China
Africa
Explore chapters and articles related to this topic
Flood hazard assessment
Published in Saowanit Prabnakorn, Integrated Flood and Drought Mitigation Measures and Strategies, 2020
In this study, the ArcSWAT 2012 interface is adopted to set up and parameterize the model, which requires a large number of input data for developing its database and constructing the model. Data of the physical and chemical properties of all soil types and layers (i.e., texture, infiltration rate, saturated hydraulic conductivity, etc.), as well as continuous time-series records of climatic data from 1985–2015, were compiled and embedded in the SWAT database (the .wgn and .sol files). The DEM and its stream network were used in the watershed delineation process. With a threshold drainage area of 100,000 ha, the watershed is first divided into sub-basins. Then, the sub-basins are subdivided into hydrologic response units (HRUs) based on unique combinations of land use, soil type, and slope characteristics (Arnold, Kiniry, et al., 2012). Consequently, the study area is comprised of 63 sub-basins (Figure 4-2) and 379 HRUs. The simulation was executed at a daily time step from 2005 to 2014, of which 2008–2011 was for calibration and 2012–2014 for validation. The first three years, 2005–2008, are a warm-up period (the equilibration period to mitigate the initial conditions) and was excluded from the analysis (Abbaspour et al., 2015).
Hydrologic models and geographic information systems
Published in Stephen A. Thompson, Hydrology for Water Management, 2017
GIS is used not just to parameterize models, but as a hydrologic model itself. Stuebe & Johnson (1990) automated the SCS curve number method for the GRASS GIS. First they manually digitized SCS soil survey sheets to create a hydrologic soil group layer. Next they created a landcover layer using vegetation data from the LANDSAT satellite, and land use data from existing aerial photography. These layers were combined using GRASS to create hydrologic response units (HRU), and the GIS calculated and assigned curve numbers to each HRU. In addition they had the GIS automatically delineate watershed boundaries using Digital Elevation Model (DEM) data. A DEM is a file containing digital elevation data. A Digital Line Graph (DLG) file contains digital representations of linear features (lines and polygons) including stream channels, roads and political boundaries. They compared the GIS-generated runoff estimates to manual runoff calculations in six watersheds. The watersheds had land use/land cover varying from farmland to forest. The GIS-generated runoff estimates were similar in four of the six watersheds (Table 13.8). The mean difference in these four basins was 9.5%. In two basins (basin 3 and 7 in Table 13.8) the mean error was 30.5%. The authors feel that most of the error in these two watersheds resulted from the automated watershed delineation procedure. DEM data contain subtle elevation errors which can influence watershed boundary delineation in areas with flat terrain. A one-meter error in the DEM data may be sufficient to route water in a different direction in flat areas, whereas a similar one-meter error in a region of steep slopes has minimal effect on the direction of runoff. Since runoff volume is sensitive to the contributing area, errors in basin delineation cause errors in runoff estimation. Table 13.8 shows that basin areas for watersheds 3 and 7 are underestimated by the GIS-delineation method.
Limitation of automatic watershed delineation tools in coastal region
Published in Annals of GIS, 2018
Watershed delineation is the perhaps the first and foremost activity in any hydrological analysis. The delineation can be performed manually from the toposheets or automatically using Digital Elevation Model (DEM) data. Automatic watershed delineation has gained huge popularity since the availability of DEM data worldwide. A number of procedures have been developed by multitudinous researchers to delineate stream networks and watershed areas automatically from DEMs. Amongst these, the D8 flow routing method (Mark 1977; OCallaghan and Mark 1984) has found widespread use in flow direction generation (Jenson 1985; Tarboton, Bras, and Rodriguez-Iturbe 1991; Turcotte et al. 2001; Rahman, Arya, and Goel 2010; Mendas 2010; Li et al. 2011; Charrier and Li 2012; Khan et al. 2014; Gopinath, Swetha, and Ashitha 2014)
Development of hybrid wavelet-ANN model for hourly flood stage forecasting
Published in ISH Journal of Hydraulic Engineering, 2018
Ashlin Ann Alexander, Santosh G. Thampi, Chithra N. R.
In order to select the significant preceding hour precipitation values that influence the succeeding water level values, a proper hydrological study of the catchment was carried out to determine the time of concentration of the catchment. HEC-Geo HMS tool in ArcGIS 10.2.1 software was used to perform the hydrological study of the catchment. SRTM DEM (30 m spatial resolution) of the study area was downloaded from the United States Geological Survey (USGS) Earth Explorer website. Using this, terrain pre-processing and basin pre-processing were carried out to derive stream and watershed characteristics and the time of concentration of the catchment. In terrain pre-processing, terrain data were given as input to compute the flow direction, flow accumulation, for stream definition and for watershed delineation. Then the tools in basin pre-processing menu were used to modify the sub basin to meet the study objective. After sub basin delineation, physical characteristics of the basin like the slope, contributing area, length and longest flow path were determined. The hydrological parameter, viz. time of concentration of the catchment was estimated in accordance with the Natural Resources Conservation Service (NRCS) TR55 methodology (USDA 1986). The information required for calculating the travel time like two-year 24 h rainfall amount, slopes and flow distance of precipitation excess on the land surface for three flow regimes (sheet flow, shallow concentrated flow and channel flow) were used to estimate the time of concentration. Based on the estimated time of concentration, the significant preceding precipitation values for estimating succeeding water levels was found out to be four hours.