ESTIMATION OF RAT PESTS ATTACK IN OIL PALM PLANTATION USING BLOCK KRIGING

• Main Author: Wedy Yeni Octaviani, Resti
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/38949
• Institution: Institut Teknologi Bandung

In oil palm plantation, pests are problem that affects the productivity of palm plants. In this thesis, pest that affect oil palm productivity is rat. Case study is carried out on an oil palm plantation in Sulawesi. The spatial relationships of these plantations can be displayed with variance of the difference in the number of rat attacks between two locations, that spatial relationships called semivariogram. Estimated number of rat pests can be obtained using block kriging. Calculations of block kriging to obtain estimation of the number rat pests in an area will use the covariogram function. Covariogram function is obtained by using the relationship between covariogram function and semivariogram function. The best semivariogram model is chosen by the smallest value of sum of square error (SSE). Covariogram function then used to estimate the required covariance in the block kriging. The results of kriging estimation of each model then compared, SSE kriging would determine the best semivariogram model. In this case, model with smallest SSE model is the model that gave smallest SSE kriging. The model is Gauss, with SSE for 16 discretization points is 400.420. The estimated value of ???? from the Gauss model is 10.1 units. This means in a radius of 10.1 units, the number of rat pests between locations is similar. Thus, it is necessary to watch out for areas that have high rat pests toward the surrounding area within the radius.