We want to figure out how land use has changed between 1978 and 2005 for several counties in southeastern Pennsylvania. We are mainly interested in the urbanization of agricultural and forested areas. You may have noticed that the land cover categories and coded values are different for the 1978 and 2005 datasets. Since we are interested in comparing land use change, we will need to standardize these categories before we can compare them. We also want to remove extraneous information from our datasets to make them easier to work with. We will use the Reclassification Tool in Spatial Analyst to perform both of these tasks simultaneously.
We will reclassify both of the input raster data layers using the standardized codes below. Codes 1, 2, and 3 collapse the existing detailed categories into broader categories. The "NODATA" (ALL CAPS) category allows us to ignore all of the land cover categories that we are not using in our analysis.
Value | Category |
---|---|
1 | Developed Land |
2 | Agricultural Land |
3 | Forested Land |
NODATA | All Other Values |
The tables below show the original land cover codes from the 1978 and 2005 land cover grids, associated descriptions, and the new codes we will use to reclassify the data.
Original value | Original Category | NEW Reclass Value |
---|---|---|
11 | Residential | 1 |
12 | Commercial and Services | 1 |
13 | Industrial | 1 |
14 | Transportation, Communications... | 1 |
15 | Industrial and Commercial Complexes | 1 |
16 | Mixed Urban or Built-up Land | 1 |
17 | Other Urban or Built-up Land | 1 |
21 | Cropland and Pasture | 2 |
22 | Orchards, Groves, Vineyards | 2 |
23 | Confined Feeding Operations | 2 |
24 | Other Agricultural Land | 2 |
31 | Herbaceous Rangeland | NODATA |
32 | Shrub and Brush Rangeland | NODATA |
33 | Mixed Rangeland | NODATA |
41 | Deciduous Forest Land | 3 |
42 | Evergreen Forest Land | 3 |
43 | Mixed Forest Land | 3 |
51 | Streams and Canals | NODATA |
52 | Lakes | NODATA |
53 | Reservoirs | NODATA |
54 | Bays and Estuaries | NODATA |
61 | Forested Wetland | 3 |
62 | Non-forested Wetland | NODATA |
72 | Beaches | NODATA |
73 | Sandy Areas other than Beaches | NODATA |
74 | Bare Exposed Rock | NODATA |
75 | Strip Mines, Quarries, and Gravel Pits | NODATA |
76 | Transitional Areas | NODATA |
Original value | Original Category | NEW Reclass Value |
---|---|---|
14 | Roads | 1 |
21 | Row Crops | 2 |
24 | Pasture/Grass | 2 |
41 | Deciduous Forest | 3 |
42 | Evergreen Forest | 3 |
43 | Mixed Deciduous and Evergreen | 3 |
50 | Water | NODATA |
51 | Streams and Canals | NODATA |
52 | Lakes | NODATA |
61 | Forested Wetlands | 3 |
62 | Emergent Wetlands | NODATA |
70 | Bare; Unclassified Urban/Mines, Exposed Rock, Other Unvegetated Surfaces | NODATA |
111 | Residential Land; 5-30% impervious | 1 |
112 | Residential Land; 31-74% impervious | 1 |
113 | Residential Land; 74% < impervious | 1 |
121 | Institutional/Industrial/Commercial Land; 5 - 30% impervious | 1 |
122 | Institutional/Industrial/Commercial Land; 31 - 74% impervious | 1 |
123 | Institutional/Industrial/Commercial Land; 74% < impervious | 1 |
124 | Airports | 1 |
241 | Golf Courses | 1 |
750 | Active Mines/Significantly Disturbed Mined Areas | NODATA |
1111 | Residential Land; 5 - 30% impervious; Deciduous Tree Cover | 1 |
1112 | Residential Land; 5 - 30% impervious; Evergreen Tree Cover | 1 |
1113 | Residential Land; 5 - 30% impervious; Mixed Tree Cover | 1 |
1121 | Residential Land; 31 - 74% impervious; Deciduous Tree Cover | 1 |
1122 | Residential Land; 31 - 74% impervious; Evergreen Tree Cover | 1 |
1123 | Residential Land; 31 - 74% impervious; Mixed Tree Cover | 1 |
1131 | Residential Land; 74% <impervious; Deciduous Tree Cover | 1 |
1132 | Residential Land; 74% <impervious; Evergreen Tree Cover | 1 |
1133 | Residential Land; 74% < impervious; Mixed Tree Cover | 1 |
1211 | Institutional/Industrial/Commercial Land; 5 - 30% impervious; Deciduous cover | 1 |
1212 | Institutional/Industrial/Commercial Land; 5 - 30% impervious; Evergreen tree cover | 1 |
1213 | Institutional/Industrial/Commercial Land; 5 - 30% impervious; Mixed tree cover | 1 |
1221 | Institutional/Industrial/Commercial Land; 31 - 74% impervious; Deciduous Tree Cover | 1 |
1222 | Institutional/Industrial/Commercial Land; 31 - 74% impervious; Evergreen Tree Cover | 1 |
1223 | Institutional/Industrial/Commercial Land; 31 - 74% impervious; Mixed Tree Cover | 1 |
1231 | Institutional/Industrial/Commercial Land; 74% < impervious; Deciduous tree cover | 1 |
1232 | Institutional/Industrial/Commercial Land; 74% < impervious; Evergreen tree cover | 1 |
1233 | Institutional/Industrial/Commercial Land; 74% < impervious; Mixed tree cover | 1 |
After all of the time periods share common land cover codes, we can calculate how much change has occurred in each category over time using the workflow below:
It is important to remember to double-check the environment settings within the Spatial Analyst tool pane, as ArcGIS sometimes ignores the global environment settings. A general rule of thumb is to always be certain of the environment settings used in your analysis, as they are critical to your results.
Notice how the extent setting we used clipped the raster to a much smaller area, and the mask setting we used assigned values of NoData to all of the areas that are both outside our study area boundary and within the extent.
Also, notice the grey areas within our study area. These are places that we reclassified the original land cover to "NoData." Keep in mind that you could also do the opposite of what we did – you can reclassify cells with starting values of "NoData" to other values.
Make sure you have the correct answer before moving on to the next step.
The cell counts in your RC_lu_1978.tif should match the examples below. If your data does not match this, go back and redo the previous step. You can double-check settings and rerun the tool in the Results window.
You’ll need right-click the RC_lu_1978.tif in Contents pane > Attribute table to see the Count attribute.
How did the extent, mask, and cell size settings affect the output raster? You can view the cell size settings by right-clicking on the output raster > Properties > Source > Cell Size.
Make sure you have the correct answer before moving on to the next step.
Your LU_2005_RC grid should match the example below. If your data does not match this, go back and redo the previous step.
Since you know the cell size and number of cells with each unique value, you can easily calculate the total area within each land cover category for the entire study area. Note that you need to use the area of the cell, not the length, when making these calculations.
In the next step, we will use the "Tabulate Area" tool to create a table with the areas of each land cover type within each county. We will repeat this for both time periods. The "Tabulate Area" tool will automatically generate column names based on the values in the input table. Since we will have two datasets with the same land cover codes, we need to be able to keep track of each year’s corresponding table. To do this, we will add new fields to each reclassified raster attribute table and populate them with a combination of the study year and the land cover code.
Make sure you have the correct answer before moving on to the next step.
Your reclassified attribute tables should have their ID values populated as shown below. If your data does not match this, go back and redo the previous step.
OID | Value | Count | LU | ID |
---|---|---|---|---|
0 | 1 | 3762517 | DEV | 1978_Dev |
1 | 2 | 16778194 | Agr | 1978_Agr |
2 | 3 | 11264313 | For | 1978_For |
OID | Value | Count | LU | ID |
---|---|---|---|---|
0 | 1 | 6730640 | Dev | 2005_Dev |
1 | 2 | 10679480 | Agr | 2005_Agr |
2 | 3 | 14529196 | For | 2005_For |
Now that we have reclassified the land cover data with standardized categories and created unique IDs, we can begin our land use change analysis. We need to calculate the area for each of the three land cover categories within each county for each time period. To do this, we will use the "Tabulate Area" tool. This tool calculates cross-tabulated areas between two datasets. This tool summarizes one dataset within regions specified by a second data set.
Open the "TA_1978.dbf" table in your map. Notice the names of the columns. What are the units of the tabulated areas?
Make sure you have the correct answer before moving on to the next step.
Your tabulated area tables should match the examples below. Both of the tables should have 19 records and 5 columns. If your data does not match this, go back and redo the previous step.
OID | FIPS_CODE | A_1978_Dev | A_1978_AGR | A_1978_FOR |
---|---|---|---|---|
0 | 025 | 60044775.0275 | 140076533.603 | 764621228.308 |
1 | 029 | 283115147.234 | 1221605255.57 | 451162509.312 |
2 | 041 | 108895573.345 | 864337176.226 | 448917213.30 |
3 | 043 | 142616070.306 | 607690006.007 | 604278011.426 |
4 | 071 | 187182416.169 | 1895813331.92 | 374069736.292 |
OID | FIPS_CODE | A_2005_Dev | A_2005_AGR | A_2005_FOR |
---|---|---|---|---|
0 | 025 | 85861467.5195 | 101130340.659 | 771302030.31 |
1 | 029 | 557661328.688 | 673340415.676 | 659276515.52 |
2 | 041 | 239732484.125 | 630922677.737 | 531524170.633 |
3 | 043 | 236418388.277 | 400440924.138 | 703767243.941 |
4 | 071 | 433833969.022 |
1367605682.98 |
608866798.469 |
We will use the Join function to create a "master table” that contains the information from both of the Tabulate Area tables and the attributes of the counties. Since a joined table contains only virtually referenced information, we will export this dataset, thus permanently saving the joins.
Make sure you have the correct answer before moving on to the next step.
Your attribute tables should match the examples below. If your data does not match this, go back and redo the previous step.
FID | county_Nam | FIPS_code | A_1978_Dev | A_1978_agr | A_1978_for | A_2005_Dev | A_2005_agr | A_2005_for |
---|---|---|---|---|---|---|---|---|
0 | Carbon | 025 | 60044775.027 | 140076533.603 |
764621228.308 |
85861467.5195 |
101130340.659 |
771302030.31 |
1 | Chester | 029 | 283115147.23 | 1221605255.57 | 451162509.312 | 557661328.68 |
673340415.676 |
659276515.52 |
2 | Cumberland | 041 | 108895573.349 |
864337176.226 |
448917213.304 | 239732484.125 | 630927677.737 | 531524170.633 |
3 | Dauphin | 043 | 142616070.306 | 607690006.007 | 604278011.426 | 236418388.277 | 400440924.138 | 703767243.941 |
4 | Lancaster | 071 | 187182416.169 | 1895813331.92 | 374069736.292 | 433833969.022 | 1367605682.98 |
608866798.469 |
5 | York | 133 | 144054770.453 |
1585057521.64 |
609945566.225 |
451855625.63 |
1025137908.13 |
837911659.598 |
6 | Philadelphia | 101 | 322253436.539 | 9576897.87968 | 4153583.7766 | 275763193.022 | 18073754.4639 | 31790615.6787 |
7 | Lebanon | 075 | 70680207.6218 | 582750540.97 | 277862437.60 | 134141054.505 | 423053914.346 | 364983291.654 |
Sometimes your calculated values will have too many digits to be stored in a long integer field. In these situations, you can use a data type of "float" instead.
As we saw in Lesson 2, it is much easier to compare numbers using percent areas vs. calculated areas. In this step, we are going to calculate the percent change within each land use type between 1978 and 2005.
([ tot land use in later time] - [ tot land use in earlier time]) / [TotAreaSqm]) * 100
Make sure you have the correct answer before moving on to the next step.
Your calculated values should match the example below. If your data does not match this, go back and redo the previous step. I have only included the values for Adams County. You may need to sort your results to find this county.
Create a map layout with the 4 map frames below. (Note: You will not turn in these maps. However, you will need to consult them to complete the Lesson 5 Quiz).
In Lesson 5, we used the Reclassify Tool to collapse complex categories into simpler versions. We also used it to eliminate portions of our starting data that we did not need for our analysis using the "NoData" code. Can you think of any other ways you could use this tool?
That’s it for the required portion of the Lesson 5 Step-by-Step Activity. Please consult the Lesson Checklist for instructions on what to do next.
Try one or more of the optional activities listed below.