= 1738 — 
Summary of Application to: 
Obtaining Supplemental Input for the Intensive Management 
of Pine Forests in North Carolina 
Darrel L. Williams 
Background and Objective of the Test: Commercial forest companies in the United States are 
faced with the challenge of producing more wood, over shorter periods of time, on a dwin- 
dling land base due to rapidly increasing demands for wood and wood fiber products. They 
are investing a considerable amount of manpower, time and money in research to develop 
improved forest management practices to meet this challenge. This research has led to the 
adaptation and refinement of intensive management practices such as harvesting natural for- 
est stands and replanting prepared sites with genetically superior seedlings. The result has 
been that the “seedling to harvest” cycle in the Southern Pine Region of the U.S. has been 
reduced from approximately 45 years, to 20 or 25 years. Accurate monitoring of the growth 
status of these artificially regenerated stands is of prime importance to these companies due 
to the increased initial investments of intensive management. Conventional monitoring tech- 
niques, such as ground observations and low altitude aerial photography, can be costly and 
the results variable because of the time required to complete data collection over large areas. 
It was hypothesized that remote sensing techniques using imagery from Landsat-type satel- 
lites would provide an economical means of obtaining supplemental input to forest inventory 
systems. The Information Transfer Laboratory at NASA/GSFC participated in a cooperative 
project with Weyerhaeuser Company's North Carolina Region to investigate the applicability 
of Landsat digital analysis for this purpose. 
Study Site: A 24,300 hectare forest tract in North Carolina was chosen as the study site. 
This site falls within the Southern Pine Region, which is essentially co-extensive with the 
Atlantic Coastal Plain physiographic province: The most prominent vegetational feature of 
much of this region is evergreen trees growing on the rather low-lying, poorly drained soils. 
Intensive forest management practiced in the study area for several years has resulted in a 
full spectrum of forest cover conditions, such as recent clearcuts, various stages of growth 
following artificial regeneration of pine, and natural stands of both pine and hardwood. Ap- 
proximately 305 kilometers of logging access roads dissect the site and this extensive road 
system was helpful for quickly and accurately identifying any given forest compartment 
within the tract (Fig. 4.1). Black and white, color, and color IR aerial photography were also 
available, as well as detailed maps and records showing stand age and species composition by 
compartment. 
Approach and Results: Winter (26 February 1974) and summer (30 August 1973) Landsat- 
1 imagery of the study site, designated as path 15 and row 35, were analyzed individually 
and then geometrically registered and overlaid in order to take advantage of temporal changes 
in the forest canopy. Prior to the detailed digital classification of the imagery, a contrast 
stretched color IR composite image of the study area was generated using the 26 February 
1974 Landsat data. Visual photointerpretation and delineation of stand conditions was gen- 
erally found to be easier using this color composite image than interpreting a black and white 
aerial photograph taken approximately 20 days after the Landsat overpass (Fig. 4.1). There 
is also a high degree of similarity in the spatial information contained in both types of imagery 
at a common scale. 
    
  
   
   
    
  
  
   
    
    
   
   
  
     
   
    
  
  
   
   
  
  
    
  
  
    
   
  
  
  
  
   
     
Compt 
selecti: 
then cl 
band v 
wood 
deline: 
canop 
situati 
this ty 
sure a 
growt 
In the 
sulted 
tribut 
simul! 
of cro 
two L 
tral si 
Febru 
The a 
samp! 
forest 
IR st: 
96% : 
an ov 
result 
Fore 
the hi 
Thes 
mon: 
mati: 
Furt 
Ri 
I tri