Water Availability Analysis of the Cijurey Reservoir Using Synthetic Discharge Generation with the Thomas-Fiering Method

Abstract

Reliable inflow information is essential for reservoir planning, particularly when historical hydrological records are limited. This study evaluates the water availability of the Cijurey Reservoir using historical inflow discharge for 1991–2020 and synthetic discharge generated for 2021–2040 using the Thomas–Fiering method. Monthly statistical parameters derived from the historical record were used to generate synthetic discharge, while a 1,000-member Monte Carlo ensemble was applied to quantify stochastic uncertainty. Model performance was evaluated using ensemble statistics, monthly mean comparison, flow-duration curves, boxplots, and histograms. The ensemble-average annual water availability was 85.11 million m³/year, closely matching the historical mean of 85.07 million m³/year, indicating no systematic wet or dry bias. The generated series also reproduced the main seasonal and distributional characteristics of the historical inflow, with relatively high water availability during January–April and lower availability during July–September. Although the generated annual variability was slightly lower than the historical variability, the historical standard deviation remained within the 90% ensemble interval. These results demonstrate that the Thomas–Fiering method, when combined with Monte Carlo simulation and multi-diagnostic validation, provides a practical and statistically defensible framework for generating plausible inflow scenarios for the Cijurey Reservoir. The resulting ensemble can serve as a hydrological basis for subsequent reservoir water-balance and reliability analyses.