In this study, numerical simulations using the Environmental Fluid Dynamics Code model were conducted to elucidate the effects of flow structures in the recirculation zone on solute storage based on the junction angle. Numerical simulations were performed at a junction angle of 30° to 90° with a momentum flux ratio of 1.62. The simulation results revealed that an increase in the junction angle caused the recirculation zone length and width to increase and strengthened the development of helical motion. The helical motion increased the vertical gradient of the mixing layer and the mixing metric of the dosage curves. The recirculation zone accumulated the solute as a storage zone, which formed a long tail in the concentration curves. The interaction between the helical motion and recirculation zone affected the transverse mixing, such that the transverse dispersion had a positive relationship with the helical motion intensity and a negative relationship with the recirculation zone size. Transverse mixing exhibited an inverse relationship with the mass exchange rate of the recirculation zone. These results indicate that the transverse dispersion is replaced by mixing due to strongly developed storage zones.

This study aims to analyze the influences of momentum ratio (M_r) and confluence angle (α) on the transverse dispersion in an urban scale confluence channel from the numerical simulation results using the Environmental Fluid Dynamics Code model. By changing the momentum flux and confluence angle from the simulation results, the analysis focused on the relations between the vertical variations of transverse velocity and transverse dispersion. The high momentum tributary aligned the mixing interface toward the outer bank and created a strong helical motion, which transported the contaminated water along the channel bed and inflows into the recirculation zone. The high momentum ratio induced the large vertical shear in transverse velocity with a strong helical motion and increased the transverse dispersion. However, the helical motion persistence rapidly decreased as the flow reached downstream and led to a decrease in the transverse dispersion for the large confluence angle. Thus, the transverse dispersion coefficient increased with a high momentum ratio and low confluence angle, and the dimensionless transverse dispersion coefficient was in the range of 0.39–0.67, which is observed in meandering channels, for Mr > 1 and α = 45°.

The advection–dispersion equation has been widely used to analyze the intermediate field mixing of pollutants in natural streams. The dispersion coefficient, manipulating the dispersion term of the advection–dispersion equation, is a crucial parameter in predicting the transport distance and contaminated area in the water body. In this study, the transverse dispersion coefficient was estimated using machine learning regression methods applied to oversampled datasets. Previous research datasets used for this estimation were biased toward width-to-depth ratio ($W/H$) values ≤ 50, potentially leading to inaccuracies in estimating the transverse dispersion coefficient for datasets with $W/H$ > 50. To address this issue, four oversampling techniques were employed to augment the dataset with $W/H$ > 50, thereby mitigating the dataset’s imbalance. The estimation results obtained from data resampling with nonlinear regression method demonstrated improved prediction accuracy compared to the pre-oversampling results. Notably, the combination of adaptive synthetic sampling (ADASYN) and eXtreme Gradient Boosting regression (XGBoost) exhibited improved accuracy compared to other combinations of oversampling techniques and nonlinear regression methods. Through the combined ADASYN–XGBoost approach, it is possible to enhance the transverse dispersion coefficient estimation performance using only two variables, $W/H$ and bed friction effects ($U/U^{*}$), without adding channel sinuosity; this represents the effects of secondary currents.

Objectives: The USA-Mexico border area faces an elevated diabetes burden, with prevalence 2-3 times higher than national averages in both countries. Yet, no forward-looking projections exist beyond 2021, limiting healthcare system planning. We provide the first multi-indicator, multi-model forecasts of diabetes burden through 2030 across border states. Methods: Diabetes mortality, prevalence, and disability-adjusted life years (DALYs) data were extracted from the Global Burden of Disease Study 2021 for all ten border states in both countries (1990-2021). Eight different time-series models including ARIMA, generalized additive models, generalized linear models, Prophet, and n -sub-epidemic framework were used to project burden from 2022-2030. Models were evaluated and compared using weighted interval scores, with forecasts stratified by age, sex, and geography. Results: Multi-model ensembles project 15-17 million people with diabetes, 23,000-26,000 annual deaths, and 1.6-1.8 million Disability-Adjusted Life Years (DALYs) due to diabetes in USA-Mexico border states by 2030. Mexican border states face accelerating mortality among adults aged 20-59 years (15-58% increases), while USA states show prevalence increases of 35-52% in youth and 47-49% in older adults. Males experienced consistently higher projected burden than females. ARIMA demonstrated superior performance for mortality and prevalence forecasting, while ensemble methods performed better for DALYs projections. Conclusions: Diabetes burden is projected to increase 30-50% across border states by 2030, driven by distinct regional patterns requiring divergent public health responses. Mexican states require community health worker programs, subsidized medications, and mobile clinics for mortality prevention, while US states must expand school-based screening and nutrition program capacity to address rising youth prevalence. These projections provide actionable intelligence to support coordinated cross-border efforts in diabetes prevention and management. Strengths and Limitations of This Study: Strengths: – Multi-model ensemble approach (eight models) with rigorous comparative evaluation using standardized metrics, providing transparency on performance variability across outcomes and demographics. – Comprehensive stratification by age group (five categories), sex, and geography enables demographically targeted healthcare planning, addressing a gap in prior border health forecasting studies. – Explicit quantification of forecast uncertainty through 95% prediction intervals and formal calibration assessment, identifying outcome-specific and geographic instances of substantial uncertainty. Limitation: – Forecasts use statistically modeled disease burden estimates (GBD 2021) rather than primary surveillance data, potentially introducing measurement error and limiting precision in capturing true state-level variation. – Time-series models assume historical trends continue unchanged, unable to account for policy interventions, technological innovations, or epidemiological shocks. Forecasts are most reliable in short term.