SIMULATING PEATLAND SURFACE MOVEMENT MEASURED USING GPS IN NORTHERN MINNESOTA

Nine GPS and 3 hydro-meteorologic monitoring stations were installed in the Red Lake Peatland of northern Minnesota. GPS antennas mounted to small trees were used to monitor the movement of the surfical peat, revealing seasonal changes in surface elevation synchronous with hydraulic head data. In 2009, surface elevation changes during the spring freshet were 20 and 4 cm at a fen and bog site, respectively. Following the spring rebound, peat surface at the fen dropped about 6 cm over 4 months, while the bog surface position remained relatively stable. During a snow event in Dec. 2009, the bog peat surface dropped 4 cm, returning it to its starting (Jan.) elevation. Several hypotheses have been proposed by previous researchers to explain surface movement including: expanding and contracting biogenic gas pockets, dilation and compression due to the flux of water to/from peat pores, and buoyancy effects. To evaluate the importance of buoyancy to peatland surface movement, a simple 1-D dynamic model of a peat column has been constructed based on a viscoelastic (Kelvin) model. The peat column is conceptualized as ten blocks connected to each other and the underlying mineral sediment. The effective mass of each block is dependent on the peat water content, peat density, gas content, and the proportion of the block below the water table. Simulations of the fen (2.8 m thick peat column) were performed by varying the water table position, based on measured hydraulic head, over a one year period. Peat elasticity, viscosity, and gas content were varied in these simulations to assess the sensitivity of the models to these poorly constrained parameters. Varying the peat's elasticity from 0.5 to 1.5 GPa (gas content of 10%) resulted in annual surface fluctuations ranging from a few centimeters to about 14 cm, less than the observed variability but consistent with observed patterns. This model neglects the impact of snow loading and the effect of gas pressure within the peat, which would likely increase surface elevation variability and produce the rapid changes observed in our data sets. Our numerical modeling suggests that buoyancy effects contribute to observed seasonal changes in peat deformation and that changes in gas content within peat deposits strongly influence the magnitude of seasonal surface oscillations.