The effects of temporal differences between map and ground data on map-assisted estimates of forest area and biomass

When areas of interest experience little change, remote sensing-based maps whose dates deviate from ground data can still substantially enhance precision. However, when change is substantial, deviations in dates reduce the utility of such maps for this purpose. Remote sensing-based maps are well-established as means of increasing the precision of estimates of forest inventory parameters. The general practice is to use maps whose dates correspond closely to the dates of ground data. However, as national forest inventories move to continuous inventories, deviations between map and ground data dates increase. The aim was to assess the degree to which remote sensing-based maps can be used to increase the precision of estimates despite differences between map and ground data dates. For study areas in the USA and Norway, maps were constructed for each of two dates, and model-assisted regression estimators were used to estimate inventory parameters using ground data whose dates differed by as much as 11 years from the map dates. For the Minnesota study area that had little change, 7-year differences in dates had little effect on the precision of estimates of proportion forest area. For the Norwegian study area that experienced considerable change, 11-year differences in dates had a detrimental effect on the precision of estimates of mean biomass per unit area. The effects of differences in map and ground data dates were less important than temporal change in the study area.