Sedimentation Patterns in the Ganges-Brahmaputra Delta System*

Irina Overeem1, K. Rogers1, P. Passalacqua2, A. Canestrelli3, S. Cohen4, and K. Matin5

Search and Discovery Article #50994 (2014)**

Posted August 11, 2014

*Adapted from oral presentation given at 2014 AAPG Annual Convention and Exhibition, Houston, Texas, April 6-9, 2014
**AAPG © 2014 Serial rights given by author. For all other rights contact author directly.
1
CSDMS, University of Colorado at Boulder, Boulder, Colorado, USA (irina.overeem@colorado.edu)
2
Department of Geosciences, Penn State, University Park, Pennsylvania, USA
3
Department of Geography, University of Alabama, Tuscaloosa, Alabama, USA
4
Civil, Architectural and Environmental Engineering Department, University of Texas at Austin, Austin, Texas, USA
5
University of Dhaka, Dhaka, Bangladesh

Abstract

The monsoon-driven Ganges-Brahmaputra river system transports large amounts of sediment from the Himalayas to the delta. The two rivers
combined bring ∼1 billion tons into Bangladesh as reconstructed from observations at Hardinge Bridge and Bahabadur gauging stations from
the late 1950's onwards. Downstream spatial distribution of sediment flux into the deltaic distributary channel network and deposition rates
onto the floodplain and delta plain are remarkably unconstrained, yet of critical importance to the understanding of the overall delta sediment
budget. We numerically model daily incoming sediment flux with a distributed hydrological basin model, WBM-SED. The model uses re-
analysis climatology to calculate the water balance and routes water and sediment throughout the Ganges and Brahmaputra drainage basins.
The estimated flux provides a boundary condition to the lowland sedimentary system. We present a simple approach to sediment routing over
the delta distributaries and into tidal channels; we use channel network characteristics to distinguish between three orders of channels and route
suspended load according to their plan view dimensions.

In the tidal delta, we reverse our simple routing scheme with sediment flux coming from the nearshore water. Characteristics of associated
islands, such as nearest-edge to water distance, are determined for each of the characteristic channel order classes. We then use two cross-
sectional process models, AquaTellUS and FV-SED, to calculate cross-channel sediment flux deposited on delta islands during river flooding
and tidal flooding. Monsoonal flooding and a high tidal range are highly efficient mechanisms to re-deposit sediment onto the delta plain,
especially in areas of high channel connectivity in the coastal zone. This finding is corroborated by our field data on sedimentation rates in the
coastal zone, which highlighted that over a single monsoonal season as much as 1 cm/yr of sediment is deposited widespread in the tidally-
controlled areas of the abandoned Western delta. Preliminary field data in the fluvial-dominated reach shows higher sedimentation rates locally
(∼5 cm/yr), but exhibits a more spatially varied sedimentation pattern. These results are comparable to rapid near-channel sedimentation as
indicated by the numerical modeling. Our simplified concept helps highlight unknowns in the delta plain storage term in the source-to-sink
system of the Ganges-Brahmaputra Delta.

Selected References

Allison, M.A., 1998, Historical changes in the Ganges-Brahmaputra delta front: Journal of Coastal Research, v. 14, p. 480-490.

Cohen, S., A.J. Kettner, J.P.M. Syvitski, and B.M. Fekete, 2013, WBMsed, a distributed global-scale reverine sediment flux model: Model
description and validation: Computers & Geosciences, v. 53, p. 80-93

Coleman, J.M., 1969, Brahmaputra River: channel processes and sedimentation: Sedimentary Geology, v. 3, p. 129-139.

Goodbred, S.L., and S.A. Kuehl, 1999, Holocene and modern sediment budgets for the Ganges-Brahmaputra river system: Evidence for
highstand dispersal to flood-plain, shelf, and deep-sea depocenters: Sedimentary Geology, v. 121/3, p. 239-258.

Goodbred, S.L., and S.A. Kuehl, 1998, Floodplain processes in the Bengal Basin and the storage of Ganges–Brahmaputra river
sediment:an accretion study using 137Cs and 210Pb geochronology: Sedimentary Geology, v. 121, p. 239-258.

Islam, M.R., Y. Yamaguchi, and K. Ogawa, 1999, Remote sending analysis of spatial and seasonal variations of suspended sediments in the
Ganges and Brahmaputra rivers and in their estuary: Proceedings of the Thematic Conference on Geologic Remote Sensing, v. 13/2, p. 11.109-
11.116.

Kottke, B., T. Schwenk, M. Breitzke, M. Wiedicke, H.R. Kudrass, and V. Spiess, 2003, Acoustic facies and depositional processes in the upper
submarine canyon Swatch of No Ground (Bay of Bengal), in V. Ittekkot, H.R. Kudrass, D. Quadfasel, and D. Unger, eds., The Bay of Bengal:
SeepSea Research Part II Topical Studies in Oceanography, v. 50/5, p. 979-1001.

Michaels, K.H., H.R. Kudrass, C. Hubscher, A. Suckow, and M. Wiedicke, 1998, The submarine delta of the Ganges-Brahmaputra: Cyclone-
dominated sedimentation patterns: Marine Geology, v. 149, p. 133-154.

Passalacqua, P., H. Sangireddy, and C. Stark, 2013, The analysis of geomorphic features in the digital terrain era: GSA Abstracts with
Programs, v. 45/7, p. 722.

Rogers, K.G., S.L. Goodbred, Jr., and D.R. Mondal, 2013, Monsoon sedimentation on the ‘abandoned’ tide-influenced Ganges-Brahmaputra
delta plain: Estuarine, Coastal and Shelf Science, v. 131, p. 297-309.
Presenter’s notes: One of the largest depositional features on Earth.
 Ganges-­‐Brahmaputra  Delta  System  
Source  to  Sink    
 
Depositional setting
1.  Bengal Basin: tectonically
deformed from continental
collision of India into Eurasia
2.  Asian monsoon: 80% of
Qw and 95% of Qs from May-
Sept. Peaks in August.
3.  Large sediment discharge:
Qs ~ 992 x 106 tons/y
4. ~80 cm sea level set up
5. Tidal range: 2-6 m
Tidal velocities: 1-4 m/s
6. Recurring cyclones every ~2
years
 Estimated  Source  and  Sink  Terms  

1000  
Floodplain storage of 300
million ton/year

300  
Exis7ng  es7mates  based  on  stra7graphic  
analyses,  geochronologic  da7ng  of  core  
material  
100   700  
(KoBke  et  al.  2003;  Goodbred  and  Kuehl  1999;  
Allison  1998;  Michels  et  al.  1998;Rogers  et  al.,   200  
2013)  
 Objective  

ì  What  are  deposi7onal  paBerns  over  a  single  flood  season  

and  over  ~50  years?    

ì  What  are  longitudinal  trends  in  sedimenta7on?  

ì  What  can  simple  numerical  models  tells  us  about  the  
lateral  distribu7on  of  sediment  over  the  delta  plain?  
 Channel  Widths  
88°E 89° 90° 91°

ì  Channel  widths  in  Ganges-­‐ log(A)

Brahmaputra  delta  vary  over  4  

10.57
4.40
orders  of  magnitude   24°
N

ì  Largest  channels  5-­‐7  km,  small  

7dal  creeks  ~1m  
23°

ì  We  use  the  network  analysis  to  

quan7fy  dimensions  for  3  
channel  orders  to  scale  floodplain  
22°
N

sedimenta7on  experiments   0 25 Km 100

  Figure 15: Channel networkColormap  

of the oGBJ
f  Island  
Delta area,  
with only  channels>57m  
channels of width ≥
(Passalacqua  
(colormap based on the logarithm of island et  area).
al.,  2013)  
The effect of eliminating t
 0 25 Km 100

Figure 11: Channel network of the GBJ Delta and map of island aspect ra

Nearest-­‐Edge  Distance  to  Channel  

γmin = 2.7. The majority of the elongated islands is located near the co
main rivers (Ganges, Padma, Meghna), but the distribution of aspect ratio
heterogeneous.

Fluvial  

Tidal  

ap of the logarithm of nearest-edge distance

rtest straight-line distance from the nearest
Figure 12: Channel networkNearest-­‐Edge  
of the GBJ Delta Distance  
and map to  Channel  (L)  
of the logarithm of ne
ny land pixel within the delta. The smallest
(Passalacqua  
L. The nearest-edge distance is calculatedet  as
al.,  
the 2013)  
shortest straight-line distanc
Presenter’s notes: WBMsed is a distributed global-scale riverine sediment flux model. Model description and validation by
Sagy Cohen, Albert J. Kettner, James P.M. Syvitski, and Bala´zs M. Fekete (2013, Computers & Geosciences, v. 53, p.80-93).
 WBM-­‐water  and  sediment  load    
daily  variability  
4
x 10
12
Ganges River

Ganges  River        
Brahmaputra River
Meghna River

10 Brahmaputra  River  
Daily  Discharge  in  m3/sec  

Meghna  River  
Daily Average Discharge in m3/s for 1960

0
0 50 100 150 200 250 300 350 400
DOY

All rivers are dominated by the Indian Summer Monsoon.

Model predictions compare well with observed data.
(e.g Coleman, 1969; Webster et al., 1998; Islam et al, 1999)
 Intra-­‐annual  variability  in  influx  of  
water  and  sediment  –  50  years  
11
x 10
10
Ganges River

Ganges  River        
linear
Brahmaputra River
9

8
Brahmaputra  River  
7
Annual Total Water Discharge (m3)

y = − 1.7e+09*x + 3.8e+12
4

0
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010
Simulation Year

1960                                                                        1980                                                                                2000                                                              

Brahmaputra  mean    Q=2.01  *  104  m3/sec  (Jian  et  al.,  2009)  

WBM  predicts      QB  =  2.31*  104  m3/sec  
Ganges  mean      Q=1.14  *  104  m3/sec  (Jian  et  al.,  2009)    
WBM  predicts      QB  =  1.26*  104  m3/sec  
 Model  the  fluvial-­‐dominated    delta  

Variable Discharge
Sea level (t) and Sediment input (t)
WBM model dictates
magnitude and
variability of annual
flood
Flowpath (t)
Simple approach:
models flood events
only, no tides, no
storms.
 AquaTellUs  Model    
∂Hx ∂F ∂Fer m
= = kc ( x ,t ) S( x ,t ) Q ( x ) ( t )
∂t ∂x ∂x
Topography (H) depends on Erosion depends on slope (S) and
sediment flux (F) discharge (Q) in fluvial domain,
F in  (0 ) grainsize-independent

F in  (x)

∂F k
F d e p o  (x) F o ut  (x)
sed ( x ,t )
= sed
F
∂x u
( x ,t )

F e ro  (x) ( x ,t )

Sedimentation depends on
sediment flux (F) en de
x -1 x x + 1
streampower (u), ksed is
grainsize dependent.
downstream
stroomafwaarts
 Lateral  Sedimentation  
Basic principles of sedimentation across channel belt and floodplain:
exponential with distance from channel (Pizutto, 1987; Goodbred &
Kuehl, 2000).

Variability in floods creates Gaussian distribution; and error function

solution (Paola, 2000; Overeem, 2005).

−(y − µ )2
1 2 −t 2
F( y) =
2 πσ
e 2σ 2 erf ( y) =
π
∫e dt

y = horizontal distance normal to channelbelt

σ = standard deviation across sedimentation zone
µ = position of flowpath axis
€
 Flood  deposition  maps-­‐  50  yrs  
Cumulative 50 Year Flood Event Deposition in m

2nd  order  
1.5

X  =  10  
channels  
20

40
X  =  40   1
(1-­‐2  km)  
 
60

Deposition in m
90  km  

80

X  =  90   1)  Lateral  

sedimenta7on  
0.5

100

>2500  m,  h  ~  
1.5m  
120

2)  Rapid  
0
140

160
sedimenta7on  
causes  natural  
avulsions,  
180 −0.5

60  km  
20 40 60 80 100 120
 50  yr  deposi7on  >1.5m  
Overbank  10 sands  amalgamate   X  =  10  
g  
nin

9
m  Fi

60
65
strea

45
55
40 50

8
75
70 80
n
Dow

7
50  yr  deposi7on   ~  0.2  -­‐1m  
X  =  40  
2nd  order  channels  disconnected  
6 60
45
55
65 70
40 85
75 80

5 50

50  yr  deposi7on  ~  0.1  -­‐0.7m  

2nd  order  channels  disconnected   X  =  90  
1
35

65
55
40
80 85
50 60

30
25 45
75

70
90

0
30 60 90 120 150 180 210 240 270 300
Median Grainsize [mu]
 Flood  deposition  maps  –  50  years  
Cumulative Flood Event Deposition in m
1.5

20
3rd  order  channels  
40
(500-­‐250m)  
 
1
90  km  

60

80
Lateral  deposi7on  s7ll  
0.5
~1500m,  but  finer  grainsizes  
 
100

120
Connec7vity  of  sediments  is  
140
0
lower,  due  to  more  stable  
160
channel  belts  
 
 
180 −0.5

60  km  
20 40 60 80 100 120
 Observations  from  Sediment  Traps  

               
                                               
2012 mixed tidal-fluvial delta
plain sediment trap sites
 Observed  Sedimentation  Rates    
9       mass  flux   1.4  

8      7Be  inventory  
1.2  
Mass  accumulaFon  (g  cm-­‐2)  

7Be  inventory    (dpm  g-­‐1  cm-­‐2)  

7  
1  
6  

5   0.8  

4   0.6  

3  
0.4  
2  
0.2  
1  

0   0  
5.1.D   5.2.D   5.3.B   7.1.A   7.2.D   7.3.A   8.1.A   8.1.C  

Regional  ver7cal  accre7on:    2.3±0.9  cm  y-­‐1  

 Conclusions  and  Future  Steps    

ì  Sedimenta7on  rates  in  7dal  delta  are  ~1.1  cm/yr.  Preliminary  results  for  fluvial-­‐dominated  
part  of  delta  plain  are  as  much  as  2.3  cm/yr.  

ì  Modeled  sedimenta7on  has  a  strong  longitudinal  grain  size  trend;  highest  aggrada7on  and  
sandiest  near  ‘apex’  and  again  near  coastal  boundary.  Model  predicts  strong  downstream  
fining;  predominantly  a  consequence  of  sediment  availability.  Sil7er,  lower  channelbelts  
occur  towards  the  coastal  floodplain.    

ì  Model  predicts  significant  interconnec7vity  of  sandy  sediments  in  2nd-­‐order  channel  
simula7ons,  channels  or  3rd  order  contribute  to  floodplain  but  appear  more  isolated  treads..  

ì  In  future:      combine  nearest-­‐edge  maps,  channel  width  classes  and  channel  sedimenta7on  
paBerns.  

ì  In  future:  design  similar  experiments  with  7dal  channel  model  (FV-­‐Sed)