We know these sources are the primary sources for sediment, but we have few constraints on their relative importance. Our initial goal is to constrain these numbers in a general sense, averaging them over decadal timescales. The ultimate goal, though, is to develop a sediment routing model that would account for the dynamic nature of the system. Different storms are likely to liberate sediment from different parts of the landscape. For example, we have observed very little sediment in ravines during light storms in which overland flow is not generated. However, we expect that larger, more intense rainfall events will result in a threshold response from the ravines that would liberate large amounts of fine sediment.