Summary of Application to: 
Shifting Cultivation in the Forests of Northern Thailand 
Lee D. Miller Kaew Nualchawee Craig Tom 
Background and Objectives of the Test: The area of the Golden Triangle at the corner inter- 
section of China, Burma, Laos and Thailand has traditionally been a rugged, inaccessible 
terrain covered with tropical forest. Shifting cultivation, swidden or cut and burn agriculture, 
as it is variously called, has always been a permanent feature of this rugged forest landscape. 
Subsistence agriculture with some small amount of intermixed cash crops consisting primarily 
of opium and dry land rice have been historically grown on small recently cleared forest sites. 
Two or three years of cultivation usually exhausts the residual soil nutrients and the field 
locations are shifted by clearing new forest plots. The small, abandoned fields regrow into 
forest and in twenty years can not readily be discerned from old growth stands except for the 
size of tree stems. Historically, shifting cultivation did not occur in all forest types. The 
earlier practitioners preferred certain types of forests, as they indicated the presence of more 
desirable soil types. Thus, certain types of forests were only minimally impacted by the pro- 
cess. Originally these selected old growth forests and those allowed to regrow sufficiently 
between shifting cultivation cycles contained valuable quantities of teak. 
Continually increasing international demand for illicit opium, coupled with the increasing 
pressure of population growth in this area, has substantially decreased the cycle time of shift- 
ing cultivation while simultaneously increasing the area under cultivation at any given time. 
The process is no longer in an ecological balance over time. All forest types are being cut and 
burned on cycles of six or seven years instead of the past process which selectively used more 
resilient forest types on cycles of greater than 20 years. Suddenly all of the forests are dis- 
appearing and the abundant tropical rainfall and unavoidable erosion of these large non- 
forested areas is making the process irreversible. Large scale examples of the results of this 
disastrous process already exist in the extensive desert areas which have been created in the 
northeastern portion of the Amazon Basin. 
The research reported here has been underway for approximately seven years to determine 
the application of remote sensing to the measurement, monitoring, and projection of spatial 
patterns of shifting cultivation in this region. The specific objectives of the research discussed 
here concerns: (a) tests of the various methods for computer classification of the forested 
areas subjected to shifting cultivation and (b) the use of these results and related information 
collected from airphotos and existing maps to model the occurrence and spread of shifting 
cultivation so that its causes and control may be more clearly understood. 
Fortunately, a persistent dry season occurs in the area during January, February and March 
and excellent Landsat imagery and airphotos have been collected during these months. The 
work reported upon here has not been directly spohsored, although various peripheral aspects 
of it have been underwritten by Colorado State University, the Mekong Secretariat of the 
UN/ESCAP, Kasetsart University, and NASA/GSFC. The interpretation of the causes and 
implications of the shifting cultivation process noted above is purely that of the authors. 
Study Site: The particular study site selected for detailed analysis is located about 45 kilom- 
eters northwest of Chiang Mai, the major city in the northern region of Thailand, and lies just 
     
      
   
  
  
  
  
  
  
  
  
  
   
   
  
  
  
  
  
  
  
   
   
  
  
  
  
   
   
  
  
  
   
   
     
sout 
lies 
anni 
cific 
(24 
rect 
earli 
cells 
ages 
Apt 
may 
erin 
thes 
lanc 
plet 
and 
plat 
pla 
ing 
wel 
and 
con 
par 
pre 
cot 
fou 
tog 
Th 
ill 
ha 
«T.