185 
:o various 
> of 
:ale and 
:er has 
>rie s. 
onventional 
;. On 
nail 
lods of 
was 
volume 
1:5000 
:ions were 
.ed on 
ion of 
the bulk 
irregular 
;ion is 
of trees 
ill scale 
trees. 
sful upto 
this 
5 S On 
tree aerial 
jmpt has 
ability of 
) used 
s and 
3 teak 
) height 
and ground 
and volume 
verted 
nd best 
le S.E. 
level 
eparation 
. volume 
ttempted 
using photo height and photo crown closure as independent variables for 
teak trees against mean ground height and plot volume as dependent 
variables . Various combinations with mean photo height and mean ground 
height, mean ground dbh, plot volume and so also mean crown closure 
and mean crown diameter were tried. Best results were obtained with 
mean photo height and mean crown closure with plot volume. The S.E. 
of the regression line varied from 5% to 3 0%. JOSHI ( 5 ) used photo 
height and photo crown closure as independent variables for dry decidu 
ous forests on various scales. The S.E. of the estimate was 3 0%. 
It was noticed that the inclusion of parameter photo crown closure did 
not significantly reduce the error of the regression line as almost all 
the variation was accounted for by using photo tree height as the indepen 
dent variable. No significant difference with different scales was 
noticed. 
"Area Determination" 
In resource inventories the usual practice has been to transfer 
the delineated forest types on maps and calculate the area of various 
types by dot grid of varying intensity. Significant gains in precision, 
time and cost have been reported by this method TOMAR ( 12 ). This 
has necessarily reduced the error of scale variation inherent in aerial 
photographs but the error of misinterpretation and also of the maps ie 
transfer error etc. were not taken into account. TIWARI ( 7 ) carried 
out investigations on the use of photo plot sampling method for Sal 
forests of Dehra Dun. He accounted for the misinterpretation and scale 
variation errors and compared the adjusted proportions with those 
obtained by purely conventional methods. No significant difference was 
obtained and use of method and aerial photographs was advocated for 
working plan preparations. JOSHI ( 4 ) estimated the area proportions 
of forest types on scales 1:5,000 to 1:20,000 and indicated no signifi 
cant difference. 
"Direct Estimation of Growing Stock" 
Investigations were carried out for Sal forests on 1;20,000 
scale. Initial stratification into three volume classes of nearly 100 
cum per ha. range was done on the basis of photo height, crown 
closure, crown diameter and texture of the tree canopy. Volume in 0.1 
ha. circular plots was estimated on a number of photo plots on the 
basis of stereograms covering the entire range of volume. Mean volume 
in each stratum was adjusted by regression of photo volume and ground 
volume of double plots. The S.E. of estimate varied from 5% to 2 0%. 
The method gave encouraging results and better precision can be obtain 
ed with larger scale and recent photographs. 
RESOURCE INVENTORIES 
Existing conventional management plans were made for small 
units of forest areas and since they were directed towards scientific 
management of forests and fulfilment of local demands they were unable to 
supply data on wood resources for large industrial units. With a greater 
number of species becoming of economic importance, the need for regional in 
ventories which could provide sufficiently reliable data for large industrial