| Summary: | Landsat TM and ETM+imagery can be used to detect of mangrove forest condition.
This research located in mangrove forest area of Banyuasin Regency, South Sumatera. The
aims of this research are: (1) to know mapping accuracy of land use and density of mangrove
forest. (2) to know damage level and distribution area of mangrove forest using Landsat
TM/ETM imagery.
Landuse classification using supervised classification method.. Mangrove forest
density using Normalized Difference Vegetation Index (NDVI). Mapping of mangrove forest
damage level using qualitative dan quantitative methods. Qualitatif method was done by
overlaying landcover and canopy density, while quantitative method was done by giving the
score and the weighting of each criterion (land cover, canopy density and soil resistence to
abrasion). The result of land use classification and mangrove forest mapping was evaluated
by obtaining a classification error matrix to assess its accuracy.
The results showed that the accuracy of landuse mapping of 94.16%, mangrove density
of 89,68%. Mapping accuracy of mangrove forest damage using qualitative method of 90%,
and 85% by using quantitative method. Based on result of qualitative method, mangrove
forest is good of 112,228.699 ha, rather damage of 26,651.060 ha, and corrupted of
42,232.152 ha (year 2003), while in the year 2007 obtained good mangrove forest of
105,878.643 ha, rather damage of 27,243.544 ha, and damage of 48,087.382 ha. Results of
quantitative method, mangrove forest is good of 127,285.879 ha, rather damage of
17,978.649 ha,and damage of 35,847.383 ha (year 2003). In 2007,mangrove forest is good of
124,138.950 ha, rather damage 18,433.115 ha, and damage of 38,637.404 ha. Mangrove
forest declined from initial estimate 145,771.479 hectares in 2003 to 140,461.735 hectares in
2007 which represent a decrease of 5,309.68 ha in a 4-yrs period. During 2003-2007,
mangrove forest of Banyuasin Regency changing its form to plantations, settlements, open
land, rice field, swamp, shrimp pond, and shrubs,
|
|---|