Remote sensing and highresolution landscape models in predictive geological mapping

1
Remote sensing and high-resolution landscape
models in predictive geological mapping

• Bob Gozzard1 and Richard Langford2
• 1Geological Survey of Western Australia
• 2Safety, Health and Environment Division, DoIR

WALIS Forum, Perth, February 2005
2
Outline of talk

• How geology is expressed
• Classical mapping v. traverse mapping
• Remote sensing data
• Digital elevation data and derivatives
• The Kalgoorlie experience

3
How geology is expressed

• Changes in geology are mirrored in the landscape
  as changes in landform
• Strong correlation between geology and landscape
• Breaks and changes of slope have geological
  meaning
• Observing these variations in slopeis at the
  core of classical geological mapping

4
Classical geological mapping

• Uses direct observation of landforms from
  detailed topographic maps and meticulous field
  observation of the relationships between geology
  and landform
• Mapping is based on identifying breaks and
  changes of slope, rates of change of slope, and
  delineating slope form lines
• All features have geological meaning
• Possible to trace geological boundaries across
  country

5
Tracing geological boundaries
6
Geologists field slip
7
Traverse mapping

• Interpretation of phototone, vegetation changes,
  and a visualization of topography on stereoscopic
  aerial photographs
• Spatially and spectrally uncontrolled environment
• Supplemented by field traverses and targeted
  mapping
• Effective, but intrinsically limited

8
Classical v. traverse

• Classical mapping
• Holistic interpretation of near surface geology
• Takes advantage of the strong correlation between
  geology and landscape features
• Costly and time consuming
• May not be practicable in many countries

• Traverse mapping
• Less reliance on and use of landscape features
• Pragmatic alternative
• Significant field component

9
Present day demands

• Higher quality and larger scale maps
• High quality 3D data to model and describe the
  landforms that underpin good geological
  interpretation
• Move away from uncontrolled sources
• More focussed field activities
• Effectively integrated spatially controlled
  datasets can be used as a predictive tool to map
  geology

10
Remote sensing data

• Landsat TM is, and will continue to be, the
  mainstay for geologists
• D-stretch 754, ratios 5/74/74/2
• Radiometrics
• Airborne magnetics
• ASTER geological heir to Landsat TM
• Hyperspectral mineral mapping (HyMap)
• Radar especially multi-frequency,multi-polariza
  tion

11
Digital elevation data

• Geoscience Australia 9-second DEM
• Soft photogrammetry of stereo aerial photography
• SPOT PAN
• RADARSAT
• ASTER
• Radar interferometry TOPSAR, SRTM
• Radar altimeters

12
DEM derivatives

• Contours high-resolution, 0.5 m
• Slope most important given the strong
  correlation between landform and geology all
  surfaces are composed of slopes
• Aspect lineations or regional slopes
• Profile convexity breaks and changes of slope
  importance in the classical methodology
• Plan convexity measures the amount of
  dissection of a landscape
• Shaded relief models pseudo 3D view

13
Bringing it all together the Kalgoorlie
experience

• Identify and characterize regolith
• Landsat TM D-stretch 754, ratios 5/74/74/2
• 0.5 m-resolution orthophotos resampled to 2.5 m
• DEM derivatives
• 0.5 m and 2 m contours
• Slope
• Aspect
• Profile curvature
• Shaded relief model

14
Orthophotograph
15
Landsat TM D-stretch 754
16
Slope image
17
Profile curvature
18
Conclusion

• Classical mapping method recognizes the
  importance of landscape in the mapping process
• High-resolution DEMs and their derivatives
  provide invaluable and detailed information on
  the landscape
• Classical techniques can be adapted to integrate
  remotely sensed data and DEMs
• Rapid creation of high quality maps

19
Take home messages

• Dont forget or underestimate the importance of
  landscape in the geological mapping process
• The availability of high-resolution DEMs has
  opened the door for a return to established
  morphological mapping techniques in the
  geosciences