CEE6440 GIS in Water
Resources Term paper
Floodplain
Management using ArcGIS and HEC-RAS
Jongho Keum
Abstract
Rain and snowmelt cause rivers and lakes to raise water surface and even overflow their banks and damage to adjacent land known as floodplain. Therefore, floodplain is very important in river management because it is a guideline where the flood will extend to and which properties are likely to be damaged. In this study, geographic information system (GIS) is coupled with one-dimensional hydraulic model, the U. S. Army Corps of Engineers Hydrologic Engineering Center¡¯s River Analysis System (HEC-RAS), in order to develop floodplain maps of a part of Logan River in the City of Logan, Utah. As input, this paper requires a digital elevation model (DEM), streamline, and streamflow for HEC-RAS simulation. The process begins with export channel cross sectional geometry to HEC-RAS followed by simulation of water surface elevation using HEC-RAS. After that, floodplains are mapped with the results of calculated water surface for each flood and each cross section. Accordingly, floodplain management methods are presented and recommended for urban, non-residential, and natural area.
1.
Introduction
Floodplain is one of the unique properties of river. Although two rivers and their larger tributaries share a number of common features and glacial melt water drainage history, significant differences in river discharge and slope, floodplain width, and sediment load strongly affect flood response (SAST, 1994). Therefore, floodplain has to be identified for each river and each flood as well.
Accurate and current floodplain maps can be the most valuable tools for avoiding severe social and economic losses from floods. Accurately updated floodplain maps also improve public safety. Early identification of flood-prone properties during emergencies allows public safety organizations to establish warning and evacuation priorities. Armed with definitive information, government agencies can initiate corrective and remedial efforts before disaster strikes (Chapman and Canaan, 2001).
Computer software packages have played an important role in water resources engineering. With development of flow analysis software, U.S. Army Corps of Engineers Hydrologic Engineering Center River Analysis System (HEC-RAS) hydraulic model, it is possible to perform one-dimensional river flow analysis easier. And, geographic information systems (GIS) is developed to analyze spatial and temporal attributes and to measure values which users are interested in by associating various geographical attributes.
The objective of this study is focused on assessment and prediction of the extent of given floods with integration one-dimensional hydraulic model with geographic terrain map. And, this can be a decision maker to establish a floodplain management.
Logan River basin located in the City of Logan, Utah shown in Figure 1 is selected as the study area of this term project. The Logan River drains eastern part of the Middle Bear-Logan watershed and originates as a high mountain stream in the Bear River Range in Idaho (Bear River Watershed Information System, 2009). The Logan River watershed area is approximately 1,500 km2. Flow pattern shall be changed because of dam structure. In this study, floodplain is mapped for urban area of the City of Logan, for specific, from First Dam to Logan River Golf Course. The main land use of study range near Logan River is residential area.
Figure 1. Study area, Logan
River, UT
2.
Software and Data Requirements
2.1
Water Surface Elevation Calculation
In this study, water surface elevation is calculated using Hydrologic Engineering Center River Analysis System (HEC-RAS) developed by the U.S. Army Corps of Engineers. HEC-RAS is a successor developed under the Microsoft Windows operating system to HEC-2 which was the standard stream hydraulic analysis software since it was released in 1964. HEC-RAS is capable of calculating one-dimensional water surface elevation for steady gradually varied flow in natural or artificial channels. The one-dimension means that HEC-RAS does not consider lateral or vertical variation.
The calculation method for gradually varied flow can be represented as graphical-integration method, direct-integration method, and step method. And, step method is classified into direct step method and standard step method. The standard step method can calculate water surface of various sections such as natural channel cross sections, while the direct step method can calculate that of uniform cross section only. Accordingly, HEC-RAS calculates water surface profile by standard step method.
|
|
|
Figure 2. Channel sections
for standard step method |
Energy balance equation between two cross sections for standard step method is as following equation 1 and 2.
where,
V :
average velocity,
g :
gravitational acceleration,
L :
discharge weighted reach length,
C : expansion or contraction loss coefficient
The unknown water surface elevation at a cross section is determined by an iterative solution of equation 1 and 2. The computation procedure is as follows (USACE, 2008).
¡¤ Assume a water surface elevation at the upstream cross section.
¡¤ Based on the assumed water surface elevation, determine the corresponding total conveyance and velocity head.
¡¤
Compute
¡¤ Solve equation 1 for WS.
¡¤
Compare the computed value of
2.2
Geographic Information System
Geographic information systems is defined broadly as a computer system for collecting, processing, integrating, and analyzing information related to some portion of the earth (Rhind, 1988). A narrow definition of GIS is a systematic approach to collecting, storing, manipulating, analyzing and displaying geographically referenced information, using a combination of computer hardware, software, personnel, and organizational procedures. The most popular GIS software package throughout the world is the ArcGIS system which is developed by ESRI in Redlands, California. ArcGIS is the integrated system including previous version of ArcInfo and ArcView. ArcInfo is a coverage model while ArcView is a shape file model.
GIS represents real world with these types of digital data : vector, raster. Vector data are defined spatially with point which is a pair of x and y coordinates, line which is a sequence of points, and polygon which is a closed set of lines. But, raster data are described by a cell grid so that one cell has one value for given attribute. So, point, line, and polygon of vector data can be represent by one cell, linear consecutive cells, and zone of cells, respectively.
In addition to vector or raster dataset, triangulated irregular network (TIN) is used to visualize geography in GIS software. The TIN format is efficient to store data because the resolution adjusts to the parameter spatial variability and triangle is the only polygon that makes a plane. For these reasons, TIN is used for analysis of land surface terrain including three-dimensional analysis.
2.3
HEC-GeoRAS
HEC-GeoRAS is a GIS tool for support of HEC-RAS using ArcGIS, i.e. an ArcGIS extension package developed by the Environmental Systems Research Institute, Inc. (ESRI) and the U.S. Army Corps of Engineers (USACE) Hydrologic Engineering Center (HEC). The main object of HEC-GeoRAS is to create an HEC-RAS import file which contains geometric data from digital terrain model and analyze spatially terrain with results exported from HEC-RAS.
The processes to integrate GIS with HEC-RAS using HEC-GeoRAS can be presented as Pre-RAS process, HEC-RAS process, and Post-RAS process. HEC-RAS process represents obviously water surface calculation. Pre-RAS means a process which should be conducted before HEC-RAS calculation. That is, channel geometry is exported for HEC-RAS. Post-RAS means a process which should be conducted after HEC-RAS calculation, i.e. results which shows water surface profile for each given floods is imported to GIS and spatially visualized.
3.
Application
3.1
Pre-RAS
United States Geological Survey (USGS) serves digital elevation model of the United States in resolution of 1 arc-second, 1/3 arc-second, and 1/9 arc-second, but 1/9 arc-second map has been developed only a few part. Because high resolution data can define geometry more specific, 1/3 arc-second DEM is used for building channel cross-sections.
HEC-GeoRAS prefer TIN for analyzing geometry. 1/3 arc-second DEM from USGS seamless server is converted to TIN using 3D Analyst Tools in ArcGIS. Cross-sections for calculation of water surface, flowpath, and banks are created using converted TIN data (Figure 3).
Figure 3.
Pre-RAS process
|
|
|
|
Figure 4.
Sample cross-sections exported from ArcGIS |
|
4.2
HEC-RAS
The nearest streamflow gage station is the site name of Logan River above State Dam, Near Logan, UT and its site number is USGS01019000. For statistical analysis, 56 years of daily streamflow data from 10/01/1953 to 11/22/2009 was collected from USGS web server. Weibull plotting position (equation 3) is used to calculate probability of streamflow.
The
historical maximum streamflow and 25%, 50%, and 75% quartile of streamflow is
selected for floodplain calculation, and, in return period terms, the quartiles
are 4-year, 2-year, and 1.33-year frequency, respectively. The historical maximum streamflow was
52.593
|
|
|
|
|
|
|
Figure 5.
Water surface calculation results for historical maximum flood |
|
4.2
Post-RAS
The final process of floodplain mapping is Post-RAS process of which primary goal is to delineate the floodplains. Exported file from HEC-RAS contains water surface profiles for each flood along the distance from streamline to both left and right floodplain boundaries. And, this file is imported to ArcGIS in order to map the floodplain. Figure 6 shows created flooding extent and depth. The blue cell is for the historical maximum streamflow and the red cell is for 25% quartile streamflow. The darker cells represent the deeper inundation depth and vice versa.
Figure 6.
Floodplain visualization (2-D view)
Figure 6 above shows just planimetrical schematic using DEM, the basemap. It is different from three-dimensional view which uses TIN as a data model. By definition of floodplain, when water surface elevation of a cell exceeds digital elevation model, then the cell is under flood and the elevation gap between two parameters would be the flooding depth. However, the three-dimensional floodplain visualization is sometimes useful because two-dimensional view in Figure 6 could not show the actual landscape. Therefore, three-dimensional floodplains are also visualized using ArcScene which serves three-dimensional view with TIN data.
|
(a) historical maximum streamflow |
(b) 25% quartile streamflow |
|
(c) 50% quartile streamflow |
(d) 75% quartile streamflow |
|
Figure 7. Floodplain visualization (3-D
view) |
|
As a result of floodplain mapping, lower basin near Logan River Golf Course is the most flood-prone area. And, some urban area, especially over left bank of Logan River, is also vulnerable. Therefore, floodplain management is needed for these areas.
5.
Floodplain Management
The common framework for dealing with floods comprises of four general types of activities: modifying flooding, modifying susceptibility to flooding, modifying the impacts of flooding, and preserving the natural and beneficial functions of floodplains (Simonovic, 2001). The first activity, modifying flooding is related to structural floodplain management, in other words, the concept is let water do not come to people and their properties. Next, modifying susceptibility to flooding means non-structural flooding mitigation activities which is let people do not live or construct buildings very close to river. Third, modifying the impacts of flooding is an activity which can spread the damage instead of reducing the amount of damage caused by flooding. Last activity can be said no action, so that the river nature and benefits of floodplains can be preserved.
For Logan River, the lower part after the First Dam, the study area of this paper is mainly residential area with some non-residential area such as golf course and park. And, the upper part above the First Dam is undeveloped natural area. Economic analysis should be done for suggestion of floodplain management because costs and benefits of each method can be numerical through economic analysis. The economic analysis consists of damage costs, benefits from construction of flood mitigation facilities, and moving costs to avoid flooding.
However, in this study, guidelines for decision making of floodplain management are only under consideration, so economic analysis does not conducted due to the limited time and information for this term project. Generally, structural floodplain management can be applied to the most of urban area when the cost of flood damage exceeds the construction anti-flooding facilities or moving costs including real properties such as land and house cost but construction is cheaper than moving. Non-structural floodplain management can be applied to the non-residential area when the cost of flood damage does not exceed flood preventing costs. And, for the undeveloped area, to keep the original floodplain is the best way to avoid flooding damage so that river environment can be preserved as the original state.
6. Conclusions and Discussions
Floodplain evaluation and delineation is very important to floodplain management, but floodplain varies with each flood. Developments of computer software such as ArcGIS and HEC-RAS save a lot of time and resources. In this study, floodplains of Logan River were delineated to recommend proper floodplain management method. Lower part of Logan River in the study area is determined as the most flood-prone area which is consists of both urban and non-residential area. Structural floodplain management is suggested to urban area, and non-structural management is suggested to non-residential area. And, preservation of original state of river is recommended for a section which river environment is still undeveloped.
This study uses the finest available DEM of 1/3 arc-second which is about 10m grid. The resolution is too rough to delineate floodplain accurately. Because 1/9 arc-second, 3.3m grid cell, DEM is under development, the procedure applied in this study will give more accurate mapping with high-resolution DEM. In addition, economic analysis for decide floodplain management is neglected in this study. Therefore, high-resolution DEM and economic analysis make the decision of floodplain management more reasonable.
References
¡¤ Bear River Watershed Information System. ¡°http://www.bearriverinfo.org¡±. (visited at Nov. 23 2009).
¡¤ Chapman, J. B. and W.D. Canaan (2001). ¡°Flood Maps are Key to Better Flood Damage Control.¡± CE News, March 2001, p.58-60
¡¤ Chuntian, C, Chau, K.W., Chunping, O. (2002) . ¡°Flood Control Management System for Reservoirs as Non-structural Measures.¡± International Workshop on Non-structural measures for water management problems, 18-20 October 2001. United Nations Educational, Scientific and Cultural Organization (UNESCO). p.108-119.
¡¤ National Weather Service (NWS). ¡°Test Basins¡± 12 Oct 2009 <http:// http://www.nws.noaa.gov/oh/hrl/dmip/test_basins.html>.
¡¤ Olsen, J. Rolf (2006). ¡°Climate Change and Floodplain Management in the United States.¡± Climatic Change. 76(3-4). p.407-426.
¡¤ Rhind, D. W. (1998). A GIS research agenda. International Journal of Geographical Information Systems. 2(1). p.23-28.
¡¤ Scientific Assessment and Strategy Team (SAST) (1994). ¡°A blueprint for Change Part V : Science for Floodplain Management into the 21st Century.¡± Administration Floodplain Management Task Force.
¡¤ Simonovic, S. P. (2002). ¡°Two New Non-Structural Measures for Sustainable Management of Floods.¡± International Workshop on Non-structural measures for water management problems, 18-20 October 2001. United Nations Educational, Scientific and Cultural Organization (UNESCO). p.65-81.
¡¤ Tate, Eric C., et all (2002). ¡°Creating a Terrain Model for Floodplain Mapping.¡± Journal of Hydrologic Engineering. 7(2). p.100-108.
¡¤ U.S. Army Corps of Engineers (USACE) Hydrologic Engineering Center (HEC) (2008). ¡°HEC-RAS Hydraulic Reference Manual version 4.0¡±