Proceedings Article | 26 July 2007
KEYWORDS: Geographic information systems, Analytical research, Shape analysis, Fractal analysis, Ecology, Remote sensing, Image resolution, Landsat, Classification systems, Earth observing sensors
In this paper Landsat images, including MSS in 1978, TM in 1988 and ETM in 2001, and ancillary data were selected to
investigate the landscape pattern change in Niyanghe watershed. Firstly, the images were classified respectively to get
the three landscape maps; then some landscape pattern analytical indices such as number of patch, shape index,
contagion, fractal dimension, Shannon index and evenness index were introduced to describe the temporal-spatial change
of landscape pattern in the study area assisted by spatial analytical techniques of GIS. The results showed that bare-land
was the main landscape type. During the periods of 1978-1988 and 1988-2001, the increased arable land area was the
largest, and the increase speed was the fastest compared with all other landscape types, the significant newly-increased
arable land derived from bare-land and grassland. The area of grassland increased rapidly. The area of bare-land
decreased gradually every year, whereas bare-land was mainly converted into arable land and grassland. Forest land area
has decreased recently because of being reclaimed into arable land. From 1978 to 2001, the extent of fragmentation,
evenness and diversity increased, contagion index decreased conversely, shape of landscape types became more and
more complicated. The main driving factors for landscape pattern change were human factors such as population growth,
consumption demands, economy development, transportation, policy, and so on.
