6.9.7. Zonal statistics of land cover values within buffers
With this binary classification, we can now do a simple statistical analysis of the land cover within each buffer polygon layer. Below, the Zonal Statistics as Table tool will be used to summarise the values of the land cover layer within the buffers created in section 126.96.36.199. and output the results into a table. Specifically, we will calculate the mean percentages of non-irrigated arable land, coniferous forest, and mixed forest within a 2 km radius of the honey bee colony locations.
- Open the Zonal Statistics as Table Tool (Spatial Analyst Tools > Zonal > Zonal Statistics as Table).
- Input raster or feature zone data: buffer2000.shp (browse to or add to *.mxd).
- Zone field: FID (Unique identifier).
- Input value raster: one of the newly created binary raster images (e.g. lc_class211.tif).
- Output table: stats211.
- Check the Ignore NoData in calculations option.
- Choose the Statistics type MEAN (Fig.54).
- Click OK.
- Open the Results window (Geoprocessing > Results) and double click the Zonal Statistics as Table option (since that was the last process completed).
steps 50 to 57 for the other input rasters by altering the input dialog for the
classes 312 and 313 (choosing appropriate filenames for the results).
Now look at the table results in the TOC (Fig. 55). If you cannot see
the tables in the TOC, make sure you are in List
by Source view instead of List by
Drawing Order (section 6.3.4.). In
each table you will see columns for Rowid, FID (unique identifier corresponding
to buffer ID), COUNT (number of pixels of particular land cover class inside
the buffers), AREA (calculation of square meters of land cover class within the
buffers), and MEAN (in percentage). In Fig. 55, the mean values directly
correspond to the percentage of non-irrigated arable land within each buffer
(uniquely identified by its FID value).
Fig. 54. The Zonal Statistics as Table dialog box.
Fig. 55. The results of the Zonal Statistics as Table calculation.