PENGARUH RESOLUSI SPASIAL PADA CITRA PENGINDERAAN JAUH TERHADAP KETELITIAN PEMETAAN PENGGUNAAN LAHAN PERTANIAN DI KABUPATEN WONOSOBO

Sigit Heru Murti

Abstract


Untuk menunjang pengembangan sektor pertanian di Indonesia, data tentang luas dan sebaran lahan pertanian secara spasial merupakan informasi yang sangat penting. Penelitian ini bertujuan untuk mengkaji kemampuan dan ketelitian citra penginderaan jauh multisensor dan multiresolusi untuk memetakan lahan pertanian di sebagian wilayah Kabupaten Wonosobo dengan pendekatan ekologi bentanglahan. Pemilihan daerah kajian didasari olehkondisi topografi yang bervariasi dengan ukuran lahan pertanian yang tidak begitu luas.

Data penginderaan jauh yang digunakan dalam penelitian ini adalah : (a) citra Landsat ETM+ yang direkam tanggal 21 Agustus 2002, (b) citra Terra ASTER VNIR yang direkam tanggal 29 September 2003, dan (c) citra ALOS AVNIR-2 dengan tanggal perekaman 29 September 2006. Penggunaan ketiga citra tersebut mempertimbangan perbedaan resolusi spasial  ketiganya. Metode pemetaan penggunaan lahan pertanian yang digunakan adalah klasifikasi multispektral terselia yang diintegrasikan dengan pendekatan ekologi bentanglahan.

Hasil penelitian menunjukkan bahwa ketelitian citra untuk interpretasi penggunaan lahan pertanian di daerah penelitian adalah: (a) ketelitian Landsat ETM+ 89,30%, (b) ketelitian ASTER VNIR 91,49%, dan (c) ketelitian ALOS AVNIR-2 adalah 93,62%. Pengaruh resolusi spasial terhadap ketelitian interpretasi sangat besar, ditunjukkan dengan semakin tinggi resolusi spasial citra, semakin tinggi pula ketelitian hasil interpretasinya, yang berujung pada semakin tinggi ketelitian pemetaannya.

Kata kunci :  penginderaan jauh, multisensor, multiresolusi, ketelitian interpretasi, pemetaan lahan pertanian

ABSTRACT

In the effective management and development of Indonesian agricultural sector, information about the distribution of agricultural land in spatial context becomes critical. The aim of this research was to understand the capability and accuracy of remote sensing data at various spatial resolutions to map agricultural land in some part of Wonosobo Regency based on landscape ecological approach. Wonosobo was selected as the study area due its topographic variation and not-so-vast agricultural area. Remote sensing data used in this research were Landsat 7 ETM+ acquired on 21st August 2001, ASTER VNIR acquired on 29th September 2003, and ALOS AVNIR-2 recorded on 29th September 2006. The spatial resolutions of the three images became the basis of the image selection.  Maximum likelihood classification was integrated with landscape ecology approach to perform the mapping. Moreover, we also incorporated several spatial data such as topographic map (RBI) and field survey data to improve mapping process. The results showed that better spatial resolution delivered better map accuracy. It was shown by the accuracy assessment of agricultural land map derived from ALOS AVNIR-2 which reached 93.62%. ALOS’s results outperformed Landsat 7 ETM+ and ASTER with only 89.3% and 91.49% accuracy respectively. The results of this research highlighted the benefit of having higher spatial resolution for agricultural land mapping.

Keywords: remote sensing, spatial resolution, accuracy assessment, agricultural

Full Text:

PDF

Article Metrics

Abstract view : 522 times
PDF - 2239 times

Refbacks

  • There are currently no refbacks.


Copyright (c) 2015 JURNAL ILMIAH GEOMATIKA

Geomatika Indexed by:

 

Copyright of Geomatika