Home
Project Background
Method
Results
Discussion of Uncertainty
Sources & Acknowlegements

 

image source: http://www.salmonriver.org/

 

 

METHODOLOGY

Land Use Map Comparison and Accuracy Analysis

The percent land use change over the years 1966-2001 was calculated by creating two error matrices. One error matrix to determine the land use change between 1966 and 1986 and another error matrix to determine the land use change between 1986 and 2001. This was done by following a modified method published the University of Alberta. The results were used to create two sets of maps depicting the areas that changed from and to agriculture land between the three time periods.

Two different land use data sources were used in this analysis. The 1966 data was from the Government of Canada Land Inventory and the 1986 and 2001 data was from Land Surveys done by the Township of Langley. Due to the use of two different data sources for three different time periods the land use classification schemes were different (Table 1).

 
 

Table 1: Original Land Use Classification Schemes for each time period

1966
1986
2001
Urban built-up area
Single Family Dwelling
Agriculture
Orchards and vineyards
Agricultural Activities
Comercial/Service Use
Improved pasture and forage crops
Grain and Forage
Golf Course
Unimproved pasture and range land
Grain
Hobby Farm
Productive woodland
Vegetables
Industrial Use
Non-productive woodland
Root Crops and Tubers
Institutional Use
Provincial Parks
Forage Crops
Land in Transition
Grazing
Military Area
Small Fruits and Berries
Mineral Extraction
Beef
Mobile Home Park
Dairy
Not in use
Commercial Activities
Park
Unused Agricultural Land
Recreational Use
Vacant Residential
Residential Use
Vacant Commercial
Transportation and Communication
Industrial Activities
Utility
Railway
Roads
Airports
Golf Course
Institutional and Civic

In order to compare the data sets the land use attributes were subjectively reclassified to have any combination of the following seven land use classes (Table 2):

  • Agriculture
  • Residential
  • Commercial
  • Parks and Recreation
  • Transportation and Communication
  • Resource and Industrial
  • Government and Institutional

The reclassification method was done simply by comparing the original land use classes with the new land use classes. The old land use class was changed to the new land use class that best described it. An excel file summarizing the reclassification is located here.

Table 2: Summary of reclassified land use attributes for each time period

1966
1966 Reclassified
1986

1986 Reclassified

2001
2001 Reclassified
Urban built-up area Residential Single Family Dwelling Residential Agriculture Agriculture
Orchards and vineyards Agriculture Agricultural Activities Agriculture Commercial/Service Use Commercial
Improved pasture and forage crops Agriculture Grain and Forage Agriculture Golf Course Parks and Recreation
Unimproved pasture and range land Agriculture Grain Agriculture Hobby Farm Agriculture
Productive woodland Resource and Industrial Vegetables Agriculture Industrial Use Resource and Industrial
Non-productive woodland Parks and Recreation Root Crops and Tubers Agriculture Institutional Use Government and Institutional
Provincial Parks Park and Recreation Forage Crops Agriculture Land in Transition (4 areas) Residential (3 areas); Transportation and Communication (1 areas)
  Grazing Agriculture Military Area Government and Institutional
Small Fruits and Berries Agriculture Mineral Extraction Resource and Industrial
Beef Agriculture Mobile Home Park Residential
Dairy Agriculture Not in Use Agriculture
Commercial Activities Commercial Park Parks and Recreation
Unused Agricultural Land Agriculture Recreational Use Parks and Recreation
Vacant Residential Residential Residential Use Residential
Vacant Commercial Commercial Transportation and Communication Transportation and Communication
Industrial Activities Resource and Industrial Utility Resource and Industrial
Railway Transportation and Communication  
Roads Transportation and Communication
Airports Transportation and Communication
Golf Courses Parks and Recreation
Institutional and Civic Government and Institutional

Each land use shapefile was converted to raster format and combined into two rasters called “Combine66_86” and “Combine86_01” using the raster calculator in ArcGIS. The combine rasters displayed all possible land use combinations for each 30 m x 30 m cell. The attribute tables for each combine raster were imported into Microsoft Access as a database file. Error matrices for each time period 1966-1986 and 1986-2001 were created using a cross tab query wizard in Microsoft Access. These matrices were then imported into Microsoft Excel where they were manipulated to assess the accuracy between the annual land use maps (percent land use that remained the same over specific time periods). The percent land use change from other to agriculture and agriculture to other (“other” is defined by the six remaining land use classes) was calculated to determine the accuracy of critical Nitrate source areas within the Salmon River Watershed.

Maps of the agriculture land use within the Salmon River Watershed were created using ArcGIS to visualize these calculations. A raster for each land use time period (1966, 1986, and 2001) was created from the land use shapefiles. The values representing the seven land use classes in each raster’s attribute table were reclassified in the following manner using the spatial analyst tool:

  • 1966 Agriculture value = 1; Other land use classes = 0
  • 1986 Agriculture value = 2; Other land use classes = 0
  • 2001 Agriculture value = 4; Other land use classes = 0

The rasters were added together using raster calculator (raster66 + raster86 + raster01), depicted below. Cells with a value of 3 represent constant agricultural land use between 1966 and 1986. Cells with a value of 6 represent constant agricultural land use between 1986 and 2001. Cells with a value of 7 represent constant agricultural land use between 1966 and 2001.

Figure 1: Diagrammatic representation of agriculture land use values

Slope Analysis

A Digital Elevation Model for the Salmon River Watershed was retrieved from the University of British Columbia library DMTI geospatial e-data. Slope for the Salmon River Watershed was calculated using the surface analysis calculator in ArcGIS. A slope greater than or equal to 5% was determined to be critical using the Universal Soil Loss Equation (USLE). An excel document containing the calculaions is located here. This equation relates potential long term average annual soil loss to factors such as rainfall and runoff (R), soil erodibility (K), slope-length gradient (LS), crop/vegetation and management (C), and support practice (P):

 
A = R x K x LS x C x P
Equation 1

The LS factor represents a ratio of soil loss under given conditions to that at a site with the "standard" slope steepness of 9% and slope length of 72.6 feet. The steeper and longer the slope, the higher is the risk for erosion. Therefore, the higher the risk for nutrient rich soil and manure applied as fertilizer to be eroded during rainfall events.

Figure 2: Determination of the critical slope value based on USLE slope data for a slope length of 31 m (raster resolution of 30 m).

Critical Nutrient Source Area Analysis

Critical nutrient source areas were determined using ArcGIS by overlaying the constant agriculture land use rasters on the critical slope raster. The land was noted as a critical nutrient source area where agriculture land use was present on critical slope areas with stream interception nearby. These areas have been deemed critical due to their proximity to surface water, and an increased chance of surface runoff occuring due to the slope.