NATURAL RESOURCE MAPPING USING GIS: COASTAL AND WATERSHED APPLICATIONS

In the riverine environment, a new project, sponsored by the Environmental Protection Agency through NOAA, is underway with the purpose of classifying river reaches of smaller river and streams for oil spill sensitivity. A model, the Reach Sensitivity Index (RSI), has been developed on South Carolina rivers and tested in the Leaf River basin of Mississippi. There are two primary criteria upon which the RSI is based: 1) the containment and recovery of the oil; and 2) the vulnerability and sensitivity of the associated wetlands. These two applications provide users with detailed maps and data tableson natural resources and response information. Current research is underway by many organizations to expand on the ESI principle to develop desktop mapping tools for identifying critical habitats, spill response, and damage assessment. Future goals include applying the ESI concept to the entire coastal zone and the RSI to the entire watershed, for analytical and management purposes. The ESI and RSI are building blocks for comprehensive land use analysis.

INTRODUCTION

THE ENVIRONMENTAL SENSITIVITY INDEX (ESI)

• Southeast Alaska
• Cook Inlet and Kenai Peninsula
• Northern California
• Central California
• Southern California
• Northern Lake Michigan
• Southern Lake Michigan
• Lake Superior
• Lake Huron
• Upper Coast of Texas
• Mississippi
• Alabama
• Delaware, New Jersey, Pennsylvania
• South Carolina
• Florida (consisting of 5 atlases)
• North Carolina

The ESI data are comprised of three general types of information:

• Habitats:Example Classification(APPENDIX A)
  • Intertidal shoreline habitats which are ranked according to a scale relating to sensitivity, natural persistence of oil, and ease of cleanup.
  • Subtidal habitats which are utilized by oil-sensitive species or are themselves sensitive to oil spills, including eelgrass beds, kelp, and coral reefs.
• Human-Use Resources:List of Elements and Sub-elements(APPENDIX B)
  • Specific areas that have added sensitivity and value because of their use by humans, such as beaches, parks and marine sanctuaries, water intakes, and archaeological sites.
• Biological Resources: List of Elements and Sub-elements(APPENDIX C)
  • Including oil-sensitive plants and animals.

The spatial data (coverages) consist of base map, human-use, and biology groupings. The base map coverages include map index polygons (INDEX); hydrology lakes, streams, and coastal shorelines (HYDRO); and ESI shoreline classifications and sensitive habitats (ESI). The human-use coverages are managed lands (MGT) and economic and cultural resources (SOCECON). These coverages contain relational data (stored in the ORACLE table SOCECON) such as the type of feature, the name of the feature, and the geographic and attribute source for the feature. The biological coverages are based on ELEMENT (BIRDS, FISH, HABITATS, M_MAMMALS, NESTS, REPTILES, SHELLFISH, and T_MAMMALS) and also contain relational attribute data.

In ORACLE, both the human-use and biology data reference a SOURCES table which stores metadata (compliant with the Federal Geographic Data Committee's Content Standards for Metadata). Although there may be numerous sources used in compiling these data, we have limited our digital representation to a single geographic source and a single attribute source. In the metadata which accompanies each ESI atlas a complete list of the sources used in generating these data is documented. The biology data are stored in three tables (BIORES, SPECIES, and SEASONALITY).

The BIORES table contains:

• the link to the coverages (RARNUM)
• the species identification number (SPECIES_ID)
• the concentration or abundance (CONC)
• the seasonality identification number (SEASON_ID)
• the geographic and seasonality sources (G_SOURCE and S_SOURCE)
• the biological element (ELEMENT)

The SPECIES table contains:

• the species identification number (SPECIES_ID)
• the species name (NAME)
• the scientific name (GEN_SPEC)
• the two letter State abbreviation (if the species is listed) (STATE)
• the State and/or Federal status (S_F)
• the threatened and/or endangered status (T_E)
• the date the list was published (DATE_PUB)
• the biological element (ELEMENT)
• the biological subelement (SUBELEMENT)

The SEASONALITY table contains:

• the biological element (ELEMENT)
• the species identification number (SPECIES_ID)
• the seasonality identification number (SEASON_ID)
• the monthly presence in January (JAN) through December (DEC)
• the breeding or other important life stages (BREED1 through BREED4)

These relational tables are structured to enable users to query for specific features and also minimize data duplication. From a GIS standpoint, it is important to remember that the coverages are stored by data element and the attributes are stored in their entirety, not by element (e.g., there is no BIRDS table), but by function.

THE REACH SENSITIVITY INDEX (RSI)

The classification of small rivers and streams for oil spill sensitivity involves:

• an evaluation of existing classification schemes
• determining flow characteristics for representative watersheds
• performing a regional assessment of the rivers and streams of the southeastern United States to derive the RSI classification scheme
• evaluating existing national, regional, and state data sources
• classifying stream reaches for the Leaf River, MS watershed
• generating 1:100,000 scale maps showing reach classification, wetlands, threatened and endangered species, managed lands, archaeological and historical sites, water intakes, access and collection points, spill sources, transportation routes, and pipelines

River classification schemes can be grouped into geomorphological, whole river, and longitudinal zonation schemes. The geomorphological classification of streams is based on physical properties such as the slope, bedload/suspended sediment ratio, and discharge. These qualities manifest themselves into braided versus single channels, sinuous versus straight, etc. The whole river concept identifies the water source and the regional physiography. The longitudinal approach identifies a ranking of streams, such as Horton's stream order classification, or the contents of the stream environment such as bottomland hardwoods and the corresponding hydrologic conditions.

To operationalize these classification concepts, a regional assessment of the rivers and streams of the southeastern U.S. was performed from which a classification scheme appropriate for inland areas was proposed. The geographic area of concern was the piedmont and coastal plain region which contains most of the facilities of concern for OPA 90 planning in EPA Region 4. It was decided early in the project that the stream classification system would be developed for normal and seasonally high water levels (annual flooding conditions), and that extreme flood events would not be addressed. The reach classification is based on how the water and oil are expected to behave under normal and annual flood conditions. From this regional analysis, a reach was defined as:

a distinct and uniform characteristic within a stretch of stream which is based on spill response modes and potential ecological and/or socioeconomic impacts from a spill. The boundary of a reach is usually marked by an abrupt change in morphology usually due to stream gradient.

Theorized and field verified (in South Carolina and Mississippi) the RSI classification scheme is:

• 1 -- Navigable. Quiet water pool with currents (<0.5 knots). Narrow to absent flood plain. Channel sides bedrock, clay, or sandy mud. No associated wetlands. Oil could be recovered from water surface or diverted to and collected against low-sensitive banks.
• 2 --Small, non-navigable. Moderate gradient and currents. Narrow to absent flood plain. Sandy, relatively straight channel. No associated wetlands. Underflow dam could be constructed or oil could be collected against low-sensitive banks.
• 3 -- Navigable. Moderate gradient and currents (<1.5 knots). High, relatively narrow flood plain. Sandy, relatively straight channel. No associated wetlands. Oil could be diverted to and collected against low-sensitive banks, but with greater difficulty than in the case of class 2 because of the larger size of the stream.
• 4 -- Small, non-navigable. Moderate to steep gradient. Fast currents. Narrow to absent flood plain. Some bedrock and rapids in relatively straight channel. No associated wetlands. May be able to construct underflow dam if stream is small enough. Otherwise, oil could not be collected but would be moved quickly through area. Water column impacts likely.
• 5A -- Navigable. Relatively steep gradient and strong currents (>1.5 knots). Narrow or absent flood plain. Some bedrock and rapids in relatively straight channel. No associated wetlands. Because of large size, would be impossible to dam stream. Oil could not be collected, but would be moved quickly through area. Water column impacts would be expected, greater than those in class 4, with significant fish kills possible.
• 5B -- Small, non-navigable. Moderate gradient and currents (<1.5 knots). Relatively high and narrow flood plain. Sandy, sinuous channel. Upper bottomland hardwood forests on associated flood plain that are periodically flooded, but not highly vulnerable to oiling. If oiled, less sensitive than lower elevation wetlands. May be able to construct underflow dam if stream is small enough. In places, may be possible to collect oil against low-sensitive banks.
• 6 --Navigable. Moderate gradient and currents (<1.5 knots). Moderately high and wide flood plain. Sandy, sinuous channel. Upper bottomland hardwood forests on associated flood plain that are periodically flooded and are generally more extensive and more vulnerable to oiling than those in class 5B. If oiled, less sensitive than lower elevation wetlands. In places, may be possible to collect oil against low-sensitive banks.
• 7 -- Navigable. Low gradient and variable currents (usually <1.5 knots). Wide and low flood plain. Stream hugs old valley wall with steep banks composed of muddy sediments or bedrock against the wall. Other side of channel has leakage of waters into an associated wide cypress-tupelo swamps, which is highly sensitive and vulnerable to oiling. It should be possible to collect oil against the low-sensitive banks adjacent to the high wall.
• 8 -- Navigable. Low gradient and variable currents (usually <1.5 knots) with flow mostly confined to relatively straight channel with well-defined low banks. Wide and low flood plain. Associated wide cypress-tupelo swamps which are highly sensitive wetlands and are vulnerable to oiling at normal high water. Because of channel confinement of the main flow of the stream, may be possible to direct oil to a collection point further downstream.
• 9 -- Navigable. Low gradient and variable currents (usually <1.5 knots). Wide and low flood plain. Channel meandering with numerous leakage points into adjacent ox-bows and associated wide cypress-tupelo swamps which are highly sensitive vulnerable to oiling at all times except extreme low water. Points of leakage into adjacent swamp would be difficult to close. It may be possible to direct oil to collection points in quiet water ox-bows, but recovery and storage would be very difficult.
• 10A -- Navigable. Low gradient and variable currents (usually <1.5 knots). Wide and low flood plain. Main channel highly sinuous and, in places, anabranching with abundant leakage points and bifurcations into adjacent wide cypress-tupelo swamps which are highly sensitive and vulnerable to oiling at all times except extreme low water. Virtually impossible to close all points of leakage. Potential collection points almost nonexistent.
• 10B -- Small, non-navigable. Low gradient and variable currents (usually <1.5 knots). Moderately wide and low flood plain that contains extensive cypress-tupelo swamps. Small channels highly anabranched and full of windfall logs and limbs. Leakage common all along the anastomosing channels into the adjacent swamp. The highly bifurcated stream flows into highly sensitive wetlands that are present along all channels. Because it is not navigable, standard collection and containment techniques untenable. Potential area of wetland impact less than in class 10A.

To visualize the characteristics of the RSI classification scheme plan and cross-section views were drawn and pictures (RSI = 9B) were taken at all field sites. The differences, and levels of sensitivity, are dramatically apparent when field work is performed which is based on well-known, theoretical, models.

Once the proposed classification scheme was approved, the next step was to apply it to the Leaf River watershed in Mississippi. The project is currently at the digitization stage and will be completed in August, 1996, with the production of sensitivity maps at a scale of 1:100,000, which will include the reach sensitivity classification, sensitive biological and human-use resources, potential spill sources, and access and collection points for response operations.

CONCLUSIONS

• involvement from local experts who are intimately knowledgeable of specific resources
• incorporation of existing data in the form of maps, tables, and digital geospatial data
• a coupling of multidisciplinary scientists with GIS experts to design and develop workable projects

The Environmental Sensitivity Index and Reach Sensitivity Index are examples of applications which have been developed for a specific application (oil spills), but can be used for other applications as well. The classification schemes are both based on similar ranking scales, they use similar topological data structures, and the geographic base are national data sets (example). The ESI shorelines are based on 1:24,000 USGS 7.5 minute quadrangles and the RSI stream networks are based on 1:100,000 EPA Reach Files.

In the coastal environment, the ESI data may be used to develop applications such as analyzing impacts due to sea level rise, emergency preparedness, beach front management, and other natural resource management and protection. In the riverine environment, the RSI data may be used to develop watershed management applications such as non-point pollution models. Together, the RSI may be combined with the ESI to develop an application which investigates the natural progression from riverine to marine environments and the effects of various land use practices and urbanization.

APPENDICES

APPENDIX A

• 1. Exposed Vertical Sea walls
• 2. Exposed Scarps in Clay
• 3A. Fine-Grained Sand Beaches
• 3B. Scarps and Steep Slopes in Sand
• 4. Medium- to Coarse-Grained Sand Beaches
• 5. Mixed Sand and Gravel (Shell) Beaches
• 6A. Gravel (Shell) Beaches
• 6B. Exposed Riprap
• 7. Exposed Tidal Flats
• 8A. Sheltered Man-Made Structures
• 8B. Sheltered Scarps in Marsh/Mud
• 9. Sheltered Tidal Flats/Oyster Beds
• 10A. Salt and Brackish-Water Marshes
• 10B. Freshwater Marshes
• 10C. Freshwater Swamps
• 10D. Mangroves

APPENDIX B

• Recreation:
  • Beach
  • Boat Ramp
  • Boating/Fishing
  • Diving Area
  • Marina
  • Park
• Management Area:
  • Marine Sanctuary
  • National Park
  • Preserve/Reserve
  • Refuge
  • Wildlife Management Area
• Resource Extraction:
  • Aquaculture Site
  • Commercial Fishery
  • Log Storage Area
  • Mining
  • Subsistence
  • Water Intake
• Cultural:
  • Archaeological Site
  • Historical Site
  • Indian Reservation

APPENDIX C

• Marine Mammal:
  • Dolphin
  • Manatee
  • Polar Bear
  • Sea Lion
  • Sea Otter
  • Seal
  • Walrus
  • Whale
• Terrestrial Mammal:
  • Bear
  • Deer
  • Mustelid (small mammal)
  • Rodent
• Bird:
  • Alcid
  • Diving Coastal Bird
  • Gull/Tern
  • Petrel/Fulmer
  • Raptor
  • Shorebird
  • Wading Bird
  • Waterfowl
• Fish:
  • Anadromous Fish
  • Beach Spawner
  • Kelp Spawner
  • Nursery Area
  • Special Concentration
• Shellfish:
  • Abalone
  • Clam
  • Conch/Whelk
  • Crab
  • Lobster
  • Mussel
  • Oyster
  • Scallop
  • Shrimp
  • Squid/Octopus
  • Echinoderm
• Reptile/Amphibian:
  • Alligator
  • Turtle