Main Article Content
The availability of information on the degree criticality of accurate land has particular significance in forest and land rehabilitation program that watershed priorities which will be rehabilitated can be know. From the above problem needs a way to determine the priority watershed that will be rehabilitate. The method used in this research is a K-Modes Clustering. K-Modes Clustering gives a model of dataset into clusters where data on a cluster that has the same characteristics and has different characteristic from other clusters based on land inquiries level parameters. From this research obtained watershed group with low scores in the area of protected forest. It is found in cluster 2 that has criticalland criteria in form of the moderate land cover, the sloping slope, the verti severe erosion harzard and the poor management.