1. Type your search term into the search bar above the map. 
2. Choose the parameter for your search from the dropdown menu. 
   • Contains: Show results that contain your search term.
   • Fuzzy: Show results with roughly similar spelling. The first results will be exact matches, and then places with less and less similar spelling will be listed.
   • Exact Match: Show results that match the exact term you entered, and no more.
   • anps_id: Show results with IDs from ANPS that match your search term.
3. Use the checkboxes to choose if you want to search the ANPS Gazetteer, NCG Gazetteer, Layers, or all three. 
4. Select Search. 

Your search results will be shown on the map. Click a place to see its details.

To apply filters and restrictions to your search, make an advanced search. 

1. Click the arrow next to the search bar. The advanced search pane will drop down.

Once you’ve set all the restrictions to your search that you want, select Search. Your search results will be shown on the map. Click a place to see its details.

If you think you will be revisiting the same search in the future, you can save it for later. 

1. Select Login at the top right and login. 
2. Follow the instructions to make a simple search or make an advanced search.
3. On the right, click List to swap to a list view.
4. On the search results page, select Save your search. 

You can find all your saved searches in My searches if you are logged in. 

1. In the green menu, Select Layers to view layers or Multilayers to view multilayers. 
2. Sort the list by Name, Size, Type, Content Warning, Created, or Updated by selecting the relevant heading. 

From here you can select the name of a layer or multilayer to see its details, or select View Maps to open it on a map. 

1. If you don’t already have a login, select Login at the top right. 
2. Click Register and fill in your details.
3. You will be emailed a confirmation. Check your junk/spam/trash folders if it doesn’t arrive in a few minutes. Some organisations hold this kind of message till the next day, so you may need to be patient.
4. Follow the instructions in the confirmation email.
5. Contact tlcmap@newcastle.edu.au if there are any problems.
6. Once you have logged in you can create layers and multilayers.

To create your own layer, you can just add one or two places, or upload a file of thousands of records.

1. In the user menu, select My layers. 
2. Select Create Layer. 
3. Fill in the form. You only need to enter a title and description. Everything else is optional and you can come back and add in more details later if you want. 
4. Set to private if you don’t want anybody to see it until you are ready, or to urgently remove it from public view. You cannot view the map of the layer if it is set to private.
5. Select Create Layer at the bottom. 

1. In the user menu, select My Multilayers. 
2. Select Create Multilayer. 
3. Fill in the form. You only need to enter a title and description. Everything else is optional and you can come back and add in more details later if you want. 
4. Select Create Multilayer at the bottom.

You’ll see a page showing the details of your new multilayer. To complete the multilayer, you need to add layers to it.

1. Select Add a Layer below the multilayer’s metadata table. 
2. Choose whether you want to add a layer from All public layers or just My layers (i.e. layers you’ve created yourself). 
3. Choose a layer from the dropdown menu. 
4. Select Add. 
5. Repeat these steps to add as many layers as you like. The multilayer will save automatically whenever you add a layer. 

To share any search result, layer, multilayer, or map, simply copy and paste the URL.

To share the raw data for a search result or layer, select Download to download the data as a CSV (works with Excel), KML or GeoJSON file. You can then send this file to others. 

Simply choose ‘ROCrate’ from the list of Download options for any search result or layer.

TLCMap is not a research data repository but we do make it easy to deposit your mapping research data in an official research data repository. ROCrate is an open standard format for saving research data with metadata in a way that can be accessed into the future, and which can be easily uploaded to a research data repository.

An ROCrate is simply a zip file containing the data (in this case the CSV, KML and GeoJSON exports of your search results, layer or multilayer) and a metadata description of it, all in open standard formats. You can edit the metadata using an ROCrate editor like Describo.

Search results and layers may change over time, so if you have based a paper on information and TLCMap you should probably ensure you have an ROCrate of it and that this is deposited following institutional research data requirements.

Note that TLCMap doesn’t mint DOIs – the repository you deposit in or your institution may mint DOIs.

Software developers can use web services API to conduct searches and access layers remotely in their code.

For those who know a little HTML, a map can be embedded in a webpage by using an iframe. 

The following code:

<iframe src="https://views.tlcmap.org/v2/journey.html?load=https%3A%2F%2F.tlcmap.org%2Flayers%2F721%2Fjson%3Fline%3Dtime" width="90%" height="400"> </iframe>

will embed a map in a web page like this:

Some web publishing systems, or organisational web security policies, may have constraints against embedding content from other sites. Check that the page you aim to embed the map in will allow this. Sometimes there is a workaround,eg: WordPress may block any embed except Youtube and similar, but you can insert a custom HTML block.

These are common statistics that are fast for the computer to process. Some of these results can be seen on a map by clicking ‘View Map’. Results can be downloaded for further analysis in spreadsheets or a GIS.

Total Places

The number of places in this layer.

Area

The area of the ‘convex hull’ of the layer.

Convex Hull
The ‘convex hull’ is the shape, or polygon, you’d make if you drew a line around all the outermost places in the layer. The coordinates of this polygon are provided in case you want to use them in a GIS system. You may wish to compare the area of one type of information with another.

Density

The number of places per square kilometre.

Centroid

The midpoint of the layer.

Bounding Box

Bounding Boxes are often used in describing map layers, this is a box drawn along lines of latitude and longitude that includes all points in the layer.

Most Central Place

The place closest to the centroid of the layer.

Most Distant Place From Centre

The place furthest from the centroid of the layer.

Distribution – Average Distance from Centroid

This measure of distribution can give you a sense of whether, or how much, places are spread out, evenly spaced, or mostly concentrated around a point, or far from the middle.

Distribution – Average Distance from Centroid / Area of Convex Hull
By dividing the Average Distance from Centroid by the area of the convex hull, we can get a measure of how spread places are, or how distant or concentrated they are on the middle – in a way that is more comparable to other layers. For example, we might want to compare a layer that covers a very large area, with one that covers a relatively small area to find out if the pattern of one sort of thing is more concentrated or more spread out. The Average Distance from Centroid for the first layer, would be very large compared to the other, but this enables us to consider it as a proportion. For example, let’s say you are wondering if music performances are spread out, or cluster around a place that might be famous for live music, and whether this is different in rural compared to urban areas. Music venues in rural areas are probably always much further away from each other than in the city simply because spaces are larger, but by treating it as a proportion of the whole area, you can make a comparison of distribution between small and big areas.

The advanced statistics are more complicated and take longer to run. For large datasets you may need to wait a minute for the results. Some of these are common statistics for which there are plenty of good explanations on the internet.

Min distance between 2 places
The 2 places that are closest together. After finding the shortest distance of all places to any other place in the layer, this is the shortest of all.

Max distance between 2 places
The two places that are furthest apart.

Average distance between places
Generally, places are this far apart from each other. After finding the distance between each place to every other place, this is the average distance between places.

Median distance between places
This is the median of all the distances between places. If you put all values in order, the median is the middle one. If the distances in kilometres between each place and each other place were 1,2,3,5,9,29,145 then the median distance is 5. This is useful when comparing with the average to get a sense of the distribution. The average of this layer is about 27.7. The median is much less than that. This means there are more lower values than higher values. This means distances between places tend to be closer together, but there’s a few ones very far away.

Standard Deviation of distance between places
The standard deviation also indicates the distribution. It’s how much places tend to deviate from the average. If it is low it means most distances are about the average. If it is large it means the distances tend to very different. This is useful for comparison. Eg: if the standard deviation of one layer is much higher than another, it means places in the first layer are more scattered and in second layer are more clustered close to each other.

Average Min distance between neighbouring places
The average distance between places and every other place may not be what you really want to know. To understand how close to each other places are, you may not be interested in the closeness of every other place, but only the closeness of the nearest place. It might be interesting to know how close campsites, waterholes, or neighbouring homesteads tend to be, because that would provide an indicator of the relative attractiveness of this kind of country to that, or indicate population density, or provide insights into historical behaviour. So this is the average of the minimum distance between each place and each other place – ie: it is the average distance of each place to its closest neighbour.

Median Min distance between neighbouring places
This is the median distance between each places’ closest neighbour, rather than all other places. See ‘median’ above for how this can be used with the average to understand distribution.

Standard Deviation of Min distance between neighbouring places
This is the standard deviation of distances between each places’ closest neighbour, rather than all other places. See above for an understanding of standard deviation.

Cluster analysis is a way to identify places that are closely grouped together. There are many different methods available, so we provide some of the common and easier to understand methods. The resulting clusters often depend greatly on the parameters you set. It’s often not a case of the analysis simply giving you a definitive answer of what the clusters are, but of tweaking the parameters to interactively interpret what clusters make the most sense and why.

Detailed descriptions of these methods can be found on the internet, but here are some brief summaries:

DBScan

DBScan creates clusters by finding places within a set distance from each other. If a place is further than that distance from any other place in the cluster, it is placed in a separate cluster. Eg: If the distance is set to 4km and A is 6km away from B, then A and B are not in the same cluster. If C is 4km from A, then C is in the same cluster as A. If C is also 3km from B, then they all join up and A, B and C are all in the same cluster.

You set the distance. This means you may do a bit of trial and error in finding a good distance to make sensible clusters. Sometimes you may find that there is a threshold at which clusters start to form, and at which every point is alone in a cluster. You may also find that the clusters don’t change much within a certain range, which would make you confident that those are the clusters. This tweaking of parameters for clustering can work in tandem with your knowledge of the field to confirm, deny, or illuminate unexpected patterns.

One advantage of this method over some others is that it can find clusters that are unusual shapes, such as a long cluster curling around a small central cluster, or donut shaped clusters.

Bear in mind that what we understand as ‘near’ or ‘far’ is not only measured in kilometres. Travel time might be important and 1km is very different in mountains to plains. The concept of how far a person might feel is normal to travel. In the city, 1 hours travel time is a long way, while in the outback, it is normal to travel that far. This might skew clusters such that urban or mountain areas that would be many clusters are grouped into one, while rural or flat clusters aren’t identified as clusters.

You can also set the ‘Number of Neighbours’. This is an optional parameter enabling you to set the maximum number of places that can be in a group. As the groups are built up, even if a place is within the set distance, if the maximum number of places is exceeded it will be placed in a different group. This may not be relevant for many cases but can be handy for assigning a certain number of places, for example to a group of people to work on.

K Means
This is one of the most common clustering methods. With K Means you set how many clusters you want to group places into, and the algorithm will find a good fit for each place into a cluster. The way this is done is by forming small groups and finding the average distance to the centroid of that group. Places are shuffled around until the average distance of each place to the centroid of its cluster is lowest. This means the clusters produced are those that are the most tightly packed around a central point. This is likely to identify clumped clusters, whereas DBScan can identify oddly shaped clusters.

Places can be clustered according to how close together they are in space, or in time. If there are places in a layer that don’t have dates, they aren’t included in the analysis.

The way we have implemented temporal clustering is similar to the DBScan spatial clustering method. You set an amount of time. If events have dates within that amount of time, then they are added to the same cluster. If they are separated by a span of time greater than the set amount of time, they are in separate clusters. You can specify time in years, days or both. Note that this doesn’t adjust for leap years or non leap years. It may be informative to compare spatial clustering with temporal clustering. This may show a progression of events across different places at different times, or bursts of activity in the same place at different times.

A common question is whether one sort of thing is close to or far from another sort of thing. Note that asking whether A places are close to B places is not the same as asking if B places are close to A places. For example, asking if music venues are usually close to restaurants is different to asking if restaurants are usually close to music venues. We might expect that music venues are close to restaurants as people often want to eat when they are enjoying a night out at music. On the other hand, people going out to eat aren’t always on their way to hear live music. We might expect that music venues are usually close to restaurants, but that there are many restaurants far from music venues. Another way to imagine this is if you have a layer where you only have places within one city (A), while you have another layer of some other type of place for the whole country (B). Asking if A is close to B can give a sensible answer, while asking if B is close to A, is asking if places all over the country are close to the city you have data for.

You may be specifically interested in how close one sort of thing is to another, for example to see how far homesteads generally were from towns, to get some sense of how many hours or days ride it would typically have taken. Eg: if the Average Min Distance were 20km, we might infer homesteads were typically within a day’s ride from towns. In other cases, these figures might be more meaningful by comparison – are music venues or sports venues closer to restaurants? Are music venues closer to restaurants or train stations and bus stops?

To analyse closeness, go to a layer for which you want to know how close they are to places in another layer, and chose ‘Closeness Analysis’ from the dropdown. Then, in the box, start typing the name of some other layer that you want to compare to and choose from the options. Click ‘Analyse’.

The following results can give you an indication of typical ‘closeness’ of one thing to another. The two layers are ‘A’ and ‘B’.

Max Distance
The maximum distance between any place in A and any place in B.

Average Min Distance
After finding the minimum distance from each point in A to any other point in B, this is the average minimum distance between A and B. This is probably the clearest and simplest indicator of how close A typically is to B.

Min Min Distance
The nearest any place in A is to a place in B. Of all the shortest distances from any place in A to any place in B, this is the shortest.

Max Min Distance
The furthest any place in A is to the nearest point in B. Of all the minimum distances from A to B, this is the furthest.

Median Min Distance
Of all the shortest distances from places in A to places in B, if you put them in order from shortest to longest, this is the middle one.

Average Min Distance / Area
This provides a common measure of ‘closeness’ regardless of scale. This is the Average Min Distance, described above as a proportion of the total area. By making it a proportion of the total area, we can compare the ‘nearness’ of one thing to another, with the ‘nearness’ of one thing to another, but regardless of scale. Eg: at a national scale you might regard things as ‘close’ to each other if they are within 10km. At a city scale they might be close to each other if they are within 1km. Another way of thinking about this is if you were to look at a national scale map and see with your eyes that points in A were close to points in B rather than scattered randomly, and you also could see something like that but at a smaller scale, you would think of both cases that places in A were close to places in B, even though in the first case are much further than the distances in the second case.

Min Min Distance / Area
This provides a common measure of the Min Min Distance regardless of scale. See above for a description of why we divide by area to arrive at a proportion.

Max Min Distance / Area
This provides a common measure of the Max Min Distance regardless of scale. See above for a description of why we divide by area to arrive at a proportion.

Median Min Distance / Area
This provides a common measure of the Median Min Distance regardless of scale. See above for a description of why we divide by area to arrive at a proportion.

When you are creating a layer, TLCMap may fail to upload all records for two reasons: 

• Errors: We attempt to identify any problems with the data, such as badly formatted dates or coordinates, and report them. However, some records may be imported before the error, leaving the job half done. Check the row after the last one that was successfully added for any potential problems (e.g., badly formatted coordinates or dates, blank ‘name’ or ‘title’ etc). 
• Very large layers (e.g., more than 5000 records): Uploading large layers can take a minute to process, so please be patient. In some cases, the layer is simply too large to handle and not all the layer is added. Also, be aware that visualisations of very large layers may take 30 seconds or so to load. This is a limitation of the web that we cannot easily resolve. 

Solutions

• If all records were not added, you can simply upload another file containing the missing ones (corrected). Uploading another CSV simply adds the new records to the same layer. 
• Try breaking the file into several smaller files and uploading them one by one. 

Very large amounts of information (usually around 5000 search results) may cause processing errors and time outs. 

If you are looking at one of your own layers, it may not display if it is set to ‘private’. The 3D visualiser is built using the ArcGIS javascript API. If the data is private the ArcGIS server can’t access it. 

Solutions

• Try breaking the file into several smaller files and uploading them one by one. 
• Set your layer to ‘public’. 

Every now and then, all the code for a page cannot be loaded quick enough, and TLCMap stalls. 

Solution

This is usually fixed by reloading the page, or forcing a hard refresh (hold down ‘shift’ and click the page refresh button).

Sometimes text on the web displays with question marks, little squares in it, or is garbled (eg: d?j? vu) This is a common problem on the web, especially when text is cut and pasted from MS Word or Excel. This tends to happen for text that isn’t in the basic Latin alphabet, such as letters with accent marks (eg: déjà vu), or for non-Latin writing systems (eg: ??) or special symbols (eg: © ?). 

TLCMap uses UTF-8 character encoding to ensure support for the full set of UTF-8 characters, which includes almost all languages and writing systems of the world, including many dead languages and many Indigenous writing systems, as well as a wide variety of symbols such as maths symbols and music notation. UTF-8 is now the de facto standard character encoding and used by most systems, especially on the web. 

If you are saving a CSV file from Excel, unfortunately, depending on the version or situation, it may save the CSV file in ‘ANSI’ encoding instead of ‘UTF-8’, which is the de facto standard character encoding.  

Solutions

• When saving a CSV file from Excel, under ‘Save as type:’ select ‘CSV UTF-8’. 
• If you need to get more advanced, try the free text editor Notepad++ which has tools for inspecting and changing character encodings. 

A recent problem with Google Chrome is that it disabled 3D rendering, which the 3D terrain view and Temporal Earth depend on. This type of issue could happen in other browsers too. In some cases, the map simply doesn’t show up, and in others it gives an error message that might mention WebGL. 

Solutions

• In Chrome, go to chrome://flags and set ‘WebGL Draft Extensions’ to ‘Enabled’. 
• For other web browsers, visit http://get.webgl.org to verify that your browser supports WebGL. If it doesn’t, upgrade your browser. If it does, you may need to change your settings. Use your favourite search engine to find how to enable WebGL in your browser, e.g. by searching for the phrase ‘Enable WebGL’. 
• 3D maps and Temporal Earth can take some time to load, so allow 20 seconds to start seeing something, especially on the first visit, or if you have a slow connection or old computer. Also, try refreshing the page.