| Summary: | Mangroves in Phang Nga Bay, Thailand and in Matang Mangrove Reserve, Malaysia serve a variety of crucial ecosystem services. However, they are threatened by various natural and human-influenced factors such as tsunami damage and development in recent decades. This project provides a look at how distribution and status of mangrove forests have changed over time and how mangrove health changes over time. Selected Landsat 5 TM images from 2000 to 2010 were analyzed to classify the land use changes by object-oriented method using feature extraction and by supervised classification. The expansion in urban development and agriculture is concerning for both Thailand and Malaysia according to the literature review (Gopal and Chauhan 2006; Giri et al. 2008). The Phang Nga Bay mangroves experienced significant 6.3% decline from 2003 to 2010 according to the supervised classification with tasseled-cap transformation. The Matang mangroves experienced a 3.95% decline from 2000 to 2010 according to the supervised classification. Although these mangroves are declining at a slower rate than the reported national and global average, the rate of decrease is still concerning compare to other Southeast Asian mangroves. We also examined the overall characteristics such as EVI, NDVI, GPP, and NDWI using Google Earth Engine to compare the overall patterns in the two study areas. There is no significant difference in EVI between the two study areas. The EVI value is 0.54 for the site in Thailand and 0.52 for the site in Malaysia. NDVI is higher for mangroves in the Thai site (0.61) than the Malaysian site (0.42). Mangroves at the Malaysian site has higher GPP and NDWI. The mean GPP for the site in Malaysia is 354 kg*C/m^2, while the mean GPP is only 217 kg*C/m^2 for the site in Thailand. The trend in GPP can be fit into an ARIMA(1, 0, 1)*(1, 0, 0)46 model for the Thai site and an ARIMA(2, 0, 1)*(1, 0, 0)46 model for the Malaysia site. The NDWI values are 0.149 and 0.137 for the Malaysian site and the Thai site correspondingly. The derived indices (tasseled cap, NDVI, and SAVI) were used to classify the mangrove areas into subclasses. An EO-1 Hyperion imagery from 2014 was examined to classify mangrove types in the Thai study area. We were able to classify mangroves into edge, island, riverine, estuary, and inland types based on the good spectral bands. A spectral library for the region or field data is necessary for more exact species classification. In terms of management, the local conservation departments and national park services in Thailand need to reach out more frequently to the local community and educate the fishermen and hoteliers about the ecosystem services of mangroves. It can be worthwhile for Matang forest managers to test the mixed block method with managed and natural mangrove patches to sustain biodiversity and ecological function of mangrove forests.
|
|---|