#TITLE_ALTERNATIVE#

• Main Author: UTARI (NIM : 25313039), RIRIN
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/30524
• Institution: Institut Teknologi Bandung

<p align= "justify">Flow discharge data has an important role in the management and development of water resources in the river. Flow discharge data must be available in time series and accurate, therefore, any empty period is not allowed. To fill the empty period of data, a model that can fill or estimate the empty flow discharge is needed. Discharge generation models that were used in this study are Spatial Correlation Model, Chain Markov Model, FJ. Mock Model, NRECA Model and ARIMA Model. The aim of this research was to find the suitable discharge generation model for monthly discharge in Upstream Cikapundung Watershed. Best model determination was calibrated by comparing the value of correlation, KAR, and RMS. From the simulation of discharge generation model, the greatest value of correlation (r) occurred when using Spatial Correlation Model, which is equal to 0.87, then followed by FJ. Mock, NRECA, Chain Markov and ARIMA Model. The result of average absolute error with KAR value for all models ranged from 0.004 to 0.009. The smallest value, 0.004, was achieved by Spatial Correlation Model. The last calibration is by comparing the average absolute error by RMS. From this calibration, Spasial Corelation Model achieved the minimum average absolut error among other models as well. Spatial Correlation Model and ARIMA model can be used to fill the empty period, discharge data generation and to forecast the flow rates in the future. Meanwhile, FJ. Mock, NRECA, and Chain Markov Model can only be used to complete and generate discharge data.<p align= "justify">