| dc.description.abstract | Data and information either having high validity or high reliability are still limited. This is might due to the difficulty to obtain the data and cost reason. The difficulty to survey the private forest is affected by several of the following factors, i.e. small size of private forest, spreadly distributed and converted easily to another land use. To overcome those dificuties, it is important find out a survey method that can be accurately estimate the private forest condition and potency. Survey methods examined in this research include social-economics and biophysic-based method, image-based method using Normalized Dtflerence Vegetation Index (NDVI), and method based on integration of basic data and image. Correlation test were performed for all private forest variables. Then stratification was established based on the selected factors. A method is recognized as accurate if it has low coefficient of mean variation, this describes a homogenity strata, and high in coefficient of variation of among strata, showed each strata has specific value that different with the other ones. Maximum difference between coeficient of mean variation and coefficient of variation of among strata was investigation. The result showed that basic data method has higher accuracy than another to method in which dominant variables are rural density and listance to city. | id |
