![[logo] US EPA](../gif/logo_epaseal.gif){kind=logo}
Specific Design Information - Illustrative Examples
This section provides detailed information on creating GRTS Designs. In addition, a presentation; Generalized Random Tessellation Stratified (GRTS) Spatially-Balanced Survey Designs for Aquatic Resources, by A.R. Olsen is available as a pdf file (5.6 MB). The presentation includes information on sample frames, GRTS theory and GRTS implementation. Software packages required to create a GRTS design include: psurvey.design, R statistical package, and ArcGis.
The survey design is a plan for selecting the sample appropriately so that it provides valid data for developing accurate estimates for the entire population or area of interest. At this point in the design process it is assumed that the folloing have been completed:
- Established the monitoring objectives
- Identified Resource Characteristics
- Identified Institutional constraints requiring consideration
- Defined the target population
- Specified Subpopulations, Domains and Stratification
- Created a sampling frame for all units of the target population from which to select the sample
- Selected a GRTS survey design for the selection of sample of units from this frame
- Selected a strategy for allocation of sampling efforts in space and time
- Established sample size(s)
Additional information on each of these items is available on Design Specifics
Information is presented for several GRTS designs; examples for streams, estuaries, lakes and ecoregions
Streams
This example is based on a small watershed in Western Oregon, the Luckiamute Watershed. The target population is perennial and intermittent streams/rivers within the watershed boundaries, [streams]. Multi-density categories are based on Strahler stream orders. Frame materials consist of GIS stream network coverage from PNW portion of RF3. Four GRTS designs are included as examples of 1) equal probability selection [equal_strms], 2) stratification by perennial/intermittent, with equal probability selection within strata [strat_strms], 3) stratification by perennial/intermittent, with unequal probability of selection by Strahler category, plus an over sample [unequal_strms], 4) panels over time sites, with stratification by perennial/intermittent, unequal probability of selection by Strahler category, plus a perennial stream over sample [panel_strms].
Design File Documentation for this example is available on design_streams and for downloading (pdf ). Input files, psurvey.design package, and output files are available [design_files]. The R software package psurvey.design is available for downloading. A published map document is available for downloading, requires ArcReader to view, explore and print (a free software package from ESRI).
Estuaries
This example is based on small estuaries along the Oregon Coast. This example illustrates an area sample for an estuarine resource. The target population is Oregon State estuaries less than 100 km2. Multi-density categories are based on estuarine area: <1 km2, 1-10 km2, 10-100 km2. A GRTS design is included that has unequal probabilities of site selection, two panels (total 50 sites), and an over sample of 50 sites. Maps of the sites are: Panel Sites, Over Sample Sites
Design File Documentation for this example is available on design_est and for downloading (pdf ). Input files, psurvey.design package, and output files are available [design_files]. The R software package psurvey.design is available for downloading. A published map document is available for downloading, requires ArcReader to view, explore and print (a free software package from ESRI).
Lakes
This example is based on lakes and reservoirs in EPA Region 1, New England States: CT, MA, NH, RI, and VT. Lake locations are given by their centroid. Data originally from National Hydrology Database [lakes]. The target population is lakes within the 5 state boundaries, NE Lakes. Four GRTS designs are included as examples of; 1) equal probability selection all lakes and reservoirs (n=300) [equal_lakes], 2) stratification by State, with equal probability selection within state [state_map], 3) unequal probability of selection based on lake area (n=300) , plus an over sample (n=200) [unequal_area], 4) six panels over time, with unequal probability based on lake area (n=50), plus an over sample (n=120) [panel_lakes]. Multi-density categories are based on lake surface area: 0-1 ha, 1-5 ha, 5-10 ha, 10-50 ha, 50-500 ha, and 500-70,000 ha.
Design File Documentation for this example is available on design_lake and for downloading (pdf ). Input files, psurvey.design package, and output files are available [design_files]. The R software package psurvey.design is available for downloading. A published map document is available for downloading, requires ArcReader to view, explore and print (a free software package from ESRI).
Ecoregions
This example is based on Omernik level 3 ecoregions within Utah. This example illustrates an area survey design. The target population is the area within Utah State [State Map]. Four example GRTS design are included; 1) equal probability selection (n=115) [equal_eco], 2) stratification by ecoregion, with equal probability selection within state (n=115) [strat_eco], 3) unequal probability of selection by ecoregions (n=115) , plus an over sample (n=200) [unequal_eco], 4) five panels over time, with unequal probability based on ecoregion area (n=50), plus an over sample (n=100) [panel_eco]. Multi-density categories are based on ecoregion area: Northern Basin and Range, Wyoming Basin, Central Basin and Range, Wasatch and Uinta Mountains, Colorado Plateaus, Southern Rockies, Mojave Basin and Range.
Design File Documentation for this example is available on design_eco and for downloading (pdf ). Input files, psurvey.design package, and output files are available [design_files]. The R software package psurvey.design is available for downloading. A published map document is available for downloading, requires ArcReader to view, explore and print (a free software package from ESRI).