Direct-tangible costs in flood zones simulated using the HEC-RAS 2-D hydraulic model – the Arazkuseh River, Golestan Province

Introduction

Flood events are the most complex natural hazards that endanger human and animal lives, social and economic settings, and environmental resources more than any other natural disaster. This phenomenon is caused by the water flow exceeding the river's channel capacity. The expansion of flood zone along river banks in recent years due to climate change and inappropriate use of natural resources is associated with irreparable socio-economic and environmental damages. Simulation of potential flood zones is crucial for management purposes of flood prone areas. Hydraulic models are proved to be useful in simulating flood zones, identifying hotspot areas, and thus, estimating potential damages. The Arazkuseh River is situated downstream of three watersheds, namely Minodasht, Narmab and Nodeh Khandooz. It is prone to flooding during periods of heavy rainfall in the watersheds. The aims of this research are to assess the performance of the HEC-RAS 2-D hydraulic model in simulating the flood zone for the event of March 17, 2019, and to estimate the direct-tangible damages incurred in a 9-km river reach from the Arazkuseh River due to floods in different return periods.

Materials and Methods

The HEC-RAS software has the ability to calculate water level in rivers while considering hydraulic structures. To estimate the velocity vectors, the two-dimensional diffusion wave was used, which takes into account more stable numerical solutions and reduces the calculation time. A Digital Elevation Model (DEM) map with a resolution of one-meter was used to create input terrain data for this model. The flood event hydrograph on March 17, 2019 with a peak discharge of 355 cubic meters was recorded at the Arazkuseh hydrometric station located at the joint outlet of the upstream watersheds. Manning's roughness coefficient values estimated based on field observations in the channel and flood plain were also calibrated with the index F during the evaluation of the model's performance. In order to evaluate the performance of the HEC-RAS 2-D model, the outputs of the model for the flood event on March 17, 2019 were compared with the flood zones identified by the Sentinel-2 satellite images at two different days (23 March and 2 April, 2019). The Pilgrim's computational method was used to identify temporal distribution model of the design rainfalls for the Arazkuseh watershed in different return periods. Additionally, the CoKriging geostatistical method was used to estimate the spatial pattern of the design rainfalls. Thus, the hydrographs simulated by the HEC-HMS hydrological model for the design rainfalls were considered as inputs to the HEC-RAS software. Following identifying the elements exposed to flooding, direct and tangible damages caused by the simulated floods to different land uses were estimated through collecting information from different sources accompanied by field observations.

Results and Discussion

The F index values (79% and 71% for 23 March, 2019, and 2 April, 2019, respectively) indicate that the HEC-RAS 2-D model has an acceptable performance in simulating the flood zone areas in the Arazkuseh River. However, the area of the simulated flood zone shows an overestimation compared to the flood zone observed by the Sentinel-2 images. The overestimation of the flood zone areas by the HEC-RAS 2-D model can be related to the accuracy of the DEM map and the Manning's roughness coefficient estimation. Analysis of the flood zone for the 100-year return period, as a base flood, reveals that crops, trees, dirt roads, residential areas, and asphalt roads are most likely to experience inundation, respectively. Even in 10-year return period, crop lands are likely to place in the flood zone due to gentle slope and the proximity to the river bank. The highest amount of direct-tangible costs for the 100-year flood, is associated to crop lands, residential areas, trees, dirt road, and asphalt road with values of 20889, 8650, 7503, 2250, and 1750 million Iranian Rial, respectively.

Conclusion

The use of DEM data with appropriate spatial resolution is very important in creating terrain data and simulating flood zones in two-dimensional models. Damages to agricultural products and costs attributed to the removal of sediments and cleaning up in crop lands will be significant due to the spreading out of the flood water in this land use. The high damage incurred to the residential areas is because of costly repairs required after flooding and also the price of houses' contents. The total area of land uses which are exposed to 100-year flooding is about 23 hectares and the total damage imposed is approximately, 41042 million Iranian Rial. Therefore, due to the expansion of the residential areas along the river, it is necessary to reduce the hazard of flooding, to enhance adaptive capacity and coping capacity, and to decrease the level of exposure. In crop lands, managers and practitioners should reduce the amount of damages to this land-use by appropriate actions such as flood insurance promotion and introducing resistant varieties to inundation in the study area.