I've been tasked to produce a map using statistics of people living in Sydney. Basically I am transforming a list of names and addresses in to a map.
Council doesn't have the base maps to cover the entire Sydney metroplitan area, so as a Spatial Information Professional I have to source freely accessible information from a variety of sources. There isn't a lot of time to request spatial information through formal channels and I'm expected to have something by the end of the week.
I needed a base map so I searched the NSW Spatial Data Catalogue and found the NSW Land and Property Information's public base map layer.
This gives me a decent quality base map to work off without having to prepare one from scratch! Below is the description of the layer:
The next task was to analyse the data supplied by the stakeholder. Names and addresses are confidential so that's all I am going to say about it. The list is a spreadsheet dump from one of the corporate systems. Luckily there isn't a lot of names and I can just format it to suit my purposes. Next I begin the task of summarising the data, formating and transforming it into tables that can be imported into a file geodatabase.
Rather geocoding street addresses which would take a long time, I opted to summarise the data based on suburbs using the NSW state wide locality dataset. This was obtained from data.gov.au. This is handy because political and administrative boundaries are frequently changing.
First I isolate the geographic extent to the Sydney region. I assumed most of the people on the list are within 60km from a fixed location based on the client's description. Any outliers are omitted from display.
Then I use the
Summarise table function in ArcGIS to export a temporary table counting the total records in people living in each suburb. Next I join this temporary table to the locality dataset using the
suburbs field as the key. I had to make sure the fields in both tables have been concatenated, otherwise the database join in ArcGIS will put out errors.
Now comes the fun bit.
Visualising quantitative statistics is a mixture of science and art. Depending on how you prepare the symbology, you can present different interpretations of the very same data.
In this job I presented two different map symbology using the same dataset and suggested them to senior management for usage. One is
Categories, Unique Values using shades of colour for intensity; the other is
Quantitaties, Graduated Symbols using size of shapes to define the number of people in each suburb (centroid).
Categories, Unique Values
Click to enlarge
Quantitaties, Graduated Symbols
Click to enlarge
The main challenge now is fixing up the map layout. There are discrepancies such as missing labels and odd shaped blobs due to topological features 'stacking' on top of each other due to the scale of these maps. Fixing these are time consuming but mandatory. Overall an unusual service request, but satisfying mapping job.