VIRGINIA FOREST RESOURCES AND 
INDUSTRIES 
Appendix 
oo 
Survey Methods 
Field Inventory 
The field inventory of the timber resources of Vir- 
ginia was made in 1940. To facilitate analysis and 
use of the data, the State was divided into five survey 
units, varying in size from 4.3 million acres to 6.4 
million acres, namely, the Coastal Plain; the northern 
Piedmont, the southern Piedmont (combined in this 
report) ; the northern mountains, and the southern 
mountains (combined in this report). 
In the field survey, crews gridironed the State with 
compass lines spaced 10 miles apart. At intervals of 
one-eighth mile along each 
sample plots were established. Records obtained on 
line, one-fourth-acre 
31,390 plots form the basis for computing the areas 
Of these, 18,- 
087 were forest plots, and here detailed measurements 
devoted to various kinds of land use. 
and observations were made concerning the number, 
size, and species of trees, the forest type and stand con- 
ditions, degree of fire damage, density and distribution 
Data for 
construction of volume tables were collected by sup- 
of reproduction, growth, and site quality. 
plementary sampling by J. W. Girard, of the Forest 
Service, who also determined timber-cull percentages. 
In estimating forest area there were two possible 
sources of error: (1) errors in classifying field plots or 
in compiling the data, and (2) sampling errors. The 
first arise from mistakes of judgment or technic and 
were minimized by the exercise of care and skill, even 
though it was seldom possible to evaluate them. In 
the Survey of Virginia, every effort was made to main- 
tain a high order of accuracy in the collection and 
In the field this took the form 
of frequent checks and a continuous program of train- 
compilation of data. 
ing. In the office the work was organized to permit 
automatic machine verification of the more important 
operations. 
The sampling method used did not permit compu- 
tation of sampling errors of the area estimates under 
correct statistical procedures. However, comparisons 
of forest area obtained by the Forest Survey plot 
method of sampling with the same forest area planim- 
etered from aerial photographs indicated that the plot 
method of sampling gave results within 2 percent of 
the planimetered area for forested areas as small as 1.5 
million acres. These empirical checks indicate that 
the Forest Survey estimate of the forest area of Vir- 
ginia should be rather precise and that estimates for 
the three major physiographic regions of the State were 
also reliable. 
In estimating timber volumes, the possible sources 
of error included (1) and (2) above and, in addition, 
(3) inaccurate measurements of tree diameter, height, 
or cull, and (4) bias resulting from impreper construc- 
tion, selection, or use of tree volume tables. As in the 
case of forest area determinations, every effort was 
made to obtain accurate measurements through fre- 
quent checks and training. The volume tables used 
also were checked and were found to give reasonably 
accurate results. It was not possible to compute 
sampling errors of the volume estimates, but on the 
basis of subsequent experience with random samples 
of roughly comparable intensity and the computation 
of sampling errors, it seems safe to assume that the 
error of the estimated saw-timber volume in the State 
does not exceed =2.5 percent. 
The reliability of one statistic as compared to an- 
other presented in the same or a related table can be 
judged roughly by its relative magnitude. In general, 
the larger values warrant greater confidence, while the 
smallest should be considered indicative rather than as 
absolute quantities. 
Increment 
Measurements for growth calculations were. ob- 
tained from increment borings made in a mechanically 
selected sample of all trees over 3 inches in diameter. 
In general, computational procedures consisted of de- 
termining the volume of small trees that grew to mer- 
chantable size during the period and of increasing the 
sample tree diameters by the measured diameter 
growth of the preceding decade. ‘The differences be- 
56 Miscellaneous Publication 681, U. S. Department of Agriculture 
