ABSTRACT:Lately, computerized hydrologic models have become an essential tool not only for a better understanding of the hydrologic cycle but also for a faster problem solving in hydrology, such as the ungauged catchments. Since Covasna river, which is a left tributary of the Black river, is monitored only by two gauging stations, an approach for runoff modelling at the ungauged outlets within the catchment, would be extremely beneficial. Thus, the main purpose of this study is to conduct a test of the ability of such a simulation modeling system for flood prediction within the upstream catchment area of Covasna river. The one chosen for carrying out this study, HEC-HMS, is known to be a reliable model with reasonable approximations. Throughout the research, ArcHydro and HEC-GeoHMS extensions are used to delineate the watershed, and to obtain the hydraulic length, the CN grid and other parameters. This study is also a solution for the problem of long simulation time of the model, by following first, the work flow implemented in ArcGis 10.2 through Model Builder.

There is an increasing need to assess and quantify the impact of land use/land cover changes, especially on surface runoff, due to rapid population growth. This study aims to investigate the land use/land cover (LULC) changes over time, for an intense rainfall event in Țibleș, Runc and Sălăuța watersheds, and their impacts on surface runoff for various antecedent moisture conditions (AMC). The GIS-based SCS-CN method and the CORINE land cover (CLC) databases for 2000, 2006 and 2012 laid the foundation for this research. Results indicated that even small land cover changes can significantly affect runoff on the short time scale, with quantitatively different effects regarding moisture conditions. The reduction in forest cover due to agricultural intensification and the conversion from pasture to cropland (especially between 2000 and 2006) resulted in higher surface runoff volumes. These changes mostly affected the middle and downstream catchments of the main rivers which means that over the years, the soil water retaining capacity has decreased.

Romania is among the European countries which are still susceptible to flood disasters. Climate change, the lack of a sustainable land management and floods approach are the most important driving factors of flood risk. Accordingly, additional research to improve or develop various methods for predicting the flash flood events, especially within densely populated watersheds, is required. This paper aims to provide a further assessment of the reliability of a distributed hydrological model developed within a GIS framework The model was developed for use in small river basins and so forth, another objective is to provide insight into whether it is acceptable for use in larger-sized watersheds than the ones it was created for. The methodology is presented by computing the discharges at the entrance of six human settlements located within the Bistra river watershed, a left-bank tributary of Barcău river (Bihor department, Romania). The chosen study area is frequently impacted by flooding, endangering the socio-economic activities, human settlements and infrastructure. A distributed hydrological modeling approach and the SCS-CN method are used in this research. The results revealed a good correlation between the estimated and measured hydrographs at the gauging points of Pădurea Neagră (upstream) and Chiribiș (located at the downstream end of the drainage basin).

The NRCS-CN (Natural Resources Conservation Service curve number) method, developed by the USDA (U.S. Department of Agriculture) is among the most widely used for the estimation of surface runoff from watersheds. Ever since its introduction in the 1950s, although it has been used to a great extent by engineers and hydrologists, the applicability of the original procedure used to determine its main parameter, the curve number (CN), to various regions with diverse environmental conditions, is still subject to many uncertainties and debates. This study presents a comparative analysis of different methods applied to determine curve numbers from local data in four watersheds located in the central part of Romania, within the mountain region surrounding the Brașov Depression. The CN values were not only computed using rainfall–runoff records from 1991 to 2020, but also determined from the standard NRCS tables documented in the National Engineering Handbook part 630 (NEH-630), for comparison purposes. Thus, a total of 187 rainfall–runoff data records from the study watersheds and five different methods were used to assess the accuracy of various procedures for determining the CN values, namely: tabulated CN (CN values selected from NRCS tables, TAB), asymptotic fitting (AF) of both natural and ordered data, median CN (MD), geometric mean CN (GM) and the arithmetic mean CN (AM) methods. The applicability of the aforementioned methods was investigated both for the original fixed initial abstraction ratio λ = 0.2 and its adjustment to λ = 0.05. Relatively similar results were found for the curve number-based runoff estimates related to the field data analysis methods, yet slightly better when the λ was reduced to 0.05. A high overall performance in estimating surface runoff was achieved by most CN-based methods, with the exception of the asymptotic fitting of natural data and the tabulated CN method, with the latter yielding the lowest results in the study area.

Hydrologic modelling studies usually involve data series with a large temporal scale, especially in Romania, focusing on a long-term impact analysis. Nevertheless, event-based runoff models are essential tools for short-term purposes such as flash flood forecasting. Suitable methods or models must be considered in order to ensure the validity of such research based on parameter calibration to a particular area. Therefore, a comparative analysis of methods must be conducted first, in order to determine the optimal ones that can be used for future data prediction. The aim of the present study is to apply and validate the MIKE HYDRO River modeling system - the UHM module, through a comparative analysis of the SCS, Generalized SCS and Proportional Loss methods available, to a small-sized mountainous watershed, where no research has been conducted in this field. To this end, three spring rainfall events were chosen, but with different antecedent moisture conditions, in order to examine how well the chosen methods can reproduce the available observations in such circumstances. The SCS method yielded the highest quality performance, but the Proportional Loss method has also proven effective under these conditions.