Terrestrial Flood Risk and Climate Change in the YallahsRiver, … · 2015. 8. 18. · INTRODUCTION Types of flooding in Jamaica 1. PAST FLOOD EVENT 2. RETURN PERIOD ANALYSIS 3 Wilson

Terrestrial Flood Risk and Climate Change in the Yallahs River, Jamaica

An assessment of future flood risk using hydrodynamic models driven by using hydrodynamic models driven by

projections of future climate

Matthew Wilson1, Arpita Mandal2, Michael Taylor3,

Christopher Burgess3, Jayaka Campbell3 and Tannecia Stepphenson3

1. Department of Geography, University of the West Indies, St. Augustine, Trinidad & Tobago

2. Department of Geography and Geology, University of the West Indies, Mona, Jamaica

3. Climate Studies Group, Department of Physics, University of the West Indies, Mona, Jamaica

Introduction

• Flooding is one of the major natural hazards affecting Jamaica.– Has been negatively affected by severe weather events

– Repeated flooding (2004, 2007, 2008, 2010) has been very costly both in terms of lives and livelihoods.

• Analysis of the number of hurricanes/ tropical storms affecting Jamaica and the variability of the rainfall pattern

INT

RO

DU

CT

ION

1. P

AS

T F

LOO

D E

VE

NT

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

Wilson et al., Terrestrial Flood Risk and Climate Change in the Yallahs River, Jamaica

WCRP VAMOS/CORDEX Workshop on Latin-America and Caribbean. CORDEX LAC: Phase I - South America. September 11-13, 2013, Lima, Perú 2

affecting Jamaica and the variability of the rainfall pattern has shown changes in trends in the years 2001-2012. – Shifts in traditional peak rainfall months of May and September-

October has shifted to June and November in some years.

– Shorter duration high intensity rainfall associated with stationery fronts and troughs

– IPCC outlook shows an increasing likelihood of more intense hurricanes, which would result in increased frequency of flooding due intensive rainfall.

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

Groundwater induced Depression Riverine

Types of flooding in Jamaica

INT

RO

DU

CT

ION

1. P

AS

T F

LOO

D E

VE

NT

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3Wilson et al., Terrestrial Flood Risk and Climate Change in the Yallahs River, Jamaica

WCRP VAMOS/CORDEX Workshop on Latin-America and Caribbean. CORDEX LAC: Phase I - South America. September 11-13, 2013, Lima, Perú

Storm Surge Urban Runoff

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

GRAPHIC ID 1

Reported flood events

INT

RO

DU

CT

ION

1. P

AS

T F

LOO

D E

VE

NT

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS



Data from Mona Geoinformatix Institute, CEAC Solutions

Ltd, ODPEM and Water Resources Authority of Jamaica

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

GRAPHIC ID 2

Study site: Yallahs River

INT

RO

DU

CT

ION

1. P

AS

T F

LOO

D E

VE

NT

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS



Title of Graphic : Upland catchment in south-east

Jamaica, draining the Blue Mountains.

Major riverine flooding associated with Hurricane

Gustav (28 August 2008).

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

GRAPHIC ID 3

1. Past flood analysis 2. Current flood risk 3. Future flood risk

24 hr rainfall for

Tropical Storm

Gustav

(28 August 2008)

Historical rainfall data (24 hr, 1992-2012, Mavis Bank gauge)

Extract annual 24 hr maximum

Fit Gumbel distribution

24 hr rainfall for 25, 50,

100 yr return periods

Baseline climate model: PRECIS

IDF Curves:

present

Future climate

model: ECHAM

Scaling of current return

periods

IDF Curves:

future

IDF Curve difference

Title

Flood

analysis

method

summary

INT

RO

DU

CT

ION

1. P

AS

T F

LOO

D E

VE

NT

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

GRAPHI

C ID 4



Temporal downscaling

(NRCS Type II/ III distribution)

Inundation extents, depths and flood wave timings

Gustav catchment

model

Current return periods

catchment modelHEC-HMS

Future return periods

catchment model

Topography

River network

Land use/ cover

Water balance

Gustav hydraulic modelCurrent return periods

hydraulic modelLISFLOOD-FP


hydraulic model

Floodplain

topography

Channel

geometry



15 min. rainfall

Channel inflow Channel inflow Channel inflow

15 min. rainfall



15 min. rainfall

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

0.000

0.100

0.200

0.300

0.400

0.500

0.600

0.700

0.800

0.900

1.000

0 6 12 18 24

Fra

ctio

n o

f 2

4-h

r ra

infa

ll

Hour

NRCS 24-hour Rainfall Distributions

Type 1

Type 1a

Type 2

Type 3

INT

RO

DU

CT

ION

1. P

AS

T F

LOO

D E

VE

NT

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS



0

10

20

30

40

50

60

70

80

90

0 6 12 18 24

Ra

infa

ll (

mm

)

Hour

Mavis Bank, Jamaica, 28 Aug 2008 – Tropical Storm Gustav (419 mm

24-hour rainfall)

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

INT

RO

DU

CT

ION

1. P

AS

T F

LOO

D E

VE

NT

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS



HEC-HMS

model

structure,

Yallahs River

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

1. Past flood analysis: Tropical Storm Gustav

• Landfall in Jamaica 28 August 2008

• Major impact (US $210 M), including flooding in Yallahs

• 24-hr data for Mavis Bank and Ramble gauges

• NRCS Type II method used to generate 15-minute rainfall

• Flows generated from rainfall

0

20

40

60

80

100

0 6 12 18 24

Ra

infa

ll (

mm

)

Mavis Bank (419 mm 24-hour rainfall)

20

40

60

80

100

120

Ra

infa

ll (

mm

)

Ramble (298.4 mm 24-hour rainfall)

1. P

AS

T F

LOO

D E

VE

NT

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

INT

RO

DU

CT

ION



• Flows generated from rainfall using HEC-HMS

0

20

0 6 12 18 24

Hour

http://en.wikipedia.org/wiki/File:Gustav_2008_track.pnghttp://en.wikipedia.org/wiki/File:Gustav_29_August_2008.jpg

Jamaica

Jamaica

http://en.wikipedia.org/wiki/File:GustavCuba.gif

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

1. Tropical Storm Gustav: Results

Maximum flood depth

HEC-HMS

15-minute

rainfall

LISFLOOD-FP

Generation of flow data for upstream

input and subcatchment inflows

Routing of flood wave along main channel

and across floodplain

1. P

AS

T F

LOO

D E

VE

NT

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

INT

RO

DU

CT

ION



(m)

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY



Tropical Storm Gustav:

Prediction of flood extent and

depth over time

(m)



depth over time



(m)



depth over time



(m)



depth over time



(m)



depth over time



(m)



depth over time



(m)





depth over time

(m)





depth over time(m)

1. Tropical Storm Gustav: ResultsTime of initial inundation Time of maximum inundation

Early flooding

in narrow

section of

valley (before

flood peak)

Flood peak

upstream

occurred ~1 hour

it reached

downstream

1. P

AS

T F

LOO

D E

VE

NT

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

INT

RO

DU

CT

ION



Hours(from start)

Hours(from start)

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

2. Estimation of current flood risk

• Daily data (24-hour) obtained for 1992-2012 from Jamaica Met Service for Mavis Bank and Ramble– Data prior to 1992 unavailable due to loss of records in fire

– Some historical data (1922-1966) obtained from “Jamaica Weather Reports” from NOAA central library (http://docs.lib.noaa.gov/rescue/data_rescue_jamaica.html)

– Annual maximum extracted:

600

1. P

AS

T F

LOO

D E

VE

NT

INT

RO

DU

CT

ION



2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

0

100

200

300

400

500

600

RA

INF

ALL

(M

M)

Annual maximum

24-hour rainfall

MAVIS BANK

RAMBLE

Tropical Storm Gustav

Hurricane Ivan

Tropical Storm Nicole

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

2. Estimation of current flood risk

• Return periods calculated for 24-hour rainfall

• 24-hour maximum temporally downscaled using NRCS Type-II rainfall– IDF curves generated for return period calculations

1200.00

IDF Curve, Mavis Bank, data set from 1994-2012

700Return period, Mavis Bank

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

1. P

AS

T F

LOO

D E

VE

NT

INT

RO

DU

CT

ION



0.00

200.00

400.00

600.00

800.00

1000.00

1.00 10.00 100.00 1000.00 10000.00

Ra

infa

ll I

nte

nsi

ty (

mm

/hr)

Duration (minutes)

5-Predicted

10-Predicted

25-Predicted

50-Predicted

100-Predicted

0

100

200

300

400

500

600

0 20 40 60 80 100 120

24

-ho

ur

Ra

infa

ll (

mm

)

Return period (years)

24 hr rainfall estimated from 1946-

1985

24 hr rainfall period 1992-2006

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

Topographically

constrained

flooding

2. Estimation of current flood risk: results

• HEC-HMS run for 25, 50 and 100 year rainfall events

• Flows from HEC-HMS routed through the LISFLOOD-FP model to predict inundation extents and depths

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

1. P

AS

T F

LOO

D E

VE

NT

INT

RO

DU

CT

ION



Right: Map of exceedance probability:

Green: 4% (25 year return period)

Blue: 2% (50 year return period)

Red: 1% (100 year return period)

Main difference

between flood

areas

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

Depth difference:

100-year event maximum

depth minus 25-year

event maximum depth.

2. Estimation of current flood risk: results

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

1. P

AS

T F

LOO

D E

VE

NT

INT

RO

DU

CT

ION



Although the flood

extents are similar,

water depths are up

to 2 m higher on the

floodplain during a

100-year event

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

ISS

UM

MA

RY

3. Estimation of future flood risk

0

2

4

6

8

10

12

14

16

18

0 2 4 6 8 10

Inte

nsi

ty (

mm

/hr)

ECHAM-2071-99, B2

5 yr RP

10 yr RP

25 yr RP

50 yr RP

100 yr RP

0

2

4

6

8

10

12

14

16

18

0 2 4 6 8 10

Inte

nsi

ty (

mm

/hr)

ECHAM-2041-70, B2

0

2

4

6

8

10

12

14

16

18

0 2 4 6 8 10

Inte

nsi

ty (

mm

/hr)

Baseline (PRECIS)

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

1. P

AS

T F

LOO

D E

VE

NT

INT

RO

DU

CT

ION



3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

IS

0 2 4 6 8 10

Duration (days)

0 2 4 6 8 10

Duration (days)

0 2 4 6 8 10

Duration (days)

-100%

-80%

-60%

-40%

-20%

0%

20%

0 2 4 6 8 10

Pe

rce

nta

ge

ch

an

ge

Duration (days)

Change factors for medium (2040 to 2070) and long term (2071 to 2099)

5 RP - mt

10 RP - mt

25 RP - mt

50 RP - mt

100 RP - mt

5 RP - lt

10 RP - lt

25 RP - lt

50 RP - lt

100 RP - lt

• Percentage difference between baseline climate (PRECIS) and future climates (ECHAM) calculated for rainfall events down to 24-hour

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

SU

MM

AR

Y

3. Estimation of future flood risk: results

• 25, 50 and 100 year rainfall scaled by

change factor from ECHAM and

temporally downscaled using NRCS

Present return periods

(analysis of gauge data)

• HEC-HMS run for future return periods and flows routed through the LISFLOOD-FP model to predict inundation extents and depths

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

1. P

AS

T F

LOO

D E

VE

NT

INT

RO

DU

CT

ION




(ECHAM, 2070-2099, B2)3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

IS2

. RE

TU

RN

PE

RIO

D A

NA

LYS

ISS

UM

MA

RY

3. Estimation of future flood risk: results

No reduction in

inundation risk

in valley

2. R

ET

UR

N P

ER

IOD

AN

ALY

SIS

1. P

AS

T F

LOO

D E

VE

NT

INT

RO

DU

CT

ION



Present annual

exceedance probability

2070-2099 annual

exceedance probability

(ECHAM, B1)

Lower risk

downstream

Maps of exceedance probability:

Green: 4% (25 year return period)

Blue: 2% (50 year return period)

Red: 1% (100 year return period)

3. F

UT

UR

E C

LIM

AT

E A

NA

LYS

IS2

. RE

TU

RN

PE

RIO

D A

NA

LYS

ISS

UM

MA

RY

Documents

Terrestrial Flood Risk and Climate Change in the YallahsRiver, … · 2015. 8. 18. · INTRODUCTION Types of flooding in Jamaica 1. PAST FLOOD EVENT 2. RETURN PERIOD ANALYSIS 3 Wilson