CONTENTS 



Page 



Overall appraisal 649 



Possible errors are of three kinds __ 649 



All data are not equally reliable 650 



Forest land, timber volume, and growth 650 



Forest Survey provided data for 33 States 650 



Special surveys made in 15 States and Coastal 



Alaska 651 



Adequacy of data depends chiefly on intensity of 



sampling 651 



Nonsampling errors cannot be measured 654 



Comparisons with 1945 estimates 655 



Estimates for Interior Alaska are only indicative. 655 



Timber utilization 655 



Data provided mainly by the Census 656 



Saw-log data have major effect on sampling 



accuracy 656 



Nonsampling errors are possible 657 



Plant residues data from special study 658 



Forest protection 658 



Fire damage data most adequate -. 658 



648 



Page 

 Forest protection — Continued 



Growth loss calculated in several ways 658 



Reliabilit}' of estimates rests on extensive body 



of data 659 



Productivity 660 



A special study of productivity was made 660 



Sampling procedure affects reliability 660 



Nonsampling errors may have occurred 664 



Forest land and timber ownership 664 



Ownership data tied to Forest Survey 664 



Sampling errors apply to part of ownership data. 664 



Some ownerships difficult to classify 665 



Forest tree planting 665 



Timber resources of North America and the world.. 665 



Future demand and supply 666 



Future demand closely related to future economic 



expansion 666 



Future supply data involve additional assump- 

 tions 666 



4 



