Main Article Content
Abstract
Satellite remote sensing data provide up to date valuable information on landuse/cover existing
condition. Developing standardized and methodology with sufficient accuracy, for assessment the spatial
distribution of agricultural land is prime needed. Landsat Thematic Mapper were used to detect and assess the
spatial distribution of wetland rice and land use/ cover in Lampung Province as a case study by applying digital
analysis Hybrid (supervised) classification approach.
To validate or the accuracy detection is to used a statistically approach sampling design (which are
consist of point sampling accuracy and area sampling accuracy) in the selected sample blocks and sample
segments. Area sampling accuracy mainly stressed to assess the accuracy wetland rice spatial distribution.
To determine the correctness of land use/ cover types is assigned to that pixel matches the true
classification of ground segment observation represented by pixel value of digital satellite images. The result of
land use/ cover analysis and classification were compared with the ground data observation contain accuracy
detection ranging from 76,7 % (bushes) and 100% (forest). Wetland rice accuracy detection have about 84,5%
accuracy. Wetland rice spatial distribution analysis over the segments and sample blocks, were compared with
the ground data assessment and observations have only less than 6% and 21,7% deviation in flat areas and
sloping areas respectively. Increasing on slope steepness, and the variety plant growth/ vegetation will be
followed by increasing deviation. High accuracy detected existing agricultural land use very helpful for
strengthening food security and site selection potential areas for agricultural commodities extensification.
Keywords: remote sensing, detection, digital analysis, Wetland rice, landuse/cover, accuracy assessment