Geoscience Reference
In-Depth Information
Figure 7.2
Location maps for the two case studies described in the text: (a) the location of the River Wye catchment
(case study one) and the 618 matched monitoring sites (case study two), and (b) the main rivers and towns in the
Wye catchment.
©
Crown copyright, Ordnance Survey. All rights reserved.
(Cressie, 1993) was used to produce complete
coverage across the Wye catchment at a resolution
(pixel size) of 500 m for both variables; the
10% of pixels with the highest BOD or extent
of resectioning was then selected (Figure 7.3).
Although conventional two-dimensional kriging of
this type overlooks the river network structure and
nesting of sub-catchments, it serves to illustrate the
principles. More valid methods are currently under
development (Peterson
et al
., 2007; Peterson and
Ver Hoef, 2010).
The highest BOD concentrations were found
along the main channel in the lower Wye
catchment and also in the River Frome to the north
east (Figure 7.3). This pattern reflects the influence
of Hereford, the largest town in the catchment,
and those areas with high concentrations of
arable and dairy farming. Similarly, channel re-
sectioning was most intense in the east of the
catchment and closely associated with the main
river channel, especially between Hereford and
Ross-on-Wye (Figure 7.3). The areas comprising
the top 10% for both pressures overlapped, with
most of the resectioned area also having elevated
BOD concentrations (Figure 7.3). The coincidence
between pressures is readily apparent when stored
in the form of separate data layers in a geographical
information system, where the different maps can
be superimposed.
The occurrence of multiple stressors identified
in this way immediately highlights potential 'hot
spots' which could become the focus of river and
catchment management interventions, especially
if the stressors have synergistic relationships
(Newson, 2010). From a research perspective,
background information on the distribution of
potential stressors could be invaluable for field
sampling design, whether to avoid confounding
factors or directly study multiple stressors. In
these situations, monitoring data represent a
rich,
yet under-used source of
background
information.
