Applying Geospatial Technologies to Weed Mat Monitoring and Mapping: The Ythan Estuary, NE Scotland - PowerPoint PPT Presentation

1 / 37
About This Presentation
Title:

Applying Geospatial Technologies to Weed Mat Monitoring and Mapping: The Ythan Estuary, NE Scotland

Description:

... is about 71 ha and there are 115 ha of mudflat and 70ha of mussel beds and sand. ... substrate map showing sand, mud, mussel beds and other known 'habitat' ... – PowerPoint PPT presentation

Number of Views:185
Avg rating:3.0/5.0
Slides: 38
Provided by: davidr82
Category:

less

Transcript and Presenter's Notes

Title: Applying Geospatial Technologies to Weed Mat Monitoring and Mapping: The Ythan Estuary, NE Scotland


1
Applying Geospatial Technologies to Weed Mat
Monitoring and Mapping The Ythan Estuary, NE
Scotland
Annual Meeting 2003 GIS in Coastal and Marine
Areas Thursday 6th March 2002
  • David R. Green Stephen D. King
  • Centre for Marine and Coastal Zone Management
  • University of Aberdeen, Scotland, UK

http//www.abdn.ac.uk/cmczm
2
This presentation examines the potential for
using colour vertical aerial photography to map
the spatial location, distribution, extent, and
changes over time in weedmat coverage found in
the Ythan estuary, North East Scotland.
  • Mapping of weedmats in the Ythan estuary has
    previously been undertaken by Raffaelli et al. -
    concluding that although fluctuating from year to
    year there appears to be a gradual trend towards
    an overall increase in the extent of weedmats
    over time.
  • Using Raffaelli's work as a reference source the
    objectives of the current work were to develop
    an up-to-date practical methodology using image
    processing and GIS (Geographic Information
    Systems) software to provide information about
    weedmat location and extent over time and to
    provide quantitative estimates of weedmat
    coverage for comparison with earlier studies.
  • Archival colour aerial photography for four years
    (1989, 1992, 1994 and 2000) was scanned into a
    computer. The application of image processing
    techniques resulted in a series of mosaiced
    images that were used as the basis for visual
    interpretation and mapping of the location and
    extent of weedmats for each year of photography.
  • Input of the maps into a Geographical Information
    System (GIS) enabled the derivation of
    quantitative estimates of the areal coverage of
    macro-algal weedmats for each year and the
    creation a series of maps showing change.

Introduction Summary
3
  • The Ythan estuary (57oN, 2oW) is the area of
    study. It lies approximately 20 km to the north
    of Aberdeen on the east coast of Scotland, and is
    tidal extending for approximately 8km in the
    general direction of Ellon.
  • The estuary averages about 300m in width, and has
    an average depth at high water of 2.5m with tidal
    range of between less than a meter to 3m or more
    (Raffaelli, 1998, p. 138).
  • According to Raffaelli (1999, p. 164) The low
    water channel is about 71 ha and there are 115 ha
    of mudflat and 70ha of mussel beds and sand.
  • The Ythan drains a catchment of about 650 km2 of
    intense arable agricultural land. The study area
    chosen for this research is the entire intertidal
    zone for the estuary.

Study Area
4
  • A significant visual feature of the Ythan estuary
    in NE Scotland in recent years has been the
    presence of benthic macro-algae (Enteromorpha,
    Ulva and Chaetomorpha) forming extensive 'green'
    'mats' over the estuary's intertidal flats.
  • One of the main reasons for the apparent observed
    increase in the areal extent of macro-algal
    weedmats is thought to be related to the amount
    of nitrogenous fertiliser applied to the Ythan
    catchment over time.
  • Studies in the Ythan estuary have been undertaken
    to (a) identify the location and distribution
    the weedmats (spanning a time period from the
    late 1940s to the present day), and (b) try to
    establish what is causing the weedmat.
  • The primary reason for undertaking such studies
    is that the algal weedmats can have a dramatic
    impact on the invertebrates in the underlying
    sediment as well as on the shorebird populations.

Macro-Algal Weedmats
5
Report Preparation
I N P U T O U T P U T
Establish Background to Research Objectives and
Aims
Meeting 1
Study Area - Context and Setting Environmental
Problem - Defined Information Required Scene
Model of Phinn et al. (1999)
  • Data Available
  • Suitability for Task i.e. algal weedmat monitoring

Project Structure
Data Processing
  • Manual Photo-interpretation
  • Digitising
  • Digital Image Pre-Processing
  • GIS layers

Meeting 2
Analysis
Reports and Conclusions
Future Work
Meeting 3
6

Geospatial Methodology
7
  • In order to study the location, distribution and
    areal extent of macro-algal weedmats, over time,
    a number of field survey- and aerial
    photo-interpretation-based studies have been
    undertaken to develop cost-effective
    methodologies for repeat environmental data
    acquisition.
  • The methods developed and used are well
    documented in Raffaelli and Way (1996) and
    include
  • the use of aerial photographs acquired from light
    aircraft flying at 1000-1500 feet
  • photo-interpretation using vertical photographs
  • the projection of colour photographic slide
    transparencies for manual tracing and area
    assessment with the aid of mm graph paper and a
    planimeter (see e.g. Raffaelli and Plomer, 1989).
  • With developments in remote sensing (digital
    image processing) and GIS software and hardware
    technology the task of deriving spatial and
    temporal information on macro-algal weedmats has
    become far easier than was previously possible,
    allowing for rapid data capture, storage,
    handling and interpretation of remotely sensed
    photography and imagery in a computer
    environment.
  • There is considerable potential to (a) repeat
    and check the results from previous studies, (b)
    build up digital databases comprising
    geo-rectified and mosaiced aerial photography and
    vector map layers that can be used as a baseline
    for future monitoring and mapping exercises, and
    (c) derive quantitative information more quickly.
  • Young et al. (2000, p. II-286) lend additional
    support for a 'remote sensing approach' to
    weedmat monitoring and mapping .'The rapid
    development of benthic macroalgae on estuarine
    mudflats during the summer growing season makes
    difficult the accurate documentation of spatial
    and temporal distributions of such algae from
    ground surveys alone'.

Remote Sensing GIS
8

The overall objective of the current work is to
devise a practical remote sensing-based
methodology to extract spatial information
(location and extent) about macro-algal weedmats
in the Ythan estuary derived from multi-temporal
archival aerial photography of varying different
scales acquired from different sources.
  • More specifically
  • aerial photo-interpretation of archival aerial
    photography (panchromatic and colour) of varying
    scales and dates obtained from local sources to
    identify macro-algal weedmats
  • data capture of selected archival aerial
    photography in the form of scanned raster digital
    datasets
  • geo-correction of the scanned aerial photography
    in a digital image processing system (Erdas
    Imagine 8.4)
  • stitching or mosaicing of the geo-corrected
    aerial photography (Erdas Imagine 8.4/PanaVue)
  • multi-temporal mapping of the location and areal
    extent of macro-algal weedmats in the Ythan
    estuary, Aberdeenshire, Scotland (Erdas Imagine
    8.4 and ESRI ArcView 3.2)
  • comparisons between the information derived with
    other, earlier results using different
    methodologies to map from aerial photographs
  • creation and documentation of a series of
    deliverables including digital aerial
    photographs, interpretations, analyses, maps, and
    other datasets

Objectives
9
  • Typically Enteromorpha forms vast 'mats' that lie
    on the surface of estuarine mud- and sand-flats.
    These mats can either be very fine coverings
    (thin) or several cms deep (thick).
  • Although it is relatively easy to identify these
    extremes on aerial photographs densities of cover
    in-between are more difficult.
  • The water retention characteristics of the
    species and also the presence of thin layers of
    water covering the algae have been found to lead
    to a reduction in the overall surface reflectance
    making identification on photography and imagery
    quite difficult
  • Infrared reflectance is usually high, and on CIR
    photography the range of colours on the film
    varies from a dull blue/grey when saturated (with
    water) to deep red when present in thick healthy
    'green' mats.
  • Where found in free-floating mats, Enteromorpha
    may also be deposited on other vegetation types.
  • In a recent study around Portsmouth, Baily et al.
    (2002) found that weedmats may form a crust of
    dead material on the upper surface, leading to
    similar reflectance characteristics to gravel
    deposits.
  • Raffaelli et al. (1999, p. 108) note that In
    situations like the Ythan estuary, Scotland,
    Chaetomorpha is much more intimately associated
    with the sediment than are Enteromorpha, Ulva, or
    Cladophora, with substantial amounts of biomass
    anchored firmly beneath the surface

Macro-algal Weedmats
10
Photographic Coverage
1992
1989
2000
1994
11

A small number of vertical 35mm colour aerial
photographs were taken in August 1994 using a
model aircraft. These are for part of the Sleek
of Tarty area only and do not cover the Ythan
catchment. They are large scale, cover a small
area and show macro-algal weedmats very clearly.
Model Aircraft Photography
12
  • The aerial photography selected, together with
    any additional data and information (where deemed
    appropriate), were transformed into a digital or
    computer compatible format.
  • Conversion of the aerial photography from an
    analogue to digital format makes it directly
    compatible with digital OS map data and any other
    GIS layers available from other sources e.g. the
    sediment map from Raffaelli et al (1989) and
    Stapleton and Pethick (1996).
  • Data capture of the analogue format aerial
    photographs (verticals) was carried out using a
    standard desktop scanner (Agfa Snapscan 1212).
    Scanning at a resolution (X and Y direction) of
    150dpi (colour) the photography was captured and
    initially stored in a .TIF (F) format (Tagged
    Image File Format).
  • Choice of the scanning resolution was primarily
    based upon generating a filesize for scanned
    photographs that was considered manageable (also
    bearing in mind the need to subsequently mosaic
    and geo-correct the images). Whilst the choice of
    a high scanning resolution retains more of the
    detail contained in the original document, a
    compromise usually has to be struck in practice
    between the relative gain in information content,
    and the final filesize, as well as considerations
    of filesize manageability for display,
    transmission and storage.

Data Preparation
13
Scanning
14
  • Ideally all photography should be geo-corrected
    first to remove any inherent distortions if a map
    is to be the end product. The result will be a
    geometrically correct map (for each date) showing
    the macro-algal weedmat distribution which can
    then be used as a layer or coverage in
    ArcInfo/ArcView 3.2.
  • Subsequently the digital raster aerial
    photographs were input to the Erdas Imagine 8.4
    Digital Image Processing and GIS software system
    where they were
  • Geo-referenced (Projection/Spheroid/ Datum
    Transverse Mercator/Airy/OS GB 1936).
  • Geo-referencing was achieved using digital OS
    vector map data tiles for the designated study
    area
  • Geo-referenced and mosaiced imagery were also
    made available in the GeoTIFF format 'cropped'
    using an on-screen digitised boundary (ArcView
    3.2 shape file) of the Ythan estuary catchment
    area

Geo-rectification
15
All of the images selected for study were
geo-corrected using the Erdas Imagine 8.4 image
processing software referenced to OS digital map
data using between 4 and 16 GCPs per image, a
polynomial transformation (orders 1 and 2), and
resampled using nearest neighbour. Where
possible, the overall RMSE (Root Mean Square
Error) was reduced to lt1. One objective of this
work was to create a composite photographic image
for the entire estuary for each date of
photography available. To do this requires that
each geo-corrected aerial photograph be stitched
or mosaiced together. This was undertaken in
Erdas Imagine 8.4 using the Mosaic option. Whilst
some results were good, others resulted in
relatively poor colour balancing between each
photograph making up the mosaic. Subsequently,
mosaicing was also tested in a software product
PanaVue (http//www.panavue.com) with markedly
improved colour balancing results.
Geocorrection Mosaic
16

Mosaic
17
  • In order to conserve computer disk storage
    requirements for the digital aerial imagery, and
    to speed up the display of the backdrop for the
    purposes of digitising, a digital vector map
    outline of the Ythan estuary high water mark
    (HWM) (captured as an ESRI .shp (Shape file
    format) via on-screen digitising using ArcView
    3.2) was used to define an image 'cookie cutter'
    delimiting the boundary of the Ythan estuary
    study area.
  • The filesizes of the resulting raster datasets
    were significantly reduced. However, in practice
    it was generally found that the full image mosaic
    provided more informational 'context' for
    interpretation.

Filesize
18
  • Default color balancing of the mosaiced image
    files completed in Erdas Imagine 8.4 was
    generally deemed to be satisfactory, although not
    all yielded good results due to the differences
    between individual images making up the mosaic
    (arising from the sun-sensor-ground angles,
    sensor view angle, and possibly small atmospheric
    changes or differences, the camera, and
    photographic processing).
  • The mismatch in colour between one photograph and
    another is, for the most part, not a serious
    problem for mapping the weedmats, although in a
    few cases the resulting image colour balance
    would benefit from closer matching.
  • Further investigation of this potential visual
    interpretation problem revealed the availability
    of some other mosaicing software (PanaVue from
    Canada) that helps to markedly improve the colour
    balancing between individual photographs and the
    resulting mosaic. PanaVue also allows manual
    positioning of the individual photographs making
    up the mosaic.

Colour Balancing
19
  • Each cropped mosaic image of the Ythan estuary
    was input to the ArcView 3.2 GIS software package
    in Imagine (.IMG) format. Units of macro-algal
    weedmatting were mapped using the online
    interactive digitising tools.
  • Consistency in the interpretation (drawing
    boundaries and identifying the category) is
    important, hence the work was carried out by a
    single individual to (a) digitise, and (b)
    interpret.
  • The minimum mappable unit (mmu) is often used as
    the basis for determining a 'cut-off' point or
    threshold for deciding which map units to
    digitise and which not to digitize.
  • To some extent the choice of unit to map also
    depends upon
  • original photographic scale
  • quality of the aerial photography/mosaic
  • degradation of the photographic image e.g. due to
    scanning/screen display
  • image contrast and sharpness between the feature
    of interest and the background or surrounding
    area/surface (sometimes difficult towards the
    edges of aerial photographs)
  • Image magnification
  • Interactive zooming (in and out of an image)
  • eyesight of the interpreter
  • experience of the interpreter
  • quality of the display/original (electronic
    colour versus reflective colour)
  • Some assistance in the interpretation exercise
    can also be offered through (a) reference to
    contextual information and (b) an examination of
    the hardcopy aerial photographs

Interpretation
20
  • In general, weedmats appear to be a distinct
    'emerald green' colour in contrast to the 'pale
    brown/greybrown' colour of the 'sand/mud'
    background. Some, however, appear 'brownish' in
    colour in the digital imagery and visual checks
    had to be made with the original photographic
    prints.
  • The general location of the weedmats in each
    mosaic also seem to correspond to the 'habitat
    conditions' e.g. sheltered marine environments,
    middle to low intertidal zone, calm, protected
    harbours etc.
  • Overall the weedmats do appear to show a strong
    visual contrast with the bottom sediments (sand
    and mud). Only where they overlap darker bottom
    sediments e.g. mussel beds/sea weed does it pose
    difficulty in accurately delineating weedmat
    units (e.g. where they overlap). Some help is,
    however, provided by examining a sediment map for
    the area by Stapleton and Pethick (1996).

Photo-interpretation Key
21

In order to narrow down the decision-making
process (weedmat/not weedmat) contextual
information (e.g. text sources, maps etc.) can be
added into the mapped layers when digitising the
weedmat boundaries. For example, a bottom
sediment or substrate map showing sand, mud,
mussel beds and other known 'habitat' factors.
Contextual Information
22

Change Detection
23

Spectral Profiles
24

'On-screen' digitising of features on aerial
photography can be undertaken directly onto a
single or mosaiced image backdrop displayed in
Arc/Info, ArcView 3.2 or Erdas Imagine 8.4.
Accompanying each digitised map layer in
ArcView 3.2 is an attribute table comprising the
attributes of the feature type e.g. polygon/area,
and an ID number for each algal weedmat unit
(polygon) digitised, together with any other
information associated with the polygon.
ArcView 3.2 GIS
25
  • Originally it was anticipated that the weedmat
    maps derived by Raffaelli et al. for the Ythan
    estuary would all be made available for this
    work. Using the display and overlay functionality
    of a GIS it would be possible to undertake the
    following
  • check the quantitative areal measurements of
    weedmat coverage for the Ythan estuary as
    obtained by Raffaelli et al. and as shown on the
    accompanying maps and histograms
  • overlay weedmat maps for the same year obtained
    from different sources (Raffaelli et al. Green
    and King) to compare and contrast both the
    interpretations and the quantitative estimates of
    weedmat cover
  • Unfortunately only one map for 1986 was
    available. It was a very generalised paper
    hardcopy black and white map, and had been
    derived from oblique colour photographic slide
    transparencies of the Ythan estuary.

Some Problems
26

An alternative approach, therefore, was to
consider some of the weedmat maps made by
Raffaelli et al. for the South Quay area at
Newburgh (other areas are also shown in Raffaelli
and Way, 1996). Although these were similarly
rather generalised maps they provided one way of
'comparing and contrasting' interpretations and
quantitative data. There are, however, some
limitations to this approach that must be
considered when interpreting the results.
  • the maps sourced from Raffaelli are of a very
    small format, generalised and appear to be the
    result of work by several different interpreters
  • comparison of the temporal maps through use of
    overlay techniques reveals that they are indeed
    quite generalised and moreover they do not
    entirely co-register or have the same locational
    reference points
  • other sources of error may have arisen in the
    derivation of the original maps e.g. different
    interpreters, different scales of photography. It
    is more than likely that all the maps contain
    errors, but it is difficult to assess this
    quantitatively or to be able to report it easily
  • inspection of the additional reports by Raffaelli
    et al. (1988, 1989, 1996) reveals some useful
    information about the areal extent of weed mats
    as measured on aerial photography taken at
    different points during the season.
  • Raffelli and Plomer (1989, p.1) also indicate
    that the 1989 survey was carried out two weeks
    later than in previous years resulting in the
    weed mat being past its peak.
  • the fact that the survey of 1986 was mapped from
    50 colour transparencies via projection onto
    paper (Raffaelli and Nicol, 1988) suggests some
    possible sources of inaccuracy as far as the
    mapping and weedmat area estimates are concerned.

Comparisons
27
However, bearing these limitations in mind it is
possible to locate four ground control points on
each map that appear to be the 'same' and to
geo-correct each of the maps, sufficient to
facilitate overlay with the geo-corrected aerial
photographs and mosaics, as well as with the maps
for each date. This allows some comparisons to be
made between different years and potentially
forms a basis within the GIS for showing change
over time that can then be mapped as a 'change'
map (using different colours and shadings) for
each year, or to be animated. Reference to the
photographic interpretations of the weedmats does
reveal, however, that
  • spatial location of the weedmats is often similar
    for each year
  • there are definite differences in areal weedmat
    coverage from year to year (as might be expected)
  • comparisons between the different interpretations
    available show that strong similarities exist but
    that differences do arise possibly due to the
    interpreter's choice of the minimum mappable unit
    (mmu), the accuracy of the rectified imagery
    being used to map from, the weedmat units which
    are identified and drawn around (greater or
    lesser detail), and the details of the boundaries
    drawn (see Green and Hartley, 2000). This
    suggests that quantitative measurements of the
    areal weedmat coverage - in terms of how many
    metres squared were present at each date - may be
    larger or smaller than they really are at any one
    time, or alternatively may be less than they
    really are due to the time of year that the
    photography was acquired and the interpretation.

Comparison
28
  • The input of geo-referenced datasets to ArcView
    3.2 makes it possible to undertake some
  • simple GIS analyses to
  • show the spatial distribution and pattern of
    algal weedmats over time (overlay analysis)
  • show the changes in the areal distribution of the
    weedmats between e.g. time 1 and time 2 (T1-T2 -
    a simple form of change detection or
    differencing)
  • estimate the total areal coverage of weedmats in
    metres squared for each date
  • estimate the percentage change over time

An ArcView extension, XTools, was downloaded and
installed to select polygons in a theme or
coverage that has geographic co-ordinates and to
derive e.g. area measurements in metres squared,
acres or hectares for each polygon. It is also
possible to make use of the Swipe, Blend and Fade
tools in the Erdas Imagine 8.4 Viewer to provide
a visual basis for observing changes in weedmat
distribution, pattern and coverage over time. A
number of different maps using the View, Layout
options can be derived for the Ythan estuary from
the environmental map and image databases created.
GIS Map Output
  • Photomosaic and Vector Overlay
  • Weedmat Coverage Maps for each year
  • Composite Weedmat Coverage Maps
  • Classes of Weedmat Coverage for each year

29

1989
1992
GIS Map Output
2000
1994
30

1989
1992
GIS Map Output
1994
2000
31

Results
32

Photo-interpretation experience suggests that if
aerial photography were to be flown again, for
repeat monitoring and mapping of weedmats ideally
it is
  • best to acquire either high quality large scale
    vertical colour aerial photographs at a scale of
    13,000-15,000, or high quality large scale
    vertical colour infrared (CIR) aerial photographs
    at a scale of 13,000-15,000 in a 9"x9" format
    preferably with fiducial marks and a GPS (note
    other examples of work on the use of remote
    sensing for weedmats have made use of scales of
    17,200, 18,000, 110,000, 112,000 and 116,666)

In both cases the high spatial resolution of the
photography enables an interpreter to(a) both
see and locate the algal weed-mats because of the
high contrast between the 'vegetation' and the
substrate (mudflats, sandflats) and (b) map
the boundaries of the identified units with
better accuracy

Recommendations
It is suggested that if more aerial photography
is to be acquired in the future for an extension
of this work that
  • For comparisons with previous aerial
    photographically derived data and information
    that it is acquired at approximately the same
    time as in previous overflights using the same
    camera setup and film at the same scale
  • Some aerial overflights are also made for other
    times during the season to try to ascertain
    whether or not there is an optimum period for
    aerial photographic acquisition to assist in the
    identification and mapping process. The optimum
    overflights could (within reason bearing in mind
    aircraft/pilot availability, prevailing weather
    conditions, cloud cover, etc.) be established by
    examining the previous aerial photography
    acquired and other information sources and
    graphs such as provided by Raffaelli et al.
    (1999, p. 108).

33

  • Colour infrared (CIR) and/or multispectral aerial
    photography is considered in the future for such
    studies, the former offering the opportunity to
    maximise the identification and separability of
    the macro-algal communities from the substrate.
    Young et al. (1999) also suggest the use of
    colour negative film.
  • Other sources of remotely sensed imagery are
    considered as alternatives to aerial photography
    that potentially offer a similar (or better
    spatial resolution) and better spectral
    resolution
  • It would be highly appropriate to gather some
    field or ground-based data and information
    coincident with the aerial photography to assist
    in characterising the photographic appearance of
    the macro-algal weedmats for different localities
    and densities (see for example work by Young et
    al. (1998))

Recommendations
34
  • Future Studies are now underway to examine the
    problem of weed mat monitoring and mapping in the
    Ythan
  • As part of this current work due account has also
    been taken of the long-term requirements of the
    Ythan Project
  • The catalogue of datasets used for this project
    provides a useful reference and record for any
    future work and for any further comparisons
    between digital datasets
  • The digital datasets generated also form a useful
    resource and baseline database for the Ythan
    estuary, which could eventually form the basis
    for a GIS for the Ythan
  • Subsequently the map datasets could also be
    utilised within an online mapserver e.g. ArcIMS
    to provide a means for provision of public access
    to environmental information about the Ythan
    estuary and its surrounds
  • This work has also provided a basis upon which it
    is now possible to compare and contrast the work
    already undertaken by Raffaelli et al. on the
    Ythan by creating new weedmat map
    interpretations for comparison with the original
    maps and aerial photographs and for digital
    capture of the Raffaelli et al. datasets for
    integrated analysis and comparison with new
    interpretations
  • Compatibility of the geo-corrected photographic
    datasets (single and mosaiced) with other digital
    GIS datasets (e.g. SNH and other data) will also
    be useful for future GIS analyses
  • The practical methodology developed here will
    also be useful for any subsequent comparisons
    with information derived from other remotely
    sensed image sources e.g. CASI, ATM and LIDAR.
    Additionally, the information derived can also be
    analysed within a GIS. LIDAR data, for example,
    offers the opportunity to acquire digital height
    data and to derive a Digital Elevation Model
    (DEM).

Future Studies
35
  • This may subsequently be useful for image
    display, interpretation and classification,
    particularly if a 3-D terrain model with imagery
    draped over the top has value for visualisation
    purposes in the public domain.
  • Assuming that one of the long-term goals of this
    work is to generate information about weedmat
    coverage and extent over time, it may be
    appropriate to consider the need to develop a
    similar or compatible methodology for mapping
    weedmats from different sources of remotely
    sensed data/imagery that may be compared with the
    work done here and by Raffaelli et al. This will
    be particularly important to consider if any
    observations are to be made about spatial pattern
    and distribution over time as well as for
    quantitative change detection.
  • Based upon the information provided in the
    documentation available, analysis of CASI imagery
    and LIDAR data may also be used in future studies
    to determine the role that they (remote sensing
    and heighting data respectively) can play in
    monitoring, identifying, and mapping macro-algal
    weedmats. CASI imagery, for example, offers finer
    multi-spectral resolution than aerial photography
    that may be of help in the identification of the
    macro-algal weedmats. Similarly with Airborne
    Thematic Mapper (ATM) which has successfully been
    used by NERC for macro-algal weedmat monitoring
    and mapping in the south of England.
  • The application of digital image processing of
    aerial photography is not straightforward and
    classification of the weedmats versus the
    surroundings is not always successful and
    repeatable. This is a view similar to that of
    Young et al. (1999, p.1176) who observed that
    'the use of digital image processing techniques
    alone to classify. was problematic'.. because
    macroalgae, for example, share the spectral
    characteristics of other classes in the visual
    portion of the electromagnetic spectrum.
    Alternative approaches might be to utilise
    Definien's ECognition software that incorporates
    contextual information into the spectral
    classification. A solution proposed by Young et
    al. (1999, p. 1176) is 'to use these techniques
    as a first step in the classification process,
    followed by a visual examination of the imagery,
    and to edit the results in a traditional approach
    to photointerpretation'.

Future Studies
36

Baban, S.M.J., 1997. Environmental Monitoring of
Estuaries Estimating and mapping Various
Environmental Indicators in Breydon Water
Estuary, U.K., Using Landsat TM Imagery.
Estuarine, Coastal and Shelf Science, Vol. 44,
pp. 589-598. Baily, B., Collier, P., Fontana, D.,
and Pearson, A., 2002. Mapping the Inter-tidal
Vegetation of Southern English Coastal Harbours
for Pollution Monitoring and Habitat Conservation
Using False Colour Infrared Photography.
24p. Bolam, S.G., Fernandes, T.F., Read, P., and
Raffaelli, D., 2000. Effects of Macro-algal Mats
on Intertidal Sand flats An Experimental Study.
Journal of Experimental Marine Biology and
Ecology. Vol. 249(2000) 123-137. Clinton, P.J.,
Young, D.R., Robbins, B.D., and Specht, D.T.,
2000. Issues in Digital Image Processing of
Aerial Photography for Mapping Submersed Aquatic
Vegetation. In, Proceedings of the Sixth
International Conference Remote Sensing for
Marine and Coastal Environments, Charleston SC,
May 1-3, 2000. Veridian ERIM International, Ann
Arbor MI, ISSN 1066-3711. Vol. 2, pp.
292-297. Daverhog, M., and Lindstrom, A., 2001.
Remote Sensing of Filamentous Algae in Shallow
Waters Along the Swedish West Coast. EU Life
Algae. 50p. Appendices. Gorman, M. and
Raffaelli, D. (1993). The Ythan Estuary.
Biologist (1993). Vol. 40/110-13. Green, D.R.,
and King, S.D., 2000. Practical Use of Remotely
Sensed Data and Imagery for Biological and
Ecological Habitat Monitoring in the Coastal
Zone. JNCC/English Nature. 69p. Green, D. R.,
King, S. D., Morton, D. C. 1998. Small-scale
airborne data acquisition systems to monitor and
map the coastal environment. Proceedings of the
5th International Conference Remote Sensing for
Marine and Coastal Environments, San Diego,
California, 5-7 October 1998, 1479-1486. Jernakof
f, P., Hick, P., Ong, C., Hosja, W., and Grigo,
S. 1996. Remote Sensing of Algal Blloms in the
Swan River. CSIRO Blue-Green Algal
Multi-Divisional Program. CSIRO Marine
Laboratories Report 226. 39p. Kidwai, S., and
Raffaelli, D. 2000. Food Selection and Growth
Patterns of Crangon crangon L. in the Presence of
Macro-algal Mats Field Experiment on the Ythan
estuary (Aberdeenshire, Scotland). Pakistan
Journal of Zoology. Vol. 32(3) 201-205. Pugh,
K.B. (1993). The Nutrient Status of the Ythan
Catchment and Estuary. NERPB, Technical Report
93/1. Raffaelli, D. (2000). Interactions Between
Macro-Algal Weedmats and Invertebrates in the
Ythan estuary, Aberdeenshire Scotland. Marine
Research (2000). Vol. 5471-79. Raffaelli, D.,
Balls, P., Way, S., Patterson, I.J., Hohmann, S.,
and Corp, N. (1999). Major Long-Term Changes in
the Ecology of the Ythan estuary, Aberdeenshire,
Scotland How Important are Physical Factors?
Aquatic ConservationMarine and Freshwater
Ecosystems. Vol. 9(1999) 219-236. Raffaelli, D.,
1999. Impact of Catchment Land-Use on an
Estuarine Benthic Food Web. In J.S. Gray et al.
(Eds.) Biogeochemical Cycling and Sediment
Ecology. Kluwer Academic Publishers, The
Netherlands. Pp. 161-171. Raffaelli, D.G.,
Raven, J.A., and Poole, L.J. (1998). Ecological
Impact of Green Macro-algal Blooms. Oceanography
and Marine Biology an Annual Review 1998. Vol.
36 97-125. Raffaelli, D., and Way, S., 1996.
Long-Term Changes in the Ecology and Usage of the
Ythan Estuary. Scottish Office Report.
19p. Raffaelli, D. 1998. The Community Ecology of
the Ythan Estuary. In The Natural History of the
Ythan Ed M.L. Gorman. University of Aberdeen pp.
138-149. Raffaelli, D., and Passoni, L., 1992.
The Distribution of Weed Mats (Enteromorpha) on
the Ythan Estuary. Unpublished Report. August
1992. Raffaelli, D., Limia, J., Hull, S., and
Pont, S. (1991). Interactions Between the
Amphipod Corophium Volutator and Macro-algal Mats
on Estuarine Mudflats. Journal of the Marine
Biological Association UK (1991). Vol.
71899-908 Raffaelli, D., Hull, S., and Milne, H.
(1989). Long-Term Changes in Nutrients, Weedmats
and Shorebirds in an Estuarine System. Cah. Biol.
Mar (1989). Vol. 30259-270. Raffaelli, D., and
Plomer, F., 1989. The Distribution of Weed Mats
in the Ythan Estuary in 1989. Unpublished Report.
NCC 3p. Raffaelli, D., and Nicol, A., 1988. The
Distribution of Enteromorpha Weed in the Ythan
Estuary. Unpublished Report. 4p. Stapleton, C.,
and Pethick, J., 1996. Coastal Processes and
Management of Scottish Estuaries. III The Dee,
Don and Ythan Estuaries. Scottish Natural
Heritage Review No. 52. Ripley, H., Jones, W.,
Scheibling, R., 2002. Mapping Invasive Species
Assessing the Impact of Species Invasions on
Coastal Ecosystems Using Remote Sensing.
Backscatter. Fall/Winter 2002. pp. 8-12. Stove,
G.C., 1978. The Hydrography, Circulation and
Sediment Movements of the Ythan Estuary.
Unpublished PhD Thesis. University of
Aberdeen. Young, D.R., Cline, S.P., Specht, D.T.,
Clinton, P.J, Robbins, B.D., and Lamberson, J.O.,
2000. Mapping Spatial/Temporal Distributions of
Green Macroalgae in a Pacific Northwest Coastal
Estuary via Small Format Color Infrared Aerial
Photography. In, Proceedings of the Sixth
International Conference Remote Sensing for
Marine and Coastal Environments, Charleston,
South Carolina, May 1-3, 2000. Veridian ERIM
International, Ann Arbor, Michigan, ISSN
1066-3711, Vol. 2, pp. II-285-II-291. Young,
D.R., and Specht, D.T., Robbins, B.D., and
Clinton, P.J., 1999. Delineation of Pacific
Northwest SAVs from Aerial Photography Natural
Color or Color Infrared Film. In, Proceedings
From Image to Information. 1999 ASPRS Annual
Conference, Oregon Convention Center, Portland,
Oregon, May 17-21, 1999. pp. 1173-1178. Young,
D.R, Specht, D.T., Clinton, P., II Lee, H., 1998.
Use of Color Infrared Aerial Photography to Map
Distributions of Eelgrass and Green Macroalgae in
a Non-Urbanized Estuary of the Pacific Northwest,
USA. In Proceedings of the Fifth International
Conferenceon Remote Sensing for Marine and
Coastal Environments, Vol. II37-45, 1998, ERIM
International Inc., Ann Arbor, pp. II-37-II45.
References
37
Thank You
In conclusion, the combination of digital data
capture, conversion and input capabilities,
digital image processing and GIS technology
offers considerable potential for integrating
archival map and image datasets, as well as
providing the potential to create a baseline
GIS-based system as the basis for future
multi-temporal monitoring and mapping of
environmental change of macro-algal weedmats in
the Ythan.
  • David R. Green Stephen D.King
  • Centre for Marine and Coastal Zone Management
  • Department of Geography and Environment
  • School of Resources, Environment and Society
  • University of Aberdeen, Elphinstone Road
  • Aberdeen, AB24 3UF, Scotland, UK
  • Tel. 44 (0)1224 272324
  • Fax. 44 (0)1224 272331
  • Email. d.r.green_at_abdn.ac.uk / s.d.king_at_abdn.ac.uk
  • Internet. http//www.abdn.ac.uk/cmczm
Write a Comment
User Comments (0)
About PowerShow.com