3 
area of 32,000 hectares. These aerial photographs were used not only for area 
and volume compilation but also for mapping purpose. 
The photographs were examined by the photo interpreter under stereoscopes, 
and delineated by forest type and stand-size class as well as nonforest land use 
type and problem area class. The information was transferred to the overlay 
with radial line plotters and KEK plotter, depending on the degree of tilt of the 
photographs. 
About 300 field sample plots were randomly selected on the overlay by forest 
stand-size class for ground measurement. It gave a sampling error of approxi- 
mately 6 7%. 
The survey started in September 1955 and will be completed in July, 1956. 
d) Land use and Forest Resource Survey of Taiwan. 
The Land Use and Forest Resource Survey is a Chinese and American coope- 
ative project sponsored by the Joint Commission on Rural Reconstruction with 
technical assistance from U.S. Forest Service. Technical personnel were furnish- 
ed by the Forest Administration, the Forest Research Institute, the Agricultural 
Research Institute and the National Taiwan University. 
The objectives of the survey are to yield (1) reliable forested land area and 
timber volume, including current growth, mortality and other data needed to 
formulate the forest policy so as to assure a proper management of forest re- 
source of Taiwan toward a continuous harvest of indigenous forest products, 
(2) present status of land use to facilitate establishing land use policy, so as to 
assure proper land use, eliminating soil loss and promote agricultural produc- 
tion and power development. 
In the survey, the aerial photogrammetric technique with double sampling 
procedure was applied. Twenty-four strips of aerial photography was flown in 
an east-west direction at a right angle to the general topography. The photo- 
graphy was done with a 24-inch focal length, minus blue filter and infra-red 
film on scales of 1/10,000 to 1/5,000. 
Photo sample plots were randomly selected on the sample strips, totalling 
37,495. They are first classified by forested and nonforested land and then 
stratified by volume class and conservation problem area class for the two major 
land use types. A total of 545 ground plots were chosen at random from the 
photo plots in each stratum. Photo plot was basic sample to obtain forested and 
nonforested areas and timber volume; and ground plot was used to control the 
photo plots, as well as to obtain ground information. 
Each photo plot was carefully studied by the interpreter in the third di- 
mension. Forest type was identified, tree height was measured with a parallax 
wedge, crown density, with a density scale; crown diameter, with a micrometer. 
And the per hectare volume was thus determined. Nonforested plots were classi- 
fied by land use type; all data observed and measured on the photographs was 
recorded such as slope, drainage pattern, type and intensity of erosion, soil depth 
and texture etc. In addition, all plots were classified by conservation problem 
area class based upon the above information. 
The sampling errors of the total forested land area and the total timber 
volume in Taiwan were + 1.59, and + 1.6 9, respectively. 
The land use and forest type and stand-size map at scale of 1/50,000 was 
compiled from the aerial photographs of 1/40,000—1/50,000. In making the