Category Archives: Blue Mountains

Blue Mountains Revealed: Prototype Production

Following on from exploring the layer combinations, described in the previous post, some final changes were made to the base raster from which the map tiles will be split. Firstly, to facilitate geo-referencing the base grid, an equal vertical and horizontal grid offset was applied at the rendering stage. This produces a map-like view that has the vertical aspect enhanced through the Linear Framework. Secondly, a reduced tonal range was used for the hypsometric tinting; this improves the clarity of the lower elevations, however this needs further work so as to enhance the contrast. The size of the resulting raster for the dataset area was 9055 × 13224 pixels.

The image was geo-referenced using gdal_translate to create a GeoTiff which was imported into MapBox’s TileMill to create an MBTiles SQLite file. This file was then expanded using mbutil to create a directory structure of tiles to be rendered through a client-side mapping library, in the prototype instance, Leaflet was chosen. It should be noted that no generalised elevation datasets were used to create the smaller scale zoom levels. In the case of this prototype, the down-sampling of TileMill is the only process applied. An investigation into generalisation across zoom levels would improve clarity and user experience. The image below links to the prototype map interface, this is not optimally hosted, so some of the tiles may take a few moments to resolve as you move through the zoom levels. This version is purely concerned with terrain browsing, subsequently, no linkage to cultural land-use or human settlement/constructed features is made.

Map Interface
Map Interface: Click to Open

 

 

Rendering Mass with a Structuring Framework

To effectively communicate the form of terrain, a number of visual cues need to be rendered so our brain can construct an internal representation from the image presented to our retina. The aim of the Blue Mountains Revealed project was to achieve this through balancing clarity of depiction with an illustrative/cartographic aesthetic. This post will illustrate the separate cue layers that go into the final rendered Blue Mountains tileset. For a boarder review of relief depictions techniques please see Whelan (2011) or refer to Cartographic Relief Depiction, by Eduard Imhof.

The rendered tileset uses a combination of mass (continuous shading) and line to suggest and delineate features. Surface altitude and slope, both object-based visual cues are represented through a continuous tonal shade whilst a view-dependent line-set is created from detecting edges in a specially rendered z-buffer (described in an earlier post).

Frontal breaks of slope are suggested by the continuous shading of terrain slope, whereas rear-facing breaks of slope are highlighted through the linear cues.  A hypsometric weighting orders the surface into a visual hierarchy. No interpolation across the shaded areas was applied and, as previously mentioned, the low resolution data patches have had an impact on the slope rendering. These artefacts in slope can be seen in Fig 6. The various layers are shown below along with a combination of the slope and the linear framework.

 

Refs:

Imhof, E. (H. J. Steward (Ed)), 1982, Cartographic Relief Representation, Walter de Gruyter

Whelan J. C., 2011, Web-based Vector Terrain Exploration, Proceedings of SVG Open 2011, Cambridge MA

The Blue Mountains Revealed: A change in rendering location

Up until now my Blue Mountains project has been exploring browser-based vector terrain rendering.  This has been building on my interest in high-resolution SVG terrain exploration since I presented a platform at The IEEE Theory and Practice of Computer Graphics a little over 10 years ago (Whelan and Visvalingam, 2003). As there was no native browser support for SVG at the time, rendering was performed through a plug-in from Adobe, with the SVG being produced on demand by a Java Servlet.

As browser rendering and processing capabilities improve, client-side rendering of terrain (and other map elements) have featured in popular web mapping environments. Apart from the development of realistic 3D rendering through Google Earth and Google Maps, the recent experiments from rapidly expanding start-up, Map Box, provide some interesting cartographic examples (see: Vector Terrain, Dynamic Hill Shading). Equally the vector rendering of map data through libraries such and PolyMaps and D3, and the experimental work documented by Michal Migurski place dynamic, vector rendering firmly in the focus for future mapping interfaces. This is especially true on the increasingly high-resolution display surfaces available or where complete stylistic freedom is required.

Whilst these developments are pushing forward this field, some of the rendering layers I have been exploring in the Blue Mountains Project are, at this stage, not suited to browser-based rendering. Therefore to ensure responsive interaction across platforms, and given the relatively static nature of the terrain data, the project is going to take a different direction and look to pre-render and serve terrain tiles to the client. An off-line tile cache will also be available for mobile applications while out of data coverage. Details of the render layers and prototype tile-server will be posted soon.

Ref: Whelan J. C. & Visvalingam M., 2003, Formulated silhouettes for sketching terrain, In Theory and Practice of Computer Graphics 2003, 90-96.

 

We are only as good as our source

The creative commons elevation data sets available from Geoscience Australia offer an improved product compared to the data sensed by their source provider, NASA.

The 1 second (30 meter resolution) SRTM derived datasets (which can be downloaded from the National Elevation Data Framework portal) are those used in the Blue Mountains Project being documented on this blog. The origin of the dataset is from a shuttle mission flown by NASA in 2000 to scan the surface terrain of the earth at 30m resolution.

However, the nature of scanning the Earth’s terrain at an angle of 45 degrees results in some problems, most notably where there are steep cliffs. The raw SRTM data have had a number of algorithmic techniques applied to provide an enhanced product for Australia and this is well documented in the excellent user guide produced by Geoscience Australia (Gallant, J.C., Dowling, T.I., Read, A.M., Wilson, N., Tickle, P., Inskeep, C. (2011) 1 second SRTM Derived Digital Elevation Models User Guide. Geoscience Australia) [11.6 MB PDF]

Whilst exploring certain key areas of the Blue Mountain data in the project visualization platform it can be seen that some of these corrective techniques have introduced some peculiarly high elevations over the area of the cliff face along the Grose Valley near Blackheath. This is also documented in Gallant et. al. (2011) p. 37 [11.6 MB PDF].

Data Errors in Grose Valley NSW. Data: (c) Commonwealth of Australia (Geoscience Australia) 2012; Image (c) John Whelan (2013)
Data Errors in Grose Valley NSW. Data: (c) Commonwealth of Australia (Geoscience Australia) 2012; Image (c) John Whelan (2013)

 

Data Errors in Grose Valley NSW. Data: (c) Commonwealth of Australia (Geoscience Australia) 2012; Image (c) John Whelan (2013)
Data Errors in Grose Valley NSW. Data: (c) Commonwealth of Australia (Geoscience Australia) 2012; Image (c) John Whelan (2013)

This level of error will clearly effect any surface dependent visualization and analysis. Areas of the valley floor have also been inserted from data derived from a lower resolution source, so, how to improve this area remains a question.

When sensing and measuring we are only as good as our measuring tools and when auto-rectifying and processing we are only as good as our algorithm design. When a trained human is in the loop, in the case of manual contour interpolation, a number of problem points are still present, but a high-level of intelligent adaption and cartographic insight is also there. This semantic understanding applied to automatic generalisation and visualization is a still a challenge in digital cartography.

The Blue Mountains Revealed

In October 2011 I presented a paper at the SVG Open conference (Cambridge MA) describing a web-based terrain visualization interface. This exploratory prototype was an illustration of how a web browser can render a high quality vector-based depiction of an area of land that the user has chosen from a familiar 2D map interface. Various depiction styles were explored and the ability to move between differing zoom levels implemented.  The prototype used a UK (Ordnance Survey) mapping API and data recently released under the UK’s Open Data Initiative (headed by Sir Tim-Berners Lee and Prof. Nigel Shadbolt). The aim of the project was to reveal the form of the land in the most aesthetically pleasing and communicative way (and to test browser rendering options).

Given my current location in Sydney, the recent release of Geoscience Australia’s elevation data set, under Creative Commons Licence, and the popularity of the Blue Mountains National Park near Sydney I am busy exploring future developments of this concept.  This project has timely significance given the fact that 2013 is the bicentenary of the first recorded European crossing of the Blue Mountains by Blaxland, Lawson and Wentworth in 1813, an event significant in the development of the Sydney area.

Below is a raster screenshot of an initial output, rendered in the browser (Google Chrome) from a set of vectors. As this is parsed set of vectors, they can be restyled and transformed on the client side or printed locally at a high resolution or large format.

The view is of a large area centred on Mount Solitary. The image shows projected profiles overlaid with a set of vector cross-sections tinted with respect to distance.  This aerial perspective is similar to the approach adopted by cartographic depictions of terrain such as the incredible panoramas in Baedeker’s 1893 guides shown below.

Source:  Baedeker K. 1893, Switzerland, Dulau and Co.

Further test outputs can be seen on Flickr as can photographs taken on a research trip earlier in 2012. More outputs from the project will be posted here until a project site is established.

 

Physical browsing and passive discovery

I am a big believer in stumbling upon things; being in an environment where something catches your eye that you were perhaps not exactly looking for, but which takes you off to discover other things / stories. This is different to active search. This is great to do by physically wandering around areas of a new environment with particular attributes, say, the cultural quarter of an unfamiliar city. I have also been keen to investigate how digital technology can assist, but not remove, the element of serendipity in discovery. Sometime ago I supervised an Honours Dissertation looking at mobile, passive resource discovery, (Stabeler, 2006), and I revisited this with another Honours Student (Harding, 2011) who looked at digital navigation across physical objects in similar and differing contexts. Both interesting, digital, geospatial search prototypes.

When I arrived in Sydney, my first exploration of the area I am staying in was to walk down the main street (Norton Street in Leichhardt) and to call into various shops and cafes. After walking into Berkelouw Books, I realised there was an extensive second-hand department upstairs. As I have started research for a terrain visualization project related to the Blue Mountains National Park (the project itself came from a serendipitous discovery in book from 1815, shortly before I left the UK), but did not have any fixed view on what I was looking for yet, I thought I would see what I could find.  One of the first books I saw was “Landscape Art and the Blue Mountains”, by Hugh Speirs, this turned out to be a good reference to start with when looking at the history of the visual depiction of this region.

The human search engine of physical browsing and passive discovery is a useful tool when it comes to making unintended but often useful discoveries. There are still many interesting areas to explore when looking at improving search in a digital geospatial context.