Precipitation design basis calculation updating using best fit program

The current study on the calculation of design basis value for precipitation has been based on an assumption that the maximum daily precipitation (mm/day or mm/24-hour) for regions in Beji, Bangsri, Jepara, Balong and Jatisari follows Generalized Extreme Value (GEV)distribution of Type 1 or also known as Gumbel Distribution. This study update the distribution through Best Fit 4.5 program available as a trial version from Palisade (www.palisade.com). This is a 10-day evaluation program. Upon execution, the program outputs the most fitting distribution according to one of the Chi-square, Anderson-Darling (A-D), or Kolmogorov-Smimov (K-S)tests. It is found that no tall data follows GEV distribution. A - D and K - S tests resulted in similar result and the K - S test was finally chosen as the most suitable one a sit does not require a large number of data and not particularly sensitive to segmentation of data. The result of the design basis calculation (mm/day)is 285 (Bangsri), 285 (Beji), 203 (Jatisari), 248 (Jepara) and 255 (Keling). Percentage difference ranging from 2 to 25 %, which means that the accuracy in determining the data distribution could significantly affect the prediction of design basis for precipitation. (author)