Artistic 3D Visualizations

Today I want to highlight the work of the Senseable Lab at MIT, from a brief review of their work I’d say they seem to specialise in the area of real time 3D visualization and sensor input.

Beautiful Design Ideas: From an artistic ‘this is a work of art’ point of view their ideas are novel, fun and highly engaging, see this TED talk for examples
I really relate to the water building, I hang out on the South Bank in London and there’s a similar water sculpture there that is hugely popular (clip). Also, I’d LOVE to have some of those helicopter pixels in my lectures to illustrate geography concepts like earthquake waves to students.
Artistic 3D Visualizations of Singapore: This year the Sensable team have produced a project collecting real time data from Singapore and visualizing it. Here are some examples as a clip:


Looking from the angle of information communication there’s lots to like:
  • Engaging animations. The graphics draw the viewer in to find out more, they’re certainly engaging and artistically beautiful. I’m sure their exhibition at Singapore Art Museum was a sucess.
  • Elegant Time lines: They show time as a playhead moving against a timeline or against a bar chart illustrating relevant data. These elegant graphics are minimalist and communicate effectively without making the animation too busy visually. In a lot of ways they remind me of Tufte’s sparklines.
  • 3D Data Visualized Well? I’ve previously praised their technique in the of visualising 3D data using altitude, color and opacity at the same time as a way of getting over the problems of 3D thematic maps.
Beautiful but Ineffective? However, I worry that beyond looking attractive, these visualisations fail to communicate the data effectively. Two example issues that occurred to me:
  • Double 3D = Busy: In the heat vs energy consumption visualisation I think trying to show 2 sets of 3D data at once with the top layer of data partly obscuring the bottom layer doesn’t work well.
  • Where’s the Rain?: In the rainfall taxi visualisation by having the rain plot in 3D above the ground its difficult to relate where its actually falling on the ground.

I raise these issues without any evidence that they are actually problems, the only way of doing that is to conduct users tests. On the research page of the Singapore project Sensable discuss technical innovations and I admit in a real time visualisation project these are significant and important. However, there is no mention of user tests, given the amount of time and money that has gone into producing these animations wouldn’t it be a good idea to find out if they actually work?

Global Warming Map – Review pt II

continued from Part I last week

Layer Control: The plugin version has no layer control available to control the individual impact layers (you can turn them on and off as a group) so you can’t get around the multiple areas problem I outline above. You can control impact layers in the GEarth file but it’s confusing:
there are 6 levels of folders and way too many elements in this folder view. It could be a lot simpler which would improve usability.
Non-Spatial Data and Closing Balloons: In the above screenshot you can see a folder titled ‘background information’ is provided. It’s non-spatial data so it shouldn’t be in GEarth at all – better to link to a set of web pages elsewhere showing the same information and avoid cluttering up the GEarth layers panel.
While I’m talking about ‘Background info’ its sub-folders (even though they don’t have a folder icon, that’s what they are) have pop up balloons with a black background so you can’t see where to click the cross to close the balloon.
Poor Balloon Design: There are a number of issues with the pop up balloons;
Information Density: IMHO balloons are there to hold extra information (including images or videos) that cannot be presented easily as a traditional map symbol/key system. If you read the actual text in the forest fire example below the information you are actually getting is: ‘high forest fire danger projected to affect every populated continent’. Unnecessarily it then tells you where the areas are that are affected – this is a map, that’s what the polygons are communicating, no need to say it again in text. So there’s little new information in the balloon, it’s interesting to compare with the amount of data you get in placemark balloons in the Dafur project
Large Fonts: The top design is from the Impacts section in the GEarth plugin, the font is way too large decreasing the total amount of text that can be put in the balloon. Its less of an issue in the GEarth file with the text more reasonably proportioned, I suspected someone re-used the KML in the main file in the plugin without customising it to this different presentation. Kirsty confirms that this was so and says that she asked for the large font size to ‘give more prominence to the headline messages we were communicating with this map’. My answer is that with smaller fonts you can add extra detail towards the bottom of a balloon without forcing the user to click a link, users will close a balloon happily when they’ve read enough information. Again, the Dafur project is a good example.
Lack of Design Coherence: The two balloons from two different sections also don’t marry visually, as can be seen above. This isn’t good graphical design, you should have a consistent look across a project (e.g. across all pages of Kirsty explains that this is because the data comes from two different organisations (the Foreign Office and the Met Office) who do very different things (act on science, research science respectively). Kirsty says this difference needs to be ‘absolutely clear[in the mind of the user]’. I can understand that need but if the difference is so important, why isn’t it explained anywhere and why do the different data sets share a map in the first place?

Placemarks in Placemarks: In Google Earth opening ‘FCO Climate Change’ folder and then clicking the Spanish placemark you get an Inception-esque Google map within a GEarth map as captured in the screenshot below. Bizarrely you can open placemarks within a placemark leading to visual confusion.
Temperature Map Key: The temperature map overlay just about works in the original map as there is a key showing what the colors mean. However, in the GEarth file the key has become too small to read (its not available in the plugin). Most users will guess the heat colours corresponding to hot and cool temperatures but a readable key adds value.
Temperature Map Contours: The temperature map has heat contours on it out in the sea. These aren’t labelled and I would predict that most users won’t know what they are so it ends up just cluttering up the view.
Down-under Problem: I mentioned the fact you can’t compare Australia and the UK at the same time in a virtual globe in the skim review.
Talking Heads: The project links to video clips of experts talking about their specialisms within the topic area, they’re presented as talking head with no visual aids. Compare the style with a trailer from Wonders of the Solar system (one of the most successful science series the BBC has had)

Unless you are an exceptionally talented speaker (I’m certainly not) it’s very difficult to maintain a user’s attention as a talking head. Notice in the second clip that you never see Brian Cox against a static studio screen, the BBC are always trying to help the viewer visualise the point he’s making by putting him on location or providing a visualisation.

You don’t need a large travel and/or CGI budget to achieve this sort visual interest as I hope I showed in my latest climate change talk using GEarth as my main visualisation tool.
Kirsty comments that “they just wanted scientists to appear as the ordinary people they are”, that’s an understandable aim but I don’t think it relates to the ‘visual aids’ point I’ve made.


This is not a failure as a project, I think its main positive is that the ‘what happens with 4 deg warming’ approach is a great choice to communicate climate science to the public. However, if the design had been thought through more it could have been a lot better.
I wasn’t a huge fan of the original map version of this project (e.g it had the same area problem) but it was OK. The vast majority of the issues I raise here have been introduced by moving the content to GEarth without thinking through what the pros and cons are of GEarth client/plugin as a presentation tool.
In relation to the above paragraph, Kirsty asks “This is the key point. The question is really, can GEarth be used to successfully display this map? … …the challenge we had was to present an existing map, already well-known, in the GEarth format”. My answer is that no, GEarth is not the best choice of medium if the data (as in this case) relates to inter-continental scales because of the downunder problem. I may elaborate in the future on what IMHO GEarth is best at.

Global Warming Map – Review ptI

This is the first half of a two part review.
I did a skim review last month of the Foreign office’s recent Google Earth project showing the effect of 4 degrees average temperature rise. In short, its a great topic to work with but they could have done much better in terms of geoweb usability.

The project was originally produced as a 2D map which has been transferred to a Google Earth plugin and GEarth file presentation with more data. Click on the image to get to the original

I’ll be reviewing the presentation in the GEarth plugin (visible on the main page) and the theGEarth file which is linked to from the main page.

I’ve been in contact with the Kirsty Lewis the project manager (I’m not certain of her exact role) who’s commented on a draft of this review.

Press: Its got a lot of press attention e.g. this piece on channel 4 news.

Review Pros

Icons: The ‘Impacts’ icons (round and multi-colored) are clear, clean and use good symbols so the user can guess what they represent before actually clicking them. I like the use of color as well, by choosing less intense colors they’ve allow differentiation between impact layers without being overpowering.

Topic: Climate change is a global problem so Google Earth represents a good choice of medium. I especially like the overarching approach they’ve taken: ‘what will be the effect of a 4 degree rise in global average temperature’, because certainly here in the UK with our damp, cold winters a 4 degree rise in average temperature seems attractive if you think of it only in a shallow manner.

Acronyms and Jargon: Throughout the project both videos and maps, they steer clear of science acronyms and jargon which is good to see in a science communication project.

Review Cons
Use Points not Areas: The project marks areas with a colored ring and provides an icon of the same color nearby that can be clicked for more information. Thus we click a tap icon to find out that droughts in southern Europe are becoming more common. It would be better to give specific examples of droughts at several points as you could have then involved photos and a human scale story of the global problem. E.g. show a photo of a Farmer in Spain with dry soil running through his fingers and a personal story about how his farm is being affected.
Kirsty commented that she thought that this ‘specific location’ approach would lead to users incorrectly thinking that climate change is completely to blame for specific problems and that explaining the concepts would be complex and unwieldy. I see her point with this, climate change is often one of a number of factors producing a problem such as arid farms and climate change actually affects the probability of drought in a given area, it’s incorrect to say it produces droughts. However, I think you could add caveats that would work around these issues without overloading the content – I explain the issues to 14 year old students in 5 minutes using a betting analogy.

Multiple Areas at the same time: Another problem with the area approach is that its confusing to show multiple overlapping areas at the same time as can be seen in this screen shot of the GEarth plugin version:

There are multiple solutions to this confusing view, e.g. annotate areas with a color fill and white border and allow only one layer at a time to be viewed, I’ve produced a mock up below:

of course, it’s easier still to use points instead of areas as I recommend above.

Kirsty’s answer to this is that the circles make the point that areas are overlapping. That’s a fair point but you need to make it less visually complex when first viewed by a user. The circles would work better if they were introduced added one by one in a GEarth tour with an audio narrative (a concept I explore in in a book chapter).
She also thinks that its important that the GEarth presentation is visually similar to the original map so people understand the links between the projects and that removing the circles would severe that link. I think she’s correct that branding is important but that doesn’t mean you have to re-use all the details – just keeping the icons set is enough to link the GEarth and Map presentations in the user’s mind.
second half next week…