Oops, I mapped too many, and also not enough of the Channel Islands…

How I used the Merge, Multipart to Singlepart, and Dissolve geoprocessing tools in ESRI’s ArcGIS Online to build digital maps of the British Red Cross’ Presidents’ Network (and to correct the Channel Islands).

I’ve just finished working on an interactive digital map for the British Red Cross’ Presidents’ Network — volunteers who act as local ambassadors for the British Red Cross across the UK.

Map Snapshot — This map details the vacant (red) and occupied (green) Presidential Counties, with embedded links to information about British Red Cross services and reach in each area — all of which will support the Presidents in their vital roles, as well as enhance internal communications between the Presidents and other British Red Cross staff and volunteers.

Since we upgraded our ESRI ArcGIS mapping platform to the cloud based ArcGIS Online in 2020, this has allowed us to produce a much greater breadth of interactive web maps and apps, both publicly and privately accessible online.

For this task, I needed to re-build an existing map detailing the boundaries of each Presidential County. Not all are true Counties — some are larger areas, mostly made up of multiple Local Authority Districts (LADs), and sometimes multiple historic counties.

For how I did this, read on.

Interim Layers (here’s one I made earlier…)

Thinking ahead to a time when a boundary may change, I decided it was worth spending the time now to build an interim layer, by grouping the various component LADs.

I re-exported an Office of National Statistics (ONS) LADs shapefile previously saved to our ArcGIS Online portal. In this new layer, I defined which of the 382 UK LADs belongs to which of the 57 Presidents’ Counties, by adding an additional field to the attribute table called “PCounty”.

By styling the new layer according to the “PCounty” field in the attribute table, each Presidents’ County appears in a distinct colour. For example, the Presidents’ County of Gogledd Cymru / North Wales (shown here in red) is made up of the Welsh Local Authority Districts (LADs) of (from North West to South East) Isle of Anglesey, Gwynedd, Conwy, Denbighshire, Flintshire and Wrexham.

At a quick glance, the interim “Presidents’ Network LADs” layer looked complete. However, as is very often the case in our line of work, the quirks of UK, Devolved and Crown geography had struck again.

The Presidents’ Network covers the Isle of Man and the Channel Islands. As Crown Dependencies, these are not part of the UK, so were not included in the ONS’ LADs shapefile — although the UK government does hold some responsibility for public services and relations, and the British Red Cross has some presence. Trust me, it’s complicated.

Using the Merge tool

Having loaded another layer containing the Crown Dependencies shapefiles, I created 2 temporary layers by selecting and exporting just the Channel Islands, and just the Isle of Man. Then, I used the merge tool to combine these 2 temporary layers with the Presidents’ Network LADs layer.

However, after completing the “PCounty” field in the attribute table for the 2 newly added Crown Dependencies, I realised that the attribute table only had 56 rows, meaning that 1 of the 57 Presidents’ Counties was missing.

Comparing to the existing map, I realised that the Channel Islands were not one Presidents’ County. Jersey and Guernsey are 2 individual Presidents’ Counties, and Sark, Herm and Alderney were not part of the Presidents’ Network at all. Oops.

Using the Multipart to Singlepart tool

No matter, I still had the temporary Channel Islands layer from the merge. Now to separate the group of islands (1x multipart feature) into individual islands (5x singlepart features).

The purple layer (above left) is 1 multipart layer (multiple islands which have been grouped). Note that there is a single row in the attribute table highlighted turquoise, which corresponds to the group of islands.

The pink layer (above right) is split into a 5 single parts (individual, ungrouped islands). The “new” separate island of Guernsey is highlighted in turquoise on the map and in the attribute table. Note that there are several rows in the attribute table: each one represents one island of the Channel Islands.

The new single part features were named in a newly added attribute field called “Island” and saved as a permanent export for future use. Then they were merged with the main Presidents’ Network LADs layer. This included all 5 islands for the same reason we were keeping the LADs breakdown — should any more Channel Islands need to be added to the Network in the future.

Using the Dissolve tool

Now to make the layer which will ultimately be displayed within the final product. I re-exported the “Presidents’ Network — LADs” layer and saved it as the “Presidents’ Network — Counties” layer.

Then I deleted the 3 Channel Islands that are not part of the President’s Network, to get the correct number of Presidents’ Counties (57) and completed a dissolve. This redrew the boundaries along a chosen field in the attribute table (in our case “PCounty”), removing the component LADs.

The Dissolve tool has removed the 6 component Local Authority Districts seen above, which make up the Presidents’ County of Gogledd Cymru / North Wales (shown here in red).

The moral of the story here, is that no matter how complicated the geography behind your mapping, or the combination of shapefiles or layers you have to use, GIS tools allow you to make the mapping products you need. It might just take you a few tries to get there!

Next steps…

Now we have the Presidents’ Network map and the data behind it in ArcGIS Online, we can look to develop the map further. For example, we could add layers featuring other British Red Cross geographies, such as where our services or internal boundaries are. This will help support our Presidents, Vice Presidents and Patrons in their crucial roles of championing the British Red Cross locally — leading to greater awareness of our work, fundraising, and advocating for better support for people in crisis.