Topography Condition Study Area

25 25 RGB images. Visual interpretation procedure is semi-automatic method using on screen digitizing. Landsat 1990 Landsat 2001 Landsat 2009 Landsat 2001 Images Correction Geometric and Atmospheric Images Classification Accuracy Assessment Land Cover Maps Change Detection Change Detection Analysis Ground Truth Data No Figure 3.9 Image classification processes; four multi temporal images was used to produce land use and land cover classes. Visual interpretation was done by observing the pattern of visible object on the imagery; the object such as river, settlement, and road network are very helpful to assist us to map the vegetation or land cover. The vegetation mapping is performed by delineating the outer boundary of pixels that have same pattern, then it was classified by using an support maps such as land cover maps, topographic, concessions, and vegetation as a reference maps. Based on the existing condition of land cover type in study area, the Landsat images were classed into 6 major classes. The classes are: 1. Forest Land: Area with high density of trees which include primary dry land forest, secondary dry land forest, swamp forest, mangrove, and plantation forest. 26 26 2. Agriculture: Area used for both annual and perennial crop cultivation, and the scattered rural settlements are closely associated with the large sized cultivated field. 3. Shrubs Land: Area covered with shrubs, bushes and small trees, with little wood mixed with some grasses. 4. Water Body: Area which remains water logged and swampy throughout the year, the man made dam, the rivers with its main tributaries, and the lake. 5. Build up: Area with high density of settlement that including high density township residences, and urban area. 6. Barren land: Area dominated by grass and small number of small trees.

3.3.1.2 Change Detection

To identify the differences between two or more land cover maps, post classification and matrix analysis was performed during image processing stage. The matrix analysis is comparing the area of each class in each land cover map, and consists of with two kinds of values; the diagonal matrix contains unchanged value while the other cell contain with a value that have been changed. Second step is generating the probability of changes between classes. Figure 3.10. Change detection procedure Wijanarto, 2006

3.4 Hydrological Modeling

3.4.1 General Description of HEC-HMS

HEC-HMS model was designed to simulate the precipitation-runoff processes of dendrites watershed systems Fleming, 2009. It’s designed to be applicable in a wide range of geographic areas for solving a broad range of Image 1 Image 2 Registration and Calibration Interpretation ‐ Land cover ‐ NDVI Classification Registration and Calibration Interpretation - Land cover - NDVI Classification Transition Matrix Trend Analysis Prediction