Kin­gussie Flood Study Baseline Modelling

31 August 2020

AECOM Qual­ity information

Pre­pared by Checked by Approved by

Mor­ag Hut­ton Sally Homon­cik Sally Homon­cik Seni­or Hydro­lo­gist Seni­or Geo­mor­pho­lo­gist Seni­or Geomorphologist

Revi­sion History

Revi­sion Revi­sion date Details Author­ized Name Pos­i­tion 1 09/07/2020 Draft for com­ment 2 31/08/2020 Final issue based on cli­ent com­ments 3 Dis­tri­bu­tion List

Hard Cop­ies PDF Required Asso­ci­ation / Com­pany Name

AECOM Pre­pared for: The High­land Council

Pre­pared by: AECOM E: morag.​hutton@​aecom.​com

AECOM Lim­ited 1 Tan­field Edin­burgh EH3 5DA UK

T: +44 131 301 8600 aecom​.com

© 2020 AECOM Lim­ited. All Rights Reserved.

This doc­u­ment has been pre­pared by AECOM Lim­ited (“AECOM”) for sole use of our cli­ent (the ​“Cli­ent”) in accord­ance with gen­er­ally accep­ted con­sultancy prin­ciples, the budget for fees and the terms of ref­er­ence agreed between AECOM and the Cli­ent. Any inform­a­tion provided by third parties and referred to herein has not been checked or veri­fied by AECOM, unless oth­er­wise expressly stated in the doc­u­ment. No third party may rely upon this doc­u­ment without the pri­or and express writ­ten agree­ment of AECOM.

AECOM

Table of Contents

1. Intro­duc­tion 6
2. Pro­ject Back­ground 7 2.1 Site Vis­it 9 2.2 His­tor­ic Flood­ing and SEPA flood maps 9 2.3 Hydro­power scheme 10
3. Flu­vi­al Hydro­lo­gic­al Assess­ment 12 3.1 Meth­od­o­logy over­view 12 3.2 Catch­ment descriptors 13 3.3 Total catch­ment hydro­logy 13 3.4 Cli­mate change 14 3.5 Flu­vi­al sub-catch­ment hydro­logy 14 3.5.1 Delin­eation and rep­res­ent­a­tion 15 3.5.2 ReFH2 recon­cili­ation 16 3.6 Flood Study inflows 16
4. Joint prob­ab­il­ity 18 4.1 Run mat­rix 18 4.2 River Spey levels 18 4.3 Joint prob­ab­il­ity con­clu­sions 19
5. Hydraul­ic mod­el­ling 20 5.1 Exist­ing Upstream Hydro­power Mod­el 20 5.2 1D/2D Town Mod­el Schem­at­isa­tion 20 5.2.1 One dimen­sion­al chan­nel mod­el 20 5.2.2 Two dimen­sion­al flood plain mod­el 22 5.2.3 Ground truth­ing 22 5.2.4 Mod­el runs para­met­ers 23 5.3 Mod­el amend­ments — cre­at­ing a new baseline 23 5.3.1 Block­age of bridges 23 5.3.1.1 Block­age by dif­fer­ence in sur­vey 24 5.3.1.2 Block­age to match 50% AEP gauge level 24 5.3.2 Chan­nel capa­city 24 5.3.3 New baseline con­clu­sions 25 5.4 Veri­fic­a­tion 25 5.5 Sens­it­iv­ity ana­lys­is 26 5.5.1 Flow 27 5.5.1.1 SEPA recom­men­ded uplift of 20% 27 5.5.1.2 Manning’s rough­ness 27 5.5.1.3 Froude lim­it 27 5.5.1.4 Bridge para­met­ers 28 5.5.1.5 Block­ages 28 5.5.1.6 Remov­al of hydro­power scheme 30
6. Res­ults 32 6.1 Baseline 32
7. Con­clu­sions 34 7.1 Mod­el inflows 34 7.1.1 Upstream mod­el hydro­logy 34 7.1.2 Town mod­el 34 7.1.3 Joint prob­ab­il­ity 34 7.2 Hydraul­ic mod­el­ling 34 7.2.1 Exist­ing upstream hydro­power mod­el 34 7.2.2 1D/2D town mod­el 35 AECOM 7.2.3 New baseline 35 7.2.4 Sens­it­iv­ity test­ing and veri­fic­a­tion 36 7.3 Baseline flood risk 36
8. Next Steps 37 Appendix A — Site pho­to­graphs 38 Appendix B — August 2014 flood event pho­tos 39 Appendix C — Hydro­power scheme 40 Appendix D — Topo­graph­ic sur­vey 41 Appendix E — Hydraul­ic mod­el build 42 Appendix F — SEPA cor­res­pond­ence 43 Appendix G — Mod­el res­ults 44 Appendix H — Floodmaps 45

Fig­ures

Fig­ure 2 – 1: Study area 8 Fig­ure 2 – 2: Schem­at­isa­tion of the hydro­power scheme 11 Fig­ure 3 – 1: Sub­catch­ments and inflow loc­a­tions into mod­el 15 Fig­ure 4 – 1: River Spey level bound­ar­ies 19 Fig­ure 5 – 1: 1D node loc­a­tions 21 Fig­ure 5 – 2: Pho­to­graph from the August 2014 event 26 Fig­ure 5 – 3: Block­age loc­a­tions 29 Fig­ure 6 – 1: 0.5% AEP flu­vi­al flood event 33

Tables

Table 3 – 1: AEP and return peri­od equi­val­ent 13 Table 3 – 2: FEH catch­ment descriptors to con­flu­ence with River Spey 13 Table 3 – 3: Peak flow estim­a­tion from the 2015 Flood Study to the down­stream extent of the Gyn­ack Burn 14 Table 3 – 4: Sub catch­ment ReFH2 inflow uplifts 16 Table 3 – 5: Peak flows to be used in the mod­el­ling exer­cise 17 Table 5 – 1: Res­ults of ground truth­ing (LiD­AR vs topo­graph­ic sur­vey) 23 Table 5 – 2: Peak flows with and without hydro­power scheme 30

AECOM

1. Intro­duc­tion

AECOM have been com­mis­sioned by The High­land Coun­cil (THC) to under­take a Flood Study in Kin­gussie, Scot­land. Kin­gussie is in close prox­im­ity to two water­courses; the River Spey lies to the south of the town and the Gyn­ack Burn runs through the centre. Fre­quent flood­ing from both these sources has affected Kin­gussie and the sur­round­ing area and this Flood Study aims to devel­op an under­stand­ing of the baseline flood mech­an­isms and asso­ci­ated damages.

Pre­vi­ous stud­ies have been under­taken by AECOM (formerly URS) on behalf of THC in 2012 and 2015 to invest­ig­ate flood risk in Kin­gussie and to devel­op options for alle­vi­at­ing it. These stud­ies con­tained hydraul­ic mod­el­ling ele­ments of both the upstream catch­ment as well as through the town.

Since the pre­vi­ous Flood Stud­ies were under­taken, the baseline con­di­tions in the upstream catch­ment have changed due to the con­struc­tion of a Hydro power scheme on Loch Gyn­ack, which as well as provid­ing a power source, also provides a sec­ond­ary bene­fit of flood atten­u­ation. This Flood Study aims to assess the impact of this flood atten­u­ation on peak flows in com­bin­a­tion with a more detailed mod­el­ling approach through town so that an updated under­stand­ing of baseline flood risk can be established.

As part of this stage of the Flood Study, a baseline dam­age assess­ment will also be under­taken to estab­lish the likely costs asso­ci­ated with the cur­rent flood risk. The eco­nom­ic assess­ment is not included in this report.

Under­stand­ing the baseline flood con­di­tions and eco­nom­ics will allow an informed decision to be made regard­ing pro­gres­sion of the Flood Study to an option­eer­ing stage.

AECOM 6

1. Pro­ject Background

The study area is out­lined in Fig­ure 2 – 1 below and encom­passes the town of Kin­gussie. The upstream catch­ment that con­tains the Loch Gyn­ack hydro power scheme, is not with­in the study area for assess­ing flood risk but will be mod­elled to determ­ine flows into the study area. The pur­pose of this study is to identi­fy areas at risk in Kin­gussie from flu­vi­al flood­ing dur­ing cur­rent day and cli­mate change scenarios.

The main flu­vi­al flood risk in the study area is from the Gyn­ack Burn which ori­gin­ates upstream in rur­al land before run­ning through the centre of Kin­gussie and join­ing with the River Spey down­stream of the town. Whilst much of the catch­ment is rur­al in nature, there are sev­er­al hydro­power schemes loc­ated upstream of Kin­gussie. All but one of these schemes take water from the Gyn­ack Burn and return it to the water­course imme­di­ately down­stream, main­tain­ing the flood mech­an­ism and peak tim­ings. How­ever one of the schemes diverts flow from the Gyn­ack Burn and tem­por­ar­ily stores it in Loch Gyn­ack and by doing so, provides peak flow atten­u­ation, affect­ing both the peak flow and tim­ings of a flood event.

Pre­vi­ous Flood Stud­ies set out to invest­ig­ate how the Gyn­ack Loch Hydro­power Scheme could affect peak flows through Kin­gussie. At the time of the most recent pre­vi­ous study (2015), the scheme had not been con­struc­ted and was in the ini­tial design stage. Indic­at­ive details such as weir lengths and crests were provided by Pit­main Estate and used to con­struct a 1D hydraul­ic mod­el of the lat­er­al weir, diver­sion, loch and out­fall struc­ture. Full details of the scheme can be found in Sec­tion 2.3.

The scheme has now been con­struc­ted and the con­struc­tion draw­ings have been provided by THC. These more detailed dimen­sions and elev­a­tions have been used to update the exist­ing hydro scheme mod­el now that the finer details have been resolved. Oth­er minor changes to the mod­el as a res­ult of updated inform­a­tion has also been under­taken to improve stability.

For the pur­poses of this study, the mod­el­ling has been split into 2 parts. The fol­low­ing has been adopted;

• 1D mod­el of the upstream diver­sion chan­nel and Loch Gyn­ack to the con­flu­ence with the River Spey from the pre­vi­ous 2015 Flood Study but with minor adjust­ments. Whilst this mod­el exten­ded to the Spey, the sec­tion through town was in 1D only and based on older inform­a­tion so there­fore not appro­pri­ate for use in this study. There are two sep­ar­ate mod­els, one with the hydro­power scheme included and one without. This mod­el is used to determ­ine inflows into the town mod­el below. Flow from these mod­els was extrac­ted above town to be used as the inflow into the new 1D/2D model;

• Newly cre­ated 1D/2D mod­el through­out the town cre­ated for this Flood study, extend­ing from Old Dis­til­lery Road, 300m upstream of High Street to the con­flu­ence with the River Spey.

Full details of how these mod­els were con­struc­ted and how they inter­act can be found in Sec­tion 5.

AECOM 7

Fig­ure 2 – 1: Study area

AECOM 8

2.1 Site Visit

A site walkover was under­taken in July 2019 to estab­lish gen­er­al topo­graphy and con­straints on and around the Gyn­ack Burn. This assess­ment exten­ded through town to the con­flu­ence with the River Spey. It was not pos­sible to view the upstream ele­ments of Loch Gyn­ack Hydro Scheme at the time of the vis­it. Dur­ing the walkover, a review of pos­sible flood flow routes and assess­ment of the viab­il­ity of poten­tial options was also undertaken.

Upstream of Kin­gussie, around the golf course, the chan­nel is seen to be incised with a nat­ur­al pro­file. There are sev­er­al sec­tions where the water­course has cut into the bed­rock, form­ing water­falls and pools. Along the entire upstream reach from the golf course to the town, the bed mater­i­al is either bed­rock or stones of vari­ous sizes, with small 5cm dia­met­er stones as well as much lar­ger 0.5m+ dia­met­er mater­i­al. Banks are lightly veget­ated around the water­course and treelined fur­ther up the bank.

The Gyn­ack Burn through town is can­al­ised but main­tains a semi nat­ur­al pro­file and riverb­anks for the major­ity. Some sec­tions of the banks have been rein­forced with gabi­on bas­kets, primar­ily upstream of the rail­way bridge. At the time of the vis­it, the banks were veget­ated with scrub, ferns and some lar­ger trees. Bed mater­i­al was seen to gen­er­ally con­sist of small to medi­um smooth stones. Upstream of bridges, depos­its of these stones were observed with­in the chan­nel, form­ing small islands. This was par­tic­u­larly notice­able at the rail­way bridge. Bed mater­i­al has also been dredged from the water­course at vari­ous loc­a­tions, primar­ily upstream of struc­tures, and some of this mater­i­al has been depos­ited on the banks.

Down­stream of Kin­gussie, between the town and the con­flu­ence with the River Spey, the chan­nel is can­al­ised along the B970, with lightly veget­ated banks and small to medi­um sized bed mater­i­al. As the water­course approaches the River Spey, the chan­nel forms a more nat­ur­al pro­file and is seen to vary its course fre­quently due to the act­ive nature of the water­course. Bed mater­i­al depos­its are built up along either side of the water­course with min­im­al veget­a­tion in some parts and dens­er trees and shrubs in oth­ers. An embank­ment runs along a sig­ni­fic­ant reach of the left hand bank to pro­tect agri­cul­tur­al land south of the High School.

There are a total of 4 bridge cross­ings through Kin­gussie, High Street, Spey Street, the rail­way line and an access bridge to the High School. These bridges vary sig­ni­fic­antly in their capa­cit­ies, with sof­fit levels above bed level ran­ging from 0.7m to 2.1m at the time of sur­vey. The bridges with the low­est sof­fit clear­ance are the rail­way bridge and school access bridge. It should be noted that sed­i­ment depos­ition upstream of bridges is a known issue and capa­cit­ies of the bridges change frequently.

A gauge was noted on the water­course upstream of the Spey Street Bridge. This gauge is level only.

Pho­to­graphs can be found in Appendix A.

2.2 His­tor­ic Flood­ing and SEPA flood maps

Flu­vi­al flood­ing in and around Kin­gussie is pre­dicted by the SEPA online Flood Risk Man­age­ment Maps¹ (FRM maps) from both the Gyn­ack Burn and the River Spey. Water is shown to exit the Gyn­ack Burn upstream of the High Street Bridge as well as oth­er loc­a­tions around Spey Street and the rail­way bridge. This flow is seen to extend south­wards, affect­ing roads, prop­er­ties and the rail­way line, before join­ing flood­wa­ter from the River Spey. The River Spey is seen to inund­ate large por­tions of the flood­plain to the south of Kin­gussie, extend­ing up to the High School.

The SEPA floodmaps are backed up with the his­tor­ic flood reports set out in The Flood Risk Man­age­ment Strategy as well as anec­dot­al accounts from SEPA and THC, where flood­ing has been noted at prop­er­ties, com­munity facil­it­ies, agri­cul­tur­al land and trans­port net­works. Flu­vi­al flood­ing has caused par­tic­u­lar issues on Spey Street, Gyn­ack Street, the rail­way line and at the High School. Anec­dot­al and pho­to­graph­ic evid­ence has shown that block­age of struc­tures, both from sed­i­ment and from woody debris, plays a sig­ni­fic­ant role in flood­ing in Kin­gussie, with all three of the lower struc­tures, Spey Street, rail­way line and school access bridge fre­quently becom­ing blocked.

SEPA and THC have provided the fol­low­ing records:

• Novem­ber 2019 — Gyn­ack Burn burst its banks, caus­ing flood­ing and the rail­way line to close;

• July 2019 — rail­way line closed;

¹ http://​map​.sepa​.org​.uk/​f​l​o​o​d​m​a​p​/​m​a​p.htm

AECOM 9

• Decem­ber 2015 – Rail­way line closed;

• Janu­ary 2015 – Sta­tion Road flooded and the rail­way line was closed;

• August 2014 — Ex-hur­ricane Ber­tha — flood­ing to sev­er­al roads and prop­er­ties: Sil­ver­f­jord Hotel, Sta­tion Road, Spey Street, Kin­gussie High School and the bowl­ing green;

• Janu­ary 2008 – flood­ing from High Street Bridge to Spey Street Bridge mak­ing the road impassable;

• Decem­ber 2006 — Flood­ing upstream of all 3 bridges flood­ing Spey Street, Gyn­ack Street, Mar­ket Lane, Ruthven Road and Kin­gussie High School;

• Janu­ary 2005 — flood­ing on Spey Street mak­ing it impass­able. Prop­er­ties on Gyn­ack Street, Spey Street and Kin­gussie High School were all threatened;

• Janu­ary 1989 and Feb­ru­ary 1990- Levels in the River Spey reached 224.27m and 223.87m respect­ively, flood­ing fields and part of Kingussie.

Pho­to­graphs of some past flood events can be seen in Appendix B.

2.3 Hydro­power scheme

Pit­main Estates have installed a hydro­power scheme in the upstream reach of the Gyn­ack Burn, part of which includes a diver­sion chan­nel which was installed in con­junc­tion with THC. This scheme diverts water from the main chan­nel by means of a lat­er­al weir arrange­ment. Flow then travels down the diver­sion chan­nel to Loch Gyn­ack where it is atten­u­ated by an out­fall weir before being used for energy gen­er­a­tion. Levels in excess of the out­fall weir are dis­charged back into the Gyn­ack Burn via a small channel.

This scheme was primar­ily imple­men­ted for energy gen­er­a­tion but a sec­ond­ary bene­fit of flood peak atten­u­ation is also real­ised through the diver­sion of flows into Loch Gyn­ack. This diver­sion may be able to impact both peak flow as well as peak timing.

Due to bank erosion on the diver­sion chan­nel, the hydro­power scheme is cur­rently not oper­a­tion­al and is there­fore also not atten­u­at­ing peak flows. Bank erosion pro­tec­tion is cur­rently being designed and for the pur­poses of this study, the hydro­power scheme is assumed to be oper­a­tion­al. A sens­it­iv­ity test will be under­taken to assess the impact should the scheme not be operational.

Draw­ings of the scheme can be found in Appendix C. Fig­ure 2 – 2 dis­plays a schem­at­isa­tion of the hydro­power scheme.

AECOM 10

Fig­ure 2 – 2: Schem­at­isa­tion of the hydro­power scheme

AECOM 11

1. Flu­vi­al Hydro­lo­gic­al Assessment

Giv­en pre­vi­ous stud­ies had been under­taken in the area, they were first assessed to determ­ine wheth­er the find­ings were suit­able for use with­in this study. After dis­cus­sions with SEPA, it was agreed that the hydro­logy under­taken in the 2015 FRA was appro­pri­ate for use in this study. The delin­eated sub­catch­ments for the Gyn­ack Burn catch­ment, along with their catch­ment descriptors and the FEH Rain­fall Run­off derived peaks at the con­flu­ence with the River Spey for recon­cili­ation pur­poses were there­fore used with­in this study.

As the hydro­power scheme mod­el has been updated from the 2015 study, small amend­ments were made to the subcatchment’s SPR val­ues so that recon­cili­ation of flows to the 2015 peaks at the con­flu­ence with the Spey could be achieved. Cli­mate change was also updated in line with cur­rent guidance.

Below is a sum­mary of the hydro­logy used in this study. Full details can be found in ​‘Kin­gussie Flood Study Update, 2015’.

3.1 Meth­od­o­logy overview

The Flood Estim­a­tion Hand­book (FEH) gives guid­ance on rain­fall and river flood fre­quency estim­a­tion in the UK and also provides meth­ods for assess­ing the rar­ity of not­able rain­falls or floods. A num­ber of meth­ods of flood estim­a­tion are presen­ted, includ­ing the FEH stat­ist­ic­al meth­od and the FEH rain­fall-run­off meth­od. Sub­sequent pub­lic­a­tions have presen­ted the ReFH and ReFH 2 rain­fall-run­off meth­od, updat­ing the FEH rain­fall-run­off method.

The stat­ist­ic­al meth­od con­sists of two parts; estim­a­tion of the medi­an annu­al flood (QMED), i.e. the flood event with an annu­al exceedance prob­ab­il­ity of 50% (1 in 2 year return peri­od), and the deriv­a­tion of a pooled or single- site growth curve. The growth curve is then mul­ti­plied by the QMED estim­ate to provide a flood fre­quency curve for the sub­ject site for a range of AEP events.

This meth­od, under­taken using WIN­FAP soft­ware, relies on deriv­ing a rep­res­ent­at­ive growth curve for the sub­ject site from a pooled group of hydro­lo­gic­ally sim­il­ar catch­ments for which there is gauged inform­a­tion. This means that the accur­acy of the meth­od and res­ult­ing flow estim­ate depends on there being a suf­fi­cient num­ber of sim­il­ar catch­ments con­tained in the gauging sta­tion data­base. Sim­il­ar­ity is judged using a dis­tance meas­ure derived from the dif­fer­ence in flood­plain extent (FPEXT), rain­fall (SAAR) and catch­ment area (AREA) between the sub­ject site and the gauging sta­tion sites. The meth­od assumes that the flood stat­ist­ics with­in the peri­ods of record in the pool­ing group are rep­res­ent­at­ive of the flood­ing régime in the future, i.e. that the data is sta­tion­ary. How­ever, the meth­od is based on actu­al observed flood data, and is there­fore con­sidered to be more robust than the more con­cep­tu­al rain­fall-run­off meth­ods for the major­ity of cases.

The best estim­ate of QMED is determ­ined using flood data at the site if such loc­al data exists. Altern­at­ively, if no such data exists, QMED can be estim­ated from FEH catch­ment descriptors and improved by data trans­fer from a suit­ably hydro­lo­gic­ally sim­il­ar donor gauge.

When the 2015 hydro­logy was under­taken, ReFH2 had just been released and was not accep­ted by SEPA due to the lack of Scot­tish and smal­ler catch­ments. There­fore, the FEH rain­fall-run­off meth­od was under­taken for com­par­is­on with the Stat­ist­ic­al ana­lys­is. SEPA have agreed that for this study, the ReFH2 meth­od was not required to be undertaken.

Giv­en the size of the catch­ment to the con­flu­ence with the river Spey is 22km², both the stat­ist­ic­al and rain­fall- run­off meth­ods were suit­able and both were under­taken for com­par­is­on. Flow estim­ates for the whole catch­ment were determ­ined at the down­stream extent of the study area for flow recon­cili­ation pur­poses. A range of return peri­ods were required. These included the 50%, 20%, 10%, 5%, 2%, 1%, 0.5%, 0.2% and 0.1% AEP events.

The chosen meth­od would pro­duce peak flow estim­ates at the down­stream extent of the mod­el that could then be used to recon­cile the vari­ous sub­catch­ment inflows. Recon­cili­ation is a use­ful means of estab­lish­ing flows as estim­ates are based on the over­all lar­ger catch­ment, rather than the smal­ler sub­catch­ments, which reduces uncer­tainty in the outputs.

Through­out this report, flood­ing events will be described in terms of their Annu­al Exceedance Prob­ab­il­ity (AEP). Table 3 – 1 sets out how these AEP events cor­res­pond to flood return periods.

AECOM 12

Table 3 – 1: AEP and return peri­od equivalent

Annu­al Exceedance Prob­ab­il­ity (AEP) event Return Peri­od 50% 2 20% 5 10% 10 4% 25 2% 50 1% 100 0.5% 200 0.2% 500 0.1% 1000

3.2 Catch­ment descriptors

In the 2015 Study, the catch­ment descriptors were obtained from the FEH CD-ROM. Table 3 – 2 dis­plays the key catch­ment descriptors of the Gyn­ack Burn to the con­flu­ence with the River Spey.

Fur­ther details regard­ing the catch­ment and it’s descriptors can be found in the 2015 report.

Table 3 – 2: FEH catch­ment descriptors to con­flu­ence with River Spey Total catch­ment descriptors Catch­ment para­met­ers NGR 275700,800550 AREA 21.85 ALT­BAR 567 ASP­BAR 136 ASPVAR 0.4 BFI­HOST 0.413 DPL­BAR 6.19 DPS­BAR 180.8 FARL 0.95 LDP 10.37 PROP­WET 0.68 SAAR 1230 SAAR4170 1206 SPRHOST 56.96 URBEXT1990 0.0003

3.3 Total catch­ment hydrology

Table 3 – 3 dis­plays the 2015 Flood Study peak flows estab­lished using the Stat­ist­ic­al and Rain­fall-run­off methods.

The rain­fall-run­off meth­od was found to pro­duce the high­er of the two estim­ates. Giv­en the rel­at­ively small size of the catch­ment, and there­fore lack of sim­il­ar catch­ments with­in the Stat­ist­ic­al ana­lys­is, as well as the atten­u­ation upstream, the FEH RR meth­od was deemed appro­pri­ate. This approach was con­firmed by SEPA for use in this study.

AECOM 13

The FEH RR peaks were there­fore used to recon­cile the sub­catch­ment flows at the down­stream extent of the model.

Table 3 – 3: Peak flow estim­a­tion from the 2015 Flood Study to the down­stream extent of the Gyn­ack Burn

Annu­al Exceedance Prob­ab­il­ity Stat­ist­ic­al Ana­lys­is Peaks FEH R‑R Peaks (AEP) event (m³/​s) (m³/​s) 50% 12.67 14.59 20% 16.78 20.33 10% 19.72 24.49 4% 23.93 30.58 2% 27.52 36.06 1% 31.55 41.28 0.5% 36.11 46.93 0.2% 43.09 57.16 0.1% 49.22 67.20

3.4 Cli­mate change

Whilst the ori­gin­al cur­rent day hydro­logy was deemed suit­able for use in this study, cli­mate change uplifts had to be reas­sessed due to updated research and guidance.

The United King­dom Cli­mate Pro­jec­tions 2018 (UKCP18) data­set was pub­lished in Decem­ber 2018 and out­lines updated prob­ab­il­ist­ic pro­jec­tions of cli­mate change impact for the 2020’s, 2050’s and 2080’s based on vari­ous emis­sions scen­ari­os and prob­ab­il­ity per­cent­iles. United King­dom Cli­mate Pro­jec­tions (UKCP09) is a pre­vi­ous ver­sion of the pro­jec­tions and has been super­seded by the 2018 projections.

Out­lined in their Flood Mod­el­ling Guid­ance for Respons­ible Author­it­ies, SEPA com­mis­sioned CEH to under­take a study assess­ing Scot­tish catch­ments vul­ner­ab­il­ity to cli­mate change. With­in this study the UKCP09 uplift pro­jec­tions were run through mod­els to provide flow uplifts for hydraul­ic basins. This exer­cise has not how­ever been under­taken using the UKCP18 data. It is there­fore still appro­pri­ate to con­sider using uplift val­ues from the CEH report in Flood Studies.

SEPA have also pub­lished guid­ance on cli­mate change allow­ances for Flood Risk Assess­ments for new devel­op­ments. Whilst the guid­ance for Flood Stud­ies applies a more adapt­ive approach to cli­mate change uplift, rather than set val­ues as spe­cified for FRAs, it is use­ful to under­stand the uplifts applied across all guid­ance to gath­er a com­plete picture.

Based on SEPA’s FRA guid­ance, an uplift of 24% flow or 35% rain­fall is recom­men­ded for the area around Kin­gussie. Giv­en the size of the catch­ment, below 30km², the FRA guid­ance recom­mends than rain­fall be uplif­ted so the flow uplift can be dis­coun­ted in this case. For com­par­is­on, the CEH report, which is based on UKCP09 data, states uplifts of 24% and 33% for the 67^th and 90^th per­cent­ile high emis­sion scen­ario 2080s respect­ively. It should be noted that the CEH report fig­ures are uplifts on flow rather than rain­fall so is not dir­ectly comparable.

For the pur­pose of this flood study, a 35% uplift in rain­fall, in line with SEPA’s FRA guid­ance, has been adop­ted for the cli­mate change scen­ario which is com­par­able to a 90^th per­cent­ile high emis­sion uplift. This is con­sidered to be a rel­at­ively con­ser­vat­ive uplift. As a fur­ther sens­it­iv­ity check, a 20% uplift in flow will be applied as set out in SEPA mod­el­ling guidance.

3.5 Flu­vi­al sub-catch­ment hydrology

With­in the 2015 mod­el­ling exer­cise, sub­catch­ment inflows were applied to sev­er­al points in the mod­el rather than apply­ing the full catch­ment flow at the upstream extent. Apply­ing the full catch­ment flow at the upstream extent of the mod­el was con­sidered to be overly con­ser­vat­ive and would mean that the effects of the hydro­power diver­sion scheme could not be prop­erly assessed.

The hydro­lo­gic­al assess­ment dis­cussed Sec­tion 3.2 and 3.3 was under­taken for the total catch­ment of the Gyn­ack Burn to the down­stream bound­ary of the mod­el. These peak flows were cal­cu­lated so that the sub­catch­ment flows could be recon­ciled in the mod­el, to match this down­stream peak estimate.

AECOM 14

The fol­low­ing sec­tions out­line the recon­cili­ation pro­cess which was under­taken again in this Flood Study due to the changes in some of the ele­ments of the hydro­power model.

3.5.1 Delin­eation and representation

Sub­catch­ments were defined in the pre­vi­ous 2015 study and remained unaltered with­in this study. Fig­ure 3 – 1 dis­plays the sub­catch­ments that feed into the model.

A total of 5 sub­catch­ments were iden­ti­fied. Some of these sub­catch­ments were water­courses such as the Allt à Bhreac-ruighe and some were run­off areas with no asso­ci­ated water­course. The run­off areas were iden­ti­fied as sep­ar­ate sub­catch­ments to allow flow to be added to spe­cif­ic sec­tions in the mod­el so that flow was not overrep­res­en­ted in the upper portions.

In the pre­vi­ous 2015 Flood Study, catch­ment descriptors were down­loaded for the trib­u­tary areas from the FEH CD-ROM. Run­off area descriptors were estab­lished using an inter­me­di­ate catch­ment assessment.

Each sub­catch­ment was rep­res­en­ted using an FEH unit which provided both the peak flow for each sub­catch­ment as well as hydro­graph shape.

Fig­ure 3 – 1: Sub­catch­ments and inflow loc­a­tions into model

AECOM 15

3.5.2 ReFH2 reconciliation

The FEH sub­catch­ment inflows were iter­at­ively scaled and run with­in the 1D upstream baseline mod­el (no hydro­power scheme) by modi­fy­ing the SPRHOST val­ues by the same per­cent­age across all sub­catch­ments, until the flow at the down­stream extent of the mod­el matched the peak flow estim­ates out­lined in sec­tion 3.3.

This res­ul­ted in the FEH SPR val­ues being adjus­ted by between 10% and 18.5% from the down­loaded descriptors. Table 3 – 4 dis­plays the per­cent­age change in SPR val­ues from default and res­ult­ant peak flow at the con­flu­ence with the Spey. Once adjus­ted, these descriptors were used for the upstream mode, with and without the hydro­power scheme, which would both be used to determ­ine inflows into the 1D/2D mod­el of the town developed for this Flood Study. It should be noted that the upstream inflows exclud­ing the hydro­power scheme were to be used as a sens­it­iv­ity check and that the mod­el that includes the hydro­power scheme gen­er­ated the baseline flows for the 1D/2D mod­el through­out town.

Table 3 – 4: Sub catch­ment ReFH2 inflow uplifts

AEP event % % decrease of default SPR value to Res­ult­ant peak flows at down­stream match down­stream flow estim­ate extent of mod­el (m³/​s) to match FEH flows 50% 10 14.6 20% 13 20.4 10% 13 24.4 4% 14 30.5 2% 14 36.2 1% 15 41.4 0.5% 17 47.0 0.2% 17 57.1 0.1% 18.5 67.2

3.6 Flood Study inflows

The flow hydro­graphs at sec­tion GYNA01_1016 of the upstream hydro­power scheme mod­el were extrac­ted and used as the inflow to the sep­ar­ate 1D/2D mod­el that was con­struc­ted as part of this Flood Study. This node was selec­ted as it was the fur­thest sur­veyed upstream sec­tion in the 1D/2D mod­el and all flow was with­in chan­nel. Table 3 – 5 dis­plays the flows that will be used in this Flood Study.

AECOM 16

Table 3 – 5: Peak flows to be used in the mod­el­ling exercise

AEP event % Peak flows into the 1D/2D mod­el (m³/​s) 50% 11.84 20% 16.73 10% 20.21 4% 25.29 2% 30.43 1% 34.79 0.5% 40.20 0.5% + CC 47.58 0.2% 50.26 0.1% 59.47 0.1% + CC 70.51

AECOM 17

1. Joint prob­ab­il­ity

4.1 Run matrix

The Gyn­ack Burn flows into the River Spey down­stream of Kin­gussie. Water levels in the River Spey have the poten­tial to affect water levels in the Gyn­ack Burn, and there­fore increase flood­ing in Kin­gussie; the worst-case situ­ation being con­cur­rent peak flow and matched return peri­ods in both watercourses.

Because of the sig­ni­fic­ant dif­fer­ence in catch­ment areas, the storm dur­a­tion that res­ults in the largest peak flows on the Gyn­ack Burn is likely to be sig­ni­fic­antly short­er than the River Spey. Flood events on the River Spey will arise from longer dur­a­tion, and there­fore less intense, rain­fall than the short dur­a­tion (likely spa­tially lim­ited) high intens­ity rain­fall that would give rise to the highest flows in the Gyn­ack Burn. This means that it is rel­at­ively unlikely that a sim­il­ar return peri­od event would occur on both water­courses at the same time and also rel­at­ively unlikely that the peaks would be con­cur­rent. Whilst event match­ing is unlikely, it is import­ant to explore the effect of the down­stream bound­ary fur­ther so that appro­pri­ate levels are applied.

An ana­lys­is of joint prob­ab­il­ity of coin­cid­ence in peaks would require gauged data on all water­courses. How­ever, as this data is not avail­able on the Gyn­ack Burn, a sens­it­iv­ity assess­ment will be under­taken using a range of sim­u­la­tions to determ­ine how levels in the Gyn­ack Burn are influ­enced by levels in the River Spey.

The below sim­u­la­tions were run as they covered the extremes of the scen­ari­os. The Gyn­ack 50% AEP/​Spey 0.5% AEP was not run after assess­ing the Gyn­ack 50%/Spey 3.33% AEP run.

Low flow in Gyn­ack Burn High flow in Gyn­ack Burn

• Gynack50% x Spey­bank­full Gynack0.5% x Speybankfull
• Gynack50% x Spey3.33% Gynack0.5% x Spey3.33%
• Gynack0.5% x Spey0.5%

The res­ults of the above sim­u­la­tions were com­pared to estab­lish wheth­er water levels in the River Spey impacted water level in the Gyn­ack Burn dur­ing both a lower and high­er event.

4.2 River Spey levels

As no new sur­vey was under­taken on the River Spey as part of this Flood Study, the chan­nel was not mod­elled in 1D and flow was not applied to the water­course. Instead, water levels for the River Spey were taken from the A9 1D/2D mod­el, provided by Fairhurst. These levels were provided between 500m upstream of Ruthven Road and the A9 cross­ing down­stream at reg­u­lar cross sec­tion spa­cing for the 3.33% and 0.5% AEP events. A level to rep­res­ent bank full was also derived from the LiDAR.

Approx­im­ately 1.5km of the River Spey was included in the 2D ele­ment of the mod­el and water levels were seen to vary. For that reas­on, 2 levels on the River Spey were selec­ted from the Fairhurst levels; one at the upstream extent and one at the con­flu­ence with the Gyn­ack Burn. These levels were then applied along the length of the Spey, with the Gyn­ack Burn dic­tat­ing where one level stopped and one level began as shown in Fig­ure 4 – 1.

AECOM 18

Fig­ure 4 – 1: River Spey level boundaries

4.3 Joint prob­ab­il­ity conclusions

When assess­ing the 0.5% AEP event on the Gyn­ack Burn, chan­ging the down­stream bound­ary on the River Spey for the 3 events out­lined in sec­tion 4.1, was found only to impact on the Gyn­ack water levels up to cross sec­tion 18, 70m down­stream of Kin­gussie High School. Water levels were found to match upstream of sec­tion 18 regard­less of levels on the River Spey.

Equally, when assess­ing the 50% AEP event on the Gyn­ack, water levels were found to only be influ­enced down­stream of sec­tion 19, which is 150m down­stream of Kin­gussie High School, when alter­ing the River Spey level between a bank­full and 3.33% AEP event. Levels upstream of sec­tion 19 were unaf­fected regard­less of levels on the River Spey.

Based on these find­ings, it was con­cluded that water levels on the River Spey did not affect water levels on the Gyn­ack through­out the town and there­fore did not affect flood­ing with­in Kin­gussie which is the main focus of this study. It was deemed to appro­pri­ate to apply bank­full levels on the River Spey for all mod­elled return peri­ods on the Gyn­ack Burn.

Floodmaps of the Spey only levels can be found in Appendix H.

AECOM 19

1. Hydraul­ic modelling

The mod­el­ling has been split into