IN-SITU MONITORING OF BLUFF STABILITY IN SOUTH EAST WISCONSIN

Rising water levels in Lake Michigan are causing wave action to reach greater elevations on beaches and, in some instances, directly eroding coastal bluffs. Once over steepened at the toe, bluff failure is controlled by hillslope processes and material properties (e.g., strength, angle of internal friction and cohesion, and unit weight). Slope failure estimates are difficult to constrain from static measurements, and in order to better understand the governing processes and improve predictions properties need to monitored during an actual failure.

We have developed a new set of instruments, which were deployed along the Lake Michigan shore from July 14 to December 16 2016. These instruments record the following metrics: GPS-L1 position, soil moisture, and extension using two spool-style extensometers at a sub-second sample interval with ca. 20-micron resolution over a range of 0-3000 mm. Two sites were situated on low bluffs composed of sand overlying late Wisconsin till. The third site was situated on a high bluff composed of two clay-rich late Wisconsin tills interbedded with thin sand.

All three BADGERs recorded some amount of creep over the observational period but one instrument placed on a low bluff recorded a sudden bluff failure on Dec 13, 2016 of a 15-meter vertical section. Prior to the sudden failure there was ca. 2.5-hour period of accelerated creep at a rate of 11.2 mm/hr. This pre-failure creep rate was orders of magnitude higher than background creep rate of 2.5x10^-3 mm/hr. Approximately 15 m³ of sediment was mobilized during this one failure. Wind speeds and wave action were relatively constant leading up to the failure. Soil moisture was measured at 98% saturation in the upper 10 cm of the bluff during the time of failure. In the 36 hours prior to failure a 7-degree C temperature drop from -1 to -8 C was recorded. Such a drop-in temperature while the ground was saturated with liquid water suggests that frost heave ice lenses in the sediment may have formed. The formation of ice lenses would have provided additional tensile stresses in the sediment that may have exceeded the yield strength of the slope material. Although dynamic failure is sudden there was a period of accelerated but non-catastrophic creep for several hours prior to a sudden failure, which needs to be accounted for in numeric simulation of bluff stability.