Remote Sensing Application with Validation Test for Inland Waters Detection in Loa Kulu Minapolitan Area, Kutai Kartanegara Regency

Loa Kulu minapolitan is an area producing land fisheries products in Kutai Kartanegara Regency, Indonesia. The demand for fishery products both within and outside this region is continuously increasing. However, intensification efforts seem to unable to fulfill market demand. Furthermore, the efforts are constrained by information unavailability for aquaculture potential land. Land detection for inland waters in research location is the first step to provide manageable water land to increase fisheries production in this region. This paper is preliminary research aimed to draw an existing map of inland waters distribution at the research site. Google image interpretation and field observations techniques are used to detect the spatial distribution of inland waters. Moreover, the methods for validation processes or data accuracy are point sampling and area sampling accuracy. The analysis results of point sampling accuracy in this study were 80.99% (acceptable result). The percentage of bias at point sampling accuracy is 19.01% (acceptable result). Total bias value which is caused by differences in time taken for two data types. Besides, the object size is smaller than the image resolution ability. While the results of the area sampling accuracy in this study indicate that swampland use has an accuracy of 100%, rice fields 99.91% and rivers 99.50% (acceptable result). In conclusion, study result is considered relevant to be the initial description of inland waters distribution map as an input for research on determining potential land for inland aquaculture development to increase fisheries production at the research site