Extreme rainfall in June 1949 and November 1985 triggered numerous large debris flows on the steep slopes of North Fork Mountain, eastern West Virginia. Detailed mapping at four sites and field observations of several others indicate that the debris flows began in steep hillslope hollows, propagated downslope through the channel system, eroded channel sediment, produced complex distributions of deposits in lower gradient channels, and delivered sediment to floodwaters beyond the debris-flow termini. Based on the distribution of deposits and eroded surfaces, up to four zones were identified with each debris flow: an upper failure zone, a middle transport/erosion zone, a lower deposition zone, and a sediment-laden floodwater zone immediately downstream from the debris-flow terminus. Geomorphic effects of the debris flows in these zones are spatially variable. The initiation of debris flows in the failure zones and passage through the transport/erosion zones are characterized by degradation; 2300 to 17 000 m 3 of sediment was eroded from these zones. The total volume of channel erosion in the transport/erosion zones was 1·3 to 1·5 times greater than the total volume of sediment that initially failed, indicating that the debris flows were effective erosive agents as they travelled through the transport/erosion zones. The overall response in the deposition zones was aggradation. However, up to 43 per cent of the sediment delivered to these zones was eroded by floodwaters from joining tributaries immediately after debris-flow deposition. This sediment was incorporated into floodwaters downstream from the debris-flow termini causing considerable erosion and deposition in these channels.

A model of cave sedimentation and genesis is used to gain greater resolution and accuracy in the calculation of an incision rate for Cheat River, West Virginia. Maze caves along the river and their primary sediments were created and deposited beneath base level. Single conduit caves are largely unrelated to base level and their sediments are derived from overlying strata. A magnetostratigraphic record is reported for cave sediments within the canyon. The magnetostratigraphy of each sample is plotted versus elevation relative to base level and depositional environment (vadose or phreatic). The resulting chart accurately depicts the range of error associated with using cave sediments as indicators of previous base-level positions. This technique can be applied within any future studies using cave sediments for deriving incision rates of rivers. The calculated incision rate of Cheat River within the study area is between 56.0 and 63.2 mm/k.y.

No abstract

Several glacial geologists working in Maritime Canada have developed a view of maximum late Wisconsin glaciation in which the Gulf of St. Lawrence and the Gulf of Maine were substantially ice-free, and ice calving into these embayments came from local ice domes, except for Laurentide ice calving into the Gulf of St. Lawrence along the Laurentide Scarp of Quebec. In contrast, most glacial geologists working in New England believe that Laurentide ice overrode all of New England and terminated as marine ice streams in the Gulf of St. Lawrence and the Gulf of Maine, and as terrestrial ice lobes in the Hudson River valley, the Connecticut River valley, and Great South Channel between Nantucket Shoals and Georges Bank. The view of New England glacial geologists is supported by numerous radiocarbon dates. Glaciological models of ice dynamics along an ice-sheet margin consisting of local ice domes, ice streams, ice lobes, ice shelves, and calving bays have been applied to the Maritime Canada/New England region. These models show that the viewpoint of glacial geologists in Maritime Canada is compatible with conditions during deglaciation, but that the viewpoint of glacial geologists in New England is compatible with conditions during the late Wisconsin maximum. The undated ice margin north of the Gulf of Maine favored by Maritime Canadians cannot coexist with the dated ice margin south of the Gulf of Maine mapped by New Englanders. The ice divide between a Laurentide ice dome in Noveau Quebec or Hudson Bay and a local ice dome in the northern Appalachians must have had a saddle over the St. Lawrence River valley. A saddle is incompatible with the late Wisconsin glacial geology, but is consistent with the glacial features formed during retreat. 139 on June 21, 2015 specialpapers.gsapubs.org Downloaded from 140 T. Hughes and Others

The probabilistic mapping of landslide occurrence at a high spatial resolution and over a large geographic extent is explored using random forests (RF) machine learning; light detection and ranging (LiDAR)-derived terrain variables; additional variables relating to lithology, soils, distance to roads and streams and cost distance to roads and streams; and training data interpreted from high spatial resolution LiDAR-derivatives. Using a large training set and all predictor variables, an area under the receiver operating characteristic (ROC) curve (AUC) of 0.946 is obtained. Our findings highlight the value of a large training dataset, the incorporation of a variety of terrain variables and the use of variable window sizes to characterize the landscape at different spatial scales. We also document important variables for mapping slope failures. Our results suggest that feature selection is not required to improve the RF modeling results and that incorporating multiple models using different pseudo absence samples is not necessary. From our findings and based on a review of prior studies, we make recommendations for high spatial resolution, large-area slope failure probabilistic mapping.