Abstract
Estimating organic carbon fluxes in data-limited tropical island catchments is constrained by sparse discharge records and concentration measurements. For the Changhua River, Hainan Island, we compared three calendar-month runoff-coefficient models based on the mean, median, and geomet-ric mean with a linear model incorporating current- and previous-month precipitation. Models were fitted to monthly records from 2011–2023 and evaluated against all 12 months of 2024. Holdout root mean square errors were 42.669, 39.544, 38.382, and 31.801 mm, respectively. The dual-precipitation model performed best, with a Nash–Sutcliffe efficiency of 0.763 and a percentage bias of −10.220%. Its fixed parameters were applied to September–November 2025 precipitation. The resulting monthly equivalent runoff coefficients were transferred to Ningyuan and Lingshui using their precipitation and drainage areas. Estimated runoff volumes were multiplied by equally weighted mean concentrations from three October–November surveys. Conditional dissolved or-ganic carbon exports were 2763.373, 730.024, and 655.773 t C for Changhua, Ningyuan, and Lingshui, respectively; corresponding particulate organic carbon exports were 571.135, 201.645, and 441.045 t C. These estimates depend on unverified catchment transferability and concentration rep-resentativeness, particularly for unsampled September. They apply to a limited observation period and do not establish annual carbon budgets or carbon sequestration.