Canadian Forest Service Publications
Updating Canada’s National Forest Inventory with multiple imputations of missing contemporary data. 2017. Magnussen, S., Stinson, G., Boudewyn, P. Forestry Chronicle 93(03): 241-245.
Year: 2017
Issued by: Pacific Forestry Centre
Catalog ID: 38949
Language: English / French
Availability: PDF (request by e-mail)
Available from the Journal's Web site. †
DOI: 10.5558/tfc2017-030
† This site may require a fee
Abstract
Canada’s National Forest Inventory (NFI) is facing an issue of spatial imbalance in photo interpreted data from 400 ha photo-plots available for estimation of state and change. Multiple imputations (MI) of missing data is therefore considered as a means to mitigate a potential bias arising from spatial imbalance, and—to a lesser degree— improve the precision relative to what can be achieved with the subset of plots having current data. In this study we explored MI with data from three study sites located in the provinces of Quebec, Ontario, and Saskatchewan. Specifically, we looked at state at time T2 and change between T1 and T2 in cover-type area proportions and in per unit area stem volume. At each location we found significant T1 differences in these attributes between plots with and without T2 data. A MI procedure with 20 replications of stochastic model-based imputations of missing data was therefore effective as a way to mitigate a bias that would arise if T2 inference was based exclusively on plots with T2 data. Possible differences between the T2 and T1 photo interpretation, paired with no efficient stratification of disturbed and undisturbed plots, largely eliminated expected gains in precision from the MI boosting of the effective T2 sample size. Despite recognized limitations, we recommend MI as an effective tool to counteract an emerging spatial imbalance in the NFI.
Plain Language Summary
Canada’s National Forest Inventory (NFI) was designed with 400 ha sampling units located on a nominal grid of 20 x 20 km. Data would be collected from a photo-interpretation. The original intensions were a re-measurement cycle of 10 years. With half the units measured every five year in a spatially balance design with two interpenetrating panels. For various reasons the actual scheduling of interpretation of sample units have resulted in a spatial imbalance in units available with concurrent data. Assessment of state and change derived from spatially imbalanced units can lead to bias and poor precision. To mitigate the bias problem and possibly achieve an improvement in precision this study sets out to evaluate the use of multiple imputations (MI) in Canada’s NFI. In MI a suite of models are estimated and used to stochastically update attribute values for units without concurrent data. Stochastic imputation preserves the uncertainty in an imputation and therefore requires replications. We evaluated MI with data from three study sites in Quebec, Ontario, and Saskatchewan. Our recommendation is to use MI to counter an anticipated spatial imbalance in NFI data. Expectations of an improved precision were not met.
