2.3. Image analysis
The satellite sensors used for this study, Thematic Mapper (TM) on Landsat 5 and Enhanced Thematic Mapper (ETM+) on Landsat 7, have a spatial resolution of 30 m × 30 m for six identical spectral bands, including the for water penetration important blue band. TM data are available since 1982 and ETM+ data since 1999 creating the possibility to map changes in the coastal environment over a relatively long period. The width of ground area covered by the different sensors (swath width) in one overpass is 185 km, and for each of these satellites a coverage cycle is achieved in 16 days.
The digital spectral values from the September 2001 Landsat ETM+ image were compared to the vegetation coverage assessed by the field survey that month. This analysis demonstrated a clear relationship, and in order to further examine the sensitivity of the satellite data spectral response in relation to vegetation coverage on the bottom of Chwaka Bay, an unsupervised classification (Mather, 2004) was conducted using the visible bands of the satellite image from September 2001. By forcing the classification to produce 10 classes and assuming that these classes represented the vegetation cover (0–10%, 10–20%, etc.) it was possible to create a remote sensing based estimate of the mean vegetation coverage to compare to the results of the September 2001 field survey.
2.3. Image analysis
The satellite sensors used for this study, Thematic Mapper (TM) on Landsat 5 and Enhanced Thematic Mapper (ETM+) on Landsat 7, have a spatial resolution of 30 m × 30 m for six identical spectral bands, including the for water penetration important blue band. TM data are available since 1982 and ETM+ data since 1999 creating the possibility to map changes in the coastal environment over a relatively long period. The width of ground area covered by the different sensors (swath width) in one overpass is 185 km, and for each of these satellites a coverage cycle is achieved in 16 days.
The digital spectral values from the September 2001 Landsat ETM+ image were compared to the vegetation coverage assessed by the field survey that month. This analysis demonstrated a clear relationship, and in order to further examine the sensitivity of the satellite data spectral response in relation to vegetation coverage on the bottom of Chwaka Bay, an unsupervised classification (Mather, 2004) was conducted using the visible bands of the satellite image from September 2001. By forcing the classification to produce 10 classes and assuming that these classes represented the vegetation cover (0–10%, 10–20%, etc.) it was possible to create a remote sensing based estimate of the mean vegetation coverage to compare to the results of the September 2001 field survey.
การแปล กรุณารอสักครู่..
