A 4,700-year record of lake level and isostasy for Lake Michigan

Four relative lake-level curves (RLLCs), produced from five sires bordering Lake Michigan, show similar timings of high and low lake-levels during the late Holocene. However, glacial isostatic-adjustments and possibly tectoni sm experienced at each site are superimposed on these records of relative l ake-level change. This effect causes the RLLCs to diverge from each other w ith time. The absolute magnitudes of lake-level fluctuations for the late H olocene can only be determined by quantifying and subtracting the component of vertical ground-movement from each RLLC. Both an exponential rate and a constant-rate equation for a shoreline under going isostatic adjustment were used to model vertical movement for each si te. Results show that for at least the last 4,000 calendar years (cal BP) o f record, vertical movement in Lake Michigan has obeyed both types of equat ions. The two models yield similar results because rates of vertical moveme nt of the shorelines around Lake Michigan are small and the time frame for which lake-level data are available is so short that the exponential nature of isostatic change is not expressed. Except for the southern shore of Lak e Michigan, all the study sires have experienced uniform isostatic uplift c onsistent with trends reported by the Coordinating Committee on Great Lakes Basic Hydraulic and Hydrologic Data (1977) and Tushingham (1992). The sout hern shore of Lake Michigan, however, experienced a change in uplift rate r elative To the port Huron outlet about 1,400 cal BP. The residuals between the calculated rates of vertical movement at each sir e and its corresponding RLLC are a record of water-level change experienced at each sire. Within the resolution of the technique used to construct the RLLCs, all the residual curves should be, and are, similar. A Fourier smoo thing of the combined residual curves yields a "eustatic" fake-level curve for Lake Michigan over the past 4,700 cal BP. The results of the Fourier sm oothing identify major lake-level fluctuations such as the Nipissing II and Algoma phases of ancestral Lake Michigan. The technique also resolves lowe r magnitude and shorter duration quasi-periodic lake-level fluctuations of about 160 years (120 to 200 years).