Landscape-Ecological Map using Three
Dimensional Vegetation Structure and Micro
Landform Classification detected by LIDAR
Survey Data
ICC 2009
Nov. 18th 2009
Escuela Militar, Santiago, Chile
Mamoru KOARAI, Takayuki NAKANO,
Junko IWAHASHI and Hiroshi P. SATO
(Geigraphical Survey Institute, JAPAN)
Content
• Introduction
• Studied area (Mt. Rausu, Shiretoko
Peninsula) and LIDAR survey
• Three dimensional vegetation map using
LIDAR data
• DEM analysis of LIDAR survey data
• Overlay analysis between micro
topography and vegetation
• Conclusion
Introduction
/The authors have been doing the project of trial produce
of landscape-ecological map of Shiretoko Peninsula,
where is Natural Heritage Area of Japan, Hokkaido
Island.
/Basic legend of landscape-ecological maps consists of
the combination of vegetation classification and
landform classification.
/In this presentation, the authors introduce three
dimensional vegetation structure map using LIDAR
survey data for legend of landscape-ecological map.
/This research project have been promoting by the
budget of the Ministry of Environment.
Background
(characteristics of LIDAR survey data)
/Last pulse LIDAR data in winter season is useful for detection of micro
landform under forest area
/Vegetation classification has been down using three dimensional vegetation
structure detected by the difference between LIDAR data in two seasons
summer
winter
Studied area and LIDAR survey
Mt. Rausu, Shiretoko Peninsula
World Natural Heritage Area of Japan
Studied area
●Studied area
Byobuiwa
profile1
South-East foots of Mt. Rausu,
Shiretoko Peninsula
(1km×4km)
●Airborne laser survey
(LIDAR Survey)
Tomariba
profile2
Haimatsubara
0.5m grid DSM and DEM
in summer season
(Sept.5th , 2008)
site4
Satomidai
2m grid DSM and DEM
in spring or autumn season
(June 6th ,2004,Oct.19th ,2004)
LIDAR
survey area
Byobuiwa
Betula ermanii
Tomariba
Pinus pumila,Betula ermanii ,low tree
Haimatsubara
Quercus crispula Blume,Pinus pumila, Sasa
Satomidai
Abies sachalinensis, Quercus crispula Blume,Betula ermanii,Sasa
profile 1
profile 2
profile 3
Three dimensional
vegetation structure
by LIDER data
study area
single layer
profile 1
profile 3
profile 2
pinus pumila
Summer
DSM(left),
DEM(right)
(0.5m grid)
Three dimensional vegetation map using LIDAR data
histogram of vegetation height on each colony
35000
Sasa-betula ermanii Cham
ササ-ダケカンバ
betula ermanii Cham
ダケカンバ
Abies sachlinesis-Quercus crispula Blume
トドマツ-ミズナラ
Acer pictum Thunb- Quercus crispula Blume
エゾイタヤ-ミズナラ
30000
画素数
number
grid
25000
over 1.5m under 6m--low tree
over 6m under 10m--middle tree
over 10m-------high tree
20000
15000
10000
5000
0
0
5
10
vegetation
height(m)
樹高(m)
15
20
Hs-Hw
Hs
Hs
Hs-Hw
Hw
Dw
Ds
Hw
Hw
Grass, pinus
pumila, bare
Hs
Evergreen trees
Deciduous trees (Single layer)
Deciduous trees (Multiple layer)
Hs<1.5m, Grass, pinus pumila, bare; Hs≧1.5m, Trees
Hs-Hw<3m (always Hs≧7m), Evergreen trees; Hs-Hw>=3m, Deciduous trees
If Hs≧7m, crown:
Dw≧10m, thick,
Dw<10m, thin.
Hs≧10
m, High
10m>Hs≧7 m,
Medium
If Hs≧10m, crown: Ds≧10m, thick; Ds<10m, thin.
Hw<5m, Single layer
Hs≧10
m, H.
10m>Hs
≧6m,
Med.
Hw≧5m, Multiple layer
Hs<6m, Hs≧10
m, High
Low
10m>Hs≧1.5
m, Med. &
Low
LIDAR vegetation map of Mt. Rausu
Vegetation classification using
three dimensional structure
detected by LIDAR data
deciduous or evergreen
deciduous (single layer or multiple layers)
three categories
×
vegetation height
deciduous (single) --- three categories (low, medium, high)
deciduous (multiple) --- two categories (low/medium, high)
evergreen --- two categories (medium, high)
+
crown thickness of high tree (thick, thin)
three categories
=ten categories
+
bare/grass/pinus pumila
=eleven categories
, pinus pumila
Relationship between LIDAR vegetation map and Actual Vegetation Map
published by Ministry of Environment
1
Vaccinium-Pinus pumila
2
Betula ermanii Cham
Sasa- Betula ermanii Cham
3
classification by LIDER
vegetationレーザデータによる植生区分
Alnus crispa
4
Acer pictum Thunb-Quercus crispula Blume
, pinus pumila
Abies sachalinensis-Quercus crispula Blume
5
高山低木群落
コケモモ-ハイマツ群落
雪田草原
エゾマツ-トドマツ群集
ダケカンバ-エゾマツ群落
ダケカンバ群落(Ⅱ)
ミヤマハンノキ群落
ササ-ダケカンバ群落
ササ群落(Ⅱ)
トドマツ-ミズナラ群落
エゾイタヤ-ミズナラ群落
ヤマハンノキ群落
ササ群落(Ⅳ)
オオヨモギ-オオイタドリ群落
シラカンバ-ミズナラ群落
チシマザサ-クマイザサ群落
硫気孔原植生
自然裸地
6
7
8
9
10
11
0
200000
400000
600000
800000
面積(㎡)
area
(㎡)
1000000
1200000
1400000
Projected plan of the crowns
corner of survey site
position of trees
LIDER vegetation map
bare,grass,pinus pumila
deciduous (single)
deciduous (multiple)
evergreen
species
Abies sachalinensis
Quercus crispula blume
Sorbus commixta hedl
Acer japonicum Thunb
Betula ermanii Cham
others
Cross section on the plot
DEM analysis
of LIDAR survey data
×10 6
(a) gradient
×10 6
×10 6
(b) convexity
(c) roughness
histogram of gradient,convexity and roughness
of each grid size DEM (0.5m, 2m and 50m)
automatic land form classification
using each grid size DEM
automatic land form
classification
using 0.5m grid DEM
s・cv・sm
s ・ c v・ ro
s・cc・sm
s ・ c c ・ ro
m・cv・sm
m・cv・ro
m・cc・sm
m・cc・ro
g・cv・sm
g ・ c v・ ro
g・cc・sm
g ・ c c・ ro
automatic land form
classification
using 2m grid DEM
automatic land form
classification
using 50m grid DEM
s・cv・sm
s ・ c v・ ro
s・cc・sm
s ・ c c ・ ro
m・cv・sm
m・cv・ro
m・cc・sm
m・cc・ro
g・cv・sm
g ・ c v・ ro
g・cc・sm
g ・ c c・ ro
0.5m grid
s・cv・sm
s ・ c v・ ro
s・cc・sm
s ・ c c ・ ro
m・cv・sm
m・cv・ro
m・cc・sm
m・cc・ro
g・cv・sm
g ・ c v・ ro
g・cc・sm
g ・ c c・ ro
2m grid
gradient
g:gentle
m:middle
s:steep
convexity
cc:concave
cv:convex
50m grid
roughness
ro:rough
sm:smooth
0.5m(g・cc・ro)
0.5m(g・cc・sm)
0.5m(g・cv・ro)
0.5m(g・cv・sm)
0.5m(m・cc・ro)
overlay of 0.5m grid and
2m grid automatic
landform classification
0.5m(m・cc・sm)
0.5m(m・cv・ro)
0.5m(m・cv・sm)
0.5m( s・ cc・ ro)
2m( g・ cc・ ro)
2m(g・cc・sm)
2m( g・ cv・ ro)
2m(g・cv・sm)
2m(m・cc・ro)
2m(m・cc・sm)
2m(m・cv・ro)
2m(m・cv・sm)
2m( s・ cc・ ro)
2m(s・cc・sm)
2m( s・ cv・ ro)
2m(s・cv・sm)
0.5m(s・cc・sm)
0.5m( s・ cv・ ro)
0.5m(s・cv・sm)
g : gentle
m : middle
s : steep
cc : concave
cv : convex
0.5m( g・cc・ ro)
0.5m(g・cc・sm)
ro : rough
sm : smooth
0.5m( g・cv・ ro)
0.5m(g・cv・sm)
0.5m(m・cc・ro)
0.5m(m・cc・sm)
0.5m(m・cv・ro)
0.5m(m・cv・sm)
0.5m( s・ cc・ ro)
2m( g・ cc・ ro)
2m(g・cc・sm)
2m( g・ cv・ ro)
2m(g・cv・sm)
2m(m・cc・ro)
2m(m・cc・sm)
2m(m・cv・ro)
2m(m・cv・sm)
2m( s・ cc・ ro)
2m(s・cc・sm)
2m( s・ cv・ ro)
2m(s・cv・sm)
0.5m(s・cc・sm)
0.5m( s・ cv・ ro)
0.5m(s・cv・sm)
ratio of each automation
landform classification
(0.5m grid and 2m grid)
Overlay analysis between micro topography and vegetation
50m(g・cc・ro)
50m(g・cc・sm)
50m(g・cv・ro)
50m(g・cv・sm)
50m(m・cc・ro)
50m(m・cc・sm)
50m(m・cv・ro)
overlay of 50m grid
automatic landform
classification
and 1/50,000 actual
vegetation map
50m(m・cv・sm)
50m(s・cc・ro)
50m(s・cc・sm)
50m(s・cv・ro)
50m(s・cv・sm)
g : gentle
m : middle
s : steep
cc : concave
cv : convex
50m(g・cc・ro)
50m(g・cc・sm)
ro : rough
sm : smooth
50m(g・cv・ro)
50m(g・cv・sm)
50m(m・cc・ro)
50m(m・cc・sm)
50m(m・cv・ro)
50m(m・cv・sm)
50m(s・cc・ro)
50m(s・cc・sm)
50m(s・cv・ro)
50m(s・cv・sm)
ratio of 50m grid
automatic landform
classification
on each colony of
1/50,000 actual
vegetation map
2m(g・cc・ro)
2m(g・cc・sm)
2m(g・cv・ro)
2m(g・cv・sm)
2m(m・cc・ro)
2m(m・cc・sm)
2m(m・cv・ro)
overlay of 2m grid
automatic landform
classification
and 1m grid LIDAR
vegetation map
2m(m・cv・sm)
2m(s・cc・ro)
2m(s・cc・sm)
2m(s・cv・ro)
2m(s・cv・sm)
g : gentle
m : middle
s : steep
cc : concave
cv : convex
2m(g・cc・ro)
ro : rough
sm : smooth
2m(g・cc・sm)
2m(g・cv・ro)
2m(g・cv・sm)
2m(m・cc・ro)
2m(m・cc・sm)
2m(m・cv・ro)
2m(m・cv・sm)
2m(s・cc・ro)
2m(s・cc・sm)
2m(s・cv・ro)
2m(s・cv・sm)
ratio of 2m grid
automatic landform
classification
on each category of
1m grid LIDAR
vegetation map
Conclusion
• The authors produce three dimensional vegetation map
using LIDAR data on south-east foots of Mt. Rausu on
Shiretoko Peninsula.
• LIDAR vegetation map is correspond to Actual Vegetation
Map and ground survey data.
• Result of overlay between automatic landform
classification and vegetation classification shows
vegetation on Mt. Rausu depend on elevation.
• The authors will produce legend of landscape-ecological
map combined with three dimensional vegetation structure
and micro landform classification for the evaluation of
biodiversity in Natural Heritage Area.
ダウンロード

原生自然地域(知床)での 景観生態学図と生物多様性データの連携