Final Project Report – Identifying Flood
Prone Areas on
The winter after Hurricane Ivan, the City of Gulf Shores experienced unusually heavy rainfall totals. Groundwater levels were abnormally high when two spring storm fronts (referred to as the April 2005 storms) one week apart caused localized flooding in areas that were not designated as a flood zones (with some exceptions) and had not flooded from previous large storms or hurricanes. These areas are mostly above +10 elevation and are not normally affected by hurricane storm surges. They normally percolate water into the ground without problems, but these storms revealed that they can flood when there is a combination of already high groundwater followed by heavy storms or series of storms. The most recent FEMA Flood maps do not show these areas as being in a high risk flood area except for one low area which is subject to storm surge in a major hurricane. One subdivision in the area had several houses flood and the streets remained flooded for more than a week.
Aerial Photograph of Study Area with property boundaries (Aerial without property boundaries, ProjLoc2)
Locate local depressions which are subject to flooding in these types of events and locate possible drainage projects to minimize flooding in these areas in the future. Determine partially developed or undeveloped areas that have flooded in the past without affecting anyone so that the problems can be resolved when they are developed. Identify areas that may flood harmlessly but when full may overflow onto a road or other area can locate possible drainage projects to allow these areas to drain without adversely affecting other areas. Identify easements or parcels that may need to be obtained to provide drainageways for these areas. Note in the photograph above the wide (330') Department of Conservation property which surrounds the easement for the State Highway Department's right-of-way for Fort Morgan Road. This undeveloped strip has served in many locations as a large percolation bed for developed areas on each side of Fort Morgan Road.
The map above is a categorized color scheme map of the Triangular Irregular Network (TIN) created from Baldwin County's LIDAR data. The study area above is approximately 3.25 miles by 1.4 miles and was clipped from a larger TIN of the Gulf Shores area using Edit TIN of the TIN Creation tools in 3D Analyst Tools, the default color range of this categorized color scheme elevation map helps gives a good feel for the landform of the area. Some of the depressions in the area can be spotted from this map alone. I created several raster maps from this TIN using the 3D Analyst Tools extension TIN to Raster, found in the Conversion category. The two rasters I created were rasters with 20' wide and 50' wide cells.
A hillshade map of the study area created from the 20' wide cell raster created using the Hillshade extension of the Raster Surface category of 3D Analyst Tools. The hillshade view gives a good visual impression of the landforms. Note that Fort Morgan Road can be seen fairly clearly throughout the study area. The actual area of interest in the study area is the area north of Fort Morgan Road and south of the wetlands surrounding Oyster Bay to the north. The boundary between the uplands and the lower wetlands to surrounding Oyster Bay can clearly be seen on this map. (The four distinct square areas in the center of the map are sanitary sewage lagoons.) (The areas south of Fort Morgan Road drain without any real problems to Little Lagoon to the south.)
Using the Contour extension under the Raster Surface category of 3D Analyst Tools, I created a contour map to attempt to define the boundaries of the depressions using the 20' wide cell raster. Small variations in the nearly flat depressions created a large number of small area circular or oddly shaped areas that are distractions from determining the actual landforms. I had hoped to use selected contours to represent the boundaries of the depressions, but the large number of contours created in what are actually fairly flat areas made the contours hard to use. The absence of contours best represents flat areas, not a multitude of small circular lines in close proximity to each other that all representing the same elevation. I did use a categorized color scheme for the contour lines to help identify areas of similar elevation, which did help reveal the major landforms, but this view was not any more useful than the cloropleth TIN map. The problem with the excess topographic lines is one of the possible projects I had considered initially, since the same problems exist with the City's topographic shapefiles. A way of smoothing the landforms, especially in areas with very low slopes would help in creating smaller, more user friendly topographic data.
The Hydrology category of the Spatial Analyst Tools contained a number of useful tools in locating the areas of interest in the study area. After creaton of the rasters, the Flow Direction extension calculates the direction in which each cell will flow. The Sink extension uses the raster it creates to find local depressions. The map above shows the results of this extension overlaid on the 50' wide cell raster grid. The white areas are the highest elevations and the black are the lowest. The multi colored points are the areas calculated to be sinks. (Ignore the colors, they were color categorized on a meaningless field so that they would show up better on top of the black and white background.) A very large number of local sinks were found, making analysis difficult. I did try the Focal extension in the Neighborhood category of the Spatial Analyst Toolsto smooth the rasters, and while it did decrease the number of by about 50%, it was just as effective to use larger cell grids to smooth out the flat areas. Although there are a large number of local sinks shown above, when you ignore the local sinks in the darkest areas (which are mostly wetland areas) and focus on the pattern of local sinks in the light gray and dark gray areas between Fort Morgan Road and the wetlands to the north, they do help to identify the areas of interest.
The Watershed extension in the Hydrology category of Spatial Analyst Tools calculates the boundaries of watersheds in a raster after the Flow Direction and Flow Accumulation extensions are used. The first extension calculates the direction in which each cell will flow and the second calculates the sum of the cells that contribute to that flow. The extension can also utilize "pour points" to inject flows of a specified amount in one or more locations, however, the default mode was more applicable to this study. The map above is color categorized into 15 different watersheds from the highest to lowest accumulation. The colors were chosen more to distinguish the various areas than to show the flow direction. Careful comparison of this map to the previous map above helped to identify the areas of interest.
Areas of Interest
Using the two maps above, I was able to identify the following areas of interest in the study area.
Outfall A – This improvement already exists, it is a newly constructed drainage outfall south of
Outfall B – Existing drainage outfall ditch (appears to be adequate as is)
Outfall C – Newly purchased City property, for future public boat launch and park – Construct new drainage outlet, piped or open ditch to provide new outfall for runoff from Area 12
Outfall D – Existing outfall ditch to Little Lagoon and one small 18" pipe crossing the road. – Upsize pipes under Fort Morgan Road and improve outfall ditch to handle runoff from Area 10.
Outfall E – New drainage easement acquired from proposed future development – Create open ditch, construct new drainage pipes under highway, and construct east-west drainage ditches on the north side of the highway to collect water from Areas 4-6 (and possibly 7?). The new east-west ditch on the north side of Fort Morgan Road should be a broad mildly graded swale to take advantage of the highly percolating sandy soils in this area.
Outfall F (Not Shown) – Existing outfall ditch to the west of the study area. The drainage improvements to the subdivision in Area 1 consisting of new piping to this ditch. Note: The western edge of the study area is a Planned Unit Development (PUD) annexed into the City consisting of approximately 100 acres of residental subdivisions around a golf course. Almost 40 acres surrounding the golf course are wetlands. The drainage system for the area was designed to utilize the golf course lakes for stormwater storage which would overflow into the wetlands before eventually draining into Mobile Bay. This golf course PUD area did not experience flooding during the April storms.
This analysis is more of a qualitative analysis of
landforms to identify potential problem areas than a quantitative analysis than
an engineer might perform with specialized hydrology software which can
calculate stormwater flows for many different frequency storm events. Reviewing the Soil Conservation Service data
on soils types for the study area showed that the uplands area of the study
area to be almost all
The most useful extension that would help this study would be a tool that smoothed out variations in areas with very low average slopes. This would also help in the creation of more user friendly contour maps without averaging out real variations of grade in the areas where the average slopes are higher. A modification to the Focal extension could accomplish this or a modification interpolating extension such as Kriging or Inverse Distance Weighted might be used. It is important to note that the data is only considered accurate to +/- 6”, so smoothing these small variations in flat areas would not degrade the real accuracy of the data. Specialized hydrology software could utilize the existing TIN data without problems since the volume of these small depressions is too low to have much effect on their calculations.