REPORT OF COMMISSION VII 
GVII-67 
a specific need and to pave the way for improving efficiency of photo interpreta 
tion. The particular objective in this paper is to discuss volume estimates of 
forests from aerial photos and methods of improving such estimates. 
The major purpose of timber cruising is to determine the total volume of 
wood on a given area. In such studies many characteristics of the forest are 
important, such as species and size of trees, condition and quality of timber, 
growth, destruction or cutting for wood products, area by size classes, forest 
types, and soil types. These items and perhaps many others are studied to 
gether with volume, but usually incidental to the study. Hence, the variability 
of wood volume and timber size are the critical characteristics of the forest 
upon which the design of a timber cruise is usually based. 
Foresters have found that aerial photos are an aid in estimating wood volume 
and stand size. Accordingly aerial photos have become a basic consideration 
in the design of timber cruises. Where a high degree of accuracy is not needed, 
the experienced photo interpreters can omit all or a large part of the field work 
usually required. One investigator, Moessner (6), has reported exceptional 
accuracy but many foresters consider such reports to be chance occurrences. 
Even where accuracy is very important the work done by the experienced 
photo interpreter reduces the amount of field work several times. The accuracy 
with which the interpretations can be made is not fully established, however, 
and is a subject of much speculation, conjectures, and discussion. This is not 
surprising in view of the many variables which can effect the accuracy. 
Methods of Estimating Wood Volume 
To obtain volume foresters use aerial photos in two ways. First, some photo 
interpreters estimate volume directly from the aerial photos; that is, some 
photo interpreters furnish specific volume estimates from photo examination 
without field study; at present these interpreters are far in the minority. Most 
photo interpreters rely upon combinations of both photo and ground work. A 
complete break-away from ground plot examination is not now generally 
acceptable, but it appears that research will develop techniques that will mark 
edly reduce the number of ground plots needed. 
Certain interpreters gain an advantage in aerial photo interpretation by 
stratified sampling designs. By this design a heterogeneous population of volume 
is divided into homogeneous volume classes or stratum, and a random sample 
of ground plots is drawn from each of these homogeneous volume classes. The 
photo interpreter classifies the forest area by these homogeneous volume classes. 
The random samples of ground plots in each homogeneous volume class are used 
for estimating the mean volume of the volume class. These methods withstand 
rigorous examination of the critics. 
Tests of Accuracy 
In the 1930’s there were no studies to show the accuracy of estimating vol 
ume from aerial photos. But during this period many foresters were getting 
interested. However it was not until after World War II that some progress 
was made. The standard deviation for homogeneous strata became a basis for 
judging the efficiency of the use of aerial photos, and is an excellent scheme 
when specific volumes are not estimated. When specific volumes are estimated, 
both on the photos and on the ground, then direct comparisons can be made be 
tween these estimates. This, of course, is a more precise test of the accuracy of 
photo estimates of volume than a comparison of the timber volume for a large 
area based on ground samples that are independent of photo techniques.