“Total-Balancing†an inventory: A method for unbiased inventories using highly biased non-sample data at variable scales

Kim Iles


The described here method can provide unbiased estimates and sampling errors with increasingly precise polygon information from non-sample sources that are often free and readily available. It is extremely flexible, and it appears to be in line with the main trend of modern sampling – you first estimate using any information available, and then you sample to adjust those estimates. At later dates, further readjustment can be done at will, as long as the total is maintained. The situation with forest inventory is very similar to mapmaking. For many years the only acceptable method of improving a map was to start over with a fresh sheet of paper and do the entire job again with great fidelity to current map accuracy standards. Those days are over. The same is true of forest inventory. A better concept is “let’s just change the parts that are not good enough”. The other parts change so little as not to be noticed. Stratified cruises with standard descriptions for multiple stands and repeating the process every 20 years while ignoring the existing inventory are old and outdated processes. In an age where information pours down upon us from every direction, it is time we started to use it effectively. It is so easy to insure statistical unbiasedness, good polygon estimates, and valid sampling errors by the progress described here that it is hard to imagine why anyone would strike the old forest inventory off the records and independently do it again from a standing start. MCFNS-1:10-13.


Inventory, Unbiased, Mean, Estimators, constrained estimator,

Full Text:



  • There are currently no refbacks.


© 2008 Mathematical and Computational Forestry & Natural-Resource Sciences