Technical Report 11

15 3. FLOOD INVESTIGATIONS A thorough analysis of flooding on the Ashley River Floodplain and impacts on land use has required a number of different studies to be undertaken. Cumulatively these form the basis of the floodplain investigation. These studies included: Historical floods (discussed in previous chapter) Risk assessment Flood hydrology River cross-section survey Topographic survey of the floodplain Floodplain roughness (hydraulic resistance) assessment Channel roughness assessment Hydraulic computer model of rivers, streams and floodplain Hydraulic modelling of a range of breakout scenarios for current land use Impact of land use changes, in Kaiapoi area, on the floodplain. 3.1 RISK ASSESSMENT As discussed above, the Ashley Floodplain Risk Assessment has been reported in Report No. U08/1. The assessment included examination of the geomorphology, analyses and performance of the existing system, potential stopbank failure (overtopping or lateral erosion) locations (refer Figure 2) and associated likelihood of failure for a range of flood scenarios. The structural strength of the existing scheme has been the subject of close scrutiny. The main conclusions are: The stopbanks are designed and built to a uniform standard, but they are not designed to withstand overtopping by floodwaters. The greatest risk of failure of the system is from erosion of the stopbanks, and such a failure could occur during a flood significantly less than the design flood. Stopbank erosion is dependent, not only on discharge, but also on other variables. It is not possible to predict the exact point of erosion failure, although constant vigilance, and a pro-active maintenance programme can lessen the probabilities of erosion failures. Following an examination of the scheme performance and current theoretical scheme capacity (approximately 3000 m 3 /s), it is predicted that there is a 50% chance of a breakout at that magnitude. The risk assessment analysis also indicates there is potential, albeit low, for breakouts to occur in flows as low as 2000 m 3 /s (approximately 10% AEP). The adopted median breakout scenario indicates there is approximately a 7% chance of a breakout from the Ashley River in any one year. For the purposes of this floodplain investigation, breakouts in floods smaller than the 1% AEP (100 year return period event) have not been modelled, although it needs to be recognised they could occur. Environment Canterbury Technical Report

16 Breakout location, could be anywhere along the stopbanked system, including multiple breakouts. Obviously for the purposes of this investigation and to reduce the numerous scenarios to be considered, the most likely breakout locations have been selected. This is based on geomorphology, historical breakout locations, and where freeboard on the stopbanks is minimal. Similarly, the magnitude of the breakout flows (peak and volume) could cover quite a range. A mid-range estimate has been adopted for these, and are based on the premise that the residual flow in the Ashley River will be approximately 3000 m 3 /s. The greatest breakout threat and the largest outflow is likely to occur at the upper end of Zone A, where the stopbank system commences and where the freeboard is minimal. Environment Canterbury Technical Report 13

17 Figure 2: Potential failure zones and historical breakouts, Ashley River Environment Canterbury Technical Report

18 3.2 FLOOD HYDROLOGY Flood estimates for the Ashley River catchment were prepared for Environment Canterbury, by NIWA (National Institute of Water and Atmospheric Research) in Discussions with NIWA concluded there was likely to be little change in the 2002 estimates. The NIWA analysis used two methods to estimate flood flows. Firstly, the Regional Flood Estimation method, using maximum flood data from nearby river catchments was used. Secondly, analysis of the Rangiora Traffic Bridge, (Environment Canterbury s recording site) annual flood peaks was made, but with only 12 years of continuous record here, this analysis cannot be used with much confidence. Therefore the Regional Flood estimates have been adopted as the design flows (i.e. total catchment flows). The estimated design flows at Rangiora for a range of probabilities are tabulated below: Event Probability Peak Flow m 3 /s Mean Annual Flood 1% AEP (100 year return period) 0.5% AEP (200 year return period) 0.2% AEP (500 year return period) 0.1% AEP (1,000 year return period) 860 3,470 4,050 5,300 6, CLIMATE CHANGE Climate change, particularly as a result of global warming is predicted to result in an increase in peak runoff and elevated sea levels. Latest indications for Canterbury are an estimated mid range temperature increase to 2090 of approximately 2 o C. While there is still uncertainty over the magnitude of this and its impact, the latest (2008) Ministry for Environment Guidelines suggest approximately 15% increase in storm rainfall depths. For the purposes of this investigation, the resulting flow increase is within the error estimate of the above design flows. In addition there is the uncertainty of the breakout magnitude, location, and duration. Therefore the adopted design flows have not specifically included an allowance for climate change. However, some sensitivity analysis of breakout flows has been modelled. Sensitivity analysis of the predicted 0.5m (mid-range) sea level rise (by 2100) has also been undertaken. 3.3 HYDRAULIC MODEL DESCRIPTION Flows over a floodplain are multi-directional and thus are difficult to predict. To determine flood levels (in river and on floodplain), flow patterns and velocities on the floodplain, to the required level of accuracy, and in a realistic manner, a 1D/2D hydrodynamic computer model (Mike Flood) was used. The Mike Flood (DH1 software) modelling system combines onedimensional (Mike 11) for the main rivers, and two-dimensional (Mike 21) modelling for the floodplain, in one modelling system. The linked Mike 11 and Mike 21 models, allow flood waters to move between the river network and the floodplain, for example, where floodplain overflows are returned to the main river. Environment Canterbury Technical Report 15

19 3.3.1 One Dimensional Channel Network Model The 1D (Mike 11) model includes the main rivers which impact on the Ashley Floodplain, i.e. Ashley, Kaiapoi and Waimakariri Rivers. The Mike 11 model of the Ashley River extends from the sea to approximately 22km up river (u/s Okuku River confluence). Surveyed cross-sections used in the model are typically 800m apart. The last full cross-section survey was in Approximately 25% of the crosssections were resurveyed in 2003, to compare with the LiDAR data (measures above water points only). A comparison of these cross sections with the 1997 survey showed that bed levels were typically m lower. However, because the 2003 survey was not a full survey, and there was uncertainty in the remaining cross-sections, it was felt prudent to base the Ashley River modelling on the 1997 cross-sections. This is likely to result in conservative flood levels in the river, especially since shingle extraction in specific reaches is still continuing. A full channel cross-section survey is programmed for 2008/2009. A Manning number of 0.03 has been used for the open channel bed resistance, and variations in resistance to vegetation, have been accounted for using relative resistances. Manning n values of up to 0.12 have been used for heavily vegetated berm areas. Two bridges (Rangiora Traffic/Cone Road and State Highway 1) have also been included, to take account of channel cross-section changes, submerged soffit, and pier losses. The Ashley River 1D model has been calibrated with the August 1986 event (1400 m 3 /s) and the December 1993 event (1900 m 3 /s). Using an open channel bed resistance value of 0.03 gave reasonable agreement, with average observed levels at the cross-sections. (Note: the observed values of the left and right bank flood levels are quite variable but could be partly due to difficulty in identifying flood marks, subsequent to the event). Predicted flood levels, for a range of these in the Ashley River (with no breakouts) is shown in Appendix 2. Kaiapoi River A large portion of the floodplain overland flow drains into the Kaiapoi River. The lower 1 km of the river, upstream of the Waimakariri River confluence is included in the 1D Mike 11 model. Cross-section data has been obtained from a few isolated surveys and from the LiDAR survey. The Kaiapoi River upstream of the Mike 11 sections, is included in the 2D model, with appropriate provision made for bed levels. The Kaiapoi River stopbank crest levels are included in both the Mike 11 cross-sections and the 2D model. Waimakariri River The Kaiapoi River drains into the Waimakariri River. Cross-sections and channel resistance from previous modelling on the Waimakariri River were incorporated into the model. Although the Waimakariri River is most unlikely to be in major flood at the same time as a major event on the Ashley River, a mean annual flood (1500 m 3 /s) has been assumed. The downstream water level, has been based on a tidal cycle, with a maximum level of 1.7m (msl). This is a relatively high tide (based on historic observations) and includes some allowance for storm surge, which is very likely to occur during a major rainfall event in the Ashley catchment. Environment Canterbury Technical Report

20 Minor River Branches A number of minor branches were also included in the Mike 11 model to simulate other physical connections between floodplain and rivers/streams. These include: Return flow channel through the gap in the stopbanks under the railway bridge near Rangiora. Saltwater Creek to Waimakariri River. Courtney Stream to the Kaiapoi River, with non return weir, to simulate tide gate. McIntosh s Drain to the Kaiapoi River, with non return weir, to simulate tide gate Two Dimensional Floodplain Model Floodplain Topography To realistically model floodplain flows, with any degree of accuracy, good topographic data, including features such as banks, terraces, overland flow channels, roads and railway embankments is essential. For the Ashley Floodplain this high resolution topographic data was obtained from a LiDAR survey (aerial laser scanning) flown in July While the specified accuracy was +0.15m, on clear open ground, analysis with GPS surveys, showed the accuracy was a lot better than this. The detail, including historic flow paths and stopbanks, of the LiDAR data can be seen in the grey scale image of the floodplain in Figure 3. Water levels and flows on the floodplain are resolved on a rectangular grid. The size of the grid is a function of the level of detail required, model stability, and computer capacity and speed. A 10m grid was chosen for this study, which for the size of the floodplain (190 km 2 ) is considered to be a good resolution. Checks were made with the full topographic survey, to ensure important topographic features such as banks, terraces, roads and railways were correctly represented in the 10m grid. Where necessary, e.g. stopbanks, levels in the 10m grid were modified to correctly represent these features. However, generally there was good agreement between the full topographic data set and the representative 10m grid (bathemetry). Environment Canterbury Technical Report 17

21 Figure 3: Ashley River Floodplain Topography Environment Canterbury Technical Report

22 Floodplain Flows As described previously stopbank breach scenarios in the Ashley River aren t necessarily a result of direct overtopping, but more likely from lateral erosion. It was therefore considered the most appropriate method to introduce flood overflows (through stopbank breaches) onto the floodplain was as source inflows distributed over 10 grid cells (for stability) at each breakout location. As discussed in Section 3.1 Risk Assessment, potentially these breakout locations could be anywhere along the stopbanked system. The source inflows, represented as outflow hydrographs, however, are located at what are considered the most likely locations. Floodplain Roughness (Surface Resistance) Floodplain flows and depths are significantly influenced by the hydraulic resistance of the ground cover and other obstructions, such as buildings and trees on the floodplain. Normal practice is to assign varying resistance values, e.g. Mannings n to the various surfaces of the floodplain by interpretation of aerial photographs and ground survey. For this investigation, using rain data from the LiDAR survey (e.g. first and last returns and intensity data), NIWA have developed a very detailed surface map of floodplain resistance. (Refer Appendix 3). Typically Mannings n values varied from 0.02 (roads) to over 0.15 (dense vegetation). The roughness detail is represented in the Mike 21 model, in a 10m grid. Some sensitivity analysis was undertaken of surface roughness Mike Flood Model The 1D (Mike 11) and the 2D (Mike 21) models are dynamically coupled together in the Mike flood module. As discussed above these links/couples are at the Rangiora railway bridge into the Ashley River, Kaiapoi River, Courtney Stream, McIntosh s Drain and Saltwater Creek. 3.4 MODEL SIMULATIONS AND RESULTS All model simulations were based on a one second timestep, to ensure stability, and results were saved every 20 minutes (real time). The floodplains on the north and south sides of the Ashley River were modelled separately to reduce simulation time. Simulations on the largest southern floodplain however still took approximately 60 hours. Ideally any model should be calibrated with historical flood events. This has been done for the 1D Mike 11 model, as described in Section However, it was not possible to obtain any data to calibrate the 2D Mike 21 model with. Historical flood maps of the 1950 and 1953 floods do exist, however it is not known what the magnitude of the multiple breakouts were. The main variable in the floodplain model is surface resistance, and this data obtained independently from NIWA, is consistent with commonly used values. Sensitivity analysis of floodplain roughness has also been undertaken. Simulations have been undertaken for the 1%, 0.5% and 0.2% AEP (i.e. 100, 200 and 500 year) events, for both existing landuse and numerous development scenarios in the Kaiapoi area. Sensitivity analyses of a number of scenarios has also been undertaken. Environment Canterbury Technical Report 19

23 % AEP (100 Year Return Period) Scenarios The risk assessment analysis shows that for a 1% AEP river flood event the most likely breakout scenario is for a single breakout in Zone A. Two scenarios have been modelled a breakout at the top of Zone A (downstream of Okuku River confluence) and at breakout Zone B (Breakout Bank Rangiora) immediately downstream of Zone A. Potentially also a breakout could also occur on the northern floodplain (Zone E or F) downstream of Rangiora, east of Sefton. The 1% AEP river flow (upstream of breakouts) is estimated to be 3470 m 3 /s. The mid-range estimate for a breakout flow at A or B is 500 m 3 /s, and for E or F 125 m 3 /s. Ashley River and breakout flow hydrographs are shown in Appendix 4. Figure 4 shows Breakout A results in generally shallow flooding (0 0.5m) across a large part of the floodplain. Rangiora is mainly free from flooding (from the Ashley), although it is likely there will be significant local runoff. Much of the floodwaters enter the Southbrook and Cam River which drains into the Kaiapoi River. East of Bramleys Road and Flaxton flood waters pond in the Flaxton Swamp area to depths up to 2.5 metres. Ground levels in this area are particularly low, at just over 1m above mean sea level. Flood waters cross the northern motorway and flooding is mainly contained on the west side of Old North Road on the east side of the Kaiapoi River. (Note: there is a ridge of beach deposits about 3m high to the east of Old North Road.) The majority of the floodwaters eventually drain into the Kaiapoi River, which has a peak flow of approximately 100 m 3 /s. The majority of the existing urban area of Kaiapoi is free from flooding, with just isolated pockets of floodwaters in some low lying streets. However, it is quite likely there will be additional flooding from local stormwater. Figure 5 shows an alternative Breakout B, for the 1% AEP scenario. Although the breakout flow is the same as A, the resulting flood extent is a lot less. This is because approximately 40% (220 m 3 /s) of the breakout flow returns back to the Ashley River through the gap in the stopbank, near the railway line. The majority of the outflow bypasses most of Rangiora to the east, entering the North Brook and Cam River. Ponding depths in the Flaxton Swamp area are up to 1.5m deep. Other breakout flows spread to the east, towards Woodend and Waikuku, and across SH1. The extent however is relatively small. There is also a possibility of a breakout (E or F) onto the northern floodplain of approximately m 3 /s with flood waters crossing State Highway 1 and draining into Saltwater Creek (Refer Figure 6). The combined flood scenarios for the 1% AEP extent is shown on the composite map as Figure 7. While this map represents the likely potential total area to be inundated in a 1% AEP flood, it is most unlikely to occur in the same flood event. Environment Canterbury Technical Report

24 Figure 4: Environment Canterbury Technical Report 21

25 Figure 5: Environment Canterbury Technical Report

26 Figure 6: Environment Canterbury Technical Report 23

27 Figure 7: Environment Canterbury Technical Report

28 % AEP (200 Year Return Period) Scenarios In this river flood event, again, there are obviously a range of breakout scenarios. The most likely have been assessed as combined breakouts to the south, A and B, or combined breakouts to the south and north, A and E/F. The 0.5% AEP river flow (upstream of breakouts) is estimated to be 4050 m 3 /s. The combined breakout at A and B are 750 m 3 /s and 400 m 3 /s respectively. These flow hydrographs are shown in Appendix 4. Figure 8 shows the combined breakout for A and B. Flooding is fairly extensive, although Rangiora township is relatively flood free in this scenario. Depths in the Flaxton Swamp area are over 3m. Flood depths in the existing urban area of Kaiapoi, north of the river are up to approximately 0.8m. Flooding towards Waikuku and Woodend is similar to the 1% AEP scenario. An alternative breakout scenario is Zones A and C combined (see Figure 9). As can be seen there is a small reduction in flooding in the Kaiapoi area. Flood depths in the Waikuku area are typically 1 1.5m, although Waikuku Beach settlement, on the sandhills is only partially flooded to relatively shallow depths. Figure 10 shows the predicted flood extent with a breakout on the north side of the river (representing Zone E & F). For this scenario flood depths are generally under 0.5m, with greater depths adjacent to SH1. Floodwaters cross SH1 and drain into Saltwater Creek and the Ashley River. The potential flood scenario for the 0.5% AEP breakouts modelled is shown in the composite map in Figure 11. Environment Canterbury Technical Report 25

29 Figure 8: Environment Canterbury Technical Report

30 Figure 9: Environment Canterbury Technical Report 27

31 Figure 10: Environment Canterbury Technical Report

32 Figure 11: Environment Canterbury Technical Report 29

33 % AEP (500 Year Return Period) Scenario The estimated 0.2% AEP event of 5,300 m 3 /s, is predicted to result in over 40% of the flow spilling onto the floodplain. The risk analysis shows that there is likely to be multiple breakouts (three), potentially, all to the south or to the south and north. The most likely scenario is shown in Figure 12. The predicted peak breakout flows are 1750 m 3 /s (Zone A), 350 m 3 /s (Zone B), and 160 m 3 /s (Zone C/D). The breakout to the north (zone F) is similar to the 0.5% AEP scenario. Flow hydrographs are shown in Appendix 4. Obviously with such large volumes of floodwaters out of the river on the floodplain, the flooding is widespread. Still only parts of Rangiora are flooded and to depths generally under 0.5m. Flood depths in the Flaxton Swamp area are over 3m. The Kaiapoi River is at full capacity and banks are overtopped. Peak flow in the river is approximately 380 m 3 /s. The majority of urban Kaiapoi is flooded with depths south of the Kaiapoi River and east of the motorway around 2.5m deep. The urban area on the north side of the Kaiapoi River is predicted to be flooded in the range of metres. Environment Canterbury Technical Report

34 Figure 12: Environment Canterbury Technical Report 31

35 3.4.4 Simulations for Kaiapoi Development Scenarios A number of growth options are currently being considered by the Waimakariri District Council for Kaiapoi. The majority of these potential growth areas are within the Ashley floodplain and some areas are also within the Waimakariri River floodplain. Numerous combinations of growth options have been modelled to ascertain the potential effects. Figure 13 shows the potential flood depths currently within the possible growth areas in the 0.5% AEP (200 year return period) event. If these areas were developed, subject to various consents, it is likely they would be raised above flood level. Figure 14 shows the increase in flood depth on adjacent areas, due to this filling, are up to 1m in existing urban areas. This is due to loss of storage and floodwater displacement/diversion. In the 0.2% AEP (500 year return period) event the flood depths and extent within these potential growth areas is obviously greater. This is shown in Figure 15 The impact of filling these areas to above the 0.2% AEP flood level is shown in Figure 16 The greatest increase in depth is approximately 1.2m in the existing urban area. A range of other what if development scenarios have also been investigated to determine potential effects. Environment Canterbury Technical Report

36 Figure 13: Environment Canterbury Technical Report 33

37 Figure 14: Environment Canterbury Technical Report

38 Figure 15: Environment Canterbury Technical Report 35

39 Figure 16: Environment Canterbury Technical Report

40 3.4.5 Model Sensitivity Analyses A number of what if scenarios were modelled to determine the sensitivity of various model parameters. These included a: 0.5m sea level rise Modified channel roughness in the lower Waimakariri River Modified floodplain roughness in urban Kaiapoi, and Modified breakout flows Sea Level Rise Scenario As discussed in Section 3.2 predicted climate change impact on sea level is a 0.5 m rise (mid-range) by In the 0.5% AEP (200 year return period) breakout scenario, sea level rise had negligible impact on peak floodplain water levels. This is because the peak levels behind the stopbanks are dominated by the major breakout flows. However, as floodwaters from the low lying areas will only be able to drain out during low tides, the time of inundation will increase. Also in times of local rainfall events flooding in the Kaiapoi area, e.g. McIntosh s Drain catchment and Courtney Stream is likely to be worse, as floodwaters will only be able to drain out when the tide level is lower. Channel Roughness Lower Waimakariri River The adopted channel roughness (0.017) for the Waimakariri River was obtained from calibration with observed flood levels in the Waimakariri River. The sensitivity analysis was based on increasing this relatively low value to This resulted in an increase in maximum water level at the Kaiapoi/Waimakariri River confluence of approximately 100mm. As for the sea level rise scenario, the change in flood depth on the floodplain is negligible, as the levels are dominated by the ponding behind stopbanks and breakout volumes. There would be some small effect on flood duration however, depending on relative timing of Waimakariri peak flows, arrival of peak breakout flows from the Ashley and tide times. Modified Floodplain Roughness The floodplain resistance was obtained through independent analysis developed by NIWA. In areas of obstruction, e.g. trees, buildings, flood levels can vary depending on the adopted resistance value. To determine the impact of higher floodplain resistance in the Kaiapoi urban area a generalised value equivalent to Mannings n of 0.1 was used in the 2D model. Generally as can be seen in Appendix 5 there was only a relatively minor change in maximum flood depth in the 0.2% AEP (500 year return period) event. This was expected as the velocities in the Kaiapoi area are relatively low. Where the velocities are higher, however, the increase in maximum depth is higher, up to 0.3m over a small area on the north side of the Kaiapoi River. The what if roughness value of 0.1 is regarded as conservative as it generally lumps the urban area as one. In reality the resistance value for roads and other open spaces will be a lot lower. It is concluded that the NIWA resistance analysis, which assigns relative values within the urban areas differently, are realistic and there is no need to adopt the more conservative values. The sensitivity analysis, however, does identify the impact of increased obstructions on the floodplain. Environment Canterbury Technical Report 37

41 Modified Breakout Flows As discussed in previous sections there is some uncertainty in the magnitude and location of the breakout flow. The analysis, however, is based on what is considered the most likely midrange estimates. Some sensitivity analysis of these flows however has been undertaken. Firstly, the 0.5% AEP breakout flows at locations A and B were increased by 20%. (Note: Climate change predictions suggest a potential 15% increase in peak runoff by 2080.) The increase in flood depths (refer Appendix 6) upstream of Kaiapoi and Flaxton is generally less than 0.1 m. However, in the ponding areas around Kaiapoi, where the greater outflow volumes have more impact, the flood extent and depths are greater. Secondly the breakout flow hydrographs for A an B were switched, in the 0.5% (200 year return period) event. This resulted in slightly less floodwater on the main floodplain in total due to increasing flow through the railway stopbank gap at Rangiora. There was obviously more extensive flooding near Rangiora and in the Waikuku Woodend areas. However, because of the reduced total volume on the floodplain, flood depths and extent were slightly reduced in the Kaiapoi area. This scenario, however, is considered less likely than that originally modelled (i.e. Figure 8). 4. DISCUSSION Despite the existing stopbanked system on the Ashley River, this investigation has shown there is still significant residual risk on the floodplain. The current stopbank capacity is approximately 3000 m 3 /s equivalent to a 2% AEP (50 year return period) event. Stopbank breaches and outflows onto the floodplain, could theoretically occur anywhere along the length of the stopbanks. However, as described above, the modelling undertaken is based on the most likely locations, based on the geomorphology, historical breakouts, and where freeboard is minimal. The difference in flood extent and damages between the 1% AEP (100 year return period) and the 0.2% AEP (500 year return period) events is significant. The 0.2% AEP scenario results in over four times more water on the floodplain than the 1% AEP event. The breakout analysis and floodplain modelling indicates that generally Rangiora is relatively safe from Ashley River flooding. Waikuku and Woodend Beach settlements are likely to be flooded in some scenarios, although many of the houses at Waikuku Beach are located on higher sandhills. Flooding to Kaiapoi and peripheral areas however is potentially extensive and deep. There are a number of reasons for this. Kaiapoi is at the lower end of the Ashley River fan; floodwaters pond behind the Kaiapoi and Waimakariri River stopbanks; drainage of the floodwaters are impeded by elevated tide and Waimakariri River levels and much of the land is very low lying and just above mean sea level. Obviously, with any study such as this, when modelling physical systems there are a number of assumptions which need to be made. Hydrology (discussed in Section 3.2) has been assessed a number of times, with the latest being NIWA s 2002 analysis. Obviously with a relatively short local record there are uncertainties, particularly when estimating extreme events. Some sensitivity analysis of breakout flows, however, has been undertaken. Environment Canterbury Technical Report

42 Topography the LiDAR data has provided very good detail of the ground surface. However, when converting this data to a grid for modelling purposes, there can be some loss of definition. Checks have been made to ensure critical ground levels have been appropriately incorporated into the 10m grid. Breakout location and magnitude as indicated this is based on the risk assessment. While the most likely locations and midrange estimates for magnitude have been selected, a large number of scenarios are possible. Some sensitivity analysis has been undertaken of these. River bed levels cross-sections from the last full survey in 1997 were used in the river modelling. Checks with a selection of more recent surveys reveal some small changes in bed level. A further survey planned for 2008/2009 will provide a more up-to-date picture. However, there could be significant aggradation of the channel following a major flood event. Channel and floodplain roughness the values chosen are based on calibration and typical values, based on previous investigations and accepted values. Major changes in channel roughness and floodplain land use would affect channel and floodplain depths. Some sensitivity analysis has been undertaken of floodplain roughness. Using the above parameters the river and floodplain model has been built to represent reality as closely as possible. The model has been peer reviewed by DHI Water and Environment and found to be well developed, provides a realistic description of the physical situation and suitable for the purpose. A range of sensitivity analysis has been undertaken to test the validity of some of the assumptions. Not surprisingly the largest variation is a result of modified breakout flows and location. To test the validity of the river flow estimates, the March 1986 South Canterbury rainfall event (estimated to be year return period rainfall) was routed through a rainfall/runoff model). This predicted a peak flow of approximately 3800 m 3 /s which compared quite well with the estimated 0.5% (AEP 200 year return period) flow for the Ashley River of 4050 m 3 /s. (Note the South Canterbury flood event, which caused widespread flooding and damage was produced by weather conditions, which could also occur in the Ashley catchment.) Numerous flood scenarios have been modelled in conjunction with a range of development options in Kaiapoi. As discussed in Section inappropriately located development can have an adverse effect on adjacent areas. The predicted flood depths in some of these potential areas is also quite high, and would be regarded as high hazard locations. High hazard is generally defined as where the flood depth is greater than 1 metre or where the product of depth and velocity is greater than 1. (Refer to Appendix 7). Figures 17 and 18 show the hazard categories for the Ashley floodplain (south side) for the 0.5% AEP (200 year return period) and 0.2% AEP (500 year return period) events. In the 0.2% AEP event a large part of Kaiapoi and adjacent areas are classified as high hazard, due to ponded depths over 1 metre. A discussion on appropriate land use within the Ashley floodplain is not the intention of this report. The floodplain investigations undertaken, however, will hopefully aid decision-makers to make informed decisions on future land use. Environment Canterbury Technical Report 39

43 The floodplain maps and associated depths will also provide good information when setting minimum floor levels for any new dwellings located on the floodplain. The perception of a 200 and 500 year return period event is that they are very extreme and will never occur in my lifetime. While the probability of these events occurring in any one year is very low, the probability of occurrence over a person s lifetime, for example, is relatively high. For example, a 200 year return period event, has a 35% chance of occurring over a 70 year period. A table showing the likelihood of flooding for a range of different return periods is shown in Appendix 8. Environment Canterbury Technical Report

44 Figure 17: Environment Canterbury Technical Report 41

45 Figure 18: Environment Canterbury Technical Report

46 5. GLOSSARY Alluvial Fan: A cone-shaped deposit of alluvium (shingle and silt deposition) made by a stream or river where it runs out onto a level plain. The fans generally form where the stream or river issue from the hills onto the lowlands. Annual Exceedance Probability (AEP): The chance of a flood of a given or larger size occurring in any one year, usually expressed as a percentage. For example if a peak flood discharge of 500 m 3 /s has an AEP of 5%, it means there is a 5% chance (i.e. a chance of one-in-twenty) of a peak flood discharge or larger occurring in any one year. AEP is the inverse of return period, expressed as a percentage. Geomorphology: The study of the topographic and geologic form of the earth s surface and the changes that take place in the evolution of land forms. LiDAR: (Light Detection and Ranging) produces very detailed rapid survey of ground surface. Measurement is made from aircraft utilising laser technology and able to measure up to 180 million data points/hour. Standard accuracy is in the order of + 150mm or better. One Dimensional Model: Using river channel cross sections along a river, flows and depths are calculated to discrete points along the channel. Flow is in one direction only, i.e. x direction along the channel. Orographic Rain: Rainfall (often intense) resulting when moist air is forced to rise by hills/ranges/mountains. Residual Risk: The hazard a community is exposed to after floodplain management measures have been put in place. For example, for stopbanked areas the residual risk is the continuing flood hazard of stopbanks being overtopped and the associated consequences. Return Period: The average time period between floods, equivalent to or exceeding a given magnitude. For example, a 100 year return period flood has a magnitude expected to be equalled or exceeded on average of once every one hundred years. Such a flood has a 1% chance of being equalled or exceeded in any given year, i.e. 1% AEP. Often used interchangeably with recurrence interval or flood frequency. Two Dimensional Model: Using a rectangular grid of the floodplain topography flows are calculated in both the x and y direction within each grid point. Depths are also calculated within each grid point. Environment Canterbury Technical Report 43

47 Environment Canterbury Technical Report

48 Appendices Environment Canterbury Technical Report 45

49 Appendix 1: Floodplain Geomorphology Environment Canterbury Technical Report

50 Appendix 2: Ashley River Flood Levels ASHLEY RIVER BANK LEVELS & FLOOD LEVELS No Breakout Scenario (bed resist 0.03, Cone Rd & SH1 bridges modelled, 1997 cross sections) (modelled Tony Oliver August 2007) x/sect M11 Chainage Left Bank Right Bank Design Q 2400 m3/s 2400 m3/s 3000 m3/s 3000 m3/s 1% AEP 3470 m3/s 3470 m3/s Min FB Min FB Min FB Environment Canterbury Technical Report 47

51 Appendix 3: Floodplain Roughness Environment Canterbury Technical Report

52 Appendix 4: Ashley River and breakout flow hydrographs Environment Canterbury Technical Report 49

53 Environment Canterbury Technical Report

54 Environment Canterbury Technical Report 51

55 Appendix 5: Floodplain Roughness Sensitivity Analysis Kaiapoi Area Environment Canterbury Technical Report

56 Appendix 6: Breakout Flow Sensitivity Analysis Environment Canterbury Technical Report 53

57 Environment Canterbury Technical Report Appendix 7: Hazard Category Definition

58 Appendix 8: Hazard Category Definition Flood Probability AEP = Annual Exceedance Probability i.e. the chance of a flood that size occurring in any one year Event Probability of occurring in period % AEP (return period) 10 yr period 30 yr period 70 yr period 5% (20 yr) 40% 80% 97% 2% (50 yr) 20% 50% 77% 1% (100yr) 10% 25% 50% 0.5%(200yr) 5% 15% 33% 0.2% (500yr) 2% 6% 12% For example there is 25% chance that a 1% AEP (100 year return period)flood will occur within a 30 year period Environment Canterbury Technical Report 55

59 Environment Canterbury Technical Report

60 Environment Canterbury Technical Report

61