Saltmarsh tidal channels have often been recognized as being stable landscape features, despite highly sinuous planforms, severely undercut banks, and high rates of bank erosion. In an effort to solve this paradox, a saltmarsh tidal channel in the San Francisco Bay was monitored from March 1995 to March 1996. The short-term rate of bank erosion was measured using erosion pins and found to be 57 +/- 10 mm yr(-1) on the outside banks of meander bends. In addition, a long-term maximum lateral migration rate of 23 +/- 23 mm yr(-1) was estimated from aerial photos, producing a dimensionless channel migration rate (defined as the rate of migration divided by channel width) of 0.5% yr(-1). The difference in the rates of lateral migration and bank erosion is attributed to the persistence of failed bank material (slump blocks) in the channel. The slump blocks induce sedimentation, protect the banks, and prevent further bank erosion. A published stability analysis method for undercut banks is applied to determine a maximum overhanging width. Using the measured compressive and tensile strengths of rooted bank material, 16.55 +/- 1.16 Wa and 2.93 +/- 0.71 kPa, respectively, the maximum width of an undercut bank is calculated to be 0.69 m. The average width of slump blocks measured in the field is 0.67 +/- 0.25 m. A simple numerical model predicting the rate of lateral migration is derived using the results from the stability analysis and data from sedimentation and erosion pins inserted throughout the channel. This model accurately predicts a rate of 23 +/- 3 mm yr(-1).
