Document Type : Original Article
Authors
1
Researcher, Atmospheric science center, Iranian National institute for oceanography and atmospheric science, Tehran, Iran
2
Iran meteorological organization , Rasht, Iran
3
Assistant Professor, Department of Hydrology and Water Resources, Faculty of Water and Environmental Engineering, Shahid Chamran University of Ahvaz, Ahvaz, Iran
4
5Associate professor, Iranian National institute for oceanography and atmospheric science, Tehran, Iran
10.22034/jcr.2025.458336.1651
Abstract
This study investigates flood forecasting and examines the effect of the resolution and type of Digital Elevation Models (DEMs) on flood peak simulation using the WRF-Hydro model in the Polrood Basin, located on the southwestern coast of the Caspian Sea. The primary objective of this study is to analyze the impact of different DEMs on improving flood predictions and optimizing the parameters of the WRF-Hydro hydrological model, which plays a crucial role in simulating hydrological dynamics and forecasting flood events. The use of more accurate and high-resolution DEMs can significantly improve prediction accuracy.
In this study, DEM data from three different models were selected. Two hydrological DEMs (HYDROSHEDS and MERIT) and one topographic and geological DEM (USGS) with a 30-meter resolution were used as input data for the simulations. These models were chosen based on their specific applications in hydrological analyses. The HYDROSHEDS and MERIT models are particularly designed for hydrological applications and can effectively simulate surface runoff and water accumulation features. On the other hand, the USGS model is more suitable for simulating geological and topographic features, as it focuses on fine-scale terrain and geological characteristics.
The study focuses on three distinct flood events that occurred on March 27, 2016, April 1, 2019, and March 21, 2021, in the Polrood Basin. These events were specifically selected to illustrate how the effect of different DEMs on flood prediction varies under different climatic and hydrological conditions. The results of the model were carefully analyzed for each event, and relative error in the simulations was assessed.
For each DEM, various parameters such as hydraulic conductivity (REFDK), infiltration parameter (REFKDT), and Manning’s roughness coefficient (MANN) were independently optimized. These parameters play a key role in determining the accuracy of the WRF-Hydro model’s flood predictions. Specifically, hydraulic conductivity and infiltration parameters are critical in surface runoff generation and flood discharge, while Manning’s coefficient influences flow travel time and its velocity. Thus, optimizing these parameters can significantly enhance flood predictions and improve the model’s overall performance.
The results of the simulations indicate that the MERIT DEM performed the best in simulating flood peaks. The relative error in river discharge calibration for the three flood events was -2.3%, -0.15%, and -5.92%, respectively. This indicates that the MERIT DEM is able to simulate floods with high accuracy. Following this, the SRTM DEM provided good results with relative errors of 0.45%, 4.4%, and 14.2%, making it another suitable option for flood prediction, despite slightly higher errors compared to MERIT.
The study also reveals that the REFDK parameter and Manning’s roughness coefficient had the most significant impact on optimizing the results. Manning’s parameter, in particular, has a significant effect on flow travel time. The lower the value of this parameter, the faster the flow passes, and the higher the volume of flow generated.
Based on these findings, the study concludes that using hydrological DEMs (such as HYDROSHEDS and MERIT) can significantly improve the results of the WRF-Hydro model. These models, due to their high accuracy and ability to simulate hydrological features in detail, are well-suited for use in flood simulations, especially in flood-prone basins.
Finally, the study emphasizes the importance of selecting appropriate DEMs and optimizing hydrological parameters in improving flood prediction accuracy. It is recommended that future studies combine DEM models with up-to-date hydrological and climatic data to provide more accurate flood forecasts for flood-sensitive and vulnerable regions.
The parameters REFDK and Manning’s coefficient (MANN) had the most significant impact on optimizing the model results. REFDK influences the rate at which water infiltrates the soil, while Manning’s coefficient affects the speed at which water flows through channels. The study found that Manning’s parameter significantly affects the flow transit time; the lower the MANN value, the faster the transit time and the greater the flow produced. This insight is crucial for flood management as it helps in understanding how changes in channel roughness can alter flood dynamics.
Based on the WRF-Hydro model outputs and the observed data, it is evident that using hydrological DEMs can significantly improve the accuracy of flood simulations. Hydrological DEMs, such as HYDROSHEDS and MERIT, are specifically designed to represent the hydrological features of the terrain, making them more suitable for flood modeling than general topographic DEMs like SRTM.
The combination of numerical weather models with hydrological models enhances the lead time for flood warnings and provides more reliable flood predictions. The WRF (Weather Research and Forecasting) model, when combined with the WRF-Hydro hydrological model, offers a powerful tool for predicting floods in flood-prone basins. The integration of these models allows for the incorporation of real-time weather data into hydrological simulations, improving the accuracy and reliability of flood forecasts.
This study suggests that for the development and implementation of flood warning systems in the country’s flood-prone basins, the WRF model should be used in conjunction with the WRF-Hydro model. The combined approach not only improves flood prediction accuracy but also provides valuable information for flood management and mitigation strategies. By using advanced modeling techniques and high-quality DEM data, authorities can better prepare for and respond to flood events, ultimately reducing the impact of floods on communities and infrastructure.
In conclusion, this research highlights the critical role of DEM resolution and nature in flood modeling. The findings underscore the necessity of selecting appropriate DEMs based on their intended use in hydrological modeling. High-resolution, hydrologically accurate DEMs like MERIT significantly enhance the performance of flood simulation models like WRF-Hydro. By optimizing key parameters and using advanced modeling tools, it is possible to achieve more accurate flood predictions, which are essential for effective flood management and disaster preparedness.
Keywords
