Geovisualisation and Geocommunication

We still distinguish between fundamental map types, like topographic vs. thematic maps, and the latter take many different forms like choropleth maps for normalized data or proportional symbol maps showing absolute values.

To fully appreciate the wide scope and rich spectrum of visualisation approaches for geographic observations, phenomena and processes, take a leisurely evening to explore this current collection of best practice examples:

Just in case you're interested in some glimpses into the long and diverse history of cartography, you might spend another evening with this collection:

Web maps, as digital online maps, are not simply digitized cartographic products. While printed maps need to provide a full context to readers - with title, orientation, legend and all - web maps are frames for interaction, for a dialog between users and a well designed platform.

By zooming out we establish spatial context, clicking or tapping on features provides access to underlying data, switching on a legend offers insights into semantics. Experiment with the map on the right, and if you want to know more about its theme, check out the  Social Progress Index .

Our main objective in this module is to use feature (and image/raster) services to create shareable web maps. This requires considerations about:

• target audience
• message to be conveyed
• presentation context (standalone, with comment)
• media and device characteristics
• etc

Our main tool in the first half of this module will be the already well known ArcGIS Online 'Map Viewer', supporting the key workflow elements of

• opening a feature layer
• exploring data and distributions
• classification
• assigning symbology
• labeling and popups
• constraining views (e.g. scale range)
• sharing the web map with your audience

Qualitative = categorical = nominal data like land use, geological units or vegetation types usually will be presented directly, maybe sometimes categories will be aggregated.

Quantitative variables like elevation, temperature or densities can be presented with continuous scales, but are more effectively communicated when classified into groups of similar observations

Classification, or categorization, is based on the universal simplification principle of less-is-more. Aiming at reducing a high number of observations into a lower, manageable number of classes combining similar observations, the communication and understanding of (spatial) patterns is supported.

Classification is based on simple, and also advanced statistical techniques, and more recently has become a target domain of machine learning.

In cartography, data classification of continuous measurements is using methods like natural breaks (Jenks), quantile, equal interval, defined interval, standard deviation etc., but also uses multivariate techniques or semantic aggregation of similar observations.

Different classification techniques can lead to significant differences in the perception of phenomena and therefore have to be selected carefully considering data characteristics as well as communication aspects with the target audience.

Before we start assigning symbols to nominal categories or classes of metric variables, we need to think about the basic elements of the language of visualisation, sometimes referred to as visual variables.

Color is considered one of the 'strongest' visual variables, its signal dominating over other symbology characteristics. Specific colours are defined using different color models which can be numerically and geometrically represented.

For screen and display use, 'active' color schemes like Red-Green-Blue mixing of primary colors is well established, with the individual components R-G-B quantified in a 1-byte range of 0-255. E.g. a bright green could be coded as 0-255-81, or in hexadecimal notation as 00ff51. 255-255-255 obviously would represent pure white.

As an alternative, the H-S-V color model can be applied. While in R-G-B all three components are equivalent, Hue serves as a 'lead' color variable, modified by the Saturation and Value (dark to light, i.e. black to white) components:

Spatial phenomena with fixed, 'qualitative' categories like soil type, ethnic groups or land cover are visualized with discrete colors clearly separating classes from each other, and avoiding transitions between colors symbolizing individual classes.

In some cases, predefined 'colormaps' are already attached to data sets to support the use of standardized legends, like in this case of a global  Sentinel-2 land cover  service shared in the Living Atlas portal.

Continuous phenomena and variables, like elevation, slope in degrees, surface temperature or precipitation are either classified with break points, or displayed with continuous legends using color ramps.

The opportunity of zooming and panning around the world, between local and global scales is a great feature of web maps. But - it is easy to get lost, not to speak of the disorientation which can easily happen on a virtual globe.

Luckily, tools for assisting the user and sometimes even constraining navigation are readily available, like:

• finder maps
• restricted extents
• min/max scales
• block rotation
• zoom to selection
• bookmarks
• home button

We want users to interact and explore, but we also need to avoid them getting lost!

Popup windows triggered by clicking or tapping on a feature are a key mechanism allowing users to explore the actual data behind map features.

• Deactivate popups for background layers
• Display only variables of interest
• 'Customize' all popups, no irrelevant fields
• Use charts / pictures in popups
• Consider working with Arcade for finetuning

Highlighting particular categories or classes by clicking on legend entries is a powerful technique assisting with the interpretation and analysis of spatial data.

Using these kinds of approaches demonstrates the dynamic and interactive character of communicating through web maps.

The power and value of perspective views is based on a much more impactful perception of spatial features and context, in comparison to flattened maps.

Several cues build viewers' 3D impression:

1. scale reduction from front to back
2. convergence of parallel lines
3. overlap and obscured spaces
4. shades and shadows, where employed
5. less detail in the background
6. reduced colour saturation towards back, fuzziness
7. changed perspectives with movement

In a Digital Earth context, exploring und understanding any area of interest or the entire globe by moving the view point, by looking at it from different angles is a fundamental concept. It is already anchored in the way Digital Earth was communicated as a ' spinning virtual globe '.

Explore the setting of mountain biking routes, and observe yourself as you 'learn' about this area by changing the view and using different angles:

Perspective views today are increasingly employed not only to understand topography, but also its 'furniture': buildings, vegetation and other 'detail', like avatars.

This type of 'local scenes' is widely used by planners, architects and whenever a local built environment is at the center of a dialog among various stakeholders.

Perspective views as 'scenes' also are a natural environment for 3D analyses, like line-of-sight and viewshed.

Getting back to Digital Earth's 'spinning virtual globe', a perspective view on our entire planet has one drawback - we only can see one hemisphere at any moment - and many advantages: all distortions due to map projections are avoided, global spatial relations and distances are more intuitively understood, and other than a 'hardware' globe we still can zoom, identity features and drape layers.

Explore navigating a virtual globe:

You of course might be familiar with different virtual globes, like switching Google Maps into ''globe view', or:

 Storymaps  are web apps developed through a 'builder' or 'wizard' supporting a no-code approach, i.e. authors can focus on the storyline and content presentation.

Typical elements of storymaps are:

• Text and imagery
• Videos and live embeds
• Maps and scenes
• ad hoc 'express maps'

These are structured into

• Sidecars
• Galleries
• Timelines etc

Most importantly, though, designing a storymap needs a clear vision of what to tell to who (audience), why, and with what intended outcome.

Before starting the compilation of a storymap, it is essential to establish a design outline. Like with other multimedia developments, sketching out a storyboard along a 'red thread' is good practice.

 Dashboards  link different simultaneous view around one task with each other, so as to explore interdependencies, monitor processes or better understand spatial phenomena.

Like the eponymous car dashboard, there is a clear purpose and user focus, but no preset operating sequence.

Dashboards are built from web maps/scenes, charts, lists, numerical or graphic indicators and selectors for filtering. Most importantly, these elements are inter-linked, users can interact with widgets in any order to focus an useful views.

Based on anticipated user interaction and underlying feature services (including real time data streams), the dashboard author will outline the different views (map, chart, table, ...), how they are linked, what kinds of actions are triggered by selectors and in-view selections etc

Most importantly, a logical narrative needs to be established what kind of learning and understanding, what decision processes are supposed to be facilitated with a dashboard.

Excellent guidance and a fast track to standalone apps is being offered by instant apps. Starting from a web map, authors are shepherded through a step-by-step definition process using a choice of design templates.

After a first 'instant' app setup, authors are given the opportunity to finetune and adjust details - or immediately share their app with the intended audience.

This authoring environment does not constrain its users into particular types of designs or templates, but allows complete free designs with all kinds of elements and widgets, on desktop or mobile platforms.

Such a fully customized user experience obviously requires more 'design thinking' and development effort, we therefore will only 'dip our toe into the water' in order to appreciate the power of this environment.

Just to complete perspectives reaching above and beyond this module, we point anyone interested to the option of developing apps running natively on Android or iOS platforms.

Up to now we have worked with generic apps like QuickCapture or Survey123 or Field Maps running projects for a particular purpose, or with web apps executed with the help of a browser - like in this module.

Whenever standalone native apps for installation through app stores are required, App Studio is the way to go. Again, this is mentioned here for orientation only :-)