Geographic Usability Workshop and absence

Sorry no post this week, I have one brewing about GIS teaching in UK schools which will come out soon. Meanwhile, I'm trying to put something together for this UK based workshop about Geographic Information Usability (via Muki Hacklay) which looks like it could be a very useful forum.

London Data Store, Snow Icons and Isle of White

London Data Store: A lot of map related news has caught my eye over the last couple of days. Last night the mayor of London promised a data store for London: my neogeographer friends everywhere cheered at the thought of all the new shiny mashups they'll make and went to sleep dreaming of being buried in £10 notes rather than snow.

Snow Volunteered Geographic Information: Talking of snow, there have been a slew of UK snow maps that deserve a look. A trending topic on twitter has been #uksnow, (explanation) where people publish the snowfall conditions in their postcode. There are a number of terms for this kicking around including Volunteered Geograpic Information (VGI) and crowd sourced maps but the best way of describing it as a map wiki: people add their own data to a map and we all benefit from the result.



Ben Marsh has a lovely map website where he takes these tweets and maps them as clickable items. To the right of the main screen (which I've grabbed above) is a list of latest tweets by people and on the left is an explanation and key. Its a lovely implementation, pulling in real time tweets but I think his icons could be improved. You get a choice of an icon made up of white spots showing snow intensity by number of dots (shown outside red box above) or red snowflake icons illustrating snow intensity by size of icon (in red box above). The white icons are difficult to see over the background and you can't really tell the difference between multiple icons showing small snow falls and a few blizzard icons. The red icons stand out well but the smallest of them are not really visible despite being red. A possible solution would be to fade out the google maps background (white translucent layer behind), increase the size of the smallest icon and make the red snowflakes white but with black borders, similar to the beige circles I discuss in another post.



Steven Gray has done a similar experiment using old BBC snow icons (which I was talking about on Tuesday). The one in the screen shot shows snow tweets from the last hour. You can see from the northern most cloud just off the East coast that there are snow icons beneath the black clouds. These get lost against the pale map and could also do with black borders. Steven hasn't sized the icons to illustrate snow intensity which gives the map a different look.



Steven's also produced a thematic map on MapTube where he's colored postcodes by the average snow report over 24 hours. This is my favourite of the three, I think it gets around the problems of icons by combining them into a thematic layer. As with the icon examples though, it would benefit if the background data was paled out to enhance the blue shading.



Finally Ed Parson's had a nice visualization that I thought I'd publish as a KMZ. I've emphasised the SW of England in this view as snow there is really rare.

BBC Weather Maps

Southern England where I'm based is about to have a unusually big snowfall. Hellish for anyone trying to travel tomorrow but fine for me as I'm working from home for the next 3 days with a fridge full of holiday left overs :) I drove into London from Southampton today so I was keeping a close look on the BBC weather forcast to check I was going to beat the snow into town. Here's a screen shot:

BBC weather maps are clear and crisp. They also rate highly IMHO for using well chosen colors e.g: blue indicates frost and subtle blobs of white illustrate snow fall (on the forecast film clip only, not in the screen shot above). Where they don't do so well is thinking how these elements work together on the map. In the screen shot above we have a number of competing shades of blue all ilustrating different things:

A] Sea covered with cloud
B] Sea with clear skies
C] Frost
D] Snow
E] Background to a Temperature label (Deg C)
F] Rain



Poor Layer Coloring: They have 6 layers on one map plotted as different shades of blue! The map could be improved by replacing blue shading with the use of symbols, interestingly, this is how they BBC used to do it a few years ago (see above).

Related post:
I've criticised the BBC for overcomplicating maps of elections here.

GEarth Tours: Power first, Usability second

The most interesting thing I learnt last week at AGU was about the strategy Google adopted with tours. The Googler who built tours told me they were aiming for power in the first release rather than making them very usable. That explains the current situation where you can easily create a simple tour (my tutorial) , where complex tours which control the timeline and animate objects are possible (see latest Google Tour below)

but where producing such involved tours is very complicated because you have to write KML code directly. I suspect this means that Google are going to work on making tours much more user friendly in the future, as per the positive characteristics I noted about ArcGIS Explorer (AGX).

In my experience people get very excited about the possibilities of technology in teaching and often the benefits aren't realised because teachers are over optimistic about the difficulties and what can be achieved but I predict students will love producing complex presentations based on GEarth tours, it's going to be great.

AGU, Tutorial Use and Harrower on Icons

I'm just about to come home from the AGU conference. Its been really worthwhile making the trip, a lot of people I bumped into were complimentary about this blog which is good to hear as blogging is quite a solitary experience a lot of the time. An unexpectedly high number of people have also been complimentary about my tutorials (includes links to older tutorials) and said they use them with their students. That's also really exciting to hear and gives me pause for thought about what else to publish in the future.

I have lots of thoughts, links and ideas the conference which will go into blog posts in the near future, however a gem of a link was given to me by Andrew Turner; Cartography 2.0 is an online text book about developing interactive maps by Mark Harrower. I've just had a quick first read about user interfaces (UIs) and it's well worth devoting 10 minutes to. Most people developing maps won't be building UIs which is a fair bit of the discussion but his comments on icon design and his respect for 'Don't make me think!' by Krug are spot on.

Google Earth Tour User Test

This is the PowerPoint of the talk I'm giving at AGU today about user testing of Google Earth tours.

I tried to get it in a more friendly format but the flash version lost the notes.

I'm going to try and video myself doing it as well.

Free Climate Change Data?

Ed Parsons has written a post about making Global warming data free for the public. He makes the point:

"Unfortunately people have lost confidence is both politicians and I’m afraid scientists to provide unbiased analysis of data on Climate Change, perhaps we now need to better educate people as to how to look at climate change data themselves"

I'm not sure about 'lost confidence in scientists' but lets leave that for another day. I totally agree that releasing machine readable data is good for society as a whole (with some ethical exceptions) and I also agree with Ed that educating the public about climate change is crucial in the coming years. Ed continues:

"and to make this data available without spin or interpretation so that people can make their own minds up."

He then goes on to discuss EDGAR, a project to make geotagged emission data freely available. He ends:

"And before any climate scientists out there claim that this is ridiculous and that the general public cannot be expected to deal with such complex tools and concepts, ask a surveyor or cartographer if they expected that the general public would be building the only detailed global digital maps a few years ago ?"

The General Public and Complex Tools: I don't think Ed's analogy holds water. It's relatively easy for me to walk out of my front door, turn on a GPS, map some roads and upload that data to OSM to help 'build detailed global digital maps', because a knowledgeable community thought up the initiative and provided the framework by which data could be added (Muki's post expands this idea). In the same way, I can view visualisations of global warming data in Google Earth because scientists have collected data, processed it and worked out the best way to present it. I find it hard to believe 'bedroom scientists' will have the skills to do the same task. You only have to look at the EDGAR website to appreciate the skills and knowledge necessary. I have an MSc in Earth Science and I don't know details such as relative contributions of different global warming gases (e.g. methane, carbon dioxide and nitrous oxide) or whether CO2 from forest fires should be included and if not, how do you take it out of the calculation.

Part of the screen shot Ed used in his post

Climate Change in Google Earth: GEarth is a fantastic tool to inform the public about climate change issues but IMHO its not a lack of raw data that's the main problem, I think its a lack of understanding about GeoWeb usabilty. You need go no further than the screen shot Ed provides to see this in the EDGAR visualisation*:

• They've failed to show what the units are in the key
• They've used a palette of colours that is difficult to view if you are green/red colour blind
  
• They're showing global data on a Virtual globe - I can't compare the map of Australia with the UK in the same view.

All this gets in the way of what would otherwise be a very compelling story.

Putting my Money where my Mouth is: In a couple of weeks I'm going to present and help chair the Virtual Globes session at AGU , my paper will be on best practices in using Google Earth tours to communicate science (including climate change) to the public. More of that later of course...

*It may be someone in EDGAR has just thrown the data into GEarth as an experiment in which case my criticisms are a little unfair - I have no evidence this is a published project and I don't know what other materials are in the project apart from the screen shot. I *really* want to encourage them to publish on Google Earth and using other GeoWeb tools, I think they have a great story to tell but I'd advise they think about usability before they do.