"coal
es of
ound
igital
is for
eated
vith a
ed by
eriod.
areas
Xf the
> area
istinct
gwall
of the
ult to
e may
future
ven if
ars of
Juence
due to
ons of
of the
alfield
ure |).
idence
epends
se. A
part of
estone.
ind the
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004
development of crown holes (Young and Lawrence, 2002).
This mechanism of failure is associated with geological
discontinuities and faults (Wigham, 2002).
fissuring associated with
geological discontinuities
jJ - 3
t
n + /
tension fractures ” ^
e I nem -
^ fi
Td iM pei ni
m
pressure ridges.
"S
ut . TUN i C 4 f
crushed pillar
Figure 1. Surface expression df subsidence due to collapse of
support pillar in underground coal workings (after
Bell and de Bruyn, 1999)
Using examples of subsidence observed during field surveys of
the Houghton-le-Spring area it has been possible characterise
the different subsidence features in terms of their topographic
expression and appearance. The disruption of groundwater
flow and surface hydrology, and the break up of the surface
caused by subsidence affect the condition and species of
vegetation growing in the proximity of subsidence features.
Water can pool in surface depressions, saturating the soil.
Biomass may be increased due to the water until vegetation
becomes water logged beyond critical tolerance levels resulting
in stress and reduced productivity (Steven et al. 1992),
Conversely, fissuring may lead to preferential drainage of an
area, resulting in vegetation stress or creating a niche
environment for a different species.
1.3 Airborne remote sensing
It has been observed that some examples of subsidence
observed in the field have a distinct topographic expression but
no anomalous spectral response. Conversely, examples have
also been noted that are too small to be observed at the scale of
the aerial photography but do have associated spectral
anomalies in hyper spectral imagery. Many examples where
found to have both a topographic and a spectral expression.
This research integrates the two disciplines of photogrammetry
and hyper spectral image processing in order to maximise the
value of airborne imagery for subsidence mapping and
monitoring.
Subsidence research is active in both areas but tends to proceed
in isolation. This work has involved developments in both
areas and secks to demonstrate that an integrated approach,
processing aerial photography and hyper spectral imagery in
parallel and combining the results, maximises the value of the
individual data sets.
715
Conventional land surveying techniques such as levelling have
been routinely employed to investigate subsidence but are not
suitable for the repetitive surveying at a regional scale
necessary to monitor the Durham coalfield. Coal Authority
abandonment plans indicate areas of pillar and stall workings
across the extent of the coalfield and subsidence may occur in
any of these areas. Photogrammetry offers a viable alternative
to land surveying for the production of surface models and the
extraction of profiles and is a proven tool in subsidence
investigations. In the absence of old survey records,
photogrammetric processing of archive photography is the only
viable method for generating historic surface profiles for
comparison with contemporary data.
The application of remote sensing for mapping surface spectral
properties is well established. Thermal imagery can be
employed to investigate soil moisture anomalies and the
application of hyper spectral imagery in vegetation studies is a
proven tool, although with limited application to subsidence
studies.
2. DATA ACQUISITION
Three survey flights were made over the Houghton-le-Spring
test site in the Durham coalfield, in September 2002, March
2003, September 2003 by the Natural Environment Research
Council’s (NERC) Airborne Remote Sensing Facility (ARSF)
aircraft. The September surveys involved the simultaneous
capture of aerial photography with a Wild RC10 survey camera
and acquisition of imagery with a Daedalus 1268 Airborne
Thematic Mapper ^ (ATM) and Compact Airborne
Spectrographic Imager (CASI-2). No photography was
acquired during the March flight due to financial restrictions.
The two September flights were flown at different altitudes
(1800 m and 1200 m), providing photography and hyper
spectral imagery at different scales under comparable ground
conditions. The photographs were subsequently scanned using
a Vexcel, Ultrascan 5000 photogrammetric scanner for
processing in a digital photogrammetric workstation.
The flights were timed to capture data after the winter, a
prolonged wet period, and after the relatively dry summer, in
order to assess what conditions are most suitable for the
identification of subsidence features, and to characterise the
temporal variation of the natural vegetation cover.
An Analytical Spectral Devices (ASD) field spectroradiometer
from the NERC Equipment Pool for Field Spectroscopy (EPFS)
was used to acquire ground spectra from calibration targets
during the survey flights. Field spectra were acquired over
tarmac, grass and an artificial white tarpaulin target. Spectral
profiles were also captured across an area of fissured ground to
enable simulation of CASI-2 spectra for processing algorithm
development and subsequent comparison of results.
3. TOPOGRAPHIC EXPRESSION OF SUBSIDENCE
Photogrammetry is an established technique in subsidence
investigations (Faig 1984, LaScola 1988) which has benefited
from recent advances in photogrammetry and the development
of new digital sensors (Spreckels, 1999). It offers a viable
method for monitoring mining subsidence of working mines
though the accuracy and limitations of the technique are not
well documented and work is still needed to quantify the
accuracy and precision of photogrammetry in subsidence
research.
