The use of multispectral remote sensing in the management of the North York moors

This thesis examines the use of multi-spectral remotely sensed data in the management of the North York Moors, an upland area of heather moorland in northern England. A series of ground radiometer surveys and airborne simulations are analysed to determine the relative importance of spatial, spectral and temporal resolution as characteristics of earth resources satellites in this environment. Particular reference is made to the potential for selecting and combining data from the Landsat MSS, TM and the SPOT HRV sensors. The results show that spectral resolution can be critical in isolating and recognising elements of the moorland community by their spectral response, especially at the most detailed levels of vegetational description. Temporal resolution has little effect on the discrimination of targets within the heather dominated areas but affects the separability of the major communities of heather, bracken and sedges. Change in spatial resolution has no clear effect on the spectral uniformity and spectral separation of the elements of the heather dominated areas. The interaction between spectral and spatial resolution is more important in isolating the major communities, where the requirement for spatial precision is balanced against the need to suppress spectral variation within the moorland. The hypothesis that multi-spectral remotely sensed data can provide critical information on the distribution and status of moorland vegetation is not refuted in this thesis. Remotely sensed data would make the greatest contribution to management if linked to other spatial data as part of a Geographical Information System. In the absence of such a formal structure satellite imagery can still provide a regular and unique inventory of the moorland habitat which will increase the efficiency of management.