Floods are the most common natural hazard on earth and are the natural phenomenon of extreme weather conditions, often violent. With an increased population growth along with global warming and climate change, a disaster like flooding worsened yearly. Fighting natural disasters like flooding can be possible with comprehensive risk analysis
One of the flood mitigation measures is the development of Flood Inundation Mapping [FIM]; an inundation can be defined as a spread out of an expanse of water that submerges the land, it occurs when channel capacity fulfills, and water flows out of the channel. Flood inundation simulation simulates severe flood events, developing hazard maps, and risk mitigation measures worldwide are discussed
A vital tool for flood monitoring is the HECRAS software, and remote sensing (RS) technology, in combination with geographic information systems (GIS), is used for developing Flood Inundation Mapping [FIM]
The flood inundation map of the floodprone areas is carried out using HECRAS 2D models with the incorporation of ArcGIS tools
Bhīma is a significant tributary of the Krishna River and one of the two major rivers of Maharashtra state, with the other being the Godavari. Bhīma originates at BhīmaShankar in the Sahyadri Ghats at the elevation of MSL 700m. The banks of Bhīma are densely populated and form fertile agricultural areas. The river is prone to frequent flooding due to the heavy monsoon season.
Land use Land cover (LULC) has primary importance along with other spatial databases like soil type, lithology, hydrology soil group in assessing flood risk
The digital elevation model (DEM) and runoff excess quantity plots have been evaluated using ArcGIS with an extension of HECGEORAS. The digital elevation model (DEM) is a raster dataset that gives information about the topography. It is constructed using data obtained from remote sensing techniques and sometimes data gathered from land surveying. The accuracy of the Digital elevation model (DEM) plays a vital role in flood simulation and floodplain mapping
The HECGEORAS is a spatial extension of ArcGIS to develop geospatial data to be used in HECRAS. This extension works as an interface to transfer data between ArcGIS and HECRAS. HECRAS is a river modeling computer program used for the simulation of hydraulics of water flow in natural rivers as well as other channels. HECRAS is a proven efficient tool for analyzing flood models and inundation maps
In the current study, the Bhīma River reach under study was defined. In preprocessing phase consisting of the preparation of model input, geometry data is done with the help of HECGEORAS in ArcGIS. The highresolution elevation data DEM is converted to a triangular irregular network (TIN) elevation model. This TIN model is in the raster format in ArcGIS. The quality of the TIN model is based on its cell size or resolution. For more accurate analysis smaller cell size is preferred because the smaller the cell size more excellent the resolution resulting in high accuracy. The current study uses a resolution of 30 m x 30m shuttle radar topography mission (SRTM) DEM of Bhīma River. The HECGEORAS extension obtains the required features like stream centerline, bank stations, and crosssections. In the processing phase, the geometry attributes created with HECGEORAS are imported into HECRAS, and analysis has been carried out. In the postprocessing phase, the outputs of the HECRAS model are imported in ArcGIS, and water surface TIN was created for a 100year profile based on water surface elevation (WSE). The conceptual framework of the methodology adopted is shown in
The flood frequency analysis plays a vital role in the case of stream/river hydrology. Past flood event records need to be evaluated to check their future occurrences. Flood problems can be quantitively and qualitatively assessed by estimating the frequencies of the flood
In the frequency of a hydrologic event, the annual peak flow value is the probability that a value will be equaled or exceeded in any year. The flood frequency analysis is one of the significant studies of river hydrology conducted based on maximum instantaneous flow
X_{T} = mean +K*std. deviation.
Flood frequency analysis is done based on maximum flood discharge recorded at Phulgaon, Maharashtra, India gauging station between 1992 to 2018 for 5, 10, 25, 50 and 100year return period using Gumbel Distribution (EV1).
Shuttle radar topography mission (SRTM) DEM coupled with Sentinel1 Synthetic Aperture Radar data is used to prepare flood hazard layers in terms of the probability of flood inundation in the GIS platform
In HECRAS, three specific parameters must be needed. The first one is stream geometry, the second is flow data, and the third one is the Model plan. The stream geometry data is in the RASGIS file. For the analysis, additional parameters like Manning's roughness coefficients which are taken from the LULC map for every crosssection, channel contraction, and expansion coefficients, are required. After putting all geometry data, steady flow data is given to the model, and finally, by assigning the Model plan, the model has been given a run. The model outputs are then imported back to GIS for the development of flood inundation mapping.
Simulation for onedimensional flow under steady conditions has been performed for possible discharge of return periods of 5, 10, 25, 50, and 100 years. The discharge evaluated for various return periods using the Gumbel flood frequency method has been used to simulate steady HECRAS model. The maximum water surface elevation for each profile has been simulated, and maximum water surface elevation of 25year and 100year return periods have been plotted against elevations of the right bank and left bank.
The sensitivity analysis for a model shows how stable the model is. Sensitivity can be checked by outputs developed by a model corresponding to change in the input variables, and it can be plotted to give the best fit of the model. Sensitivity is the rate of variation in one parameter regarding change in another parameter. It is a mathematical tool for the developing, calibrating, and validating a hydraulic model
The primary response parameters to be used to quantify the sensitivity of the onedimensional flow model are Manning’s coefficient and crosssection spacing
The hydraulic simulation model and geographic information system are highly effective for floodplain mapping. Sustainable and economic development of the wider geographic area is achieved through a systematic approach to floodplain management. The present research aims to develop flood inundation maps on the left and right bank of Bhīma River in Maharashtra state, India, and it includes integrating GIS and HEC RAS tools.
The reoccurrence interval of hydrologic data is estimated to match the number of stimulating actions to their frequency of existence over the use of probability circulations^{ }






Mean (Xm) 
1290.87 
1290.87 
1290.87 
1290.87 
1290.87 
Stand Deviation (Sy) 
894.51 
894.51 
894.51 
894.51 
894.51 
Frequency Factor (K_{T}) 
0.89 
1.57 
2.44 
3.08 
3.72 
Skewness (Cs) 
0.73 
0.73 
0.73 
0.73 
0.73 
Yn 
0.53 
0.53 
0.53 
0.53 
0.53 
Sn 
1.09 
1.09 
1.09 
1.09 
1.09 
Peak Discharge (X_{T})(m^{3}/s) 
2085 
2700 
3477 
4053 
4626 
Geometry data with the main channel is shown in
The parameters used are river station (RS), stationelevation data, stream centerline locations of the left and suitable bank stations, downstream reach lengths, expansion, contraction coefficients, Manning’s roughness coefficients, and details of hydraulic structures. The crosssections are perpendicular to the river line and must be stretched to ensure all the flood water is occupied within the crosssectional area. Accuracy in geometry and flow controls increases the model’s efficiency. The floodplain mapping is done with the help of georeferenced shapefiles in ArcGIS. The elimination of errors can be done efficiently in ArcGIS while importing the outputs of the simulations. Floodplain was then delineated for different recurrence intervals. Water surface TIN was obtained for a 100year return period in postprocessing, as shown in
The flood inundation depth map (refer to
It has been observed that the left bank representing the east side of the river is more vulnerable to water spills than the right bank representing the west side of the river. The probability of the number of cross sections on both sides of banks has been estimated to check the vulnerable zone for high water levels for different return periods along the study reach.
It is noticed that the left bank number of crosssections overtopped is 849 (77%) of total crosssections for 25 the year return period, and for the 100year return period number of crosssection overtopped are 995 in number (90%) of total crosssections.
In the case of the rightover bank, the number of crosssections overtopped for a 25year return period is 712, which is 64% of total crosssections. For a 100year return period, the number of crosssections overtopped is 893 which is 81% of total crosssections.
The probable peak discharge for return periods of 25 and 100 years is estimated from Gumbel frequency distribution, and it has been simulated under steady conditions. Maximum water surface levels have been compared for the elevations of both the left (east) and right (west) banks of Bhīma River for all cross sections. The cross sections having elevation lesser than relative water surface elevation on the corresponding bank have been considered unsafe, and these will have the probability of water spill. The percentage of crosssection prone to spilling of water during high discharge on the left bank is 77 % and for the right bank is 64 % for 25year return period while for 100year return periods it has come out as 90% On the left bank and 81 % on the right bank. For all the return periods, it has been observed that the left bank of the Bhīma River is more prone to water spills due to higher water levels in the river than the right bank.
Sensitivity analysis was performed keeping Manning’s roughness (Range 0.02 to 0.045) as a variable, and its effect has been observed on the other parameters; this shows changes in response parameters like average velocity, Froude number, hydraulic depth, and water surface elevations. Manning’s roughness sensitivity analysis results are summarized in






Left Flood Plain 
Central Channel 
Right Flood Plain 
25yr 
100yr 
25yr 
100yr 
25yr 
100yr 
25yr 
100yr 
0.025 
0.02 
0.025 
10.5 
12.2 
13.2 
20 
10.4 
12.7 
4.20 
4.38 
0.03 
0.025 
0.03 
12.5 
15.7 
15 
21.8 
13.8 
15.48 
5.30 
5.38 
0.035 
0.03 
0.035 
14.23 
19.5 
18.6 
22.5 
15.6 
17.5 
6.48 
6.2 
0.04 
0.035 
0.04 
17.5 
22.40 
23.2 
30.2 
19.7 
21.8 
7.68 
8.4 
0.045 
0.035 
0.045 
19.75 
23.61 
25 
32.57 
21 
23 
8.65 
9.61 
From the sensitivity analyses, it has been observed that the response parameters established projected guiding changes in ranged from + 8 to 33% approximately. The rate of change of Manning's roughness coefficient considering 25 years of return period increases from 0.02 to 0.045. The average velocity and average Froude number decreased by 19.75% (i.e., from 4.47 m/s to 2.32 m/s) and 25% (i.e., from 0.96 to 0.43), respectively, while the average water surface elevation (WSE) increase by 21% (i.e., from 549.46 m to 551.78 m), and hydraulic depth increases by 8.65% (i.e., from 1.62 to 2.71m).
The rate of change of Manning's roughness coefficient considering 100 years of return period increases from 0.02 to 0.045. The average velocity and average Froude number decreased by 23.61% (i.e., from 5.46 m/s to 3.24 m/s) and 25% (i.e., from 0.85 to 0.43), respectively. In comparison, the average water surface elevation (WSE) increases by 23% (i.e., from 551.72 m to 553.93 m), and hydraulic depth increases by 8.65% (i.e., from 2.47 to 3.81 m).
The Sensitivity analysis indicates that the model is performing appropriately. As the roughness parameter increased, there was a decrease in Froude Number, average total velocity, and increased hydraulic depth and water surface elevation.
Integrating a hydraulic simulation model and geographic information system is a general and trustworthy approach for floodplain mapping, and it is used in the Bhima River basin in this paper. This approach can be helpful to the decisionmakers in catchment management in various scenarios like flooding, catchment degradation, erosion, and landslips. It is further helpful in the sustainable and economic development of the wider geographic area by providing systematic and consistent information applicable to the catchment. The current study aimed to develop flood inundation maps on the left and right banks of Bhīma River, including integration of GIS and HECRAS tool along Bhima River, reach located in the Maharashtra state of India. The digital elevation model (DEM) of Bhīma River is used to develop the geometric data and delineation of floodplain maps using the HECGEORAS tool. Flood frequency analysis for the Bhima River basin using Gumbel Probability distribution indicates flood magnitudes of 2085 m^{3}/s, 2700 m^{3}/s, 3477 m^{3}/s, 4053 m^{3}/s, 4626 m^{3}/s for the return periods of 2, 5, 10, 25, 50, and 100 years respectively.
It is observed that around 30.8 Km^{2} (37.46 %) and 41.65 Km^{2} (50.66%) area gets inundated for the 25 years and 100year return period, respectively. The flood inundation depth classification reveals that around 17 % to 23% of the total flooded area may get inundated due to water depth of 3.6 m to 5.6 m, and around 1 % to 3.5% area may get inundated by water depth of less than 0.3.m for the return period of 25yrs and 100yrs respectively.
In case of analysis of water surface elevations, the possibility of a percentage of crosssection prone to spilling of water during high discharge on the left bank may be around 77 % and around 64 % on the right bank for the return period of 25 years. Similarly, it may be around 90 % on the left bank and 81 % on the right bank for the 100 years return period. It has been observed that the left bank of the Bhīma River is more prone to water spills due to higher river water levels than the right bank.
The sensitivity analysis with the variations in roughness coefficient (n) ranging from +20% to 20% from its base value recommends that the model is working suitably as the roughness parameter was increased for a 25year return period from 0.02 to 0.045; Froude Number decreases  by 25% (i.e., from 0.96 to 0.43), and the average total velocity decreases by 19.75% (i.e., from 4.47m/s to 2.32 m/s) and increases in hydraulic depth 8.65% (i.e., from 1.62 to 2.71m).and water surface elevation increases by 21% (i.e. from 549.46 m to 551.78 m), also for 100year return period the roughness parameter was increased for 25year return period 0.02 to 0.045 Froude Number decreases 25% (i.e., from 0.85 to 0.43) and average total velocity decreases by 23.61% (i.e., from 5.46 m/s to 3.24 m/s) and increases in hydraulic depth by 8.65% (i.e., from 2.47 to 3.81 m) and water surface elevation increases by 23% (i.e from 551.72 m to 553.93 m). The simulation model developed in this paper has shown effectiveness in floodplain mapping.
The authors wish to acknowledge the Hydrology Project, Water resources department Government of Maharashtra, India, (Hydrology Data Users Group), for providing the hydrologic data used for this study. The author also acknowledged the Sardar Patel College of Engineering, Mumbai, for providing funds for this project under MHRD's TEQIP III, India.