The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B6b. Beijing 2008 
of individuals, communities, and even whole ecosystems. The 
landscape ecology method is relatively new, and has gradually 
become an important tool for studying ecological security. It 
has shown considerable promise, at both whole nation and even 
global scales.All these methods were able to achieve a 
quantitative analysis of the ecological security status by using 
the statistical data, but lacked the information of pattern 
graphical. Thus it is essential to establish a regional system for 
assessment ecological security that can provide a quantitative 
and more effective assessment. 
Figure 1 .PSR concept model of OECD 
2. METHODS 
2.2 Establishment of the assessing model based GIS 
As ecological security assessment needs a large and real-time 
data support and it also needs to feed back its result to the 
policy maker on time, so it requires the support from the system 
that could provide information instantly and truly. RS&GIS 
technology meets this requirement. Geographic information 
system (GIS), as a useful tool to analyze and manage spatial 
information, has a superior advantage in ecological security 
assessment. The RS data, such as satellite images and aerial 
photo of different periods, can be utilized to realize dynamic 
Remote sensing inspection of land cover changes and acquires 
dynamic or mutative information to update and perfect the data 
correlative with environment management. And the huge 
amount of multi-temporal and multi-sensor data of earth’s 
surface can be acquired and used to analyze in order to achieve 
reliable, precise and satisfying results. 
2.1 Indicators system 
This paper took the ‘pressure-state-response’ model (PSR) that 
has been proposed by OECD and the United Nations UNEP 
(Qian et al., 2001) as the foundation of an indicator system for 
assessing ecological security (Figure l).The PSR approach is a 
causal one that covers causes and effects influencing a 
measurable state. In this sense, three categories of indicators are 
distinguished (OECD, 1993): 
• Indicators of environmental Pressure describe pressures 
on the environment originating from human activities, 
including quality and quantity of natural resources (e.g., 
emissions, mining of raw materials, fertilizer input). 
• Indicators of environmental conditions (State) are 
designed to describe the status quo of the environment 
and the quality and quantity of resources and their 
changes over time (e.g., forest area, protected areas). 
• Indicators of societal Response show to which degree 
society is responding to environmental changes and 
concerns. This could be the number and kind of measures 
taken, the efforts of implementing or the effectiveness of 
those measures. Responses may range from public (e.g., 
legislation, taxation, promotion) to private sector 
activities (e.g., reduced consumption, recycling) (Linser, 
2001). 
The PSR model has proven to be a logical, comprehensive tool 
to picture environmental issues from an anthropocentric 
perspective. Instead of observing a single phenomenon or 
problem a causal model of causes, impacts and effects on the 
environment is generated. This approach is rather powerful in 
communicative and opinion-building processes (Wolfslehner, 
2007). 
Based on the indicators combined with the weight calculated 
using the AHP model, we developed the following formula for 
the ecological security index (ESI): 
ESI = X fV i * s i 
Where Sj represents the results for indcator i, and Wj represents 
the weight of indicator i. 
2.3 Remote sense information extraction 
Environment monitoring is an important aspect of remote 
sensing application. The use of remote sensing as a data source 
for environmental management is increasing. In middle-scale 
area, the rapid and macro monitoring of vegetation cover in 
mountainous area of Xishuangbanna is based on remote sensing 
responsive spectral characteristic analysis. This paper extract 
Normalization Difference Vegetation Index selected (NDVI) to 
identify the vegetation cover states. NDVI is the most extensive 
application of a vegetation index. NDVI is linear related with 
distribution of vegetation density and the state of vegetation 
growth and spatial distribution of vegetation in the best 
direction factor. NDVI calculate formula: 
NDVI = 
NIR-R 
NIR + R 
where , NIR is Near-Infrared spectral band and R is red 
spectral band. 
3. APPLICATION 
3.1 Case study area 
Xishuangbanna is located in southern Yunnan Province, in 
southwest China (24 10 to 22 40 N, 99 55 to 101 50 E) 
bordered by Laos in the south and southeast and Burma in the 
southwest. It is a famous region in China for its diverse flora, 
fauna and beautiful landscape. It is estimated that about 
5000species of higher plants (16% of those in China) exist in 
this area of 19,200 km 2 (0.2% of that in China), of which nearly 
1000 species of wild plants can be directly utilized by the 
people. This area has one of the greatest diversities of species in 
China and is of great importance in the maintenance of regional 
biodiversity (Cao and Zhang 1997; Shanmughavel et al. 2001). 
In the early 1950s, about 60% of the total area of 
Xishuangbanna was covered by forest (Xu 1985). Recent 
investigation has revealed that this has declined greatly due to 
the rapid expansion of the local population and irrational