Canadian Forest Service Publications

A Markov chain method for simulating bulk density profiles in boreal peatlands. 2014. Thompson, D.K.; Waddington, J.M. Geoderma 232-234(2014):123-129.

Year: 2014

Issued by: Northern Forestry Centre

Catalog ID: 35529

Language: English

Availability: PDF (request by e-mail)

Available from the Journal's Web site.
DOI: 10.1016/j.geoderma.2014.04.032

† This site may require a fee

Mark record


Bulk density is a key determinant of numerous physical characteristics in peat including hydraulic conductivity, smouldering combustion vulnerability, and water retention in the unsaturated zone. A Markov chain model based on peat type (primarily Sphagnum, sedge, and sylvic peats) was applied to the depth-wise structure of boreal peatlands using 143 cores from western Canada as source data. Bulk density and peat type were modelled in 2160 simulated peat profiles by driving Markov chains associated with bulk density distributions by peat type and depth. The model closely reproduced the expected change in bulk density between vertically adjacent peat horizons. Markov-derived peat profiles showed somewhat greater variance in organic matter load in the upper 135 cm compared to observed cores due to the lack of whole-profile bias or trends in density. The method and derived Markov chains shown here have utility in hydrological modelling, regional carbon estimates, and in the modelling of disturbances such as wildfire.

Plain Language Summary

This article presents a new mathematical modelling approach for predicting the density and other properties of peat in western Canada. The model works at a scale of a square metre in area, and accurately predicts the type of peat (decomposed vegetation from different types of plants) and density of the peat for the top 135 cm from the surface. Instead of assuming that all the peat is of equal density (as is the case in some models) we are instead able to accurately reproduce real world trends in density where more dense peat at the bottom transitions to lower-density peat at the surface. This model has broad applications in wildfire carbon emissions accounting, hydrological modelling, and estimating how much carbon is stored in Canada's boreal peatlands.