<![CDATA[Nowadays, Breaks for Additive Seasonal and Trend (BFAST) method based on time series of Moderate Resolution Imaging Spectroradiometer (MODIS) satellite data is increasingly used to monitor the temporal dynamics of vegetation changes. Nevertheless, sensitivity of the BFAST method for detecting the vegetation cover changes based on the choice of vegetation indices and land cover types has not been widely investigated. Breaks for Additive Seasonal and Trend (BFAST) method has applied to MODIS 16-day Enhance Vegetation Index (EVI) and Normalized Difference Vegetation Index (NDVI) composites images (2000-2014) of three land cover types (Urban and Built-Up, Evergreen Broadleaf Forest and Savannah) within Australia. Overall, the number and time of changes detected in the three land cover types differed with both time series data because of the data quality due to the cloud cover. As conclusion, the EVI is more sensitive than NDVI for detecting the seasonal and abrupt changes for the land cover which has the dense vegetation and large canopy background such as evergreen broadleaf forest. Furthermore, NDVI is more reliable to detect the seasonal and abrupt changes that occurred in land cover types which have sparse vegetation such as urban, built-up area and savannah.Keywords: Additive Model, BFAST, EVI, NDVI, MODISABSTRAKSaat ini, Metode Breaks for Additive Seasonal and Trend (BFAST) berdasarkan data satelit Moderate Resolution Imaging Spectroradiometer (MODIS) telah banyak diaplikasikan untuk melakukan monitoring terhadap perubahan dinamis dari tutupan vegetasi. Namun, sensitifitas BFAST untuk mendeteksi perubahan vegetasi berdasarkan pilihan indeks vegetasi dan jenis tutupan lahan yang berbeda belum banyak dilakukan. Metode Breaks for Additive Seasonal and Trend (BFAST) telah diaplikasikan dengan menggunakan data Enhanced Vegetation Index (EVI) dan Normalized Difference Vegetation Index (NDVI) dari satelit MODIS 16-harian terhadap tiga jenis tutupan lahan (perkotaan dan lahan terbangun, hutan berdaun lebar dan padang rumput) di wilayah Australia untuk periode data tahun 2000 - 2014. Secara umum, hasil deteksi metode BFAST berbeda untuk setiap tutupan lahan baik dari segi jumlah dan waktu yang dipengaruhi oleh kualitas data karena adanya tutupan awan di lokasi penelitian. Dapat disimpulkan bahwa EVI lebih sensitif digunakan dalam mendeteksi adanya perubahan musiman dan mendadak pada tutupan lahan dengan vegetasi yang rapat dan berkanopi lebar seperti hutan tropis. Sedangkan NDVI lebih sensitif digunakan untuk mendeteksi komponen musiman dan perubahan mendadak terutama untuk tutupan lahan yang memiliki vegetasi jarang seperti perkotaan, lahan terbangun dan padang rumput.Kata kunci: Additive Model, BFAST, EVI, NDVI, MODIS]]>
