Abstract
This thesis investigates the misfit valleys of Jersey, Channel Islands. Data on the geology, topography and Quaternary history of the Island provides the geomorphological context to the study. The valleys are suggested to be misfits as they either lack any evidence of recent fluvial activity, or the streams that flow though them are so small that they could not logically be expected to have eroded the valleys. The discharge record of the one gauging station in Jersey is analysed in order to determine whether supporting evidence of stream underfitness can be drawn from these data. This record is too short (only 7 years) to allow meaningful conclusions. Therefore the Jersey precipitation record is used as a proxy for discharge. Examination of the precipitation record shows that trends of increasing winter precipitation and decreasing summer precipitation are apparent. These changes are related to those of the surrounding region and possible causes are suggested. To determine if the valleys were the product of processes operating during the 130 year period of the precipitation record, the Soil Conservation Service (1973) ‘curve numbers’ approach is used to estimate catchment discharge from the precipitation record. A runoff-erosion model is used to estimate the possible amount of erosion produced by these estimated discharges. From this it is suggested that valley formation could not be reasonably expected to have occurred over this time period. Extrapolation of this record to cover a longer period suggests that valley formation was unlikely to have occurred during either the past 1000 years, or over the duration of the Holocene. Given that valley formation is unlikely to have occurred under contemporary climates, it is necessary to consider how formation might be possible. This involves an extensive review of existing theories of misfit valley formation. By a process of elimination it is concluded that valley formation was most likely to have occurred during the cold periods of the Quaternary. The current form of the valleys is most probably a result of processes operating during the most recent, Devensian, cold period. It is suggested that valley formation was a result of a combination of powerful spring snow melt floods and the presence of permafrost that would reduce the amount of infiltration of snowmelt. A review of geomorphological processes in analogous catchments suggests this mechanism is feasible. In order to gain as much support as is possible for this theory, an extensive investigation of the morphometry of the valleys is performed. This aims to demonstrate that the morphometry of the Jersey catchments is dissimilar from other temperate valleys but comparable to periglacial permafrost catchments. Before this is possible, a historical review of morphometry and fluvial geomorphology is conducted to provide an academic context to this research. This traces the origins, development and application of morphometric methodology to fluvial geomorphology, the concept of drainage density, and the impacts of fractals on fluvial morphometry. A range of models of drainage network structure is discussed, including so called ‘Hortonian analysis’, the infinite topologically random networks model, and the optimal channel network model. The morphometry of the valley networks is examined, together with the methodology used to collect these data. This analysis investigates the structure of the drainage networks, stream orientations, basin areas, gradients, length and width. The variation of these with geology is examined, and it is concluded that there are no statistically significant variations. The possible variation of drainage density with a range of morphometric parameters is investigated. Again, no statistically significant variations are found within and between catchments, between differing geologies, or with drainage basin area. Possible explanations for this are suggested: that the presence of permafrost during valley formation reduced the impact of geological differences; that the extremely high discharges during valley formation rendered such differences irrelevant; or that the close proximity of the catchments means that great variations in other environmental factors did not occur. The use of the Jarvis E-index is discussed and applied to the Jersey catchments, suggesting that whilst this is a useful measure of network ‘compactness’, it is of limited value. The theoretical issues involved with use of drainage density-discharge relationships are discussed together with the validity of this methodology. This concludes that whilst this approach does have problems, it is of value and may be applied to the Jersey catchments. The topological structure of the valley networks is investigated, specifically whether valley lines and networks are fractal features. This is determined by calculating the fractal dimension of the valleys. If a non-integer value is produced then the valleys may be suggested to be fractal, claimed to be the most energetically efficient form. Estimation of the fractal dimension of the stream lines can be determined using five methods. These are, using the exponent in the main stream length-catchment area relationship, derivation from the ‘Horton ratios’, the Richardson or divider method, a modification of this, and functional box counting. This investigation shows disagreement between the various methods when applied to any given valley. This is interpreted to mean that the various methods can not be applied globally, and that the Jersey valleys are not fractal features. A range of explanations is given to account for this, including the periglacial origin of the Jersey valleys in contrast to the temperate origins of most other valleys used to estimate the fractal dimension. This origin of the valleys is suggested to lead to a network structure that whilst being energetically efficient, is not fractal. Asymmetric valley cross sections are often interpreted as being indicative of valley formation under periglacial conditions, as such this would provide important supporting evidence for a periglacial origin for the Jersey valleys. An extensive review of existing theories on the formation of such asymmetric valleys points to the wide range of theories to account for these forms. Research on Jersey suggests that the valley cross sections are indeed asymmetric. This utilised two techniques, firstly using 1:10,000 scale maps, a series of transects were taken across the valleys, and contour heights and separations were used to measure slope gradients. Secondly a digital elevation model was developed for one catchment (St. Peter’s) and a GIS was used to investigate asymmetry within this catchment. This research suggests that there is no single simple trend of one slope aspect being consistently steeper across the whole of the Island. (Although research in St. Peter’s catchment suggests that north and east facing slopes are steeper than any other aspect.) This is taken to be the result of differing radiation receipts leading to differing amounts of slope activity and hence gradients during periglacial climates. A range of possible mechanisms is proposed for how such differing receipts result in differing slope gradients, and various interpretations are proposed, including analogues involving chaos theory. Finally, an attempt is made to estimate the discharge that formed the misfit valleys. A variety of existing morphometric methods of discharge estimation is applied, producing a range of discharge estimates. Many of these approaches are not directly applicable to the hypothesised periglacial origin of the Jersey valleys. Application of these methods is conducted in order to gain the widest possible range of discharge estimates. These estimates are compared to data from periglacial catchments in Alaska to determine whether these estimates are plausible values for periglacial catchments. This allows the immediate rejection of an array of methods, resulting in a selection of approaches that produce discharge per unit area values similar to the Alaskan data. These values suggest that the fluvial discharges that formed the Jersey valleys were some 4 to 6 times greater than contemporary discharges.