Friday 30 December 2016

Project update

Well since I wrote in my last post about rearranging my two Flickr accounts, I have been busy - I went away to Dunedin for a few days over Christmas - but I have now completed a monumental task of downloading almost 200 individual albums from the enzedrail Flickr albums and uploading them to the kahukowhai albums. The upload totals show that approximately 12200 photos were uploaded to the new albums. The reason there were so many albums (a great many of them contain only a few photos) is that the rail photos were classified by location and accordingly, there are a lot of locations on the railway network in NZ. They are all grouped at the end of the albums list under the RAIL: prefix, but there are about 180 albums in that classification as notable albums have been moved into the list of general albums. Supposedly there are nearly 50,000 photos now on that site and only 2700 on the NZ Rail Maps flickr site (enzedrail).

The outstanding work that was waiting to be completed when the photos were over on enzedrail will now be pushed ahead and that is to go through all the albums and add the contents to RAIL: albums where the locations are known, this will result in more complete album contents with all of the locations for every photo being better known. In the main part the Transport <year> series of albums are the ones to be gone through to arrive at these locational classifications because a lot of them haven't been classified in this way up until now. But also there are some rail albums that are for different things like locomotive classes and they haven't been pulled out of the other albums yet. This is a lot easier to do with all the photos in one place instead of being spread across two different Flickr accounts as they have been.

Meanwhile work goes on with the maps themselves. As it was known that there were some bugs in 2.17 version of Qgis and as there is a new 2.99 master version I decided to upgrade the MainPC installation to 2.99 which has happened. But although one issue with the rule based styles is gone from this edition, there is a new problem when adding records to a table which contains a value map in one of its fields because the values in these fields keep getting trashed and reset to Null when you save the new records. I updated at the same time my VM to 2.18 from 2.16 and found the rule based style editing problem still happening. So I cloned this VM and removed Qgis completely then installed the LTR version, 2.14.10, as there was really only the choice of 2.14, 2.18 or 2.99 to install.

So at the moment for example I have been doing some detailing of the MSL with adding the km pegs and changing the bridge details to show the bridge numbers. Well this is tricky because to add km posts, I have to use 2.14. To straighten up the routes I have to use 2.99. So I have to run 2.14 in a VM and 2.99 on the computer itself, both with the same project and data files open, although it is a different version of the project file because as usual the later version upgrades the project file to a new format. Kind of messy but it's what I have to do at the moment until the project developers fix the add records problem in 2.99.

Thursday 22 December 2016

Best way to share maps online

As this project evolves, there will be two main ways that maps will be published:
  • As volumes online (CreateSpace)
  • As individual maps
Currently the individual maps are stored on Flickr as it appears to have the best system for displaying the images that make up the maps.

I have two Flickr accounts, one for rail photos (enzedrail) and one for general photos (kahukowhai). In order to keep with these two accounts and the best usage of the resources, I plan to proceed by using enzedrail only for the maps, and kahukowhai for everything else. This actually makes a lot of sense, since the maps really need their own site, while the other rail photos should be amalgamated with the rest of my photos.

As of today enzedrail's photo albums will be reorganised. This has been done with the map albums being renamed with the region they are in and grouped by the region. So that looks quite good, and I have uploaded some new map sets as well.

It will take quite a while to shift the photos across to kahukowhai.Fortunately I have found a free FlickrDownloadr application to download the albums one at a time and then they can be uploaded one at a time through the upload page. Since Flickr decided to charge for their Uploadr tool, I was tempted to sign up for a Pro subscription for long enough, but have decided to manually upload as the downloader only brings down one album at a time anyway.

Wednesday 16 November 2016

NZ Rail Maps From Here

As most will be aware this blog has not had a post in about 2 months, and in fact I have not done any mapping activity in that time. For the most part I have needed some time to look at where this project is going, and how it fits in with other new interests I have developed in that time.

In the last couple of days I have reaffirmed my interest in developing the Otago Central Railway volume of maps which has had a lot of activity over the past two years. As noted in a post to the project's FB group recently this is the highest priority of development at present but it will go on probably without significant aerial photography usage as I had done previously. Part of this is that the increased cost of getting photography from Archives New Zealand, whilst still significantly cheaper than what we have been able to get from other sources, is conflicting with other financial imperatives for me at this time. 

So the main resource to be used for the ongoing development is the chainage charts and possibly some recent LINZ aerial imagery where desirable to pinpoint specific locations of interest but not the historic photography except for what I have already obtained (Clyde-Cromwell).

At the same time as developing the Otago Central volume I expect to push the volumes for the rest of the country into some sort of completion, this is because Otago Central is a high numbered volume in the series of volumes that cover the whole country and it would make a lot of sense to have the lower numbered volumes completed first before releasing the Otago Central volume. The lower numbered volumes will not receive any revision at this time, except for km peg and bridge numbering for currently open routes.

So that is a plan for the next couple of years roughly and the key intention is to push this project towards "publication" of some sort of a completed set of volumes for the whole country within a 2-3 year timeframe.

Wednesday 3 August 2016

Railmaps of Dunedin

Well I have been pretty quiet for the past week and that will be the pattern from here on in with occasional updates as time permits. Although in this case a lot of that is because of changing computers around as well. It was decided to get more use out of the secondary PC seeing as it is actually the newer and faster one, so I am still partway through the job of having given it the three head video card and third screen, and the extra memory (currently it has 16 GB and the older computer has 12, but this is because of the older computer's memory not working in the newer computer, so I intend to add another 8 GB to this newer computer when the opportunity permits, to ensure it has plenty of memory). Rearranging all the files etc is still in the process of happening due to the amount of work involved.

I have been working on maps of the Dunedin area and these will soon go into the Otago-Southland maps project so that they can be published as the earlier MSL series did. At the moment I am working on Merton, Seacliff, Waitati, Sawyers Bay, Port Chalmers, Ravensbourne and Dunedin. 

Here is a photo from Google Earth in Waikouaiti.

Just north of the township you have what looks like and in fact is an abandoned piece of highway complete with the centre line marked in and here you can see the closed level crossing. In the background you have the current highway that was once linked into this other old highway via this level crossing. 

The other end of that old highway is a bit further north here at Tumai.

Here we have Bridge 197, the old Tumai Overbridge over the MSL and in the background the new section of highway that bypasses this twisty section over that narrow bridge and also eliminates the level crossing shown in the previous picture. The main issues were in fact at that overbridge with its narrowness plus the two sharp curves on its approaches, and at the level crossing.

The highway realignment involved 3.5 km of new highway and cost $7 million to build. It was opened by Minister of Transport Annette King in September 2007. Projects like this rarely happen in the South Island these days as National has focused on sucking as much money as possible into the RONs.

Tuesday 12 July 2016

Global Contributions and Sources

The global contribution and source list has been published here. It will be updated as necessary.

Global source list

Currently the blog is getting a global source and reference list assembled. This page which will be published sometime later this week, will acknowledge all of the source contributions made to this project.

Monday 27 June 2016

MSL Part 5, Onehunga, Palmerston North

Well now that I have checked to see what aerial coverage exists in Canterbury Maps I have been fortunate to discover full coverage of Rakaia in 1974. I am now considering mapping the rail yards as the quality of this is very good. Burnham and Rolleston were also well covered in the same run of aerial photography. There is actually still a lot of work to do with the aerial coverage around the Christchurch area. At the moment I am still continuing to work my way south and not actually sure when I will do these areas.

There has been a lot of debate today about the Onehunga Branch because of the dropping of a proposed extension to Auckland Airport. This brings into focus the truncation of the branch when it ceased to be a freight line to the port. I will have a look at historic Auckland aerial photography to see if there is enough to draw the full Onehunga Port yard as quite a lot of tracks are still visible on the wharves there.

With a bit more work some aerial footage of Palmerston North in 1945 has also turned up and shows quite a bit of detail of the old central railway route so that is another thing to look forward to in future.

Sunday 5 June 2016

Wairarapa Line [0A]: Volume 6 Progress Update 1

Well after a few days work I have completed maps of the Wairarapa Line. There are 95 individual maps covering roughly 170 km of line. The scale varies from 1:2222 to 1:3333 for the Wellington-Speedys Crossing section (55 km) which has a lot more detail in it as it includes all of the suburban stations, branches and deviations that occurred along the Hutt Valley and Rimutakas route. This 55 km takes more than two thirds of all the maps. From Featherston to Woodville the map scale is reduced to 1:9999 with only two exceptions - the realignment of the track at Bridge 80, around the 99 km peg, at 1:2490 and the Woodville map at 1:6666. One important change in the data layout of the maps which was taken (aside from the earlier reclassification of milepegs as "stations") has been to combine bridges and tunnels into a single table. The reason for this which is particularly relevant for the Wairarapa Line is that subways under the track tend to be classified as bridges by Kiwirail even though we have used a tunnel (subway) symbol for them and thus put them in as tunnels. So they will use the subway symbol but will have a B series number on the map. I intend to implement this change in all of the NZ projects (6 for the whole country) but will not be rushing them all through as I did with the milepegs change. One useful thing done with the milepegs change was to implement the new styles as global styles so that they are at least accessible across every project instead of having to duplicate the style changes in 6 projects.

One detail omitted from the Wellington maps has been the ferry terminals with their linkspans which are classified as bridges by Kiwirail. This detail will have to be left until a future time, such as when the NIMT gets done.

My attention has now turned to the Main South Line from Christchurch to Invercargill. This will take in a number of stations in Christchurch where I have previously drawn historical yard layouts from the Canterbury Maps footage. When I changed the maps data layout in a major way last year and combined a lot of separate tables together, unfortunately a lot of the job of pulling the yard layouts into the new tables was not completed for some reason. Consequently this is one of the jobs that now has to be done because this data is notably absent from the maps and should not be. For example in Lyttelton where the MSL starts, I put together the maps only in March last year and all the data was in the maps that were published on the Flickr site, yet I have still not imported it into the new data tables. And Lyttelton is far from being the only station in the Christchurch suburban area where this is the case. In fact I am sure in all the yards that will be the case. The new tables were created in which to store the historic yard layouts but the data still has to be put in. So this one is actually quite a big task which is going to take several weeks to complete.

The major reason that is now emerging for the mess is that the version of Qgis that I used at the time would not paste table data properly and often filled in nulls for a lot of columns in the data tables. This means I am having to painstakingly transcribe some of the pasted data. And this also explains why some of the pastes required have not already been done. They were left until now, at which point the Qgis problem has largely been fixed. After a lot of work pulling together data from a lot of tables I have managed to pull the Lyttelton yard maps together except for the current yard which has eluded me and may have to be put back in from an original source, hopefully I have it already drawn in KML. 

Then finally it will be back onto the Otago Central as I have let that one slide since last year. Part of the reason is I have had a lot of personal stuff to deal with this year, but it has been worth waiting to get the chainage charts that I have just received.

UPDATE: Well I managed to find everything I needed for Lyttelton eventually after scrabbling around for a few more "old" data tables. And then I started marking out the details heading towards Heathcote. Once getting the map done as far as Heathcote, again some details had to be picked up from old tables and put into the current table structure. I have now got to Woolston but in the main these old details are just buildings/structures and siding locations as it seems some of the track data for stations other than Lyttelton was completed previously in the new data structure. Things have been a little slow lately but with new commitments in some other areas I will have to accept that there will be a lesser pace during term time. One good thing earlier this year was finding a suitable software package to automatically sync the project's Google Drive as Google does not produce its own client for Linux. This required a small license fee to purchase (OverGrive) but it seems to be doing a proper job. 

Tuesday 31 May 2016

Recording mileposts

Milepost recording on NZ Rail Maps has been going for some months, maybe a year and a half, and the positions come from the Kiwirail Alcam data. Mileposts have been recorded up until now as features, with the distance recorded in the caption field. It has now been decided to store them as stations. Six new fields have been added to the stations tables, being two sets of three fields. When a milepost is recorded, it can have two sets of attributes if it is a zero peg or some other peg that records the attributes pertaining to two different lines. An interesting example that is not a zero peg is found on the first few km out of Wellington where the mileages for the Wairarapa Line and NIMT run side by side on the same physical tracks. However the Alcam map does not show anything for the Wairarapa Line until 1.8 km, while other sources confirm the duality of mileposts from 0 km. 

The reason for putting the data into the Stations table is that the same labels formula can be used for both mileposts and stations, which allows distance data to be recorded for stations as well. The label formula looks like this:

trim(concat("caption", ' ', "Distance",' ', "Unit", ' ', "Label", ' ',"Distance 2",' ', "Unit 2", ' ',"Label 2", ' '))

In simple terms this is outputting a string that consists of seven of the nine fields of the table joined together, with spaces between them. Provided we only put data into fields where it is actually needed, for example the units and labels are left null where the distance has not been filled in, then the label will be formatted correctly. The design of the fields allows for older type mileposts that are actually in miles or chains to be input for lines where these were the known distances. On old S&I diagrams signals and interlocked devices outside station limits were named using a number made up from miles and chains, and these can also be input using this system as long as the correct units are stated.

The reason for changing the data storage from the features table to the stations table was that the means of displaying the data was reviewed. Previously a typical milepost caption looked like ...k. where the dots represent numbers and the k represents kilometres. For example 187k5 was the caption with the letter k being where the decimal point would normally be. The problem with this representation is that while 187.5 km is clear and unambiguous, 187k5 could be either 187.5 km or 187.05 km, although it was always assumed that it mean the former value. Converting the data to actual numbers also makes it easier to sort and in theory should make it easy to put the data into a table in a hardcopy or electronic publication. The distance fields are stored as a maximum of 7 digits with 4 digits after the decimal point, but when they are fed into the concat function as above, trailing zeroes are eliminated.

A milepost is not "current" or "former" as it is not necessarily a physical representation. However signals will be current or former.

Monday 2 May 2016

Road/Rail Bridges on the Hokitika/Ross Line [2]



The second location of a combined bridge is at the Taramakau River as mentioned above, being Bridge 13, between 14 and 15 km. It is lattice girder steel with six spans each 121 feet long (total length around 220 metres). It is founded in the bed on steel cylinders which must suggest a heavy flow is a common feature of this river.


Because the bridge is on the main highway to Christchurch it has tended to get a lot of attention. Obviously while New River was easily dealt with, Taramakau has attracted more than 80 years of promises; for example in 1965 both this bridge and the one at Arahura were going to be replaced within five years. A couple of ideas that didn't make it were an extra lane on top and on the sides, because of insufficient strength in the foundations. Still, a clip on cycle/pedestrian lane was attached late last year for the West Coast Wilderness Trail even though the new bridge has been planned for building later this year. There was a bridgekeeper based there for many years, whose job was to close gates at each end to keep road traffic back, but this position must have come to a close with the reduction in trains as there is currently only one or two per day. 


The gate at the south end is a prominent feature of this 1987 photo along with the bridge keeper's hut. The bridge was recently repainted red after being in this silver paint scheme for many years. The clearance shown as 3.44 metres in my picture was later increased to the present 4.60 metres by raising the top of the bridge and the overhead wires were also removed.

Here is an aerial view from Land Information New Zealand around five years ago.


NZTA says the new bridge will be about 30 metres downstream (to the left of this photo) and will also cross over the railway line with an overbridge to the south.

Road/Rail Bridges on the Hokitika / Ross Line

Road/rail bridges in New Zealand have been and are of three types: 
  • Single shared deck. The most common type but also the most difficult to operate as it disrupts train and traffic equally when one mode must have exclusive use. Bridges either had bridgekeepers with gates at each end to close off the traffic, or else trains were subject to a 10 km/h maximum speed in order to ensure there was no hazard to other types of traffic given the length of some of the bridges. Although several of these bridges are still in place around New Zealand, there are only three currently active including the one at Camerons mentioned below. When that is converted to rail-only use, probably next year, the two on the Taieri Gorge Railway will be the last remaining operational in New Zealand.
  • Double decks. Three were built in the early part of the 20th century and all of them had the road on the lower deck and the train on the upper deck. In all cases the road deck was for a single lane of traffic. All three still stand but none have both decks in use for road and rail. The Awatere River bridge on the MNL was converted to rail only use in the early 2000s. The Ongarue River bridge on the Stratford Okahukura Line has only been used by road traffic since the line was mothballed about five years ago; and the Ohinemuri River bridge at Karangahake has had the upper rail deck removed some 25 years ago and replaced by a pedestrian bridge as part of the Hauraki Rail Trail.
  • Parallel decks. Two were built in the early part of the 20th century and a third was constructed about three years ago at Arahura as mentioned below. These bridges have substructure (piers) that are shared between two separate decks (superstructure). The only real disadvantage compared to separate bridges is the need to achieve a similar alignment and height in the road and rail. The first two were respectively at Inangahua Junction (still dual use, with road since the mid 1920s and rail since about 1940) and Westshore (built around 1930, currently only the road deck is used and only by pedestrians, although the rail line is still in place). 


The branch line from Greymouth to Hokitika and Ross dates from 1879 when work first began but it didn't get to Hokitika itself until 1893. Then a few years later in 1901 it was decided an extension further south was a good idea and the line was opened to Ruatapu in 1906. The final section to Ross was finished in 1909. This remained open for 71 years until the railway south of Hokitika closed in 1980. The line to Hokitika as it remains in use today has therefore been serving its district for the past 123 years. Although there are sidings nowadays at Kumara and Houhou, as far as I know the only source of traffic for the line is the Westland Dairy factory right at the end. The siding there predates the factory and was probably built to reach a nearby sawmill.

The line originally had four combined bridges of which only one remains today, at the Taramakau River. A replacement road bridge at this location is going to be built starting later this year but the original will remain in use by trains. This article documents the four bridges. One of the original four bridges at Arahura was replaced by a new parallel deck combined structure.

As the railway heads south from Greymouth the first combined bridge location is at New River between the 12 and 13 km pegs. This is Bridge No. 11.


My historical notes (compiled and published 22 years ago) said it was bridge 12 but as that location can be seen further south, this must have been mistaken. Apart from that the only information I got was that the bridge was originally built with five spans with a total length of 204 feet (about 60 metres). The new road bridge was built alongside and opened in December 1937. By taking a look at the Google street view coverage we can see that the bridge photographed at that time (2009) appeared to fit this description, with it having one very short span at the southern end, three Howe Truss spans and another span (supposedly the longest) at the north end. Both the visual inspection and the Google ruler cast doubt upon the length mentioned above so it is possible there is another error there or else one of the ends was filled in at some stage with the overall length possibly more like 50 metres. I understand the bridge was rebuilt sometime between 2010-2013 and therefore no longer has the Howe Truss spans upon it. This has been the case with a number of bridges on the line since Kiwirail was renationalised in 2008 and there are now not remaining any Howe Truss spans anywhere on this line, or supposedly anywhere in the Kiwirail network.



The second location of a combined bridge is at the Taramakau River as mentioned above, being Bridge 13, between 14 and 15 km. It is lattice girder steel with six spans each 121 feet long (total length around 220 metres). It is founded in the bed on steel cylinders which must suggest a heavy flow is a common feature of this river.


Because the bridge is on the main highway to Christchurch it has tended to get a lot of attention. Obviously while New River was easily dealt with, Taramakau has attracted more than 80 years of promises; for example in 1965 both this bridge and the one at Arahura were going to be replaced within five years. A couple of ideas that didn't make it were an extra lane on top and on the sides, because of insufficient strength in the foundations. Still, a clip on cycle/pedestrian lane was attached late last year for the West Coast Wilderness Trail even though the new bridge has been planned for building later this year. There was a bridgekeeper based there for many years, whose job was to close gates at each end to keep road traffic back, but this position must have come to a close with the reduction in trains as there is currently only one or two per day. 


The gate at the south end is a prominent feature of this 1987 photo along with the bridge keeper's hut. The bridge was recently repainted red after being in this silver paint scheme for many years. The clearance shown as 3.44 metres in my picture was later increased to the present 4.60 metres by raising the top of the bridge and the overhead wires were also removed.

Here is an aerial view from Land Information New Zealand around five years ago.


NZTA says the new bridge will be about 30 metres downstream (to the left of this photo) and will also cross over the railway line with an overbridge to the south.

The third example of this type of bridge was until about three years ago found at Arahura River. It was originally constructed with seven 80 foot Howe Truss wooden spans and another seven much shorter plain spans just under 15 feet, giving it a total length around 200 metres. As far as I know it never had a bridgekeeper. The location is around the 31 km peg and it is Bridge No. 28.


The foundations of the original bridge were partly wooden piles and partly steel cylinders. Reconstruction brought along a new bridge which has the separate road and rail decks sharing the same piles in the riverbed. It is thus New Zealand's newest example of a combined bridge. The location of the original railway bridge is occupied by the present road bridge deck. One of the Howe Truss spans has been preserved in a reserve area adjacent to the south end of the bridge.


In this 1987 photo it can be seen some of the plain spans were at the south end with their wooden piles, while the first of the Howe Truss spans is resting on some iron cylinders. Probably these cylinders were placed in order to allow for where the water flow was heaviest in the riverbed.

Here is an aerial view of the old bridge from the Kiwirail ALCAM maps.


The replacement structure as seen in LINZ's aerial photography. A roundabout for traffic was added at the south end to combine a level crossing with a T junction. 


The fourth of the bridges was immediately south of Hokitika and crossed the Hokitika River. This bridge was easily the longest of the four, having 16 80-foot spans and 76 20-foot spans, totalling around 850 metres, making it one of the longest railway bridges in New Zealand. The road across it was always considered to be a local road and the biggest problem for NZR was getting enough money from local authorities to pay for the upkeep of the road deck. It is interesting that the original highway bridge dating from 1878 in wood and replaced 1938 in concrete was placed at Kaniere, some 3 km upstream, where the length was able to be halved. One assumes the pros and cons of building the massive length at Hokitika itself was well canvassed prior to construction. The winning tender to build the bridge was confirmed in 1902 at 26,541 pounds, which today is equivalent to about $4.6 million. By 1909 NZR was demanding 150 pounds in annual maintenance contribution from ratepayers ($24,000) but a poll of the same ratepayers resulted in the Hokitika Borough Council offering 90 pounds ($14,000). By the time the bridge closed NZR could only demand $300, which was the equivalent of 150 pounds at the time decimal currency was introduced, and not at all the same as 150 pounds in 1909. This was clearly manifestly insufficient for such a long bridge in 1980.

Because of the cost wrangles, the bridge was one of only a few road and rail bridges that for all its life had transverse decking with the rails raised above the deck level. Think of it as a deck made up of railway sleepers with the rails sitting on top. In order to cross, a road vehicle has to straddle the rails. This is difficult with vehicles that have insufficient clearance underneath as some sports cars would for example. This problem was dealt with on most other road rail bridges where the money was available by laying an extra layer of longitudinal decking planks on top of the sleepers which brought the deck up to be virtually flush with the top of the rails, as you can see on the photo of the Arahura bridge above. 

The arrangement of decking led to the locals calling it "the longest xylophone in the world" as the sleepers would rattle under the wheels of passing vehicles making quite a lot of noise in the process. By the early 1970s the road deck was in pretty bad shape but NZR was still only getting $300 a year while spending more than $7000. In 1977 the Ross Branch was proposed for closure due to the high costs of bridge maintenance; the actual date was set for November 1980 and $10,000 was spent patching the bridge up to keep the line open. The road deck of the bridge was closed to traffic in 1979 as it was by then unsafe. Tenders were called to demolish it in 1982 and the work was completed two years later. The TV film "Bad Blood" apparently used this bridge to film a scene of a police car heading south from Hokitika in 1981 (set 40 years earlier). 

In the early 1990s NZTA decided to build a new highway bridge in the same location as the original and this appears to have opened around 1992/3. This resulted in State Highway 6 being relocated onto a new route between Hokitika and Ruatapu. South of Ruatapu the highway follows its original route. The new highway reused the railway formation from Hokitika to the Mahinapua Creek bridge, 














Saturday 30 April 2016

Road/Rail Bridges on the Hokitika/Ross Line [1]

Road/rail bridges in New Zealand have been and are of three types: 
  • Single shared deck. The most common type but also the most difficult to operate as it disrupts train and traffic equally when one mode must have exclusive use. Bridges either had bridgekeepers with gates at each end to close off the traffic, or else trains were subject to a 10 km/h maximum speed in order to ensure there was no hazard to other types of traffic given the length of some of the bridges. Although several of these bridges are still in place around New Zealand, there are only three currently active including the Taramakau mentioned later in this series. When that is converted to rail-only use, probably next year, the two on the Taieri Gorge Railway will be the last remaining operational in New Zealand.
  • Double decks. Three were built in the early part of the 20th century and all of them had the road on the lower deck and the train on the upper deck. In all cases the road deck was for a single lane of traffic. All three still stand but none have both decks in use for road and rail. The Awatere River bridge on the MNL was converted to rail only use in the early 2000s. The Ongarue River bridge on the Stratford Okahukura Line has only been used by road traffic since the line was mothballed about five years ago; and the Ohinemuri River bridge at Karangahake has had the upper rail deck removed some 25 years ago and replaced by a pedestrian bridge as part of the Hauraki Rail Trail.
  • Parallel decks. Two were built in the early part of the 20th century and a third was constructed about three years ago at Arahura as mentioned later in this series. These bridges have substructure (piers) that are shared between two separate decks (superstructure). The only real disadvantage compared to separate bridges is the need to achieve a similar alignment and height in the road and rail. The first two were respectively at Inangahua Junction (still dual use, with road since the mid 1920s and rail since about 1940) and Westshore (built around 1930, currently only the road deck is used and only by pedestrians, although the rail line is still in place). 


The branch line from Greymouth to Hokitika and Ross dates from 1879 when work first began but it didn't get to Hokitika itself until 1893. Then a few years later in 1901 it was decided an extension further south was a good idea and the line was opened to Ruatapu in 1906. The final section to Ross was finished in 1909. This remained open for 71 years until the railway south of Hokitika closed in 1980. The line to Hokitika as it remains in use today has therefore been serving its district for the past 123 years. Although there are still sidings nowadays at Kumara and Houhou, as far as I know the only source of traffic for the line is the Westland Dairy factory right at the end. The siding there predates the factory and was probably built to reach a nearby sawmill.

The line originally had four combined bridges of which only one remains today, at the Taramakau River. A replacement road bridge at this location is going to be built starting later this year but the original will remain in use by trains. This article documents the four bridges. One of the original four bridges at Arahura was replaced by a new parallel deck combined structure.

As the railway heads south from Greymouth the first combined bridge location is at New River between the 12 and 13 km pegs. This is Bridge No. 11.


My historical notes (compiled and published 22 years ago) said it was bridge 12 but as that location can be seen further south, this must have been mistaken. Apart from that the only information I got was that the bridge was originally built with five spans with a total length of 204 feet (about 60 metres). The new road bridge was built alongside and opened in December 1937. By taking a look at the Google street view coverage we can see that the bridge photographed at that time (2009) appeared to fit this description, with it having one very short span at the southern end, three Howe Truss spans and another span (supposedly the longest) at the north end. Both the visual inspection and the Google ruler cast doubt upon the length mentioned above so it is possible there is another error there or else one of the ends was filled in at some stage with the overall length possibly more like 50 metres. I understand the bridge was rebuilt sometime between 2010-2013 and therefore no longer has the Howe Truss spans upon it. This has been the case with a number of bridges on the line since Kiwirail was renationalised in 2008 and there are now not remaining any Howe Truss spans anywhere on this line, or supposedly anywhere in the Kiwirail network.

Sunday 24 April 2016

Stations on the Springburn Branch

Canterbury Maps has aerial coverage of Springburn, Mount Somers, Cavendish and Anama at present. It is possible this could be added to at a later date as they keep expanding their coverage. 

I have looked at the possibility of getting aerial scans of the remaining stations from Archives New Zealand, however I have had to rule this out because it is all small scale stuff. In fact Canterbury Maps' 1956 coverage that I have used for the above stations is at a scale of 1:18,000 which is impossible to work with on a contact print. And in fact on the CM site it is too hard to see all of the details especially for Anama. This is also probably the reason why some of the later coverage of the area is also at insufficient resolution, for example the early 1980s coverage.

When I drew up the maps of various parts of railway stations around Christchurch I relied mainly on coverage from the 1960s and 70s which was sharp enough to show individual railway tracks. In particular, the 1973 coverage of Christchurch Station was extremely sharp. I have cross referenced the survey number from that and found it was taken at a scale of 1:10,000. A lot of other, less clear coverage, was taken at much smaller scales. It is no coincidence because I believe Canterbury Maps is scanning off contact prints the same as I have sourced from Archives New Zealand. The problem is that the use of small scale imagery is widespread in a lot of areas of New Zealand. It costs a lot to get the original negatives scanned and that cost is impossible for me to achieve. 

Here in any case are maps of the four stations (plus one) which I believe are accurate (in some cases drawing on additional resources) but which may contain some errors because the aerial coverage is insufficient to confirm all the details exactly.


Springburn, based on 1956 aerial with additional diagram from 1930s. When the line opened to Mount Somers in 1885, that was intended as the terminus and it was not until 1889 that the line was extended to Springburn. The four houses are for locomotive running staff, not traffic staff. The locomotive depot for the line was located at Springburn but the lack of traffic there saw no need to have a stationmaster. In later years with a reduction in train services the loco depot was closed as trains were operated from Ashburton, but the turntable was left in place for turning steam engines until the line was cut back to Mount Somers in 1957.


Buccleugh. There was nothing to see on 1956 aerial coverage because the station closed 1952.


Mount Somers. This was the terminus between 1885-1889 and again in 1957-1968. The Mount Somers tramway shown coming into the yard at the top of the map closed in 1943.


Cavendish. Originally planned as the terminus of the line, as reaching Mount Somers required a long bridge across the Ashburton River. 


Anama. A local sawmill was served by the line here.

The entire set of updated maps for the whole line can be viewed here:

The next lines to be looked at are the Southbridge and Little River branches.



Wednesday 30 March 2016

Springburn 1956


Here is the layout of Springburn drawn off the track plan I posted earlier. I am going to update the Mount Somers plan and what I can from other stations so I have an up to date set of maps.

As the diagram shows the only thing left on site today is the platform. The loading bank and the last house remained on site until the last 10 years or so.

Tuesday 29 March 2016

Springburn Branch Maps

Well after my last post back in the beginning of the year, things did slow down a bit. A fair amount of mapping work did get done, but not as much as I expected. But I have worked on a lot of different things since then.

Currently I am focusing on the Springburn Branch because Canterbury Maps turns out to have good coverage from 1956 of a part of it (unfortunately not all of it) but enough to cover Mount Somers to Springburn with the existing layout data we already have from the NZMRG. In other words I can use their diagrams in combination with the aerial photos to draw actual layouts to scale etc.

This past weekend things have slowed down a bit due to changing my main computer over to Linux (Mint 17.3). But right now I am back up to speed with all the software installed that I need for this task because Qgis and some other packages all run on Linux and a lot of the other stuff just happens in web browsers anyway.

Here is a screendump of Qgis 2.14 running on Linux


The only issue with migration has been changes in paths, as the file path structure was not copied exactly the same from the old computer. I was able to open the project in a text editor and bulk replace the changed paths to 189 layers otherwise it would have been a lot of fun to have to browse to all the layers' new locations to find them in the GUI. 

We can accurately size and place buildings from the aerial photo which is a big improvement on the previous schematic from NZMRG that was not to scale. The purply coloured lines and shapes still have to be put into the right places to produce the finished result.

Here is the aerial photo the above were traced from. This is a 1956 image from Canterbury Maps.


There is some additional detail not present in the Canterbury Maps image that was traced from a track diagram from the 1930s. What is interesting about this 1956 image is that four houses could still be seen even though the engine depot apparently was closed in the 1930s as far as I know. It appears only train crew were based at Springburn which must have been a flag station with no actual stationmaster. It is somewhat odd that Mt Somers did not draw the extra facilities but I guess Springburn was the obvious location at the end of the line for engine servicing, but did not have enough traffic to be attended.

Here is a part of the track layout diagram referred to.



Sunday 3 January 2016

Current Projects on NZ Rail Maps

During the current holiday period I am working on a lot of cleanup projects and other activities. Some of these include:
  • Reorganising and restructuring several projects, particularly Northland-Auckland and Canterbury-Westland, that have data in the old single purpose layer structures.
  • Adding information to a number of projects, especially Canterbury-Westland and Palmerston North Gisborne Line. New information from Kiwirail's GIS includes bridge locations and kilometre pegs.
These projects will be progressed from now through to the middle of January as due to it being a holiday period this is the best type of activity for this time. As various people are taking holidays and won't be available to contribute to the project group, the Otago Central work won't resume until towards the end of January.

The Palmerston North Gisborne Line and the lines on the West Coast are all being updated with bridges, distances and track layouts at stations. This time around the PNGL is being done all the way back to Palmerston North, so it won't just be the Gisborne line that gets updated; instead, the Napier line will be fully specced out as well. However there are no plans for new volumes of publications at this stage. On our Map Volumes Collection page that you can see a link to at the bottom of every page of this blog, you can see what has been published and there will be no revisions to those volumes. However, individual maps will be published and uploaded to the Individual Maps Collection that is also linked from the bottom of blog pages.

The focus for publications remains CreateSpace and the Otago Central volume which has top priority and will be assembled as soon as the maps have been completed for that project. My goal is to have that work finished by the end of 2016.

Friday 1 January 2016

Welcome to 2016

This blog has been slightly redesigned and in particular the stuff at the bottom each page has been improved. Amongst other things you will find
  • A set of links to all the sites that this project uses. The explanation of these is as follows
    - Individual map collection - Flickr slides of maps that have been published so far. 
    - Map volumes collection - what has been published so far as PDFs in complete volumes covering all lines in particular regions.
    - Project Files collection - this is where the project files are stored and can be obtained from, with one exception detailed below.
    - KML Files collection - these files are used with Google Earth and are basically historical as they have not been updated for many years and consequently contain a lot of inaccuracies. They were the forerunner to this project. I still use them all the time with the map development so that I can quickly find a particular location on the railways.
    - There will soon be a link to the online version of the map key so that you can find the keys quickly and easily. Older volumes use obsolete styles as there have been 32 major revisions of the keys, but it is largely stable now reflecting the maturity of the project's development.
  • A subscribe form for getting an RSS link to the posts. 
  • A follow by email form for getting updates emailed via Feedburner.
  • Site stats are not up there with Enzed Transport yet as the blog was only started this year, but are growing.
  • Links to my other blogs. This is just for consistency between all my blogs and the other blogs will not be of interest to many readers except for Enzed Transport.
As you will notice the project files collection link in particular is a new one, the bitly URL being http://bit.ly/NZRailMapsProject. This is a new bitly URL pointing to the new Google drive I recently set up and which is synchronised as it happens with updates from project work. The idea has been to have all of the project files in this location, whereas the previous online storage in OneDrive only stored the layers that I had created myself. With the change in the layers structure in general it has also been a good time to reorganise the storage. OneDrive has been dropped in capacity by MS to only 5 GB considering that the OneDrive I used for the project used to have 25 GB in it, which was guaranteed by MS never to be diminished as it was a special promotion, MS sucks. So the OneDrive storage is going to be removed eventually I think. Before MS deletes it. Due to the fact MS has got greedy again and dropped most of the useful stuff as well as capacity from OneDrive which is a huge backflip of late.

There is one exception to the GD storage folder in terms of project file storage, files that won't be available from that Drive for the present. There isn't quite enough capacity in the GD for all the project files completely, so the decision was taken to move the terrain relief files, which are big GeoTiffs, outside the GD storage structure. These files are not essential to a map; they just draw some nice 3D imagery in the background that fills in terrain; and they are the only files in the project that aren't vector graphics (they are raster files). So that is how they come to be excluded from online public storage. The files can of course be downloaded from Linz on the Koordinates website which is where I got them from. You will still see terrain relief on maps I produce, you just won't see the files in the online Google Drive public storage of the project.

Anyway the holiday period of the present is a great time to push things along and I am taking a look on multiple fronts. One of those is to update the maps of the entire PNGL from Palmerston to Gisborne with the bridge and distance data from Alcam. So I am also working with that.