North Western - Neagh Bann CFRAM Study

UoM 36 Hydrology Report

IBE0700Rp0009

rpsgroup.com/ireland

North Western – Neagh Bann CFRAM Study

UoM 36 Hydrology Report DOCUMENT CONTROL SHEET

Client OPW

Project Title North Western – Neagh Bann CFRAM Study

Document Title IBE0700Rp0009_UoM 36 Hydrology Report_F03

Document No. IBE0700Rp0009

DCS TOC Text List of Tables List of Figures No. of This Document Appendices Comprises 1 1 130 1 1 4

Rev. Status Author(s) Reviewed By Approved By Office of Origin Issue Date

B. Quigley D01 Draft U. Mandal M. Brian G. Glasgow Belfast 08/11/2013 L. Arbuckle

B. Quigley F01 Draft Final U. Mandal M. Brian G. Glasgow Belfast 31/03/2014 L. Arbuckle B. Quigley F02 Draft Final U. Mandal M. Brian G. Glasgow Belfast 11/08/2015 L. Arbuckle

B. Quigley F03 Draft Final U. Mandal M. Brian G. Glasgow Belfast 08/07/2016 L. Arbuckle

rpsgroup.com/ireland

Copyright

Copyright - Office of Public Works. All rights reserved. No part of this report may be copied or reproduced by any means without prior written permission from the Office of Public Works.

Legal Disclaimer

This report is subject to the limitations and warranties contained in the contract between the commissioning party (Office of Public Works) and RPS Group Ireland

rpsgroup.com/ireland

NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

TABLE OF CONTENTS

LIST OF FIGURES ...... IV LIST OF TABLES ...... VI ABBREVIATIONS ...... IX 1 INTRODUCTION ...... 1 1.1 OBJECTIVE OF THIS HYDROLOGY REPORT ...... 3 1.2 SUMMARY OF THE AVAILABLE DATA ...... 3 1.2.1 Summary of Available Hydrometric Data ...... 3 1.2.2 Summary of Available Meteorological Data ...... 5 2 METHODOLOGY REVIEW ...... 8 2.1 HYDROLOGICAL ANALYSIS ...... 8 2.2 USE OF METEOROLOGICAL DATA ...... 9 2.3 DESIGN FLOW ESTIMATION ...... 9 2.3.1 Index Flood Flow Estimation ...... 9 2.3.2 Growth Curve / Factor Development ...... 11 2.3.3 Design Flow Hydrographs ...... 11 2.4 HYDROLOGY PROCESS REVIEW ...... 12 2.5 CATCHMENT BOUNDARY REVIEW AND HYDROLOGICAL ESTIMATION POINTS ...... 14 2.5.1 Catchment Boundary Review ...... 14 2.5.2 Hydrological Estimation Points ...... 16 3 HYDROMETRIC GAUGE STATION RATING REVIEWS ...... 18 3.1 METHODOLOGY ...... 18 3.2 RATING REVIEW RESULTS ...... 19 3.3 IMPACT OF RATING REVIEWS ON HYDROLOGICAL ANALYSIS ...... 22 4 INDEX FLOOD FLOW ESTIMATION ...... 24 4.1 MODEL 1 – BALLINAMORE ...... 25 4.2 MODEL 2 – ...... 28 4.3 MODEL 3 - BALLYBAY ...... 31 4.4 MODEL 4 – ...... 35 4.5 MODEL 5 – BUNDORAN AND TULLAGHAN ...... 40 4.5.1 Model 5a ...... 40 4.5.2 Model 5b ...... 40 4.5.3 Model 5c ...... 40 4.6 INDEX FLOOD FLOW CONFIDENCE LIMITS ...... 43 5 FLOOD FREQUENCY ANALYSIS AND GROWTH CURVE DEVELOPMENT ...... 44 5.1 OBJECTIVE AND SCOPE ...... 44 5.2 METHODOLOGY ...... 44 5.2.1 Selection of Statistical Distribution ...... 44

IBE0700Rp0009 i Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

5.2.2 Forming a Pooling Region and Groups ...... 44 5.2.3 Growth Curve Development ...... 44 5.2.4 Limitations in the FEH and FSU Studies ...... 45 5.3 DATA AND STATISTICAL PROPERTIES ...... 45 5.3.1 Flood Data ...... 45 5.3.2 Pooling Region Catchment Physiographic and Climatic Characteristic Data 49 5.3.3 Statistical Properties of the AMAX series ...... 51 5.4 STATISTICAL DISTRIBUTION ...... 52 5.5 GROWTH CURVE ESTIMATION POINTS ...... 54 5.6 POOLING REGION AND GROUP FOR GROWTH CURVE ESTIMATION ...... 55 5.6.1 Pooling Region ...... 55 5.6.2 Pooling Group...... 56 5.7 GROWTH CURVE ESTIMATION ...... 56 5.7.1 Choice of Growth Curve Distributions ...... 56 5.7.2 Estimation of Growth Curves ...... 57 5.7.3 Examination of Growth Curve Shape ...... 58 5.7.4 Recommended Growth Curve Distribution for UoM 36 ...... 62 5.8 RATIONALISATION OF GROWTH CURVES ...... 64 5.8.1 Relationship of Growth Factors with Catchment Characteristics ...... 64 5.8.2 Generalised Growth Curves ...... 65 5.8.3 Results for different parent pooling regions for UoM 36 ...... 69 5.8.4 Estimated Growth Factors for UoM 36 ...... 71 5.8.5 Comparison of the at-site growth curves with the pooled growth curves ...... 73 5.8.6 Growth factors for all HEPs in the UoM 36 ...... 77 5.9 COMPARISON WITH FSR GROWTH FACTORS ...... 83 5.10 GROWTH CURVE DEVELOPMENT SUMMARY ...... 83 6 DESIGN FLOWS ...... 85 6.1 DESIGN FLOW HYDROGRAPHS ...... 85 6.1.1 FSU Hydrograph Shape Generator ...... 85 6.1.2 FSSR 16 Unit Hydrograph Method...... 86 6.2 COASTAL HYDROLOGY ...... 88 6.2.1 ICPSS Levels ...... 88 6.2.2 ICWWS Levels ...... 89 6.2.3 Consideration of ICPSS and ICWWS Outputs ...... 90 6.3 JOINT PROBABILITY ...... 91 6.3.1 Fluvial – Fluvial ...... 91 6.3.2 Fluvial – Coastal ...... 92 7 FUTURE ENVIRONMENTAL AND CATCHMENT CHANGES ...... 94 7.1 CLIMATE CHANGE ...... 94 7.1.1 UOM 36 Context ...... 94

IBE0700Rp0009 ii Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

7.1.2 Sea Level Rise ...... 95 7.2 AFFORESTATION ...... 96 7.2.1 Afforestation in UoM 36 ...... 96 7.2.2 Impact on Hydrology ...... 99 7.3 LAND USE AND URBANISATION ...... 101 7.3.1 Impact of Urbanisation on Hydrology ...... 104 7.4 HYDROGEOMORPHOLOGY ...... 107 7.4.1 Soil Type ...... 107 7.4.2 Channel Typology ...... 108 7.4.3 Morphological Pressures - Land Use ...... 114 7.4.4 Arterial Drainage (Channelisation) ...... 116 7.4.5 Localised Pressures ...... 118 7.5 FUTURE SCENARIOS FOR FLOOD RISK MANAGEMENT ...... 119 7.6 POLICY TO AID FLOOD REDUCTION ...... 120 8 SENSITIVITY AND UNCERTAINTY ...... 121 8.1 UNCERTAINTY / SENSITIVITY ASSESSMENT MODEL BY MODEL ...... 122 8.2 CONCLUSIONS OF SENSITIVITY ANALYSIS ...... 124 9 CONCLUSIONS ...... 125 9.1 SUMMARY OF THE RESULTS AND GENERAL PATTERNS ...... 125 9.2 RISKS IDENTIFIED ...... 126 9.3 OPPORTUNITIES / RECOMMENDATIONS ...... 127 10 REFERENCES: ...... 129

IBE0700Rp0009 iii Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

LIST OF FIGURES

Figure 1.1: UoM 36 AFA Locations and Extents ...... 2

Figure 1.2: Hydrometric Data Availability ...... 4

Figure 1.3: Meteorological Data Availability ...... 7

Figure 2.1: Hydrology Process Flow Chart ...... 13

Figure 2.2: Drowes / Lough Melvin Catchment Boundary Review ...... 15

Figure 2.3: UoM 36 Catchment Boundary Comparison ...... 16

Figure 4.1: UoM 36 Watercourses to be Modelled ...... 24

Figure 4.2: Model 1 HEPs and Catchment Boundaries ...... 25

Figure 4.3: Model 2 HEPs and Catchment Boundaries ...... 28

Figure 4.4: Model 3 HEPs and Catchment Boundaries ...... 32

Figure 4.5: Model 4 HEPs and Catchment Boundaries ...... 36

Figure 4.6: Model 5 HEPs and Catchment Boundaries ...... 42

Figure 5.1: Locations of 248 Gauging Stations ...... 46

Figure 5.2: Relative frequencies of catchments sizes (AREA) within the Pooling Region 1 (54 stations) ...... 50

Figure 5.3: Relative frequencies of the SAAR values within the Pooling Region 1 (54 stations) ...... 50

Figure 5.4: Relative frequencies of the BFI values within the Pooling Region 1 (54 stations) ...... 51

Figure 5.5: L-Moment Ratio Diagram (L-CV versus L-Skewness) for 54 AMAX series in Pooling Region 1 ...... 52

Figure 5.6: Spatial distribution of the HEPs and GC_EPs on modelled watercourses in UoM 36 ...... 55

Figure 5.7: L-moment ratio diagram (L-skewness versus L-Kurtosis) ...... 57

Figure 5.8: Pooled Growth Curve EP 56- (a) EV1 and GEV distributions; (b) GLO distributions ...... 60

Figure 5.9: Comparison of EV1, GEV and GLO growth curves on the EV1-y probability plot (Growth Curve EP No. 56) ...... 63

Figure 5.10: Relationship of growth factors with catchment areas for 74 HEPs ...... 64

IBE0700Rp0009 iv Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 5.11: Relationship of growth factors with SAAR for 74 HEPs ...... 64

Figure 5.12: Relationship of growth factors with BFI for 74 HEPs ...... 65

Figure 5.13: Relationship of growth factors with catchment areas (for 337 growth curve estimation points) ...... 66

Figure 5.14: Comparison of growth curves for different parent pooling regions ...... 70

Figure 5.15: GLO growth curves for all Growth Curve Groups (5 No.) ...... 71

Figure 5.16: Growth Curve for GC Group No. 4 with 95% confidence limits ...... 73

Figure 5.17: The at-site and pooled frequency curves along with the 95% confidence intervals ...... 75

Figure 6.1: Various AEP Hydrographs for Tributary HEP 36_1762_2_RA (Model 3) ...... 86

Figure 6.2: Location of ICPSS Nodes in Relation to Coastal AFAs ...... 88

Figure 6.3: Draft ICWWS potential areas of vulnerable coastline ...... 89

Figure 6.4: Typical 1% AEP Coastal Boundary Makeup (to Staff Gauge Zero) ...... 90

Figure 7.1: CORINE 2006 Forest Coverage in UoM 36 Compared to the rest of Ireland ...... 97

Figure 7.2: Forest Coverage Changes in UoM 36 ...... 98

Figure 7.3: UOM 36 CORINE Artificial Surfaces (2000 / 2006) ...... 103

Figure 7.4: UoM 36 Soil Types (Source: Irish Forest Soils Project, FIPS – IFS, Teagasc, 2002) ...... 108

Figure 7.5: Channel Types of UoM 36 in National Context (Source: WFD Channel Typology dataset) ...... 110

Figure 7.6: Channel Slopes of UoM 36 in National Context (Source: WFD Channel Typology dataset) ...... 111

Figure 7.7: WFD Channel Typology UoM 36 ...... 112

Figure 7.8: Changes in Channel Slope UoM 36 ...... 113

Figure 7.9: UoM 36 Land Use (CORINE 2006) ...... 115

Figure 7.10: Arterial Drainage Schemes within UoM 36 Modelled Catchments ...... 117

IBE0700Rp0009 v Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

LIST OF TABLES

Table 1.1: Fluvial and Coastal Flood Risk at each AFA 3

Table 2.1: UoM 36 Models with Gauging Stations 10

Table 2.2: Summary of Catchment Boundary Review 14

Table 3.1: Existing Rating Quality Classification for Rating Review Stations in UoM 36 19

Table 3.2: AMAX Series Data Before and After Rating Review 20

Table 3.3: Summary of Rating Review Effects and Mitigation 22

Table 4.1: Qmed Values for Model 1 26

Table 4.2: Qmed Values for Model 2 29

Table 4.3: Qmed Values for Model 3 34

Table 4.4: Qmed Values for Model 4 38

Table 4.5: Qmed Values for Model 5 41

Table 5.1: Alternative parent pooling regions 45

Table 5.2: Hydrometric Station Summary for Pooling Region 1 (54 sites) 47

Table 5.3: Summary of Catchment physiographic and climatic characteristics of Pooling Region (54 sites) 49

Table 5.4: Statistical properties of 54 AMAX Series in Pooling Region 1 51

Table 5.5: Summary results of probability plots assessments (EV1, LN2, GEV & GLO distributions) for all 54 AMAX series in Pooling Region 1 53

Table 5.6: Summary of the catchment characteristics associated with the 118 HEPs 54

Table 5.7: Growth curves shape summary 59

Table 5.8: Catchment descriptors for all pooled sites for GC EP No. 56 59

Table 5.9: Frequency curve shapes of the individual site’s AMAX series associated with the Pooled Growth Curve EP 56 61

Table 5.10: Estimated growth factors for Growth Curve No. 56 62

IBE0700Rp0009 vi Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Table 5.11: Growth curve estimation summary 67

Table 5.12: Growth Curve (GC) Groups 69

Table 5.13: Growth factors for range of AEPs 71

Table 5.14: Estimated percentage standard errors for growth factors (XT) for a range of AEPs (source FSU Work- Package 2.2 “Frequency Analysis” Final Report – Section 13.3) 72

Table 5.15: Hydrometric gauging stations located on the modelled watercourses in UoM 36 hydrometric area 74

Table 5.16: Growth factors for all 118 HEPs for a range of AEPs for UoM 36 77

Table 5.17: Study growth factors compared with FSR growth factors 83

Table 6.1: ICPSS Level in Close Proximity to UoM 36 AFAs / HPWs 89

Table 6.2: Initial Screening for Relevance of Joint Probability 92

Table 7.1: Afforestation from 2000 to 2006 99

Table 7.2: Allowances for Effects of Forestation / Afforestation (100 year time horizon) 100

Table 7.3: Population Growth in UoM 36 (Source: Central Statistics Office of Ireland (CSO)) 101

Table 7.4: Population Growth within Urban AFAs (Source: Source: Central Statistics Office of Ireland (CSO)) 102

Table 7.5: Urbanisation Growth Indicators 104

Table 7.6: Potential Effect of Urbanisation on Qmed Flow on the Cavan River (at confluence with Annalee River) 105

Table 7.7: Potential Effect of Urbanisation on Qmed Flow in UoM 36 106

Table 7.8: Channel Types and Associated Descriptors 109

Table 7.9: UoM 36 Allowances for Future Scenarios (100 year time horizon) 119

Table 8.1: Assessment of contributing factors and cumulative effect of uncertainty / sensitivity in the hydrological analysis 122

IBE0700Rp0009 vii Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

APPENDICES APPENDIX A UOM 36 Hydrometric Data Status Table 1 Page

APPENDIX B Rating Reviews 13 Pages

APPENDIX C NAM Output 2 Pages

APPENDIX D Design Flows for Modelling Input 22 Pages

IBE0700Rp0009 viii Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

ABBREVIATIONS

AEP Annual Exceedance Probability

AFA Area for Further Assessment

AFF At-site Flood Frequency

AMAX Annual Maximum flood series

AREA Catchment Area

BFI Base Flow Index

CFRAM Catchment Flood Risk Assessment and Management

CORINE Coordination of Information on the Environment

C4i Community Climate Change Consortium for Ireland

DTM Digital Terrain Model

EV1 Extreme Value Type 1 (distribution) (=Gumbel distribution)

EPA Environmental Protection Agency

FARL Flood Attenuation for Rivers and Lakes

FEH Flood Estimation Handbook

FRA Flood Risk Assessment

FRMP Flood Risk Management Plan

FSR Flood Studies Report

FSSR 16 Flood Studies Supplementary Report No. 16

FSU Flood Studies Update

GC Growth Curve

GDSDS Greater Dublin Strategic Drainage Study

GEV Generalised Extreme Value (distribution)

GLO General Logistic (distribution)

GSI Geological Survey of Ireland

HA Hydrometric Area

HEFS High End Future Scenario (Climate Change)

HEP Hydrological Estimation Point

HPW High Priority Watercourse

HWA Hydrograph Width Analysis

IBE0700Rp0009 ix Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

IH124 Institute of Hydrology Report No. 124

IPCC Intergovernmental Panel on Climate Change

LA Local Authority

LN2 2 Parameter Log Normal (distribution)

L-CV Coefficient of L variation

MPW Medium Priority Watercourse

MRFS Mid Range Future Scenario (Climate Change)

NBIRDB Neagh Bann International River Basin District

NDTM National Digital Terrain Model

NWIRBD North Western International River Basin District

OD Ordnance Datum

OPW Office of Public Works

OSi Ordnance Survey Ireland

PCD Physical Catchment Descriptor

PFRA Preliminary Flood Risk Assessment

Qmed median of AMAX flood series

Qbar / QBAR mean average of AMAX flood series

RBD River Basin District

RFF Regional Flood Frequency

ROI Region of Influence

SAAR Standard Average Annual Rainfall (mm)

SuDS Sustainable Urban Drainage

UAF Urban Adjustment Factor

UoM Unit of Management

IBE0700Rp0009 x Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

1 INTRODUCTION

The Office of Public Works (OPW) commissioned RPS to undertake the North Western – Neagh Bann Catchment Flood Risk Assessment and Management (CFRAM) Study in March 2012. The North Western – Neagh Bann CFRAM Study was the sixth and last CFRAM Study to be commissioned in Ireland under the EC Directive on the Assessment and Management of Flood Risks, 2007 as implemented in Ireland by SI 122 of 2010 European Communities (Assessment and Management of Flood Risks) Regulations, 2010.

The North Western IRBD covers an area of 12,320 km2 with approximately 7,400 km2 of that area in the . It includes two Units of Management (UoMs), UoM 01 (Donegal) and UoM 36 (Erne). It takes in all of as well as parts of Leitrim, Cavan, Monaghan, Longford and Sligo. There is a high level of flood risk within the North Western IRBD, with significant coastal flooding in County Donegal as well as areas of fluvial flooding throughout the district.

The Neagh Bann IRBD covers an area of 8,120 km2 with approximately 2,010 km2 of that area in Ireland. It represents one single Unit of Management, UoM 06 (Neagh Bann).

This hydrology report covers UoM 36 which includes hydrometric areas 35 and 36. It covers an area of 2,742 km2 and includes the majority of as well as areas of counties Leitrim, Monaghan, Longford, Donegal and Sligo. The principal river in UoM 36 is the Erne (which drains part of County Cavan before crossing the border into near ). In addition to the Erne River system, there are numerous smaller rivers and streams including the Annalee, Woodford and Finn rivers. Lakes in UoM 36 include Lough Oughter, Lough Melvin and Lough Gowna as well as numerous other smaller lakes.

UoM 36 is predominantly rural with the largest urban areas being Cavan town and Ballyshannon. The fertile soils of the Erne basin are capable of supporting intensive agriculture.

Within UoM 36 there are five Areas for Further Assessment (AFA) as shown in Figure 1.1. These are: Cavan Town and Ballyconnell in County Cavan, Bundoran & Tullaghan in County Donegal, Ballinamore in County Leitrim and Ballybay in County Monaghan. There are no Individual Risk Receptors identified for further assessment within UoM 36.

IBE0700Rp0009 1 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 1.1: UoM 36 AFA Locations and Extents

In accordance with the North West – Neagh Bann River Basin Districts Catchment-based Flood Risk Assessment and Management (CFRAM) Study, Stage II Project Brief (hereinafter referred to as the North Western – Neagh Bann CFRAM Study Brief) only those areas not afforded protection by existing or planned schemes are considered in full as part of this Study. For other areas within AFAs benefiting from existing flood relief schemes, assessment under the North Western – Neagh Bann CFRAM Study will be limited to development and appraisal of maintenance and management options and the consideration of any implications associated with potential development as identified in relevant spatial planning documents. No such areas have been identified within the North Western – Neagh Bann study area. It should be noted that areas subject to minor works are not considered as having schemes in place. The AFAs and the flood risk sources to be considered within UoM 36 as part of this study are listed in Table 1.1 which indicates a predominance of fluvial flood risk with only Bundoran and Tullaghan AFA also at risk from coastal flooding.

IBE0700Rp0009 2 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Table 1.1: Fluvial and Coastal Flood Risk at each AFA

AFA Fluvial Coastal AFA Fluvial Coastal

Bundoran and   Cavan  - Tullaghan

Ballyconnell  - Ballybay  -

Ballinamore  - Total 5 1

1.1 OBJECTIVE OF THIS HYDROLOGY REPORT

The principal objective of this Hydrology Report is to provide detail on the outputs from the processes of hydrological analysis and design flow estimation. The details of the methodologies used and the preliminary hydrological analysis are provided in the Inception Report ‘IBE0700Rp0004_UoM 36 Inception Report_F02’ (RPS, 2013). This report provides a review and summary of the methodologies used as well as details of any amendments to the methodologies since completion of the Inception Report. The report will provide details of the results of the hydrological analysis and design flow estimation and summarise the outputs from the analysis which will be taken forward as inputs for the hydraulic modelling. Discussion will be provided within this report on the outputs in terms of the degree of confidence which can be attached to the outputs and the opportunities for providing greater certainty for future studies, including opportunities for improving the observed data used to inform the study.

This report does not include details of the data collection process, flood history within the AFAs or methodology and results from the historic flood analysis as this is contained within the Inception Report for UoM 36.

1.2 SUMMARY OF THE AVAILABLE DATA

1.2.1 Summary of Available Hydrometric Data

Hydrometric data is available at 41 hydrometric gauge station locations within UoM 36 as shown in Figure 1.2 below. Thirty-six of these stations have water level and flow data available, two of which are operated by the Department of Agriculture and Rural Development (DARD) Rivers Agency (Northern Ireland).

IBE0700Rp0009 3 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 1.2: Hydrometric Data Availability

IBE0700Rp0009 4 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Of the 36 stations with flow data available, 12 stations are located on watercourses to be modelled or just upstream of the modelled reach. Seven of these stations were rated under FSU as having a rating classification such that there was sufficient confidence in the rating for use within FSU (A1, A2 or B). Three of the stations were given a rating of A1 indicating that there is confidence in the rating up to at least 1.3 x Qmed. This is the highest classification for stations taken forward for use within FSU and indicates certainty in flood flows recorded for extreme flood events (above Qmed). A further three of the stations were rated as A2 indicating confidence in flows up to a maximum of 1.3 x Qmed. One of the stations has a rating of B indicating that confidence in the rating is limited to Qmed. This is the minimum classification for which stations were taken forward for use within FSU and indicates that there is uncertainty in the flood flows recorded for extreme flood events (above Qmed).

1.2.2 Summary of Available Meteorological Data

Observed rainfall data from a number of different sources is available within and in close proximity to UoM 36 (refer to Figure 1.3 overleaf):

 Met Éireann daily and hourly rainfall gauges within the North Western IRBD and beyond. There are two hourly gauges within UoM 36 - Clones and hourly stations are located in Counties Monaghan and Cavan.

 National Roads Authority sub-daily precipitation sensor information has become available since the project inception phase. Data has been received for two locations within UoM 36. The information consists of varying time steps but generally at 20 minutes and 1 hour spacing. The information is of unknown accuracy as the sensor technology has been developed primarily for the identification of precipitation type rather than high accuracy rainfall recording.

 The UK Met Office daily and hourly rainfall gauge information for gauges within Northern Ireland but in close proximity to the border has become available since the project inception phase. Two hourly gauges are in proximity to the extents of UoM 36 at Derrylin, and Castlederg.

Historical time series rainfall data can be used as an input to catchment scale hydrological rainfall runoff models to simulate a continuous flow record within a catchment. High resolution temporal data is required to achieve the required accuracy within the hydrological models and as such hourly time series data is required. Daily rainfall data is not considered to be of a high enough temporal resolution to be used as direct input for hydrological modelling on its own but can be used along with the hourly data to inform the spatial distribution of hourly rainfall data within the catchments. In relation to UoM 36 the only hourly rainfall station used in hydrological analysis is the Met Éireann hourly gauge at Clones (1951 – 2008). It is the closest gauge to the only rainfall runoff model that was constructed (to inform the hydraulic model for Ballybay AFA) and is of high enough temporal resolution and accuracy to be of use whereas as the Ballyhaise record is so short that it would be of limited use in deriving an extended AMAX series through rainfall run-off modelling. The hourly rainfall dataset from the Clones gauge is

IBE0700Rp0009 5 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL processed and supplied by the national meteorological authority (Met Éireann) and is considered to be of high accuracy. Given that this data is being used as input to a rainfall run-off model which is used as a complimentary design flow estimation technique, checking for errors in the data is undertaken through the calibration of the rainfall run-off model and only where it is found that calibration to hydrometric data could not be achieved. Further details of the rainfall runoff model constructed within UoM 36 are in Section 4.1 and Appendix C.

A data collection meeting held prior to commencement of the Study (between RPS, HydroLogic, OPW and Met Éireann) identified an opportunity for exploring the use and benefits of rainfall radar data in hydrological analysis. Trials undertaken within the Eastern CFRAM Study area demonstrated that there were benefits to be had by using gauge adjusted radar as opposed to using rain gauge data only to drive rainfall runoff models. RPS reviewed the extents of the radar coverage in relation to the NW – NB Study area and found there to be some coverage of the NW Study area from both the Met Office radar at Castor Bay and the Met Éireann radar at Dublin Airport. However the study area is generally well gauged and there were only a few locations where the use of high resolution rainfall data would be required to produce a simulated, calibrated and extended hydrometric gauge record. In addition there is already high temporal resolution rainfall data from the hourly gauge in close proximity to the potentially benefitting catchments. Processing of the radar records into gauge adjusted, gridded and catchment aggregated time series was therefore not considered necessary within the Study area.

In addition to the observed historical rainfall data available at the aforementioned rain gauge locations, further meteorological information is required as input to hydrological models namely observed evaporation, soil moisture deficits and potential evapotranspiration data. Historical time series data is available for these parameters at Met Éireann synoptic weather stations. The locations at which historical data is available are generally the same as for hourly rainfall data. Figure 1.3 shows the locations of all of the rain gauges available and the availability of historic information at the hourly rainfall gauges.

IBE0700Rp0009 6 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 1.3: Meteorological Data Availability

IBE0700Rp0009 7 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

2 METHODOLOGY REVIEW

The methodologies for hydrological analysis and design flow estimation were developed based on the current best practice and detailed in the UoM 36 Inception Report. In the intervening period there have been a number of developments both in best practice, and the hydrological analysis tools which are available such that it is prudent that the overall methodology is reviewed and discussed. As well as a review of the methodology this chapter seeks to discuss amendments to the catchment boundaries that have become apparent and must be considered in the hydrological analysis.

2.1 HYDROLOGICAL ANALYSIS

The main tasks of hydrological analysis of existing gauge data have been undertaken based on the best practice guidance for Irish catchments contained within the Flood Studies Update. The analysis of the data available from the hydrometric gauge stations shown in Figure 1.2 has been carried out based on the guidance contained within FSU Work Packages 2.1 ‘Hydrological Data Preparation’ and 2.2 ‘Flood Frequency Analysis’ and is detailed in Chapter 5. This analysis was undertaken prior to the receipt of survey information which would have allowed the progression of the North Western – Neagh Bann CFRAM Study gauge station rating reviews identified within the UoM 36 Inception Report. Following completion of the rating reviews there was not found to be significant uncertainty in the ratings at two of the three stations however at one station the previous rating was found not to be relevant to the present day hydrological conditions. The rating reviews, the new rating relationships and the consequences of the rating reviews for hydrological analysis are discussed in detail in chapter 3 of this report. The following elements of hydrological analysis have been assessed against the potential impact of uncertainty in the rating and mitigation measures and / or re-analysis undertaken to ensure the robustness of the hydrological analysis:

 Gauged Index Flood Flow (Qmed) – Where there has been shown to be uncertainty in the

rating within the range of flows up to and around Qmed, the Annual Maxima (AMAX) flow series

has been re-processed using the revised rating. The use of the gauged Qmed in design flow estimation is further discussed in 2.2.1.

 Single site (historic) flood frequency analysis – As the estimated frequency of a flood event is a function of the ranking of the event within the AMAX series, and this will not change following re-processing of the AMAX series, this will have little impact on the outputs of this study.

 Growth Curve Development – The inclusion of gauge years within pooled flood frequency analysis that have a high degree of uncertainty could have a skewing effect within the frequency analysis but the effect will be diluted within a group (where it is assumed other gauge years have a high degree of confidence). The cumulative effect of uncertainty in both directions at multiple gauges may also have a cancelling out effect within a pooling group and as such it is not necessary to re-analyse the pooling groups. However where growth curves

IBE0700Rp0009 8 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

are based on a single site analysis where it has been shown that there is uncertainty in the rating, the single site analysis has been re-analysed with the re-processed AMAX data based on the revised rating relationship.

2.2 USE OF METEOROLOGICAL DATA

Chapter 1.2.2 details where high resolution temporal rainfall data required as input to rainfall runoff models is available within UoM 36 and discusses how the availability of hourly stations within close proximity to potentially benefitting catchments negates the need for radar data. The good availability of meteorological data, both daily and hourly within and in close proximity to UoM 36 provides the high temporal resolution data needed for driving the one rainfall runoff model that has been undertaken at station 36150. This was the only location identified where rainfall data could be used within a calibrated, hydrological model to enhance the available flow data through simulation of a long term record. Elsewhere, the good availability of A1 and A2 stations already provides high confidence in flow data such that there is no need for additional hydrological modelling.

Within the NW-NB CFRAM Study methodology rainfall runoff data is used within calibrated (to hydrometric gauge data) hydrological catchment models to provide additional simulated catchment flow data to bring greater confidence to statistical design flow estimates and provide additional (simulated) historical flow data for model calibration. In the case of UoM 36, most of the hydrometric stations located on modelled watercourses have a high level of confidence associated with them. To this end, only one rainfall runoff model was deemed to have potential benefit in augmenting the AMAX series and associated confidence in flow data of Station 36150 on the Shantonagh River. Hourly rainfall data from the Met Eireann Clones gauge was used to provide detailed temporal resolution in this model.

2.3 DESIGN FLOW ESTIMATION

The estimation of design flows is based on the best practice guidance for Irish catchments generally as outlined in the Flood Studies Update (FSU) and supplemented with other methodologies where these are considered more appropriate. The methodologies for estimation of the various elements which make up the design flow estimates to be used for hydraulic modelling are detailed below.

2.3.1 Index Flood Flow Estimation

Estimation of the Index Flood Flow is required for all catchments and sub-catchments to be analysed under the CFRAM Study with each sub-catchment defined by a Hydrological Estimation Point (HEP). The preferred methodologies for estimation of design flow vary depending on the size, whether or not the catchment is gauged and also based on how the runoff from the catchments impacts upon the AFA. However a comprehensive, hierarchical approach is being taken to index flood flow estimation whereby all the specified methodologies available at each HEP are employed to estimate the index flood flow and to provide robustness to the estimates. For example, in the first instance, the FSU 7-

IBE0700Rp0009 9 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL variable ungauged catchment descriptor equation (Work Package 2.3) is used to calculate an estimate of the Index Flood Flow at all HEPs and where available, gauge records, rating reviews and other applicable methodologies are used to adjust / improve the estimate as the design flow estimation is developed. The hierarchy of preferred methodologies is discussed as follows.

2.3.1.1 Gauged Index Flood Flow (Qmed)

HEPs have been located at all hydrometric gauging stations where flow data is available. In the case of UoM 36 there are 12 gauging stations with flow data located directly on (or just upstream of) modelled watercourses, three of which are subject to a review of the rating using hydraulic modelling. Following rating review it can be considered that these gauging stations will have confidence in the rating at Qmed or above. All five designated fluvial models include watercourses which are gauged as indicated by Table 2.1.

Table 2.1: UoM 36 Models with Gauging Stations

Model AFA Station FSU Rating Review Number Classification 1 Ballinamore 36028 Aghoo - No 2 Ballyconnell 36027 Bellaheady A2 Yes 3 Ballybay 36150 Shantonagh Br. - Yes 36018 Ashfield A1 No 4 Cavan 36036 Kilconny - No 36019 Belturbet A1 No 36037 Urney Bridge - No 36010 Butlers Bridge A1 No 36031 Lisdarn A2 Yes 36016 Rathkenny B No 36018 Ashfield A1 No 5 Bundoran and 35029 Mullanaleck Br. - No Tullaghan 35071 Lareen A2 No

2.3.1.2 Ungauged Index Flood Flow (Qmed)

At all catchments the ungauged catchment descriptor based method FSU WP 2.3 ‘Flood Estimation in Ungauged Catchments’ has been used, to derive estimates of Qmed, including small ungauged catchments. This is in accordance with recently published guidance “Guidance Note 21 - CFRAM guidance note on flood estimation for ungauged catchments”. This guidance note drew on the finding that alternative methods for small catchments (Flood Studies Report, NERC, 1975; IH Report 124, Marshall and Bayliss, 1994) do not have enough empirical support in Ireland and draw on older and cruder datasets than FSU. Therefore, in the first instance, the FSU 7-variable ungauged catchment

IBE0700Rp0009 10 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL descriptor equation (Work Package 2.3) is used to calculate an estimate of the Index Flood Flow at all HEPs and where available, gauge records or catchment runoff models are used to adjust / improve the estimate as the design flow estimation is developed.

The FSU methodology outlined in WP 2.3 recommends that estimates based on the seven parameter catchment descriptor equation are adjusted based on the most hydrologically similar gauged site. The adjustment factor is applied to the regression equation estimate at the subject catchment and can be described in simple terms as the gauged Qmed divided by the regression equation estimated Qmed at the most hydrologically similar gauged site. Hydrological analysis tools developed by OPW as part of the FSU identify 216 gauge locations which are described as ‘Pivotal Sites’ following analysis of the data available as part of FSU WP 2.1 ‘Hydrological Data Preparation’.

2.3.2 Growth Curve / Factor Development

Growth curves have been developed based on single site and pooled analysis of gauged hydrometric data based on the FSU methodology set out in Work Packages 2.1 and 2.2. Full details and discussion of the results can be found in Chapter 4.

2.3.3 Design Flow Hydrographs

The design flow hydrograph methodology for the NW-NB CFRAM Study centres around FSU Work Package 3.1 ‘Hydrograph Width Analysis’ and uses the tools developed by OPW for analysing flood hydrographs at gauged sites. Since the completion of the Inception Report the methodology for deriving design flow hydrographs has been developed further following the release of the FSU Hydrograph Shape Generator (version 5). As such the hydrograph shapes are generated based on the following methods:

1. At HEPs representing larger catchments (generally 10km2 or larger) within UoM 36 hydrographs will be generated using the recently released Hydrograph Shape generator (version 5) developed by OPW. This tool increases the list of Pivotal Sites from which median hydrograph shape parameters can be borrowed based on the hydrological similarity of the Pivotal Site when compared to the subject site. The release of version 5 of this tool has increased the pool of Pivotal Sites to over 150. RPS trialling of this version of the FSU Hydrograph Shape Generator in CFRAMS has found that the generated hydrograph shapes provide a reasonably good fit when compared to the observed and simulated (NAM) hydrographs across the Eastern and South Eastern Study areas.

2. At HEPs representing smaller catchments (generally less than 10km2) it may not be possible to find a suitable Pivotal Site from which a comparable hydrograph shape can be borrowed, particularly for the very small sub-catchments representing tributary headwaters. In this instance hydrograph shapes have been generated using the Flood Studies Supplementary Report (FSSR) 16 Unit Hydrograph method.

IBE0700Rp0009 11 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Design hydrographs have been developed at all HEPs. It was originally intended that at the smallest inflow / tributary HEPs that continuous point flows could be input. However analysis of this method found that the hydrograph was critical in some of the smallest watercourses which are restricted by culverts / bridges where flood volume as opposed to flood flow becomes the critical characteristic of a flood. Examples of this are urban watercourses within AFAs where existing culvert and channel structures may surcharge and as such the event flood volume may be a critical factor. Application of continuous point flows on the upstream reaches of the hydraulic models could lead to an unrealistic build up of water behind culvert structures where this is the critical flood mechanism.

2.4 HYDROLOGY PROCESS REVIEW

Following developments in best practice and guidance documents and the refinement of RPS methodology through its application on the NW-NB CFRAM Study the hydrology process has been amended slightly from that which has been presented in the UoM 36 Inception Report (summarised previously in Figure 5.2 of report IBE0700Rp0002_UoM 36 Inception Report_F02). The revised process flow chart which has been applied in carrying out the hydrological analysis and design flow estimation for UoM 36 is presented in Figure 2.1. It is worth noting that the core methodologies employed within the Study are statistically based. These approaches do not require the identification of critical storms as the method ensures that the correct frequency conditions are achieved through checking the developing modelled hydrograph moving down through the catchment and adjusting the timings and peaks on the lateral inflow and tributary point inflows where necessary. This is the process shown in boxes 14 and 15 within Figure 2.1.

IBE0700Rp0009 12 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 2.1: Hydrology Process Flow Chart

IBE0700Rp0009 13 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

2.5 CATCHMENT BOUNDARY REVIEW AND HYDROLOGICAL ESTIMATION POINTS

In line with the CFRAM Study Stage 1 Project Brief (ref. 2149/RP/002/F, May 2010) Section 6.3, RPS delineated the catchment boundaries at HEPs using the FSU derived ungauged and gauged catchment boundaries as a starting point. In addition to the FSU delineated catchments, sub-catchments relating to cross border catchments were also provided by Rivers Agency and where provided these tended to capture cross-border catchments more accurately and as such were used as the starting point for review. For details of the full methodology for undertaking this review see UoM 36 Inception Report Section 5.3.2.

2.5.1 Catchment Boundary Review

Following the completion of the review process a number of the catchment boundaries were amended and in several cases the catchments boundaries were changed by more than 10%. Table 2.2 gives a summary of the changes in the catchment area at CFRAMS HEP points when compared to the equivalent FSU / Rivers Agency catchment from which they were derived.

Table 2.2: Summary of Catchment Boundary Review

Change in Catchment Area Number of HEPs % of HEPs

New Catchment Delineated 21 18

No change 13 11

0 – 10% 60 51

Greater than 10% 24 20

Total 118 100

Not all the catchments related to HEPs that are required to be considered within UoM 36 were previously delineated. Some of the catchments relate to small streams and land drains which were too small to be considered under FSU and as such RPS defined these previously un-delineated HEP catchments using a combination of mapping, aerial photography and the National Digital Height Model (NDHM). In addition many of the cross border catchments were not captured accurately appearing to be cut-off at either the border / boundary of UoM 36 or at the extents of the NDHM. As discussed, the Rivers Agency provided catchment boundaries for all of the cross border catchments where these eventually discharged to the sea in Northern Ireland which aided delineation of cross border catchments in particular. One example of a catchment which was required to be delineated across the border was the Drowes River / Lough Melvin catchment at the upstream extent of the Bundoran and Tullaghan AFA. A large portion of Lough Melvin is located in Northern Ireland and the catchment which drains to this portion is not considered in the FSU ungauged catchment

IBE0700Rp0009 14 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

shapefiles used as the basis for catchment delineation. The FSU and Study delineated catchments are shown in Figure 2.2.

Figure 2.2: Drowes / Lough Melvin Catchment Boundary Review

The review concluded that 18% of HEPs required new delineation. For the rest of the HEPs, the pre-existing FSU ungauged catchments were already accurately delineated in 62% of cases but the remainder required modification by more than 10% since they were found not to be representative of the NDHM, the mapping or draft survey information. The most common reason for amendment in the case of UoM 36 was replacement of the FSU catchment with the equivalent Rivers Agency catchment, which occurred in 78 cases without further RPS change and a further 36 cases with additional changes made by RPS.

Figure 2.3 provides an overview of the difference between CFRAM Study catchments (RPS) and the FSU catchments from which they originated. Finalised catchment boundaries have been issued to OPW such that they can inform other studies and neighbouring UoM boundaries.

IBE0700Rp0009 15 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 2.3: UoM 36 Catchment Boundary Comparison

2.5.2 Hydrological Estimation Points

HEPs are defined within hydraulic models at their upstream limits, at tributary confluences (greater than 5km2), at intermediate locations along modelled reaches, at hydrometric stations and at their downstream limits. For full details on this refer to the UoM 36 Inception Report. Each defined HEP was given a Node Identification Code for the CFRAM Study termed “NODE_ID_CFRAMS”. The starting point for this ID code was the FSU NODE ID at which gauged or ungauged catchment descriptors are defined. This ID is in three parts as follows:

e.g. 36_1234_1 where ‘36’ denotes the relevant hydrometric area, ‘1234’ denotes the river ID and ‘1’ denotes the position of the FSU node along the river centreline (for gauged HEPs, the “36_1234” notation is replaced with the station number). This NODE ID was used in the first instance for HEP identification but was adapted for the CFRAM Study under the following conditions:

 36_1234 - catchment descriptors and catchment area are based on FSU database;

 36_1234_RPS – catchment descriptors and catchment area are based on FSU database but the catchment area has been edited by RPS;

IBE0700Rp0009 16 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

 36_1234_RA – catchment descriptors and catchment area are based on FSU database but the catchment area has been replaced with that of the catchment provided by Rivers Agency;

 36_1234_RARPS – catchment descriptors and catchment area are based on FSU database but catchment area has been replaced with that of the catchment provided by Rivers Agency and has been further edited by RPS.

The IDs for each HEP are tabulated for each hydraulic model in Chapter 4 and Appendix D.

IBE0700Rp0009 17 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

3 HYDROMETRIC GAUGE STATION RATING REVIEWS

As a follow on from the recommendations of Work Package 2.1 of the FSU, a task was included in the North Western – Neagh Bann CFRAM Study brief to undertake further rating review of a subset of hydrometric gauging stations. Following the completion of the risk review stage and finalisation of the AFA locations three hydrometric stations were specified for rating review. The three stations to be taken forward for review were chosen for rating review by OPW as they had available continuous flow data, were located on (or just upstream or downstream of) watercourses to be modelled and were deemed under FSU Work Package 2.1 as currently having a rating quality classification that could be improved upon (i.e. there may be some uncertainty in the rating at extreme flood flows).

3.1 METHODOLOGY

The methodology for carrying out rating reviews entails the following general steps:

1. Gauge station reach of watercourse is surveyed in detail (site visit, cross sections and LiDAR survey). Rating review survey is prioritised ahead of survey required for hydraulic modelling.

2. A hydraulic model is constructed of the reach of the watercourse from sufficient distance upstream to a sufficient distance downstream of the gauge station. Where rating review reaches have been modelled separately from the main AFA model, the main AFA model will be calibrated to the results of rating review to ensure consistency.

3. Spot gauged flows are replicated within the model using design flow hydrographs and model parameters adjusted within realistic limits in order to achieve the corresponding recorded water levels at the gauge station location.

4. When calibration is achieved flows are increased from zero to above the highest design flow (>0.1% AEP event) and the corresponding modelled water levels at the gauge location are recorded.

5. The stage (water level minus gauge station staff zero level) versus discharge results are plotted to determine the modelled stage discharge (Q-h) relationship.

6. The existing Q-h relationship is reviewed in light of the modelled relationship and the existing reliable limit of the Q-h relationship is extended up to the limit of the modelled flows. In some cases where the existing Q-h relationship has been extrapolated beyond the highest gauged flow (for practical reasons) the modelled Q-h relationship may vary significantly and as such the reliability of the existing gauged flood flows is called into question.

Three hydrometric stations have been specified for this analysis within UoM 36 and are shown in Table 3.1.

IBE0700Rp0009 18 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

3.2 RATING REVIEW RESULTS

The current rating quality classification assigned under the FSU for each station (if available) and whether the rating review indicated that there is significant uncertainty in the existing rating, defined as a difference in

Qmed of more than 10%, is stated in Table 3.1.

Table 3.1: Existing Rating Quality Classification for Rating Review Stations in UoM 36

Station Final Station Rating Quality Significant Uncertainty Station Name Number Classification Identified in current rating

36027 Bellaheady No current rating (previously A2) Yes

36031 Lisdarn A2 No

36150 Shantonagh Bridge Not rated under FSU No

A1 sites – Confirmed ratings good for flood flows well above Qmed with the highest gauged flow greater

than 1.3 x Qmed and/or with a good confidence of extrapolation up to 2 times Qmed, bank full or, using suitable survey data, including flows across the flood plain.

A2 sites – ratings confirmed to measure Qmed and up to around 1.3 times the flow above Qmed. Would have at least one gauging to confirm and have a good confidence in the extrapolation.

B sites – Flows can be determined up to Qmed with confidence. Some high flow gaugings must be

around the Qmed value. Suitable for flows up to Qmed. These were sites where the flows and the rating

was well defined up to Qmed i.e. the highest gauged flow was at least equal to or very close to Qmed,

say at least 0.95 Qmed and no significant change in channel geometry was known to occur at or about the corresponding stage.

C sites – possible for extrapolation up to Qmed. These are sites where there was a well defined rating up

to say at least 0.8 x Qmed. Not useable for the FSU.

U sites – sites where the data is totally unusable for determining high flows. These are sites that did not possess 10 years of data or more, had water level only records or sites where it is not possible to record flows and develop stage discharge relationships. Not useable for FSU.

As well as the uncertainty in the existing ratings some gauging station ratings are limited such that they do not cover the range of flood flows other than through extrapolation of the stage discharge relationship. As a result of this all of the AMAX series level data has been re-processed into AMAX flow data using the revised rating derived from the rating review models and the revised AMAX series flow data presented in Table 3.2 below. Full details of the individual rating reviews can be found in Appendix C.

IBE0700Rp0009 19 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Table 3.2: AMAX Series Data Before and After Rating Review

36027 36031 36150 Bellaheady Lisdarn Shantonagh Bridge

Exist (m3/s) RR (m3/s) Exist (m3/s) RR (m3/s) Exist (m3/s) RR (m3/s)

1974 27.07 1975 23.56 5.89 1976 25.55 5.91

1977 23.56 5.25 1978 21.63 6.47 1979 24.22 6.50

1980 26.56 7.02 1981 24.38 6.32 1982 26.39 7.57 1983 26.90 6.33 1984 23.40 6.38 1985 23.07 7.28 1986 26.56 6.41 1987 25.21 13.70 1988 17.46 7.24 6.94

1989 26.39 6.39 6.20 1990 28.11 6.38 6.20

1991 29.86 6.69 6.47 1992 31.43 7.34 7.16 1993 32.23 5.80 5.69 1994 32.23 5.85 5.73 1995 30.44 7.00 6.74 1996 37.40 6.43 6.24 1997 38.87 5.06 5.02 1998 30.05 5.88 5.76 1999 15.35 7.59 7.64

2000 34.05 6.02 5.88 14.39 13.0 2001 33.03 6.96 6.70 2002 37.40 8.87 9.08 5.09 5.0 2003 31.83 6.67 6.45 10.9 10.23 2004 42.96 8.32 8.47 20.09 18.05 2005 36.34 5.64 6.04 5.94 2006 31.47 7.59 10.32 9.80 2007 50.91 10.44 16.77 15.16 2008 31.43 8.15 9.76 9.36 2009 32.23 12.06 19.74 17.75 2010 5.18 8.41 8.15

2011 8.17 47.30 52.68 2012 5.11 2013 8.69

Qmed 25.38 33.54 6.47 10.9 10.23 FSU 24.38 6.45 - - % Diff. 32.2 0.3 6.2

Denotes data taken forward for use in FSU. Rating considered to have confidence up to at least

1.3 x Qmed (both A1 stations) IBE0700Rp0009 20 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

At the Bellaheady gauging station (36027) it has not been possible to review the existing rating for which a classification of A2 was given. This is because the rating classification applies to the period before the canal was refurbished in the early 1990s. Since then the OPW have abandoned the rating and only one spot gauging is available to calibrate a modelled rating. This spot gauging was taken during a period of flooding along the Erne system in October 2009. Nevertheless this study has sought to develop a rating for the period post canal refurbishment such that flow information can be derived from the water level data that has been collected since 1992. Under this new modelled rating the median flood flow (Qmed) has been found to have increased significantly (32%) since canal refurbishment (Table 3.2). However there is a high degree of uncertainty surrounding this flow due to the reliance on one spot gauging for calibration and the potential for a backwater effect from downstream tributaries to affect the rating on this canalised, flat reach. Although the canal has been restored, leading to the OPW abandoning the flow rating of this gauging station, it is not considered that this would have such a significant effect on catchment run-off as to lead in an increase in

Qmed by up to one third above the previous value for which there was a high degree of certainty at Qmed. As such it is recommended that pre canal refurbishment value is retained for design flow estimation purposes.

At Lisdarn (36031) the current EPA rating has been applied since 14/12/87 when there was a shift in the staff gauge zero. Therefore the rating review AMAX values begin at 1988. The rating review exhibited good agreement between the modelled and existing EPA developed rating curve. At Qmed there was found to be

0.3% difference and as such the observed Qmed value can be taken forward for design flow estimation with confidence.

Similarly the rating review at the Shantonagh Bridge gauging station (36150) exhibited good agreement between the modelled and existing EPA developed rating curve. At Qmed there was found to be less than

10% difference and as such the observed Qmed value can be taken forward for design flow estimation with confidence.

IBE0700Rp0009 21 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

3.3 IMPACT OF RATING REVIEWS ON HYDROLOGICAL ANALYSIS

As discussed in Chapter 2, Methodology Review much of the hydrological analysis was undertaken prior to survey information at the relevant gauging stations being available such that the rating reviews can be carried out. As such it is necessary to quantify the potential impact on the hydrological analysis and identify where re-analysis or mitigation to minimise the potential impact is required. The various elements of the hydrological analysis and design flow estimation are listed below and a summary of the potential impact and the proposed mitigation measures is detailed (Table 3.3).

Table 3.3: Summary of Rating Review Effects and Mitigation

Hydrological Potential Effects of Uncertainty in the Potential Mitigation Analysis Rating Impact

Most uncertainty with poor rating likely at flood flows and as such there could be Re-assess Q for FSU Gauged Q uncertainty in AMAX series. Will affect Medium med med classified sites of C or U Qmed at sites with a classification lower than B. An issue where an ungauged catchment is adjusted based on a pivotal site with Ungauged high uncertainty. As Pivotal Sites are Low None required Q med taken from A1, A2 & B classification they are unlikely to be affected.

Flood frequency is a function of the Frequency re-analysis not ranking of events within the AMAX series, required. Historic flood the position in the ranking is unlikely to be frequency affected by adjusting all the values of the Medium Where event flows are analysis series (i.e. unless just adjusting a specific used for calibration historic gauge period) but the flood flow figure flows must be re- must be revised if used for calibration. calculated The inclusion of gauge years within pooled flood frequency analysis that have a high degree of uncertainty could skew At gauges where there has the pooled frequency analysis but the been shown to be effect will be diluted within a group (where uncertainty, re-assess Growth curve Medium / it is assumed other gauge years have a single site analysis to development Low high degree of confidence). The check that it is within 95th cumulative effect of uncertainty in both percentile confidence limits directions at multiple gauges may also of the pooled analysis. have a cancelling out effect within a pooling group. Calibration of hydraulic models is undertaken at extreme flood flows where Hydraulic highest degree of uncertainty could be Reassess calibration event model High present. Model calibration therefore flows where necessary calibration dependent on upper limits of gauge rating.

IBE0700Rp0009 22 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Hydrological Potential Effects of Uncertainty in the Potential Mitigation Analysis Rating Impact

Hydrograph Uncertainty would affect values but semi- Shape dimensionless shape will not change (Q is Low None required Generation expressed factorially from 0 to 1).

IBE0700Rp0009 23 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

4 INDEX FLOOD FLOW ESTIMATION

The first component in producing design flows within the majority of best practice methods widely used in the UK and Ireland is to derive the Index Flood Flow which within the FSU guidance is defined as the median value of the annual maximum flood flow series or Qmed. The methodologies being used in this study are detailed in the UoM 36 Inception Report and are reviewed in Chapter 2 of this report. As discussed the methods combine best practice statistical methods. This chapter details the Index Flood Flow estimation at each of the HEPs within UoM 36 on a model by model basis, including a discussion on the confidence and comparison of the outputs from the considered methodologies. There are five models included in UoM 36 and these are shown in Figure 4.1.

Figure 4.1: UoM 36 Watercourses to be Modelled

IBE0700Rp0009 24 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

4.1 MODEL 1 – BALLINAMORE

Ballinamore AFA (County Leitrim) is affected by the Ballinamore, Ballyconnell Canal (now part of the Shannon Erne Waterway) which is fed by the Yellow River and St John’s Lough upstream. The AFA is located on the Canal itself which then progresses in a southerly direction before turning north- eastwards towards . Model 1 terminates at the downstream end of Garadice Lough at Ballincur Bridge. The total catchment area of Model 1 at this point is almost 200km2 and there is a significant proportion of forested area just upstream of the AFA (20%). Model 1 also constitutes three HPW reaches which affect the northern portion of Ballinamore AFA before joining together and entering the Canal in the town centre.

The HEPs and associated sub-catchments of the Ballinamore Model are shown in Figure 4.2.

Figure 4.2: Model 1 HEPs and Catchment Boundaries

There is one gauging station within Model 1 (36028, Aghoo) operated by OPW which is located on the Ballinamore and Ballyconnell Canal downstream of Ballinamore AFA. It has 37 years of water level and flow data (1974 – 2011) but it is not rated under FSU for use as a pivotal site. The AMAX series provided by OPW was of water level only suggesting an unreliable rating. It has therefore not been pursued for use as a pivotal site for adjustment of initial Qmed estimates using catchment descriptors

IBE0700Rp0009 25 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

(FSU WP 2.3), particularly since there is an A2 rated station located on the same river further downstream (Stn no. 36027).

Station 36027 (Bellaheady) is an FSU A2 rated station located downstream on the Woodford River 3 (which is part of the Canal) within Model 2. The associated Qmed is 24.4m /s taken forward following rating review (although this indicated some uncertainty) and due to its location downstream of Model

1, it has been used as the pivotal site for adjustment of initial Qmed estimates (i.e. Qmed pcd) at HEPs on the Canal itself. For the modelled tributaries of the Canal, Station 36021 (Kiltybartan) has been used as the pivotal site. This station is located on the Yellow River (which feeds the Canal) and can be considered to be the most representative gauged data of the tributary catchments upstream of the AFA. It is also geographically closest to the HEPs on these watercourses. It is a B rated station under 3 FSU with a Qmed of 23.37m /s.

The estimated Qmed values for the various HEPs within Model 1 are shown in Table 4.1. Note that there are instances of decreasing Qmed values despite increases in catchment area. This is due to

Physical Catchment Descriptors (PCDs) other than area dominating the estimated Qmed flow rate. In the case of the modelled tributary flowing into the AFA extents the upstream node (36_756_1_RA) has a higher Qmed than the most downstream node (36034) due to the lower reaches significantly flattening which would be expected to attenuate the flow of runoff from the upland catchment. This is also the case from nodes 36_2275_1_RA to 36_2275_3 on the main channel just downstream of the

AFA. The most downstream node on the main channel (36091_RA) has a lower Qmed than the next node upstream (36_2082_RA) due to the attenuation effect of Garadice Lough in between. It must be noted that the downstream nodes discussed here are check point flows within the model and the reduction in flows represents an attenuating effect that would generally be expected to be re-created within hydraulic models. However the exact impact of these attenuating effects is captured much more accurately and site specifically in the hydraulic model where it is modelled using physical measurements of the channel and attenuating structures as opposed to the FSU ungauged catchment descriptor estimation model (used here) which is based on more general catchment characteristics. As such the attenuation effect captured within calibrated hydraulic models would be expected to take precedence over the effect captured within these estimates.

Table 4.1: Qmed Values for Model 1

2 3 Node ID_CFRAMS AREA (km ) Qmed (m /s) Preferred Estimation Methodology

36_756_1_RA 13.32 3.87 FSU (Adjusted – 36021)

36_1625_2_RA 3.92 0.37 FSU (Adjusted – 36021)

36_1625_4_RARPS 4.50 0.44 FSU (Adjusted – 36021)

36034_RA 20.22 3.25 FSU (Adjusted – 36021)

36_2098_2_RA 133.48 16.18 FSU (Adjusted – 36027)

IBE0700Rp0009 26 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

2 3 Node ID_CFRAMS AREA (km ) Qmed (m /s) Preferred Estimation Methodology 36_2274_2_RA 138.61 16.38 FSU (Adjusted – 36027)

36_2274_3_RARPS 138.65 18.59 FSU (Adjusted – 36027)

36_2275_1_RA 159.48 19.27 FSU (Adjusted – 36027)

36_2275_2_RA 159.65 19.02 FSU (Adjusted – 36027)

36_2275_3_RA 159.82 18.69 FSU (Adjusted – 36027)

36028_RA 166.85 19.34 FSU (Adjusted – 36027)

36_2082_2_RA 179.36 20.03 FSU (Adjusted – 36027)

36091_RA 199.82 14.84 FSU (Adjusted – 36027)

Note: Flow highlighted in yellow represent total flows at that point in the model rather than input flows

IBE0700Rp0009 27 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

4.2 MODEL 2 – BALLYCONNELL

Ballyconnell AFA (County Cavan) is located on the Woodford River which forms part of Ballinamore Ballyconnell Canal East (Shannon – Erne Waterway). Model 2 comprises the Woodford River from the downstream limit of Model 1 and continues downstream through Ballyconnell and on to Upper Lough Erne. This river forms the international boundary between Ireland and Northern Ireland downstream of Ballyconnell until it reaches Upper Lough Erne, which in itself is a cross-border lake. The Rag River is a significant tributary of the Woodford River with its confluence approximately 2.5km upstream from Upper Lough Erne. Ballyconnell AFA is also fluvially affected by four small relatively steep watercourses which are sourced in the forested foothills of and join the Woodford River (Canal) from the west and north- west within the AFA itself. These watercourses are also included within Model 2. The total catchment area of Model 2, at the point where the Woodford River enters upper Lough Erne is 455km2.

The HEPs and associated sub-catchments of the Ballyconnell model are shown in Figure 4.3.

Figure 4.3: Model 2 HEPs and Catchment Boundaries

Station 36027 (Bellaheady) is an FSU A2 rated station located on the Woodford River upstream of

Ballyconnell AFA (refer to Section 4.1). It has been used as the pivotal site for adjustment of initial Qmed

IBE0700Rp0009 28 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL estimates at HEPs on the Woodford River itself although the rating review indicated that there is some uncertainty associated with the Qmed value at this gauging station (see Chapter 3). For the smaller modelled tributaries a review of pivotal site options indicated no clear trend of upwards or downwards adjustment of initial Qmed pcd estimates using catchment descriptors. In other words the seven geographically closest and seven most hydrologically similar pivotal sites produced Qmed results with a high degree of scatter above and below the initial estimate. As such it was considered inappropriate to adjust the initial Qmed estimates based on catchment descriptors. The resulting estimated Qmed values for the various HEPs within Model 2 are shown in Table 4.2.

Table 4.2: Qmed Values for Model 2

AREA 3 Preferred Estimation Node ID_CFRAMS 2 Qmed (m /s) (km ) Methodology

36091_RA 199.82 14.84 FSU (Adjusted - 36027)

36_2534_3_RARPS 101.92 21.80 FSU (Adjusted - 36027)

36_1511_4_RA 3.87 0.67 FSU (Unadjusted)

36_527_8_RA 5.88 1.38 FSU (Unadjusted)

36027_RA 330.22 24.38 Observed (Gauging Station)

36_656_5_RA 8.65 5.75 FSU (Unadjusted)

36_1576_1_RARPS 0.91 0.84 FSU (Unadjusted)

36_2379_1_RARPS 1.13 0.85 FSU (Unadjusted)

36_2379_2_RARPS 2.74 1.97 FSU (Unadjusted)

36_1415_4_RARPS 4.50 3.10 FSU (Unadjusted)

36_1285_3_RARPS 13.53 7.38 FSU (Adjusted - 36027)

36_1834_1_RA 1.07 0.92 FSU (Unadjusted)

36_1834_3_RA 2.52 1.65 FSU (Unadjusted)

36_2589_2_RA 355.46 26.56 FSU (Adjusted - 36027)

36_10003_RARPS 369.89 27.56 FSU (Adjusted - 36027)

36_10002_1_RA 394.66 29.29 FSU (Adjusted - 36027)

36_2565_2_RA 55.16 2.32 FSU (Adjusted - 36027)

36_10002_D_RA 454.56 33.44 FSU (Adjusted - 36027) Note: Flow highlighted in yellow represent total flows at that point in the model rather than input

IBE0700Rp0009 29 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

It is evident from Table 4.2 that peak flow rates do not always correlate well with area or the sum of the inputs upstream. This is due to catchment descriptors other than area dominating the flow rate. Within Model 2 the FARL PCD, representing the effect of flood attenuation due to reservoirs and lakes, varies widely and generally has the effect of reducing the peak flow rate on the main channel (attenuation). Catchments where 2 there is much less attenuation due to reservoirs and lakes are likely to have a much higher Qmed / km . One such tributary catchment is the River represented by the HEP 36_2534_3_RARPS entering the upstream reach of Model 2. This catchment contains much less attenuation than the main channel catchment and as a result has a much higher Qmed, despite a smaller area. Furthermore this catchment will have a much quicker response time (time to hydrograph peak) than the main channel and as such the peak in the main channel and this tributary are not likely to coincide. This also explains why the sum of the Qmed values, which represent hydrograph peak values, entering the upstream reaches of the model does not equate to the downstream main channel Qmed value.

IBE0700Rp0009 30 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

4.3 MODEL 3 - BALLYBAY

Ballybay AFA (County Monaghan) is located on upper reaches of the Dromore River which is a significant tributary of Upper Lough Erne. Model 3 includes the Dromore River reach from Ballybay to Hydrometric Station 36018 at Ashfield approximately 19km downstream. The Dromore river flows through several lakes along this modelled reach including White Lough, Closeagh Lough, and Drumore Lough. Fluvial flood risk also emanates from four small watercourses which drain the drumlin landscape surrounding Ballybay before joining the Dromore River within the village itself, including the Shantonagh River which flows into Ballybay from the north-west. These watercourses are also included in Model 3. The total catchment area of Model 3 at its downstream limit is 220km2.

The HEPs and associated sub-catchments of the Ballybay model are shown in Figure 4.4.

IBE0700Rp0009 31 F03 NW-NB CFRAM Study UoM 36 Hydrology Report –FINAL

Figure 4.4: Model 3 HEPs and Catchment Boundaries

IBE0700Rp0009 32 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Hydrometric Stations 36030 and 36070 are located on the Dromore River and White Lough respectively but record water level only.

Hydrometric Station 36150 is operated by Monaghan County Council and is located on the Shantonagh River approximately 860m upstream of the upstream limit HEP 36_30_4_RA and represents a catchment area of 36.2km2. It was not rated under FSU and was subject to a CFRAM Study rating review in accordance with the project brief (refer to Chapter 3). The highest gauged flow at the station is 4.77m3/s with 3 a gauged Qmed value of 10.9m /s (based on 11 years of AMAX data). Therefore, confidence in observed data is limited to values no higher than 0.43 x Qmed.

The results of the CFRAM Study rating review as outlined in Chapter 3, indicates a difference in Qmed of 6% when comparing the gauging station AMAX series with the CFRAM Study rating review series. This is less than 10% difference and is not considered significant enough to warrant changing the gauge Qmed value. The station did record a water level for the significant flood event of 25th October 2011 and the flow value based on the original rating was 47.3m3/s. The CFRAM Study rating review flow for the same event was 52.6m3/s which provides some confidence in the performance of the station’s rating for flood flows.

A rainfall runoff model (NAM) was also constructed to simulate the hydrological behaviour of the catchment to the gauging station (36150) and to bring further confidence to the Qmed value. As discussed in Section 1.2.2, the hourly rainfall data from the nearby Met Eireann rain gauge station at Clones was considered of high enough resolution spatially and temporally to be used as input data for the NAM model. The model was constructed and calibrated to lower to medium flows where corresponding rainfall data was available between 2000 and 2008. Satisfactory calibration was achieved in terms of mass-balance and flow. The 3 resulting Qmed value was 10.74m /s based on a much longer simulated AMAX series from 1951 to 2008.

This further increases confidence in the Qmed value and brings it into play for use as a pivotal site within Model 3. It is a relatively small catchment in terms of area, and was considered appropriate for use as a pivotal site for all HEPs located on the tributaries of the Dromore River.

Station 36018 (Ashfield) is located on the Dromore River at the downstream limit of Model 3 and is operated by OPW. It is an FSU A1 rated station with high confidence in flow values up to at least 1.3 x Qmed which is 16.25m3/s. This station was used as a pivotal site for all HEPs located on the Dromore River.

The resulting estimated Qmed values for the various HEPs within Model 3 are shown in Table 4.3. The decrease in flows moving downstream on the Dromore River between HEPs 36030_RA and 36070_RA is due to the attenuation effect of White Lough. It must be noted that the reduction in flows represents an attenuating effect that would generally be expected to be re-created within hydraulic models. However the exact impact of these attenuating effects is captured much more accurately and site specifically in the hydraulic model where it is modelled using physical measurements of the channel and attenuating structures as opposed to the FSU ungauged catchment descriptor estimation model (used here) which is based on more general catchment characteristics. As such the attenuation effect captured within calibrated hydraulic models would be expected to take precedence over the effect captured within these estimates.

IBE0700Rp0009 33 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Table 4.3: Qmed Values for Model 3

AREA 3 Preferred Estimation Node ID_CFRAMS 2 Qmed (m /s) (km ) Methodology

36_767_6_RA 47.57 9.63 FSU (Adjusted – 36150)

36024_RA 49.05 9.77 FSU (Adjusted – 36150)

36074_RA 50.90 10.28 FSU (Adjusted – 36150)

36_10001_U_RARPS 0.04 0.01 FSU (Adjusted – 36150)

36_710_Trb_RARPS 0.36 0.11 FSU (Adjusted – 36150)

Simulated / Observed (NAM) at 36.17 10.74 36150 gauging station

36_30_4_RA 37.54 10.99 FSU (Adjusted – 36150)

36_10000_U_RARPS 0.93 0.42 FSU (Adjusted – 36150)

36_10000_RA 2.19 0.96 FSU (Adjusted – 36150)

36_30_8_RA 40.65 11.39 FSU (Adjusted – 36150) & 36151_RA

36_1691_3_RA 11.29 3.76 FSU (Adjusted – 36150)

36_1691_8_RA 14.39 4.40 FSU (Adjusted – 36150)

36_2116_4_RA 9.39 2.27 FSU (Adjusted – 36150)

36030_RA 119.25 16.79 FSU (Adjusted – 36018)

36070_RARPS 126.51 13.78 FSU (Adjusted – 36018)

36_734_2_RA 9.24 1.59 FSU (Adjusted – 36018)

36153_RA 137.72 14.91 FSU (Adjusted – 36018)

36_1762_2_RA 27.10 3.66 FSU (Adjusted – 36018)

36_2308_4_RA 4.09 0.21 FSU (Adjusted – 36018)

36072_RA 191.19 14.16 FSU (Adjusted – 36018)

36_624_6_RA 24.55 3.33 FSU (Adjusted – 36018)

36018_RA 220.40 16.25 FSU (Adjusted – 36018)

Note: Flow highlighted in yellow represent total flows at that point in the model rather than input flows

IBE0700Rp0009 34 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

4.4 MODEL 4 – CAVAN

Cavan town is located at the upstream end of the Catchment which terminates at Upper Lough Erne. The AFA is directly affected by the Cavan River which is a tributary of the Annalee River. Fluvial flood risk also emanates from a dense network of smaller unnamed watercourses that flow through the AFA draining the surrounding lands before joining the Cavan River within the AFA extent. These are also included within Model 4 and are named according to relevant townland as follows: Aghnaskerry, Annageliff, Cullies, Curragho, Derrychamp, , Drumcrauve, Drumherrish, Gartnasillagh, Moynehall, Reask, and Sweelan. The Keadew watercourse marginally affects the northern part of Cavan AFA at its upstream end, but is a direct tributary of the Annalee River, not the Cavan.

The Cavan River joins the Annalee River approximately 5km downstream from the AFA. The Annalee River is also part of Model 4 as far upstream as its confluence with the Dromore River. Model 4 then extends upstream along the Dromore River until the downstream limit of Model 3 (refer to Section 4.3). The Annalee is a tributary of the River Erne. The River Erne begins at Lough Oughter and flows northwards towards Upper Lough Erne. The Annalee joins the River Erne approximately 3 km downstream from its confluence with the Cavan River. Model 3 extends along the Annalee to where it joins the River Erne and continues downstream to its termination at Upper Lough Erne. At this point, the total catchment area of Model 3 is 1514km2 making it the largest model in UoM 36. The overall catchment is very much characterised by its lakes with one of the largest being Lough Oughter located to the west of Cavan town at the upstream end of the River Erne. In terms of the AFA itself, there are seven smaller lakes located either on the modelled tributaries of the Cavan River or just upstream of the modelled extent. These include Lough Sweelan, Green, Killymooney, Drumgola, Beaghy, and Shantemon. Figure 4.5 overleaf indicates the Model extents, HEPs and catchment boundaries.

IBE0700Rp0009 35 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 4.5: Model 4 HEPs and Catchment Boundaries

IBE0700Rp0009 36 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Model 4 contains seven hydrometric stations that have water level and flow available. These are listed in Table 2.1 (Section 2.3.1.1). Five of these stations are rated under FSU.

Station 36018 (Ashfield) is located on the Dromore River at the upstream limit of Model 4 (refer to Section 4.3 for details). It has not been used as a pivotal site within Model 4. It was considered for use in adjusting the initial Qmed estimation for HEP 36_1102_5_RA which represents the unmodelled portion of the Annalee River where it joins Model 3 (refer to Figure 4.5). However a review of 14 pivotal site options for this HEP revealed no clear trend for upwards or downwards adjustment. Indeed the geographically closest station,

36016 located approximately 2km downstream of the HEP yields a Qmed result above the 68%ile confidence upper limit. Since the remaining pivotal site options do not robustly support such significant upwards adjustment it was decided to retain the initial Qmed estimate based on catchment descriptors in this instance. Station 36016 will be used as a gauged check point during the hydraulic modelling phase of this Study. Its purpose will be to ensure the observed flows are being simulated by the model at this location and the inflows from upstream HEPs including 36_1102_5_RA will be revisited if necessary to achieve this.

Station 36016 (Rathkenny) is located on the Annalee River upstream of Cavan AFA. It is an FSU B rated 3 station with Qmed of 50.7m /s based on 14 years of AMAX data. For reasons discussed above, it has not been used as a pivotal site within Model 4 but the observed data will serve as a gauged check at this location during the hydraulic modelling phase.

Station 36031 (Lisdarn) is an A2 rated station located on the Cavan River downstream of Cavan town. Qmed is 6.45m3/s based on 29 years of AMAX data. It is also a CFRAM Rating Review station. It has been used as a pivotal site in the adjustment of initial Qmed estimates of HEPs representing the Cavan River within the AFA extent.

Station 36010 (Butlers Bridge) is located on the Annalee River just upstream of its confluence with the 3 Cavan River. It is an A1 rated station with a Qmed of 66.8m /s based on 50 years of AMAX data. It has been used as a pivotal site in the adjustment of HEPs representing the Annalee River / River Erne including HEP 36037_RA located downstream which is a gauging station without reliable flow data. It will also serve as a gauging station check point during the hydraulic modelling phase as previously discussed.

Station 36037 (Urney Bridge) is located on the River Erne downstream of its confluence with the Cavan and Annalee Rivers. It was not rated under FSU and an AMAX series has not been provided. It has not been used as a pivotal site.

Station 36019 (Belturbet) is located on the River Erne approximately 9km upstream from Upper Lough Erne. 3 It is an A1 rated station under FSU with a Qmed of 89.95m /s based on 47 years of AMAX data. It has been used as pivotal site in the adjustment of HEPs representing the River Erne and will also serve as a gauging station check point during the hydraulic modelling phase as previously discussed.

Station 36036 (Kilconny) is located on the River Erne just downstream of Station 36019. It was not rated under FSU and an AMAX series has not been provided. It has not been used as a pivotal site.

IBE0700Rp0009 37 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

The resulting estimated Qmed values for the various HEPs within Model 3 are shown in Table 4.4. As was the case for the aforementioned HEP tributary 36_1102_5_RA, there are several ungauged HEPs within Model

4 that have not had their initial FSU Qmed pcd adjusted using a pivotal site. The reason for this is that a review of all pivotal site options (seven geographically close, and seven hydrologically similar) revealed no clear trend for upwards or downwards adjustment. Whilst use of geographically close stations are generally preferable it was not considered applicable in these cases to significantly increase Qmed estimates of smaller tributary catchments based on larger main channel catchments since the pivotal site review did not support it. That said the pivotal sites located within Model 4 will serve as gauged data check points during the hydraulic modelling phase. As previously discussed, their purpose will be to ensure the observed flows are being simulated by the model at these locations and the inflows from upstream HEPs will be revisited if necessary to achieve this.

Table 4.4: Qmed Values for Model 4

2 Qmed Node ID_CFRAMS AREA (km ) 3 Preferred Estimation Methodology (m /s)

36018_RA 220.40 16.25 Observed (Gauging Station) 36_1102_5_RA 278.53 19.98 FSU (Unadjusted)

36016_RA 508.18 50.70 Observed (Gauging Station) 36_228_2_RA 102.06 16.24 FSU (Unadjusted) 36_596_8_RA 80.21 10.36 FSU (Unadjusted) 36_2398_2_RA 11.18 2.69 FSU (Unadjusted)

36_78_1_RARPS 0.08 0.05 FSU (Unadjusted) 36_789_2_RA 7.02 2.66 FSU (Unadjusted) 36_422_4_RA 9.60 3.37 FSU (Unadjusted)

36_1522_11_RA 10.66 3.49 FSU (Unadjusted) 36_674_7_RA 6.87 1.55 FSU (Unadjusted) 36_1328_3_RA 11.51 1.23 FSU (Unadjusted)

36010_RA 776.55 66.80 Observed (Gauging Station) 36_706_1_RA 14.66 3.03 FSU (Unadjusted) 36_769_U_RARPS 0.59 0.21 FSU (Unadjusted) 36_769_1_RARPS 1.01 0.34 FSU (Unadjusted) 36_760_U_RARPS 0.79 0.33 FSU (Unadjusted) 36_1676_2_RA 1.85 0.52 FSU (Unadjusted)

36_1652_1_RA 2.93 0.85 FSU (Unadjusted) 36_1652_3_RA 5.68 1.77 FSU (Unadjusted)

IBE0700Rp0009 38 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

2 Node ID_CFRAMS AREA (km ) Qmed Preferred Estimation Methodology m3/s) 36_68_1_RARPS 0.46 0.16 FSU (Unadjusted) 36_68_2_RARPS 1.40 0.46 FSU (Unadjusted) 36_1678_2_RPS 1.44 0.21 FSU (Unadjusted) 36_1678_3_RA 3.53 0.28 FSU (Unadjusted) 36_1984_1_RA 3.14 0.55 FSU (Unadjusted) 36_2232_U_RPS 0.10 0.04 FSU (Unadjusted) 36_1611_U_RARPS 0.004 0.01 FSU (Unadjusted) 36_1611_1_RARPS 0.48 0.27 FSU (Unadjusted) 36_1114_2_RA 1.74 0.23 FSU (Unadjusted) 36_254_6_RA 25.51 4.98 FSU (Unadjusted) 36_1922_U_RARPS 0.01 0.01 FSU (Unadjusted) 36_1922_9_RARPS 4.02 1.99 FSU (Unadjusted) 36_1921_Inter_1_RARPS 59.90 6.34 FSU (Adjusted – 36031) 36_1921_Inter_2_RA 59.92 6.34 FSU (Adjusted – 36031) 36031_RA 60.71 6.45 Observed (Gauging Station) 36_1113_3_RA 3.59 1.55 FSU (Adjusted – 36031) 36_743_1_RARPS 0.24 0.10 FSU (Unadjusted) 36_973_1_RARPS 0.01 0.01 FSU (Unadjusted) 36_973_2_RARPS 1.06 0.27 FSU (Unadjusted) 36_892_1_U_RARPS 0.10 0.01 FSU (Unadjusted) 36_882_6_RARPS 4.14 1.03 FSU (Unadjusted) 36138_RA 82.28 6.52 FSU (Adjusted – 36031) 36_189_3_RA 82.32 6.52 FSU (Adjusted – 36031)

36023_RA 866.32 74.36 FSU (Adjusted – 36010) 36037_RA 1457.11 93.22 FSU (Adjusted – 36019) 36_1017_2_RA 14.05 0.47 FSU (Unadjusted) 36_1951_3_RA 1482.02 92.22 FSU (Adjusted – 36019) 36019_RA 1486.38 89.95 FSU (Adjusted – 36019) 36036_RA 1486.70 89.99 FSU (Adjusted – 36019) 36_1904_9_RA 14.20 1.04 FSU (Unadjusted) 36_2367_3_RA 1511.50 90.20 FSU (Adjusted – 36019) 36_2286_D_RARPS 1514.15 90.24 FSU (Adjusted – 36019) Note: Flow highlighted in yellow represent total flows at that point in the model rather than input flows

IBE0700Rp0009 39 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

4.5 MODEL 5 – BUNDORAN AND TULLAGHAN

Bundoran town (County Donegal) and Tullaghan village (County Leitrim) are two coastal urban centres represented by Model 5. Fluvial flood risk emanates from several small coastal watercourses that flow through the AFA extent before reaching Donegal Bay, all of which are included in Model 5. These are divided into Model 5a, 5b and 5c.

4.5.1 Model 5a

Bundoran is directly affected by two watercourses. The smaller watercourse flows through Drumacrin townland and through the north of Bundoran town before meeting Donegal Bay at Bundoran Strand. It has a total catchment area of 1.9km2 and constitutes Model 5a which is ungauged.

4.5.2 Model 5b

The larger watercourse in Bundoran is the Bradoge River which has a catchment area of 36km2 where it meets Donegal Bay within Bundoran town. Model 5b is ungauged.

4.5.3 Model 5c

The largest of the Models is the Drowes River and tributaries which have fluvial flood risk potential for Tullaghan. It flows from Lough Melvin which is a large lake (21km2 in area) within County Leitrim. The total catchment area of the Drowes River where it meets Donegal Bay is 261km2.

The HEPs and associated sub-catchments of the Bundoran and Tullaghan model are shown in Figure 4.6.

There are two hydrometric stations located on the Drowes River upstream of the Model extent with flow data available. Station 35029 (Mullanaleck Bridge) is not rated under FSU. Station 35071 (Lareen) has an FSU 3 classification of A2 with a Qmed of 26.29m /s based on 31 years of AMAX data. The latter station has been used as a pivotal site in the adjustment of initial Qmed estimates of HEPs representing the Drowes River within Model 5c.

The resulting estimated Qmed values for the various HEPs within Model 5 are shown in Table 4.5. There are several ungauged HEPs within Model 5 that have not had their initial FSU Qmed estimates adjusted using a pivotal site. As previously discussed, the predominant reason for this is that a review of all pivotal site options (seven geographically close, and seven hydrologically similar) revealed no clear trend for upwards or downwards adjustment. Furthermore the use of the geographically closest station (35071) was also ruled out on the basis that its flow is heavily attenuated (Lough Melvin) and is therefore unsuitable for data transfer to HEPs not on the River Drowes with similar FARL conditions. It should also be noted that many of the FSU catchment descriptors for model 5 which represent cross border catchments emanating in Northern Ireland did not consider significant portions of the catchment which are located across the border. In particular the majority of the Drowes River / Lough Melvin catchment is contained within Northern Ireland yet the contributing area of the catchment contained within Northern Ireland has not been considered within the

IBE0700Rp0009 40 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL calculation of the FSU catchment descriptors. As such catchment descriptors required for the Study catchments have either been re-calculated (such as FARL) or borrowed from adjacent catchments where appropriate.

The pivotal sites located within Model 5c will serve as gauged data check points during the hydraulic modelling phase. As previously discussed, their purpose will be to ensure the observed flows are being simulated by Model 5c at these locations and the inflows from upstream HEPs will be revisited if necessary to achieve this.

Table 4.5: Qmed Values for Model 5

AREA Preferred Estimation Node ID_CFRAMS Q m3/s) (km2) med ( Methodology

Model 5a

35_4230_U_RARPS 0.55 0.22 FSU (Unadjusted)

35_4230_3_RA 1.89 1.02 FSU (Unadjusted)

Model 5b

35_4241_U_RA 25.42 7.10 FSU (Unadjusted)

35_4239_1_RA 26.97 7.62 FSU (Unadjusted)

Model 5c

35_195_1_RA 248.12 26.29 FSU (Adjusted – 35071)

35013_RA 250.09 26.57 FSU (Adjusted – 35071)

35_2665_U_RARPS 0.05 0.02 FSU (Unadjusted)

35_2665_6_RARPS 1.98 0.80 FSU (Unadjusted)

35_4067_3_RARPS 257.24 28.29 FSU (Adjusted – 35071)

35_1000_1_RPS 0.03 0.02 FSU (Unadjusted)

35_2282_U_RA 1.77 0.74 FSU (Unadjusted)

35_2327_4_RA 3.71 1.77 FSU (Unadjusted)

35_1056_2_RARPS 261.00 29.11 FSU (Adjusted – 35071) Note: Flow highlighted in yellow represent total flows at that point in the model rather than input flows

IBE0700Rp0009 41 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 4.6: Model 5 HEPs and Catchment Boundaries

IBE0700Rp0009 42 F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

4.6 INDEX FLOOD FLOW CONFIDENCE LIMITS

All five models in UoM 36 have hydrometric stations located within them. There are eight stations which are FSU rated with data deemed to be of a high enough quality such as to be taken forward as pivotal sites within FSU (confidence in the rating at Qmed). Four of these stations are rated A1 and a further two are rated A2 which means there is high certainty in flood flows above Qmed. Therefore UoM 36 can be considered as relatively well gauged in comparison with other Units of Management within the North Western and Neagh Bann Study areas such as UoM01 in the North-West (Co. Donegal) which is sparsely gauged.

The rating reviews in UoM 36 involve two A2 stations and one non-FSU rated station which has high quality data. The review of the Lisdarn FSU station (36031) and the non-FSU station (36150) did not indicate significant impact on the gauged Qmed value. The rating review undertaken at the Bellaheady station (36027) found that the A2 rating was no longer applicable since the canalisation of the Woodford River. However the rating review had very little post canalisation data against which to calibrate and the derived Qmed value could not be validated. In summary the rating review Qmed value, although highly uncertain, would indicate low confidence in the Qmed values along this reach.

The FSU method for Flood Estimation in Ungauged Catchments (WP 2.3) is the preferred methodology for the estimation of the index flood flow in ungauged catchments. In the first instance the index flood flow has been estimated using this method at all HEPs. The estimates are then adjusted where possible based on observed flow data with confidence at the index flood flow (Qmed). Data is applied from sites, in order of preference, on the modelled watercourse, just upstream or downstream of the modelled extents or from remote sites which have a gauging station representing a catchment that is deemed to be hydrologically or geographically similar to the subject site. There are three models where some HEPs display no clear trend for upwards or downwards adjustment based on pivotal site review, and it was therefore considered appropriate not to adjust initial Qmed pcd based on catchment descriptors. Such HEPs are generally small in area and not comparable to the geographically or otherwise hydrologically similar pivotal site options.

All design flows will be reviewed again during the hydraulic analysis phase in line with Figure 2.1 and index flows revisited if model calibration outputs deem it necessary.

IBE0700Rp0009 43 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

5 FLOOD FREQUENCY ANALYSIS AND GROWTH CURVE DEVELOPMENT

5.1 OBJECTIVE AND SCOPE

This chapter deals with the estimation of flood growth curves for the UoM 36 Unit of Management (Hydrometric Areas – HA35 and HA36) of the North Western – Neagh Bann CFRAM study areas. The estimated growth curves will be used in determining the peak design flood flows for all HEPs located on the modelled tributary and main river channels within the UoM 36 study area.

The scope of this chapter includes:

(i) Selection of a statistical distribution suitable for regional flood frequency analysis, (ii) Selection of pooling region and groups, and (iii) Growth curve estimation.

5.2 METHODOLOGY

5.2.1 Selection of Statistical Distribution

The suitable distributions for the Annual Maximum (AMAX) series for all hydrometric gauging sites located within UoM 36 were determined based on the statistical distribution fitting technique described in the Flood Studies Update (FSU) Programme Work Package 2.2 “Frequency Analysis” (OPW, 2009), UK Flood Estimation Handbook (FEH) (Institute of Hydrology, 1999) and 1975 Flood Studies Report (NERC, 1975).

5.2.2 Forming a Pooling Region and Groups

The pooling group associated with each of the growth curves was formed based on the Region-of- Influence (ROI) approach (Burn, 1990) recommended in FSU (2009). The region from which the AMAX series were pooled to form a pooling group for each of the growth curves was selected based on the similarity in catchment characteristics (both in terms of climatic and physiographic) in the neighbouring geographical region.

5.2.3 Growth Curve Development

Growth curves for each of the HEP locations were developed / estimated in accordance with the methodologies set out in the FSU, FSR and FEH studies. The Hosking and Wallis (1997) proposed L-Moment theories were used in estimating the parameters of the statistical distributions. The growth curve estimation process was automated through development of a FORTRAN 90 language based computational program.

IBE0700Rp0009 44 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

5.2.4 Limitations in the FEH and FSU Studies

There is no explicit guidance provided in FEH or FSU for dealing with the issues surrounding the production of a large number of growth factors within a river system and the associated problems with consistency and transition from growth curve to growth curve. For UoM 36, a catchment characteristic based generalised growth curve estimation method, as discussed later in Sections 5.7.4 and 5.8, was used to deal with this real world problem.

5.3 DATA AND STATISTICAL PROPERTIES

5.3.1 Flood Data

The AMAX series for all hydrometric gauging sites located within UoM 36 (HA36 & HA35) and also in the neighbouring river catchments (HA01, HA03, HA07 & HA26) were obtained from the OPW, EPA and DARD Rivers Agency (Northern Ireland). In addition, the AMAX data series used in FSU research (216 AMAX series for the entire country) were also obtained from OPW to form a pooling region for growth curve analysis. Figure 5.1 illustrates the spatial distribution of the gauging sites for which AMAX records were collected (a total of 248 gauging sites including Northern Ireland). The record lengths for these gauging stations vary from 7 to 70 years (up to year 2011) with a total of 7,750 station-years of AMAX values. The UoM 36 study area has 613 station-years of AMAX values from 19 hydrometric gauging sites.

There are climatic differences between the UoM 36 study area and other parts of the country. Restricting the choice of pooling stations to this study region, should ensure an additional degree of homogeneity. However, given the small number of AMAX values (613 station-years) available in the study area, the pooling region outside this area could prove to be useful. Three alternative extended parent pooling regions, as outlined in Table 5.1, have therefore been considered for estimating the growth curves for UoM 36. The parent pooling region which provides the highest growth curve, i.e. the most conservative flood estimate, will be adopted as the basis for the design floods. This has been explained further in Section 5.8.3.

Table 5.1: Alternative parent pooling regions

Pooling Records No. of Region Hydrometric Areas (Station- Stations No. years) 1 HA03, HA07, HA26 & HA36 54 2035

HA01, HA03, HA07, HA26, HA35 & 2 74 2786 HA36

All stations (inc. Rivers Agency NI 3 248 7750 HA01, HA02 & HA03)

IBE0700Rp0009 45 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Rivers Agency NI

Figure 5.1: Locations of 248 Gauging Stations

Table 5.2 presents the locations details, record lengths and some of the catchment characteristics of the hydrometric stations located in Pooling Region 1. In this region 2035 station-years of AMAX values are available from 54 gauging sites. The record lengths range from 9 to 70 years. Stations with FSU classification of B or above are also identified. 28 stations have A1 & A2 rating quality classification (refer to Chapter 3 for the definition of the rating quality classifications of the hydrometric gauges).

IBE0700Rp0009 46 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Table 5.2: Hydrometric Station Summary for Pooling Region 1 (54 sites)

Record Gauge Area SAAR Stations Waterbody Location Length BFI FARL Rating (Km2) (mm) (Years) Classification 03010 BLACKWATER MAYDOWN BRIDGE 43 964.93 1008.00 0.395 0.976 FEH 03017 UPPER BANN DYNES BRIDGE 15 315.94 1023.00 0.449 0.974 FEH DERRYMEEN 03022 BLACKWATER BRIDGE 18 183.49 1143.00 0.460 0.977 FEH 03024 CUSHER GAMBLES BRIDGE 29 170.94 995.00 0.365 0.992 FEH 03033 UPPER BANN BANNFIELD 38 101.64 1261.00 0.471 0.951 FEH 03043 OONA SHANMOY 26 88.59 1003.00 0.400 0.974 FEH 03051 BLACKWATER FAULKLAND 29 143.20 1083.30 0.472 0.953 A2 07001 TREMBLESTOWN TREMBLESTOWN 50 151.31 913.24 0.700 0.996 A2 07002 DEEL [Raharney] KILLYON 51 284.97 920.53 0.780 0.929 A2 BLACKWATER A1 & B 07003 (ENFIELD) CASTLERICKARD 51 181.51 809.22 0.649 1.000 BLACKWATER A2 07004 (KELLS) STRAMATT 53 245.74 1007.88 0.619 0.772 07005 BOYNE TRIM 52 1332.17 879.71 0.721 0.983 A1 07006 FYANSTOWN 49 177.45 936.67 0.552 0.990 A2 07007 BOYNE BOYNE AQUEDUCT 50 441.18 870.98 0.663 1.000 A1 & B 07009 BOYNE NAVAN WEIR 34 1658.19 868.55 0.713 0.911 A1 BLACKWATER A1 & A2 07010 (KELLS) LISCARTAN 51 699.75 948.29 0.658 0.798 BLACKWATER A2 & B 07011 (KELLS) O'DALY'S BR. 49 281.74 1003.32 0.678 0.965 07012 BOYNE SLANE CASTLE 70 2460.27 890.06 0.678 0.893 A1 07017 MOYNALTY ROSEHILL 11 70.64 991.74 0.516 0.993 ‐ 07023 ATHBOY ATHBOY 9 100.10 950.81 0.717 0.995 ‐ BLACKWATER ‐ 07033 (KELLS) VIRGINIA HATCHERY 30 124.94 1032.22 0.439 0.893 26001 SHIVEN BALLINAMORE 18 240.30 1050.00 0.530 n/a ‐ 26002 SUCK ROOKWOOD 58 641.45 1067.03 0.610 0.980 A2 26005 SUCK DERRYCAHILL 56 1085.00 1054.40 0.560 0.980 A2 26006 SUCK WILLSBROOK 58 184.76 1120.64 0.540 0.970 A1 26007 SUCK BELLAGILL 58 1207.22 1045.62 0.650 0.980 A1 26008 RINN JOHNSTON'S BR. 55 280.31 1035.47 0.610 0.860 A1 BLACK [South A2 26009 Leitrim] BELLANTRA BR. 40 98.22 1018.79 0.538 0.936 26012 BOYLE TINACARRA 53 519.92 1142.97 0.686 0.823 A1

IBE0700Rp0009 47 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Record Gauge Area SAAR Stations Waterbody Location Length BFI FARL Rating (Km2) (mm) (Years) Classification 26014 LUNG BANADA BRIDGE 31 215.14 1198.70 0.634 0.944 B 26015 ESLIN CORRASCOFFY 38 59.50 1030.00 0.600 n/a ‐ 26019 CAMLIN MULLAGH 56 252.96 979.62 0.540 0.990 A1 26021 INNY BALLYMAHON 35 1098.78 945.25 0.830 0.810 A2 26022 FALLAN KILMORE 38 61.88 915.82 0.580 1.000 A2 26108 BOYLE BOYLE ABBEY BR. 20 527.32 1142.66 0.733 0.825 B BALLINDARRAGH ‐ 36005 COLEBROOKE BRIDGE 34 313.59 1156.00 0.421 0.987 DRUMRAINEY ‐ 36007 SILLEES BRIDGE 32 166.30 1332.00 0.495 0.888 36010 ANNALEE BUTLERS BR. 55 771.73 967.55 0.632 0.845 A1 36011 ERNE BELLAHILLAN 54 336.30 979.63 0.759 0.801 A2 36018 DROMORE ASHFIELD 55 234.40 950.12 0.650 0.850 A1 36019 ERNE BELTURBET 52 1491.76 971.21 0.786 0.753 A2 36021 YELLOW KILTYBARDAN 32 23.41 1569.64 0.330 1.000 A2 36027 WOODFORD BELLAHEADY 36 333.81 1373.34 0.660 0.720 ‐ ‐ 36028 WOODFORD AGHOO 18 170.20 1334.99 0.637 0.814 36031 CAVAN LISDARN 34 63.77 910.43 0.497 0.958 A2 36037 ERNE URNEY BRIDGE 13 870.10 961.64 0.623 0.641 ‐ 36044 ANNALEE LISNACLEA 26 160.50 996.97 0.704 0.686 ‐ 36071 L. SCUR GOWLY 21 68.03 1314.65 0.644 0.823 B 36072 DROMORE L. NEW BR. 9 204.80 950.72 0.694 0.777 ‐ 36073 ST. JOHN'S LAKE WOOD ISLAND 16 100.10 1283.67 0.665 0.716 ‐ 36076 L. SILLAN 30 52.90 1006.67 0.697 0.718 ‐ DERRYGOONEY ‐ 36078 LAKE DERRYGOONEY 31 74.80 980.38 0.703 0.678 36079 L. BAWN CORLEA 32 65.40 981.77 0.703 0.697 ‐ 36080 L. GOWNA CLOONE 33 36.10 995.48 0.838 0.603 ‐

The various calculations steps associated with the growth curve estimations for UoM 36 have been presented only for Pooling Region 1. The results for all pooling regions are compared in Section 5.8.3.

IBE0700Rp0009 48 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

5.3.2 Pooling Region Catchment Physiographic and Climatic Characteristic Data

In addition to the AMAX series, some catchment physiographic and climatic characteristics information including the catchment sizes (AREA), Standard Average Annual Rainfall (SAAR), catchment Base Flow Index (BFI) and the Flood Attenuation by Reservoirs and Lakes (FARL) Index for all 248 stations were also obtained from OPW.

Table 5.3 presents a summary of the catchment characteristics for all gauging sites in Pooling Region 1 (contains 54 no. stations). Catchment sizes range from 23 km2 to 2460 km2 with a median value of 210 km2. SAAR values range from 809mm to 1570 mm with a median value of 1003 mm. BFI values vary from 0.330 to 0.838.

Table 5.3: Summary of Catchment physiographic and climatic characteristics of Pooling Region (54 sites)

Characteristics Minimum Maximum Mean Median

AREA (km2) 23.41 2460.27 410.92 209.97

SAAR (mm) 809.22 1569.64 1042.65 1003.16

BFI 0.330 0.838 0.609 0.636

FARL 0.603 1.000 0.884 0.940

The relative frequencies of the AREA, SAAR and BFI values within the 54 stations are also presented in Figures 5.2, 5.3 and 5.4 respectively.

It can be seen from Figure 5.2 that the majority of the catchment areas in the selected sites fall in the range 50 to 350 km2. Figure 5.3 shows that the SAAR values in the majority of the stations range from 800 to 1200 mm and similarly, Figure 5.4 shows the relative frequency of the BFI values within the 54 catchments. It can be seen from this figure that the BFI values in the majority of the 54 catchment areas range from 0.35 to 0.75.

IBE0700Rp0009 49 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

AREA

0.20 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 Relative frequency Relative 0.02 0.00 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500 2600 2700

Area (km2)

Figure 5.2: Relative frequencies of catchments sizes (AREA) within the Pooling Region 1 (54 stations)

SAAR

0.45

0.40

0.35

0.30

0.25

0.20

0.15 Relative frequency Relative 0.10

0.05

0.00 700 900 1100 1300 1500 1700 SAAR (mm)

Figure 5.3: Relative frequencies of the SAAR values within the Pooling Region 1 (54 stations)

IBE0700Rp0009 50 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

BFI

0.250

0.200

0.150

0.100 Relative frequency Relative

0.050

0.000 0.30 0.40 0.50 0.60 0.70 0.80 0.90 BFI

Figure 5.4: Relative frequencies of the BFI values within the Pooling Region 1 (54 stations)

5.3.3 Statistical Properties of the AMAX series

Table 5.4 provides a summary of the statistical properties of the AMAX series for all 54 gauging sites 3 in Pooling Region 1. The median annual maximum flows (Qmed) range from 3.83 to 191.40 m /s with an average value of 41.30 m3/s. The L-CV values range from 0.063 to 0.317 with an average value of 0.158, while the L-Skewness values range from -0.083 to 0.429 with an average value of 0.146. This is less than the theoretical L-Skewness of EV1 distribution.

Table 5.4: Statistical properties of 54 AMAX Series in Pooling Region 1

Parameters Minimum Maximum Average Median

Record Lengths (years) 9 70 38 36

3 Mean Flow (m /s) 4.00 212.13 43.15 25.85

3 Median Flow (m /s) 3.83 191.40 41.30 23.73

L-CV 0.063 0.317 0.158 0.147

L-skewness -0.083 0.429 0.140 0.137

L-Kurtosis 0.015 0.377 0.158 0.150

Figure 5.5 shows the L-CV versus L-Skewness diagram for the 54 AMAX series in the Pooling Region 1.

IBE0700Rp0009 51 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

L-CV versus L-Skewness 0.5

0.4 HA03_stations 0.3 HA07_stations 0.2 HA26_stations 0.1 HA36_stations Skewness ‐ L 0

‐0.1

‐0.2

‐0.3 0 0.1 0.2 0.3 0.4 L‐CV

Figure 5.5: L-Moment Ratio Diagram (L-CV versus L-Skewness) for 54 AMAX series in Pooling Region 1

From Figure 5.5 it can be seen that there is a fair degree of variance within the L-Moment ratios across the different hydrometric areas within Pooling Region 1. The variability and skewness of the pooled data for any particular site will generally be reflective of the proportion of data taken from any particular hydrometric area. For example for sites where data is heavily pooled from within HA36 itself the L-moment ratio L-CV will tend towards the lower end indicating a lower variability within the pooled data which will then be reflected in the growth factor estimates.

5.4 STATISTICAL DISTRIBUTION

The individual gauging site’s AMAX series were fitted to four flood like distributions, namely EV1, GEV, GLO and LN2 distributions. The EV1 and LN2 distributions are two-parameter distributions while the GLO and GEV distributions each have three-parameters.

The choice of distributions used for this study was guided by the findings in the FSU Report (September, 2009). In the case of 2-parameter distributions, the FSU Work Package 2.2 report states (Section 4.2, page 40) “It can be deduced from the linear patterns that Irish flood data are more likely to be distributed as EV1 or LN2 rather than Logistic distribution (LO) among 2-parameter distributions”. Therefore the elimination of LO as a 2-parameter distribution is robustly based on a study of all relevant Irish data. Also, FSU concentrated on GEV and GLO from among the available 3- parameter distributions. The lack of emphasis on LN3 by FSU was possibly based on the L-Kurtosis vs. L-skewness moment ratio diagram (FSU WP 2.2 Report, Figure 3.10, page 30) and that one could

IBE0700Rp0009 52 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL be used as a surrogate for the other. Then, because of the overwhelmingly central role traditionally played by GEV in flood frequency analysis, the FSU decided to base its analysis using the GEV rather than LN3. The same reasoning was adopted for the present study.

Based on the visual inspections of the probability plots of all 54 AMAX series in Pooling Region 1, it was found that the three-parameter distributions provide better fits to the majority of the 54 AMAX series. Between the GEV and GLO distributions, the GLO distribution was found to be the most appropriate distribution for design purposes when considered against individual sites (see Section 5.7.4). For the GLO distribution, 41 out of 54 frequency curves showed concave upward shape, 4 concave downward and 9 straight lines. For the GEV distribution, 16 showed concave upward shape, 29 showed concave downward and 9 are of straight line type.

Table 5.5 presents the summary results of the visual assessments of the probability plots for all 54 AMAX series in Pooling Region 1. It should be noted here that one reason for the change of concavity (upward and downward) shapes seen in GEV and GLO is due to the difference in abscissa used in the probability plots i.e. EV1y = -ln{-ln(1-1/T)} for GEV distribution and GLOy = -ln{1/(T-1)} for GLO distribution.

Table 5.5: Summary results of probability plots assessments (EV1, LN2, GEV & GLO distributions) for all 54 AMAX series in Pooling Region 1

No. distribution in each quality ranks (1, 2 & 3) Distri butio Fitted line type Rank 1 n Rank 2 Rank 3 (very (good) (fair) good)

10 27 17 All straight line EV1 

 All concave upward (At Log LN2 15 25 14 n scale)  9 – straight line (GEV type I)  16 – concave upward (GEV Type II) GEV 17 25 12  29 – concave downward (GEV Type III)

 9 – straight line, GLO 24 23 7  41 – concave upward &  4 – concave downward

IBE0700Rp0009 53 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

5.5 GROWTH CURVE ESTIMATION POINTS

In order to estimate the peak design flows for each of the 118 HEPs located on the modelled watercourses in UoM 36, using the ‘index-flood’ method (FEH, 1999; FSU, 2009), growth curves for each of the HEPs are required. The selection of the HEPs was based on the hydraulic model conceptualisation of the modelled watercourses within each of the AFAs in UoM 36. For the integration of hydrological input to the hydraulic model and also for the calibration and verification of the hydraulic models the HEPs were identified at the following locations on the modelled watercourses:

- HEPs at the upstream limit of model, - HEPs where tributaries enter the modelled channels, - HEPs at gauged stations on modelled channels, - HEPs at intermediate points on the modelled channels, and - HEPs at downstream limit of model.

The details of the selection process for the HEPs are discussed in the UoM 36 Inception Report (section 5.3). Table 5.6 presents a summary of the catchment characteristics associated with the 118 HEPs in UoM 36. The catchment areas vary from close to zero (at the top of modelled tributaries) to 1514 km2. The SAAR values range from 895 to 1550mm while the BFI values vary from 0.354 to 0.786.

Table 5.6: Summary of the catchment characteristics associated with the 118 HEPs

Catchment Minimum Maximum Average Median descriptors

2 AREA (km ) 0.004 1514.15 152.47 14.13

SAAR (mm) 895 1550 1077 965

BFI 0.354 0.786 0.556 0.535

Based on the similarity of the catchment characteristics of these HEPs with the selected gauging sites located within Pooling Region 1, growth curves for all HEPs with areas greater than 5 km2 were estimated. HEPs with areas less than 5 km2 generally represent the small upstream limits of the tributaries and their catchments clustered within and around the AFA extents. Almost 95% of the selected gauging sites in the pooling region have catchment areas more than 5 km2. Therefore, the pooling groups for the HEPs with catchment areas less than 5 km2 would not be the homogeneous groups and so the errors in the estimated growth curves would be larger. Based on these considerations, 74 HEPs (out of 118) were initially selected as points for the estimation of growth curves within UoM 36. However as discussed in Section 5.8.2, this was extended to 340 with the

IBE0700Rp0009 54 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL addition of a further 222 Growth Curve Estimation Points (GC_EPs) in order to aid rationalisation of the growth factors. Figure 5.6 shows the spatial distribution of these HEPs on the modelled watercourses in UoM 36.

Figure 5.6: Spatial distribution of the HEPs and GC_EPs on modelled watercourses in UoM 36 Note: GC No. 56 (River Dromore at is used as an example in Section 5.7.3

5.6 POOLING REGION AND GROUP FOR GROWTH CURVE ESTIMATION

5.6.1 Pooling Region

Based on the similarity of climatic characteristics, it has been initially decided that the AMAX series from the river catchments located within the UoM 36 study area and also from the neighbouring hydrometric areas HA03 (Bann), HA07 (Boyne) & HA26 (Shannon-upper) will be pooled to form a pooling group for growth curve estimation. However, given the comparatively small number of AMAX values (2,035 station-years) available in these HAs, the pooling region outside this region could prove to be useful. Based on this, two additional alternative extended parent pooling regions have also been considered for estimating the growth curves for UoM 36. The details of all three pooling regions were

IBE0700Rp0009 55 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL outlined in Table 5.1. The parent pooling region which provides the highest growth curve, i.e. the most conservative flood estimate, will be adopted as the basis for the design floods. This has been explained further in Section 5.8.3.

The values of AREA, SAAR and BFI encountered in the 118 HEPs are summarised by their minimum, maximum, average and median values in Table 5.6. Comparison of these with the selected stations’ corresponding catchment characteristics show a good overlap, which indicates that the selected stations provide good coverage for the range of catchments encountered in the HEPs in UoM 36.

5.6.2 Pooling Group

Pooling groups can be formed on the basis of geographical proximity to the subject site. However in the UK FEH study (1999) it was found that such pooling groups were less homogeneous than those formed by the Region of Influence (ROI) approach of the type proposed by Burn (1990). The ROI approach selects stations, which are nearest to the subject site in catchment descriptor space, to form the pooling group for that subject site. In the FSU studies a distance measure in terms of three catchment descriptors of AREA, SAAR and BFI was used in forming a pooling group. The recommended distance measure in the FSU studies is:

2 2 2  ln AREA  ln AREA   lnSAAR  lnSAAR    BFIBFI   i j   i j   i j  dij  7.1       2.0     ln AREA    ln SAAR    BFI  (5.1)

Where i is the subject site and j=1,2,….M are the donor sites.

In this study, the pooling group was formed based on the above distance measure. The size of the pooling groups was determined based on the FEH recommended 5T rules (i.e. the total number of station-years of data to be included when estimating the T-year flood should be at least 5T). The donor sites associated with this pooling group size are selected based on the lowest distance measures among the available gauging sites in the pooling region.

5.7 GROWTH CURVE ESTIMATION

5.7.1 Choice of Growth Curve Distributions

In the ‘index-flood’ method one of the major assumptions is that the frequency distributions at different sites in the pooled group are identical apart from a scale factor, which is the median flow (Qmed).

As discussed in Section 5.4, the three-parameter GEV and GLO distributions were found to be the better suited distribution for most of the 54 AMAX series than the two-parameter distributions. Furthermore, it can be seen from the L-moment ratio diagram for these 54 AMAX series as shown in Figure 5.7, that the GEV distribution is providing the line of best fit to L-moment ratios of AMAX series

IBE0700Rp0009 56 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL from pooling region no. 1, since the theoretical values of the GEV distribution’s L-Skewness and L- Kurtosis pass centrally through the observed L-moments ratios of the 54 AMAX series.

L-moment ratio diagram

0.5

Observed 0.3 GEV kurtosis ‐ GLO L

0.1 EV1

‐0.1 ‐0.3 ‐0.2 ‐0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 L‐skewness

Figure 5.7: L-moment ratio diagram (L-skewness versus L-Kurtosis)

Based on the above, the GEV distribution can be adopted as the best candidate distribution for the regional growth curve for UoM 36. However, since the probability plots show that the GLO distribution is also suitable, this distribution is also considered as a candidate distribution for the regional growth curve estimation. Although the two-parameter distributions exhibit more bias in the regional flood frequency estimates as compared to the three-parameter distributions, the two-parameter EV1 distribution is also used in the growth curve estimation process for comparison purposes and to replace the GEV or GLO growth curve when the shape displayed by either of these two distributions is concave downward in order to avoid potential underestimation of extreme event growth factors.

5.7.2 Estimation of Growth Curves

The algebraic equations of the EV1, GEV and GLO growth curves and associated parameters are given below:

EV1 distribution:

Growth Curve: xT   /11lnln2lnln1 T  (5.2)

t Parameter:   2 (5.3) t2   2lnln2ln

IBE0700Rp0009 57 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

where, t2 is the L-coefficient of variation (L-CV) and  is Euler’s constant = 0.5772.

GEV distribution:

k   k  T   Growth Curve: x    ln2ln1  , k  0 (5.4) T    k   T 1 

The parameters k and  are estimated from sample t2=L-CV and sample t3=L-skewness as follows: [Hosking & Wallis (1997, p.196)]

2 2 2ln  9554.28590.7 cck where c   (5.5) 3 t3 3ln

kt3   (5.6) k k 2 kt k  2112ln1

GLO distribution:

 k Growth Curve: xT T  111 , k  0 (5.7) k

The parameters k and  are estimated from sample t2=L-CV and sample t3=L-skewness as follows [Hosking & Wallis (1997, p.197)]:

2 sinkkt  tk 3 and   (5.8)  22 sin kttkk

The pooled regional values of the t2 (L-CV) and t3 (L-skewness) have been estimated as the weighted average values of corresponding at-site sample values weighted by the at-site record lengths. These values were equated to the expressions for these quantities written in terms of the distribution’s unknown parameters as given above and the resulting equations are solved for the unknown parameters.

5.7.3 Examination of Growth Curve Shape

Growth curves for all of the selected 74 HEPs for a range of AEPs were estimated in accordance with the above methodologies. An examination of the derived shapes of the growth curves showed that, because of the fixed shape distribution, the EV1 growth curves are of straight-line type for all 74 HEPs, while in the GEV and GLO distribution cases growth curves take either the concave upwards (upward bend) or concave downwards (downward bend) shapes based on the skewness of the pooled group. In the GEV distribution case, 33 out of 74 curves, showed concave downward shape,

IBE0700Rp0009 58 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

34 showed concave upward shape and 7 showed almost a straight line; while in the GLO distribution case, all 74 curves showed the concave upward shape (see Table 5.6).

Distribution Growth Curve Shape

EV1 All straight lines

33 - concave downward GEV 34 – concave upward 7 – straight line

GLO All concave upward

Table 5.7: Growth curves shape summary

An assessment of the suitability of the three growth curve distributions was undertaken by examining the suitability of these distributions in fitting the AMAX series in the pooling groups associated with all 74 HEPs. In other words, for a particular HEP, the pooled growth curves, based on EV1, GEV and GLO, were superimposed on the standardised probability plots of the AMAX series which form the pooling group (typically 10 to 12 such series). A visual comparison of the suitability of the growth curves for each of the 80 HEPs selected for growth curve analysis was made and recorded. As an example, HEP No. 56 (Dromore River at Claragh) was selected to illustrate the composition of one pooling group (refer to Figure 5.6).

In estimating the pooled growth curve for HEP No.56, 507 station-years of records from 11 sites were pooled. Figure 5.6 shows the location of this HEP. Table 5.8 shows the catchment characteristics, statistical properties and estimated distance measures for each of the sites from the subject HEP.

Table 5.8: Catchment descriptors for all pooled sites for GC EP No. 56

Record Specific Hydrometric AREA SAAR Qmean L- length 2 BFI 3 Qmean L-CV L-kur dij stations (km ) (m /s) 3 2 skew (years) (mm) (m /s/km )

36072 9 204.80 950.72 0.694 20.26 0.099 0.202 0.004 0.296 0.090

36018 55 234.40 950.12 0.650 15.48 0.066 0.131 0.110 0.146 0.175

07011 49 281.74 1003.32 0.678 26.71 0.095 0.245 0.175 0.096 0.511

07002 51 284.97 920.53 0.780 19.63 0.069 0.186 0.072 0.078 0.516

36044 26 160.50 996.97 0.704 11.89 0.074 0.085 0.036 0.191 0.532

07004 53 245.74 1007.88 0.619 19.82 0.081 0.149 0.159 0.151 0.539

07001 50 151.31 913.24 0.700 20.34 0.134 0.317 0.294 0.134 0.552

07006 49 177.45 936.67 0.552 22.22 0.125 0.179 0.113 0.055 0.598

36011 54 336.30 979.63 0.758 18.33 0.055 0.117 0.031 0.189 0.616

IBE0700Rp0009 59 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Record Specific Hydrometric AREA SAAR Qmean L- length 2 BFI 3 Qmean L-CV L-kur dij stations (km ) (m /s) 3 2 skew (years) (mm) (m /s/km )

26019 56 252.96 979.62 0.540 22.89 0.090 0.152 0.282 0.172 0.636

26008 55 280.31 1035.47 0.610 23.72 0.085 0.122 0.132 0.238 0.778

Subject site (Growth - 220.40 950.67 0.692 - - 0.171* 0.143* - - Curve EP- 56) *Pooled regional values

It can be seen from the above table that the subject site’s catchment characteristics are well placed within the pooled sites’ catchment descriptor space. The subject site has an upstream catchment area of 220.4km2, SAAR and BFI values of 950.67mm and 0.692 respectively which are located approximately at the median locations of the pooled sites’ corresponding values.

The estimated pooled average L-CV and L-Skewness are 0.171 and 0.143 respectively. This suggests that the pooled growth curve would follow a distribution which has L-Skewness slightly less than that of the EV1 distribution (0.167).

Figure 5.8 shows the estimated EV1, GEV and GLO growth curves for the GC EP No. 56. The GEV growth curve is a concave downward shaped curve, while the GLO curve is concave upward shaped.

Figure 5.8: Pooled Growth Curve EP 56- (a) EV1 and GEV distributions; (b) GLO distributions

An assessment of the at-site GEV and GLO growth curves were carried out through a visual inspection of their individual probability plots. A summary of this assessment is provided in Table 5.9.

IBE0700Rp0009 60 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Table 5.9: Frequency curve shapes of the individual site’s AMAX series associated with the Pooled Growth Curve EP 56

Individual at-site growth curves Hydrometric Comparison of performances stations GEV (EV1y Plot) GLO (Loy Plot) (visual) Moderate concave Mild concave GEV fits slightly better 36072 downward downward Mild concave Mild concave upward GLO fits slightly better 36018 downward 07011 Straight line Mild concave upward GLO fits slightly better Mild concave Both fit equally well to the Mild concave upward 07002 downward observed records Mild concave Mild concave upward GLO fits slightly better 36044 downward Mild concave Moderate concave GLO fits slightly better 07004 downward upward Moderate concave Both fit equally well to the Mild concave upward 07001 upward observed records Mild concave Both fit equally well to the Mild concave upward 07006 downward observed records Moderate concave Mild concave upward GEV fits slightly better 36011 downward Moderate concave Both fit equally well to the Mild concave upward 26019 upward observed records Mild concave Mild concave upward GLO fits better 26008 downward

The above assessment shows that both the GEV and GLO distributions fit the observed at-site records quite well at all eleven sites with a slightly better performance by the GLO distribution. In the case of GEV distribution eight sites showed concave downward shaped curves (mild to moderate), two concave upward and one site showed a straight line. While in the GLO distribution case, 10 showed concave upward and one showed a concave downward curve. This suggests that, the shape of the pooled growth curves in the case of GEV distribution can be expected as concave downward while for the GLO distribution it would be concave upward.

Table 5.10 shows the estimated growth factors for a range of AEPs for Growth Curve No. 56. The estimated 1% AEP growth factors for the EV1, GEV and GLO distributions are 2.102, 2.024 and 2.120 respectively.

IBE0700Rp0009 61 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Table 5.10: Estimated growth factors for Growth Curve No. 56

AEP (%) EV1 GEV GLO

50 1.000 1.000 1.000

20 1.295 1.292 1.264

10 1.490 1.478 1.445

5 1.678 1.651 1.631

2 1.921 1.867 1.897

1 2.102 2.024 2.120

0.5 2.283 2.175 2.364

0.1 2.703 2.510 3.032

5.7.4 Recommended Growth Curve Distribution for UoM 36

The following factors were considered to select an appropriate growth curve distribution for the UoM 36 area:

(i) Suitability of a distribution in fitting the individual at-site records, (ii) No. of distribution parameters, and (iii) Shape of the pooled growth curve

A visual examination of the at-site frequency curves for all 54 gauging sites when considered individually showed that the AMAX series for most of these sites can be described slightly better by the GLO distribution than by the EV1 and GEV distributions.

The number of distribution parameters also plays an important role in deriving an appropriate growth curve. The fixed skewness two-parameter distributions generally suffer from large biases, particularly at the upper tail of the distribution. The three-parameter distributions, in contrast, suffer from larger standard error though they are less biased. However this standard error is generally reduced by the pooled estimation process. The use of two-parameter distributions such as the Gumbel distribution is not therefore recommended in regional frequency analysis (Hosking and Wallis, 1996). The use of a two-parameter distribution is beneficial only if the investigator has complete confidence that the at-site distribution’s L-Skewness and L-Kurtosis are close to those of the frequency distributions. As discussed in Section 5.7.1, the L-CV and L-Skewness of most of the sites in the Pooling Region differ from those of the theoretical values of the EV1 distribution. This suggests that a three-parameter distribution would be more appropriate to describe the growth curves for UoM 36.

IBE0700Rp0009 62 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

The shape of the growth curve also plays an important role in the design and operation of the flood management scheme for a river catchment. It is generally not considered appropriate to have a growth curve with the concave downward shape. A significant number of the GEV growth curves showed concave downward shape (33 out 74). In contrast, all 74 GLO growth curves are of concave upward shape.

The estimated 1%-AEP GLO growth factor is slightly greater than the GEV growth factor, for almost all 74 growth curves by an amount of 0.1 to 5% (see Table 5.10 for growth curve No.56). This is largely due to the concavity noted above. Figure 5.9 shows a comparison of the GEV, GLO and EV1 growth curves for growth curve No.56, all plotted in the EV1 probability plot. It can be seen that the GLO distribution displays some concave downwards behaviour at the lower end of the plot but this represents event frequencies which are higher (more regular) than the Qmed event and as such are not relevant in a flood frequency context. The GEV growth curve at EP no. 56 displays slight concave downwards behaviour but its similarity to the 2 parameter EV1 distribution suggests that the effect of the shape parameter k is minimal. As flood frequency becomes more extreme (higher EV1 reduced variate) the three distributions can be considered to represent different projections of extremity for future flood magnitude with the GLO distribution giving a more conservative estimate of magnitude for extreme events and the GEV resulting in a less conservative estimate.

Comparison of Growth Curves

3.5

3.0 GEV EV1 2.5 GLO 2.0

1.5

Growth factors 1.0

0.5

0.0 -4-202468 EV1 reduced variate

Figure 5.9: Comparison of EV1, GEV and GLO growth curves on the EV1-y probability plot (Growth Curve EP No. 56)

Based on the above, it is recommended to adopt the GLO distribution derived concave upward shape growth curve for the subject river catchments in UoM 36.

IBE0700Rp0009 63 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

5.8 RATIONALISATION OF GROWTH CURVES

5.8.1 Relationship of Growth Factors with Catchment Characteristics

In order to reduce the number of growth curves to a practicable number, the relationship between the estimated growth factors for a range of AEPs and the relevant catchment descriptors were examined. The catchment descriptors used were the AREA, SAAR and BFI. Figures 5.10, 5.11 and 5.12 show the variations of growth factors with AREA, SAAR and BFI respectively for all 74 HEPs.

Figure 5.10: Relationship of growth factors with catchment areas for 74 HEPs

Figure 5.11: Relationship of growth factors with SAAR for 74 HEPs

IBE0700Rp0009 64 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 5.12: Relationship of growth factors with BFI for 74 HEPs

It can be seen from Figure 5.10 that the growth factors generally increase with decrease in catchment sizes. However this rate of increase is larger for the catchment areas less than 250 km2 and also for the larger AEPs growth factors. This can be attributed to the smaller upland catchment areas where catchment response time is shorter and where no flow attenuation is available. For the larger catchments flow attenuation is generally provided by lakes and wider downstream channels. For catchment areas larger than 500 km2 the growth factors do not change noticeably with the further increase in catchment area. No such particular patterns in the relationships of the growth factors with the SAAR and BFI values were found (Figures 5.11 and 5.12). It can be noted that there is a dip in the growth factors for SAAR values around 970mm evident in Figure 5.11. This is due to a cluster of HEPs on the lower reaches of the main channel of the Annalee / Erne Rivers which represent a large, highly attenuated catchment with shallow growth curve characteristics. Within the graph these HEPs are surrounded by catchments from the same geographical area (similar SAAR values) which represent smaller, less attenuated tributary catchments which tend to have steeper growth curve characteristics. This demonstrates that SAAR values may be linked to catchment centroid location but there is no consistent pattern between SAAR values and growth curve characteristics.

5.8.2 Generalised Growth Curves

Based on the findings as discussed in Section 5.8.1, growth curves for UoM 36 were further generalised based on catchment size. To examine further the relationship of the catchment size with the growth factors and also to generalise the growth factor estimates, an additional 222 growth curve estimation points with various catchment sizes were selected on the modelled watercourses. Figure 5.6 shows the spatial distribution of these points. The catchment physiographic and climatic characteristics data associated with these additional growth curve estimation points were obtained from the OPW.

IBE0700Rp0009 65 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 5.13 shows the variation of the estimated growth factors for a range of AEPs and catchment sizes for all 296 HEPs (74 HEPs plus 222 additional points). Similar catchment size-growth factor relationships were found in this case as were found in the 74 HEPs. It can be seen from this figure that the growth factors for catchment areas greater than 700 km2 do not change appreciably with the increase in catchment sizes. However, the variations in growth factors for the smaller catchment sizes are significant.

Figure 5.13: Relationship of growth factors with catchment areas (for 337 growth curve estimation points)

As a result of the above, growth curves are generalised based on ranges of catchment size as shown below:

1. AREA < 10 km2 2. 10 < AREA <= 25 km2 3. 25 < AREA < = 50 km2 4. 50 < AREA < = 100 km2 5. 100 < AREA < = 150 km2 6. 150 < AREA < = 200 km2 7. 200 < AREA < = 300 km2 8. 300 < AREA < = 400 km2 9. 400 < AREA < = 900 km2 10. AREA > 900 km2

Table 5.11 shows the estimated average and median growth factors for the above 10 categories of growth curves along with their associated group standard deviations for a range of AEPs. The number

IBE0700Rp0009 66 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL of HEPs used for the standard deviation calculation in each of the catchment size categories is presented in column 2 of Table 5.11. It can be seen from this that the standard deviations in the 1% AEP growth factors in these catchment size categories range from 0.5% to 11.1%. The highest variations were found in the smaller catchment size categories. Although the standard deviation for the 200 – 300km2 range is similar to the 150 – 200km2 range this band covers a larger range in area and as such would be expected to have a much higher standard deviation if the variation in growth factors with area was uniform. Hence, it is recommended that the growth factors for all HEPs with catchment sizes falling in the catchment area range from 10 to 200 km2 (i.e. the catchment size categories 2 to 6 as shown in Table 5.11) be estimated from the individual growth curve estimation process. In other words, separate growth curves should be estimated for all HEPs with the catchment areas falling in the range of 10 to 200 km2. All HEPs with catchment areas less than 10 km2 are considered to have the same growth curve. For the remaining categories the median growth curves will be used.

Table 5.11: Growth curve estimation summary

Growth factors No of Catchment size HEPs AEP (%) 50% 20% 10% 5% 4% 2% 1% 0.50% 0.20% 0.10% range in size range Return Period 2 5 10 20 25 50 100 200 500 1000 (years)

Average 1.000 1.258 1.444 1.642 1.710 1.938 2.194 2.486 2.934 3.328 1. AREA < 10 km2 16 Median 1.000 1.270 1.466 1.676 1.748 1.990 2.264 2.577 3.059 3.486

St. dev 0.000 0.019 0.035 0.052 0.059 0.081 0.108 0.140 0.193 0.242

Average 1.000 1.254 1.436 1.632 1.699 1.922 2.174 2.461 2.900 3.287 2. 10 < AREA <= 2 25 km 29 Median 1.000 1.261 1.449 1.651 1.720 1.951 2.212 2.508 2.964 3.366

St. dev 0.000 0.020 0.036 0.055 0.061 0.084 0.111 0.144 0.196 0.245

Average 1.000 1.263 1.452 1.656 1.726 1.960 2.225 2.526 2.991 3.401 3. 25 < AREA <= 2 50 km 20 Median 1.000 1.267 1.460 1.668 1.739 1.978 2.249 2.558 3.034 3.456

St. dev 0.000 0.011 0.019 0.029 0.033 0.045 0.059 0.077 0.105 0.131

Average 1.000 1.258 1.442 1.638 1.704 1.928 2.179 2.461 2.894 3.273 4. 50 < AREA <= 2 30 100 km Median 1.000 1.263 1.450 1.650 1.718 1.944 2.197 2.482 2.917 3.307

IBE0700Rp0009 67 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Growth factors No of Catchment size HEPs AEP (%) 50% 20% 10% 5% 4% 2% 1% 0.50% 0.20% 0.10% range in size range Return Period 2 5 10 20 25 50 100 200 500 1000 (years)

St. dev 0.000 0.007 0.013 0.019 0.021 0.029 0.038 0.049 0.068 0.086

Average 1.000 1.248 1.420 1.600 1.661 1.862 2.084 2.331 2.702 3.021 5. 100 < AREA < 2 = 150 km 33 Median 1.000 1.254 1.435 1.627 1.692 1.910 2.153 2.425 2.840 3.201

St. dev 0.000 0.015 0.029 0.047 0.053 0.077 0.107 0.145 0.208 0.269

Average 1.000 1.244 1.409 1.578 1.635 1.819 2.018 2.234 2.552 2.819 6. 150 < AREA < 2 = 200 km 31 Median 1.000 1.237 1.396 1.559 1.614 1.790 1.981 2.188 2.492 2.746

St. dev 0.000 0.013 0.023 0.034 0.037 0.050 0.064 0.080 0.106 0.128

Average 1.000 1.257 1.432 1.612 1.672 1.869 2.083 2.318 2.665 2.958 7. 200 < AREA < 2 = 300 km 21 Median 1.000 1.264 1.445 1.631 1.693 1.897 2.120 2.364 2.726 3.032

St. dev 0.000 0.013 0.023 0.034 0.038 0.051 0.066 0.083 0.109 0.132

Average 1.000 1.224 1.376 1.531 1.583 1.751 1.933 2.133 2.425 2.670 8. 300 < AREA < 2 = 400 km 27 Median 1.000 1.224 1.375 1.530 1.582 1.750 1.932 2.131 2.423 2.668

St. dev 0.000 0.002 0.003 0.004 0.004 0.005 0.007 0.008 0.011 0.013

Average 1.000 1.226 1.379 1.536 1.588 1.759 1.944 2.145 2.442 2.691 9. 400 < AREA < 2 = 900 km 54 Median 1.000 1.221 1.371 1.524 1.575 1.742 1.922 2.119 2.408 2.651

St. dev 0.000 0.012 0.021 0.031 0.034 0.045 0.057 0.071 0.093 0.111

Average 1.000 1.246 1.415 1.591 1.651 1.845 2.058 2.294 2.645 2.945 10. AREA> 900 2 km 35 Median 1.000 1.246 1.415 1.591 1.651 1.845 2.059 2.295 2.647 2.948

St. dev 0.000 0.000 0.000 0.001 0.001 0.003 0.005 0.008 0.014 0.019

IBE0700Rp0009 68 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Thus for UoM 36 the aforementioned 10 categories of catchment size have been reduced to five categories (hereafter called Growth Curve Groups) as presented in Table 5.12. The estimated growth curve types in each category are also presented in Table 5.12.

Table 5.12: Growth Curve (GC) Groups

Growth Growth curves type / Curve Catchment size range estimation process Group No.

GC01 AREA<=10km2 Use median growth curve

GC02 10 < AREA <= 200 km2 Use individual growth curve

GC03 200 < AREA < = 400 Use median growth curve km2

GC04 400 < AREA < = 900 Use median growth curve km2

GC05 AREA> 900 km2 Use median growth curve

5.8.3 Results for different parent pooling regions for UoM 36

Similar to Pooling Region 1, the generalised growth curve groups have also been estimated for the remaining parent pooling regions (Groups 2 & 3, Table 5.1). Figure 5.14 shows the comparisons of these growth curve groups for all parent pooling regions. What is most apparent is that there is very good agreement between the results using Pooling Regions 1 & 2. This would be expected given the majority of sites within Pooling Region 2 are also found within Pooling Region 1. The much larger Pooling Region 3 is affected by the full range of sites nationally with a much higher range of climatic and geographical variation and this is reflected in the deviation in the results from regions 2 & 3. It can be seen from this figure that no particular pooling region gives the largest growth curves/factors estimates for all growth curve groups. In general, Pooling Region 3 (entire country) gives the steeper growth curves for the smaller catchments growth curve groups (GC01 & GC02), while the Pooling Regions 1& 2 give the steeper growth curves for the larger catchments growth curve groups of GC03, GC04 & GC05. Pooling Region 1 (54 Sites) gives steeper curves for all larger growth curve groups of GC03, GC04 & GC05 than Pooling Region 2. The differences in the Pooling Regions 1 and 2 estimates are in the range of 0.1 to 3%, i.e. the Pooling Region 1 estimates are only 0.1 to 3% larger than that of the Pooling Region 2 estimates.

Based on the above, it is considered prudent to adopt the parent pooling region which provides the highest growth curve, i.e. the most conservative flood estimate. The design growth curves for the growth curve groups of GC03, GC04 & GC05 have been estimated from Pooling Region 1 (54 Sites), while for the remaining two growth curve groups (GC01 & GC02), the Pooling Region 3 derived estimates have been used. This is considered prudent due to a limited amount of small site data

IBE0700Rp0009 69 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL contained within Pooling Region 1. These design growth factors for a range of AEPs are presented in Table 5.13 in Section 5.8.4.

Figure 5.14: Comparison of growth curves for different parent pooling regions

IBE0700Rp0009 70 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

5.8.4 Estimated Growth Factors for UoM 36

Table 5.13 presents the estimated growth factors for a range of AEPs for each of the above growth curve groups. Figure 5.15 shows the estimated growth curves (GLO) for all growth curve groups. Note that Growth Curve Group No. 2 represents a range of individual growth curves which have not been generalised due to the variation within the group but only the median growth curve is displayed in Figure 5.15 for clarity.

Table 5.13: Growth factors for range of AEPs

GLO - Growth factors GC Catchment size Group range AEP AEP AEP AEP AEP AEP AEP AEP AEP AEP No. 50% 20% 10% 5% 4% 2% 1% 0.5% 0.2% 0.1%

1 AREA<=10km2 1.000 1.381 1.666 1.979 2.088 2.46 2.89 3.39 4.181 4.898

1.197 1.323 1.444 1.483 1.605 1.731 1.861 2.041 2.185 10 < AREA <= 200 2 1.000 to to to to to to to to to km2 1.388 1.679 2.000 2.112 2.494 2.937 3.454 4.274 5.018

200 < AREA < = 3 400 km2 1.000 1.224 1.375 1.530 1.582 1.750 1.932 2.131 2.423 2.668

400 < AREA < = 4 900 km2 1.000 1.221 1.371 1.524 1.575 1.742 1.922 2.119 2.408 2.651

5 AREA> 900 km2 1.000 1.246 1.415 1.591 1.651 1.845 2.059 2.295 2.647 2.948

6.0 AREA<=10km2 5.0 10 < AREA <= 200 km2 (Median) 200 < AREA < = 400 km2 400 < AREA < = 900 km2 4.0 AREA> 900 km2 AEP(%) 3.0

2.0

Growth factors 1.0 20% 10% 5% 4% 2% 1% 0.5% 0.2% 0.1%

0.0 0.01.02.03.04.05.06.07.0

LogisticLogistic Reduced reduced Variate vatiate

Figure 5.15: GLO growth curves for all Growth Curve Groups (5 No.)

IBE0700Rp0009 71 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

The large range observed in growth curve group 2 as shown in Table 5.13 is as a result of developing individual growth curves reflecting very different catchments within the 10 – 200 km2 area band. Within this band are both heavily attenuated catchments densely covered in lakes and waterways as well as small steep headwater catchments reflecting catchments at both extremes of growth curve behaviour.

The uncertainties associated with the above growth curve estimates are expressed in terms of 95% confidence interval of these estimates and were estimated from the following relationship:

T T  XseXileX T )(96.1)%95( (5.8)

The standard error (se) of the growth curves is estimated in accordance with the FSU recommended methodology. Table 5.14 presents the estimated standard errors in terms of percentage of the estimated growth factor for a range of AEPs. The upper and lower limits of the confidence interval were estimated using the above mentioned Eq. 5.8. For example, for the GC Group No. 4, the estimated 1% AEP growth factor is 1.922 and the associated 95% upper and lower confidence limits are 2.110 and 1.734 respectively.

Table 5.14: Estimated percentage standard errors for growth factors (XT) for a range of AEPs

(source FSU Work- Package 2.2 “Frequency Analysis” Final Report – Section 13.3)

Return Annual periods Exceedance Se (XT) % (years) probabilities (%)

2 50% 0.60 5 20% 1.00 10 10% 1.80 20 5% 2.77 25 4% 3.00 50 2% 3.90 100 1% 5.00 200 0.5% 5.94 500 0.2% 7.30 1000 0.1% 8.30

Figure 5.16 shows the estimated growth curve along with the 95% upper and lower confidence limits for GC Group No. 4.

IBE0700Rp0009 72 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

3.5

3.0 GC04:400 < AREA < = 900 km2 AEP(%) 2.5 95%ile CI

2.0

1.5

1.0 Growth factors Growth 20% 10% 5% 4% 2% 1% 0.5% 0.2% 0.1% 0.5

0.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

Logistic reduced variate

Figure 5.16: Growth Curve for GC Group No. 4 with 95% confidence limits

5.8.5 Comparison of the at-site growth curves with the pooled growth curves

The FSU programme recommended that “in the event that the at-site estimate of Q-T relation is steeper than the pooled one then consideration will have to be given to using a combination of the at- site estimate and the pooled estimate for design flow estimation”. In light of this, the at-site frequency curves (Q-T) for each of the gauging sites located on the modelled watercourses (eight gauging sites) in UoM 36 were examined and compared with the relevant pooled frequency curves. In the case where the pooled frequency curve is flatter than the at-site curve, the design growth curves/factors should be estimated from the at-site records. If the pooled growth curve is concave downward then a two parameter distribution should be fitted to the pooled growth curve so as to avoid the upper bound.

Furthermore the FSU study recommended that “If a very large flood is observed during the period of records the question arises as to whether it should over-ride any more modest estimate of QT obtained by a pooling group approach or whether a weighted combination of the pooling group estimate and the at-site estimate should be adopted. If a combination is used the weights to be given to the two components of the combination cannot be specified by any rule based on scientific evidence but must be chosen in an arbitrary, however one would hope a reasonable way.”

Table 5.15 shows the hydrometric gauges (eight gauging sites) located on the UoM 36 modelled watercourses. The final pooled growth curve group numbers associated with these gauges are also included therein.

IBE0700Rp0009 73 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Table 5.15: Hydrometric gauging stations located on the modelled watercourses in UoM 36 hydrometric area

Growth Curve Stations WATERBODY LOCATION Group No.

36010 Annalee Butlers Bridge GC04

36018 Dromore Ashfield GC03

36019 Erne Belturbet GC05

36027 Woodford Bellaheady GC03

36028 Woodford Aghoo GC02

36031 Cavan Lisdarn GC02

36037 Erne Urney Bridge GC05

36072 Dromore L. New Bridge GC02

Figure 5.17 shows the comparisons of the At-site and Regional Flood Frequency (AFF and RFF) curves for the above mentioned hydrometric gauging sites. The EV1 distribution was used for these comparisons for consistency. In addition to the frequency curves, the 95%ile confidence intervals associated with the regional estimates were also included in these plots. The EV1 straight line was used as an indicative descriptor of the at-site distribution, rather than a GEV or GLO curve, because the latter when fitted at-site, is liable to be misleading because of the large standard error involved in the shape parameter particularly. This was used for those stations where the individual AMAX series standardised growth curves were different considerably, in some cases, from the pooling growth curve. In such cases, EV1 regional growth curves were used instead of GLO curves; because the nature of the adjustment implies that an appropriate curved shape could not be determined with more accuracy than that of a straight line i.e. preserving with a curved growth curve in such cases would be an “illusion of accuracy”. It should be noted that given EV1 is a two parameter distribution it is more limited in its ability to fit data with outliers as it cannot take on a curved shape.

IBE0700Rp0009 74 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

36010_Flood Frequency Curve (EV1) 250

Observed 200 AEP At-site EV1 Regional EV1 150 95% CI /s) 3 100

Flow (mFlow 50 50% 20% 10% 4% 1% 0 -2.0 0.0 2.0 4.0 6.0 8.0 EV1 Reduced Variate

36019_Flood Frequency Curve (EV1) 36027_Flood Frequency Curve (EV1) 80 300 Series1 Observed 70 AEP 250 AEP At-site EV1 At-site EV1 60 Regional EV1 Regional EV1 95% CI

/s) 200 3 95%CI 50

150 /s)

3 40

Flow (m Flow 100 30 20 50 (m Flow 50% 20% 10% 4% 1% 10 50% 20% 10% 4% 1% 0 0 -2.0 0.0 2.0 4.0 6.0 8.0 -2.0 0.0 2.0 4.0 6.0 8.0 EV1 Reduced Variate EV1 Reduced Variate

36028_Flood Frequency Curve (EV1) 36031_Flood Frequency Curve (EV1) 60 20 18 Observed Observed AEP 50 AEP 16 At-site EV1 At-site EV1 14 Regional EV1 Regional EV1 40 12 95% CI /s) /s)

3 95% CI 3 10 30 8 6 Flow (m Flow Flow (m Flow 20 4 50% 20% 10% 4% 1%

10 50% 20% 10% 4% 1% 2 0 0 -2.0 0.0 2.0 4.0 6.0 8.0 -2.0 0.0 2.0 4.0 6.0 8.0 EV1 Reduced Variate EV1 Reduced Variate

Figure 5.17: The at-site and pooled frequency curves along with the 95% confidence intervals

IBE0700Rp0009 75 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

36037_Flood Frequency Curve (EV1) 36072_Flood Frequency Curve (EV1) 160 70 Observed Observed 140 AEP 60 AEP At-site EV1 At-site EV1 120 Regional EV1 50 Regional EV1 95% CI 95% CI 100 40 /s) /s)

3 80 3 30 60 20

Flow (mFlow 40 (mFlow

10 50% 20% 10% 4% 1% 20 50% 20% 10% 4% 1% 0 0 -2.0 0.0 2.0 4.0 6.0 8.0 -2.0 0.0 2.0 4.0 6.0 8.0 EV1 Reduced Variate EV1 Reduced Variate

Figure 5.17 (cont’d): The at-site and pooled frequency curves along with the 95% confidence intervals

It can be seen from the above frequency curves that at two sites (out of eight), the AFF curves are slightly steeper than the RFF curves, suggesting that the regional curves slightly underestimate when compared with a number of observed floods at these stations. However, these at-site growth curves fall within the 95%ile confidence limits of the estimated associated regional growth curves.

If an AFF curve lies below the confidence limits of the RFF curve, as is the case with two of the sites in UoM36 (36028 & 36037), then we consider it prudent to adopt the RFF curve as the design curve, on the basis that the observed flood record has, by chance, fallen below the regional average and that there is a chance or possibility that the record of the next 20 or 30 years will revert to resembling the RFF curve rather than reproduce a re-occurrence of the recent past. It may be the case that the flatter at-site behaviour is more representative of the subject catchment and is generally caused by the particular high degree of attenuation evident within UoM 36 whereas the pooled curves are moderated by the lumped sum of catchments within the pooling group which are unlikely to be quite as attenuated as some of the catchments found in UoM36. It has to be acknowledged that this type of decision may lead to a degree of over-design but it is recommended that this be knowingly accepted.

On the other hand if an AFF curve lies above the RFF curve, then we consider it prudent to take account of both when deciding on the design curve/flood. This could be done by calculating a weighted average of the two curves. The relative weights should be decided, on a case by case basis, following examination of the degree of difference between the two curves, including consideration of the confidence limits of the RFF curve, shape of the at-site probability plot and the number of observed large outliers in the data series.

As the at AFF curves at the two sites discussed above are only marginally steeper and are within the 95th%ile confidence limits no adjustment to the RFF curves is necessary and the design growth

IBE0700Rp0009 76 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL curves for all HEPs located in close proximity to the above stations have been estimated from their relevant regional growth curves.

5.8.6 Growth factors for all HEPs in the UoM 36

Based on the catchment sizes associated with each of the 118 HEPs, the relevant estimated growth factors for a range of AEPs are presented in Table 5.16 overleaf.

Table 5.16: Growth factors for all 118 HEPs for a range of AEPs for UoM 36

Node Growth factors (XT) No. AREA 1% AEP 0.2% AEP 0.1% AEP Node ID_CFRAMS 2 (km ) Lower Upper Lower Upper Lower Upper X X X 95%ile T 95%ile 95%ile T 95%ile 95%ile T 95%ile

1 36_767_6_RA 47.57 2.136 2.368 2.600 2.742 3.200 3.658 3.049 3.641 4.233

2 36_30_8_RA 40.65 2.608 2.891 3.174 3.622 4.227 4.832 4.168 4.978 5.788

3 36_10000_U_RARPS 0.93 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

4 36_10000_RA 2.19 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

5 36_30_4_RA 37.54 2.608 2.891 3.174 3.622 4.227 4.832 4.168 4.978 5.788

6 36_1691_3_RA 11.29 2.640 2.927 3.214 3.647 4.256 4.865 4.182 4.995 5.808

7 36_2116_4_RA 9.39 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

8 36_1691_8_RA 14.39 2.640 2.927 3.214 3.647 4.256 4.865 4.182 4.995 5.808

9 36_734_2_RA 9.24 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

10 36_1762_2_RA 27.10 2.478 2.747 3.016 3.321 3.876 4.431 3.758 4.488 5.218

11 36_2308_4_RA 4.09 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

12 36_624_6_RA 24.55 2.589 2.870 3.151 3.530 4.119 4.708 4.023 4.805 5.587

13 36_1102_5_RA 278.53 1.743 1.932 2.121 2.076 2.423 2.770 2.234 2.668 3.102

14 36_228_2_RA 102.06 1.836 2.036 2.236 2.254 2.630 3.006 2.459 2.937 3.415

15 36_596_8_RA 80.21 2.150 2.384 2.618 2.786 3.251 3.716 3.111 3.715 4.319

16 36_2398_2_RA 11.18 2.649 2.937 3.225 3.662 4.274 4.886 4.202 5.018 5.834

17 36_78_1_RARPS 0.08 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

18 36_422_4_RA 9.60 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

19 36_1522_11_RA 10.66 2.617 2.901 3.185 3.605 4.207 4.809 4.130 4.932 5.734

20 36_674_7_RA 6.87 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

21 36_1328_3_RA 11.51 2.602 2.885 3.168 3.578 4.175 4.772 4.096 4.892 5.688

IBE0700Rp0009 77 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Node Growth factors (XT) No. AREA 1% AEP 0.2% AEP 0.1% AEP Node ID_CFRAMS 2 (km ) Lower Upper Lower Upper Lower Upper X X X 95%ile T 95%ile 95%ile T 95%ile 95%ile T 95%ile

22 36_706_1_RA 14.66 2.640 2.927 3.214 3.647 4.256 4.865 4.182 4.995 5.808

23 36_769_U_RARPS 0.59 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

24 36_769_1_RARPS 1.01 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

25 36_760_U_RARPS 0.79 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

26 36_1676_2_RA 1.85 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

27 36_1652_3_RA 5.68 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

28 36_68_2_RARPS 1.40 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

29 36_68_1_RARPS 0.46 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

30 36_1678_2_RPS 1.44 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

31 36_1678_3_RA 3.53 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

32 36_2232_U_RPS 0.10 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

33 36_1611_U_RARPS 0.00 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

34 36_1611_1_RARPS 0.48 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

35 36_1984_1_RA 3.14 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

36 36_1114_2_RA 1.74 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

37 36_254_6_RA 25.51 2.597 2.879 3.161 3.558 4.152 4.746 4.065 4.855 5.645

38 36_1922_U_RARPS 0.01 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

39 36_1922_9_RARPS 4.02 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

40 36_743_1_RARPS 0.24 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

41 36_973_1_RARPS 0.01 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

42 36_973_2_RARPS 1.06 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

43 36_1113_3_RA 60.74 2.425 2.689 2.953 3.270 3.816 4.362 3.714 4.435 5.156

44 36_892_1_U_RARPS 0.10 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

45 36_882_6_RARPS 4.14 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

46 36_189_3_RA 82.32 2.175 2.411 2.647 2.826 3.298 3.770 3.161 3.775 4.389

47 36_1017_2_RA 14.05 2.440 2.705 2.970 3.249 3.791 4.333 3.665 4.377 5.089

48 36_1904_9_RA 14.20 2.610 2.894 3.178 3.584 4.182 4.780 4.099 4.895 5.691

49 36_2098_2_RA 133.48 1.627 1.804 1.981 1.866 2.178 2.490 1.973 2.356 2.739

IBE0700Rp0009 78 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Node Growth factors (XT) No. AREA 1% AEP 0.2% AEP 0.1% AEP Node ID_CFRAMS 2 (km ) Lower Upper Lower Upper Lower Upper X X X 95%ile T 95%ile 95%ile T 95%ile 95%ile T 95%ile

50 36_2274_2_RA 138.61 1.581 1.753 1.925 1.796 2.096 2.396 1.891 2.258 2.625

51 36_756_1_RA 13.32 2.049 2.272 2.495 2.648 3.090 3.532 2.958 3.533 4.108

52 36_1625_2_RA 3.92 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

53 36_1625_4_RARPS 4.50 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

54 36_2275_1_RA 159.48 1.561 1.731 1.901 1.749 2.041 2.333 1.830 2.185 2.540

55 36_2275_2_RA 159.65 1.561 1.731 1.901 1.749 2.041 2.333 1.830 2.185 2.540

56 36_2275_3_RA 159.82 1.561 1.731 1.901 1.749 2.041 2.333 1.830 2.185 2.540

57 36_2534_3_RARPS 101.92 1.858 2.060 2.262 2.294 2.677 3.060 2.510 2.998 3.486

58 36_1511_4_RA 3.87 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

59 36_527_8_RA 5.88 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

60 36_656_5_RA 8.65 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

61 36_1576_1_RARPS 0.91 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

62 36_1285_3_RARPS 13.53 2.040 2.262 2.484 2.644 3.086 3.528 2.960 3.535 4.110

63 36_1415_4_RARPS 4.50 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

64 36_1834_1_RA 1.07 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

65 36_1834_3_RA 2.52 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

66 36_2589_2_RA 355.46 1.743 1.932 2.121 2.076 2.423 2.770 2.234 2.668 3.102

67 36_2565_2_RA 55.16 1.794 1.989 2.184 2.168 2.530 2.892 2.349 2.805 3.261

68 35_195_1_RA 248.12 1.743 1.932 2.121 2.076 2.423 2.770 2.234 2.668 3.102

69 35_2665_6_RARPS 1.98 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

70 35_2665_U_RARPS 0.05 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

71 35_2282_U_RA 1.77 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

72 35_2327_4_RA 3.71 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

73 35_1056_2_RARPS 261.00 1.743 1.932 2.121 2.076 2.423 2.770 2.234 2.668 3.102

74 35_4241_U_RA 25.42 1.965 2.179 2.393 2.442 2.850 3.258 2.676 3.196 3.716

75 35_4239_1_RA 26.97 1.965 2.179 2.393 2.442 2.850 3.258 2.676 3.196 3.716

76 35_4230_U_RARPS 0.55 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

77 35_4230_3_RA 1.89 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

IBE0700Rp0009 79 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Node Growth factors (XT) No. AREA 1% AEP 0.2% AEP 0.1% AEP Node ID_CFRAMS 2 (km ) Lower Upper Lower Upper Lower Upper X X X 95%ile T 95%ile 95%ile T 95%ile 95%ile T 95%ile

78 36_2286_D_RARPS 1514.15 1.857 2.059 2.261 2.268 2.647 3.026 2.468 2.948 3.428

79 36_10002_D_RA 454.56 1.734 1.922 2.110 2.063 2.408 2.753 2.220 2.651 3.082

80 35013_RA 105.84 1.726 1.914 2.102 2.052 2.395 2.738 2.207 2.636 3.065

81 36023_RA 866.32 1.734 1.922 2.110 2.063 2.408 2.753 2.220 2.651 3.082

82 36024_RA 49.05 2.116 2.346 2.576 2.713 3.166 3.619 3.014 3.600 4.186

83 36028_RA 166.85 1.602 1.776 1.950 1.806 2.108 2.410 1.893 2.261 2.629

84 36030_RA 119.25 1.678 1.860 2.042 1.931 2.253 2.575 2.043 2.440 2.837

85 36034_RA 20.22 1.911 2.119 2.327 2.367 2.762 3.157 2.591 3.094 3.597

86 36036_RA 1486.70 1.857 2.059 2.261 2.268 2.647 3.026 2.468 2.948 3.428

87 36037_RA 1457.11 1.857 2.059 2.261 2.268 2.647 3.026 2.468 2.948 3.428

88 36070_RARPS 126.51 1.678 1.860 2.042 1.931 2.253 2.575 2.043 2.440 2.837

89 36072_RA 191.19 1.569 1.740 1.911 1.760 2.054 2.348 1.841 2.199 2.557

90 36074_RA 50.90 2.049 2.272 2.495 2.600 3.034 3.468 2.877 3.436 3.995

91 36091_RA 199.82 1.607 1.782 1.957 1.814 2.117 2.420 1.902 2.271 2.640

92 36138_RA 82.28 2.175 2.411 2.647 2.826 3.298 3.770 3.161 3.775 4.389

93 36151_RA 40.64 2.608 2.891 3.174 3.622 4.227 4.832 4.168 4.978 5.788

94 36153_RA 137.72 1.654 1.834 2.014 1.905 2.223 2.541 2.017 2.409 2.801

95 36155_RA 191.29 1.569 1.740 1.911 1.760 2.054 2.348 1.841 2.199 2.557

96 36010_RA 776.55 1.734 1.922 2.110 2.063 2.408 2.753 2.220 2.651 3.082

97 36016_RA 508.18 1.734 1.922 2.110 2.063 2.408 2.753 2.220 2.651 3.082

98 36018_RA 220.40 1.743 1.932 2.121 2.076 2.423 2.770 2.234 2.668 3.102

99 36019_RA 1486.38 1.857 2.059 2.261 2.268 2.647 3.026 2.468 2.948 3.428

100 36027_RA 330.22 1.743 1.932 2.121 2.076 2.423 2.770 2.234 2.668 3.102

101 36031_RA 60.71 2.465 2.733 3.001 3.388 3.954 4.520 3.885 4.640 5.395

102 36_2082_2_RA 179.36 1.618 1.794 1.970 1.836 2.143 2.450 1.930 2.305 2.680

103 36_10002_1_RA 394.66 1.743 1.932 2.121 2.076 2.423 2.770 2.234 2.668 3.102

104 36_10003_RARPS 369.89 1.743 1.932 2.121 2.076 2.423 2.770 2.234 2.668 3.102

105 36_789_2_RA 7.02 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

IBE0700Rp0009 80 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Node Growth factors (XT) No. AREA 1% AEP 0.2% AEP 0.1% AEP Node ID_CFRAMS 2 (km ) Lower Upper Lower Upper Lower Upper X X X 95%ile T 95%ile 95%ile T 95%ile 95%ile T 95%ile

106 36_1951_3_RA 1482.02 1.857 2.059 2.261 2.268 2.647 3.026 2.468 2.948 3.428

107 36_2367_3_RA 1511.50 1.857 2.059 2.261 2.268 2.647 3.026 2.468 2.948 3.428

108 35_4067_3_RARPS 257.24 1.743 1.932 2.121 2.076 2.423 2.770 2.234 2.668 3.102

109 35_1000_1_RPS 0.03 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

110 36_1921_Inter_1_RARPS 59.90 2.465 2.733 3.001 3.388 3.954 4.520 3.885 4.640 5.395

111 36_1921_Inter_2_RA 59.92 2.465 2.733 3.001 3.388 3.954 4.520 3.885 4.640 5.395

112 36_1652_1_RA 2.93 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

113 36_710_Trb_RARPS 0.36 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

114 36_10001_U_RARPS 0.04 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

115 36_2379_1_RARPS 1.13 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

116 36_2379_2_RARPS 2.74 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

117 36_2273_2_RARPS 0.08 2.607 2.890 3.173 3.583 4.181 4.779 4.101 4.898 5.695

118 36_2274_3_RARPS 138.65 1.581 1.753 1.925 1.796 2.096 2.396 1.891 2.258 2.625

The design flood flows for any required AEP will be calculated by multiplying the Index Flood, Qmed of

each HEP by the above estimated relevant growth factors. The Qmed at gauged sites will be estimated from the observed AMAX series supplemented with additional simulated gauge years through rainfall

runoff modelling (MIKE NAM). For the ungauged sites Qmed will be estimated from the FSU and IH 124 recommended catchment descriptors based methodologies and through the use of rainfall runoff (MIKE NAM) modelling to simulate flow records and hence produce a simulated AMAX record at the ungauged site.

It should be noted here that any uncertainties in the design flood estimates obtained from the index- flood method generally result from the uncertainties associated with both the index-flood (Qmed) and growth factor estimates. The uncertainties in the growth factor estimates can result both from the sampling variability and mis-specification of the growth curve distribution. The sampling error is considered to be small due to the larger record lengths (pooled records) used in the estimation process.

Furthermore, it should also be noted here that, any allowances for future climate change in the design flood flow estimate should be applied to the median flow estimates. Any effects of the climate change on the growth curves are expected to be minimal.

IBE0700Rp0009 81 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

IBE0700Rp0009 82 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

5.9 COMPARISON WITH FSR GROWTH FACTORS

A comparison of the estimated growth factors within UoM 36 with the FSR growth factors was undertaken as indicated by Table 5.17. All growth curves were indexed to the median annual maximum flows (Qmed).

Table 5.17: Study growth factors compared with FSR growth factors

AEP (%) 50% 20% 10% 4% 2% 1% 0.5% 0.2% 0.1% 1.197 1.323 1.483 1.605 1.731 1.861 2.041 2.185 UoM 36 1.000 to to to to to to to to 1.388 1.679 2.112 2.494 2.937 3.454 4.274 5.018 Average of UoM 36 1.000 1.320 1.553 1.891 2.183 2.517 2.899 3.495 4.028

FSR 1.000 1.260 1.450 1.630 1.870 2.060 2.250 2.620 2.750

Table 5.17 indicates that the study area growth factors (average values) are higher than the FSR growth factors. These higher values for the UoM 36 Rivers can be attributed to the Region of Influence Approach to pooling and the development of growth curves for individual catchments including a high number representing small upland and tributary catchments affecting the AFAs. The physical catchment descriptors representing these catchments lend themselves to pooling groups weighted more towards smaller catchments which tend to display steeper growth curve behaviour.

5.10 GROWTH CURVE DEVELOPMENT SUMMARY

Growth curves for all HEPs were calculated from the regional flood frequency analysis technique as recommended in the FEH, FSU and FSR studies (Region of Influence Approach).

The AMAX series for all hydrometric gauging sites located within the UoM 36 area (HA36 & HA35) and also in the neighbouring rivers catchments (HA01, HA03, HA07 & HA26) were obtained from the OPW, EPA and the Rivers Agency of Northern Ireland. In addition to these, the AMAX data series used in FSU research (216 AMAX series for the entire country except Northern Ireland) were also obtained from the OPW to form a pooling region for growth curve analysis. The selection of the pooling region for UoM 36 was based on the similarity of catchment characteristics both in terms of climatic and physiographic characteristics. Given the small number of AMAX values available in the study region HAs, two additional alternative extended parent pooling regions (including the entire country) were considered for estimating the growth curves for UoM 36. The parent pooling region which provided the highest growth curve, i.e. the most conservative flood estimate, was adopted as the basis for the design floods.

The size of a pooling group associated with each of the HEPs was determined based on the FEH recommended 5T rule (with a minimum of 500 station-years AMAX series for each pooled growth

IBE0700Rp0009 83 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL curve). The pooling process was based on the FSU recommended catchment characteristics based (AREA, SAAR and BFI) distance measures between the subject and donor sites.

The statistical distribution suitable for a pooled growth curve was determined based on a number of factors such as - the suitability of this distribution for fitting the contributory stations’ at-site AMAX series, the number of distribution parameters and shape of the growth curves (concave upward or concave downward). Four flood like distributions namely, the EV1, LN2, GEV and GLO distributions were considered. The three-parameter GLO distribution was found to be the best suited distribution in all respects and therefore was chosen as the growth curve distribution for all HEPs in UoM 36.

Initially, growth curves for each of the 118 HEPs in UoM 36 were estimated separately. Subsequently, the number of growth curves was reduced based on their relationship with the catchment areas. It was found that the growth factors generally increase with the decrease in catchment sizes. This increase rate is larger for the catchment areas less than 250 km2 and also for the larger AEP growth factors. For any catchment areas greater than 200 km2 the growth factors do not change appreciably with the increase in catchment sizes. Based on this, the following five generalised growth curve groups were recommended for the subject rivers catchments in UoM 36:

1. GC group No. 1: AREA < 10 km2

2. GC group No. 2: 10 < AREA <= 200 km2

3. GC group No. 3: 200 < AREA < = 400 km2

4. GC group No. 4: 400 < AREA < = 900 km2

5. GC group No. 5: AREA >= 900 km2

It is recommended that the growth factors for all HEPs with catchment sizes ranging from 10 to 200 km2 (Growth Curve Group No. 2) be estimated from the individual growth curve estimation process. For the remaining categories the median growth curves will be used. For all HEPs with catchment areas less than 10km2, it is recommended to use the estimated median growth factors associated with Growth Curve Group No. 1 although it must be noted that the smallest catchment from which data is pooled is 5km2.

The design growth curves for the growth curve groups GC03, GC04 & GC05 were estimated from the Pooling Region 1 (54 Sites), while for the remaining two growth curve groups (GC01 & GC02), the Pooling Region 3 derived estimates were used.

The estimated 1% AEP growth factors for UoM 36 vary from 1.731 to 2.937 depending on the catchment sizes. Growth factors for the smaller catchments are generally larger than those of the larger catchments but a high degree of variance was evident within catchments of 10 to 200km2 in size such that individual growth curves have been retained.

IBE0700Rp0009 84 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

6 DESIGN FLOWS

6.1 DESIGN FLOW HYDROGRAPHS

Following estimation of the Index Flood Flow (Qmed) and growth factors for each HEP it is possible to estimate the peak design flows for a range of Annual Exceedance Probabilities (AEPs). In addition to the total design flows estimated for each HEP, lateral inflows must be generated to represent the flow from the lateral catchment between HEPs. Catchment descriptors do not exist for the lateral inflow catchments within FSU and these have not been derived as part of this Study. The RPS methodology involves using the catchment descriptors of the total catchment at the downstream HEP with the area replaced by the difference in area between the upstream and downstream nodes / HEPs to derive an estimate of the lateral inflow Qmed based on FSU WP 2.3. In some instances where it is obvious that the catchment descriptors of the total catchment are not representative of the lateral / top-up catchment (particularly URBEXT and FARL) these have been adjusted based on orthophotography / Corine datasets. These will be reviewed as required during the hydraulic analysis stage as part of a hierarchical approach to ensuring the correct frequency conditions are achieved (i.e. the total flow in the model at each intermediate / gauging station / downstream limit HEP is correct) as we move down through the modelled catchment.

All of the design flows which will be used for hydraulic modelling input are detailed in Appendix C. The final component of estimating the fluvial design flows is to ascertain the profile of the design flow hydrograph for each HEP, i.e. the profile of the flow over time as a flood event rises from its base flow to achieve the peak design flow (rising limb) and then as the flood flow rate decreases and the watercourse returns to more normal flows (recession limb). As discussed in Chapter 2 of this report the methodology for this study has been developed further since production of the Inception Report and as such two methodologies have been used for UoM 36 to derive the design flow hydrograph shapes (widths) such that these can be applied to a range of design events:

1. FSU Hydrograph Shape generation tool (developed from FSU WP 3.1) for all HEPs representing catchments more than 10 km2

2. FSSR 16 Unit Hydrograph method for catchments less than 10 km2 where no suitable pivotal site is available

6.1.1 FSU Hydrograph Shape Generator

For all of the HEPs which represent catchments larger than 10 km2 the Hydrograph Shape Generator tool developed as an output from FSU WP 3.1 is used to derive the design hydrograph. The Hydrograph Shape Generator Tool is an Excel spreadsheet containing a library of parametric, semi- dimensionless hydrograph shapes derived from gauge records of pivotal sites using the HWA software previously discussed. Based on hydrological similarity, a pivotal site hydrograph is ‘borrowed’ and applied at the subject site (in this case the CFRAMS HEP) based on catchment

IBE0700Rp0009 85 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL descriptors. One potential issue with the use of the Hydrograph Shape Generator tool is the lack of small catchments from which suitably short hydrographs are available. This, along with overly long receding limbs on hydrographs, was particularly noticeable in earlier versions of the software but is much improved with the addition of further pivotal sites to bring the number within the library up to 145.

An example is shown in Figure 6.1 for the HEP 36_1762_2_RA which is a tributary of the Dromore River at (Model 3), representing a catchment of 27 km2. The hydrograph shape parameters have been adjusted based on the nearest hydrologically similar pivotal site, Gowly (36071) which is located in the Erne catchment.

Final Design Hydrographs 18.0

16.0

14.0 50% 12.0 20% /s

3 10% 10.0 5% 2% Flow m 8.0 1% 0.5% 6.0 0.1%

4.0

2.0

0.0 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 Timestep (Hours) Figure 6.1: Various AEP Hydrographs for Tributary HEP 36_1762_2_RA (Model 3)

6.1.2 FSSR 16 Unit Hydrograph Method

Early testing of the FSU Hydrograph Shape Generator tool found that for smaller catchments the shape that was derived appeared to be unrealistically long for some of the smaller catchments when compared to the available observed / simulated flow data for small sites. It is thought that this is as a result of a lack of pivotal sites within the library representing small catchments with only two pivotal

IBE0700Rp0009 86 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL sites representing a catchment area of less than 25 km2 included (with shape parameters identified). Based on this experience it was found that below 10km2 it was difficult to obtain a suitable pivotal site such that the duration of the hydrograph was not significantly overestimated and therefore for catchments less than 10km2 (but not limited to) an alternative but tried and tested methodology is used to derive the hydrograph. The FSSR 16 Unit Hydrograph method was used for these catchments whereby semi dimensionless hydrographs were derived with the same time-step as used for the other hydrographs within the model using the ISIS FSSR 16 UH tool. The methodology followed to derive the FSSR 16 semi dimensionless hydrograph for a subject catchment is summarised below:

1. Time to Peak of the 1 hour unit hydrograph estimated from FSU PCDs (area, MSL, S1085, SAAR & URBEXT) and adjusted for time step

2. The design storm duration is estimated as a function of SAAR and the estimated time to peak

3. An areal reduction factor is calculated as a function of design storm duration and catchment area.

4. Catchment Wetness Index is calculated as a function of SAAR.

5. A soil index is calculated using on FSR Winter Rain Acceptance Potential soil mapping

6. The Standard Percentage Runoff (SPR) is calculated as a function of the soil types within the subject catchment

7. Rainfall characteristics for the subject catchment are derived from FSU DDF gridded outputs (M5-2D & M5-25D) and FSR maps (Jenkinson’s Ratio r)

8. The outputs from steps 2 to 7 are input to the ISIS FSSR 16 boundary unit module to produce a semi dimensionless hydrograph (fitted to a peak of 1) based on Unit Hydrograph principles which can then be scaled to the various design peak flows

IBE0700Rp0009 87 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

6.2 COASTAL HYDROLOGY

Analysis of the hydrological elements which contribute to coastal flood risk has been undertaken at a national level through the Irish Coastal Protection Strategy Study (ICPSS) and the Irish Coastal Wave and Water level Study (ICWWS). This study does not seek to re-analyse these elements of coastal flood risk but rather seeks to combine them, along with the fluvial elements where applicable, such that the total combined fluvial and coastal flood risk is assessed on an AFA by AFA basis. The only AFA within UoM 36 identified as being at coastal flood risk, Bundoran & Tullaghan, experiences combined coastal / fluvial flood risk.

6.2.1 ICPSS Levels

Outputs from the Irish Coastal Protection Strategy Study have resulted in extreme tidal and storm surge water levels being made available around the Irish Coast for a range of AEPs. The locations of ICPSS nodes in relation to the Bundoran & Tullaghan AFA are shown in Figure 6.2 and the associated extreme coastal water level for the full range of coastal AEPs have been extracted from the ICPSS and are shown in Table 6.1.

Figure 6.2: Location of ICPSS Nodes in Relation to Coastal AFAs

IBE0700Rp0009 88 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Table 6.1: ICPSS Level in Close Proximity to UoM 36 AFAs / HPWs

Annual Exceedance Probability (AEP) %

2 5 10 20 50 100 200 1000

ICPSS Node AFA / HPW Highest Tidal Water Level to OD Malin (m)

NW10 2.54 2.68 2.78 2.88 3.00 3.10 3.20 3.42 Bundoran & Tullaghan NW11 2.56 2.71 2.81 2.91 3.04 3.14 3.24 3.47 (Extract from: Irish Coastal Protection Strategy Study, Phase 5 – North West Coast, Work Packages 2, 3 & 4A)

6.2.2 ICWWS Levels

The Irish Coastal Wave and Water level Study (ICWWS) is being progressed by the OPW in order to consider the potential risk associated with wave overtopping at exposed coastal locations. The study is currently ongoing but preliminary analysis has been made available for the NW-NB CFRAM Study to identify the areas within UoM 36 which have been identified as potentially vulnerable to this flood mechanism. Only a small length of vulnerable coastline at Tullaghan has been identified and this is shown in Figure 6.3.

Figure 6.3: Draft ICWWS potential areas of vulnerable coastline

IBE0700Rp0009 89 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

The study outputs will be in the form of a range of combinations of water level and wave characteristics (wave height, period, frequency and the joint probability assessed extreme water level) for each AEP.

6.2.3 Consideration of ICPSS and ICWWS Outputs

It is important to note that the outputs from both the ICPSS and the ICWWS are to be considered separately. Tidal boundaries will be applied within the 2D models at a scale and distance necessary to capture the complete effects of a dynamic tide and the propagation effects e.g. up the Drowes River. The ICPSS levels will be applied considering a range of joint probability scenarios (as detailed in 6.3.2) in order to determine the most onerous flood outline for any AEP. The levels which have been derived from the ICPSS will be applied within the 2D portion of the hydraulic (hydrodynamic) models. All ICPSS levels will be applied as the maximum level on the oscillating average tidal cycle observed at the nearest tidal gauge with the surge applied over 48 hours. A typical 1% AEP surge on tidal cycle to staff gauge zero is shown in Figure 6.4 below. Bathymetric and cross sectional survey has been undertaken within the tidal reaches of coastal models in order to accurately capture the effects of tidal propagation within the estuaries and into the tidal reaches of the watercourses where relevant. Details on the model specific application of the ICPSS levels at the coastal boundaries will be contained within the subsequent Hydraulic Modelling report.

Tidal Elevation [metres] 1% AEP Surge [metres] 1% AEP Total Water Level [metres]

Figure 6.4: Typical 1% AEP Coastal Boundary Makeup (to Staff Gauge Zero)

IBE0700Rp0009 90 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

It is important to note that the outputs from the ICWWS are not directly applicable through the standard 2D hydraulic modelling packages used for coastal flood modelling. The assessment of the volume of flood water from wave overtopping is a function of the outputs from the ICWWS (wave height, period, frequency and the joint probability assessed extreme water level), the duration of the event and the dimensions and hydraulic performance of the sea defence and foreshore. At each of the two AFAs that have been identified as vulnerable to wave overtopping, preliminary analysis will identify the location and length of sea defence / frontage which is vulnerable to wave overtopping. This section will then be assessed against the range of wave / extreme water level combinations for each AEP to determine the most onerous scenario. The total overtopping volume from the most onerous scenario for each AEP will then be assessed against the digital terrain model (LiDAR based) to ascertain the mapped flood extents, depth and hazard behind the sea defence / frontage within the AFA. Further details of the methodology for assessment and modelling of the wave overtopping flood risk will be contained within the Hydraulic Modelling report.

6.3 JOINT PROBABILITY

Joint probability is a consideration within UoM 36 in relation to the occurrence of fluvial – fluvial events (where extreme flood events on tributaries and the main channel of rivers coincide) and also at the downstream tidal reaches of the modelled watercourses where tidal – fluvial events become a consideration i.e. where the Drowes River and other minor watercourses flow into Donegal Bay at Bundoran & Tullaghan.

6.3.1 Fluvial – Fluvial

There are no major river confluence points within the AFA extents / HPW reaches of UoM 36. The confluence points of the Dromore, the Larah and Cavan Rivers with the Annalee River are all downstream of the AFA extents on medium priority watercourse. Furthermore the modelled reaches are all gauged downstream of the confluence points so the frequency conditions resulting from the confluence of these significant rivers is captured within a long term high quality gauge record. Fluvial joint probability is less of a concern on models representing smaller catchments such as those around Cavan where the critical storm in the confluencing sub-catchments is likely to be similar.

To account for fluvial joint probability in the remainder of less significant confluence points RPS has specified a high number of HEPs such that as we move down the model, i.e. past confluence points, the hydraulic modeller has to hand the design flows downstream of the confluence point such that they can check that the sum of the inflows within the tributary and the main channel are creating the correct frequency conditions downstream of the confluence point. Where these conditions are not being achieved the modeller will adjust the flows depending on the relationship between catchment descriptors of the main channel and tributary such that the joint probability relationship can be determined to create the correct frequency conditions downstream of the confluence point. This is a modelling consideration and may require an iterative approach. These adjustments will be carried out

IBE0700Rp0009 91 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL in line with the guidance provided in FSU WP 3.4 ‘Guidance for River Basin Modelling’ and detailed in the Hydraulic Modelling report.

6.3.2 Fluvial – Coastal

In terms of UoM 36, this category of joint probability is only relevant to the Bundoran & Tullaghan AFA. The RPS methodology for assessing joint probability for coastal and fluvial flooding is outlined in the CFRAM Study technical note ‘NTCG GN20 Joint Probability Guidance (RPS, June 2013)’. It advocates a stepped approach to the consideration of fluvial coastal joint probability whereby the relevance is assessed to ascertain at which sites dependence may exist and further analysis is needed:

The first stage in any Joint Probability analysis should be to ascertain whether the flooding mechanisms in any particular area, either AFA or MPW, actually warrant the consideration of the joint probability of occurrence. This screening stage should involve a review of all existing information on flooding within the area of interest, such as records of historic events or previous studies including the output from the CFRAM PFRA and the complementary ICPSS data. Where this review identifies either a significant overlap in the areas of fluvial and tidal flood risk or a proven history of significant flooding from both sources, joint probability should be considered. Where the flooding mechanism is heavily dominated by one particular source it is questionable whether joint probability analysis is justified.

An initial screening process has been undertaken on the Bundoran & Tullaghan AFA to see if further analysis is required. The results of this screening are shown in Table 6.2.

Table 6.2: Initial Screening for Relevance of Joint Probability

Evidence / Model History of Further JP AFA Name Comments No. Joint Analysis Occurrence

5 Bundoran & No Lower reaches of Drowes River tidal No Tullaghan along boundary of AFA. However even the 0.1% coastal outline does not flood land via the river. Reference to flooding during heavy rain at high tides may indicate pluvial flooding exacerbated by high tides. No real overlap in flood outlines.

Following initial screening it is not considered necessary to consider the Bundoran & Tullaghan AFA for further analysis of the joint probability relationship. This is not to say there is no evidence of a tidal

IBE0700Rp0009 92 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL influence at this location but rather that there is no known evidence of joint fluvial and coastal flood occurrence and that there are no low lying areas on the lower reaches that would be particularly sensitive to such a joint occurrence, over and above a fluvially or tidally dominant event in isolation. For this model, suitably conservative tidal downstream boundary conditions will be applied which are relatively conservative such as the highest astronomical tide, oscillating such that there is coincidence between peak tide and hydrograph. It is not thought this will lead to unrealistic downstream flood extents as there is no significant overlap of the most extreme 0.1% AEP events, when considering the PFRA and ICPSS outlines, within the AFA extents. Nevertheless this will be reviewed following initial model runs to check that this assumption is valid.

IBE0700Rp0009 93 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

7 FUTURE ENVIRONMENTAL AND CATCHMENT CHANGES

There are a number of future potential changes which may affect the outputs of this study and as such it is prudent that they are identified and their potential impact quantified such that the outputs can accommodate as much as practically possible these changes. This chapter outlines potential environmental changes such as climate change and changes to the catchment such as afforestation and changing land uses. UoM 36 represents catchments which are mostly entirely rural but it has been shown (Chapter 4) that some of them feature significant degrees of forest coverage which is known to have an effect on catchment runoff. Despite the rural nature of the catchments there are some highly urbanised catchments such as the urban watercourses flowing through Cavan and the effect of further urbanisation on the watercourses flowing through AFAs must be considered. These issues, along with potential management and policy changes are considered in this chapter.

7.1 CLIMATE CHANGE

According to the United Nations Intergovernmental Panel on Climate Change (2007) there is “unequivocal” evidence of climate change and furthermore:

"most of the observed increase in global average temperatures since the mid-20th century is very likely due to the observed increase in anthropogenic greenhouse gas concentrations."

(Climate Change 2007, IPCC, Fourth Assessment Report AR4)

Further to this carbon dioxide levels in the atmosphere were observed at over 400 parts per million in Hawaii. This is considered a milestone threshold and is at a level last thought to have occurred several million years ago when the Arctic was ice free and sea levels were up to 40m higher1.

The effects of climate change on flood risk management are obvious but in terms of fluvial flooding they are not straightforward to quantify. Changes in sea level have direct impact on coastal flooding and a range of predictions on projected rises are available. A number of meteorological projections are also available for changes in rainfall but these have a wide degree of variance particularly from season to season and are difficult to translate into river flow.

7.1.1 UOM 36 Context

Research into climate change in Ireland is coordinated by Met Éireann through the Community Climate Change Consortium for Ireland (www.c4i.ie). Research summarised in the report ‘Ireland in a Warmer World – Scientific Predictions of the Irish Climate in the 21st Century’ (Mc Grath et al, 2008)

1 http://www.theguardian.com/environment/2013/may/10/carbon-dioxide-highest-level-greenhouse- gas

IBE0700Rp0009 94 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL seeks to quantify the impact of climate change on Irish hydrology and considers the impacts of nine Irish catchments all of which were outside UoM 36 with the nearest being the Moy catchment in Mayo / Sligo. The ensemble scenario modelling from the regional climate change model predicts that between the two periods of 1961 – 2000 and 2021 – 2060 that Ireland is likely to experience more precipitation in autumn and winter (5 – 10%) and less precipitation in summer (5 – 10%). Between the periods of 1961 – 2000 and 2060 – 2099 this trend is likely to continue with increases of 15 – 20% generally, but up to 25% in the northern half of the country in autumn and drier summers of up to 10 – 18%.

The report seeks to further quantify the impact on hydrology in Ireland through the use of a HBV-Light conceptual rainfall runoff model (provided by Prof. Jan Seibert of Stockholm University) to simulate the effects of climate change on stream flow within the nine Irish catchments. The HBV-Light conceptual rainfall runoff model of the Moy catchment (HA34) was calibrated using historical meteorological data against the hydrometric gauge record at the Rahans gauging station (34001). The Moy model was found to be the best calibrated of the nine catchment runoff models when considered in terms of the R2 error measurement. Validation of the model against observed data at the gauging station found that the Moy model was moderately well calibrated when it came to simulating the annual maximum daily mean flow but that the model appeared to be underestimating mean winter flow. Following simulation of the meteorological climate change ensembles within the runoff models the following observations were made in the Moy and other catchments for the changes between the periods (1961 – 2000) and (2021 – 2060):

 Reductions in mean daily summer flow of up to 60% and increases in mean winter flow of up to 20% are the general pattern across all nine study catchments. In the Moy catchment this increase in mean winter flow was found to occur in February and March as opposed to January which was typical in the other catchments.

 Mixed results were obtained in terms of increased risk of extremely high winter flows in the Moy catchment although some other catchments such as the Feale and Suir showed risk doubling. It is thought that increased risk is more likely on catchments with a quicker response time.

 No change in annual maximum daily mean flow is apparent in the Moy catchment for all return periods but a moderate increase in risk is apparent on two of the other eight.

7.1.2 Sea Level Rise

Research from c4i summarised in the aforementioned report states that sea levels around Ireland have been rising at an annual rate of 3.5mm per year for the period 1993 – 2003 which is higher than the longer term rate of 1.8mm per year for the period 1963 – 2003. This trend is likely to be more modest in the with a ‘net trend’ (allowing for isostatic adjustment of the earth’s crust) of 2.3 – 2.7mm per year. On top of this the report notes that storm surges are likely to increase in frequency.

IBE0700Rp0009 95 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

The latest UK Climate Projections are covered in UKCP09 and put the central estimate of relative sea level rise at Belfast (to the east of UoM 36), based on a medium emissions scenario for the year 2095 at 31.6cm. The central estimate of a high emissions scenario for 2095 is 40.3cm but the predictions range from approximately 10cm to 70cm. The relative sea level rise detailed in UKCP09 allows for vertical land movement (isostatic adjustment) based on estimates taken from Bradley et al (2009). Storm surge models using the operational Storm Tide Forecasting Service (STFS) also show some increase in extreme storm surge although these rises are much less than was predicted in UKCIP02. It is not projected that the surge which could be expected to be exceeded for the 2, 10, 20 or 50 year return periods will increase by any more than 9cm by 2100 anywhere along the UK coast. It is noted however that other international climate models predict the rises to be much greater and these cannot be completely ruled out. In particular one high end surge scenario H++ combined with sea level rise infers increases in the 50 year return period extreme water level of as much as 3m by 2100 in some places around the UK.

7.2 AFFORESTATION

7.2.1 Afforestation in UoM 36

There is much legislation governing forestry practices in Ireland but it is implemented through the document ‘Growing for the Future – A Strategic Plan for the Development of the Forestry Sector in Ireland’ (Department for Agriculture, Food & Forestry, 1996). The plan points out that over the period from 1986 to 1996 afforestation saw quite a dramatic growth in Ireland from a level of approximately 70 km2 annually to almost 240 km2 annually in 1996 largely driven by a growth in private forestry activities. Within UoM 36 the current forest coverage as recorded in the 2006 CORINE land maps is shown in Figure 7.1.

IBE0700Rp0009 96 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 7.1: CORINE 2006 Forest Coverage in UoM 36 Compared to the rest of Ireland

The total forested area, including transitional woodland scrub, within UoM 36 is 245km² which is approximately 8.9% of the total area. This is below the average for the country which is approximately 10%. Forest cover in UoM 36 is densest in the west within County Donegal, Leitrim and in the north west of County Cavan. When we compare the CORINE 2006 database to the 2000 database there appears to have been an increase in the forested area as shown in Figure 7.2

IBE0700Rp0009 97 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 7.2: Forest Coverage Changes in UoM 36

IBE0700Rp0009 98 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 7.2 indicates that the increase in Forest coverage appears to be primarily due to an increase in transitional woodland scrub as opposed to actual forest. However closer inspection of the CORINE 2000-2006 datasets indicates that UoM 36 experienced an increase in both actual forest and transitional woodland scrub. The areas of forest from the two periods of the CORINE 2006 database are broken down further in Table 7.1.

Table 7.1: Afforestation from 2000 to 2006

CORINE CORINE Annualised Change 2000 2006 Change

Area % of Area % of Area % of Area % of (km²) catch. (km²) catch. (km²) catch. (km²) catch.

Forest 80.7 2.9 88.0 3.2 7.3 0.3 1.2 0.04

Transitional 149.4 5.4 157.1 5.7 7.7 0.3 1.3 0.05 Woodland Scrub

Total 230.1 8.4 245.1 8.9 15.0 0.5 2.5 0.09

Total Countrywide 6,631 9.4 7,087 10.1 456 + 0.65 76 +0.11

Table 7.1 indicates that both total forest / woodland scrub have increased in UoM 36 between 2000 and 2006 by 0.5% of the catchment area. This is slightly lower than the national average of 0.65%. Likewise, the UoM has experienced an annual increase of 0.09% compared to a national annual increase of 0.11%. If the annualised increase in afforestation were to continue for the next 100 years forest coverage in UoM 36 would more than double rising from 245 km² (8.9%) to 595 km² (18%).

The strategic plan sets out a target for the increase of forest area to 11,890 km² by 2035 in order to achieve a critical mass for a successful high-value added pulp and paper processing industry and this is the main driver behind the increases in forested area. If this value is to be realised nationally the rates of forestation will need to double in comparison to the change observed between 2000 and 2006.

7.2.2 Impact on Hydrology

A number of studies have been carried out on a range of catchments in an attempt to capture the effects of afforestation on runoff rates and water yields. The DEFRA (UK) report ‘Review of impacts of rural land use management on flood generation’ (2004) considers a number of case studies where the effects of afforestation on the catchment runoff were considered. The report concluded that the effects of afforestation are complex and change over time. A summary of the main findings in relation to afforestation are given below in relation to the River Irthing catchment in the north of England:

IBE0700Rp0009 99 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

 Water yield tends to be less from forest than pasture;

 In the Coalburn sub-catchment (1.5 km²) study peak flows were found to increase by 20% in the first 5 years and times to peak decreased, with the effect reducing over time (to 5% after 20 years). The time to peak was also reduced;

 In the overall River Irthing catchment (335 km²) the same effect was observed but to a much smaller degree.

The Coalburn catchment provides lessons which may be relevant to parts of UoM 36. The overall impact of afforestation is likely to be negligible in the larger river catchments considering the small proportion of recently forested area against the larger catchment area. The large forested areas within the catchment area of Bundoran and Tullaghan AFA are upstream of Lough Melvin and as such potential growth within these areas is not considered to have a potential significant effect on design flow estimates within the model. However the models receiving waters from areas directly upstream that are likely to see afforestation may be susceptible to the potential impacts of afforestation and as such some sensitivity analysis of the effects of afforestation would be prudent. As such it is recommended that sensitivity analysis to quantify the effects of potential afforestation is analysed at Model 1 – Ballinamore and Model 2- Ballyconnell since modelled reaches within them are located downstream of upland forest areas.

In each of these models the effects of afforestation will be modelled using the following recommended adjustments to the input parameters:

Table 7.2: Allowances for Effects of Forestation / Afforestation (100 year time horizon)

Mid Range Future Scenario High End Future Scenario (MRFS) (HEFS)

- 1/3 Tp¹ - 1/6 Tp¹ + 10% SPR² Note 1: Reduce the time to peak (Tp) by one sixth / one third: This allows for potential accelerated runoff that may arise as a result of drainage of afforested land

Note 2: Add 10% to the Standard Percentage Runoff (SPR) rate: This allows for increased runoff rates that may arise following felling of forestry

(Extracted from ‘Assessment of Potential Future Scenarios for Flood Risk Management’ OPW, 2009)

IBE0700Rp0009 100 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

7.3 LAND USE AND URBANISATION

The proportion of people living in urban areas (classified as towns with a population of 1,500 or more) has increased dramatically in recent years with a nationwide increase of over 10% in the total urban population recorded between the 2006 census and the 2011 census. The total population within the counties located within UoM 36 has increased by varying degrees since 1991 as demonstrated by Table 7.3.

Table 7.3: Population Growth in UoM 36 (Source: Central Statistics Office of Ireland (CSO))

1986 1991 1996 2002 2006 2011 Population (Number) 52,379 51,293 51,313 52,772 55,997 60,495 Actual Change Since Previous Census -1,086 20 1,459 3,225 4,498 Monaghan (Number) Population Change Since Previous -2.1 0.04 2.8 6.1 8 Census (%) Population (Number) 53,965 52,796 52,944 56,416 64,003 72,874 Actual Change Since Previous Census -1,169 148 3,472 7,587 8,871 Cavan (Number) Population Change Since Previous -2.2 0.3 6.6 13.4 13.9 Census (%) Population (Number) 27,035 25,301 25,057 25,815 28,950 31,798 Actual Change Since Previous Census -1,734 -244 758 3,151 2,848 Leitrim (Number) Population Change Since Previous -6.4 -0.97 3 12.2 9.8 Census (%) Population (Number) 129,203 128,117 129,994 137,575 147,264 161,137 Actual Change Since Previous Census -1,086 1,877 7,581 9,689 13,873 Donegal (Number) Population Change Since Previous -2.1 1.5 5.8 7 9.4 Census (%)

As indicated by Table 7.3, UoM 36 has seen moderate population rise since 1991. It is evident that the percentage of population change has been steadily increasing with an average annual growth rate of 1.2% within the counties containing AFAs since the 1991 census.

IBE0700Rp0009 101 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

None of the counties have shown an increase in the share of the rural population since 2006 and as such the data would suggest that the population growth within UoM 36 has been almost entirely within the urban centres

Table 7.4 confirms that urban population growth within the urban AFAs (population > 1500) for the period 2006 – 2011 has been quite significant ranging from 9% in Bundoran to 29.5% in Cavan over the five year census period.

Table 7.4: Population Growth within Urban AFAs (Source: Source: Central Statistics Office of Ireland (CSO))

Increase Since 2006 Urban Area Population 2011 (%)

Bundoran 2140 9 Cavan 10205 29.5

The total percentage population growth in these AFAs however is 19.3% for the period 2006 – 2011 which equates to an average annual growth rate of approximately 3.9%.

To determine if these changes translate into equivalent increases in urbanised areas we must examine the CORINE database within UoM 36. CSO Population data per town is not available for 2002 and CORINE datasets are only available for 2000 & 2006 so direct comparison is not possible. Furthermore the CORINE datasets are narrow in terms of the time period under consideration and may represent a period of rapid development. However this dataset is a further indicator of urbanisation growth and is the only dataset which measures the urban fabric, which more directly affects run-off compared to population growth. A simple comparison of the datasets from 2000 to 2006 within UoM 36 appears to show that there has been an increase in artificial surfaces within UoM 36 from 23.3 km² in 2000 to 28.7 km² in 2006 which represents an increase of just over 23% in six years (see Figure 7.3).

IBE0700Rp0009 102 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 7.3: UOM 36 CORINE Artificial Surfaces (2000 / 2006)

Closer inspection of the CORINE datasets shows that a notable proportion of this growth in artificial surfaces is due to changes outside the AFAs. 33.3% of artificial surface growth between 2000 and 2006 occurred outside the AFAs, a large proportion of which consists of the addition of quarries and sports facilities including the golf course extension at Clones. Golf courses, however, are generally permeable surfaces and although quarries are generally impermeable runoff generally collects within them and does not directly affect the AFAs. The AFAs with an increase in the extent of artificial surfaces are:

 Bundoran: 39.6% increase (5.7% annually)  Ballyconnell: 73.5% increase (9.6% annually)  Cavan: 31.8% increase (2.7% annually)  Ballybay: 25 % increase (1.7% annually)

No change in terms of artificial surface land cover was experienced between 2000 and 2006 within the Ballinmore AFA or in the catchment upstream.

The average annual growth rate in the artificial surfaces within all UoM 36 AFA extents is 4.9%.

IBE0700Rp0009 103 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

The CSO has also produced Regional Population Predictions for the period of 2011 - 2026 based on a number of scenarios considering birth rates and emigration. Under all the modelled scenarios the is set to experience strong population growth.

Under the M0F1 Traditional model, which tends to reflect longer term growth trends, the projected rise for the region in the 15 year period equals 6.3% equating to an average annual growth rate of 0.4%. Under the M2F1 Recent model, which tends to reflect more recent growth rates, the projected rise in population is 25% equating to an annual average growth rate of 1.5%. Any estimation of the rate of urbanisation should consider the three measures of recent growth which have been examined along with the projected population increases from CSO for the region. These are summarised in Table 7.5 below.

Table 7.5: Urbanisation Growth Indicators

Population in Population in Artificial Surfaces CSO M0F1 CSO M2F1 UoM 36 UoM 36 Urban (CORINE) within Population Population Counties AFAs UoM 36 AFA Projection Projection Extent 1991 - 2011 2006 - 2011 2011 - 2016 2011 - 2016 2000 - 2006

Average Annual 1.2% 3.9% 4.9% 0.4% 1.5% Growth Rate (%)

7.3.1 Impact of Urbanisation on Hydrology

The effect of urbanisation on runoff is well documented. The transformation from natural surfaces to artificial surfaces, which in almost all cases are less permeable, increases surface runoff such that it is generally faster and more intense. If for example we consider the FSU ‘URBEXT’ catchment descriptor at HEP 36_189_3_RA on the Cavan River just before its confluence with the Annalee River (Model 4, Cavan AFA) currently at 4.96% the URBEXT and apply the different growth rates outlined in Table 7.5 we could see a large degree of variance in the predicted future urban extent. At the lower end the catchment could potentially rise to 7.4% urbanised (based on growth of 0.4% per annum) and at the upper end of the growth rates the catchment could become fully urbanised (based on growth of 4.9% per annum) in the 100 year time span to be considered under the future scenarios. The latter is not considered realistic as it is based on a short comparison period of land use datasets during a time of high economic growth and property boom which has since ended. Considering all the data growth rates of 1% and 2.5% are considered appropriate estimates for the 100 year mid range (MRFS) and high end (HEFS) future scenarios respectively.

Based on the FSU equation (WP 2.3) for index flow estimation (Qmed) based on catchment descriptors the Urban Adjustment Factor (UAF) for the Cavan River catchment would vary as shown in Table 7.6

IBE0700Rp0009 104 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL if we consider growth of 1% and 2.5% as representative for the 100 year mid range (MRFS) and high end (HEFS) future scenarios respectively.

Table 7.6: Potential Effect of Urbanisation on Qmed Flow on the Cavan River (at confluence with Annalee River)

Total Catchment Growth Rate URBEXT1 UAFS2 Qmed Flow m3/s Present Day n.a. 4.96 1.07 6.5

100 Year MRFS 1% p.a. 13.42 1.21 7.3

100 Year HEFS 2.5% p.a. 58.60 1.98 12.0

Note 1: URBEXT is the percentage of urbanisation in the catchment Note 2: Urban Adjustment Factor (UAF) = (1 + URBEXT/100)1.482

Table 7.6 represents one of the more urbanised catchments within UoM 36 and as such can be considered a more onerous example of the potential effect of urbanisation. At the less onerous end catchments with no existing urbanisation could remain totally rural. There are also examples of catchments representing small watercourses on the edges of AFAs which are currently totally rural but which could become totally urbanised in 100 years time if the spatial growth of the urban fabric of the AFA occurs in the direction of that small catchment. In this scenario the application of growth rates to a URBEXT value of zero will have no effect and as such the effect could be missed using a methodology that applies factors to the URBEXT values. It must also be considered that any attempts to predict the spatial growth of AFAs on a 100 year time frame would be highly uncertain as growth rates and growth direction are dictated by complex social, economic and cultural factors which cannot be predicted far into the future.

In light of these large uncertainties it is not considered prudent to attempt to predict the varying effects of urbanisation on a HEP by HEP basis and as such it is considered prudent to apply a factor based on the average URBEXT values within the Unit of Management and the growth rates considered above of 1% and 2.5% respectively for the medium and high end future scenarios. It is still considered prudent though that small urban watercourses with catchments that emanate around the periphery of AFA extents are considered to become much more urbanised and as such will be considered as having URBEXTs of 50% for the mid range and 85% for the high end future scenarios (85% is considered the urban saturation level as some green spaces will always remain).

The effect of recent developments in sustainable drainage policy and guidance must also be considered. The move away from conventional drainage systems is likely to gather pace with the aim of these policies and systems to provide drainage for urban areas which recreates the runoff behaviour of the rural catchment in an attempt to mitigate flood risk. Sustainable drainage policy is already being implemented in Dublin through the Greater Dublin Strategic Drainage Strategy

IBE0700Rp0009 105 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

(GDSDS) but is largely in its infancy outside the capital but it would be expected to develop greatly throughout the time span of the future scenarios. Therefore the current effect of urbanisation on catchment runoff could be expected to reduce over time as sustainable drainage policy and systems develop.

There is no directly applicable data / research into the likely effectiveness of SuDS policies at reducing the impact of future urbanisation on catchment runoff in an Irish context. The paper titled ‘Performance and Design Detail of SUDS’ (Macdonald & Jefferies, 2003) outlines research undertaken in Scotland on the effectiveness of a range of different systems implemented and found that the effectiveness is dependent on the type of system implemented (source control or site / regional) but that all systems considered delivered at least a 50% reduction in peak runoff rate rising to over 80% for source control systems.

Given the development of SuDS policies in recent years it is appropriate that some allowance is made for the effectiveness of SuDS at mitigating the impact of urbanisation on peak runoff rates. It is therefore assumed that SuDS policies and systems will mitigate the impact of future urbanisation by half (50% effective) within the tributary watercourses affecting the AFAs where SuDS implementation is most likely to be focussed. The urban adjustment factors which will therefore be applied to the design flow estimates for the mid range and high end future scenarios for a typical UoM 36 catchment (shown here as a catchment average HEP) and for a small tributary catchment which may be susceptible to full urbanisation are shown in Table 7.7 below:

UAF Growth Rate URBEXT1 UAF2 (adjusted for SuDS)

HEP Average 2.97 1.04 n.a.

100 Year MRFS 1% p.a. 8.03 1.12 n.a.

100 Year HEFS 2.5% p.a. 35.09 1.56 n.a.

Tributary Catchments susceptible varies varies varies to full urbanisation

100 Year MRFS n.a. 50 1.824 1.412

100 Year HEFS 85 2.488 1.744

Note 1: URBEXT is the percentage of urbanisation in the catchment Note 2: Urban Adjustment Factor (UAF) = (1 + URBEXT/100)1.482

Table 7.7: Potential Effect of Urbanisation on Qmed Flow in UoM 36

The allowances for urbanisation are based on a robust analysis of population growth, recent increases in artificial surfaces and population projections from CSO. However this is based on extrapolation of current growth rates which are dependent on complex social, economic and

IBE0700Rp0009 106 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL environmental factors. Furthermore the estimation of the Urban Adjustment Factor under FSU is based on data from existing urban catchments and therefore does not reflect the impact of recent policy changes and changes to drainage design guidelines where the emphasis is on developments replicating the existing ‘greenfield’ flow regime through attenuation and sustainable urban drainage systems. An approach has been developed that considers an average adjustment factor for the majority of HEPs across UoM 36. These adjustment factors will translate into increases in flow of approximately 8% and 36% for the mid range and high end future scenarios respectively. Small catchments emanating from just outside AFAs which would be susceptible to full urbanisation are to be considered separately and will see their flows increase by up to 36% and 68% for the mid range and high end future scenarios respectively.

There is high uncertainty in all of these allowances as discussed above and it is recommended that they are reviewed at each cycle of the CFRAM Studies.

7.4 HYDROGEOMORPHOLOGY

Hydrogeomorphology refers to the interacting hydrological, geological and surface processes which occur within a watercourse and its floodplain. Erosion and deposition of sediment are natural river processes that can be exacerbated by anthropogenic pressures such as land use practices and arterial drainage.

7.4.1 Soil Type

Figure 7.4 overleaf illustrates the soil types that characterise UoM 36. For discussion purposes the south eastern portion is called UoM 36a and the north western portion of UoM 36 is called UoM 36b. The drumlin landscape across UoM 36a is characterised by the predominance of deep gleys. UoM 36b is characterised by peat and peaty and deep gleys.

Within UoM 36b and also the north western boundary of UoM 36a, the peaty soils would indicate relatively high susceptibility to soil erosion and can be considered a source of sediment which if accelerated due to anthropogenic pressures and given the right pathway (channel typology) can make its way to the watercourse network. The deep gleys that characterise UoM 36a indicate poorly drained soils and higher potential for surface water runoff.

There is currently ongoing research in Ireland and the UK involving modelling the risk of diffuse pollution in river catchments, including sediment transport. Recent research has focussed attention on assessing risk based on erodibility and hydrological connectivity to the river network, with land use/land cover the most common measure of erodibility. While soil type clearly has an influence on erodibility, Reaney et al. (2011) argue that an emphasis upon land cover is warranted as land cover is typically correlated with soil type (refer to Section 7.4.3).

IBE0700Rp0009 107 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 7.4: UoM 36 Soil Types (Source: Irish Forest Soils Project, FIPS – IFS, Teagasc, 2002)

7.4.2 Channel Typology

As part of national Water Framework Directive studies on hydromorphology through River Basin District projects a national channel typology dataset was defined for Irish rivers2. It classified river channels into channel types at 100m node points along each reach. Table 7.5 below outlines the four main channel types and how these relate to the four catchment descriptors used to define them; valley confinement, sinuosity, channel slope and geology.

2 (http://www.wfdireland.ie/docs/20_FreshwaterMorphology/CompassInformatics_MorphologyReport)

IBE0700Rp0009 108 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Table 7.8: Channel Types and Associated Descriptors

Channel Type Confinement Sinuosity Slope Geology

Step Pool / Cascade High Low High Solid

Bedrock High Low Variable Solid

Riffle & Pool Low - Moderate Moderate Moderate Drift / Alluvium

Lowland Meander Low High Low Drift / Alluvium

Typical undisturbed channel behaviour in terms of flow is described as follows for each of the channel types shown.

Bedrock: Boulders and cobbles often exposed, but few isolated pools Overbank flows uncommon. Morphology only changes in very large floods.

Cascade and step-pool: At low flows, many of the largest particles (boulders, cobbles) may be exposed, but there should be continuous flow with few isolated pools

Pool-riffle: Gravel bars may be exposed in low water conditions, but gravels and cobbles in riffles as well as logs and snags are mainly submerged.

Lowland Meandering: In low flow conditions some bars or islands may be exposed, but water fills the majority of the channel.

Figures 7.5 and 7.6 illustrate the channel typology and channel slope of UoM 36 in a national context.

IBE0700Rp0009 109 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

UoM 36a UoM 36b

Figure 7.5: Channel Types of UoM 36 in National Context (Source: WFD Channel Typology dataset)

IBE0700Rp0009 110 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 7.6: Channel Slopes of UoM 36 in National Context (Source: WFD Channel Typology dataset)

UoM 36a is a relatively low slope, low energy system with predominance of inland low slope lowland meandering channels. In contrast, UoM 36b is a generally characterised by relatively steep, step pool cascade channels consistent with the upland areas to the west of Ireland.

Figures 7.7 and 7.8 provide a closer look at channel types and slope within UoM 36.

IBE0700Rp0009 111 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 7.7: WFD Channel Typology UoM 36

As illustrated by Figure 7.7, Ballinamore, Ballyconnell, Cavan and Ballybay AFAs are located on lowland meandering, low energy watercourses. These drain the surrounding lakelands and drumlin landscapes. Ballinamore is at the upper end of the Erne catchment and is the first AFA downstream of the steeper step-pool cascade mountain streams of Anierin and Bencroy in County Leitrim. Bundoran and Tullaghan AFA is located on several small coastal pool-riffle type watercourses. These channel types also represent the change in channel slope from relatively steep in upland areas to relatively shallow moving downstream. Figure 7.8 indicates the change in channel slope across the UoM 36.

IBE0700Rp0009 112 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 7.8: Changes in Channel Slope UoM 36

The steepest channels are located at the mountainous areas within UoM 36b. These sub-catchments are all upstream of Lough Melvin which can be considered to dissipate the high energy of these steep watercourses. Downstream of Lough Melvin and in the other catchments affecting Bundoran and Tullaghan slope values are all low in the 0 to 0.0293 (flat to 1 in 34) category. The mountain streams on the foothills of Slieve Ainagh and Benboy in County Leitrim are the steepest within UoM 36a and form the headwaters of the Erne catchment upstream of Ballinamore. Channel slope in this area ranges from 0.23 to 0.28 (in other words 1 in 3 to 1 in 4). The remainder of UoM 36a is generally characterised by low slope rivers ranging from 0 to 0.0293 (flat to 1 in 34)

These channel types are typical of Irish catchments. Sediment transport, erosion and deposition are natural morphological processes. In larger catchments it is expected that the upper reaches will be more dynamic with erosion taking place and as the river moves to the lower lands, sediment is accumulated and transported. Sediment deposition is expected where the channel meanders and loses energy. Based on the aforementioned figures, the AFAs that could be affected by sediment deposition are:

 Ballyconnell  Ballinamore  Cavan IBE0700Rp0008 113 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

 Ballybay

7.4.3 Morphological Pressures - Land Use

Figure 7.9 illustrates the land use types within UoM 36. It is essentially rural dominated by pasture (62% of catchment area) and natural vegetation (14%). There are pockets of peat bogs in the uplands to the north west (9%) and areas of forest, again mainly in the north west, make up 11% of catchment area (refer to Section 7.1).

Drainage of bog lands and peat extraction activities potentially lead to large quantities of peat silt being discharged to the receiving waters. However peat bogs and associated drainage of the land are restricted in area and location within UoM 36 such that they do not have significant impact on the modelled watercourse catchments. Lough Melvin acts as a buffer between peat bogs and modelled watercourses downstream for the Bundoran and Tullaghan AFA. The only exception is Ballinamore AFA which is at the receiving end of mountain streams that drain the peat bogs located on the eastern foothills of Slieve Anierin and Bencroy. Therefore sedimentation due to upstream peat extraction activities is a consideration within Model 1. It is not expected that sediment transport would continue downstream from Model 1 to Model 2 (Ballyconnell) given the presence of lakes on the river channel which will act as sinks to trap sediment.

In and around the urban fabric of the AFAs it may also be possible to identify potential sources of localised impacts such as sediment loss from construction sites, road wash or channel works and if such sources are known to the Progress Group it is recommended that these are made known to the Study team ahead of the options analysis stage (refer also to Section 7.4.8).

IBE0700Rp0008 114 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 7.9: UoM 36 Land Use (CORINE 2006)

IBE0700Rp0008 115 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Pasture is the predominant land use in UoM 36. Overgrazing of soils in areas of commonage is also a source of increased geo-morphological impact due to exposed soils washing into headwaters, increasing flashiness through more rapid runoff and increased sediment load in rivers resulting in increased deposition downstream. Under the Water Framework Directive this pressure was identified as a potential risk to river morphological status in the national context but not within UoM 36, indicating that overgrazing is not an issue, certainly not from a flood risk management perspective.

Arable land coverage is very low in UoM 36 at less than 0.3% and is therefore not considered a source of soil erosion and sediment loss to watercourses.

The impact of hydro-geomorphological changes on UoM 36 ultimately applies to the performance of flood risk management options. In terms of the modelled watercourses, the recently acquired channel cross section survey data will reflect the current status of the watercourses in terms of siltation based on the measurements taken for modelling purposes. The impact of sediment transport and deposition within the AFAs highlighted here will be considered further under the hydraulic modelling of options stage of the CFRAM Study.

The transportation of sediment and subsequent deposition within the following AFAs is identified for further consideration under hydraulic modelling:

 Ballyconnell  Ballinamore

7.4.4 Arterial Drainage (Channelisation)

A further consideration in UoM 36 is the potential effect of arterial drainage on watercourse channel and floodplain geomorphology. The original Arterial Drainage Act, 1945 was a result of the Browne Commission which examined the issue of flooding and the improvement of land through drainage works and was mainly focussed on the agricultural context. Following flood events in the mid to late 1980s the emphasis on flood management shifted to the protection of urban areas and as such the Arterial Drainage Amendment Act was passed in 1995. This widened the scope of the act to cover the provision of localised flood relief schemes. The OPW have used the Arterial Drainage Acts to implement various catchment wide drainage and flood relief schemes. Arterial drainage scheme works may consist of dredging of the existing watercourse channels, installation of field drains / drainage ditches and the construction of earthen embankments using dredged material to protect agricultural land.

The extent of the modelled watercourses and their contributing catchments that are affected by arterial drainage within UoM 36 is conveyed by the Arterial Drainage Scheme and Drainage District GIS shapefiles provided by OPW. Rivers within modelled catchments that have been subject to arterial drainage schemes and subsequent channel maintenance are shown in Figure 7.10.

IBE0700Rp0008 116 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Figure 7.10: Arterial Drainage Schemes within UoM 36 Modelled Catchments

As indicated by Figure 7.10, Ballinamore, Ballyconnell and Cavan AFAs are located within modelled catchments that have been extensively arterially drained in the past. These schemes are all termed “Drainage Districts” which means that the drainage works was undertaken before the introduction of the Arterial Drainage Act in 1945, usually by local drainage boards for agricultural purposes. They do not fall under OPW’s remit for channel maintenance in terms of channel capacity for flood flow conveyance and in most cases, they have re-naturalised as a result. In terms of sedimentation of rivers, the initial drainage schemes have had the long term effect of making river courses more susceptible to bed and bank erosion in high flow conditions and resulting siltation. This was due to the removal of natural gravels and bank vegetation. However in the case of historical drainage districts, the lack of systematic and programmed maintenance has allowed re-naturalisation of channels and as such this is not a concern.

The Ballinamore Ballyconnell Canal is also included in this dataset. The canalisation of the existing Woodford River took place in the late 19th century, gradually fell into disrepair and was restored as a navigable waterway in the 1990s when it was renamed as part of the Shannon-Erne Waterway. Waterways Ireland maintains it for navigation purposes.

The hydrological analysis and design flow estimation undertaken as part of this study seek to represent as accurately as possible the present day scenario. The ARTDRAIN2 FSU catchment

IBE0700Rp0008 117 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL descriptor is included in the ungauged index flow estimation equation where applicable. In the case of UoM 36, the ARTDRAIN2 catchment descriptor is given as zero for all FSU nodes within modelled catchments in UoM 36. This supports the previous assumption that rivers historically drained before the 1945 Arterial Drainage Act have re-naturalised and as such do not further warrant hydrological consideration in this Study.

7.4.5 Localised Pressures

As well as the catchment based pressures discussed in this report, localised morphological changes can have an impact on channel capacity and the structural integrity of flood defences due to the effects of scour from high sediment loads within rivers. For example known areas of bank erosion within AFAs can undermine existing channel structures. At this stage of the study, data relating to such localised effects within AFAs has not been received for inclusion in this analysis. It is recommended that Progress Group members confirm if such data is available within their organisations that could be of use in the options development process.

IBE0700Rp0008 118 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

7.5 FUTURE SCENARIOS FOR FLOOD RISK MANAGEMENT

The OPW does not have a specific policy for the design of flood relief schemes but has produced a draft guidance note ‘Assessment of Potential Future Scenarios for Flood Risk Management’ (OPW, 2009). The document gives guidance on the allowances for future scenarios based on climate change (including allowing for the isostatic movement of the earth’s crust), urbanisation and afforestation. Table 1 from the guidance has been adapted for the purposes of this study to take into account catchment specific effects and is presented here as the basis for the design flow adjustment for the mid range (MRFS) and high end (HEFS) future scenarios.

Table 7.9: UoM 36 Allowances for Future Scenarios (100 year time horizon)

MRFS HEFS

Extreme Rainfall Depths + 20% + 30%

Flood Flows + 20% + 30%

Mean Sea Level Rise + 500mm + 1000mm

Urbanisation URBEXT multiplied by 2.73 URBEXT multiplied by 11.83 Susceptible sub-catchments Susceptible sub-catchments URBEXT = 50%4 URBEXT = 85%4

Afforestation - 1/3 Tp¹ - 1/6 Tp¹ + 10% SPR²

Note 1: Reduce the time to peak (Tp) by one sixth / one third: This allows for potential accelerated run-off that may arise as a result of drainage of afforested land

Note 2: Add 10% to the Standard Percentage Run-off (SPR) rate: This allows for increased run-off rates that may arise following felling of forestry

Note 3: Reflects growth rates of 1% and 2.5% p.a. for mid range and high end future scenarios. To be applied to FSU URBEXT Physical Catchment Descriptor (PCD) up to a maximum of 85%.

Note 4: Applied to areas of sub-catchment or tributary catchment within the AFA which are susceptible to rapid urbanisation but which at present are predominantly undeveloped (i.e. growth rates applied to existing low FSU URBEXT PCD would result in an unrealistically low future scenario URBEXT).

The peak flows for each of the future scenario design events for every HEP can be found in Appendix D.

IBE0700Rp0008 119 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

7.6 POLICY TO AID FLOOD REDUCTION

Considering the projected growth in population predicted within UoM 36 the main future change which could increase flood risk is urbanisation of the catchment. If not managed correctly rapid urbanisation could lead to large swathes of some catchments becoming hard paved and drained through conventional drainage systems which are designed to remove water from the urban area quickly and efficiently. This could have potentially significant implications for fluvial flooding as the flood flows in the watercourses and rivers would intensify. Some of the smaller watercourses in particular could become prone to flash flooding if they become urbanised.

Sustainable Urban Drainage (SuDS) policy has been about for over a decade now in the UK and Ireland. The term covers a range of practices and design options that aim to replicate the pre- development surface water runoff characteristics of the undeveloped catchment following development both in terms of water quality but more importantly, from the perspective of flood risk management, in terms of runoff peak flow, intensity and volume.

SuDS policy at a national level is outlined in the OPW document “The Planning System and Flood Risk Management” (November 2009) where guidance on its design and implementation is also provided. Typical measures include soft engineered solutions such as filter strips, swales, ponds and wetlands and hard engineered solutions such as permeable paving, ‘grey water’ recycling underground storage and flow control devices. The implementation of successful SuDS requires a joined up policy that covers planning, design, construction and maintenance. One of the biggest issues surrounding SuDS implementation is long term ownership and maintenance although the long term benefits of SuDS can be shown to outweigh the costs associated with these issues.

If a comprehensive SuDS policy is implemented covering planning, implementation and maintenance, then the impacts of urbanisation on flood flows can be substantially mitigated. The use of retrofitting SuDS in areas of flood risk will be considered as a flood risk management option in the options development phase of this CFRAM Study where appropriate.

IBE0700Rp0008 120 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

8 SENSITIVITY AND UNCERTAINTY

Hydrological analysis and design flow estimation are probabilistic assessments which originate from observed data. The long term conditions which affect the observations, whether they are climatic or catchment-based, have been shown to varying degrees to be changing over time. Further to this, the degree of uncertainty within the sub-catchments analysed under the North Western – Neagh Bann CFRAM Study varies greatly due to the quality and availability of observed data. The following factors which may affect the quality of both the analysed historic events and the estimation of the future design events are listed below:

 Hydrometric data record length and gaps  Hydrometric data quality (classified in terms of the rating confidence under FSU WP 2.1)  High quality meteorological data availability  Calibration quality of hydrological models (generally a result of all of the above)  Standard error of flow estimation (catchment descriptor based) techniques  Future catchment changes, urbanisation, afforestation, sedimentation etc.  Climate change

The above list is not exhaustive but seeks to identify the main potential sources of uncertainty in the hydrological analysis. Further to these the list of factors which could potentially affect the uncertainty and sensitivity of the assessment of flood risk under the North Western – Neagh Bann CFRAM Study is subject to further uncertainties and sensitivities related to the hydraulic modelling and mapping stages. Examples of some of the modelling considerations which will further affect the sensitivity / uncertainty of the CFRAM Study outputs going forward from the hydrological analysis are past and future culvert blockage and survey error (amongst others). These considerations will be considered through the hydraulic modelling and mapping report along with the hydrological considerations listed here to build a complete picture of uncertainty / sensitivity of Study outputs.

It is not possible to make a quantitative assessment of all of the uncertainties as some of the factors are extremely complex. Nevertheless it is important that an assessment is made such that the results can be taken forward and built upon through the subsequent phases of the study. It is also important that the potential sources of uncertainty in the hydrological analysis and design flow estimation are flagged such that the integrated process of refining the hydrological inputs and achieving model calibration can be achieved more efficiently through a targeted approach. A qualitative assessment has therefore been undertaken to assess the potential for uncertainty / sensitivity for each of the models and is provided in this chapter. The assessed risk of uncertainty is to be built upon as the study progresses through the hydraulic modelling and mapping stages. Following completion of the present day and future scenario models the assessed cumulative uncertainties can be rationalised into a sensitivity / uncertainty factor for each scenario such that a series of hydraulic model runs can be performed which will inform the potential error on the flood extent maps.

IBE0700Rp0008 121 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

8.1 UNCERTAINTY / SENSITIVITY ASSESSMENT MODEL BY MODEL

Table 8.1: Assessment of contributing factors and cumulative effect of uncertainty / sensitivity in the hydrological analysis

Model Model Uncertainty / Sensitivity – Present Uncertainty / Sensitivity – Future Scenarios Notes No. Name Day Scenario

Observed Catchment Ungauged Forest- Urban- Climate Sediment- Flow Data1 Data2 Flow ation4 isation5 Change6 ation7 Estimates3

No useable gauging stations directly on the modelled 1 Ballinamore Medium / Low Medium / Medium Low Medium Medium watercourse but A2 stations just upstream and High Low downstream. Uncertainty in relation to Bellaheady GS. Watercourses unlikely to be greatly affected by urbanisation. Catchment to north of AFA is ungauged and there is some uncertainty in the applicability of the gauge data to this catchment. This catchment could be susceptible to afforestation. Peat in upper catchment potential sources of sediment and potential for deposition at AFA in flat, meandering reaches.

2 Ballyconnell Medium / Medium / Medium / Medium Medium / Medium Medium / A2 gauging station located within modelled reaches Low Low Low but some uncertainty in Qmed value. FSU catchments High Low downstream of AFA required amendment due to border. Some uncertainty in the applicability of the gauge data to the small watercourses affecting AFA. Small tributary catchment could be susceptible to some urbanisation and afforestation. Peat in upland catchment but sediment likely to be intercepted by lakes upstream.

3 Ballybay Low Low Medium / Low Low Medium Low Gauges at upstream and downstream extents of the Low model. One A1 and one good confidence at Qmed following rating review. Some uncertainty in the applicability of the gauge data to the small watercourses affecting AFA. Smaller tributary catchments susceptible to urbanisation but effect likely to be reduced significantly through attenuation by lakes

IBE0700Rp0008 122 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Model Model Uncertainty / Sensitivity – Present Uncertainty / Sensitivity – Future Scenarios Notes No. Name Day Scenario

Observed Catchment Ungauged Forest- Urban- Climate Sediment- Flow Data1 Data2 Flow ation4 isation5 Change6 ation7 Estimates3

4 Cavan Low Low Medium / Low Medium / Medium Low High quality gauge data at downstream extents of the Low High model. Some uncertainty in the applicability of the gauge data to the small watercourses affecting AFA. Many of the mostly small catchments affecting the AFA could be affected if Cavan were to see high levels of urbanisation.

5 Bundoran & Low Medium / Medium Low Low Medium Low A2 gauging station located just upstream of AFA. High Many catchments emanating from NI and as such Tullaghan issue with catchment descriptors. These have been estimated / calculated based on NI mapping / orthophotography / nearby PCDs. Areas were required to be derived and new catchment descriptors estimated or borrowed from adjacent suitable catchments.

1 Observed flow data marked n.a. where there is no gauged data within the modelled catchment to inform the flood flow estimation for the model. Low to high reflects uncertainty in the gauged data at Qmed if available. 2 Catchment data refers to delineated catchment extents or catchment descriptors. Low to high reflects uncertainty in physical catchment descriptors or catchment delineation. 3 Ungauged flow estimates based on FSU WP 2.3. Dependent on 1 & 2. Where high quality gauge data is available along modelled reach upon which adjustment can be performed then uncertainty is considered low. Where no gauge data is available within catchment then certainty is considered medium to high. Uncertainty greater in smaller, urbanised catchments where ungauged estimation methodologies are considered to be more sensitive. 4 See Section 7.2 Considered to be low risk of uncertainty to hydrological analysis in most of UoM 36. High risk where there is significant risk of forestation of small catchment just upstream of AFA which is the dominant source of flood risk to the catchment. 5 See Section 7.3 Considered generally to be a medium to high risk of uncertainty to hydrological analysis in urban areas where potential significant, dense urbanisation is possible which would make up a significant proportion of the catchment. High risk where small catchments largely contained within the AFA extents and potentially subject to high risk of urbanisation. 6 See Section 7.1 Considered a medium risk of uncertainty to hydrological analysis in all cases due to the range of projections. 7 Sedimentation of channels causing capacity issues or localised impacts on channel structures are to be considered in options development phase of CFRAM Study where relevant. Degree of uncertainty indicated here is based on qualitative assessment of accelerated soil erosion risk due to land use pressures and pathways to watercourses. Considered under future scenarios only as present day sediment conditions are reflected by recently captured channel survey data.

IBE0700Rp0008 123 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

8.2 CONCLUSIONS OF SENSITIVITY ANALYSIS

The assessment of uncertainty and sensitivity in each category is relative within UoM 36. The assessment of uncertainty as being medium or high does not suggest that the analysis is poor but rather in the context of the design flow estimation techniques being employed in the North Western – Neagh Bann CFRAM Study that uncertainty in that category is towards the higher end of the range. For example the modelled watercourses which affect the Bundoran and Tullaghan AFA has many catchments emanating from Northern Ireland and the FSU catchment descriptors did not account for this. The risk to design flow estimation has been mitigated by re-estimating catchment descriptors for CFRAMS to ensure the inclusion of the full catchment. However, risk remains as medium/high, to acknowledge this increased uncertainty associated with the catchment descriptors. In UoM 36 the largest degree of uncertainty for the present day scenarios is attributed to this scenario. The remaining AFAs have a medium/low risk attributed to ungauged flow estimates, which is a reflection of the good availability of reliable gauging stations within UoM 36.

In the future scenarios climate change has been defined as a potential source of medium uncertainty due to the inherent uncertainties surrounding climate change science and how these will translate into changes in fluvial flood flows in Ireland. Within UoM 36 it is considered that urbanisation is not generally a source of high uncertainty in the prediction of future flood flows with the exception of Cavan. This AFA is of a size and growing such that large swathes of dense development drained through conventional drainage systems could in the future make up a large proportion of the catchment or sub-catchments. The factors which affect urbanisation are difficult to predict for a 100 year time horizon due to the complex social, cultural and economic factors which affect it. At the upper limit of the predictions large swathes of the smaller catchments on the periphery of towns could become fully urbanised which could more than double some of the index flood flows. There is also the affect of sustainable drainage to consider which adds a further degree of uncertainty depending on the extent to which it is successfully implemented.

Afforestation has been identified as a potential source of future uncertainty for two of the AFAs / models – Ballinamore and Ballyconnell. In the majority of catchments / sub-catchments in UoM 36 the proportion of recently forested land coverage is likely to be low but for AFAs which have an area directly upstream which could be newly forested over a short time frame (such as sub-catchments which are on the edge of forested area) this could be significant. The impact of hydro- geomorphological changes has been assessed and is ultimately a consideration for flood risk management options and it may be appropriate that this is considered through the modelling of options at the hydraulic modelling stage.

IBE0700Rp0008 124 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

9 CONCLUSIONS

Good hydrometric data exists within the larger channels of UoM 36 which is of sufficient quality to be of use for design flow estimation and as such there is generally a high degree of certainty in design flow estimates.

The FSU Qmed pcd equation (WP 2.3) generally tends to underestimate when compared with Qmed gauged at hydrometric stations. This has had the effect of increasing Qmed pcd estimates at HEPs where these stations are used as pivotal sites. However, in the case of smaller tributaries that are less comparable to the larger pivotal sites, alternative pivotal site options have been reviewed and adjustment factors applied where relevant.

There is good availability of meteorological data, both daily and hourly within and in close proximity to UoM 36. These provide the high temporal resolution data needed for driving the rainfall runoff model that has been undertaken at station 36150. This hydrometric station is located just upstream of a modelled reach on Model 3 (Ballybay) and given that it was not rated under FSU and had also been specified for CFRAM Rating Review it was considered that rainfall runoff (NAM) modelling would be of benefit. Elsewhere, the good availability of A1 and A2 stations already provides high confidence in flow data such that there is no need for additional hydrological modelling. The results of the CFRAM Study rating review for Station 36150 did not prove significant in terms of change to the original gauged AMAX series.

The calibration of the hydraulic models to historic flood data and observed evidence will further help to screen out design flow estimates which are not reflective of the actual behaviour of these sub- catchments.

There are many potential future changes to the catchment, margins of error and uncertainties which must be considered within the study. However the cumulative application of worst case scenarios, one on top of the other could lead to erroneous flood extents which do not take into account the diminishing cumulative joint probability of these factors. For this reason this report has separated future UoM 36 changes that have a high degree of certainty in the projections from those changes which are less certain. Future changes which have a high degree of uncertainty, along with margins of error and other uncertainties have been risk assessed individually. This risk assessment is to be taken forward and built upon through the hydraulic modelling phase with the ultimate goal of providing a single error margin for the flood extent maps on an AFA by AFA basis. This rationalised single error margin is designed to inform end users in a practical way as to the varying degree of caution to which mapped flood extents are to be treated.

9.1 SUMMARY OF THE RESULTS AND GENERAL PATTERNS

UoM 36 can be characterised hydrologically as follows:

IBE0700Rp0008 125 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

 The catchment has a wide range of climatic and physiographic characteristics. The drier, lowland areas in the Cavan River floodplain have SAAR values as low as 895 mm and as low as 900mm in the east of UoM 36, while catchments in the upland areas of Donegal and Leitrim have SAAR values in excess of 1400mm.

 Hydrometric data is of good quality and availability for larger channels but is not available for many smaller modelled tributaries.

 Meteorological data is of good availability in the catchment.

 Flood behaviour when defined in terms of the growth curve, i.e. in orders of magnitude greater than the median event, generally more extreme in the upper catchment than would have been thought based on older methodologies (FSR) although there was a wide variance in pooled frequency analysis for small to midsized catchments (10 to 200km2) with some catchments displaying flatter growth curve behaviour than the regional FSR curve.

 The 1% AEP flood event ranges from approximately 1.7 to 3 times larger than the median flood flow. This compares to approximately 2 under FSR.

Design flow estimation is the primary output of this study and has been developed based on the analysis contained in this report. This analysis is based on quality assessed observed data and the latest Irish catchment flood hydrology techniques. This analysis will require further validation through the calibration of the hydraulic models. As modelling progresses there may be some elements of the hydrological analysis that might need to be questioned and interrogated further. This is reflective of best practice in hydrology / hydraulic modelling for flood risk assessment. RPS believe that through the use of best practice statistical methods that the design flow estimation has as high a degree of certainty as is possible prior to calibration / validation and that this will save time and increase accuracy as UoM 36 moves into the hydraulic modelling phase of the CFRAM Study process. Nevertheless the modelling may necessitate the adjustment of some of the design flows and as such any adjustments made will be summarised within the Hydraulic Modelling Report.

9.2 RISKS IDENTIFIED

The main potential source of uncertainty in the analysis is due to the lack of hydrometric gauge data in the majority of smaller catchments. In addition, cross-border catchment areas and associated catchment descriptors within the existing FSU database were found not to represent the Northern Ireland portions, proving a significant risk within the Lough Melvin catchment. Other cross border catchment areas downstream of Ballyconnell were also found not to be represented accurately but this was generally found to be a smaller area of the catchment and downstream of any AFAs and as such has not been deemed a significant risk to the Study.

IBE0700Rp0008 126 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

Following this cycle of the North Western – Neagh Bann CFRAM Study the main potential adverse impact on the hydrological performance of the catchments is the effect of future changes and in particular the scope for rapid urbanisation of towns. Further rapid urbanisation of the tributary catchments around towns such as Cavan could significantly increase flood risk if this leads to development which is unsustainable from a drainage perspective.

9.3 OPPORTUNITIES / RECOMMENDATIONS

The lack of available hydrometric data on smaller catchments for use in the study highlights potential opportunities to improve the hydrological analysis further in the next cycle of the North Western – Neagh Bann CFRAM Study:

1. All of the models within UoM 36 have gauged data to inform the design flow estimation on the main channels however many of the small tributary watercourses / sub-catchments are hydrologically quite different to the main channels which within UoM 36 are heavily attenuated due to lake / canalised sections. Any of these ungauged sub-catchments would obviously benefit from the addition of a hydrometric gauge.

Recommending that new gauging stations are installed on the ungauged watercourses affecting the AFAs is unrealistic within the timeframe of this or even the next CFRAM Study cycle. Multiplied up nationally this would lead to a long list of gauging stations which would likely remain unrealised at a time when many organisations are rationalising their existing networks and may even obscure the case for those gauging stations which are more acutely needed. A more focussed exercise to identify the most acutely needed gauging stations would be more effectively undertaken following hydraulic modelling and consultation such that the AFAs which are at greatest risk, are most affected by uncertainty in the design flow estimates and which would significantly benefit from additional calibration data are identified as priorities. As such it is recommended that this exercise is undertaken following the hydraulic modelling stage.

In the interim improvements to the existing hydrometric gauge network should focus on improving the ratings through the collection of additional spot flow gaugings at flood flows at the existing stations on, directly upstream or downstream of AFAs:

 35029 Four Masters Bridge (OPW)  35071 Lareen (OPW)  36016 Rathkenny (OPW)  36021 Kiltybardan (OPW)  36027 Bellaheady (OPW)  36028 Aghoo (OPW)  36031 Lisdarn (EPA)  36037 Urney Bridge (OPW)

IBE0700Rp0008 127 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

 36150 Shantonagh Bridge (EPA)

It is assumed that the gauging stations within UoM 36 which currently have a rating of A1 will be maintained to that standard into the foreseeable future.

Furthermore there is a shortage nationally of very small and / or heavily urbanised catchment gauge data and as such new gauging stations on this type of catchment, ideally within a CFRAM Study AFA, could be progressed immediately.

2. Observed rainfall data is generally of limited use for fluvial and coastal flood risk analysis. Its primary usage in this study is in developing catchment runoff models. The main rivers in UoM 36 are generally well gauged and as such any catchment runoff models would only be of benefit in relation to the smaller tributary watercourses affecting the AFAs. In order to achieve the accuracy required this data must be of high temporal resolution (hourly) but must also be accompanied by some calibration (flow) data which is not readily available for these smaller watercourses. Rainfall data may also be useful in relation to hydraulic model calibration for estimating rainfall event return period which can be linked to flood return period. Rainfall data may also be integral to the operation of any flood forecasting system but a recommendation to provide additional infrastructure on that basis would come as a result of risk assessment and optioneering process.

Hourly rainfall data is currently being collected within HA 36 at the Ballyhaise Met Éireann station and just across the border at the UK Met Office station at Derrylin just north of Ballyconnell. Further to this the AFAs within HA36 are also within the range of the rainfall radars at Castor Bay and Dublin Airport whereby radar data following processing could provide additional high resolution rainfall data. Only the Bundoran & Tullaghan AFA in HA 35 could be considered to be totally un-served in terms of hourly rainfall data. Hourly data may be available at the Met Éireann station at Finner but was not made available for this study. This information if it became available could aid the estimation of flood frequency in relation to the smaller tributaries affecting the Bundoran & Tullaghan AFA. It may also provide data for rainfall runoff modelling of these same sub-catchments but as discussed without some observed flow data to calibrate to these would be of limited use. As such the only recommendation at this stage in relation to rainfall data is that further investigation into the data availability from the Finner station is undertaken such that hourly data may be made available for future cycles.

3. The delineation of cross-border catchments and derivation of associated FSU physical catchment descriptors should be reviewed to ensure potential errors in the data for catchments emanating from Northern Ireland is amended for future cycles.

IBE0700Rp0008 128 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

10 REFERENCES:

1. EC Directive on the Assessment and Management of Flood Risks (2007/60/EC)

2. S. Ahilan, J.J. O’Sullivan and M. Bruen (2012): Influences on flood frequency distributions in Irish river catchments. Hydrological Science Journal, Vol. 16, 1137-1150, 2012.

3. J.R.M. Hosking and J.R.W. Wallis (1997): Regional Frequency Analysis – An approach based on L-Moments. Cambridge University Press.

4. Flood Studies Update Programme – Work Package 2.1 – Review of Flood Flow Ratings for Flood Studies Update – Prepared by Hydrologic Ltd. for Office of Public Works (March 2006)

5. Flood Studies Update Programme – Work Package 2.2 – “Frequency Analysis” – Final Report – Prepared by the Department of Engineering Hydrology of National University of Ireland, Galway for Office of Public Works (September 2009).

6. Flood Studies Update Programme – Work Package 2.3 – Flood Estimation in Ungauged Catchments – Final Report – Prepared by Irish Climate Analysis and Research Units, Department of Geography, NUI Maynooth (June 2009)

7. Flood Studies Update Programme – Work Package 3.1 – Hydrograph Width Analysis – Final Report – Prepared by Department of Engineering Hydrology of National University of Ireland, Galway for Office of Public Works (September 2009)

8. Flood Studies Update Programme – Work Package 5.3 – Preparation of Digital Catchment Descriptors – Pre-Final Draft Report – Prepared by Compass Infomatics for Office of Public Works (January 2009)

9. Michael Bruen and Fasil Gebre (2005). An investigation of Flood Studies Report – Ungauged catchment method for Mid-Eastern Ireland and Dublin. Centre for Water Resources Research, University College Dublin.

10. North Western - Neagh Bann CFRAM Study – UoM01 Inception Report. Office of Public Works, 2012.

11. Flood Estimation Handbook- Statistical Procedures for Flood Frequency Estimation, Vol. 3. Institute of Hydrology, UK (1999).

12. NERC, 1975. Flood Studies Report. Natural Environment Research Council.

13. Institute of Hydrology Report No. 124 – Flood Estimation for Small Catchments (D.C.W. Marshall and A.C. Bayliss, June 1994)

IBE0700Rp0008 129 Rev F03 NW-NB CFRAM Study UoM 36 Hydrology Report – FINAL

14. Ireland in a Warmer World, Scientific Predictions of the Irish Climate in the Twenty First Century Prepared by Met Éireann and UCD (R. McGrath & P. Lynch, June 2008)

15. Growing for the Future – A Strategic Plan for the Development of the Forestry Sector in Ireland (Department for Agriculture, Food and Forestry, 1996)

16. Review of Impacts of rural land use management on flood generation (DEFRA, 2004)

17. Arterial Drainage Maintenance & High Risk Channel Designation, Draft Programme 2011 – 2015 (OPW Environment Section, 2011)

18. The Planning System and Flood Risk Management, Guidelines for Planning Authorities (OPW, 2009)

IBE0700Rp0008 130 Rev F03

APPENDIX A

UOM 36 HYDROMETRIC DATA STATUS TABLE

A1

APPENDIX B

RATING REVIEWS

BELLAHEADY (36027)

Gauge 36027 is located in County Cavan on the Woodford River Ballyconnell Canal. The river reach runs from Ballinamore AFA north east until it passes through Ballyconnell AFA and eventually joins upper Lough Erne.

Ballyconnell AFA

Ballinamore AFA Gauge Station 36027

Woodford River Ballyconnell Canal

Figure 1: Location of the Bellaheady Gauging Station

B1

There are no modelled tributaries which join the Woodford River / Ballyconnell Canal upstream of the Gauging Station. However the Lecharrownahone and the River join the channel within the Ballyconnell AFA. The confluence of the Lecharrownahone is approximately 4.4km downstream and the River Doon is approximately 4.8km downstream of the gauge station. The gauge is located on the upstream face of a road bridge. The river channel (bank to bank) is approximately 20m wide with the lowest bed level of 46.36m OD Malin. This section of the Woodford River Ballyconnell Canal was modelled as 1D only. Therefore the river banks were extended to the extent of the surveyed data. The cross section has a bank level of 51.94m OD Malin for left bank and 53.29m OD Malin for right bank, the width of the surveyed cross section at the gauging station is 101m. As stated in the 'Station gauge datum report' from OPW the gauge zero is currently 50.855m OD Poolbeg since 22/10/92 which has been converted to 48.08m OD Malin. The staff gauge zero level has been surveyed as part of this study at 48.09m OD Malin and this value has been used as the basis for this review. The gauge has continuous water level and flow records from 1974 - 2009, however continuous flow records only exist up to 1992 following the canalisation of the Woodford River and the abandonment of the rating by OPW.

B2

Figure 2: Model cross-section upstream of the bridge at gauge location (Top); Photo of gauge location (Bottom)

Figure 2 shows the gauge station cross section and bridge cross section along with a photograph of the gauge located on the upstream face of the bridge. The study reach extends approximately 13.4km upstream and 18km downstream of the gauge. There are 2 bridge structures upstream of the gauge and 6 bridges and 2 weirs downstream. The river also has 3 millraces along the modelled extents. The Woodford River / Ballyconnell Canal generally meanders but the immediate upstream and downstream approaches to the gauge are relatively straight. The one dimensional hydraulic model of the reach uses information from approximately 160 cross sections. The downstream boundary condition applied to the model was calculated as the critical flow Q-h relationship with the upstream boundary consisting

B3

of a number of different inflow hydrographs combining to give a peak flow of 62 m3/s, equivalent to an estimated 0.1% AEP event.

The Woodford River / Ballyconnell Canal was canalised in the late 19th century and restored in the 1990s. OPW do not currently have a rating for Bellaheady and it is stated on the OPW Hydro-Data website that the gauge provides 'Poor quality low flow data - to be used for indicative purposes only’. The river was canalised from 01/01/92.' The National Review under FSU Work Package 2.1 assigned data recorded from the gauge a quality classification of A2 (i.e. confidence in the rating up to approximately 1.3 x Qmed) but this refers only to the period be prior to restoration of the canal. Since the canalisation of the river in 1992 there is only one spot gauged flow available in 2009 and OPW have discontinued the use of the rating to derive flow values.

The model was calibrated to the single spot gauged flow. No calibration could be carried out by comparing model outputs with OPW rating curve information as there is no current rating. Adjustments were made to the Manning’s n values for channel and over bank roughness to reflect vegetation growth and channel roughness in order to develop a realistic model of the channel and flow conditions. The best fit rating curve was achieved with a Manning's n value of 0.04. Analysis of the results shows that water levels remain in bank up to the estimated 0.1% AEP flow.

The results of the rating review are shown in Figure 2 and Table 1. The spot gauging, RPS modelled output and RPS rating curve are plotted.

Figure 3: RPS Rating Curve

B4

Equation Min Stage (m) Max Stage (m) C a b 1 1.000 2.000 5.168 1.129 1.760 2 2.000 2.500 12.117 0.285 1.403 3 2.500 2.798 4.051 0.802 2.114 Where: Q = C(h+a)b and h = stage readings (m)

Table 1: Rating equation values for Bellaheady Gauge 36027

Figure 2 shows that the RPS rating is calibrated to within 0.09m of the only available spot gauging since the previous rating was discontinued. This is within the tolerance for a high priority watercourse 3 3 3 which is 0.2m. The spot gauging flow is 42.5m /s, 18.12 m /s greater than Qmed at 24.38 m /s.

There is a low level of confidence in the lower portion of the Q-h relationship where instability in the Q- h relationship was evident from the model. This is due to the tributaries flowing into the Woodford River / Ballyconnell Canal and, due to the flat topography of the channel, back flow affects the rating at the gauging station. This effect is not evident at higher flows as tributary flows are dwarfed by the main channel hydrograph but does suggest uncertainty in any rating at lower flows (below 20m3/s).The rating review curve is derived using 0.1% AEP event hydrographs entered at the relevant locations upstream of the gauging station to mimic actual flow conditions and flood hydrograph timing.

The rating equation which has been processed from the modelled Q – h is to be treated with caution for two reasons:

1. The rating has been calibrated based on only one spot flow gauging

2. There is evidence of instability in the rating curve below 20m3/s due to back flow from downstream tributaries.

B5

LISDARN (36031)

The gauging station at Lisdarn (36031) is located on the Cavan River 1.5km North of Cavan town, County Cavan approximately 150m upstream of its confluence with the Cullies River. The staff gauge is located on the right hand bank of a open section directly upstream of a single arch bridge. The channel is approximately 12.8m wide with a minimum bed level of approximately 51.8 OD Malin and bank levels of 54.8m OD Malin (Right bank) and 54.4m OD Malin (Left bank). The staff gauge zero ordnance level used by EPA is currently 51.929m OD Malin. The staff gauge zero level was surveyed to be 51.983m OD Malin which is a difference of 0.054m. The OPW staff gauge zero level has been used as the basis for this review to ensure consistency with the existing ratings and spot gaugings.

Figure 1: Location of the Lisdarn Gauging Station

The gauge is operated by the EPA, records start from 07/11/74 and there have been two reviews of the rating for the site. The current rating has been applied since 14/12/87 when there was a shift in the staff gauge zero and is made up of 4 equations. Therefore only the 92 spot gauges with water levels and flows recorded since 14/12/87 are applicable to this rating review. The largest relevant spot 3 th 3 gauging is 7.3462m /s which was recorded on the 29 Nov 1995. Qmed for this site is 6.45m /s.

B6

Figure 2: Model Cross-Section at Gauge Location (Top); Photo of Gauge Location (Bottom)

The rating review model extends approximately 8.5km upstream of the gauge station on the Cavan River and 6km downstream where the reach joins the Dromore River. There are six tributaries which join the Cavan River upstream of the gauge and there is also a millrace approximately 1.3km upstream of the gauge location. There are 29 bridge structures in total along the modelled extent of the Cavan River. 26 are upstream of the gauge, the closest of which is approximately 315m. 1 bridge is at the gauge location and 2 structures are downstream of the gauge the closest of which is approximately 810m away. There are also 2 weirs in the river both upstream of the gauge. The one dimensional hydraulic model uses information from 259 cross sections for the Cavan River. The downstream boundary condition applied to the model was calculated as the critical flow Q-h relationship, with a number of inflow hydrographs added upstream of the gauging station in order to replicate an estimated 0.1% AEP event at the gauging station itself. Manning's n values were adjusted to describe the

B7

channel and flood plain roughness to replicate vegetation growth and produce a realistic model of the flow conditions.

The gauging station was given a rating classification of A2 under FSU Work Package 2.1 (2004) but additional high flow spot gaugings were recommended to increase confidence. An A2 classification indicates that there is confidence in the rating up to around 1.3 times the Qmed flow. The highest flow 3 3 spot gauging is 7.3462m /s which is above the Qmed value for the station which is 6.45m /s.

The results of the rating review are shown below in Figure 3 and Table 1. The graph demonstrates the modelled Q-H relationship and shows the comparison with the OPW rating curve and spot gaugings.

Figure 3: Comparison of Existing OPW Rating Curve and RPS Rating Curve for all flows

B8

Table 1: Rating equation values for gauge 36031

Section Min Stage (m) Max Stage (m) C a b

1 0.256 0.385 59.872 0 5.755

2 0.385 0.800 5.171 0 3.193

3 0.800 1.510 3.617 0 1.587

4 1.510 1.531 1.486 0 3.757

5 1.531 2.499 6.8141 -0.4652 1.399 Where: Q = C(h+a)b and h = stage readings (metres)

Note: Sections 1 to 4 are existing OPW rating curve segments which have been retained

Figure 3 shows that the modelled curve is not capturing low flow behaviour but the curve is calibrated very well to the medium to higher spot gaugings. Regarding the low flow differences, it would not be beneficial to the study to undertake additional survey to capture the low flow behaviour as it has negligible effect on the rating at flood flow.

Figure 3 shows that the modelled curve accurately represents equation 3 of the current rating curve from 2.5m3/s up to 6.96m3/s. The modelled curve also goes through the highest spot gauging. The full extent of the first three rating equations have been maintained from the EPA curve. The fourth equation has been maintained to a stage height of1.531m, the intersection point with the modelled rating curve extension. The extent of the retained EPA equations are plotted in Figure 3 followed by the derived RPS equation 5.

This rating was carried out using the MIKE FLOOD 1D/2D modelling package. A Manning’s n value of 0.065 was applied to the cross-section for the best fit rating curve. This Manning's value describes a main channel reach which is sluggish, weedy and with deep pools which although towards the upper end is a fair description of this reach of the Cavan River. Floodplain roughness was set to 0.034 which describes high grass pastures or mature row crops. The use of these roughness values resulted in the best fit rating curve.

B9

SHANTONAGH BRIDGE (36150)

Gauge 36150 is located in County Monaghan on the Shantonagh River. The river reach runs into Ballybay town from the North West and joins the Dromore River. The Cornamucklaglass is a tributary of the Shantonagh; its confluence is approximately 2km downstream of the gauge station. The gauge is located 5m upstream of Shantonagh Bridge off the R183 road in the vicinity of Derryvalley Presbyterian Church.

Figure 1: Location of the Shantonagh Bridge Gauging Station

The river cross section is approximately 13.5m wide with the lowest bed level of 80.82m OD Malin and bank levels of 82.54m OD Malin for left bank, 84.94m OD Malin for right bank. As stated on the HydroNet website the gauge zero ordnance level used by EPA is currently 81.186m. The staff gauge zero level has been surveyed as part of this study at 81.189m OD Malin and this value has been used as the basis for this review.

B10

The gauge is managed by the EPA and is currently active. Continuous water level and derived flow records have been provided from June 1997 to February 2012. Data recorded before 1997 was not applicable as a weir was constructed in the central arch of the Shantonagh Bridge to help control low flow.

Figure 2: Model Cross-Section at Gauge Location (Top); Photo of gauge location (Bottom)

The rating review reach extends approximately 900m upstream of the gauge and 3km downstream where the reach joins the Dromore River. There are 2 bridge structures located upstream of the gauge within the modelled reach approximately 345m and 420m upstream. There are an additional 7 bridge structures between the gauge and where the Shantonagh joins the Dromore River. The closest of these is 60m downstream. The upstream and downstream approaches to the gauge are relatively straight. There are 92 cross sections included in the 1D hydraulic model for the Shantonagh reach.

B11

The upstream boundary input was set with a hydrograph with a peak flow of 54.69 m3/s equivalent to an estimated 0.1% AEP event peak flow but with the flood event base flow removed (hydrograph rising and falling to 0m3/s) such that low flow water levels were comparable to the EPA Shantonagh Bridge ratings. Manning's n valves were adjusted to describe the channel and flood plain roughness to replicate vegetation growth and produce a realistic model of the flow conditions.

The EPA have described the rating standard at Shantonagh Bridge as 'good' on the HydroNet website. The National Review under FSU Work Package 2.1 did not rate the Shantonagh Bridge gauge with a 3 quality classification. The Qmed value from EPA for the station is 10.9 m /s but the highest flow spot 3 gauging is 4.77 m /s and therefore there is no confidence in the Qmed value based on the current rating. For the purposes of the rating review the EPA rating equation is considered valid up to the level of the highest spot gauging of 0.675m. The model and survey do not suggest that there is potential for flow to bypass the gauge and immediate floodplain.

The results of the rating review are shown below in Figure 3 and Table 1. The graph demonstrates the derived RPS rating curve and shows the comparison between the EPA rating curve (which consists of two equations) and spot gaugings. The first two equations on the RPS curve have been taken from the existing EPA rating curve.

Modelled Rating

Figure 3: Comparison of Existing OPW Rating Curve and RPS Modelled Rating for all flows

B12

Min Stage Max Stage Section C a b (m) (m)

1 0 0.535 18.1896 0 2.9061 2 0.535 0.675 10.4718 0 2.0231 3 0.675 0.991 13.9461 -0.2048 1.4608 4 0.991 1.135 8.8247 0.0705 1.7 5 1.135 1.605 17.9574 -0.3761 1.4 6 1.605 1.756 33.3627 -0.8130 1.4 7 1.756 2.141 45.7043 -1.0002 1.4 Where: Q = C(h+a)b and h = stage readings (metres)

Table 1: Rating equation values for gauge 36150

Figure 3 shows that the model accurately represents the rating curve based on the lower flow gaugings up to the last gauging at 4.77m3/s. The curve continues to follow the EPA rating curve up to a flow of 7m3/s where the curves begin to diverge slightly. A Manning's ‘n’ value of 0.05 was applied to the cross section which resulted in the best fit rating curve. The results show that the floodwaters 3 break the banks at approximately 10m /s, which is slightly less than Qmed.

The weir constructed in the central arch of Shantonagh Bridge has had significant hydraulic influence on the Q-h relationship at the gauge station location. To reproduce the EPA Q-h relationship up to it's limit of reliability the model has been modified to represent the weir. Cross sections were surveyed at the upstream and downstream faces of the bridge and did not capture the high point of the weir within the middle arch. However the surveyor did survey a long section through this arch and as such the high point of the weir is known. In order to replicate the effect of the weir the downstream invert level was raised to the level of the weir invert to capture this low flow control point. This resulted in raising the invert of the section by 319mm to an invert of 81.140m OD Malin.

B13

APPENDIX C

NAM OUTPUT

C1

C2

APPENDIX D

DESIGN FLOWS FOR MODELLING INPUT

Model 1 - Ballinamore Flows for AEP AREA Model Node ID_CFRAMS 2 Qmed 50% 10% 5% 1% 0.5% 0.1% (km ) 20% (5yr) 2% (50yr) number (2yr) (10yr) (20yr) (100yr) (200yr) (1000yr) 36_756_1_RA 13.32 3.87 3.87 4.91 5.67 6.49 7.72 8.80 10.04 13.69 Model 1 36_1625_2_RA 3.92 0.37 0.37 0.51 0.61 0.73 0.90 1.06 1.25 1.80 Model 1 36_1625_4_RARPS 4.50 0.44 0.44 0.61 0.74 0.87 1.09 1.28 1.50 2.16 Model 1 Top-up flow between 36_1625_2_RA & 0.58 0.06 0.06 0.09 0.11 0.13 0.16 0.19 0.22 0.32 Model 1 3_1625_4_RARPS 36034_RA 20.22 3.25 3.25 4.08 4.66 5.26 6.14 6.90 7.73 10.07 Model 1 Top-up flow between 36_756_1_RA & 2.40 0.44 0.44 0.55 0.63 0.71 0.83 0.94 1.05 1.37 Model 1 36034_RA

36_2098_2_RA 133.48 16.18 16.18 19.54 21.71 23.86 26.82 29.18 31.67 38.11 Model 1 36_2274_2_RA 138.61 16.38 16.38 19.61 21.67 23.72 26.51 28.72 31.05 36.99 Model 1 Top-up flow between 36_2098_2_RA & 5.13 0.75 0.75 0.89 0.99 1.08 1.21 1.31 1.41 1.68 Model 1 36_2274_2_RA 36_2274_3_RARPS 138.65 18.59 18.59 22.25 24.59 26.91 30.07 32.58 35.22 41.97 Model 1 Top-up flow between 36_2274_2_RA & 0.05 0.01 0.01 0.01 0.01 0.02 0.02 0.02 0.02 0.02 Model 1 36_2274_3_RA 36_2273_2_RARPS 0.08 0.00 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.02 Model 1 36_2275_1_RA 159.48 19.27 19.27 23.11 25.50 27.83 30.93 33.36 35.86 42.11 Model 1 Top-up flow between 36_2275_1_RA & 0.53 0.09 0.09 0.11 0.12 0.13 0.15 0.16 0.17 0.20 Model 1 36034_RA

36_2275_2_RA 159.65 19.02 19.02 22.80 25.16 27.46 30.52 32.92 35.39 41.55 Model 1 Top-up flow between 36_2275_2_RARPS & 0.17 0.03 0.03 0.04 0.04 0.05 0.05 0.05 0.06 0.07 Model 1 36_2275_1_RA

D1

36_2275_3_RA 159.82 18.69 18.69 22.41 24.73 26.99 30.00 32.35 34.78 40.84 Model 1 Top-up flow between 36_2275_3_RA & 0.17 0.03 0.03 0.04 0.04 0.04 0.05 0.05 0.06 0.07 Model 1 36_2275_2_RA 36028_RA 166.85 19.34 19.34 23.42 25.97 28.45 31.76 34.35 37.03 43.73 Model 1 Top-up flow between 36028_RA & 7.03 1.00 1.00 1.21 1.34 1.46 1.63 1.77 1.91 2.25 Model 1 36_2275_3_RA 36_2082_2_RA 179.36 20.03 20.03 24.30 26.99 29.61 33.16 35.94 38.87 46.18 Model 1 Top-up flow between 36_2082_2_RA & 12.51 1.65 1.65 2.00 2.23 2.44 2.73 2.96 3.21 3.81 Model 1 36028_RA 36091_RA 199.82 14.84 14.84 18.00 19.97 21.89 24.44 26.44 28.53 33.70 Model 1 Top-up flow between 36091_RA & 20.46 1.75 1.75 2.13 2.36 2.59 2.89 3.13 3.37 3.98 Model 1 36_2082_2_RA

MRFS Flows for AEP HEFS Flows for AEP AREA Model Node ID_CFRAMS (km2) 50% 20% 10% 5% 2% 1% 0.5% 0.1% 10% 1% 0.1% number (2) (5) (10) (20) (50) (100) (200) (1000) (10) (100) (1000) 36_756_1_RA 13.32 4.83 6.12 7.07 8.09 9.63 10.98 12.53 17.08 9.56 14.84 23.08 Model 1 36_1625_2_RA 3.92 0.46 0.63 0.76 0.91 1.13 1.33 1.56 2.25 0.94 1.63 2.76 Model 1 36_1625_4_RARPS 4.50 0.55 0.76 0.92 1.09 1.35 1.59 1.87 2.70 1.13 1.95 3.31 Model 1 Top-up flow between 36_1625_2_RA & 0.58 0.08 0.11 0.13 0.16 0.20 0.23 0.27 0.40 0.17 0.29 0.49 Model 1 3_1625_4_RARPS 36034_RA 20.22 4.02 5.04 5.75 6.50 7.59 8.51 9.54 12.43 7.06 10.46 15.27 Model 1 Top-up flow between 2.40 0.54 0.68 0.78 0.88 1.03 1.15 1.29 1.69 0.96 1.42 2.07 Model 1 36_756_1_RA & 36034_RA

36_2098_2_RA 133.48 20.19 24.38 27.09 29.77 33.47 36.41 39.52 47.56 33.28 44.74 58.43 Model 1 36_2274_2_RA 138.61 20.44 24.46 27.04 29.59 33.07 35.83 38.73 46.15 33.22 44.01 56.69 Model 1 Top-up flow between 36_2098_2_RA & 5.13 0.93 1.11 1.23 1.35 1.51 1.63 1.76 2.10 1.51 2.00 2.58 Model 1 36_2274_2_RA

D2

36_2274_3_RARPS 138.65 23.16 27.73 30.65 33.54 37.48 40.61 43.90 52.30 37.65 49.89 64.26 Model 1 Top-up flow between 36_2274_2_RA & 0.05 0.01 0.02 0.02 0.02 0.02 0.02 0.03 0.03 0.02 0.03 0.04 Model 1 36_2274_3_RA 36_2273_2_RARPS 0.08 0.01 0.01 0.02 0.02 0.03 0.03 0.04 0.05 0.02 0.03 0.06 Model 1 36_2275_1_RA 159.48 23.97 28.74 31.71 34.62 38.47 41.50 44.61 52.38 38.96 50.98 64.35 Model 1 Top-up flow between 0.53 0.11 0.14 0.15 0.17 0.18 0.20 0.21 0.25 0.19 0.24 0.31 Model 1 36_2275_1_RA & 36034_RA

36_2275_2_RA 159.65 23.65 28.36 31.29 34.16 37.96 40.94 44.02 51.68 38.45 50.30 63.50 Model 1 Top-up flow between 36_2275_2_RARPS & 0.17 0.04 0.05 0.05 0.06 0.06 0.07 0.07 0.09 0.06 0.08 0.11 Model 1 36_2275_1_RA 36_2275_3_RA 159.82 23.24 27.87 30.75 33.56 37.31 40.24 43.26 50.79 37.78 49.43 62.40 Model 1 Top-up flow between 36_2275_3_RA & 0.17 0.04 0.04 0.05 0.05 0.06 0.06 0.07 0.08 0.06 0.08 0.10 Model 1 36_2275_2_RA 36028_RA 166.85 24.05 29.13 32.30 35.38 39.50 42.72 46.06 54.38 39.69 52.48 66.82 Model 1 Top-up flow between 7.03 1.24 1.50 1.66 1.82 2.03 2.20 2.37 2.80 2.04 2.70 3.44 Model 1 36028_RA & 36_2275_3_RA 36_2082_2_RA 179.36 24.93 30.23 33.57 36.84 41.25 44.72 48.35 57.45 41.25 54.94 70.59 Model 1 Top-up flow between 12.51 2.06 2.49 2.77 3.04 3.40 3.69 3.99 4.74 3.40 4.53 5.82 Model 1 36_2082_2_RA & 36028_RA 36091_RA 199.82 18.47 22.40 24.86 27.24 30.41 32.91 35.51 41.94 30.54 40.43 51.52 Model 1 Top-up flow between 20.46 2.18 2.65 2.94 3.22 3.60 3.89 4.20 4.96 3.61 4.78 6.09 Model 1 36091_RA & 36_2082_2_RA

Input flows Top‐up flows. These flows should be entered laterally Check flows. Modellers should check to make sure these flows are being reached at each HEP Some of these flows may be put in at the US point due to a small difference between US & DS flows.

D3

Model 2 - Ballyconnell Flows for AEP AREA Model Node ID_CFRAMS 2 Qmed 0.1% (km ) 50% (2) 20% (5) 10% (10) 5% (20) 2% (50) 1% (100) 0.5% (200) number (1000) 36091_RA 199.82 14.84 14.84 18.00 19.97 21.89 24.44 26.44 28.53 33.70 Model 2 36_2534_3_RARPS 101.92 21.80 21.80 27.01 30.65 34.51 40.11 44.90 50.26 65.35 Model 2 36_1511_4_RA 3.87 0.67 0.67 0.92 1.11 1.32 1.64 1.93 2.27 3.27 Model 2 36_527_8_RA 5.88 1.38 1.38 1.90 2.30 2.73 3.39 3.99 4.68 6.75 Model 2 36027_RA 330.22 24.38 24.38 29.84 33.52 37.30 42.67 47.10 51.95 65.05 Model 2 Top-up flow between 36091_RA & 18.74 1.66 1.66 2.03 2.28 2.54 2.90 3.20 3.53 4.42 Model 2 36027_RA 36_656_5_RA 8.65 5.75 5.75 7.95 9.59 11.39 14.15 16.63 19.50 28.18 Model 2 36_1576_1_RARPS 0.91 0.84 0.84 1.17 1.41 1.67 2.08 2.44 2.86 4.14 Model 2 36_2379_1_RARPS 1.13 0.85 0.85 1.17 1.41 1.67 2.08 2.45 2.87 4.14 Model 2 36_2379_2_RARPS 2.74 1.97 1.97 2.72 3.29 3.90 4.85 5.70 6.69 9.66 Model 2 Top-up flow between 36_2379_1_RARPS & 1.60 1.20 1.20 1.65 1.99 2.37 2.94 3.46 4.05 5.86 Model 2 36_2379_2_RARPS 36_1415_4_RARPS 4.50 3.10 3.10 4.28 5.17 6.14 7.63 8.97 10.52 15.19 Model 2 Top-up flow between 36_1576_1_RARPS & 0.63 0.49 0.49 0.68 0.82 0.98 1.22 1.43 1.68 2.42 Model 2 36_1415_4_RARPS 36_1285_3_RARPS 13.53 7.38 7.38 9.32 10.75 12.29 14.63 16.69 19.07 26.09 Model 2 Top-up flow between 36_656_5_RA & 0.38 0.26 0.26 0.32 0.37 0.43 0.51 0.58 0.66 0.91 Model 2 36_1285_3_RARPS 36_1834_1_RA 1.07 0.92 0.92 1.27 1.54 1.82 2.27 2.66 3.12 4.51 Model 2 36_1834_3_RA 2.52 1.65 1.65 2.28 2.75 3.27 4.07 4.78 5.60 8.10 Model 2 Top-up flow between 36_1834_1_RA & 1.45 0.98 0.98 1.36 1.64 1.95 2.42 2.84 3.34 4.82 Model 2 36_1834_3_RA

D4

36_2589_2_RA 355.46 26.56 26.56 32.51 36.52 40.63 46.48 51.31 56.59 70.86 Model 2 Top-up flow between 36027_RA_RA & 9.19 0.86 0.86 1.06 1.19 1.32 1.51 1.67 1.84 2.31 Model 2 36_2589_2_RA 36_10003_RARPS 369.89 27.56 27.56 33.74 37.90 42.17 48.24 53.26 58.74 73.54 Model 2 Top-up flow between 36_2589_2_RA & 14.43 1.32 1.32 1.61 1.81 2.02 2.31 2.55 2.81 3.52 Model 2 36_10003_RARPS 36_10002_1_RA 394.66 29.29 29.29 35.85 40.28 44.82 51.26 56.59 62.42 78.15 Model 2 Top-up flow between 36_10003_RARPS & 24.78 2.19 2.19 2.68 3.01 3.35 3.83 4.23 4.67 5.84 Model 2 36_10002_1_RA 36_2565_2_RA 55.16 2.32 2.32 2.85 3.22 3.60 4.15 4.61 5.11 6.50 Model 2 36_10002_D_RA 454.56 33.44 33.44 40.83 45.84 50.96 58.25 64.27 70.86 88.64 Model 2 Top-up flow between 36_10002_1_RA & 4.74 0.46 0.46 0.57 0.64 0.71 0.81 0.89 0.98 1.23 Model 2 36_10002_D_RA

MRFS Flows for AEP HEFS Flows for AEP AREA Model Node ID_CFRAMS (km2) 50% 20% 10% 5% 2% 1% 0.5% 0.1% 10% 1% 0.1% number (2) (5) (10) (20) (50) (100) (200) (1000) (10) (100) (1000) 36091_RA 199.82 18.47 22.40 24.86 27.24 30.41 32.91 35.51 41.94 30.54 40.43 51.52 Model 2 36_2534_3_RARPS 101.92 27.20 33.70 38.24 43.06 50.05 56.03 62.72 81.54 46.98 68.84 100.18 Model 2 36_1511_4_RA 3.87 0.83 1.15 1.39 1.65 2.05 2.41 2.83 4.09 1.71 2.96 5.02 Model 2 36_527_8_RA 5.88 1.72 2.38 2.87 3.41 4.24 4.98 5.84 8.43 3.52 6.11 10.36 Model 2 36027_RA 330.22 29.81 36.49 40.99 45.62 52.17 57.60 63.53 79.54 47.44 66.66 92.06 Model 2 Top-up flow between 36091_RA & 36027_RA 18.74 2.03 2.48 2.79 3.10 3.55 3.92 4.32 5.41 3.49 4.90 6.77 Model 2 36_656_5_RA 8.65 7.04 9.73 11.73 13.94 17.33 20.35 23.88 34.50 14.32 24.83 42.09 Model 2 36_1576_1_RARPS 0.91 1.03 1.43 1.72 2.05 2.54 2.99 3.51 5.07 2.10 3.65 6.18 Model 2 36_2379_1_RARPS 1.13 1.06 1.46 1.76 2.09 2.60 3.05 3.58 5.17 2.38 4.13 6.99 Model 2 36_2379_2_RARPS 2.74 2.46 3.40 4.10 4.87 6.06 7.12 8.35 12.06 5.04 8.74 14.82 Model 2 Top-up flow between 36_2379_1_RARPS & 1.60 1.49 2.06 2.49 2.95 3.67 4.31 5.06 7.31 3.05 5.30 8.98 Model 2

D5

36_2379_2_RARPS

36_1415_4_RARPS 4.50 3.87 5.35 6.45 7.66 9.53 11.19 13.13 18.97 7.93 13.75 23.30 Model 2 Top-up flow between 36_1576_1_RARPS & 36_1415_4_RARPS 0.63 0.61 0.84 1.01 1.20 1.49 1.75 2.05 2.97 1.17 2.03 3.43 Model 2 36_1285_3_RARPS 13.53 9.21 11.63 13.41 15.33 18.25 20.83 23.80 32.55 16.47 25.59 39.99 Model 2 Top-up flow between 36_656_5_RA & 36_1285_3_RARPS 0.38 0.31 0.40 0.46 0.52 0.62 0.71 0.81 1.11 0.57 0.89 1.39 Model 2 36_1834_1_RA 1.07 1.13 1.56 1.88 2.23 2.78 3.26 3.82 5.53 2.59 4.49 7.62 Model 2 36_1834_3_RA 2.52 2.10 2.91 3.51 4.16 5.18 6.08 7.13 10.31 5.56 9.64 16.33 Model 2 Top-up flow between 36_1834_1_RA & 36_1834_3_RA 1.45 2.08 2.87 3.46 4.11 5.11 6.00 7.04 10.17 2.57 4.46 7.55 Model 2 36_2589_2_RA 355.46 33.04 40.44 45.42 50.54 57.81 63.82 70.40 88.14 55.81 78.41 108.29 Model 2 Top-up flow between 36027_RA_RA & 36_2589_2_RA 9.19 1.12 1.37 1.53 1.71 1.95 2.15 2.38 2.98 1.82 2.55 3.52 Model 2 36_10003_RARPS 369.89 34.28 41.96 47.14 52.45 60.00 66.24 73.06 91.47 57.92 81.38 112.38 Model 2 Top-up flow between 36_2589_2_RA & 36_10003_RARPS 14.43 1.64 2.01 2.26 2.51 2.87 3.17 3.50 4.38 2.77 3.89 5.38 Model 2 36_10002_1_RA 394.66 36.43 44.59 50.09 55.74 63.75 70.38 77.63 97.20 61.54 86.47 119.42 Model 2 Top-up flow between 36_10003_RARPS & 36_10002_1_RA 24.78 2.68 3.27 3.68 4.09 4.68 5.17 5.70 7.14 4.26 5.98 8.26 Model 2 36_2565_2_RA 55.16 2.84 3.49 3.94 4.41 5.08 5.64 6.26 7.95 4.57 6.53 9.21 Model 2 36_10002_D_RA 454.56 40.86 49.89 56.01 62.27 71.17 78.53 86.58 108.31 64.82 90.88 125.35 Model 2 Top-up flow between 36_10002_1_RA & 36_10002_D_RA 4.74 0.58 0.71 0.79 0.88 1.01 1.11 1.22 1.53 0.90 1.26 1.74 Model 2

Input flows Top‐up flows. These flows should be entered laterally Check flows. Modellers should check to make sure these flows are being reached at each HEP

D6

Some of these flows may be put in at the US point due to a small difference between US & DS flows.

Model 3 - Ballybay Flows for AEP AREA Model Node ID_CFRAMS 2 Qmed 0.5% 0.1% (km ) 50% (2) 20% (5) 10% (10) 5% (20) 2% (50) 1% (100) number (200) (1000) 36_767_6_RA 47.57 9.63 9.63 12.51 14.56 16.74 20.00 22.80 25.97 35.06 Model 3 36024_RA 49.05 9.77 9.77 12.64 14.69 16.88 20.12 22.91 26.08 35.16 Model 3 Top-up flow between 36_767_6_RA & 1.47 0.37 0.37 0.47 0.55 0.63 0.75 0.86 0.98 1.32 Model 3 36024_RA 36074_RA 50.90 10.28 10.28 13.15 15.21 17.38 20.58 23.35 26.45 35.31 Model 3 Top-up flow between 36024_RA & 1.85 0.46 0.46 0.59 0.68 0.78 0.92 1.05 1.19 1.58 Model 3 36074_RA 36_10001_U_RARPS 0.04 0.01 0.01 0.02 0.02 0.03 0.03 0.04 0.05 0.07 Model 3 36_710_Trb_RARPS 0.36 0.11 0.11 0.15 0.18 0.21 0.26 0.31 0.36 0.52 Model 3 Top-up flow between 3610001_U_RARPS 0.32 0.09 0.09 0.13 0.16 0.19 0.23 0.27 0.32 0.46 Model 3 7 36_710_Trib_RARPS 36150 36.17 10.74 10.74 14.74 17.76 21.13 26.33 31.05 36.57 53.46 Model 3 36_30_4_RA 37.54 10.99 10.99 15.07 18.17 21.61 26.94 31.76 37.41 54.69 Model 3 36_10000_U_RARPS 0.93 0.42 0.42 0.58 0.70 0.84 1.04 1.22 1.43 2.07 Model 3 36_10000_RA 2.19 0.96 0.96 1.32 1.59 1.89 2.35 2.77 3.24 4.69 Model 3 Top-up flow between 36_10000_U_RARPS 1.26 0.56 0.56 0.78 0.94 1.11 1.38 1.63 1.91 2.76 Model 3 & 36_10000_RA 36_30_8_RA 40.65 11.39 11.39 15.63 18.84 22.40 27.93 32.93 38.78 56.70 Model 3 Top-up flow between 36_30_4_RA & 0.91 0.32 0.32 0.44 0.53 0.64 0.79 0.93 1.10 1.61 Model 3 36_30_8_RA 36_1691_3_RA 11.29 3.76 3.76 5.21 6.30 7.49 9.34 11.00 12.93 18.77 Model 3 36_1691_8_RA 14.39 4.40 4.40 6.10 7.38 8.79 10.95 12.89 15.15 22.00 Model 3

D7

Top-up between 36_1691_3_RA & 3.10 1.04 1.04 1.45 1.75 2.08 2.60 3.06 3.60 5.22 Model 3 36_1691_8_RA 36_2116_4_RA 9.39 2.27 2.27 3.14 3.78 4.49 5.59 6.56 7.70 11.12 Model 3 36030_RA 119.25 16.79 16.79 20.55 22.98 25.37 28.62 31.23 33.96 40.96 Model 3 Top-up flow between 36074_RA & 33.07 5.05 5.05 6.18 6.91 7.63 8.61 9.39 10.21 12.32 Model 3 36030_RA 36070_RARPS 126.51 13.78 13.78 16.87 18.87 20.82 23.50 25.63 27.88 33.63 Model 3 Top-up flow between 36030_RA & 7.26 1.21 1.21 1.48 1.65 1.82 2.06 2.25 2.44 2.95 Model 3 36070_RA 36_734_2_RA 9.24 1.59 1.59 2.19 2.65 3.14 3.91 4.59 5.38 7.78 Model 3 36153_RA 137.72 14.91 14.91 18.12 20.20 22.26 25.08 27.35 29.75 35.92 Model 3 Top-up flow between 36070_RARPS & 1.97 0.28 0.28 0.34 0.38 0.42 0.47 0.51 0.56 0.67 Model 3 36153_RA 36_1762_2_RA 27.10 3.66 3.66 5.00 5.99 7.05 8.65 10.06 11.68 16.44 Model 3 36_2308_4_RA 4.09 0.21 0.21 0.29 0.35 0.42 0.52 0.61 0.72 1.04 Model 3 36072_RA 191.19 14.16 14.16 17.02 18.80 20.53 22.84 24.64 26.51 31.14 Model 3 Top-up flow between 36153_RA & 22.28 1.89 1.89 2.27 2.51 2.74 3.05 3.29 3.54 4.15 Model 3 36072_RA 36_624_6_RA 24.55 3.33 3.33 4.60 5.55 6.58 8.15 9.55 11.17 15.99 Model 3 36018_RA 220.40 16.25 16.25 19.89 22.34 24.86 28.44 31.40 34.63 43.36 Model 3 Top-up flow between 36072_RA & 4.66 0.44 0.44 0.54 0.60 0.67 0.77 0.85 0.93 1.17 Model 3 36018_RA

MRFS Flows for AEP HEFS Flows for AEP AREA Model Node ID_CFRAMS 2 (km ) 50% 20% 10% 5% 2% 1% 0.5% 0.1% 10% 1% 0.1% number (2) (5) (10) (20) (50) (100) (200) (1000) (10) (100) (1000) 36_767_6_RA 47.57 12.03 15.63 18.19 20.92 24.98 28.49 32.44 43.80 22.35 35.00 53.81 Model 3 36024_RA 49.05 12.20 15.79 18.35 21.08 25.14 28.63 32.58 43.93 22.55 35.17 53.97 Model 3

D8

Top-up flow between 1.47 0.46 0.59 0.69 0.79 0.94 1.07 1.22 1.65 0.85 1.32 2.02 Model 3 36_767_6_RA & 36024_RA 36074_RA 50.90 12.84 16.43 19.00 21.71 25.72 29.17 33.05 44.12 23.35 35.84 54.20 Model 3 Top-up flow between 36024_RA 1.85 0.58 0.74 0.85 0.97 1.15 1.31 1.48 1.98 1.05 1.61 2.43 Model 3 & 36074_RA 36_10001_U_RARPS 0.04 0.02 0.02 0.03 0.03 0.04 0.05 0.06 0.09 0.04 0.07 0.12 Model 3 36_710_Trb_RARPS 0.36 0.13 0.18 0.21 0.25 0.31 0.37 0.43 0.62 0.30 0.52 0.87 Model 3 Top-up flow between 3610001_U_RARPS 7 0.32 0.12 0.16 0.20 0.23 0.29 0.34 0.40 0.58 0.26 0.46 0.78 Model 3 36_710_Trib_RARPS 36150 36.17 13.42 18.41 22.19 26.39 32.90 38.79 45.69 66.79 27.27 47.66 82.06 Model 3 36_30_4_RA 37.54 13.18 18.09 21.80 25.93 32.32 38.11 44.89 65.62 27.89 48.75 83.94 Model 3 36_10000_U_RARPS 0.93 0.53 0.73 0.88 1.05 1.30 1.53 1.79 2.59 1.19 2.06 3.50 Model 3 36_10000_RA 2.19 1.18 1.63 1.97 2.34 2.90 3.41 4.00 5.78 2.42 4.19 7.10 Model 3 Top-up flow between 36_10000_U_RARPS & 1.26 0.70 0.97 1.17 1.39 1.73 2.03 2.38 3.44 1.44 2.50 4.23 Model 3 36_10000_RA 36_30_8_RA 40.65 14.13 19.38 23.37 27.79 34.64 40.84 48.10 70.32 28.71 50.17 86.40 Model 3

Top-up flow between 0.91 0.70 0.96 1.16 1.38 1.72 2.03 2.39 3.50 1.72 3.00 5.17 Model 3 36_30_4_RA & 36_30_8_RA 36_1691_3_RA 11.29 4.69 6.51 7.87 9.36 11.67 13.74 16.15 23.44 9.66 16.88 28.80 Model 3 36_1691_8_RA 14.39 5.43 7.53 9.10 10.83 13.51 15.90 18.69 27.13 11.18 19.53 33.33 Model 3

Top-up between 36_1691_3_RA 3.10 1.35 1.87 2.26 2.69 3.36 3.95 4.65 6.75 3.18 5.55 9.47 Model 3 & 36_1691_8_RA 36_2116_4_RA 9.39 2.84 3.92 4.73 5.61 6.98 8.20 9.62 13.90 5.81 10.07 17.07 Model 3 36030_RA 119.25 20.81 25.47 28.49 31.44 35.48 38.71 42.10 50.78 35.00 47.55 62.38 Model 3 Top-up flow between 36074_RA 33.07 6.26 7.66 8.57 9.45 10.67 11.64 12.66 15.27 10.52 14.30 18.76 Model 3 & 36030_RA 36070_RARPS 126.51 17.09 20.91 23.39 25.82 29.13 31.78 34.57 41.69 28.74 39.04 51.22 Model 3

D9

Top-up flow between 36030_RA 7.26 1.50 1.83 2.05 2.26 2.55 2.78 3.03 3.65 2.52 3.42 4.49 Model 3 & 36070_RA 36_734_2_RA 9.24 1.98 2.74 3.31 3.93 4.88 5.74 6.73 9.72 4.06 7.05 11.94 Model 3 36153_RA 137.72 18.50 22.48 25.07 27.62 31.12 33.93 36.91 44.57 30.80 41.69 54.76 Model 3

Top-up flow between 1.97 0.35 0.42 0.47 0.52 0.58 0.64 0.69 0.83 0.58 0.78 1.03 Model 3 36070_RARPS & 36153_RA 36_1762_2_RA 27.10 4.58 6.25 7.48 8.81 10.81 12.58 14.59 20.55 9.19 15.45 25.24 Model 3 36_2308_4_RA 4.09 0.27 0.37 0.44 0.53 0.66 0.77 0.90 1.31 0.60 1.04 1.76 Model 3 36072_RA 191.19 17.51 21.05 23.26 25.40 28.25 30.47 32.79 38.51 28.58 37.44 47.32 Model 3 Top-up flow between 36153_RA 22.28 2.34 2.81 3.10 3.39 3.77 4.07 4.37 5.14 3.81 5.00 6.31 Model 3 & 36072_RA 36_624_6_RA 24.55 4.16 5.75 6.93 8.22 10.19 11.94 13.96 19.99 8.52 14.67 24.56 Model 3 36018_RA 220.40 20.10 24.61 27.64 30.76 35.18 38.84 42.84 53.64 33.96 47.72 65.90 Model 3 Top-up flow between 36072_RA 4.66 0.54 0.66 0.75 0.83 0.95 1.05 1.15 1.45 0.92 1.29 1.78 Model 3 & 36018_RA

Input flows Top‐up flows. These flows should be entered laterally Check flows. Modellers should check to make sure these flows are being reached at each HEP Some of these flows may be put in at the US point due to a small difference between US & DS flows.

D10

Model 4 - Cavan Flows for AEP AREA Model Node ID_CFRAMS 2 Qmed 0.5% 0.1% (km ) 50% (2) 20% (5) 10% (10) 5% (20) 2% (50) 1% (100) number (200) (1000) 36018_RA 220.40 16.25 16.25 19.89 22.34 24.86 28.44 31.40 34.63 43.36 Model 4 36_1102_5_RA 278.53 19.98 19.98 24.46 27.47 30.57 34.96 38.60 42.58 58.68 Model 4 36016_RA 508.18 50.70 50.70 61.90 69.50 77.26 88.31 97.44 107.42 149.45 Model 4 Top-up flow between 36018_RA & 9.25 16.25 16.25 19.84 22.28 24.77 28.31 31.23 34.43 47.91 Model 4 36016_RA 36_228_2_RA 102.06 16.24 16.24 20.06 22.74 25.54 29.61 33.07 36.94 60.35 Model 4

36_596_8_RA 80.21 10.36 10.36 13.44 15.66 18.04 21.61 24.71 28.24 52.01 Model 4

36_2398_2_RA 11.18 2.69 2.69 3.73 4.51 5.38 6.71 7.90 9.29 13.17 Model 4

36_78_1_RARPS 0.08 0.05 0.05 0.07 0.08 0.10 0.12 0.15 0.17 0.25 Model 4

36_789_2_RA 7.02 2.66 2.66 3.67 4.43 5.26 6.54 7.69 9.02 7.84 Model 4 Top-up flow between 36_78_1_RARPS & 6.94 2.63 2.63 3.63 4.38 5.21 6.47 7.61 8.92 7.76 Model 4 36_789_2_RA 36_422_4_RA 9.60 3.37 3.37 4.65 5.61 6.67 8.29 9.74 11.42 16.62 Model 4 Top-up flow between 36_789_2_RA & 2.58 0.98 0.98 1.36 1.64 1.95 2.42 2.84 3.33 4.85 Model 4 36_422_4_RA 36_1522_11_RA 10.66 3.49 3.49 4.83 5.83 6.93 8.62 10.14 11.90 17.11 Model 4 36_674_7_RA 6.87 1.55 1.55 2.14 2.59 3.07 3.82 4.48 5.26 7.59 Model 4 36_1328_3_RA 11.51 1.23 1.23 1.70 2.05 2.44 3.03 3.56 4.17 6.16 Model 4 36010_RA 776.55 66.80 66.80 81.56 91.58 101.80 116.37 128.39 141.55 177.09 Model 4 Top-up flow between 36016_RA & 36.29 3.79 3.79 4.62 5.19 5.77 6.60 7.28 8.02 10.04 Model 4 36010_RA 36_706_1_RA 14.66 3.03 3.03 4.20 5.07 6.04 7.53 8.86 10.42 14.83 Model 4

D11

36_769_U_RARPS 0.59 0.21 0.21 0.30 0.36 0.43 0.53 0.62 0.73 1.05 Model 4 36_769_1_RARPS 1.01 0.34 0.34 0.47 0.57 0.68 0.84 0.99 1.16 1.68 Model 4 Top-up flow between 36_769_U_RARPS 0.41 0.15 0.15 0.21 0.25 0.30 0.37 0.43 0.51 0.73 Model 4 & 36_769_1_RARPS 36_760_U_RARPS 0.79 0.33 0.33 0.45 0.54 0.65 0.80 0.95 1.11 1.60 Model 4

36_1676_2_RA 1.85 0.52 0.52 0.72 0.87 1.04 1.29 1.52 1.78 2.57 Model 4 36_1652_1_RA 2.93 0.85 0.85 1.17 1.41 1.68 2.09 2.45 2.88 4.15 Model 4 Top-up flow between 36_1676_2 & 1.08 0.33 0.33 0.46 0.56 0.66 0.82 0.96 1.13 1.63 Model 4 36_1652_1 36_1652_3_RA 5.68 1.77 1.77 2.45 2.95 3.50 4.36 5.12 6.00 8.67 Model 4 Top-up flow between 36_760_U_RARPS 1.96 0.65 0.65 0.90 1.09 1.29 1.61 1.89 2.21 3.20 Model 4 & 36_1652_3_RA 36_68_1_RARPS 0.46 0.16 0.16 0.22 0.26 0.31 0.39 0.46 0.54 0.78 Model 4 36_68_2_RARPS 1.40 0.46 0.46 0.64 0.77 0.91 1.13 1.33 1.56 2.26 Model 4 Top-up flow between 36_68_1_RARPS & 0.94 0.32 0.32 0.44 0.53 0.63 0.78 0.91 1.07 1.55 Model 4 36_68_2_RARPS 36_1678_2_RPS 1.44 0.21 0.21 0.29 0.35 0.41 0.51 0.60 0.71 1.02 Model 4 36_1678_3_RA 3.53 0.28 0.28 0.38 0.46 0.55 0.68 0.80 0.94 1.35 Model 4 Top-up flow between 36_1678_2_RPS & 2.09 0.17 0.17 0.23 0.28 0.34 0.42 0.49 0.57 0.83 Model 4 36_1678_3_RPS 36_1984_1_RA 3.14 0.55 0.55 0.75 0.91 1.08 1.34 1.58 1.85 2.67 Model 4 36_2232_U_RPS 0.10 0.04 0.04 0.06 0.07 0.08 0.10 0.12 0.14 0.20 Model 4 36_1611_U_RARPS 0.00 0.01 0.01 0.01 0.02 0.02 0.02 0.03 0.03 0.05 Model 4 36_1611_1_RARPS 0.48 0.27 0.27 0.37 0.45 0.54 0.67 0.78 0.92 1.33 Model 4 Top-up flow between 36_1611_U_RARPS 0.48 0.27 0.27 0.37 0.45 0.53 0.66 0.78 0.91 1.32 Model 4 &

D12

36_1611_1_RARPS

36_1114_2_RA 1.74 0.23 0.23 0.31 0.38 0.45 0.56 0.65 0.77 1.10 Model 4 Top-up flow between 36_2232_U_RPS & 1.16 0.15 0.15 0.21 0.26 0.30 0.38 0.44 0.52 0.75 Model 4 36_1114_2_RA 36_254_6_RA 25.51 4.98 4.98 6.87 8.29 9.84 12.21 14.33 16.79 24.38 Model 4 Top-up flow between 36_1984_1_RA & 20.63 4.08 4.08 5.63 6.79 8.06 10.01 11.74 13.76 19.97 Model 4 36_254_6_RA 36_1922_U_RARPS 0.01 0.01 0.01 0.01 0.02 0.02 0.02 0.03 0.03 0.05 Model 4 36_1922_9_RARPS 4.02 1.99 1.99 2.74 3.31 3.93 4.88 5.74 6.73 9.73 Model 4 Top-up flow between 36_1922_U_RA & 4.01 1.98 1.98 2.74 3.30 3.92 4.87 5.73 6.72 9.70 Model 4 36_1922_9_RARPS 36_1921_Inter_1_R 59.90 6.34 6.34 8.51 10.15 11.96 14.78 17.32 20.31 29.41 Model 4 ARPS Top-up flow between 36_706_1_RA & 4.10 0.51 0.51 0.69 0.82 0.97 1.20 1.40 1.65 2.38 Model 4 36_1921_Inter_2_R A 36_1921_Inter_2_R 59.92 6.34 6.34 8.51 10.15 11.96 14.79 17.33 20.32 31.06 Model 4 A Top-up flow between 36_1921_Inter_1_R ARPS & 0.02 0.00 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.02 Model 4 36_1921_Inter_2_R A 36031_RA 60.71 6.45 6.45 8.66 10.33 12.17 15.04 17.63 20.67 31.59 Model 4 Top-up flow between 36_1921_Inter_2_R 0.79 0.11 0.11 0.15 0.18 0.21 0.26 0.30 0.35 0.54 Model 4 A & 36031_RA 36_743_1_RARPS 0.24 0.10 0.10 0.14 0.16 0.19 0.24 0.28 0.33 0.48 Model 4 36_973_1_RARPS 0.01 0.01 0.01 0.01 0.02 0.02 0.02 0.03 0.03 0.05 Model 4 36_973_2_RARPS 1.06 0.27 0.27 0.37 0.45 0.53 0.66 0.77 0.91 1.31 Model 4

D13

Top-up flow between 36_973_1_RARPS 1.05 0.53 0.53 0.73 0.88 1.04 1.29 1.52 1.78 2.57 Model 4 & 36_973_2_RARPS Top-up flow between 36_973_1_RARPS

& 36_973_2_RARPS 36_1113_3_RA 3.59 1.55 1.55 2.14 2.58 3.06 3.81 4.47 5.25 7.58 Model 4 Top-up flow between 36_743_1_RARPS 2.53 1.43 1.43 1.97 2.38 2.82 3.51 4.12 4.83 6.99 Model 4 & 36_1113_3_RA 36_892_1_U_RARP 0.10 0.01 0.01 0.01 0.02 0.02 0.02 0.03 0.03 0.05 Model 4 S 36_882_6_RARPS 4.14 1.03 1.03 1.43 1.72 2.04 2.54 2.98 3.50 5.06 Model 4 Top-up flow between 36_892_1_U_RARP 4.04 1.05 1.05 1.45 1.75 2.08 2.59 3.04 3.57 5.16 Model 4 S & 36_882_6_RARPS 36_189_3_RA 82.32 6.52 6.52 8.48 9.91 11.43 13.72 15.71 17.99 24.60 Model 4 Top-up flow between 36_1113_RA & 12.91 1.15 1.15 1.49 1.75 2.01 2.42 2.77 3.17 4.34 Model 4 36_189_3_RA 36023_RA 866.32 74.36 74.36 90.79 101.95 113.33 129.54 142.92 157.57 197.13 Model 4 Top-up flow between 36010_RA & 7.46 0.86 0.86 1.05 1.18 1.32 1.50 1.66 1.83 2.29 Model 4 36023_RA 36037_RA 1457.11 93.22 93.22 116.15 131.90 148.31 171.99 191.94 213.93 274.81 Model 4 Top-up flow between 36023_RA & 590.78 21.88 21.88 27.27 30.96 34.81 40.37 45.06 50.22 64.51 Model 4 36037_RA 36_1017_2_RA 14.05 0.47 0.47 0.64 0.77 0.90 1.10 1.28 1.48 2.07 Model 4 36_1951_3_RA 1482.02 92.22 92.22 114.90 130.49 146.72 170.14 189.87 211.64 271.85 Model 4 Top-up flow between 36023_RA & 10.87 0.92 0.92 1.15 1.30 1.47 1.70 1.90 2.11 2.72 Model 4 36_1951_3_RA 36019_RA 1486.38 89.95 89.95 112.08 127.28 143.11 165.96 185.21 206.44 265.17 Model 4

D14

Top-up flow between 36_1951_3_RA & 4.36 0.38 0.38 0.47 0.54 0.61 0.70 0.78 0.87 1.12 Model 4 36019_RA 36036_RA 1486.70 89.99 89.99 112.13 127.34 143.18 166.04 185.30 206.54 265.31 Model 4 Top-up flow between 36019_RA & 0.32 0.03 0.03 0.04 0.05 0.05 0.06 0.07 0.08 0.10 Model 4 36036_RA 36_1904_9_RA 14.20 1.04 1.04 1.44 1.73 2.06 2.56 3.00 3.52 5.08 Model 4 36_2367_3_RA 1511.50 90.20 90.20 112.39 127.64 143.51 166.42 185.73 207.01 265.92 Model 4 Top-up flow between 36036_RA & 10.60 0.86 0.86 1.08 1.22 1.37 1.59 1.78 1.98 2.55 Model 4 36_2367_3_RA 36_2286_D_RARPS 1514.15 90.24 90.24 112.44 127.69 143.57 166.50 185.81 207.10 266.03 Model 4 Top-up flow between 36_2367_3_RA & 2.65 0.23 0.23 0.29 0.33 0.37 0.43 0.48 0.53 0.68 Model 4 36_2286_D_RARPS

MRFS Flows for AEP HEFS Flows for AEP AREA Model Node ID_CFRAMS (km2) 50% 20% 10% 5% 2% 1% 0.5% 0.1% 10% 1% 0.1% number (2) (5) (10) (20) (50) (100) (200) (1000) (10) (100) (1000) 36018_RA 220.40 20.10 24.61 27.64 30.76 35.18 38.84 42.84 53.64 33.94 47.68 65.85 Model 4 36_1102_5_RA 278.53 37.59 46.01 51.68 57.51 65.78 72.62 80.10 110.40 63.46 89.16 135.55 Model 4 36016_RA 508.18 62.66 76.51 85.91 95.49 109.15 120.43 132.77 184.72 105.47 147.86 226.79 Model 4 Top-up flow between 9.25 5.16 6.30 7.07 7.86 8.98 9.91 10.93 15.20 8.68 12.17 18.66 Model 4 36018_RA & 36016_RA 36_228_2_RA 102.06 20.28 25.04 28.39 31.88 36.97 41.29 46.11 75.34 34.88 50.73 92.56 Model 4 36_596_8_RA 80.21 12.94 16.78 19.55 22.53 26.98 30.84 35.26 64.92 24.02 37.89 79.76 Model 4 36_2398_2_RA 11.18 3.36 4.66 5.64 6.71 8.37 9.86 11.59 16.44 6.92 12.11 20.20 Model 4 36_78_1_RARPS 0.08 0.11 0.15 0.18 0.22 0.27 0.32 0.37 0.54 0.13 0.22 0.38 Model 4 36_789_2_RA 7.02 3.32 4.59 5.53 6.57 8.17 9.60 11.26 9.79 6.80 11.79 12.03 Model 4 Top-up flow between 36_78_1_RARPS & 6.94 3.29 4.54 5.47 6.50 8.08 9.50 11.14 9.69 6.73 11.67 11.90 Model 4 36_789_2_RA 36_422_4_RA 9.60 4.21 5.81 7.01 8.32 10.35 12.15 14.26 20.74 8.61 14.93 25.48 Model 4

D15

Top-up flow between 36_789_2_RA & 2.58 1.23 1.69 2.04 2.43 3.02 3.55 4.16 6.05 2.51 4.36 7.44 Model 4 36_422_4_RA 36_1522_11_RA 10.66 4.36 6.03 7.28 8.65 10.76 12.65 14.86 21.36 8.94 15.55 26.25 Model 4 36_674_7_RA 6.87 1.94 2.68 3.23 3.83 4.77 5.60 6.57 9.48 3.97 6.88 11.64 Model 4 36_1328_3_RA 11.51 1.54 2.12 2.56 3.04 3.78 4.44 5.21 7.69 3.15 5.46 9.45 Model 4 36010_RA 776.55 82.78 101.08 113.49 126.16 144.21 159.11 175.42 219.46 139.44 195.48 269.62 Model 4 Top-up flow between 36.29 4.69 5.73 6.43 7.15 8.17 9.02 9.94 12.44 7.90 11.08 15.28 Model 4 36016_RA & 36010_RA 36_706_1_RA 14.66 3.78 5.24 6.33 7.54 9.40 11.06 13.00 18.51 7.78 13.59 22.74 Model 4 36_769_U_RARPS 0.59 0.30 0.42 0.50 0.60 0.74 0.87 1.02 1.48 1.05 1.82 3.08 Model 4 36_769_1_RARPS 1.01 0.46 0.64 0.77 0.92 1.14 1.34 1.57 2.27 1.72 2.99 5.06 Model 4 Top-up flow between 36_769_U_RARPS & 0.41 0.20 0.28 0.34 0.40 0.50 0.58 0.68 0.99 0.75 1.30 2.20 Model 4 36_769_1_RARPS 36_760_U_RARPS 0.79 0.41 0.57 0.68 0.81 1.01 1.19 1.39 2.01 0.93 1.61 2.72 Model 4 36_1676_2_RA 1.85 0.65 0.90 1.09 1.30 1.61 1.89 2.22 3.21 1.34 2.32 3.94 Model 4 36_1652_1_RA 2.93 1.20 1.66 2.00 2.38 2.95 3.47 4.07 5.88 4.11 7.13 12.08 Model 4 Top-up flow between 1.08 0.47 0.65 0.79 0.93 1.16 1.36 1.60 2.31 1.61 2.80 4.75 Model 4 36_1676_2 & 36_1652_1 36_1652_3_RA 5.68 2.64 3.64 4.39 5.22 6.49 7.62 8.94 12.92 8.27 14.35 24.31 Model 4 Top-up flow between 36_760_U_RARPS & 1.96 0.97 1.34 1.62 1.92 2.39 2.81 3.29 4.76 3.05 5.29 8.96 Model 4 36_1652_3_RA 36_68_1_RARPS 0.46 0.25 0.34 0.41 0.49 0.61 0.72 0.84 1.21 0.72 1.24 2.11 Model 4 36_68_2_RARPS 1.40 0.78 1.08 1.30 1.55 1.93 2.26 2.65 3.83 1.93 3.35 5.69 Model 4 Top-up flow between 36_68_1_RARPS & 0.94 0.54 0.74 0.89 1.06 1.32 1.55 1.82 2.63 1.33 2.30 3.90 Model 4 36_68_2_RARPS 36_1678_2_RPS 1.44 0.26 0.36 0.44 0.52 0.65 0.76 0.89 1.28 0.59 1.02 1.73 Model 4 36_1678_3_RA 3.53 0.36 0.50 0.61 0.72 0.89 1.05 1.23 1.78 0.99 1.72 2.92 Model 4

D16

Top-up flow between 36_1678_2_RPS & 2.09 0.22 0.31 0.37 0.44 0.55 0.64 0.76 1.09 0.61 1.05 1.79 Model 4 36_1678_3_RPS 36_1984_1_RA 3.14 0.68 0.94 1.13 1.35 1.68 1.97 2.31 3.34 1.39 2.42 4.10 Model 4 36_2232_U_RPS 0.10 0.11 0.16 0.19 0.22 0.28 0.33 0.38 0.55 0.20 0.35 0.60 Model 4 36_1611_U_RARPS 0.00 0.01 0.02 0.02 0.02 0.03 0.03 0.04 0.06 0.02 0.04 0.07 Model 4 36_1611_1_RARPS 0.48 0.44 0.61 0.74 0.88 1.09 1.28 1.51 2.18 0.80 1.39 2.36 Model 4 Top-up flow between 36_1611_U_RARPS & 0.48 0.44 0.61 0.73 0.87 1.08 1.27 1.50 2.16 0.80 1.38 2.34 Model 4 36_1611_1_RARPS 36_1114_2_RA 1.74 0.38 0.53 0.64 0.76 0.94 1.11 1.30 1.86 0.95 1.64 2.76 Model 4 Top-up flow between 36_2232_U_RPS & 1.16 0.26 0.36 0.43 0.52 0.64 0.75 0.88 1.27 0.65 1.12 1.88 Model 4 36_1114_2_RA 36_254_6_RA 25.51 6.34 8.76 10.56 12.54 15.57 18.27 21.40 31.08 15.25 26.37 44.86 Model 4 Top-up flow between 36_1984_1_RA & 20.63 5.20 7.18 8.66 10.28 12.76 14.97 17.53 25.46 12.50 21.61 36.76 Model 4 36_254_6_RA 36_1922_U_RARPS 0.01 0.01 0.02 0.02 0.02 0.03 0.03 0.04 0.06 0.02 0.04 0.06 Model 4 36_1922_9_RARPS 4.02 3.65 5.04 6.08 7.23 8.98 10.55 12.38 17.88 7.74 13.43 22.75 Model 4 Top-up flow between 36_1922_U_RA & 4.01 3.64 5.03 6.07 7.21 8.96 10.53 12.35 17.85 7.72 13.40 22.71 Model 4 36_1922_9_RARPS 36_1921_Inter_1_RARPS 59.90 11.57 15.52 18.52 21.82 26.97 31.61 37.06 53.67 25.62 43.73 74.24 Model 4 Top-up flow between 36_706_1_RA & 4.10 0.94 1.26 1.50 1.77 2.19 2.56 3.00 4.35 2.08 3.54 6.01 Model 4 36_1921_Inter_2_RA 36_1921_Inter_2_RA 59.92 11.57 15.53 18.52 21.83 26.98 31.62 37.07 56.67 25.62 43.74 78.39 Model 4 Top-up flow between 36_1921_Inter_1_RARPS & 0.02 0.01 0.01 0.01 0.01 0.02 0.02 0.02 0.03 0.01 0.03 0.05 Model 4 36_1921_Inter_2_RA 36031_RA 60.71 12.29 16.49 19.68 23.19 28.66 33.59 39.38 60.20 29.93 51.10 91.57 Model 4

D17

Top-up flow between 36_1921_Inter_2_RA & 0.79 0.21 0.28 0.34 0.40 0.49 0.58 0.68 1.03 0.51 0.88 1.57 Model 4 36031_RA 36_743_1_RARPS 0.24 0.12 0.17 0.20 0.24 0.30 0.35 0.41 0.60 0.25 0.43 0.74 Model 4 36_973_1_RARPS 0.01 0.01 0.02 0.02 0.02 0.03 0.03 0.04 0.06 0.02 0.04 0.07 Model 4 36_973_2_RARPS 1.06 0.49 0.68 0.82 0.98 1.21 1.43 1.67 2.42 1.04 1.81 3.06 Model 4 Top-up flow between 36_973_1_RARPS & 1.05 1.13 1.56 1.88 2.24 2.78 3.27 3.83 5.54 2.04 3.54 6.00 Model 4 36_973_2_RARPS Top-up flow between 36_973_1_RARPS & 0.00 Model 4 36_973_2_RARPS 36_1113_3_RA 3.59 2.67 3.69 4.45 5.29 6.58 7.73 9.06 13.09 6.40 11.10 18.81 Model 4 Top-up flow between 36_743_1_RARPS & 2.53 2.96 4.09 4.93 5.86 7.29 8.56 10.04 14.51 5.36 9.29 15.74 Model 4 36_1113_3_RA 36_892_1_U_RARPS 0.10 0.01 0.02 0.02 0.02 0.03 0.04 0.04 0.06 0.02 0.04 0.07 Model 4 36_882_6_RARPS 4.14 1.28 1.76 2.13 2.53 3.14 3.69 4.33 6.25 2.61 4.53 7.68 Model 4 Top-up flow between 36_892_1_U_RARPS & 4.04 1.38 1.90 2.29 2.72 3.38 3.98 4.66 6.74 3.66 6.34 10.75 Model 4 36_882_6_RARPS 36_189_3_RA 82.32 11.75 15.29 17.86 20.61 24.74 28.33 32.44 44.36 23.72 37.63 58.91 Model 4 Top-up flow between 36_1113_RA & 12.91 2.06 2.68 3.13 3.61 4.33 4.96 5.68 7.76 4.15 6.58 10.31 Model 4 36_189_3_RA 36023_RA 866.32 91.56 111.79 125.53 139.53 159.49 175.97 194.01 242.72 154.22 216.20 298.20 Model 4 Top-up flow between 7.46 1.86 2.28 2.56 2.84 3.25 3.58 3.95 4.94 3.78 5.30 7.31 Model 4 36010_RA & 36023_RA 36037_RA 1457.11 114.85 143.10 162.51 182.72 211.89 236.47 263.58 338.57 171.47 249.52 357.25 Model 4 Top-up flow between 590.78 27.08 33.74 38.31 43.08 49.95 55.75 62.14 79.82 40.42 58.82 84.22 Model 4 36023_RA & 36037_RA 36_1017_2_RA 14.05 0.59 0.80 0.95 1.12 1.37 1.58 1.83 2.56 1.17 1.95 3.15 Model 4 36_1951_3_RA 1482.02 113.61 141.56 160.76 180.76 209.62 233.93 260.74 334.93 198.08 288.23 412.68 Model 4

D18

Top-up flow between 10.87 1.14 1.41 1.61 1.81 2.09 2.34 2.61 3.35 1.98 2.88 4.12 Model 4 36023_RA & 36_1951_3_RA 36019_RA 1486.38 110.82 138.08 156.81 176.32 204.47 228.18 254.34 326.70 193.21 281.15 402.54 Model 4

Top-up flow between 4.36 0.47 0.58 0.66 0.75 0.87 0.97 1.08 1.38 0.82 1.19 1.70 Model 4 36_1951_3_RA & 36019_RA 36036_RA 1486.70 110.84 138.11 156.85 176.35 204.51 228.23 254.39 326.77 193.25 281.21 402.62 Model 4 Top-up flow between 0.32 0.07 0.09 0.10 0.11 0.13 0.15 0.16 0.21 0.15 0.22 0.31 Model 4 36019_RA & 36036_RA 36_1904_9_RA 14.20 1.38 1.91 2.31 2.74 3.41 4.01 4.70 6.78 3.98 6.91 11.69 Model 4 36_2367_3_RA 1511.50 111.07 138.39 157.16 176.71 204.92 228.69 254.90 327.43 193.64 281.77 403.43 Model 4

Top-up flow between 10.60 1.06 1.33 1.51 1.69 1.96 2.19 2.44 3.14 1.86 2.70 3.87 Model 4 36036_RA & 36_2367_3_RA 36_2286_D_RARPS 1514.15 111.44 138.85 157.68 177.30 205.60 229.45 255.75 328.52 200.89 292.32 418.53 Model 4 Top-up flow between 36_2367_3_RA & 2.65 0.29 0.36 0.41 0.46 0.53 0.59 0.66 0.85 0.50 0.73 1.04 Model 4 36_2286_D_RARPS

Input flows Top‐up flows. These flows should be entered laterally Check flows. Modellers should check to make sure these flows are being reached at each HEP Some of these flows may be put in at the US point due to a small difference between US & DS flows.

D19

Model 5 - Bundoran & Tullaghan Flows for AEP AREA Model Node ID_CFRAMS 2 Qmed 0.5% 0.1% (km ) 50% (2) 20% (5) 10% (10) 5% (20) 2% (50) 1% (100) number (200) (1000) 35_4230_U_RARPS 0.55 0.22 0.22 0.30 0.37 0.44 0.54 0.64 0.75 1.08 Model 5 35_4230_3_RA 1.89 1.02 1.02 1.40 1.69 2.01 2.50 2.94 3.44 4.98 Model 5 Top-up flow between 35_4230_U_RARPS & 35_4230_3_RA 1.34 0.81 0.81 1.12 1.35 1.61 2.00 2.35 2.75 3.97 Model 5

35_4241_U_RA 25.42 7.10 7.10 9.01 10.34 11.73 13.75 15.47 17.37 22.69 Model 5 35_4239_1_RA 26.97 7.62 7.62 9.67 11.10 12.60 14.76 16.61 18.65 24.36 Model 5 Top-up flow between 35_4241_U_RARPS & 35_4239_1_RA 1.56 0.53 0.53 0.67 0.77 0.87 1.02 1.15 1.29 1.68 Model 5

35_195_1_RA 248.12 26.29 26.29 32.18 36.15 40.22 46.01 50.79 56.02 70.14 Model 5 35013_RA 250.09 26.57 26.57 32.52 36.54 40.65 46.50 51.34 56.62 70.89 Model 5 Top-up flow between 35_195_1_RA & 35013_RA 1.97 0.42 0.42 0.51 0.57 0.64 0.73 0.81 0.89 1.11 Model 5 35_2665_U_RARPS 0.05 0.02 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.11 Model 5 35_2665_6_RARPS 1.98 0.80 0.80 1.11 1.34 1.59 1.98 2.32 2.73 3.94 Model 5 Top-up flow between 35_2665_U_RARPS & 35_2665_6_RARPS 1.93 0.79 0.79 1.08 1.31 1.55 1.93 2.27 2.66 3.85 Model 5 35_4067_3_RARPS 257.24 28.29 28.29 34.63 38.90 43.29 49.51 54.66 60.29 75.48 Model 5 Top-up flow between 35013_RA & 35_4067_3_RARPS 5.17 1.07 1.07 1.31 1.47 1.64 1.87 2.07 2.28 2.86 Model 5 35_1000_1_RPS 0.03 0.02 0.02 0.03 0.03 0.04 0.05 0.06 0.07 0.10 Model 5

D20

35_2282_U_RA 1.77 0.74 0.74 1.03 1.24 1.47 1.83 2.15 2.52 3.64 Model 5 35_2327_4_RA 3.71 1.77 1.77 2.45 2.96 3.51 4.36 5.13 6.01 8.69 Model 5 Top-up flow between 35_2282_U_RA & 35_2327_4_RA 1.94 0.97 0.97 1.33 1.61 1.91 2.37 2.79 3.27 4.73 Model 5 35_1056_2_RARPS 261.00 29.11 29.11 35.63 40.03 44.54 50.95 56.25 62.04 77.67 Model 5 Top-up flow between 35013_RA & 35_1056_2_RARPS 0.05 0.02 0.02 0.02 0.02 0.02 0.03 0.03 0.03 0.04 Model 5

MRFS Flows for AEP HEFS Flows for AEP AREA Model Node ID_CFRAMS 2 (km ) 50% 20% 10% 5% 2% 1% 0.5% 0.1% 10% 1% 0.1% number (2) (5) (10) (20) (50) (100) (200) (1000) (10) (100) (1000) 35_4230_U_RARPS 0.55 0.32 0.45 0.54 0.64 0.79 0.93 1.09 1.58 1.04 1.81 3.07 Model 5 35_4230_3_RA 1.89 1.96 2.70 3.26 3.87 4.81 5.65 6.63 9.58 3.78 6.56 11.12 Model 5 Top-up flow between 35_4230_U_RARPS & 35_4230_3_RA 1.34 1.52 2.10 2.53 3.01 3.74 4.39 5.15 7.44 2.74 4.76 8.06 Model 5

35_4241_U_RA 25.42 8.79 11.15 12.81 14.53 17.02 19.15 21.51 28.09 15.73 23.53 34.51 Model 5 35_4239_1_RA 26.97 9.66 12.26 14.07 15.97 18.71 21.05 23.64 30.87 17.57 26.28 38.55 Model 5 Top-up flow between 35_4241_U_RARPS & 35_4239_1_RA 1.56 1.12 1.43 1.64 1.86 2.18 2.45 2.75 3.59 2.42 3.62 5.31 Model 5

35_195_1_RA 248.12 32.68 39.99 44.93 49.99 57.18 63.13 69.63 87.18 55.20 77.56 107.10 Model 5 35013_RA 250.09 32.92 40.29 45.26 50.36 57.61 63.60 70.15 87.82 55.61 78.13 107.90 Model 5

Top-up flow between 35_195_1_RA & 35013_RA 1.97 0.52 0.63 0.71 0.79 0.91 1.00 1.10 1.38 0.87 1.23 1.70 Model 5

D21

35_2665_U_RARPS 0.05 0.03 0.04 0.05 0.06 0.07 0.08 0.10 0.14 0.06 0.10 0.17 Model 5 35_2665_6_RARPS 1.98 1.67 2.31 2.79 3.31 4.12 4.84 5.67 8.20 4.11 7.13 12.08 Model 5 Top-up flow between 35_2665_U_RARPS & 35_2665_6_RARPS 1.93 1.65 2.28 2.76 3.27 4.07 4.78 5.61 8.10 4.07 7.06 11.97 Model 5 35_4067_3_RARPS 257.24 34.30 41.98 47.16 52.47 60.02 66.26 73.09 91.50 58.39 82.05 113.30 Model 5

Top-up flow between 35013_RA & 35_4067_3_RARPS 5.17 2.23 2.73 3.07 3.42 3.91 4.32 4.76 5.96 4.54 6.38 8.81 Model 5 35_1000_1_RPS 0.03 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.12 0.05 0.09 0.15 Model 5 35_2282_U_RA 1.77 0.93 1.28 1.54 1.83 2.28 2.68 3.14 4.54 1.90 3.29 5.58 Model 5 35_2327_4_RA 3.71 3.05 4.21 5.08 6.03 7.50 8.81 10.33 14.92 7.01 12.15 20.60 Model 5 Top-up flow between 35_2282_U_RA & 35_2327_4_RA 1.94 2.11 2.92 3.52 4.18 5.20 6.10 7.16 10.34 5.20 9.02 15.29 Model 5 35_1056_2_RARPS 261.00 36.24 44.36 49.83 55.45 63.42 70.02 77.23 96.69 59.49 83.59 115.43 Model 5

Top-up flow between 35013_RA & 35_1056_2_RARPS 0.05 0.03 0.04 0.04 0.05 0.06 0.06 0.07 0.08 0.06 0.09 0.12 Model 5

Input flows Top‐up flows. These flows should be entered laterally Check flows. Modellers should check to make sure these flows are being reached at each HEP Some of these flows may be put in at the US point due to a small difference between US & DS flows.

D22