A quick post just to highlight my first attempt at creating a GIS related tutorial. This tutorial looks at the creation of 3D models in ArcScene, modelling in Google Sketch up and finally bringing the two together.
This tutorial provides a very basic introduction to this functionality however should be fairly easy to expand upon. If you have any questions please send them through the comment box below.
So here it is. The culmination of 16 hours and several hundred contributions. The UCL map of London has provided a fascinating insight into how people remember, communicate and translate spatial information. Before I write up the full project, I suggest you start to explore some of the contribution and come up with your own conclusions as to the success of the map.
Full screen available here (Higher resolution image to follow)
Inspired by the work of Stephen Walter in his collection of hand drawn maps I have put forward a proposal to conduct a mental mapping exercise, creating a hand drawn map of London by the students and staff at UCL. Though I foresee a number of potential challenges I believe the potential outcome could be hugely impressive. I have included below the project proposal and would be grateful for any insight that may be offered.
Following the work of Stephen Walter who produced the famous London Island map, I would like to propose a similar project which was built upon the knowledge held within UCL. The foundation, being an outline of Greater London and the content being provided on a voluntary basis by members of the university. The work, then being attributed towards those who contributed.
Feasibility/Requirements: Due to the nature of the project the map canvas will need an area of high footfall with reasonably easy access. The project would be run over 7 days allowing access to the majority of interested parties. In addition, to document the production of the map, a webcam would be used to record the development of the map at set time intervals (5 to 10 minutes). In order to promote participation a completion element could be included.
Objectives/desired result: The goal over the week period would be to create a fairly comprehensive map of London combining the knowledge of many participants. The end product would be a digital reproduction which could be displayed in a web format allowing users to explore the map through panning and zooming. In addition the time-lapse data collected would provide insight into how people transfer mental knowledge from mind to paper. For example, do major networks i.e rivers and roads develop first and then the map is filled in around these?
A number of headlines have broadcast recently featuring the introduction of Google’s new indoor street view product, but is this really the answer? The product is aimed at creating a more comprehensive product and has been trialled in a number of museums previously as part of the Google Arts program. The museum project is an excellent use of the street-view technology and allows easy navigation around the museum and view/bookmarking of art pieces of interest.
So there would appear to many benefits to the idea, but, what are the downsides? Firstly, how often are the images updated? The majority of street-view imagery was collected within the last year, and while this may be suitable in slow changing environments is it going to be suitable for the rapidly changing retail market. For example, retailers appearing to stock a product now unavailable and consequently appearing out of date.
Secondly the present system requires user to access the service through Google places rather than through the current street-view interface. This does reduce the smooth transition between services although there is some speculation that this is just a matter of time with a critical mass of indoor-view services available.
I look forward to hearing people’s own views as to this development.
A new discovery for me, StatPlanet by http://www.sacmeq.org/ provides an excellent visualization platform for a reasonable number of per-loaded datasets. The standard GUI provides a number of supporting statistics including temporal change. The temporal change graph is placed above the data an works intuitively displaying the line graph based upon the current mouse position. The included feature ability to select a feature and have it remain graphed for later comparison helps to improve visual data extraction.
Screen Capture of Statpack map (source http://www.sacmeq.org/)
The StatPlanet package is one of a number provided by http://www.sacmeq.org/ and includes a number of graphical data visualizations packages which may be used to display aspatial data with equal interactivity.
The StatPlanet further increases its value through the ability to export the viewed data, not only as images but also the underlying data allowing more personal research to be conducted on the data. All in all making StatPlanet and excellent data discovery and principle testing service.
The country is bleeding. This title was first used on the flowing data Blog when displaying unemployment figures in the United States (available http://goo.gl/iMnH2). In the case of the flowing data map the unemployment data is displayed on a standard projection of the states.
US Unemployment (source Flowing Data)
In my opinion the layout of the United States makes the data fairly easy to visualise as very crudely the states could be seen as a rectangle. The UK however does not have this luxury and as a result visualising similar datasets cannot be performed so easily.
When looking to carry out a similar process on the UK using data from the Guardian data Blog and Wikipedia. In order to better visualise this data it has been structured in such a way that it may be displayed using a rectangular cartogram. The data is divided firstly by region and then electoral constituency before visualising unemployment using a similar colour ramp to that of Flowing Data. The tree map produced in R does a good job of displaying the data though it would be beneficial if spatial integrity could be included. In addition to this each constituency cell is labelled only by number and congruently either a lookup table or call-out function would be required to explore the data further.
Mapping Zombies… Though not the first idea that comes to mind when considering the power of Google. This map highlights some of the potential of the service. The map was created by Dr Mark Graham, Taylor Shelton, Monica Stephens and Matthew Zook of the Oxford Internet Institute.
Zombies Mapped. Source: Oxford Internet Institute
The map creators make the observations that the concentration of searches are reasonably well correlated with the locations of English speaking countries. An interesting further development may be to identify the corresponding words in a number of different languages in order to look at the trend on a more language inclusive level.
The Oxford Internet Institute provides many more excellent data visualisations including a large number of tree maps, twitter and Google visualizations. Below I have included an example of a tree map create by the OII.
Treemap of Academic Knowledge and Language. source Oxford Internet Institute
Throughout my education, cartograms have been one of my favorite forms of cartographic visualization. Cartograms work on the principle of deforming data based upon an external factor. In the first example, the boroughs of London have been altered based upon the mean equivalent income. As can be seen the City of London region has been significantly increased whereas Havering has been reduced.
The above example was created using Quantum GIS and the ‘Cartogram Creator’ plugin. The data used are the London Stat pack provided by the London DataStore and joined to a London boroughs shapefile from the same store. The cartogram projection used in the example is the Dougenik, Chrisman and Niemeyer cartogram. The cartogram is presented alongside a correctly presented map of London to help with comparison. A purple colour ramp due to the associated of the colour purple with wealth.
If you have any comments or suggestions of good cartograms or other maps please post them into the comment box.
This week, Aidan Slingsby of City university,presented his work on visualization of demographic information,through the OAC explorer. The application uses the OAC census classification scheme to categorize the data and uses a selection of cartograms andcolours to disseminate the maximum amount of information whilst not congesting the viewer.
OAC explorer rectangle cartogram (CityGI centre)
The screenshot above is an example of therectangular cartogtram used to more easily visualize the data. Thoughthere is significant deformation the postcode scheme allows a user to veryeasily navigate through the data down to a specific post code region. One smallissue with this is that the spread out nature of Scotland means ithas been reduced to only the top 2 rows.
One of the many impressive aspects of the map isthe use of colour. In order to create an equal viewing experience for allcategories the colours have been chosen to have equal hue so there is nodifference in colour lightness. The application has made excellent use of thisby allowing lightness to be applied when a user wishes to highlight particularcategories of interest.
During the final year research project, one of the more successful outcomes of the project was the construction of a web mapping application designed to visualize the suitability of land for shelter camps following rapid onset natural disasters. During this project I communicated with ShelterBoxUK and MapAction to look at where existing services exist and at challenges which may be encountered when working in these environments. The application as demoed in the attached video was constructed using only open source software. The reason being, the platform needed to be available, not just following a disaster but also between events to allow user to gain technical competency, identify and fix faults and finally to increase awareness.
The spatial data was hosted on a virtual server running GeoServer (An open-source map server capable of supporting WFS, WMS, WCS and is beginning to introduce WPS(Web Processing Services)). In order to access the spatial data stored on the server OpenLayers is used to display the data and apply the styles as specified in the SLD (Style layer descriptor file). The desktop processing element of the project was conducted using quantum GIS. QGIS has the ability to link to WFS servers and can also export the symbology created into the SLD file so the symbology may be replicated across the system.
Though the cartography used within the application is not great it does highlight clearly the areas of the highest suitability vs low as specified in the UN Sphere project minimum standards in disaster management. One of the interested results of the demonstration has been the growing number of views the video has received on YouTube (1314 currently).