GIS for the Study of Beach Morphodynamics
175
the existence of the coastal underground mines (Fig. 1). The thick dashed line
indicates the cliff trace digitised from 1:10000 OS maps. Boxes indicate the study
areas. The “Main” and “Yard” colliery panels close to the Dawdon beach area have
been worked during the late 1980’s. “High Main” has been worked during the 1960’s
and all the other panels prior to 1947.
This chapter describes the work that was carried out using a GIS to compile
and analyse the data. Results were then compared with theoretical values and areas of
surface impact of predicted subsidence in the area.
Data Development and Analysis
a. Survey points
The main body of the field research has been conducted by using a total station
(Geodimeter 400) to survey several times a year, three profile lines on each of the main
beaches affected by the spoil at Dawdon, Easington and Horden (Fig. 1).
Measurements were taken at regular time intervals to cover seasonal variations.
Instrument error is estimated to be less than 1 cm. OS maps (1:10000) were used as a
backcloth for the cliff line and high/low water marks.
The GIS software that was used was the ArcInfo Workstation version which
can handle both vector and raster data in surface analysis.
Survey points from beach profile data were used to investigate the
development of beach surface over the study period. From these datasets, the
corresponding TINS for every study area were constructed. Lattices representing the
elevation maps of the beaches were produced for every survey that was carried out
using bivariate quintic interpolation of the related TINS. The smoothing effects of this
method were considered appropriate since we were dealing with relatively small TINS
and map extents in each beach. The cell size was set to 10 m. This spacing was
considered appropriate taking into account the distance between survey points and
profile lines. To assess the error in the elevation data, the lattice containing only the
measured elevation at the survey points was compared to the lattice containing only the
extracted interpolated heights at the same points. The mean difference of the resulting
lattice is very close to 0, so the standard deviation can be considered as an estimate of
the RMS error. The error of any two resulting maps after overlay operations was
calculated as the square root of the sum of squared errors of the two maps. Table 1
shows the errors (m) for the resulting difference maps.
The elevation lattices for all 3 beaches were then used to produce slope and
aspect maps, one for each survey. Data collected in March and June 1991 were used as
“base” data from which changes in beach elevation for the coming months were
examined. The mean values of slope, aspect and “difference” elevation data were also
studied over the five year period, taking into account the estimated error, in order to
assess any variability of the mean size of these characteristics with time (Figs 2, 3, 4).
However, the study concentrated on identifying spatial variations and patterns.
Elevation differences were studied as “cut” and “fill” changes of the beach surface.
These changes were further analysed by assuming thresholds. Mean “cut” values were
175
the existence of the coastal underground mines (Fig. 1). The thick dashed line
indicates the cliff trace digitised from 1:10000 OS maps. Boxes indicate the study
areas. The “Main” and “Yard” colliery panels close to the Dawdon beach area have
been worked during the late 1980’s. “High Main” has been worked during the 1960’s
and all the other panels prior to 1947.
This chapter describes the work that was carried out using a GIS to compile
and analyse the data. Results were then compared with theoretical values and areas of
surface impact of predicted subsidence in the area.
Data Development and Analysis
a. Survey points
The main body of the field research has been conducted by using a total station
(Geodimeter 400) to survey several times a year, three profile lines on each of the main
beaches affected by the spoil at Dawdon, Easington and Horden (Fig. 1).
Measurements were taken at regular time intervals to cover seasonal variations.
Instrument error is estimated to be less than 1 cm. OS maps (1:10000) were used as a
backcloth for the cliff line and high/low water marks.
The GIS software that was used was the ArcInfo Workstation version which
can handle both vector and raster data in surface analysis.
Survey points from beach profile data were used to investigate the
development of beach surface over the study period. From these datasets, the
corresponding TINS for every study area were constructed. Lattices representing the
elevation maps of the beaches were produced for every survey that was carried out
using bivariate quintic interpolation of the related TINS. The smoothing effects of this
method were considered appropriate since we were dealing with relatively small TINS
and map extents in each beach. The cell size was set to 10 m. This spacing was
considered appropriate taking into account the distance between survey points and
profile lines. To assess the error in the elevation data, the lattice containing only the
measured elevation at the survey points was compared to the lattice containing only the
extracted interpolated heights at the same points. The mean difference of the resulting
lattice is very close to 0, so the standard deviation can be considered as an estimate of
the RMS error. The error of any two resulting maps after overlay operations was
calculated as the square root of the sum of squared errors of the two maps. Table 1
shows the errors (m) for the resulting difference maps.
The elevation lattices for all 3 beaches were then used to produce slope and
aspect maps, one for each survey. Data collected in March and June 1991 were used as
“base” data from which changes in beach elevation for the coming months were
examined. The mean values of slope, aspect and “difference” elevation data were also
studied over the five year period, taking into account the estimated error, in order to
assess any variability of the mean size of these characteristics with time (Figs 2, 3, 4).
However, the study concentrated on identifying spatial variations and patterns.
Elevation differences were studied as “cut” and “fill” changes of the beach surface.
These changes were further analysed by assuming thresholds. Mean “cut” values were