(1b) Legend / Categorizations 
Chairman: Ryutaro Tateishi 
Chiba University, Japan 
The discussion session on Legend / Categorizations was held from 11:00 AM to 12:30 PM on May 
31, 1995. The discussion was led and reported by the Chair, Dr. Ryutaro Tateishi of Chiba 
University, Japan. 
The group focused on land cover legends. 
One problem of land cover legends is that there are many kinds of legends in different countries and 
different projects. This fact prevents data exchange and integration. The first topic was standard land 
cover legend. Is the standard land cover legend possible? The answer was No because users 
community is diverse and there are a variety of needs for land cover. However it was suggested to 
get several standards for each of the users community. The important thing here is that a legend must 
reflect needs of users. It must be also noted that needs are not static. They will evolve over time. 
One approach to establish one of several standards is being done by IGBP Land Cover Working 
Group. The users of land cover data by IGBP are global change scientists such as climate change 
and carbon circulation. For the purpose of these studies, discrimination of tree, shrub and grassland 
is important. The problem is how to draw the boundary between them. Another suggested approach 
is to establish legend from the view point of vegetation structure, vegetation seasonal trends and 
climate, and vegetation species.(see Global Ecosystems Framework by J.S. Olson in this book) 
Hierarchical structure of categories is recommended because it provides greater versatility and 
longevity. 
It is important to remove bias in categories. Categories in areas which are not populated tends to be 
more rough than those of near populated area. Similarly, Since most existing land cover legends for 
global area have been developed by Western Hemisphere scientists, they have less detailed 
distinctions in Eastern region. This is a cultural bias. Therefore Asian input to the development of 
legends is recommended. 
Another problem in current land cover legend is spatial heterogeneity. Fuzzy logic may have a use in 
classification to cope with this problem. 
In conclusion, the development of several standard land cover legends for different users community 
was suggested. Key points in this development are users' needs, hierarchical structure, and removal 
of biases. 
do 
lab 
In 
SO 
do 
res 
la 
rec 
da 
dii 
gn 
an 
ari