The landscape is partitioned into hillslopes and unchanneled valleys (hollows), and colluvial (hillslope controlled) and alluvial (self-formed) channels. The key issue for any study of headwater catchments is the rational distinction between these elements. Accurate identification of process domain transitions from hillslopes to hollows, hollows to colluvial channels and colluvial to alluvial channels, are not obvious either in the field or from topographic data derived from remotely sensed data such as laser derived (LIDAR) digital elevation models. The research in this dissertation investigates the spatial arrangement of these landforms and how hillslope and fluvial process domains interact in two pairs of headwater catchments in southwest and central Montana, using LIDAR data. This dissertation uses digital terrain analysis of LIDAR-derived topography and field studies to investigate methods of detection, modeling, and prediction of process transitions from the hillslope to fluvial domains and within the fluvial domain, from colluvial to alluvial channel reaches. Inflections in the scaling relationships between landscape parameters such as flowpath length, unit stream power (a metric of the energy expended by the channel in doing work), and drainage area were used to detect transitions in flow regimes characteristic of hillslope, unchanneled valleys, and channeled landforms. Using the scale-invariant properties of fluvial systems as a threshold condition, magnitude-frequency distributions of curvature and the derivative of aspect were also used to detect hillslope, fluvial, and transitional process domains. Finally, within the classification of channeled landforms, the transition from colluvial to alluvial channels was detected using the presence/absence of repeating patterns in the power spectra of fluvial energy and channel form parameters. LIDAR-derived scaling relations and magnitude-frequency distributions successfully detected and predicted locations of mapped channel heads and hollows and spatial regions of process transitions. Subreaches of arguably alluvial channel conditions were also identified in power spectra. However, extrinsic forcing limits ability to detect a clear transition from colluvial to fully alluvial conditions. Headwater catchments present a mosaic of process domains, in large determined by local structure and lithology. However, process domain transitions appear detectable and statistically, though not deterministically, predictable, irrespective of setting.
展开▼
