How Many Regions Does a Nonlinear Muskingum Model Need? Bayesian Selection and Flood Validation
Year: 2026
Type: article
Abstract: This record contains the Supplemental Materials of the paper “How Many Regions Does a Nonlinear Muskingum Model Need? Bayesian Selection and Flood Validation” by G. Gurarslan (submitted to the ASCE Journal of Irrigation and Drainage Engineering). The study asks how many inflow sub-regions the partitioned nonlinear Muskingum model with lateral flow (NLMM-L) of Ayvaz and Gurarslan (2017) needs. Admissible partitions are enumerated to obtain global calibration optima. Parameters are sampled within each partition with DREAM, partition evidence is estimated by bridge sampling, and partitions are combined by Bayesian model averaging. The selected models are tested on an independent River Wye flood (January 1969). The file supplemental_materials_JIDE.pdf contains: Section S1: Sampler settings and error-model priors Section S2: Digitization of the January 1969 River Wye flood Section S3: Residual diagnostics Tables S1–S4: Posterior parameters, posterior summaries, posterior probabilities of the number of sub-regions, and the comparison of NLMM-L with LMM-L Figs. S1–S3: Calibration SSQ versus number of sub-regions, posterior distributions of the two-region model, and deterministic predictions of the January 1969 flood Notation list We will add the code and data to this record after the paper is accepted.
Author: Gurarslan, Gurhan
Institution: Pamukkale University
Language: English
Subfield: Water Science and Technology
Field: Environmental Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225467628