Does sedimentation or erosion trigger
river forestation? A numerical modelling
approach
Takashi Asaeda, Kelum Sanjaya, and Md Harun Or Rashid
Saitama University
5 th International Multidisciplinary Conference on
Hydrology and Ecology
13-16 April 2015 Vienna
Background of the study
Intensive forestation widely occurs in East Asian rivers.
deteriorates the ecosystem of gravelly and sandy bars.
affects flood protection.
changes the landscape of rivers.
Tama River (Tokyo) Ota River (Hiroshima)
Fraction of vegetation coverage
increased in Japanese rivers
1947 1947
2014 2014
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Oko
tsk
Ho
kkai
do
Toh
oku
Jap
an S
ea
Co
ast
San
riku
/Jo
ban
No
rth
ern
Kan
to
Toky
o M
etro
po
litan
Are
a
Toka
i
Ho
kuri
ku
Kii
Pe
nin
sula
Kan
sai M
etro
po
litan
Are
a
San
in
Seto
Inla
nd
Sea
Co
ast
Sou
ther
n S
hik
oku
No
rth
ern
Kyu
shu
Sou
ther
n K
yush
u
vegetation coverage 1946
1975
2010
Frac
tio
n o
f ve
geta
tio
n c
ove
rage
Vegetation cover in 1946
Vegetation cover in 2010
Vegetation cover in 1975
Asaeda et al. (2013) Proc. IAHR
1947 October 1977 December present
Channel is filled with
stony sediment
Low in vegetation
Reduced sediment
+
Channel incision
+
Vegetation invasion
60 yrs transition of the river channel An example of Sagami River near Tokyo
Sediment budget (>0.2mm)
between 1944 and 2010
Sediment harvesting
before 1964
24,000,000m3
Sediment inflow
11,000,000m3
Sagami Dam 7,000,000m3
Shiroyama Dam
2,000,000m3
3,000,000m3
Weirs 180,000m3
Miyagase Dam
1,000,000m3
Remarkable amount of sediment was harvested before
1964 and trapped by dams
500,000m3
Doshi Dam
400,000m3
Fukashiro Dam
50,000m3
Numamoto Dam
100,000m3
Sediment is deficit in the river channel
Total inflow
14.5 milli m3
Total amount
of harvesting
24 milli m3
Total amount
of trapped
10.55 milli m3
Flow into sea
???
Deficiency of
sediment in the
river channel
<<
Deposited area
Eroded area
Eroded area:
exposure of underlying
nutrient rich sediment
Example of Arakawa River
North of Tokyo
Vegetation colonization at deposited and eroded areas
Deposited area:
covered by cleansed sediment
low in nutrient and seed bank
Deficiency of sediment decreases the deposition at flood time
Reduction of deposited areas introduces earlier colonization of vegetation
The vegetation coverage increases even with the same flood frequency
Flooded at Sep 2007
View of 2010 (3 years later)
Deposited
eroded
1999
Deposited
eroded
1996
1993
Before flood
After flood
3 yrs later
Example of Kurobe River
Deposited location
Eroded location
Asaeda et al. River Res. Appl. (2014) Online
Simulation with
Dynamic Riparian Vegetation Model
(DRIPVEM)
Development processes of vegetation communities on the sediment bars
Accumulation of nutrient
Flushing of vegetation and
accumulation of cleansed sediment
Herbs
Trees hydrochory
Increase in biomass
Colonization of trees
Growth with self-thinning
Repeated process of
vegetation flushing, and
the deposition of
cleansed sediment (low
nutrient) or exposure of
surface sediment (same
as previous nutrient) with
floods
For herb development,
edaphic (nutrient and
sediment size) condition
must be processed
after floods
Tree seeds are
dispersed at floods at
floods, then grow
afterwards, decreasing
by self-thinning.
hydrology
herbaceous plants
trees
soil nutrients
flooding
flooding Cleansed sediment deposition or erosion
Recruitment of seeds
Flushing of vegetation
loss of trees
Model structure
shading
loss of herbs
mortality & decomposition
defoliation
Atmospheric fallout
N-fixation
seeds Recruitment of trees
denitrification
remaining trees
d
dt Tree density = Recruitment (flood characteristics) * Self-thinning function
- Flush(inundation, age)
Tree morphology, biomass = F(age)
Model scheme for each mesh
Geomorphology of the river channel, hourly flood level, sediment size : observed data used;
The ground surface was divided into 10mx10m mesh size
Initial condition : no vegetation at 50 years before
Simulation time step: 1 month (with recorded hourly highest flood level in the month)
Tree density and biomass for organs (module TREE)
Herb biomass = F(soil N, particle size, shading by trees)
Herb biomass (module HERB)
d
dt Soil N = Uptake by plants (biomass, N content)
N fixation (biomass, fixed N content)
Atmospheric fallout
-
+
+
denitrification -
+ decomposition (dead biomass, fixed N content)
Soil N concentration (module SOIL)
Observed
Simulated
Example of simulated results
0
10
20
30
40
50
1-5 6-10 11-15 16-20 21-25Com
positio
n o
f tr
ee a
ges
Tree age (yrs)
Observed
Simulated
Distribution of Salix
2010
Observed herb
2010
Simulated herb distribution
Simulated results
Distribution of herbs
Validation of the model
Compatibility of vegetation coverage simulation
to the observed data
Bare soil
0100020003000400050006000700080009000
1000011000
Totalsimulated
mesh
Compatiblemesh
Nu
mb
er
of
me
sh
Herbs
64.7%
75.4% Trees
80.4%
Hii River
Trees
Herbs
Bare soil
0
5000
10000
15000
20000
25000
Totalsimulated
mesh
Compatiblemesh
Nu
mb
er o
f m
esh
79.3%
71.6%
53.7%
Kuzuryu River
Observation Salix
Robinia
Simulation Salix Robinia
Arakawa River
(after 20years simulation)
Hii River (after 50 years simulation)
Observed Simulated
Age comparison
Asaeda et al. Riv Res Appli 2014
0
10
20
30
40
0 5 10 15
Avg
.Bio
mas
s(gd
w/m
2)
Time(years)
0
20
40
60
80
100
120
0 5 10 15
Avg
. b
iom
ass(
gdw
/m2)
Time (years)
0
20
40
60
80
100
0 5 10 15
Avg
. io
mas
s(gd
w/m
2)
Time(years)
0
5
10
15
20
25
0 5 10 15
Avg
.bio
mas
s(gd
w/m
2)
Time (years)
Kurobe River
Kuzuryu River Hii River Sagami River
Delay time of herb colonization
0
0.5
1
1.5
2
2.5
3
3.5
0 0.5 1 1.5 2 2.5 3 3.5
Sim
ula
ted
del
ay t
ime
of
her
b
colo
niz
atio
n (
year
s)
Observed delay time of herb colonization (years)
Kuzuryu River
Hii River
Sagami River
Kurobe River
non-deposited
Effects of deposition on the vegetation coverage
- Long period simulation results-
Vegetation coverage of periodical steady condition
0
0.2
0.4
0.6
0.8
1
0 1 2 3 4 5
Dep
osite
d a
rea
fr
action
Flood level (m)
0.8 0.75
0.7 0.65
0.6 0.55 0.5 0.45
Conditions of simulation
Geomorphology: Kuzuryu River
Flood condition: 1/year , 1/10years
Deposited area fraction: 0~1.0
Simulation period: 50 years (until periodical steady condition is achieved)
Observed condition
Vegetation cover increases
with decreasing fraction of
deposited area, as well as
the flood intensity and flood
frequency 0
0.2
0.4
0.6
0.8
1
0 1 2 3 4 5
De
po
site
d a
rea
fr
action
Flood level (m)
0.85
0.80 0.75
0.70 0.65 0.60 0.55
0.50
Every year flood
Every 10 year flood
Harvesting of sediment / termination of sediment inflows by dams
Reduction of movable sediment in the river channel
Decrease in sediment deposition area at flood time
Fast recovery of vegetation after the flood
Increase in vegetation coverage in the river channel
Possible mechanism of the increased vegetation coverage in the river channel
Summary
1. Vegetation coverage is increasing according to the present observations.
2. Vegetation colonization after a flood is delayed by the deposition of
cleansed sediment.
3. The sediment stock in the river channel has been reduced due to
harvesting and termination of sediment inflows by upstream dams,
consequently sediment deposition during floods was substantially
decreased.
4. Decreasing sediment deposition area during floods has enhanced the
vegetation coverage of the river channel.
5. A dynamic model was developed to describe the processes.
• The model consists of four interacting modules; hydrological processes,
trees, herbaceous plant biomass, and nutrient concentration of the soil.
• It describes the recruitment and the later growth of trees, herbaceous plant
biomass, the nitrogen concentration of the soil, and flushing of vegetation.
• Simulation results indicates the vegetation coverage decreases with the
fraction of deposited area as well as flood intensity and frequency.