Denudation processes and rates in the Transverse Ranges, southern California: Erosional response of a transitional landscape to external and anthropogenic forcing

[1] Quantification of denudation in the Transverse Ranges of California permits reconstruction of spatial and temporal variations in erosion that represent both the response and evolution of interacting hillslopes and channels. On the southern front of the San Gabriel Mountains, observational records of the infilling of debris basins and dams define twentieth-century landscape erosion rates averaging 1.6 and 0.9 mm yr , respectively. Although all major sediment transport occurs during intense winter storms, debris production on hillslopes is greatly enhanced by recurrent fires. Consequently, in this populated region, anthropogenic fires have augmented the natural erosion rates. We perform a global inversion to estimate the role of precipitation intensity, burned areas, and local slope on catchment denudation rates. After subtracting the effects of anthropogenic fires we estimate landscape denudation rates given natural fire ignition rates. Increased fire during the past century has augmented sediment production in debris basins by an average of 60%, and individual basins show increases up to 400%. To identify the dominant hillslope erosion processes, the volumetric contribution of landslides was estimated using repeat aerial photographs for the same time interval over which the debris basins have been operative. Between 1928 and 1973, landsliding produced only 10% of the sediment in debris basins. Even in the long term, when infrequent but volumetrically important landslides occur, bedrock landslides appear to contribute a maximum of 50% to the long-term landscape denudation. Previous mapping of soil slippage in the San Dimas Experimental Forest within the San Gabriel Mountains [Rice et al., 1969; Rice and Foggin, 1971] indicates that shallow landsliding is likely to be the dominant modern hillslope erosion process. When compared to incision rates derived from a fluvial shear stress model and to exhumation rates based on low-temperature thermochronological data [Blythe et al., 2000], modern ‘‘natural’’ erosion rates are comparable to denudation rates since the Pliocene. Comparisons of modern erosion rates suggest that debris production on hillslopes and first-order channels is directly dependent on vegetation cover and precipitation intensity. For higher-order channels (drainage areas >2 km), only major storms convey the sediments down valley. During the past century, some temporal decorrelation occurs between small-scale and large-scale catchments because hillslopeproduced sediment is stored in second-order and higher channels until major storms mobilize it. Thus the larger fluvial network damps the episodic, fire-induced hillslope sediment pulses that occur within small watersheds. At a longer temporal scale, however, uplift and denudation may have been sustained sufficiently long in much of the San Gabriel Mountains for the topography to reach a macroscale steady state. In contrast to rapidly eroding ( 2 mm yr ) ranges, for which quantification of bedrock landsliding will approximate the sediment flux, the San Gabriel Mountains occupy a niche of intermediate rates (0.1–1.0 mm yr ) in which a broad suite of hillslope processes, including shallow-seated and deep-seated landslides, debris flows, and wet and dry ravel, JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 109, F01006, doi:10.1029/2003JF000023, 2004