DOI: 10.5593/SGEM2014/B31/S12.041


M. Iosub, A. Enea, O. E. Hapciuc , G. Romanescu, I. Minea
Wednesday 1 October 2014 by Libadmin2014

References: 14th International Multidisciplinary Scientific GeoConference SGEM 2014, www.sgem.org, SGEM2014 Conference Proceedings, ISBN 978-619-7105-13-1 / ISSN 1314-2704, June 19-25, 2014, Book 3, Vol. 1, 315-322 pp

The purpose of this study is to accurately estimate the flood affected areas for Leghin village, which is situated on the Ozana river flood plain, in Romania, by combining GIS dedicated software and statistical analysis. A comparison was made, between an automated flood simulation method, designed specifically for hydrological analysis (HEC-RAS), and a manual method consisting of a combination between a GIS dedicated and an image processing software (TNTMips and Gimp), using the same input parameters. The Ozana river, located in Neamţ county Romania, has a large negative impact on the 9 villages and 1 major city (with a population of over 42 000 people), it flows through, when a flood occurs. The average flow rate was calculated at 2,2m3/s (for Leghin hydrological station), but it has reached a maximum historical value of 250 m3/s, causing significant damage. The flood analysis has been made for river flow rates, specific to two occurrence probabilities (0.1% and 5%), calculated through statistical methods, such as Pearson III, which was used to determine the occurrence probability (the method was validated for Romanian hydro-meteorological conditions) and Weibull distribution, used to identify the flow rate values, for an extended period of time. After generating the flood simulation, corresponding to a water level specific to the 0.1% probability, a number of 30 buildings subjected to flooding were identified, and a large surface of arable land was estimated to also be submerged. The areas where the vulnerable buildings are located (referring mostly to houses and sheds), are estimated to be covered by a 40 – 80 cm water level. The final results reveal the areas that are subjected to flooding, calculated through both HEC-RAS, and the combination of GIS & image processing software method, at the chosen probabilities. By comparing results from two different flood analysis methods, we are able to more accurately emphasise the areas affected by floods. This is also a good method to validate one method through the other, and observe substantial result deviations.

Keywords: flood zone, HEC-RAS, GIS, probability, simulation, statistical analysis