عنوان مقاله [English]
2D Flood Simulation Using the Nays 2D Flood Model and Comparison with the Sentinel 2 Satellite Image (Case Study: Flood of March 2019 at the end of Arazkuse River, Golestan Province)
Background and Objectives: Flooding occurrence with rare return periods in most regions of the country, and its harmful human and financial effects which are intensifying often due to human actionshas brought the issue of flood management under more attention. Numerous numerical models have been developed to simulate flooding. These hydraulic models are three types, i.e., one-dimensional, two-dimensional and three-dimensional. Due to the fact that it is necessary to simulate inundation accurately with low cost, a two-dimensional horizontal hydraulic model called Nays 2D Flood was used in the present study.
Materials and Methods: aforementioned model was used for a short period of time, it's before joining the Gorganrood River for the flood event occurred on March 20, 2019, as this phenomenon is one of the rare floods for region. For this purpose, a DEM with one meter resolution and flood hourly hydrograph were given to the model. Then an appropriate grid was determined and the roughness of the channel and flood plain were mapped using the Google Earth images in the model software environment.
Results: The results of this simulation include maps of depth and flood zone and flow rate. The results were evaluated by computing the F coefficient and comparing the area calculated by the model with the images recorded by Sentinel2 satellite on the same date. The F statistic was calculated to be 0.78 in the present study. This number indicates that there is an acceptable match between the area calculated by the model and the area observed by the satellite. The flood area is 13.95ha in the satellite image, while the simulated area was estimated 14.30ha by the model. To evaluate depth results, from rating curve data were used. The depth calculated by the model is 6.22 m on 5 April 2019. At that date, the depth of water was observed according to the rating curve measuring 6.63 meters. The error rate in this evaluation was calculated 6.18%.
Conclusion: These results suggest that the Nays 2D flood model has an acceptable function and it can also be used in future research. Since the flood movement in the flood plain is in length and width, two-dimensional models that take into account this fact are preferred one-dimensional models. The results of this study and similar research indicate the good accuracy of these models in simulating the flood zone outside the river channel. Having a high resolution DEM, the model can be used successfully without channel sections data when the flood has spread out of the channel bed.