Tools for Flood Management

Guna Paudyal

Senior Flood Management Specialist Managing Director, DHI-(India)

Knowledgebase

(data systems)

Analysis &

Modelling

Decision making &

Communication

Land-use Planning

& control

Structural

Measures

Flood

Preparedness

Emergency

Management

Tools for

Integrated

Flood

Management

Learning from past

• A judicious combination of structural flood mitigation and preparedness programs supported by information technology reduces flood disaster risks and minimizes people’s suffering

flood management decision making

Questions asked?

•What are the effects of predicted rainfall?

•What will be the water levels the next 3 days?

•Where are critical locations?

•How to operate the reservoirs?

•Controlled embankment failures? Where?

Planning

•Repair and reinforcement of embankments: where?

•How to plan for controlled flooding?

Military specialists

blew up dikes to

divert swollen cities

& save cities from

ragging floods

Integrated Flood Management

• Comprehensive amounts of data

• Unlimited number of solutions

• Need tools for analysing and decision

making

Mathematical Models

Flood Management investigations involve:

Examples Tools

Example: Krishna-Bhima basins

Maharashtra: RTDSS

Inflow forecast

Reservoir operation

Optimization

Food forecasting

Warning dissemination

Benefits

Flood disasters in Maharashtra

Maharshtra: Krishna-Bhima basins

floods of 2005-2006

District Human Losses Cattle Losses

2005 2006 2005 2006

Satara 11 23 156 239

Sangli 13 19 224 23

Kolhapur 26 26 236 80

46 major and medium reservoirs

Operated with rigid operational rule

curves: keep the reservoirs full

towards the end of rainy season.

But when heavy rain occurs in

catchments, then the reservoirs

are operated releasing sudden

floods downstream causing

damaging floods.

High Level Government commission:

Floods of 2005 and 2006 were devastating, strong needs of specific

forecasts and early warning were felt. Reservoir operations should

consider downstream flooding more explicitly.

District

Crops & Infrastructure

Land/ Structure Damages

Horticulture Total

Farmers State Crops Govt.

Nursery

Satara 329.3 2.3 63.7 152.9 0.5 548.7

Sangli 744.4 3.5 148.2 98.6 0.4 995.1

Kolhapur 2080.5 26.8 125.7 152.0 1.3 2386.3

District Human Losses Cattle Losses

2005 2006 2005 2006

Satara 11 23 156 239

Sangli 13 19 224 23

Kolhapur 26 26 236 80

Flood losses in one year (2006) 238 crore / yr.

Total Project cost = 31.85 Crores

Cost of setting up RTDAS (telemetry 300 stations) = 23 crores

Cost of forecasting system incl., software, datatbase, capacity bldg.,

(completed in 18 months + 2 years support = 8.15 cr

Cost of software (1 set of MIKE11 RT= Rs. 16 lakhs only (less than

0.50 %)

Optimized reservoir operation during flood emergencies

These losses

could have been

avoided

Krishna River Forecasting (Maharshtra)

300 telemetry

stations

Comprehensive Knowledge base system (linked to RTDAS

Reservoir operation decision

making - interactive

Dissemination system

Flood Bulletin

SMS & E-mail

alerts

15

System tetsing: 2013, 2014

monsoons

16

Day-to-day operation results

17

System performance

18

System performance 2014

Ujjani Reservoir

Date Observed

1 Day Forecast 2 Day Forecast 3 Day Forecast

Forecaste

d Differen

ce Forecaste

d Differenc

e Forecaste

d Differenc

e

01-08-2014 491.12 491.64 -0.52 491.94 -0.82 492.20 -1.08

02-08-2014 491.77 492.13 -0.36 492.33 -0.56 492.55 -0.78

03-08-2014 492.11 492.27 -0.16 492.47 -0.36 492.73 -0.62

04-08-2014 492.20 492.33 -0.13 492.54 -0.34 492.82 -0.62

05-08-2014 492.36 492.62 -0.26 493.14 -0.78 493.53 -1.17

06-08-2014 493.20 493.75 -0.55 494.06 -0.86 494.34 -1.14

07-08-2014 493.73 494.04 -0.31 494.27 -0.54 494.51 -0.78

09-08-2014 494.21 494.45 -0.24 494.55 -0.34 494.72 -0.51

10-08-2014 494.33 494.43 -0.10 494.50 -0.17 494.63 -0.30

11-08-2014 494.41 494.49 -0.08 494.53 -0.12 494.65 -0.24

12-08-2014 494.50 494.59 -0.09 494.64 -0.14 494.75 -0.25

13-08-2014 494.46 494.63 -0.17 494.67 -0.21 494.78 -0.32

14-08-2014 494.47 494.51 -0.04 494.55 -0.08 494.64 -0.17

15-08-2014 494.47 494.57 -0.10 494.59 -0.12 494.67 -0.20

16-08-2014 494.46 494.52 -0.06 494.54 -0.08 494.63 -0.17

490.50

491.00

491.50

492.00

492.50

493.00

493.50

494.00

494.50

495.00

7-30 8-1 8-3 8-5 8-7 8-9 8-11 8-13 8-15 8-17

Ujjani- Water Level Comparison

Observed

1 Day Forecast

649.00

650.00

651.00

652.00

653.00

654.00

655.00

656.00

657.00

658.00

7-30 8-1 8-3 8-5 8-7 8-9 8-11 8-13 8-15 8-17

Koyna - Water Level Comparison

Observed

1 Day Forecast

Optimization of Reservoir Operational

short term during flood emergencies

Koyna Dam model

Comparison of Simulated and Observed Discharges for

Koyna Catchment (R2=0.95, Wbl=0.00%

(Obs=5660mm/y, Sim=5660mm/y)) – near perfect

calibration with good rainfall data

6 hrs lead time)

Reservoir Operational Guidance System

Results at Arjunwad (Koyna Complex)

Joy for some Flood trouble for others

Optimum operation during the whole flood season

(Khadakwasala example)

Optimization of Reservoir Operation

(long term operation – planning)

The Krishna model could be replicated to

optimize the operation of Ajwa Reservoir

h Q

Q Q

h

h h

• Saint Venant Equations

• 6 Point Abbott-Ionescu

Finite Difference Scheme

dynamic/diffusive/kinematic

• Looped Network

Q

x

h

t

Q

t

Q

A

x

h

x + b = 0, + + gA = 0

2

Computational tools are well developed.

Hydrodynamic Module, Core of MIKE 11

Knowledge of

hydrology and

Hydraulics is a basic

requirement to a

successful modelling

Modelling of water resources

Hydrologic Model Hydrodynamic Model

Catchments/Watershed

River/Channel

Floods are generated from

Nepal (flashy catchments)

Floods are also generated

form Catchments in Bihar

For effective flood management

• Need to increase lead time of forecast

• Need to increase locations

• Need to know area inundation

• Need to analyse embankment breaches

• Early warning system

• Real time implementation

Example of Bagmati River

Flood forecasting in Bihar

Flood Forecast products

Sample flood map with LiDAR DEM

Accurate Flood

Maps Using

LiDAR DEM

Village wise

inundation

map

40.50

41.00

41.50

42.00

42.50

43.00

26-0

6-2

014

27-0

6-2

014

28-0

6-2

014

29-0

6-2

014

30-0

6-2

014

01-0

7-2

014

02-0

7-2

014

03-0

7-2

014

04-0

7-2

014

05-0

7-2

014

EKMIGHAT FORECAST

COMPARISON

OBSERVED

ONE DAY DHI

FORECAST38.5039.0039.5040.0040.5041.0041.5042.0042.5043.00

26-0

6-2

014

27-0

6-2

014

28-0

6-2

014

29-0

6-2

014

30-0

6-2

014

01-0

7-2

014

02-0

7-2

014

03-0

7-2

014

04-0

7-2

014

05-0

7-2

014

HAYAGHAT FORECAST

COMPARISON

OBSERVED

ONE DAY DHI

FORECAST

Examples of 2014 forecasts for Bagmati

River & Flood plain River & coastal

2-d urban flood 1-d river

modellign 2-d flood

modellign

The state-of-the-art for flood modeling

The

latest:

flexible

mesh 2-d

modellin

g for fast

computat

ion.

Flood Modelling of complex systems -

Chaophraya river Basin: Thailand

Fully

hydrodynamic

2-dimensional

modelling for

better flood risk

assessment

Model with present

structures

Model with future structures

(Flood protection)

Consider structural flood mitigation and non-structural

risk management

89°00´00" 89°30´00" 90°00´00" 90°30´00" 91°00´00" 91°30´00" 92°00´00"

21°30´00"

22°00´00"

22°30´00"

23°00´00"

B A Y O F B E N G A L

(PC170) D:\P0014\P5070\AVDATA\FLDRETRN.APR 27th April'98 *MMH*--A4

Legend :

0.5 - 1 m

1 - 1.5 m

1.5 - 2.0 m

> 2.0 m

0 - 0.5 m

Flood Depth

Coast Line

Polders / Embankments

5 0 5 10 15 20 25 30 Kilom eters

Bangladesh Transverse Mercator Projection

Loca tio n Ma p

N

Flood Depth, No Wind on Land, Return Period 50 Years

>2 m

1-2 m

0-1 m

Cyclone Shelter

Example from Bangladesh: High Risk Area from

flooding due to 100 yaer storm

Storm Surge Hazard Map

Flooding depth

• Quicker evacuation

• Higher evacuation ratio

• Correct direction/route to evacuate

Flood Hazard Map

Simulation of Embankment breach

scenario (Bangladesh)

Modeling of Impact of Sea level rise on

coastal area flooding (Bangladesh)

SLR 15 cm SLR 27 cm

Integrated Real Time DSS Tools Example: Bhakhra Beas Real Time Decision Support System

RTDAS, Real time platform to access RTDAS and other data sources.

Hydrological modelling (rainfall/snow-runoff), hydrodynamic

modelling, inflow forecast, flood forecast, reservoir operation, water

management)

Example of Online data access

© DHI

Other Online Data sources used in modelling –

Quantitative

Precipitation

Forecast

from

RIMES

Satellite

Precipitation

(TRMM)

Snow Cover (MODIS)

Inflow forecast Bhakra/Beas

User defined Operation Scenarios

© DHI

Forecasted release =

14,000 cusecs

WL: 21-14 Nov, 2013

WL: (increase by 0.20 ft.)

Decreased release

= 7,000 cusecs

Long Term Forecasting (Seasonel Forecasting)

Ensemble (28

events) of Inflow

to Bhakra (50%

quantile

highlighted)

Ensemble of

Reservoir Level in

Bhakra (50%

quantile

highlighted)

Optimization of reservoir release from

- To minimise d/s flooding

- Optimisation is performed if the forecast shows maximum water level in one or more

reservoir will exceed at the end of the forecast period.

- The optimisation will suggest some pre-releases from the reservoirs.

Maximum Water Level in Pong Dam (red), Forecast Water

Level (black), and Optimised Water Level (green)

Spill during Actual Event (black) and Spill suggested by

Optimisation (green)

Flood Mapping

Downstream Flood Map:

5, 10 and 15 hours after peak release from Bhakra

Benefits of RTDSS

The year 2013 was similar to 2008. Using the

RTDSS, it was possible to control the flood spills

and store the water in the reservoirs, thus limiting

flood damages, while increasing power

generation.

0

5000

10000

15000

20000

25000

30000

01-S

ep-2

01

4

02-S

ep-2

01

4

03-S

ep-2

01

4

04-S

ep-2

01

4

05-S

ep-2

01

4

06-S

ep-2

01

4

07-S

ep-2

01

4

08-S

ep-2

01

4

09-S

ep-2

01

4

10-S

ep-2

01

4

11-S

ep-2

01

4

12-S

ep-2

01

4

13-S

ep-2

01

4

14-S

ep-2

01

4

15-S

ep-2

01

4

16-S

ep-2

01

4

17-S

ep-2

01

4

Infl

ow

, C

use

cs

Date

Bhakra Inflow

Observed

2014

Large Basin scale flood forecasting

•Do the Affected people get the information?

•Do they understand the forecast ?

•Is the Information useful ?

Conclusion