Upcoming content on this blog and NZ Railmaps

Although there has not been much happening on this blog lately, there will be new content coming soon which is a republication of certain content that used to be provided on my older websites that are now closed. Some of this content has recently been archived for public viewing on NZ Railmaps but it is more appropriate to publish it on the blog rather than in that website and therefore I will be taking the opportunity to update it as well as making it available to a wider audience through this blog.

Lately the main changes in NZ Railmaps have been in the International section where I have expanded the coverage of Wales. There are now separate files for North, Central, West and South Wales. As the geographic definitions of this Welsh regionalisation are not very clear cut I have taken some suggestions from Wikipedia for the boundaries of these files. The UK mainland is very interesting from a historical railway perspective due to the large number of routes that were constructed, many of which are now closed today but of which substantial traces remain. Wales is a spread out country like New Zealand, with a similar population albeit in a much smaller land area, and due to the nature of population distribution and lack of intensive development many of the old lines have substantial remnants still easy to follow. There is also excellent Google Earth coverage, and a very good resource of old maps to trace out the lines from at the New Popular Edition project’s website, something not yet matched in New Zealand. There is also a file for Ecuador, but it is unlikely this will be updated much in future due to the lack of suitable coverage of that country.

I have determined that the standard for international maps is that all lines will be traced in the same colour and therefore their status will not be determinable from the files. Only the major features will be marked, significant bridges or tunnels where still remaining today. Furthermore, no effort is made to determine the names of any routes. These restrictions will ensure that the amount of work needed to draw these maps is kept in reasonable proportion given that these are countries which I know relatively little about and have not visited.

Map Updates Completed Recently

Since writing about Canterbury branch lines a week ago, most of the work completed has been on the Nelson Section, which has been thoroughly revised and is pretty well done for the moment, last updated today.

Due to the on-again, off-again nature of this project, there is going to be a pause in the mapping work from now for an indeterminate period due to other interests that I am spending more time in at present.

Mapping Canterbury branch lines

Since I wrote about the Otago Central, which I was working on the map of up to the first week in December, after a brief break due to end of year work pressures, I started updating the maps of Canterbury branch lines. Due to my file structure, this means the maps are in three files, being the MNL_Other, MSL_Canterbury_Other and ML_Other series representing respectively lines north, south and west of Christchurch. With work on the Midland Line file I also checked and updated the maps of the branches that are on the West Coast also. As part of the Little River branch is of course in a rail trail the opportunity was taken to restructure the presentation of the line in the same way as the Otago Central. Since I have not described this process before it is fairly straightforward and is in turn based on the way the East Coast Main Trunk file was restructured, which was purely on the basis of necessity due to the way the ECMT has changed so much over more than a century. In brief, in the past a line which had more than one use/status at different times, would have been structured into separate folders in the file for each of those uses with duplication.

For example, the OCR for most of its life was a 236 km branchline. Later on it became a 3.5 km industrial line, a 60 km heritage line, a 150 km rail trail and a 22 km closed section. Thus there would have been four separate folders (branch, industrial line, heritage line, rail trail) with an obvious duplication of information between the first folder and the other three. The new structure is to show the entire route only once in its current status, with documentation about the change in status of various sections, and a special “Closed Sections” folder for sections that are closed, including changes made to the original route as it stands today. For example the “Closed Sections” folder of the OCR file shows the Cromwell Gorge, the original Clyde station, Prices Creek deviation and three small deviations made by the Rail Trail. This structure obviously affects the Little River line and any other line that is still being used today in any form. The opportunity has been taken over the past month to use the new Spot coverage to bring all these branch line up to date starting with Southbridge, some of these areas also recently got Digital Globe cover for various areas, for example on Southbridge it has been possible to spot a few small bridges that are still in place.

Mapping the Methven line was interesting because it also has Street View coverage along about half its length. Leitch & Scott refer to “near Sherwood (16 km)…several massive bridge abutments”. Since I could not find any trace on the overhead view I used Street View to search around that area, which was unsuccessful. Consequently I believe L&S to be incorrect. Subsequently the remains of a bridge around 20 metres length was located at 43°45'30.48"S 171°57'25.67"E and Street View shows these remains clearly. Why would there actually be a bridge here? The dip in the road suggests there must have been a waterway here at some point, as far as the surrounding farmland goes it has all disappeared or been reclaimed. I am fairly certain this is the only bridge on the line that would fit the L&S description.

Also looked at have been Springburn, Fairlie, Waimate, Waiau, Whitecliffs etc. Fairlie has lots of small bridges or culverts still in place. As a general rule many early culverts look like small bridges because they have two abutments and a small “bridge” structure across the top. It wasn’t until more recent times that fully enclosed culverts became more common and on these branch lines the culverts were not replaced as they have been on main routes. Waimate was interesting in terms of redrawing the route through the Waihao Gorge which I decided had some problems particularly the latter section. In all routes the opportunity to add stations (approximate in most cases) has been taken where necessary as well as revising some route locations if I think that it might be justified. Looking at all the branches there are some abandoned bridges still in place on many of them particularly the Ross Branch in its latter section which has recently been revealed with full DG coverage. The Ruatapu-Ross section appears like a formed road but whether the numerous wooden bridges are still passable I cannot say obviously, so far searching Google has not turned up any websites with more info.

Work continues meanwhile all around the country, parts of the NIMT, PNGL and at the moment Wairarapa (including the two Rimutaka rail trails) is continuing. It’s disappointing though that NZTA has stopped publishing maps and additional information on roading projects on its website, instead you have to email a given address for this info. So I had to find out from elsewhere about the Waipukurau overbridge replacement location, and I haven’t managed to get any info as yet about the realignment around Matahorua which involves a new bridge over the railway line there.

Google Earth adds Spot coverage of North Island

A few weeks ago I wrote about the enhancement of Google Earth coverage of the South Island when Cnes/Spot imagery was added. Although it is only medium resolution, it basically displaces all that horrible low resolution Landsat coverage which has filled in the gaps of Google Earth for a long long time. The addition of this coverage makes it possible to think about completing most of the maps I have drawn because the improvement makes it possible to pick out additional geographic features that are otherwise impossible to determine.

One of my favourites in the North Island is the Kirikopuni Balloon Loop on the Dargaville Branch as seen here:

View Larger Map

There are many more Kirikopunis waiting to be opened up by the enhanced coverage, even as I noted parts of Lower Hutt still not covered by the urban footage that their council apparently hasn’t seen the light on yet. Still, it is now possible to see Waterloo station.

Major progress at NZ Rail Maps site

This is just a quick post to note that a lot of work has been done on setting up this blog, Picasa Web Albums but especially on the NZ Rail Maps site. Just about every rail line map is being gone over as time permits, though I am very busy and often it is quite cursory. However full advantage is being taken of the Spot coverage of the South Island recently added to Google Earth to update the maps, most of the rural branch line coverage in particular is being brought up to date, the Stratford Okahukura Line and nearly every other line in the North Island including parts of the NIMT have been looked at so far, ECMT etc etc. Just finished looking at the Kingston Line and some but not all of its branches but that is what has been completed to date in a short amount of time, it is a very busy time of year, we have just had a long weekend so there has been opportunity.

Elsewhere I see the mothballing of the SOL has been controversial, but I sincerely doubt that it can ever be viable in the future given its geographical disadvantages. Reality is coming home to roost in certain areas, though Auckland still seems immune to it with the rebuilding of the Onehunga Branch, the new line to Manakau, proposals for a central rail tunnel and harbour bridge duplication, electrification, and laying a line to the airport. Still I imagine the money will run out sooner or later.

One of the useful features to be added in the latest version of Google Earth is being able to access earlier coverage of areas. For example, when I looked at the old Rimutaka Incline route recently, I found there was extensive cloud cover over the parts of the old route between Upper Hutt and Kaitoke. But stepping back to earlier footage gets rid of most of that, letting me see things like the temporary siding that connected the old and new routes between Mangaroa and Maymorn.

In general at this time the revision of maps is mainly for missed features. I am not bothering with alignment corrections except where coverage resolution has been improved significantly. It is notable that there is a lot of high-res of most of the main centres now and Hamilton in particular is a big improvement. But it is ridiculous that part of Lower Hutt is still only covered by low res Landsat, and we await Spot coverage of the rest of the North Island. Alignment errors tend to occur every time coverage is updated, which can be often in some areas, and fixing these is a very time consuming process that I haven’t got time for at the moment. I would suggest the alignments might only be corrected every 5 years or so, and the main bulk of the maps perhaps annually.

Must rush but the updates of the web site will continue, gradually.

NZ Rail Maps Site Reinstated

Today I reinstated this blog and the associated NZ Rail Maps website. I used to have two rail related blogs and about half a dozen websites. Early February 2009 I decided to abandon a 25 year hobby of railfanning and take up new hobbies related to gardening and nature. This continues unabated at this point. However I have decided to reinstate this blog and the Rail Maps website. In the past nine months I have continued to pursue an interest in geography with the ongoing use of Google Earth, and have published a few articles on other blogs about rail geographical subjects. Since February there has been a significant improvement in Google Earth coverage, especially with the addition of Cnes/Spot satellite coverage of the South Island. This new coverage means that practically the whole South Island is now covered. This makes it possible to update all the unfinished maps. This will gradually be done over an unspecified period. I have done some small scale updating of a few maps, particularly the ECMT, Midland Line and its branches including Ross. So the maps will be available at the site again.

However, there are and will be significant differences in the way the maps and this blog are updated in the future:

• Blog postings will mainly be in relation to website updates. I am not writing a significant volume of other material related to this subject. Nor am I routinely undertaking research or investigation into rail related subjects
• Map updates are limited to improving existing maps when coverage improves. No significant effort will be made to try to identify points of interest that are not already marked. No completely new maps will be drawn.
• There will be no research in relation to maps. I don’t have contact with the railfan community in NZ as I am not a member of any groups associated with this community. Nor do I have time to look up other sources of information in relation to maps. Essentially the process of improving maps is limited to identifying features that can already be seen in Google Earth, rather than locating existing non visible features.
• There will not be a set timeframe in which maps are updated. Rather it will be done spasmodically as free time permits.

In addition, all other rail related websites I have produced in the past are not going to be reinstated. I have no ongoing interest in the other subjects that these websites covered. Nor will former postings to this blog or the old NZ Rail Maps blog be reinstated. Although it would be very easy to repost from source, I don’t have the source files stored anywhere and they were not backed up when the blogs were originally removed.

So that is how things stand. If you go to the website you will find some maps have been updated recently. That’s about all I have time for right now.

Mapping photos with GPS and Google Earth

Since I bought my first digital cameras four years ago, there has been widespread development in the availability of virtual globe software which lets us do many useful geographically orientated things, such as mapping points of interest and the locations of specific photographs of scenery. There are two specific services which are integrated with Google Earth for photolocation. These are Picasaweb and Panoramio. At this stage I have stuck with Picasaweb due to my dislike of having to duplicate effort in Panoramio, and its lack of sophisticated upload tools like the Picasa client software that Google produces. However, Panoramio pictures can be displayed as an integral layer in Google Earth. I hope that Google will introduce a similar level of integration of Picasaweb for those of us who prefer its capabilities so that we can submit photos directly from Picasaweb in our existing and new albums.

The main subject of this post is how to map photos using Google Earth and Picasa. In the beginning of 2008 I created my first mapped albums in Picasaweb, as the client software features the ability to geotag individual pictures. It does this by using Google Earth as a user interface to find the coordinates of a specific place which the user nominates as a photo location, and then using the EXIF standard it adds coordinate information to an existing JPEG photo. As my albums can incorporate up to several hundred pictures each, it takes a lot of time and effort to locate each individual picture. As well as that, you have to be able to work out where each picture was taken. On the train trips that I did last January, and some other trips that I have created other albums for, because my knowledge of the routes was very good, I could work this out without too much trouble. But on a recent trip I did on the Waipara river, I couldn’t work out where everything went because I didn’t know the area.

The easiest way to get around this problem is to batch synchronise data from a GPS unit with photos. A $200 handheld GPS from Garmin will automatically create a track of locations over a period of time while it is turned on. This track can be downloaded to a PC and then using appropriate software the timestamps of the GPS trackpoints can be synchronised to specific pictures using the timestamps that your digital camera embeds into each picture that it takes. The software can then save this information in a file that can be imported into Google Earth to display the photomap.

A recent trip to Quail Island was the first trial of this system for me. The only real issues encountered were technical limitations in Google Maps and Google Earth, and the GPS’s batteries going flat. My equipment was:

• Garmin eTrex H handheld GPS. This is a high sensitivity, basic unit at the lower end of the price and spec range. It is a rugged water-resistant design, and the high sensitivity means it is better to use indoors (such as inside a moving vehicle) when environmental conditions make it more difficult to receive the faint signals from the GPS satellites.
• Garmin serial PC data cable. This connects the GPS to the serial port of a PC. Your computer will need to have a 9 pin serial port (the old RS232 style). A lot of modern computers do not have these ports and you might also need to purchase a USB to serial adapter at extra cost if your PC is in this category. Garmin still provides only the RS232 interface on its most recent lower end units. This is a lot slower than USB, but it works satisfactorily and the cable can be connected and disconnected on the fly.
• GPS download and geotagging software. You do not need to buy this from your GPS manufacturer, as typically like the data cables, such products are unnecessarily expensive. A lot of GPSs are supported by third party products. I used the free EasyGPS software to download the tracks and geotag, which it does by comparing timestamps.
• A geotagging compatible web album or software system that can display your tracks and photos automatically in their correct locations on a virtual globe. I use Picasaweb, Google Earth and Google Maps as needed.

First thing is to get your GPS set up. The eTrex automatically records the track by default as long as it is switched on. Make sure the batteries have enough juice in them to last for the whole of your trip. Secondly, ensure that the clock(s) on your camera(s) are very accurately set. Get a time signal from the radio (or somewhere) and sync the clocks to that. The more accurate these are, the more accurate will be the locations of photos that can be determined by the geotagging software. Then, on the day of your trip, just turn on the GPS before your trip starts, and take your photos. Simple :)

Once you’ve returned, download photos to your PC and the track data from the GPS. Then, select the photos you want and carry out any additional processing needed. I use IrfanView to batch resize my photos and add a copyright caption which also shows the date and time the picture was taken. For public web display I resize to 960x720. The next step is to geotag. In EasyGPS you simply tell it to add photos to the correct track. This causes it to write the coordinates of each photo directly into its EXIF headers, by matching GPS and photo timestamps. I had no problems with this as I synched my cameras’ clocks within 10 seconds of GMT. The GPS gets its time automatically, of course. You just need to make sure it knows what your timezone is so it can adjust the timestamp automatically to your local time.

Once finished geotagging in EasyGPS, save the track as a GPX file for use with other software. GPX is the open GPS XML format for exchanging GPS data. My next step is to open the GPX file in Google Earth, which the current free edition has the ability to do. This lets me edit the tracks to remove parts that are irrelevant to my map. I discarded the picture information from the GPX as it is static to the hard drive locations of my PC and therefore not much use on the internet. Using Picasa client (version 2.7 is recommended), I then captioned each of the photos on my PC and then bulk uploaded them to a new Picasa Web Album of my choice. This causes Picasaweb to automatically create a map of the locations of all the photos of the album. I can then download this to my PC (using the “View in Google Earth” link) and open it in GE. I then combined the photo map from Picasaweb with the GPS tracks that I imported from my GPS to create the final map of my trip. That trip was then imported to Google Maps to allow it to be embedded for display in a blog posting. All of the thumbnails on this map will automatically link to Picasaweb to display the photo from my web album.

Quail Island trip posting with map

Digital Photography School’s geotagging how-to

Stacy and Brian’s geotagging how-to