Bushwalk Australia

Hey folks. Just a quick post about a new Budawangs topographic map I've published this year, hopefully of interest to Budawangs gurus on this forum. It's a 1:30,000-scale map with 20m contours, although there are 10m auxiliary contours wherever it's not steep. Printed on a lightweight synthetic stock, double-sided, so no worries with a bit of rain or rough treatment.

I used 2015 and 2018 lidar surveys to derive elevation (as well as a vegation/canopy underlay which might be of interest to bush-bashers). The map squeezes in as much as possible, covering everything south of Nerriga Road and north of Long Gully and Pigeon House trailheads. You could use this map for just about every Budwangs walk going; Wirritin Ridge is about all that's missing. (Though there's an inset for Currockbilly Mountain.)

A link to the map, including a legend and a few detail closeups: https://nswtopo.com/#budawangs

I'm a self-taught cartographer, starting with rogaining maps years ago. I've been doing digital topo maps for a fair while now. This is my second printed map. Still honing my craft but I'm very pleased with how this one has turned out! Paper maps are a bit anachronistic these days, I suppose, but so rewarding to make, especially when the terrain is so distinctive. Hope you guys like it!

(Thanks to admins for OKing this post.)

Well, there's a coincidence! Earlier today I followed a link from John Evan's blog to your nswtopo budawangs page.

Paper maps certainly still have their place (and my Corang & Endrick sheets are certainly showing their age). The sample pics look really good.

What's the source of the contours? and the hydrology? and the tracks? I compared bits around Styles Creek with andrewp's img map and I think your contours are smoothed a little. I guess you have 20m contours so as to not clutter the map? 10m would be better, I think, but might be hard to read on a 1:30,000 map.

I also note you have a track (marked disused) heading N off the Corang Plateau. I've just walked this - still some evidence of the footpad on the ground but hard to follow, that's for sure. I have only seen this on the Budawangs Sketch Map so am surprised to see it here.

Good luck with the project.

So contours and hydrology are all from lidar survey data, of which NSW now has exceptional coverage. In its raw form this is known as 'point cloud' data, and is astonishingly detailed! But it's also provided in a processed form as a DEM (digital elevation model): basically, a raster image with each pixel representing a corresponding elevation on the ground. Very high resolution, each pixel covering about one square metre.

Contours are created from the DEM. With resolution being so high, raw contours come out pretty jagged and ugly, so I apply some judicious filtering to the elevation model to produce slightly smoother contours, as you noticed. There's always a tradeoff between accuracy and readability. As for contour interval, the base interval is ten metres, which is useful and necessary in flat areas. But yes, in steeper areas they just get too dense. So I thin out every second contour in steeper areas to get 20m intervals. And then again in cliff areas, where contours basically break down as a metaphor, they're knocked right back to just the index contours at 100m intervals. Relief shading, vegetation change, etc then take over to depict these parts of the terrain. It's pretty tricky to get right in areas like Budawangs and the Blue Mountains, since there is such a large variation in terrain steepness. Yeah, I spend a lot of time on it!

Watercourses are also calculated from the DEM. It's really fascinating! Software calculate the slope for each pixel on the DEM. Then, basically, it simulates an amount of water being dropped on each pixel, and calculates how it accumulates as it flows down slopes and into gullies and valleys. So the watercourse lines align really well with contours, being generated from the same underlying data. For wider water features (rivers, ponds, the reservoir etc), I wrote a tool which extracts these areas directly from the raw lidar data. (The laser is absorbed by the water rather than being reflected, so there are gaps in the point cloud where the water is.)

Tracks are from all over. I started with the NSW topographic data (DTDB), but supplemented with OpenStreetMap data and Google Earth checks for roads and vehicular tracks. For walking tracks, I harvested GPX tracks from various sources on the web to double-check positioning, and make corrections where necessary. With the old disused/overgrown tracks and pads, it was a judgement call as to which ones to include. Even the really important ones can be overgrown (like the Vines-Styles track at the moment). As you noted, there are labels to indicate the more tenuous ones. And a lot of deliberation about what to do with the Corang Lagoon track, with the closure, but it exists on the ground so I put it in with appropriate labeling.

Thanks for letting me nerd out about maps for a while!

mholling wrote:Thanks for letting me nerd out about maps for a while!

No problem! and thanks for the detailed info - I can imagine the many hours you've put into it.

Very cool! Great to see the art of cartography still alive. Looking forward to receiving my map

My copy arrived yesterday (ordered Sunday, so very quick delivery). The map looks great. Huge coverage for a single double sided sheet. Paper stock finish looks as durable as I've seen. On first impression level of detail looks good and useful, maybe cliffs a bit harder to discern than other topos.

Thank you. I can't comprehend how much work would go into producing something like this.

Any tips on pen choice for the synthetic stock?

Thanks so much for the vote of confidence! Much appreciated. The synthetic stock doesn't hold pencil too well, it will smudge a bit, as will a felt-tip pen. But a ball-point pen should do the job.

The map is awesome. Excellent job. Once again thanks so much for the copy.

Chev wrote:Any tips on pen choice for the synthetic stock?

Large stationers sell Chinagraph pencils, which are made for writing on glass, metal and synthetic plastics.

My copy also arrived shortly after ordering. What a great job you've done Matt. Your catographic style really works well (using both 10 and 20m contours, not showing cliffs). Nice use of the veg mapping from the lidar! I also like the effort you've gone to in making it a physical thing vs just a digital map, not very common these days! I'd be keen to see how many you sell over time.

The lidar stuff was really fun. Not only can you get elevation from lidar, but canopy height too. In open areas you can see the structure of individual trees, it's amazingly detailed. In this case I extracted canopy height and used it to create the vegetation texture.

Offset printing is also pretty interesting. The unit cost becomes pretty affordable once you order a decent quantity. There's another map I've noticed recently, the Topograph Snowy Mountains map, which is priced a lot higher, so I'm guessing they're doing shorter/on-demand digital printing runs. More expensive but less money up front. If I can cover costs in 3-4 years I'll be happy.

Glad once again to be getting positive comments. It's a bit hard to evaluate your own work sometimes! The bar's not that high IMO, I never found the government maps particularly nice to read. And I always thought the Budawangs could do with its own map (sketch map aside). From my digital maps, I know that the 8927-3N Corang sheet is *really* popular.

Got mine on Friday just before heading up to Cooyoyo and the Castle with a group of seven scouts. Was great to use for some navigation training while on the track! I also spent quite a bit of time showing it off to another bushwalker at a very busy Cooyoyo. Great work! Agree it is so cool to have a large format map that looks great and feels durable.

Nice work on the map production!

mholling wrote:For wider water features (rivers, ponds, the reservoir etc), I wrote a tool which extracts these areas directly from the raw lidar data. (The laser is absorbed by the water rather than being reflected, so there are gaps in the point cloud where the water is.)

Might be getting into too much detail for this post but would you be willing to share any more info on your methodology of extraction of wide water features from the lidar data?

I've done a fair bit of documentation around extracting streams from DEM data (eg https://maps.ozultimate.com/wiki/qgis_hydrology), but the processes start to break down around wider water features!

Despite beiing fairly firmly fixed in Tassie now, I couldn't resist getting a copy of this. It's an excellent map; the scale is perfect and the detail works well. Time to organise a mainland trip, methinks; I've only ever done a few very short daywalks in Morton, a very long time ago.

tom_brennan wrote:Nice work on the map production!

Might be getting into too much detail for this post but would you be willing to share any more info on your methodology of extraction of wide water features from the lidar data?

I've done a fair bit of documentation around extracting streams from DEM data (eg https://maps.ozultimate.com/wiki/qgis_hydrology), but the processes start to break down around wider water features!

Thanks Tom! To get the wide river areas and lakes (e.g. Porters Creek Reservoir) I use the lidar tiles, which will exhibit 'voids' where there are water areas. (The laser reflection off the water, if any, is specular and won't generally produce a lidar return.) My water-area tool collects all the lidar points, thins them a bit, then triangulates them. Where there are voids, the triangles are long and skinny, so it detects and outlines these areas. (There are other causes of lidar voids, mostly occluded areas below cliffs, so it also checks that the outlines are align horizontally.) It works pretty well overall although rivers can look a bit odd when they are skinny. The Delaunay triangulation was difficult to code (C++) and resource-heavy since it churns through hundreds of millions of points!

Your map just arrived and I'm really impressed with it! Might buy the Larapinta collection just incase - I do love maps!

Any thoughts on what you might try doing next? Perhaps the Southern Blue Mountains/Kanangra/Nattai?

Have you thought about doing a hybrid map with current cartographic data but historical names, like adding the sketch map names (budawang, dunphy,gundungurra etc)?

Just a thought!

TomPom wrote: Perhaps the Southern Blue Mountains/Kanangra/Nattai?

Have you thought about doing a hybrid map with current cartographic data but historical names, like adding the sketch map names (budawang, dunphy,gundungurra etc)?

Just a thought!

I'd welcome this, could even extend it to adding known passes, camping caves etc, but I wonder (with absolutely no qualification whatsoever to talk on the matter) if that would constitute editorialising as opposed to simply presenting cartographic material in a user-friendly format?

But yeah a southern blueys/nattai (or even greater ettrema/shoalhaven!) map would be fantastic

Glad you're happy with your map @TomPom. I might not pursue any more printed maps for a while, as I need to get a better idea on how long it will take to recover costs. If I were to do another, I would love love love to do Snowy Mountains/Namadgi maps, since that's the area I know the best. It's a very popular area, but a bit of a crowded map market already (with the SutMap and Topograph maps).

I wouldn't feel qualified doing Blue Mountains maps as I haven't been there much for a long time now. Such a vast area, and all the canyons there would also be pretty challenging for mapping; I think @tom_brennan has experience in that regard.

As for informal and historic placenames, that's a bit of a can of worms. In the Budawangs I found references to a number of informal placenames, all the way down to names used within bushwalking clubs. Where do you draw the line? I generally just stuck to officially-named features in the Geographical Name Register, with a handful of exceptions. I decided to avoid labeling the camping caves because the NPWS Budawangs Strategy document specifically requests they not be marked on maps. I think collected wisdom like what the sketch map contains, or track notes, are not quite so important these days to be included on a map. Back in those days the map was the main source of information, but now we have the internet and Google Earth, where you can do as much pre-trip research as you want.

That's fair enough! (to both mapping and the informal place names).
I do love cross referencing the sketch maps with the topos.

The snowy compacted down to a 1:30000 scale would be great, I think future AAWT walkers would thank you as well.

Totally on board with keeping sensitive areas off maps, there's enough publicising that sort of think already unfortunately!

I've done a bunch of downloadable topo maps of the Blue Mountains for canyoning. These were done with an earlier version of NSWTopo. You can find them here: https://ozultimate.com/canyoning/maps.htm

I do plan to do some more, but it's very much for the fun of it, particularly in this day and age when many people just navigate off their phone . Probably more of the canyon areas - there are a few that are still missing.

If anyone is interested, there is also a custom 1:25000 map for the Mt Wilson/Mt Irvine Rural Fire Service, which has a lot of additional information that is not on the standard maps. You can buy them from them: https://www.mtwilson.com.au/information ... neous-maps

On an only vaguely related note, Matt - I assume you're not taking issues on GitHub for NSWTopo any more? I notice that the Issues tab is no longer there.

Yeah the software is in a bit of a funny place. It started off as a tool for downloading maps from internet servers, which was more difficult to do back then. It still works for that, but nowadays I mainly use it for my own mapping with local data (shapefiles etc) on my computer. (It's probably best used by someone who can write their own layer files.) I don't know about the issues tab but I didn't ever provide any support for the software—it maybe wasn't very wise to publish the code and create that expectation. The internet servers where map layers come from are also frustratingly prone to change, so a layer can suddenly stop working because the server was reconfigured. (The transmission lines layer being one example, as you noticed!) But I hope it's more user-friendly to use now, with good command-line help and so on.

Hopefully now that NSW has expanded its Spatial Portal, many of the layers that were prone to breaking should now become more stable! It certainly looks like that's the intention of the portal.

It's always tricky with open source software that you've primarily written for personal use. General support issues for NSWTopo tended to get raised (and assisted) in the original thread that you posted on bushwalk.com (https://bushwalk.com/forum/viewtopic.php?f=36&t=9028). Most of the problems people were having were config and setup related. I was thinking more actual code issues. That said, the code issues I've had with the latest version appear purely related to how Chrome runs under Windows, and I know you don't really test for Windows! I documented my steps in case it was useful for others at: https://maps.ozultimate.com/wiki/nswtopo - I'd try and see if I could work out a Windows fix and open a pull request, but Ruby is not a strength of mine!

In the end I bit the bullet and installed Ubuntu under Windows which got around the Chrome issues.

The Help documentation is very good (excellent compared to the average open source project!)

And great that you pushed a new version of both nswtopo and nswtopo-layers to rubygems.org, as that seems to have fixed some dependency issues I was hitting.

This is such a great venture, Matthew! I've just ordered one. This sort of product will help keep single track walking going in Australia. And it's even paper :0. Thanks for your efforts.

No worries @jtmiske, thanks for giving it a try and I hope you like it!