The adoption of aerial photographs as the primary data source effectively restricted the survey to 'surface-visible' palaeochannels only (this term is proposed to denote palaeochannels manifest at the land surface in a visually detectable form, in this case from the air). This represents only a subset (albeit a large subset) of the resource in toto. In particular, the usefulness of the technique is likely to be severely impaired in the perimarine zone, where the borehole record demonstrates the presence of both large palaeochannels and extensive ancient land surfaces, yet these features are masked from the air by deep alluvial sequences (Lillie 1998a; 1998b).
In addition, there is an inverse relationship between the age of a palaeochannel and its likely visibility from the air: processes of succession render channels progressively less visible; gradual alluvial deposition will eventually mask channels. Differing land-use patterns may greatly complicate this process: unploughed floodplain grassland, for example, appears to preserve relict dry channels almost indefinitely; channels may be artificially kept open or filled in.
The possible age-range of surface-visible palaeochannels has a major bearing on the potential usefulness of the survey. Large and Petts (1996), in their survey of 28 major cut-off channels along the middle Trent, concluded that many of these channels are in fairly early stages of succession, suggesting a relatively recent origin. Navigation work – during the 18th century in particular – may account for these features, such as the meander cut-off at Sawley, Derbyshire. However, older channels may be preserved as dry depressions in floodplain grassland, as cropmarks in arable land, or may be fossilised in boundary features (hedge-lines, ditch- and stream-lines, parish boundaries) even when the palaeochannel itself is no longer visible. Channels forming parish boundaries suggest an origin within the early medieval period. Medieval palaeochannels at Hemington, Leicestershire, with dates ranging from the 8th to the 18th centuries AD (Ellis and Brown 1998; Cooper 2003) are clearly visible on aerial photographs. A surface-visible abandoned meander at Colwick, Nottinghamshire, was active during the later Iron Age or early Roman period (Salisbury et al. 1984). Outside the Trent Valley, surface-visible palaeochannels on the River North Tyne can be dated as far back as 5000 cal BC (Moores et al. 1999). So, although the surface-visible palaeochannel record is likely to be weighted in favour of more recent channels, the evidence suggests that a significant number of older channels are present. Indeed, in certain areas the entire Holocene sequence may be represented, although this cannot be predicted throughout the course of the valley.