Definitions

Document Sample
Definitions Powered By Docstoc
					 Quantifying geomorphic and riparian land cover changes
  either side of a large flood event using airborne remote
                sensing: River Tay, Scotland
                            By Robert G. Bryant and David J. Gilvear
                                    Geomorphology 29 (1999) 307-321




                                     Definitions
Bathymetric – Measuring from the bottom of the river to the surface of the river

Lyzenga algorithm – Dr. Lyzenga is a research scientist at the University of Michigan;
1981. Remote sensing of bottom reflectance and water attenuation parameters in shallow
water using aircraft and Landsat data. Int. J. Remote Sensing 2, 71–82.

Airborne thematic mapper(ing) (ATM) – Passive multi-spectral mapping preformed from a
low flying aircraft; measurements are taking in 11 bands encompassing the visible, and
near, mid, and thermal infrared spectrum.

Iterative - To utter or do a second time or many times; to repeat; as, to iterate advice.

Delaunay triangles - a collection of edges satisfying an "empty circle" property: for each
edge we can find a circle containing the edge's endpoints but not containing any other
points (I don’t really know what this means).

Dalradian supergroup - A group of metamorphosed sediments laid down in Scotland and
Ireland between 800 and 600 million years ago.

Lacustrine – pertaining to lakes.

Dark-pixel correction - a technique which determines the pixel in the image with the
lowest brightness value. This pixel is assumed to have a zero ground reflectance such
that its radiometric value represents the additive effect of the atmospheres.
                                Summary
•The purpose of this study was an attempt to use multi-spectral imaging in a
new way to quantify changes in a dynamic river system over a large scale,
and to quantify changes in the system after a major flood event.

•Dynamic changes were quantifiable to a very fine degree and the results
were in an easily used format. The changes measured in the river system
were minor; the researchers speculated that a previous flood event may have
reduced the effect.

•This study illustrates the potential value of multi-spectral imaging




                            Research Area

•The River Tay is a wandering gravel bed river which are one of the most dynamic
components of the landscape

•The River Tay has the highest mean annual discharge (160 m3 s-1) in the UK with
a catchment of 4690km2

•Low evapotranspiration values, thin soils, impermeable catchment geology, and
two hydroelectric dams

•Land use 8% agriculture, 5% forestry, and <1% urban

•Two major flood events 1990 (1759 m3 s-1) and 1993 (1878 m3 s-1) ATM was
preformed in 1992 and 1994

•Study reach is the confluence of the Rivers Tay and Tummel, channel length 3
km, channel width 50-60 m, floodplain width 1 km with embankments, and fairly
natural geomorphic processes occur within the embankments
                                   Objectives

•Assess the feasibility of using airborne remote sensing for discriminating, detection,
and quantification of change within a number of attributes of a dynamic wandering
gravel bed river either side of a large flood event

•Quantify spatial variation and change in fluvial landforms and land cover over long
reach scales (>20 times channel width)

•Normal Methods use a reductionism framework – particular aspects of the system are
isolated and studied under uniform conditions
                                     Methods

•ATM was preformed twice once on June 12, 1993 and once on June 13, 1994
•A 1:5000 aerial photograph was taken along with the ATM
•A 1:12000 aerial photograph was taken immediately following the flood to estimate
later geomorphic changes
•Pre-processing
    •Data were initially corrected to SI units using sensor calibrating data
        •Bands 1, 9, 10, and 11 were removed due to poor signal-to-noise, poor
        calibration, or data redundancy
    •The remaining bands were then histogram normalized to those of 1992 to correct
    any variability in the radiometric calibration and sensor characteristics
    •A simple dark-pixel correction was done to further account for and remove any
    atmospheric differences
    •The results were then checked with corresponding ground spectra collected at
    the same time as the ATM
    •It was then geometetrically corrected to a digitized 1:10000 survey map




                          Methods Continued
•Detection of change was by classification comparison
    •Each image was subjected to the same process
    •Maximum likelihood classifier
    •Same number of classifiers were used on each image
    •To quantitatively map changes the matrix function of ERDAS was used
    •After multiple trials 7 distinct land cover types were aggregated from a possible
    of 18
    •Each was then checked with field survey data and photographs collected at the
    time of ATM
    •Classification accuracy was 72% for 1992 and 74% for 1994
•Bathymetric change
    •Used Lyzenga equation to derive a simple linear depth function for reflectance
    values derived from both panchromic photography and single-date multi-spectral
    imagery
    •Water depth at selected sites were collected co-incident with 1992 ATM
    •Data were radiometric corrected and histogram normalized
                                    Results
The approach used in this study allows the results to be presented in a variety of ways
•Maps of in stream geomorphology both above and below the water
•In-stream stability maps
•Maps of riparian land cover and land use change
•Quantitative data on areas of fluvial and riparian land cover and changes from one
type of fluvial feature and land cover to another
•Below water surface change as 2D perpendicular to and parallel to the channel
•The scale and region of interest can be varied to the degree of detail required and
has the ability to zoom in on an area of interest




 Maps of in stream geomorphology both above and below the
                          water
The scale and
region of interest
can be varied to
the degree of
detail required
and has the ability
to zoom in on an
area of interest




In stream stability
       maps
                                               Below water
                                               surface change
                                               as 2D
                                               perpendicular to
                                               and parallel to
                                               the channel




Quantitative data on areas of fluvial and riparian land cover and
  changes from one type of fluvial feature and land cover to
                            another
                             Overall Results

•The extent of the change in the reach, given the magnitude of the flood, was not
great.
•This may be because of the previous flood three years earlier.
•Features that showed limited change
    •Bar head deposition
    •Bar tail accretion
    •Bar dissection
    •Mid channel scour
    •Bank erosion (unvegetated)




          Vertical geomorphologic changes of 10cm were detectable




                           Discussion/Overall
•Some inconsistency between ground survey and ATM
•Quantified changes in fluvial landforms and riparian land cover
•10 cm vertical change detected
•1.5 m horizontal change detected
•Very useful results
•Too expensive for most applications
•Elaborate on how to pre-process of data

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:1
posted:12/8/2011
language:
pages:8