Category Archives: Open Data

UK Revealed: A New Web Map for Exploring the Terrain of the UK

Through applying novel terrain rendering techniques to the UK Ordnance Survey’s Terrain 50 dataset, a new web map has been produced allowing users to explore the physical features of the landscape without the distraction of cultural features. This follows on from the Geoscience Australia case studies, improving the tile-stitching methods, applying a test colour scheme and including some additional geographical features.

Wharfdale
Wharfdale and Littondale, Yorkshire Dales National Park

The resulting prototype tileset covers some 40,000 km sq. at a resolution of 50m, including the 100km grid squares NY, NZ, SD and SE. This area contains 5 National Parks and a varied geology including Chalk Uplands, Glaciated Landscapes, Limestone Pavements, Sandstone Tablelands, Vales and Moors.  Place names from the Ordnance Survey’s Vector Map District are brought in at appropriate zoom levels to orient the user and intertidal foreshore features are used to define the coastal region. The main design feature of this style of map is the addition of a vertical offset to create a 2.5D surface rendering whilst maintaining planimetric accuracy. The inclusion of the key visual cues explored in the Blue Mountains Revealed project are as described in prior posts to this blog. The image above illustrates an area of Upper Wharfdale and Littondale, Yorkshire Dales National Park, with an areal photograph below as a landscape comparator. The stepped limestone terraces are picked out well with prominent landscape features being clearly demarked.

Aerial Littondale
Aerial Littondale

The workflow used was similar to the previous prototypes: render and composite the raster tiles, re-project using gdalwarp for use with common web mapping libraries, use tilemill to add coastal vector data and rules for place-name inclusion and finally, render out the tileset and embed using Leaflet.js. There are still some registration errors with the addition of the foreshore feature-set from Vector Map District, and some problems with the varying of sea level height across adjacent terrain tiles.  The image below illustrates an effective small-scale overview of the contained valley of Swaledale in The Yorkshire Dales National Park.

Swaledale
Swaledale, Yorkshire Dales National Park

As MapBox expand their terrain rendering options (similar to the vector experiments on this blog and Whelan (2011)) and Ordnance Survey explore visualisation styles, the depiction of terrain in web-mapping platforms is being increasingly informed by a broader design agenda. The following images illustrate some areas of the UK Revealed case study, with the final image providing a link to the hosted tileset.

moors
The North Yorkshire Moors National Park with geomorphological features clearly depicted
pennines
The Central Pennines, West of Leeds, seen without their usual dense overlay of cultural features
three_peaks
The Three Peaks of the Yorkshire Dales
TheNorth
Link to the hosted prototype tileset.

References:

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

 

Testing in Tasmania

To test the rendering algorithms developed in the Blue Mountains prototype, a larger case study with a more varied geomorphology was needed. As I was in Sydney at the time and wanted to be able to explore the landscape of the case study for myself, a short trip down to Tasmania seemed like an obvious choice. The island is rich in significant geological sites and offers a diverse range of surface features, many of which are described in detail in the book “Created From Chaos” by Peter Manchester.

The photographic set “Tasmanian Landscapes” documents a brief trip around the island in late December 2013. It was summer in Australia, but only 12 degrees in Hobart with a slight flurry of snow encountered whilst crossing the Gordon River Dam. The set opens with the flight down the East Coast, capturing a great view of the granite peaks of The Hazards, shown below.

The Hazards, Tasmania
The Hazards, Tasmania

After a trip up the River Derwent from Hobart, the coastline of the Tasman Peninsula was explored, followed by The Hazards and Freycinet National Park, this time from the ground. The terrain then opened up through the Midlands, with views west towards the Central Highlands. The trip then concluded by travelling to Mount Field National Park and along the Gordon River Road into the South West Wilderness.

South West Wilderness, Tasmania
South West Wilderness, Tasmania

Elevation data for the study was extracted from the Geoscience Australia terrain portal to cover the whole of the main island of Tasmania, resulting in a dataset of 14426 by 12371 points.  No changes were made to the rendering software since the Blue Mountains prototype and, after processing the data, the resulting raster image was 43137 × 36998 pixels.

The following images show some interesting screen-captures from browsing the terrain. The prototype map interface can be found by clicking through on the last screen-capture (please note that there is a tile-stitching error running North/South from Turners Beach to Cox Bight).

Area to the North-West of Lake Saint Clair, Tasmania
Area to the North-West of Lake Saint Clair, Tasmania
Lakes Gordon and Pedder near Gordon River Dam, Tasmania
Lakes Gordon and Pedder near Gordon River Dam, Tasmania
Ben Lomond, Tasmania
Ben Lomond, Tasmania
Great Western Tiers, Tasmania
Great Western Tiers, Tasmania
Mount Field, Tasmania
Mount Field, Tasmania
Cradle Mountain, Tasmania
Cradle Mountain, Tasmania
Tasmania Tasmania Map Interface - Click on the image to be taken to the interfaceTasmania Map Interface - Click on the image to be taken to the interface Interface
Tasmania Map Interface – Click on the image to be taken to the interface

Compared to existing web mapping services, the rendered tileset conveys the form of the land with a stylistic, yet accurate and revealing depiction. From a cursory analysis, the balance between continuous mass and structuring line compliment each other well. Notable deficiencies are in the allocation of grey-scale values. The range of higher elevations over the case study is in small areas meaning that darker shading from the hypsometry layer is influencing a large percentage of the land area.  Further analysis will follow in a future post, as will the next phase of this project, working with the Ordnance Survey’s UK elevation open data product, Terrain 50.

I have worked extensively with prior Ordnance Survey elevation products, interpolated from manually drawn contour lines. The new dataset is derived from LIDAR scanned data and automatically generalised. I am interested to see how the rendering algorithms work with the new dataset and I include a reference here to existing investigative work carried out by Dr. Christopher Green at the University of Oxford analysing this data.

 

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.

Visualization and Open Data

So, I have now been back in Sydney for two weeks, have taken up a position of Visiting Scholar at the University of Sydney, (Design Lab, Faculty of Architecture Design and Planning) and have commenced teaching on two undergraduate courses. It is the first week of the academic year, the campus is very busy and it is 30 degrees.

Today, alongside Elmar Trefz, a co-lecturer, I was talking with third year Information Visualization students about balancing aesthetics and communication in visualizations. I used these examples from a static, terrain visualization project I undertook some time ago to illustrate a this.

Pencil Sketch of the Lake District
Automatic Pencil Sketch of the Lake District National Park, UK.

The image is created through selecting depth discontinuities in a high-resolution z-buffer and weighting image-level pixels according to a threshold.

The data used in the production of these plots was, at the time, subject to licencing constraints from the provider, Ordnance Survey, UK. The data is now released under a creative commons licence. This opening up of publicly funded datasets has created many new opportunities for developers and designers. However, there are still some reserved datasets that I hope the release strategy will one-day reach, especially the 1:10000 elevation data.

I am looking forward to seeing what datasets this year’s Design Computing students discover and link to drive their final projects.