Ali Shafqat Akanda, Antarpreet Jutla & Shafiqul Islam

OverviewMotivation and goals of study

Geography and Hydrology

Existing forecasting practices

Seasonal streamflow forecasts

Seasonal Cholera forecasts

Downscaling / Interpretation

MotivationBangladesh is a land of natural calamities

Costly droughts and floods each year

Droughts and floods followed by cholera

Exiting flood forecasting system has a lead-time

of only 2-6 days and no cholera forecasts.

A monthly-to-seasonal forecasts could provide

planning, preparation and preemption time.

Satellite Image of eastern South Asia showing the topography of the region and the Ganges (dark blue) and Brahmaputra (light blue) basin rivers (Courtesy: ESRI)

The Eastern Himalayas

Five countries

India

Nepal

Bhutan

Bangladesh

China

Three mighty rivers

Ganges

Brahmaputra

(Tsangpo in

China)

Meghna

The Ganges-Brahmaputra-Meghna Basin Region

Key facts about the river basins:1) 92% of the basin areas lie outside Bangladesh (thus no infrastructural solutions).2) G-B-M average annual precipitation are 1500, 2500, and 4000 mm, respectively.

Spring Onslaught of Drought

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

m3/

sec

Ganges

Brahmaputra

Spring Onslaught of DroughtLowest flow of the season in Feb and March

Only half a day of rainfall in winter on average

Salt water intrusion in Ganges estuary region

Severe water shortage in Sundarbans mangrove

Spring cholera outbreak in coastal districts

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

m3/

sec

Ganges

Brahmaputra

Then comes the Floods …

Then comes the Floods …

Intense monsoon rainfall in 4 months

(JJAS)

Peak Floods usually in August and

September

Entire country is the floodplain!

Loss of lives, crops and infrastructure

Massive outbreak of diarrheal diseases

More cholera outbreaks in Autumn

(OND)

Existing Forecasting System Flood Forecasting and Warning Centre Bangladesh Water Development Board Issues 24 and 48 hour river stage

forecasts and countrywide flood inundation maps

Using experimental 2-6 days forecasts by CFAB group in Georgia Tech, USA.

No cholera forecasts. Little coordination between Health and Water Departments.

Existing Forecasting System

Research QuestionsWhat are the key processes? Are Ganges, Meghna,

Brahmaputra governed by same moisture sources?

What are the significant predictors? How can the probabilistic forecasts be used with existing ones?

How do these river basins affect cholera outbreak patterns in spring and autumn, separate/together?

How can the forecasts be salient? Which information would be most useful to the local and real users?

Research QuestionsWhat are the key processes? Are Ganges, Meghna,

Brahmaputra governed by same moisture sources?

What are the significant predictors? How can the probabilistic forecasts be used with existing ones?

How do these river basins affect cholera outbreak patterns in spring and autumn, separate/together?

How can the forecasts be salient? Which information would be most useful to the local and real users?

What forecasts do we need?Droughts

Expected time

and amount of

rainfall

Local water

table

fluctuations

If drought year,

for how long

will it be?

Extent of

salinity

intrusion in

South

Floods

Arrival and

intensity of

monsoon rainfall

Crop impacts

(most important

variable!)

Extent of

Inundation area

and duration

Flood

displacement and

rehabilitation.

Cholera

Timing of spring

& autumn peaks.

Expected

Intensity of each

peak

Which areas are

most vulnerable?

Social, economic,

and health

impacts .

The Hydrology of G-B-M

0

10

20

30

40

50

60

70

80

% F

looded

Flood Affected AreaSource: BWDB

-3

-2

-1

0

1

2

3

1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000

No

rmal

ize

d A

nn

ual S

tre

amflo

w

Year

Ganges

FMA

J ASO

-3

-2

-1

0

1

2

3

4

1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005

No

rmaliz

ed A

nn

ual S

tre

amflo

w

Year

Brahmaputra

J FM

J JAS

The Hydrology of Cholera First outbreaks in

Spring (March – May)

during low water

availability

More outbreaks in

Autumn about 2

months after floods

(October – December)

Significant inter-

annual & intra-annual

variability

Role of Hydrology?

Role of Hydrology in Cholera

Role of Hydrology in Cholera

Cholera MAM Mean

Cholera OND Mean

Bay of Bengal JJA SST

Ganges Streamflow -0.20 0.60 0.43

Brahmaputra Streamflow -0.54 0.34 0.42

Combined Flow Volume -0.65 0.55 0.46

Flood Affected Area 0.18 0.56 0.62

Bay of Bengal DJF SST -0.57 -0.23 -

Bay of Bengal JJA SST 0.14 0.70 -

1

Role of Hydrology in Cholera Two separate hydro-climatological

processes may explain the two cholera

peaks separately

Water scarcity causes cholera in Spring,

while Water abundance causes cholera in

Autumn

A stronger drought year shows more

cholera, and A higher flood year also

shows more cholera

ConclusionSeasonal flow and cholera forecasts are vital

Emphasis needed on interpreting forecasts

Relate to local variables: water availability,

drought duration, salinity intrusion areas,

inundation scenarios, agricultural options,

areas more vulnerable to cholera

Presentation of scientific information

Socio-economic-health impacts

Future WorkObtain Complementary Evidence:

Detailed salinity intrusion analysis

Remote Sensing for flood inundation

Examine seasonal phytoplankton patterns

Employ Forecasting ModelChoice of predictor variables

Integrate Biology-Hydrology-Remote Sensing

Assessment of Socio-economic-health

impactsCholera burden/vulnerability analysis

AcknowledgementsTufts School of Engineering, Boston, MA

Water: Systems, Science, and Society

Program at Tufts University, Boston, MA

Bangladesh University of Engineering

and Technology, Dhaka, Bangladesh

International Centre for Diarrhoeal

Disease & Research, Dhaka, Bangladesh