Because peat accumulates only beneath the water table, the shape of a peat body should reflect the shape of its water table and thus the hydrology of the peat body. Three different models successfully reproduce the observed peat dome morphology, including a central bog plain. In the first model, the bog plain develops because peat accumulation is limited by anaerobic decay of peat beneath the water table. With certain simplifying assumptions, an analytic solution for this model can be obtained. The other two models are more easily investigated numerically. In the first model, the initial peat accumulation rate is limited only by plant growth and decay and is the maximum rate observed during peat dome development. As a peat dome expands laterally, peat accumulation slows because the water table ceases to rise fast enough to preserve all the available plant material. Eventually, anaerobic decay beneath the water table matches the rate of peat addition to the top of the peat body, and net peat accumulation ceases.
In the second model, peat accumulation is assumed to begin simultaneously over a wide area. The central bog plain is a transient feature; the peat dome fails to reach a steady-state condition for long after peat begins to accumulate. Peat accumulation rates in the central bog plain are either constant or vary as a function of climate.
In the third model, the height of the water table is controlled primarily by ground-water inflow from an underlying aquifer. The maximum height the water table can reach is the head in the underlying aquifer. If the area over which ground-water influx occurs is sufficiently broad, and the hydraulic conductivity is sufficiently low, most of the peat dome can develop a water table that is nearly at its maximum possible height, thus creating the central bog plain. Peat accumulation rates in this model decrease with time.
Comparison of theoretical predictions with observed peat dome morphologies shows a good correspondence between the predicted and observed profiles of the peat domes for all three models. A comparison of apparent peat accumulation with the first model, assuming no anaerobic decay, shows a good correspondence for a Sumatran core. The third model probably could be calibrated to match this core as well. A core from a German Sphagnum peat bog provides clear evidence of anaerobic decay because of an increase in growth rate with time. By reducing the thickness of early deposited layers, anaerobic decay can result in an apparent increase in peat accumulation rate, even if the actual rate of addition of the peat to the surface declines with time.