LOOKING INTO SMALLHOLDER COFFEE PLANTATION WITH ALOS DATA IN LAMPUNG PROVINCE, INDONESIA

摘要

Coffee production in Indonesia has im plications in both international and domestic economic significance. Smallholder plantation is the dominant form of coffee farming system, which can pose dif ficulty in accurate assessment of cof fee growing areas in both spatial and temporal context. The objective of this study is to investigate the performance of information extraction about the coffee plantation using the combination of ALOS/AVNIR-2 (10 m resolution MS) and ALOS/PRISM (2.5 m resolution PAN) data in Lampung Province, the heart of the country' s robusta cof fee region. I mprovement in spatial characteristics over conventional Landsat-type medium resolution satellite image data enhanced the performance to identify and isolate patchy cof fee plantations of less than 1 ha in size. Previous efforts in mapping spatial extent of cof fee plantations has been overly generalized and lim ited to smaller than 1:100,000 scale. Persisting cloud cover still is the major challenge in the area but detailed delineation of cof fee plantations made possi ble defined the areas of interest for further monitoring efforts. The use of data from different sensors onboard ALOS platform indicated that smallholder coffee plantation in Lampung Province can be differentiated from surrounding seasonal crop growing areas at mapping scale of 1:50,000 that has not been readily feasible to date.