Hurricane Planning with Satellite Imagery

• Assessing Potential Inundation From Hurricanes
• Using Satellite Data with a Storm Surge Model
• Making Damage and Recovery Assessments Using Satellite Images

Image rendered by NASA Goddard Laboratory for Atmospheres. (Data from NOAA/AVHRR)

Satellite imagery, used with geographic information systems (GIS) and physical models, can give coastal resource managers and emergency preparedness officials a wealth of hurricane-related information. Prior to a storm, remotely sensed data helps pinpoint where previous hazard events have occurred, where they are likely to occur in the future and the costs associated with historical events. After the storm, remote sensing can be used to determine the extent of landscape change and monitor the progress of recovery. Remote sensing allows a larger land mass to be studied in a shorter amount of time than is possible with traditional ground cover study methods.

Assessing Potential Inundation from Hurricanes

Storm surge (storm tide) is the most dangerous aspect of a hurricane. It is a phenomenon that occurs when the winds and forward motion associated with a hurricane and low barometric pressure pile water up in front of the storm system as it moves toward the shore. Storm surge heights and associated waves are dependent on a combination of factors, including the configuration of the continental shelf (narrow or wide), the depth of the ocean bottom, the intensity of the storm, the storm's forward speed, and the storm's direction of movement.

Storm surge flooding, high winds, and freshwater flooding are the main hazards that render an area unsafe during a hurricane. To assess the vulnerability of an area to hurricane storm surge, the maximum envelope of water must be compared to the elevation of the area to predict whether the area will be flooded during a range of storm scenarios. The storm's angle and direction of attack will be critical for modelling different scenarios. The methodology used to assess hurricane storm surge vulnerability will be briefly explained here.

Surge Inundation Areas

To estimate the extent of flooding that can be expected from a hurricane making landfall along South Carolina’s coast, results from The National Oceanic and Atmospheric Administration's (NOAA) National Weather Service Sea, Lake and Overland Surges from Hurricanes (SLOSH) model were run by the Storm Surge Group at the National Hurricane Center. Three models were used to predict the potential storm surge for the South Carolina coast: the Wilmington/Myrtle Beach basin model, the Charleston basin model, and the Savannah/Hilton Head basin model. The output from these SLOSH models consists of predictions of water heights throughout these basins created by hurricanes making landfall along the South Carolina coast.

U.S. Army Corps of Engineers SLOSH Model Basins for the Southeast

The U.S. Army Corps of Engineers merged the SLOSH model results with digital elevation models (DEM’s) up to the 9 meter (30-foot) contour of South Carolina's coast to create storm surge maps. These maps were developed by scanning U.S. Geological Survey (USGS) quadrangle sheets to create an electronic background map, and by digitizing topographic information from these quad sheets along with supplemental elevation data provided by the U.S. Army Corps of Engineers. By processing the elevation from the base maps, a ground surface model was created and merged with SLOSH model results to create a storm surge map. In areas where the water surface elevation was greater than the terrain elevation, the area was shaded. The resulting maps represent the Maximum of the Maximum (MOM) storm surge composite of hypothetical storms calculated at high tide, one depicting storms with slow forward speeds (5 and 15 miles per hour), and one depicting storms with fast forward speeds (25 and 35 miles per hour.) To see samples of other surge maps, click here.

Example of a Hurricane Surge Map for Horry County

Surge elevation, or water height, is the output of the SLOSH model. At each SLOSH grid point, the water height is the maximum value that was computed at that point. The water height is calculated relative to mean sea level of 1929, also referred to as National Geodetic Vertical Datum (NGVD), and not relative to the ground elevation. Height of water above terrain is not calculated because terrain height varies within a SLOSH grid square. For example, the altitude of a 1-mile grid square may be assigned a value of 1.8 meters (6 feet), but this value represents an average of land heights that may include values ranging from 0.9 to 2.7 meters (3 to 9 feet). In this case, a surge value of 2.5 meters (8 feet) in this square, implying, a 0.7 meters (2 feet) average depth of water over the grid’s terrain, would include some terrain without inundation and other parts with as much as 1.5 meters (5 feet) of overlying water. Therefore, the depth of surge flooding above terrain at a specific site in the grid square is the result of subtracting the actual terrain height from the model-generated storm surge height in that square. It should be noted that, even if the SLOSH model is supplied accurate data, the computed surges may contain errors of 20 percent of observed water levels.

Because tidal anomalies of about +0.3 meter (+1 foot) NGVD before the arrival of a hurricane are not uncommon, the initial water heights used for each SLOSH model run simulate conditions at high tide. In the Wilmington/Myrtle Beach and the Charleston basin models, all SLOSH runs of hypothetical hurricanes were supplied with initial datum of +1.1 meters (+3.5 feet) NGVD (ocean) or +0.9 meters (+3.0 feet) NGVD (bays and lakes.) In the Savannah/Hilton Head basin model, an additional +1.5 meters (+5.0 feet) NGVD were included in SLOSH model runs to simulate conditions at high tide.

Screen grabs of storm surge maps have been provided on this CD-ROM. However, the storm surge data had not been finalized at the time this product went to press. To inquire about data availability contact the U.S. Army Corps of Engineers.

The U.S. Army Corps of Engineers storm surge maps are to be used in combination with other GIS data or satellite imagery to enable managers to plan for major hazards events, and the analysis would only evaluate the vulnerability of an area to hurricane storm surge. This type of analysis would not evaluate the vulnerability of an area to high winds or fresh water flooding. Therefore, the surge maps alone should not be used as a tool for determining which areas are susceptible to hurricane hazards. Emergency management personnel may take factors other than storm surge into consideration when they define evacuation zones. Coastal residents should heed evacuation orders and recommendations of emergency management officials when storms threaten. Emergency management officials use storm surge maps and other resources to help carry out their mission.

Return to Top of Page

Using Satellite Data with a Storm Surge Model

County planners, coastal managers, and local communities can better prepare for the next natural disaster by learning from past experiences. Coastal Change Analysis Program (C-CAP) land cover data, coupled with results from the SLOSH model used within a GIS, can enable emergency management and community planners to better prepare for hurricane impacts on their region. Estimates of the particular land cover classes that may be inundated by each category hurricane can enable planners to better assess their region's risk and vulnerability. With this type of information, planners are better able to prioritize and target mitigation and preparedness activities for their area.

In this example, GIS data layers derived from the SLOSH model were overlaid on C-CAP land cover data for the Berkeley-Dorchester-Charleston tri-county area to assess potential land cover impacts from different category hurricanes. Each category of hurricane has a different inundation profile, with stronger hurricanes on average inundating more land than the less powerful storms. These inundation areas were derived from the SLOSH model and converted into GIS data layers. Using the 1995 C-CAP land cover data as the base, each of these inundation data layers was used to identify the amount and types of land cover expected to be inundated by each category storm. This type of analysis provides information on not only the location and size of inundation areas, but also the pervasive land cover types within these areas.

The table below shows the acreage of land cover potentially affected by each of the five hurricane storm categories. By analyzing the data, patterns begin to emerge that can be useful in many planning and preparedness applications. For instance, the acreage of high- and low-intensity developed land becoming inundated during a storm surge event more than doubles when comparing a Category I to II storm, but only rises slightly for the next three category storms. Conversely, because of its close proximity to the shore, nearly all of the estuarine emergent wetland land cover is inundated in a Category I storm. With this information, planners can begin to pinpoint the location of particular resources at risk from different category hurricanes. If information is available to assign a value-per-acre of a certain resource, estimates of potential damages can be obtained using this methodology.

The integration of satellite imagery, physical models, and GIS provides natural resource managers, community planners, and emergency preparedness officials the ability to better prepare for natural hazard events.

Return to Top of Page

Making Damage and Recovery Assessments Using Satellite Images

GIS and remote sensing technology can be used to monitor landscape change and recovery arising from a natural hazard event. In 1989, Hurricane Hugo, a Category IV hurricane, made landfall in South Carolina. Although Hugo made initial landfall near Charleston, it eventually impacted 24 of South Carolina's 46 counties, causing more than $7 billion in damage. Timber, South Carolina's largest valued agricultural crop, was deeply affected by Hugo. With saturated soils from heavy rains prior to the storm and high winds during the hurricane, thousands of acres of harvestable timber were lost. Using a time series of land cover data, a portion of this loss can be examined.

The C-CAP land cover images above depict the land cover classification for a region near the Cooper River in Berkeley County over a six-year period (1989 to 1995). The pre-Hugo land cover data show that the majority of the natural habitat in this area in 1989 was forest, which is indicated by the dark green color. Following Hurricane Hugo's landfall in September 1989, the majority of the landscape was severely altered with much of the forest changing to scrub/shrub habitat. You can see this change displayed in the 1990 post-Hugo land cover image by noting the predominance of olive green color in the image. The 1995 C-CAP land cover data, taken six years after Hugo, reveal little change from the 1990 land cover data. This indicates that the majority of the original forested land has yet to recover from the damage caused by Hugo. The dark gray and black colors in the 1995 land cover image show that the majority of the landscape remains scrub/shrub.