Urban Floodplain Park GIS Raster Data Exercise
Rotina Tian LA 604 Regional Planning Professor Martha Hunt Feb 20, 2016
Content Introduction
Page 1
Goals and Objectives
Page 2
GIS Layers
Page 2
Flow Chart
Page 3
Suitability Maps
Page 4 - 7
Overlay Suitability Maps
Page 8 - 9
Focus Areas
Page 10
Blow up Maps
Page 11
Comparative Analysis
Page 12
Introduction
This exercise is to generate two suitability maps for urban floodplain park site selection from raster data. In the last project, the same maps was done by analysis the vector data. Raster data will provide hierarchy and more flexibility to site selection, but it requires more comprehension over ArcGIS software.
1
Goals and Objectives Vector Project
Raster Project
1. Appropriate land use area for the park 1.1. Place that is identified as open space 2. Most efficient area for flood control 2.1. Place within flood plain 3. Provide good service and convenient public access for urban area 3.1. Cities that own a population over 10,000 4. Provide education opportunities 4.1 Places within 3,000 feet from schools
1. Appropriate land use type for the park 1.1. Reclassify different areas by land use type 2. Most efficient area for flood control 2.1. Places within 3,000 from flood plain 3. Provide good service and convenient public access for urban area 3.1. Cities that own a population over 6,000 4. Provide education opportunities 4.1 Places within 3,000 feet from schools
GIS Layers and Considerations Populated Area: The park need to be located in a populated city to have enough visitors. Floodplain: The park need to be built within 900 feet from floodplain; places within 300 feet distance are the most suitable areas. Landcover: The park need to be built on appropriate land cover types. Different types of wetland areas are the most suitable land type, and the highly developed area need to be avoid. Schools (from recreation facilities): Areas closer to school can provide more education opportunities. Less than 300 feet is the most suitable distance. 2
Flow Chart Schools (from recreation facilities)
Floodplain
Populated Area
Multiple ring buffer: 300, 600, 900 feet
Landcover
Multiple ring buffer: 500, 1000, 1500, 2000, 3000
Reclassify 5,3,1,0
Reclassify 5,3,1,0
Reclassify 5,4,3,2,1,0
Reclassify 5,4,3,2,1,0
Town population value
Proximity to flood plain
Landcover value
Proximity to schools
Equally Overlay:Town pupolation, Floodplain, Landcover, Schools
Weighted Overlay:Town pupolation 30%, Floodplain 40%, Landcover 20%, Schools 10%
Equally suitability map
Weighted suitability map 3
Cities
Winchester
The citis is reclassified into 4 hierarchies by their population. The city Richmond owns the largest populaton of 39,124, and is considered to be the most suitable area. Two other cities who own population larger than 10,000 are sorted into moderate suitability, and the Winchester which is the only city left to have a population over 6,000, is sorted as low suitability. All of the other cities, each has a population smaller than 5,000, are not suitable for the park location
New Castle
Richmond
High Suitability Moderate Suitability Low Suitability
Connersville
No Suitability
0 2.5 5
10
15
ÂŻ
Miles 20
4
Floodplain
A munltiple ring buffer is set for the flood plain. The offset distances are 300 feet, 600 feet and 900 feet. The areas further than 900 feet from flood plain are not to be considered, they are inefficient to release the flood.
Legend 0 2.5 5
10
15
ÂŻ
Miles 20
5
Landcover
This map shows the suitability of different types of landcover. The developed area, pasture/hay and cultived crops are sorted as moderate-low and low suitability, the different types of forests are sorted as moderate suitability, and the wetlands and open space are high suitability. The high suitable areas are mostly concentrated within the cities, and some are close to roads and rivers. The areas of low suitability or no suitability are located close to the highly suitable areas, which bring constrains to the park's location.
High Suitability Moderate-High Suitability Moderate Suitability Moderate-Low Suitability Low Suitability No Suitability
0 2.5 5 0 2.5 5
10 10
15 15
Miles Miles 20 20
ÂŻ 6
Schools
This map shows the education opportunity of different areas. The offset distances from schools are 500 feet, 1,000 feet, 1,500 feet, 2,000 feet and 3,000 feet. Areas within 1500 feet from schools are easier to access on foot, therefore are sorted into high suitability and moderate-high suitability. Areas further than 3000 feet is difficult to access on foot, so are sorted as no suitability. Accessing on foot can bring much more outdoor education opportunities for students.
High Suitability Moderate-High Suitability Moderate Suitability Moderate-Low Suitability Low Suitability No Suitability
0 2.5 5
10
15
ÂŻ
Miles 20
7
nd
Equaly Overlay Suitabiility Map 5 Criteria Equalc Overlay
6,574 24,936 88,702 603,816 584,084 3,389,479
Winchester
The equally overlay map shows the hierarchy of suitability. The value of four layers (cities, floodplain, landcover and schools) are equally added, and the sum is sorted into 6 hierarchies. It matches the criteria of the site selection. It shows strong pattern of both the bigger cities and floodplain. The high suitability is of 0.14% of the entire area, which is very limited.
New Castle
Richmond
High Suitability Moderate-High Suitability
Moderate-Low Moderate Suitability
Connersville
Moderate
Moderate-Low Suitability
Moderate-High Low Suitability No Suitability
2.5 55 00 2.5
10 10
15 15
Miles Miles 20 20
8
ÂŻÂŻ
nd
Weighted Overlay Suitabiility Map
Winchester
The map shows the hierarchy of suitability of weighted added values of each layer. The calculation is sum of 30% of cities layer, 40% of floodplain layer, 20% of landcover layer and 10% of schools layer. It matches the criteria of the park's site selection, and is very similar to the last map. The high suitability is only of 0.07% of the entire area, which is half of the equal overlay.
New Castle
Richmond
High Suitability Moderate-High Suitability Moderate Suitability
Connersville
Moderate-Low Suitability Low Suitability No Suitability 0 2.5 5
10
15
ÂŻ
Miles 20
9
Focus Area Equally overlay
Weighted Overlay
From comparison of these two results in Richmond city, it is obvious that the equal overlay layer has a wider range of suitability, and the other one is more constrained by the shape of the floodplain. The equally overlay layer is influenced by the schools layer more than it should be, because the education opportunity is not a essential criterion for park location. Therefore, the weighted overlay is more reasonable.
10
Blow up Map Equally overlay
Weighted Overlay
Richmond
No Suitability
No Suitability
Low Suitability Moderate-Low Suitability Moderate Suitability
Source: Esri, DigitalGlobe, CNES/Airbus DS, USDA, USGS, IGP, swisstopo, and the GIS
Moderate-High Suitability High Suitability
Richmond Richmond
0
500
1,000
Feet 2,000
¯
Low Suitability High Suitability Moderate-Low Suitability Moderate-High Suitability Moderate Suitability GeoEye, Earthstar Geographics, 3 AEX, Getmapping, Aerogrid, IGN, Moderate-High User Community 2 Suitability 1 High 0 Suitability
00
500 500
1,000 1,000
¯¯
Source: Esri, DigitalGlobe, GeoEye, Earthstar Source: Esri, DigitalGlobe, GeoEye, Earthsta CNES/Airbus DS,DS, USDA, USGS, AEX, Getmapping, A CNES/Airbus USDA, USGS, AEX, Getmapping, IGP, swisstopo, andand thethe GISGIS User Community IGP, swisstopo, User Community
Feet Feet 2,000 2,000
These blow up plans in downtown Richmond reveal different results from different calculation methods. The weighted overlay is more constrained than equally overlay one. The area at the bottem has moderate-high suitability in equally overlay map, while it only has moderate suitability in the other one, because the landcover layer is weighted more in equally overlay map. In fact, it is not very suitable for the flood plain park, because it is too far from flood plain. In this case, the weighted overlay method is more reasonable. 11
Comparative Analysis After studying two different ways to analysis site suitability, I have a clear understand of the differences between the vector data and raster data.The vector data itself is more complex and presice than the raster layer. When doing the analysis, it is easy to get information for each piece ofv land, while the raster data is too rough in small scale. On the other hand, the result from vector data is very limited, in some cases, it is possible even to have no suitable area at all; while raster layer can provide more flexibility to site selection. When I was doing the value calculation from previous layers in raster data, I noticed that the final result highly depended on the criteria I set to different layers. It is more subjective than the result from vector data. Therefore, it is very important to be more considerable and do more research before setting the criteria for sorting the suitability. For example, in this case, because there is not enough information for me do define a value to different landcover types, I can only decide that by my own experience. I have changed the value of some landcover types which I am not very familar with, and the result came out very different. Although the result from raster data is more flexible, it should be used very careful; the vector data includes more information of each piece of land, but it is very limited and strict. 12