20190911
15184500
1.0
FALSE
MosaicToNewRaster serap_urb2020;app_urb2020;gcpo_urb2020;gcp_urb2020 \\IGSKFAEBFSSDMS\SciDMS\SC00E6E\D2\Working\SLEUTH\20190911\se_sleuth.gdb sesleuth2020 PROJCS['NAD_1983_Albers',GEOGCS['GCS_North_American_1983',DATUM['D_North_American_1983',SPHEROID['GRS_1980',6378137.0,298.257222101]],PRIMEM['Greenwich',0.0],UNIT['Degree',0.0174532925199433]],PROJECTION['Albers'],PARAMETER['False_Easting',0.0],PARAMETER['False_Northing',0.0],PARAMETER['Central_Meridian',-96.0],PARAMETER['Standard_Parallel_1',29.5],PARAMETER['Standard_Parallel_2',45.5],PARAMETER['Latitude_Of_Origin',23.0],UNIT['Meter',1.0]] 16_BIT_SIGNED 60 1 LAST FIRST
CopyMultiple "'C:\Users\558351\Desktop\REPI\REPI_Urbanization\SLEUTH Mosaic Geodatabase.gdb\se_sleuth.gdb\sesleuth2020' RasterDataset" "C:\Users\558351\Desktop\REPI\REPI_Urbanization\New File Geodatabase.gdb" sesleuth2020 "sesleuth2020 RasterDataset sesleuth2020 #"
sesleuth2020
-567291.227000
1892168.773000
51212.078000
2396552.078000
1
file://\\IGSKFAEBFSSDMS\SciDMS\SC00E6E\D2\Working\SLEUTH\20190911\se_sleuth.gdb
Local Area Network
002
Projected
GCS_North_American_1983
Linear Unit: Meter (1.000000)
NAD_1983_Albers
<ProjectedCoordinateSystem xsi:type='typens:ProjectedCoordinateSystem' xmlns:xsi='http://www.w3.org/2001/XMLSchema-instance' xmlns:xs='http://www.w3.org/2001/XMLSchema' xmlns:typens='http://www.esri.com/schemas/ArcGIS/10.7'><WKT>PROJCS["NAD_1983_Albers",GEOGCS["GCS_North_American_1983",DATUM["D_North_American_1983",SPHEROID["GRS_1980",6378137.0,298.257222101]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]],PROJECTION["Albers"],PARAMETER["False_Easting",0.0],PARAMETER["False_Northing",0.0],PARAMETER["Central_Meridian",-96.0],PARAMETER["Standard_Parallel_1",29.5],PARAMETER["Standard_Parallel_2",45.5],PARAMETER["Latitude_Of_Origin",23.0],UNIT["Meter",1.0]]</WKT><XOrigin>-16901100</XOrigin><YOrigin>-6972200</YOrigin><XYScale>266467840.99085236</XYScale><ZOrigin>-100000</ZOrigin><ZScale>10000</ZScale><MOrigin>-100000</MOrigin><MScale>10000</MScale><XYTolerance>0.001</XYTolerance><ZTolerance>0.001</ZTolerance><MTolerance>0.001</MTolerance><HighPrecision>true</HighPrecision></ProjectedCoordinateSystem>
20190913
15590000
20190913
15590000
ISO19139
VAT_sesleuth2020
Table
16
Projected urbanization
Biodiversity and Spatial Information Center
OBJECTID
OBJECTID
OID
4
0
0
Internal feature number.
Esri
Sequential unique whole numbers that are automatically generated.
COUNT
COUNT
Integer
4
0
0
VALUE
Probabilities of urbanization
Biodiversity and Spatial Information Center
200
10 - 20% probability of urbanization
Biodiversity and Spatial Information Center
800
70 - 80% probability of urbanization
Biodiversity and Spatial Information Center
900
80 - 90% probability of urbanization
Biodiversity and Spatial Information Center
975
95 - 97.5% probability of urbanization
Biodiversity and Spatial Information Center
500
40 - 50% probability of urbanization
Biodiversity and Spatial Information Center
25
0 - 2.5% probability of urbanization
Biodiversity and Spatial Information Center
600
50 - 60% probability of urbanization
Biodiversity and Spatial Information Center
1000
97.5 - 100% probability of urbanization
Biodiversity and Spatial Information Center
1
Existing urbanization
Biodiversity and Spatial Information Center
300
20 - 30% probability of urbanization
Biodiversity and Spatial Information Center
400
30 - 40% probability of urbanization
Biodiversity and Spatial Information Center
50
2.5 - 5% probability of urbanization
Biodiversity and Spatial Information Center
950
90 - 95% probability of urbanization
Biodiversity and Spatial Information Center
700
60 - 70% probability of urbanization
Biodiversity and Spatial Information Center
100
5 - 10% probability of urbanization
Biodiversity and Spatial Information Center
VALUE
Integer
4
0
0
311B40E3-E52E-4656-9272-B8D3DFF13078
dataset
Curtis Belyea
Biodiversity and Spatial Information Center, Dept. of Applied Ecology, NCSU
Biologist, GIS and Database Specialist
919-513-7337
919-515-4454
127 David Clark Labs, Dept. of Applied Ecology, NCSU
Raleigh
NC
27695-7617
US
cbelyea@ncsu.edu
20190913
ArcGIS Metadata
1.0
Curtis Belyea
Biodiversity and Spatial Information Center, Dept. of Applied Ecology, NCSU
Biologist, GIS and Database Specialist
919-513-7337
919-515-4454
127 David Clark Labs, Dept. of Applied Ecology, NCSU
Raleigh
NC
27695-7617
US
cbelyea@ncsu.edu
File Geodatabase Raster Dataset
sesleuth2020
2016-05-23T00:00:00
Curtis Belyea
Biodiversity and Spatial Information Center, Dept. of Applied Ecology, NCSU
Biologist, GIS and Database Specialist
919-513-7337
919-515-4454
127 David Clark Labs, Dept. of Applied Ecology, NCSU
Raleigh
NC
27695-7617
US
cbelyea@ncsu.edu
This dataset represents the extent of urbanization (for the year indicated) predicted by the model SLEUTH, developed by Dr. Keith C. Clarke, at the University of California, Santa Barbara, Department of Geography and modified by David I. Donato of the United States Geological Survey (USGS) Eastern Geographic Science Center (EGSC). Further model modification and implementation was performed at the Biodiversity and Spatial Information Center at North Carolina State University.Urban growth probability extents throughout the 21st century were projected for the Southeast Regional Assessment Project (SERAP), which encompasses all or parts of the states of Alabama, Arkansas, Florida, Georgia, Illinois, Indiana, Kentucky, Louisiana, Mississippi, Missouri, North Carolina, Ohio, South Carolina, Tennessee, Virginia and West Virginia. Urban modeling for a large portion of the Appalachian, Gulf Coastal Plains & Ozarks, and Gulf Coast Prairie Landscape Conservation Cooperatives, which encompasses parts of the states of Texas, Oklahoma and Louisiana, Arkansas, Missouri and Illinois. Urban modeling for a large portion of these areas were completed earlier by the Biodiversity and Spatial Information Center as part of the Southeast Regional Assessment Project (SERAP). These data are supplementary to those projections.
Urbanization in the southeastern United States puts natural ecosystems and agro-ecosystems at risk. A picture of what future urbanization might look like, based on recent trends, may be of value to urban, conservation and transportation planners alike.
Biodiversity and Spatial Information Center; Dept of Applied Ecology at North Carolina State University, Raleigh, NC
Curtis Belyea
Biodiversity and Spatial Information Center, Dept. of Applied Ecology, NCSU
Biologist, GIS and Database Specialist
919-513-7337
919-515-4454
127 David Clark Labs, Dept. of Applied Ecology, NCSU
Raleigh
NC
27695-7617
US
cbelyea@ncsu.edu
West Virginia
Arkansas
Georgia
Pennsylvania
Indiana
Florida
North Carolina
Alabama
Mississippi
Kentucky
South Carolina
Tennessee
Ohio
Southeast
Virginia
Louisiana
Oklahoma
Texas
United States
Illinois
Missouri
Growth
SLEUTH
Urbanization
Project Gigalopolis
Prediction
Model
Development
Development
Kentucky
Indiana
Alabama
South Carolina
Georgia
Missouri
West Virginia
SLEUTH
Southeast
Florida
Growth
Virginia
Louisiana
United States
Oklahoma
Project Gigalopolis
Prediction
Tennessee
Urbanization
Model
North Carolina
Illinois
Texas
Mississippi
Arkansas
Ohio
Pennsylvania
This data set is not intended for site-specific analyses. Interpretations derived from its use are suited for regional and planning purposes only. These data are not intended to be used at scales larger than 1:100,000.
60
Version 6.2 (Build 9200) ; Esri ArcGIS 10.7.0.10450
true
-103.143808
-72.611193
21.815928
44.558369
true
-96.607701
-88.421865
29.603247
39.266971
1
-103.143808
-72.611193
44.558369
21.815928
No restrictions.
Prediction is run with the optimum set of coefficient values determined above. Output values in the Color Table were derived in order to capture probability of urbanization with a 95% confidence interval. Other important settings included: Monte Carlo Iterations = 200; Number of working grids = 6; Prediction Stop Date = 2101
Mosaicking was performed with the ArcGIS geoprocessor and the Python programming language via custom written code. For each subunit, the extent and mask environments were set to that of the study area. With these two settings, single output map algebra was used to Set null values (NoData) to 0 as such: gp.singleoutputmapalgebra_sa(“Con(Isnull(inputRaster),0,inputRaster") Cell statisctics was then performed on all subunits in the study area for each year to produce the mosaic raster for the study area for that year as such: gp.cellstatistics_sa(rasterList,studyarea_year,"MAXIMUM")
Once the initial optimal Auxillary Diffusion Multiplier was arrived at, the maximum fractional values for Area, Edges and Clusters were found. This was accomplished running calibration with the initial optimal Auxillary Diffusion Multiplier and the following coefficient values: Diffusion: 100 Breed: 100 Spread: 100 Slope: 0 Road Gravity: 100 After this step was run, the resulting log file was opened in Microsoft Excel, and the highest values for the time periods between input urban data time stamps for Fractional Area, Fractional Edges and Fractional Clusters were recorded. These were used in later calibration steps to evaluate results by weighting error indicated by Fractional Area, Fractional Edges and Fractional Clusters.
After each calibration step performed as described above, all log files produced were concatenated into a single file, which was opened in Microsoft Excel. Each of the Area, Edges and Clusters fractional measures for each time period between input urban data time stamps from the calibration were divided by their respective Maximum Fractional values determined in step 3 and summed. The absolute values of these summations are minimized through sorting, and the coefficient combinations with the lowest absolute weighted fractional error values are used as indicators for the ranges of coefficient values to be used in the next calibration step. It was observed that in initial rounds of calibration the diffusion coefficient among the best performing sets of coefficient combinations was often 1, indicating that lower values would provide better fitting results in calibration. As a result, the Auxillary Diffusion Multiplier was reduced until diffusion coefficient values other than 1 started appearing in the better performing combinations of coefficients. This step was repeated until optimal coefficient values were arrived at. The top performing combination of coefficients was then used in prediction.
After prediction was completed, the output .gif images were converted to geo-referenced .tif files. After geo-referencing was completed, projection information was defined and the .tif files were converted to ESRI format grid datasets and reclassified to remove hillshade values used as backdrop using the ArcGIS geoprocessor and the Python programming language. Values in the raster attribute table were interpreted as follows: 1 - Extent of existing urbanization at the time of the most recent input dataset for the model SLEUTH; 3 - 0-2.5% Probability of urbanization for the projected year; 4 - 2.5-5% Probability of urbanization for the projected year; 5 - 5-10% Probability of urbanization for the projected year; 6 - 10-20% Probability of urbanization for the projected year; 7 - 20-30% Probability of urbanization for the projected year; 8 - 30-40% Probability of urbanization for the projected year; 9 - 40-50% Probability of urbanization for the projected year; 10 - 50-60% Probability of urbanization for the projected year; 11 - 60-70% Probability of urbanization for the projected year; 12 - 70-80% Probability of urbanization for the projected year; 13 - 80-90% Probability of urbanization for the projected year; 14 - 90-95% Probability of urbanization for the projected year; 15 - 95-97.5% Probability of urbanization for the projected year; 16 - 97.5-100% Probability of urbanization for the projected year.
Prior to running the SLEUTH-3r model, the input ESRI format grid datasets were converted to .gif format images in order to conform to the model's requirements. Once input datasets were converted to .gif format, the first step in calibration, finding an initial optimal Auxillary Diffusion Multiplier, was performed. The scenario file for the study area was completed, with the following coefficient values: Diffusion = 100 Breed = 0 Spread = 0 Slope = 100 Road Gravity = 0 After this was run, the resulting log file was opened with Microsoft Excel, and the values for Fractional Area were examined. Ideal values for the Fractional Area were determined to be between +0.15 and 0.30 for the time periods between input urban data time stamps. If the fractional values were too low or too high the Auxillary Diffusion Multiplier was changed, according to their direct relationship, and the step was repeated.
Prior to mosaicking, grids produced from outputs of the SLEUTH model were reclassified once more. Bin values were replaced with upper-limit probability values mulitplied by 10 (to avoid the need for floating point rasters and make interpretation of values more straighforward) as described below: gp.reclassify_sa(dataset, "VALUE", "3 25;4 50;5 100;6 200;7 300;8 400;9 500;10 600;11 700;12 800;13 900;14 950;15 975;16 1000", output_dataset) where the value 3 was reclassified to 25 (upper-limit 2.5%),...16 was reclassified to 1000 (upper-limit 100%).
Calibration was initiated using the initial optimal Auxillary Multiplier, 25 Monte Carlo iterations and testing coefficient values as such, where no significant relief is present: Diffusion: 0 - 100; step interval 5 Breed: 0 - 100; step interval 5 Spread: 0 - 100; step interval 5 Slope: fixed at 25 Road Gravity: fixed at 100 due to the fact that it has been shown to have no significant relationship to fit statistics. Where significant relief is present: Diffusion: 0 - 100; step interval 10 Breed: 0 - 100; step interval 10 Spread: 0 - 100; step interval 10 Slope: 0 - 100; step interval 10 Road Gravity: fixed at 100 due to the fact that it has been shown to have no significant relationship to fit statistics.
2
11686
60.000000
17609
60.000000
true
false
-51951.227000 748292.078000
649208.773000 1804832.078000
649208.773000 748292.078000
-51951.227000 1804832.078000
298628.773000 1276562.078000
Projected urbanization
gcpo_urb2020
1000
0
Unified Code of Units of Measure
16
false
false
false
false
false
Pixel
39089
40991
60.000000
60.000000
16
1
Upper Left
FALSE
LZ77
1
pixel codes
FGDBR
TRUE
row and column
60.000000
60.000000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