Detecting deforestation with a spectral change detection approach using multitemporal Landsat data: A case study of Kinabalu Park, Sabah, Malaysia

Published Date
Journal of Environmental Management
September 2008, Vol.88(4):784–795, doi:10.1016/j.jenvman.2007.04.011

Author 

• Mui-How Phua ^a,b,,
• Satoshi Tsuyuki ^b
• Naoyuki Furuya ^c
• Jung Soo Lee ^d

• aSchool of International Tropical Forestry, Universiti Malaysia Sabah, Locked bag 2073, 88999 Kota Kinabalu, Sabah, Malaysia
• bGraduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan
• cForestry and Forest Products Research Institute, 1 Matsunosato, Tsukuba, Ibaraki 305-8687, Japan
• dDivision of Forest Management & Landscape Architecture, College of Forest and Environmental Sciences, Kangwon National University, Korea

Received 1 March 2006. Revised 20 March 2007. Accepted 13 April 2007. Available online 12 July 2007. 

Abstract

Tropical deforestation is occurring at an alarming rate, threatening the ecological integrity of protected areas. This makes it vital to regularly assess protected areas to confirm the efficacy of measures that protect that area from clearing. Satellite remote sensing offers a systematic and objective means for detecting and monitoring deforestation. This paper examines a spectral change approach to detect deforestation using pattern decomposition (PD) coefficients from multitemporal Landsat data. Our results show that the PD coefficients for soil and vegetation can be used to detect deforestation using change vector analysis (CVA). CVA analysis demonstrates that deforestation in the Kinabalu area, Sabah, Malaysia has significantly slowed from 1.2% in period 1 (1973 and 1991) to 0.1% in period 2 (1991 and 1996). A comparison of deforestation both inside and outside Kinabalu Park has highlighted the effectiveness of the park in protecting the tropical forest against clearing. However, the park is still facing pressure from the area immediately surrounding the park (the 1 km buffer zone) where the deforestation rate has remained unchanged.

Keywords

Deforestation detection

Pattern decomposition method

Spectral change detection

Landsat

Kinabalu Park

---

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

 Table 1

Table 1.

Fig. 6.

Fig. 7.

 Table 2

Table 2.

Fig. 8.

 Table 3

Table 3.

Fig. 9.

• ⁎ 
  Corresponding author. School of International Tropical Forestry, Universiti Malaysia Sabah, Locked bag 2073, 88999 Kota Kinabalu, Sabah, Malaysia. Tel.: +6088 320000x8773; fax: +6088 320876.

For further details log on website :
http://www.sciencedirect.com/science/article/pii/S0301479707001594