5. Sensitivity analysis
The constraint maps are composed of a number of layers, it is helpful to know how stable and predictable the output is at the chosen set of weightings and how changes in these can affect the output.
The sensitivity of the output to variations in the weightings of each data theme ie: environmental, industrial, technical or socio-cultural was investigated by weighting an overlay output of all the layers in each theme as per the final output except for one theme which was weighted twice as highly as the other three data themes.
Using this approach a matrix of combinations was created to show the combination of overall relative weightings used in the sensitivity test. Table 4 shows the overall relative weightings matrix resultant from the sensitivity test. The results (Figures 24-27) show that although there are regional scale differences in the outputs the overall constraint level remains higher inshore and the relative areas of minimised constraint appear in the same locations. These figures show that with a doubling of weighting on each of the themes in turn the relative output remains within similar relative ranges of constraint. Hence it is acceptable to consider that the original weighted map is an apt representation of the constraints generated by the layers and relative weightings included in the constraint model.
Table 4: Matrix showing combinations of weightings per theme. In turn each theme has been weighted twice as high as the three others to assess the effect this has on the overall output.
|Theme||Environmental weighted||Industrial weighted||Sociocultural weighted||Technical weighted|
Figure 24: Constraint output with the environmental layer set weighted at twice as much as the rest of the layers. © Crown copyright and database rights (2018) OS (100024655).
Figure 25: Constraint output with the industrial layer set weighted at twice as much as the rest of the layers. © Crown copyright and database rights (2018) OS (100024655).
Figure 26: Constraint output with the socio-cultural layer set weighted at twice as much as the rest of the layers. © Crown copyright and database rights (2018) OS (100024655).
Figure 27: Constraint output with the technical layer set weighted at twice as much as the rest of the layers. © Crown copyright and database rights (2018) OS (100024655).
The changes in each of the outputs where one data theme eg: environmental, were double weighted did not differ significantly from the output that did not alter the relative weighting of the themes. The output composed of layers weighted as per Table 1 was used for assessing the best opportunities and drawing the initial AoS.