Change detection of green spaces in Putrajaya using remote sensing technique

• Main Author: Azhar, Muhammad Nabil Asyraf
• Format: Project Paper Report
• Language: English
• Published: 2019
• Online Access: http://psasir.upm.edu.my/id/eprint/84664/1/FH%202019%2017%20IR.pdf; http://psasir.upm.edu.my/id/eprint/84664/
• Institution: Universiti Putra Malaysia

Nowadays, rapid urban expansion has changed the forms of land in most part of the world. In Malaysia, after independence in 1957, the economics and population has dramatically increased. Thus, many forestland have been converted into industrial area to support the economical demand. While in Putrajaya, the administrative capital of Malaysia were also have been affected by the country’s rapid growth since the last 17 years when it was first establishment in 19 October 1995 and declared by the Prime Minister Tun Dr Mahathir Mohamad as the Federal Territory in 1st February 2001. In pursuing Malaysian Vision of Wawasan 2020, Putrajaya was planned to be a green city and intelligent city. Hence, there is a need of mapping tool to monitor and acquire information about the green spaces changes. In this study, remote sensing technique was applied for detecting the land cover changes over a time span of 17 years. The objective of this study was to obtain information about the current status of green spaces in Putrajaya by using remote sensing (Landsat 4 and Sentinel 2) data, secondly to compare the changes of green spaces as well as the land cover at two different times (Year 2000 and 2017) and finally to produce a map of land cover types in Putrajaya. The method include making subsets from raw data image, removing noises by using Masking tool, extracting features with supervised classification tool, analyzing NDVI method, image analysis supported by accuracy assessment and ground truth activity. Supervised classification extracted four features which were Vegetation1 (grassland/shrubs), Vegetation2 (forest), Water body and URS which stands for urban rock and soil. Besides, observing NDVI (Normalized difference vegetation index) from unsupervised classification has four features which are from no, low, medium and high vegetation. All the area for each feature is determined for data in all images. Results from this study shows that for the study area of 4638 ha, the land cover changes from year 2000 to 2017 are: (Vegetation 1 +205 ha, Vegetation 2 -715 ha, Water body +153 ha and URS +357 ha). The study shows the effectiveness, accuracy of remote sensing technique in monitoring the land cover change.