INTRODUCTION Alteration of hydrology has greatly affected the abundance and functioning of wetland systems throughout the United States. California s Central Valley is a classic example of an area affected by severe wetland degradation. U.S. Fish and Wildlife Service calculations suggest that of the original 5 million acres in the Central Valley of California, only 319,000 wetland acres remain (Dahl, 1990), while recent work with satellite imagery, collected from 1986-1989, suggests there may be less than 260,000 acres left (Kempka and Kollash, 1990). The importance of the Central Valley to waterfowl habitat has made this region a priority area for the North American Waterfowl Management Plan (NAWMP). Presently, several federal, state, and private organizations expend extensive resources restoring and managing wetlands to benefit several wetland species and provide buffers to human development. Numerous variables need to be considered when selecting these locations for enhancement to ensure the maximum benefit is achieved. Many of these habitat, hydrologic, and energetic variables are spatial in nature. This project demonstrates how GIS technology can assist in the evaluation of these complex spatial relationships to benefit wetlands conservation efforts. This paper presents the methods used to develop a wetlands inventory for the Central Valley of California and describes how this inventory data is used in two GIS models developed to assist in wetland restoration efforts. Satellite imagery from two different dates were processed along with ancillary data from several sources to produce an up-to-date wetlands inventory. The inventory and other GIS data layers were integrated to generate a GIS computer model to view various scenarios for targeting wetlands restoration/enhancement projects, as well as monitoring the goals for wetlands enhancement efforts throughout specific geographic regions. The Central Valley of California is the inventory project area. The model has been applied to the Sacramento Valley portion of this region. Objectives The goal of this project is to develop a GIS database of wetland and riparian areas for the Central Valley of California and San Francisco Bay. Specific objectives include: 1) develop image processing techniques to identify wetlands using multi-temporal satellite imagery; 2) establish a baseline wetlands inventory for the Central Valley and San Francisco Bay area; and 3) develop a GIS model to track wetland restoration efforts within a North American Waterfowl Management Plan focus area and to identify priority areas for future wetlands restoration. Project Area The project area for the Wetland and Riparian Inventory was defined initially by the boundaries of three TM scenes. The valley was subset out of the three scenes using the 300 foot contour as digitized off 1:250,000 scale USGS topographic maps (Figure 1). In all, the project area includes 9,977,907 acres (15,588 sq. miles) and covers three key regions: 1) Sacramento Valley, 2) San Francisco Bay and the Sacramento and San Joaquin River Deltas, and 3) North San Joaquin Valley. The fourth region, South San Joaquin Valley, will be completed as funding allows. A map of the project area is shown in Figure 1.
Data Sources
1. Landsat Thematic Mapper Imagery - Two seasons of imagery from each
of three TM Path/Row locations were required for this project. First, images
from early-mid summer (6/28/93 and 7/7/93) were needed to identify
wetland emergents during the growing season. Second, scenes from the
previous winter (1/3/93) were required to determine the location and extent
of winter flooding. A cloud-free TM scene for the North San Joaquin area
could not be located; therefore, a SPOT scene from 11/13/90 was used along
with an older TM scene from 11/8/86 for areas not covered by the SPOT
scene. The combination of information from both the summer and winter
classified images allowed for the identification of seasonally (i.e. winter)
flooded wetland and agriculture classes.
2. National Wetlands Inventory - The National Wetlands Inventory is a U.S.
Fish and Wildlife Service mapping program that delineates detailed wetland
classes using manual interpretation of aerial photographs. The most recent
NWI data for most of the Central Valley was derived from aerial
photography collected in the late 1970s and early 1980s. For this project,
the NWI data was used along with additional datasets to stratify out areas
of potential wetlands as a precursor to the image classification.
3. Hydrography - USGS 1:100,000 scale Digital Line Graphs (DLGs) were
modified by CDFG to include River Reach attributes. This River Reach
dataset was used to build a buffer around hydrographic features for
delimiting riparian areas.
4. Agricultural Field Boundaries - A digital coverage of agricultural lands
was acquired from the California Department of Conservation Farmlands
Mapping Program. This data was used along with the NWI data to build a
mask for stratifying the imagery to reduce confusion between spectrally
similar agriculture and wetland classes.
Wetland and Riparian Inventory - Methods
The classification scheme used for this project was modeled in part after the
classification schemes found in "Classification of Wetlands and Deepwater
Habitats of the United States" (Cowardin et al., 1979) and the NOAA C-
CAP protocols (Dobson et al., 1995). A Riparian category was also added
for woody vegetation found in the river flood plains after the classification
system described in Riparian Resources of the Central Valley and
California Desert (Warner and Hendrix, 1985). The classes were organized
under three broad landcover categories: Agriculture, Wetlands, and Uplands.
The summer Landsat Thematic Mapper scenes were initially stratified into
these three general categories using a combination of ancillary datasets and
same-year winter images which were classified to identify winter flooding.
This stratification was performed to reduce confusion between spectrally
similar classes (i.e. rice and emergent wetlands) and to refine the spectral
signatures. The resulting strata were then classified into more detailed
landcover classes, including permanent and seasonally flooded wetlands. A
combination of supervised and unsupervised classification techniques were
used. Field data, aircraft reconnaissance, ancillary data, and 1:40,000 scale
NAPP aerial photographs were used to identify training sites and label
spectral clusters. Post-classification modeling was performed using ancillary
data (NWI and hydrography) to further refine the classification.
Specifically, NWI data was used to label the emergent wetlands as Estuarine
or Palustrine. Non-persistent wetlands were identified using a combination
of summer and winter imagery (i.e. emergent in summer, flooded in winter).
Finally, a three pixel buffer on 1:100,000 scale perennial streams and canals
was used to stratify the woody class into Riparian and Non-riparian.
DU GIS Model
The GIS Model was developed by Ducks Unlimited and Pacific Meridian
Resources under a NASA EOCAP grant to provide assistance in waterfowl
restoration decision-making, evaluate current waterfowl habitat, and provide
an educational tool for illustrating seasonal and between year changes in
waterfowl habitat in the Sacramento Valley. This model utilizes the vector
and raster modeling capabilities of ArcInfo and ArcGrid with a user-friendly
interface developed to guide the user through the specific model
applications. The model consists of four applications: 1) Rice Phenology
Cycle, 2) Change Detection, 3) Waterfowl Energetics Model, and 4)
Wetlands Restoration Site Analysis Model. The latter two will be discussed
here.
Waterfowl Energetics Model
The Energetics Model was developed to assist resource managers in
evaluating how well the Central Valley Habitat Joint Venture wetland
restoration/enhancement goals (CVHJV, 1990) are being met in the
Sacramento Valley under various scenarios. In the model, the acreages of
various waterfowl habitats from the Wetland and Riparian Inventory are
input along with user-defined population and energetics parameters to
determine how well the CVHJV habitat goals are being met. To do this,
equations established by the CVHJV (Heitmeyer, 1989) are used to calculate
the energy requirements of the user-defined target waterfowl population.
Next, the amount of food available to waterfowl is calculated using habitat
acreages from the 1993 Wetland and Riparian Inventory (described above)
along with user-defined assumptions about how much food is available to
and consumed per day by waterfowl in each habitat type (Heitmeyer and
Raveling, 1988; Miller et al., 1989; Heitmeyer et al., 1989). The results of
these calculations are then presented in terms of Target Use Days, Actual
Use Days Available under the given user inputs, and the percentage of
Target Use Days that were supported under the specified scenarios. Figure
2 is a conceptual diagram of the Energetics Model. Figure 3 shows an
example of the model output.
Site Analysis Model
The Site Analysis Model was designed to assist waterfowl resource
managers in evaluating potential locations for restoring wetlands and
targeting agricultural enhancement efforts. Three data layers are considered
in the analysis: 1) Distance to Refuges, 2) Distance to Wetlands, and 3)
Distance to Water Delivery. The features in these three GIS data layers can
be coded by the user in terms of their value in meeting site criteria. The
various GIS data layers can then be combined (i.e. added together) to create
a composite output map showing a site suitability value for each grid cell in
the map. Figure 4 is an example of output from the Site Analysis Model.
CONCLUSIONS Presently, Areas 1 and 2 of the Wetland and Riparian Inventory are completed. The combined use of multi-date satellite imagery and existing digital ancillary datasets has proven a useful technique for producing a baseline inventory. Initial review indicates the Energetics Model is a useful tool for tracking joint venture accomplishments throughout the Central Valley of California. Multiple GIS layers can easily be applied for similar regional landscape analysis to other Joint Venture regions throughout North America. Multiple date analysis of critical agricultural lands, such as rice, can reveal changes in landscape flooding strategies (Spell et al., 1995). Currently, the Energetics Model is being used to help CVHJV reevaluate waterfowl habitat goals. The siting model allows for quantitative assessment of land parcels for locating potential wetland restoration/enhancement projects. The model functions effectively in a GIS environment but additional data layers need to be added to improve its utility for wetlands conservation purposes. Additional variables such as soils, parcel ownership, water districts, and others will be added as funding allows to make the siting model more robust. This model will be reprogrammed to run on a desktop ArcView system to allow greater accessibility by waterfowl biologists, wetland ecologists, etc. REFERENCES Central Valley Habitat Joint Venture. 1990. Central Valley Habitat Joint Venture Implementation Plan. U.S. Fish and Wildlife Service, Portland, OR. 102 p. Cowardin, L.M., V. Carter, F.C. Golet, and E.T. LaRoe. 1979. Classification of Wetlands and Deepwater Habitats of the United States. U.S. Department of Interior, Fish and Wildlife Service, FWS/OBS-79/31. Washington, D.C. Dahl, T.E. 1990. Wetlands Losses in the United States 1780 s to 1980 s. U.S. Department of the Interior, Fish and Wildlife Service, Washington, D.C. 21 pp. Dobson, J.E., E.A. Bright, R.L. Ferguson, D.W. Field, L.L. Wood, K.D. Haddad, H. Iredale III, J.R. Jensen, V.V. Klemas, R.J. Orth, J.P. Thomas. 1995. NOAA Coastal Change Analysis Program (C-CAP): Guidance for Regional Implementation. NOAA Technical Report NMFS 123. Heitmeyer, M.E., D.P. Connelly and R.L. Pederson. 1989. The Central, Imperial, and Coachella Valleys of California. In Smith, L.M., R.L. Pederson and R.M. Kaminski, eds. Habitat Management for Migrating and Wintering Waterfowl in North America. Texas Tech. Univ. Press, Lubbock. pp. 475-505. Heitmeyer, M.E. and D.G. Raveling. 1988. Wintering Resources Use by Three Species of Dabbling Ducks in California. Final Report. Delta Waterfowl and Wetland Research Station, Portage la Prairie, Manitoba. 201 p. Heitmeyer, M.E. 1989. Agriculture/Wildlife Enhancement in California: The Central Valley Habitat Joint Venture. Trans. North American Wildlife and Natural Resources Conference. 54: 391-402. Kempka, Richard G., and R. Peter Kollash. 1990. Comparison of National Wetland Inventory and a Winter Satellite Inventory for the California Central Valley. Yosemite Centennial Symposium Proceedings. February 13- 20. Yosemite, CA. pp. 179-187. Miller, M.R., D.E. Sharp, D.S. Gilmer, and W.R. Mulvaney. 1989. Rice Available to Waterfowl in Harvested Fields in the Sacramento Valley of California. California Department of Fish and Game, 75:113-123. Spell, R., A. Lewis, R. Kempka, and F. Reid. 1995. Evaluation of Winter Flooding of Ricelands in the Central Valley of California Using Satellite Imagery. Proceedings Versatility of Wetlands in the Agricultural Landscape Conference. September 17-20. Tampa, Florida. pp. 357-366. Warner, R.E. and K.M. Hendrix. 1985. Riparian Resources of the Central Valley and California Desert. California Department of Fish and Game, Sacramento, CA.
Richard G. Kempka
Remote Sensing Manager
Pacific Meridian Resources
3074 Gold Canal Drive
Rancho Cordova, CA 95670
Telephone: (916) 852-2000
Fax: (916) 852-2200
Ruth E. Spell, Andrew T. Lewis
Remote Sensing Analysts
Pacific Meridian Resources
3074 Gold Canal Drive
Rancho Cordova, CA 95670
Telephone: (916) 852-2000
Fax: (916) 852-2200
Frederic A. Reid
Regional Biological Supervisor, Western Regional Office
Ducks Unlimited, Inc.
3074 Gold Canal Drive
Rancho Cordova, CA 95670
Telephone: (916) 852-2000
Fax: (916) 852-2200
Scott Flint, Kari Lewis
Land Conservation Planners, Natural Heritage Division
California Department of Fish and Game
1419 Ninth St.
Sacramento, CA 95814