LANDSCAPE ECOLOGY
SREM 3011
LECTURE 13
Dr Brendan Mackey
Department of Geography
The Australian National University
Water shed hydrology,the passage of precipitated water through
vegetation cover,soil and rock to the stream,Precipitation falls over
the whole watershed and is concentrated in stream channels,
Main limitation of?bucket? MI method is that it does not factor in
catchment hydrological processes,
overland flow; infiltration;
throughflow
processes are topographic-
ally driven
spatial modelling of these
processes requires a suit-
ably scaled digital elevation
model (DEM)
NB Focus here on humid,erosional landscapes,not dry depositional
landscapes
Simulation models of catchment hydrology:
- many deterministic and empirical models
- all model water flow in the catchment as a
function of:
1,Topographic characteristics and
2,Soil characteristics
- Highly parametrised models (eg,CSIRO?Topog?)
difficult to?run? or implement across landscapes
due to lack of required spatial data
- Therefore?simple? models are,popular”
“Position in topo-sequence” as an index of run-on/run-off
Examples of profiles across terrain divided into morphological types of
landform element
Slope lines overlaid on a contour map show ridge lines and course lines
,Position-in-a-topographic-sequence” an index of
whether you are shedding or receiving water
But terrain is 3D,therefore position-in-catchment
is a better description
Area above point in catchment is critical
ie,the up-slope area or up-slope contributing area
At drainage line,USCA is large
At crest,USCA is small or?
> USCA therefore > potential discharge (?) of water
through that point/location
An idealized catchment above a 20m x 10m plot:
As = A/W = Specific catchment area
= Index of unit area discharge
= Average catchment length (ACL)
A
A = USCA
w = plot width
w
tan? = slope angle = b/a
Slope for?landscape unit of analysis? eg,plot
Index of hydraulic gradient
Soil attributes affecting T (transmissivity)
- a,Soil depth and texture
- b,Soil porosity
- c,Hydraulic conductivity
b
a
Wetness index? is defined as
WI = (As * T) / tan?
where As is specific catchment area
T is transmissivity
tan? is slope angle
Based on a great deal of theory!
Certain assumptions:
a,Infiltration constant across landscape
b,Impermeable layer at fixed depth
- therefore problems in karst country & depositional
landscape
- works better in erosional landscapes
Spatial application of wetness index?
Soil attribute data generally unavailable
Therefore,assume T is uniform and equal to 1
Topographic wetness defined as an index:
TWI = As / tan? } Interpretation?
Therefore factor out soil and therefore need only
calculate two topographic characteristics:-
(1) As? f (USCA,plot width)
(2) Slope angle (tan?)
As = A / w
where As is specific catchment area
A is USCA
w is width of grid cell
w? is?constant? for small area,but varies with
latitude vis-à-vis regional studies
More than one method to calculate flows as
water can diverge or converge
ongoing research into optimum algorithm!
Calculation of Topographic Wetness Index
99 95 85 85
90 80 80 85
85 80 75 80
85 80 70 65
1 1 1 1
2 4 3 1
1 2 9 1
1 2 5 16
Elevation (DEM)
Flow Direction
Accumulated Cell Count
Slope (calculated from elevation)
Accumulated cell count
= A = USCA
Elevation Slope
(primary) (Secondary)
need appropriately
scaled DEM
Topographic Wetness Index:
-?A? and tan? are secondary terrain attributes
- TWI is example of?compound? or knowledge-based
terrain attribute
- TWI is an index not absolute values
- TWI requires calibration ie,correlation with field
data
- Comparison of TWI between landscapes/ecosystems
/regions is tricky
,climatic differences
,geological differences? soil differences
Primary topographic attributes (adapted from Speight 1974,1980)
Estimation of topographic wetness index in the field
ie,at survey plots:
* Two options
(A) derive from DEM (ex situ)
(B) measure in field (in situ)
* Problems?
- availability of suitably scaled DEM?
- DEM is always smoother than reality
- in situ okay upslope,but harder downslope
Therefore try combination
in situ for upslope plots
DEM for downslope plots
Another idea:
Substitute the flow-path length (FPL) for As
For FPL = As must assume?A? (USCA) is a rectangle
As = A/w Therefore calculate a crude approximation of TWI at a plot
TWI = FPL / tan?
where tan? is plot slope and
FPL is flow path length
As = USCA/w = ACL
TWI = DUNTC/ tan?
FPL
DUNTC?*
* DUNTC = Distance upslope to nearest crest
A
w
Climate interpolation/gridding
+
DEM-based terrain analysis
+
Remote-sensing - spectral/emittance
- geophysical
+
Digitized-existing thematic maps
Enhanced capacity for spatially distributed landscape
ecological models
Transformed ability to
model physical
environment
EOS platform
1000s more bands
& > spatial resolution
Elevation Percentile Index
- circle of specified radius moved over each grid cell
- all the grid cells in the circle ranked according to
elevation (min? max)
- % (percentile ranking) of focal cell is recorded
- index of?local? topographic position; local run-on/
run-off
- optimum radius for circle
SREM 3011
LECTURE 13
Dr Brendan Mackey
Department of Geography
The Australian National University
Water shed hydrology,the passage of precipitated water through
vegetation cover,soil and rock to the stream,Precipitation falls over
the whole watershed and is concentrated in stream channels,
Main limitation of?bucket? MI method is that it does not factor in
catchment hydrological processes,
overland flow; infiltration;
throughflow
processes are topographic-
ally driven
spatial modelling of these
processes requires a suit-
ably scaled digital elevation
model (DEM)
NB Focus here on humid,erosional landscapes,not dry depositional
landscapes
Simulation models of catchment hydrology:
- many deterministic and empirical models
- all model water flow in the catchment as a
function of:
1,Topographic characteristics and
2,Soil characteristics
- Highly parametrised models (eg,CSIRO?Topog?)
difficult to?run? or implement across landscapes
due to lack of required spatial data
- Therefore?simple? models are,popular”
“Position in topo-sequence” as an index of run-on/run-off
Examples of profiles across terrain divided into morphological types of
landform element
Slope lines overlaid on a contour map show ridge lines and course lines
,Position-in-a-topographic-sequence” an index of
whether you are shedding or receiving water
But terrain is 3D,therefore position-in-catchment
is a better description
Area above point in catchment is critical
ie,the up-slope area or up-slope contributing area
At drainage line,USCA is large
At crest,USCA is small or?
> USCA therefore > potential discharge (?) of water
through that point/location
An idealized catchment above a 20m x 10m plot:
As = A/W = Specific catchment area
= Index of unit area discharge
= Average catchment length (ACL)
A
A = USCA
w = plot width
w
tan? = slope angle = b/a
Slope for?landscape unit of analysis? eg,plot
Index of hydraulic gradient
Soil attributes affecting T (transmissivity)
- a,Soil depth and texture
- b,Soil porosity
- c,Hydraulic conductivity
b
a
Wetness index? is defined as
WI = (As * T) / tan?
where As is specific catchment area
T is transmissivity
tan? is slope angle
Based on a great deal of theory!
Certain assumptions:
a,Infiltration constant across landscape
b,Impermeable layer at fixed depth
- therefore problems in karst country & depositional
landscape
- works better in erosional landscapes
Spatial application of wetness index?
Soil attribute data generally unavailable
Therefore,assume T is uniform and equal to 1
Topographic wetness defined as an index:
TWI = As / tan? } Interpretation?
Therefore factor out soil and therefore need only
calculate two topographic characteristics:-
(1) As? f (USCA,plot width)
(2) Slope angle (tan?)
As = A / w
where As is specific catchment area
A is USCA
w is width of grid cell
w? is?constant? for small area,but varies with
latitude vis-à-vis regional studies
More than one method to calculate flows as
water can diverge or converge
ongoing research into optimum algorithm!
Calculation of Topographic Wetness Index
99 95 85 85
90 80 80 85
85 80 75 80
85 80 70 65
1 1 1 1
2 4 3 1
1 2 9 1
1 2 5 16
Elevation (DEM)
Flow Direction
Accumulated Cell Count
Slope (calculated from elevation)
Accumulated cell count
= A = USCA
Elevation Slope
(primary) (Secondary)
need appropriately
scaled DEM
Topographic Wetness Index:
-?A? and tan? are secondary terrain attributes
- TWI is example of?compound? or knowledge-based
terrain attribute
- TWI is an index not absolute values
- TWI requires calibration ie,correlation with field
data
- Comparison of TWI between landscapes/ecosystems
/regions is tricky
,climatic differences
,geological differences? soil differences
Primary topographic attributes (adapted from Speight 1974,1980)
Estimation of topographic wetness index in the field
ie,at survey plots:
* Two options
(A) derive from DEM (ex situ)
(B) measure in field (in situ)
* Problems?
- availability of suitably scaled DEM?
- DEM is always smoother than reality
- in situ okay upslope,but harder downslope
Therefore try combination
in situ for upslope plots
DEM for downslope plots
Another idea:
Substitute the flow-path length (FPL) for As
For FPL = As must assume?A? (USCA) is a rectangle
As = A/w Therefore calculate a crude approximation of TWI at a plot
TWI = FPL / tan?
where tan? is plot slope and
FPL is flow path length
As = USCA/w = ACL
TWI = DUNTC/ tan?
FPL
DUNTC?*
* DUNTC = Distance upslope to nearest crest
A
w
Climate interpolation/gridding
+
DEM-based terrain analysis
+
Remote-sensing - spectral/emittance
- geophysical
+
Digitized-existing thematic maps
Enhanced capacity for spatially distributed landscape
ecological models
Transformed ability to
model physical
environment
EOS platform
1000s more bands
& > spatial resolution
Elevation Percentile Index
- circle of specified radius moved over each grid cell
- all the grid cells in the circle ranked according to
elevation (min? max)
- % (percentile ranking) of focal cell is recorded
- index of?local? topographic position; local run-on/
run-off
- optimum radius for circle
