GE3502GE5502 Geographic and Land Information Systems

1
GE3502/GE5502Geographic and LandInformation
Systems
Lecture 14 Deriving other indices of terrain
shape and function. Terrain Analysis
2
Lecture plan

• Terrain modelling
• Morphometric indices
• Example Modelling soil erosion
• Example Modelling landslides

3
(No Transcript)
4
Mount Fuji, Japan
5
(No Transcript)
6
What is terrain modelling?

• Terrain modelling is the mathematical study of
  the ground surface, both individual landforms and
  continuous topography
• Also called terrain analysis, quantitative
  geomorphology, and (geo)morphometry newer term
  digital terrain modelling is now common
• Origins go back 150 years to contour map
  measurements by von Humboldt and later German
  geographers, and physiographic mapping.

7
What is terrain modelling?

• Many of its results are calculated from
  square-grid arrays of terrain heights, or digital
  elevation models (DEMs)
• Examples include
• maps of shaded relief,
• slope angle,
• slope curvature,
• river basin morphometry,
• various other process parameters

8
Applications

• Assessing the severity of geologic hazards
• Soil erosion,
• Model tectonic events,
• Predict the movement of surface and ground water,
  
• Many other problems in science and engineering
• Importance for the study of earth surface
  processes and how they interact with other
  phenomena is evident from the inclusion of
  terrain-modelling capabilities in standard GIS
  software such as, GRASS, ERMapper, ArcView and
  ArcInfo (ArcGIS), IDRISI, and MapInfo

9
Morphometric parameterisation

• Evans (1979) considers five terrain parameters
  that may be defined for any two dimensional
  continuous surface
• elevation (0 order)
• slope, aspect (1st order differential)
• profile (slope) convexity, plan (contour)
  convexity (2nd order differential)
• 1st and 2nd order functions have components in
  the X,Y, and Z planes
• The higher the order of the polynomial, the
  greater the number of points required to uniquely
  identify the necessary coefficients

10
2nd order derivatives Eg. Used to measure
erosion, map drainage etc.)
11
Feature approach
12
Algorithmic approach
Aspect
Slope

• SFD single flow direction - distributes flow to
  a single cell with the lowest relative elevation
  
• most common method suitable for stream network
  extraction in low resolution DEMs
• used in flowacc in ArcView, ArcGRID, r.watershed
  in GRASS, Rivertools

Flow accumulation
13
Terrain skeleton

• the network of ridges and valleys in a piece of
  topography - the points of highest and lowest
  elevations
• Peucker and Douglas (1975) identified concave
  pixels to find streams and convex pixels to find
  ridge points
• Mark (1983) used an algorithm that starts at an
  upstream drainage area and identifies
  successively lower pixels
• Advantage of this method is that it yields a set
  of drainage lines that are guaranteed to be
  connected. Peucker and Douglas' method sometimes
  yields broken segments

14
Terrain skeleton

• Verbal classification scheme to assign a category
  to each pixel PIT, PEAK, PASS, RAVINE, and
  RIDGE. The pixels are connected into a network,
  and then thinned to a set of lines.
• Two limitations to algorithmic methods
• Terrain skeleton that is extracted can only be as
  good as the original DEM
• Algorithmic methods can be led astray by noise in
  the data

15
DEM Elevation
16
Profile curvature

• indicates areas of accelerated flow (convex) and
  areas with decreasing flow velocity (concave)

17
Upslope area

• indicates area upslope from each cell

18
Tangential curvature

• represents areas of convergent (concave) and
  divergent (convex) flow

19
Topographic Wetness Index

• Represents wetness

20
Example Soil ErosionRUSLE (Revised Universal
Soil Loss Equation)

• Basic equation is
• E R K LS C P
• where E is the average soil loss, R is the
  rainfall intensity, K is the soil erodibility, C
  is the cover factor, P is the cultivation factor
  and LS is the
• topographic factor
• LS(r) (m1) A(r) / a0 m sin
  b(r) / b0 n
• where A is upslope contributing area per unit
  contour width, b is the slope, a0 22.1m, b0
  0.09, m is 0.4-0.6 and n1-1.3 or use more
  complex process equations
• Need to also have continuous coverages of DEM
  (upslope area, slope), Soils (K-factor), Rainfall
  (R-factor), Landcover (C and P factors)

21
e.g. RUSLE in ArcView

• Given data grids elevation, K, C, (P),
  constants R 220, resolution 10 m
• 1. DERIVE SLOPE
• under Analysis select derive slope, give the
  new theme name slope
• 2. MAP CALCULATOR
• build an expression
• (elevation.FlowDirection(FALSE)).FlowAccumulatio
  n(NIL)
• Evaluate, give the new theme name flowacc
• build an expression
• ((flowacc resolution/22.1).Pow(0.6))((((slop
  e0.01745).Sin)/0.09).Pow(1.3))1.6
• Evaluate, give the new theme name lsfac
• build an expression
• RKCPlsfac
• Evaluate, give the new theme name soilloss

22
C cover factor
K soil erodibility
LS topographic factor
T tolerable soil loss
23
RUSLE Output
24
Landslide studies

• terrain modelling assists in predicting locations
  of potential landslides and debris flows
• Use digital maps of slope and curvature from DEMs

25
Landslide studies

• Combine these measures of terrain form with
• Physical characteristics of rocks and soils
• Land cover
• Rainfall
• then identifying locations that might be subject
  to various types of landsliding
• Helps local planners by creating quantitative
  risk maps

26
USGS San Francisco study

• Deep-seated "slump" type landslide in San Mateo
  County
• Beginning a few days after the 1997 New Year's
  storm, the slump opened a large fissure on the
  uphill scarp and created a bulge at the downhill
  toe
• Over 250,000 tonnes of rock and soil moved in
  this landslide
• Led to detailed modelling in region
• Go to landslide mpeg (http//elnino.usgs.gov/lands
  lides-sfbay/photos.html)

27
(No Transcript)
28
Assessing the quality of DEMs

• most important data layer used in evaluating
  erosion risk is the digital elevation model
• vertical precision in cm is needed, metre
  resolution is not sufficient because it creates a
  "step" effect leading to zero slopes and
  contourlike slope pattern, ultimately
  underestimates erosion
• systematic errors such as stripes or bands can be
  present in the DEM due to resampling
• another common error is edge mismatching between
  adjacent datasets due to inaccurate projection
  conversion or wrong datum

29

• Insufficient interpolation methods
• can avoid by using increased smoothing or
  manually adding points in areas where data are
  sparse

30
Conclusions

• Quality of DEMs is critical
• Any terrain indices must have basis in landscape
  process
• Higher order morphometry not yet viable
• Use of fractal analysis promising

31
Readings

• Burrough McDonnell, (1998) Chapter 8
• DeMers, (2000) Chapter 10