ALOS AVNIR-2 DIGITAL DATA ANALYSIS FOR TURBIDITY MAPPING IN SEMARANG COASTAL AREA, CENTRAL JAVA, INDONESIA

Abstract

This study aims to determine the spectral response of ALOS AVNIR-2 satellite’s image on the turbidity of the waters and devise the algorithms for turbidity mapping in the coastal waters of Semarang City, Indonesia. The Remote Sensing Technology Center of Japan, RESTEC supported the ALOS AVNIR data and conducted the technical training on Application and Verification Project to support this research. The method used in this study is digital image processing and field survey. Digital image processing method including histogram adjustment for radiometry correction, rectification for geometry correction, cropping area of interest, masking the area for the separation of the mainland, statistical analysis to determine the spectral response of the image of water turbidity (R 2 ≥ 0.9) and algorithm model evaluation by using the Mean Normalized Bias (MNB) as an indicator of systematic error and the Root Mean Square (RMS) as an indicator of random error (Ouillon et al., 2008). The results of statistical analysis showed that the red channel and green channel on ALOS AVNIR-2 has a relatively high spectral response of the turbidity level (R 2 ≥ 0.9). The best algorithm model for the single spectral channel obtained from the red band with the MNB value is 22.65% and the RMS value is 32.09%. The best algorithm model for the dual channel spectral bands obtained from a combination of red and blue bands with the MNB value is 19.89% and the RMS value is 28.22%. Distribution of water turbidity levels in the waters of Semarang city has a range between 0.03 to 9.44 NTU. Key words: ALOS AVNIR-2, Turbidity Mapping and Coastal Area