Thursday, 23 January 2020

Palmerston North Gisborne Line [4A]: Makaraka, Mataphero & Muriwai [1]: Makaraka Branch 1 - Intro, Park Racecourse Station & Corsons Seeds Siding

It's time we started posting substantive content about the PNGL in separate titled articles, rather than in progress updates. And so this first article in a series about three peripheral stations close to Gisborne, has been grabbed out of a progress update and made into an article of its own. The remaining parts are close behind, as we have to get these maps completed in a few more days. This part has actually been updated with more information about the Makaraka Branch including additional aerial photo views, so the information in it is different from what was originally posted as part of a progress update.

The Makaraka Branch is the first section of the old Moutohora Branch, from Gisborne to Makaraka, which was Gisborne's first railway. Construction works began in 1900 at Gisborne and the first section which went out to Ormond was opened in 1902. Makaraka was the junction of the Ngatapa Branch until it closed in 1931, the actual points being a little to the west of the station and separated from it by a bridge. The Makaraka Branch is only a little over 3 km in length, whereas a number of publications have erroneously used 5 km. The reason for this is that the distance from Gisborne to Makaraka when the Moutohora Branch first opened was 3.29 decimal miles, which when converted  becomes around 5.29 km. But there was never any 5 km peg at or near Makaraka. The way this measurement is arrived at is by converting the mileage distance of Makaraka into a metric measurement, which we do not endorse in any way, because such a desktop conversion will inevitably be compared with actual kilometre pegs, which in many cases (this being a prime example) did not line up with a conversion from an imperial distance peg.

The reason this measurement is wrong is that the zero peg for the Moutohora Branch was originally at Gisborne Station. But after the PNGL was built through from the south and opened in 1943, the section from Gisborne Station to Gisborne Junction which used to be part of the Moutohora Branch, eventually became part of the main line, and the start of the Makaraka Branch was put at Gisborne Junction, 1.15 miles from Gisborne. So at that point, Makaraka was only 2.14 miles from the junction. However, the nearest mile peg to Makaraka Station was still the 3 mile peg, because the mile pegs generally are not moved when the actual distance changes. In other words, the beginning of the Makaraka Branch ("zero peg") would be actually at the 1.15 mile peg. Another example of this offset point of origin is found in Wellington, where the Wairarapa Line starts at its own 1.8 km measurement, which is also 1.8 km on the NIMT that it branches off, and other examples exist on the NZ national network, although some people claim that the lines in question are dually measured.

So the correct distances for the Makaraka Branch according to distance pegs would not have been displayed either in the position of those pegs or in any official records until metrication of NZ Railways in 1974. At that point, the start of the Branch would have truly been at a zero peg at Gisborne Junction, 0 km, with the station of Makaraka being around 3.2 km. We haven't been able to sight any actual official record of where Makaraka Station is in metric measurements, so we have to use the imperial measurements. And that means the maps actually show the location of Makaraka Station as 3.29 miles, because we use the official distances from NZR Working Timetables or other official NZR records. If the Quail Atlas, which prints the erroneous 5 km measurement that has been reproduced in a number of other publications (among them IPL Press's book"Steaming To The Sunrise", David Leitch's "Ghost Railways" book, and Wikipedia) had in fact used the metric working timetable to get their distance it would have been correctly shown as 3.x km instead. Also, the Quail Atlas shows that the main Gisborne station was the junction of the two lines, which is not actually correct.

Having gone through all this now, we have to change the distance shown for Gisborne Junction which is currently "388.52 km PNGL 0 km MB" to "388.52 km PNGL 0 km MB 1.15 M MB" so that we have made it reasonably clear that the start of the Moutohora Branch, nominally from 1943 to 1974, was at its own 1.15 mile peg. The official NZR records we have been able to access showed clearly enough that the section from Gisborne to Gisborne Junction was part of the PNGL main line during at least some of that time, but that the milepegs on the Moutohora Branch / Makaraka Branch, before 1974, remained in their original locations and thus needed to have this reduction factored in when measuring the actual length of the branch or distance of stations from origins. This being one of many examples often referred to in this blog where a distance peg is not an accurate measurement of actual distance. From our perspective, distance pegs are only accurate when they are first put in. After that point, they are not so much distance markers as reference points off which other reference measurements or coordinates can be derived. Of course, with hand held GPS devices available nowadays, these distance marker derived coordinates are not as important as they once were. To make it clear, we have changed the map labels to display the reference to the Moutohora Branch distance measurements at Gisborne Junction, Park Racecourse and Makaraka. Online copies of these maps will be updated in a day or two, and notified in the next Volume 5 Progress Update.

 Diagram map for the junction and the first part of the line up to near Park Racecourse.

So now let us take a look at the Branch itself, its stations and sidings. The first question to be addressed is the official treatment of the Makaraka Branch in NZR documents. We have not undertaken any research from Archives New Zealand, so we cannot access the type of records that are held in their collection. What we generally know about the branch is from documents we have obtained from other sources, mainly working timetables, track diagrams, S&I diagrams and Kiwirail's public GIS (the ALCAM level crossing viewer). The S&I diagram references are interesting. These diagrams are usually a useful official document for siding and yard layouts. The S&I diagram for Gisborne for 1965, the first one we can access at present, simply shows the Makaraka Branch as "Park Racecourse Siding",  a switchlocked siding off the main line, rather than an actual branch line. This depiction of the Makaraka Branch was repeated in the 1988 S&I diagram (the labelling used then was "Racecourse Park Siding") despite the "siding" going through to Makaraka at that time (for the  APMB store). The 1989 S&I diagram was practically identical, but the one issued in 1991 just shows the "siding" with no labelling, and it seems fair to assume the branch was probably almost closed at that time. It was not until 2003 that these diagrams actually labelled the "Old Makaraka Branch" as such, together with a note that the switchlock was fixed in the normal position, and as far as we know this is still the case with the most current diagram.
Extract of S&I 1102, Gisborne, 1965.

Extract of S&I 2861, 2006. The way the Makaraka Branch is drawn on the diagram shows it is not formally part of the Kiwirail network at this time.

The Makaraka Branch had just two stations, and these were where industrial sidings were subsequently provided. Park Racecourse was the first station on the Makaraka Branch, at 2.28 miles from Gisborne, or 1.13 miles from Gisborne Junction. The NZRLS 1947 North Island Working Timetable reprint shows that at that time, Park Racecourse had very few facilities, but it did possess one siding loop capable of holding 36 four-wheel wagons, which is clearly visibly in use on the aerial photography we obtained from Retrolens for the late 1940s and early 1950s when the station was still open, being in use in each case. There were also stockyards. and a passenger platform on the right side of the line, which must have been between the two tracks. The aerial photos are not clear enough to ascertain with complete certainty the exact locations of all facilities at Park Racecourse to enable them to be marked accurately on a map; only the stock yards are so indicated, whilst the siding is traced off the 1986 aerial photo as it was used by Corsons.

Park Racecourse was closed in 1959 at the same time as all other Moutohora Branch stations when the branch closed, and whilst it and Makaraka were transferred subsequently into the Makaraka Branch, there is no suggestion that it was ever formally reopened.

Corsons Seeds appears to have been first established at its present site in the 1960s, sometime after the closure of Park Racecourse, and is directly opposite the station site. The siding consists of a loop running about 300 metres from the Chalmers Road level crossing westward. It is highly likely that this siding reuses the formation of the Park Racecourse station loop siding mentioned above, including possibly the full length of same. It is even possible that this loop is the original one from the station, that was simply abandoned after closure, and reinstated into use when Corsons opened, but this cannot be confirmed in any way at this time. On Corsons premises there was formerly a loading shelter over the track; this shelter and the attached building having been demolished since the NZR corridor survey of 1986. We believe the track is still in place although heavily overgrown like the Branch itself. Aerial photos of wagons in the Corsons siding are rare, suggesting it received only light traffic, and may have been out of use altogether from the early 1980s.

Current diagram for Park Racecourse, showing Corsons Siding, the location of the stockyards, and the Corsons Seed premises. The siding appears to follow more or less the location of the Park Racecourse station siding.

1951 aerial map of Park Racecourse, then a station on the Moutohora Branch. Three wagons can be seen in the siding, along with the small stockyard. The station platform is not immediately obvious.

Park Racecourse aerial map for 1986, showing Corsons Seeds premises including loading shelter.

Aerial map of Park Racecourse for 2017, showing building changes at Corsons Seeds premises.

Wednesday, 22 January 2020

Palmerston North Gisborne Line [0O]: Volume 5 Progress Update 15

We are pushing on with all possible speed to complete this volume of maps, but some late changes have meant that the maps for the Makaraka Branch have had to be redrawn for the second time this week. This means the series of maps for this line have been generated three times in total over a period of two weeks.

This is happening as we get into the intensive work of checking and updating everything, and finding in this case and in other cases that more historical aerial photography needs to be put into the GIS, and also in the case of the Gisborne region, discovering that the higher resolution 0.125 metre Linz imagery dated 2012 can be substituted with 0.3 metre imagery dated 2017 if we need more up to date base coverage in some areas where 2012 is too far back and where the scale of a map is not large enough for the resolution difference to be material. An example of this is at Makaraka where the old Apple and Pear coolstore lately occupied by Weatherell Transport burned down in 2016 and was subsequently demolished.

The first revision of the Makaraka Branch coverage done two days ago was completed after creating the first lot of historical tiles for various stations on the branch and resulted in the total number of maps and diagrams for the Moutohora Branch being increased to 115, which included an increase of the number of diagram maps from 40 to 47 and an unknown increase in the number of aerial maps to 68. This meant that the sequence numbers of all the diagram and aerial maps above a certain level were incremented by 7 using a special script we have available. For the latest revision we maintained the existing sequences by revising the coverage of the diagrams within the first 10 sequences so that we were able to squeeze in an extra higher resolution diagram of Park Racecourse station whilst keeping practically the same amount of detail of Makaraka station despite reducing the scale of several of the diagrams there slightly. There are now 47 diagram maps and 65 aerial maps, which is a small reduction in the number of aerials but with practically no loss of detail, the reason being that we had covered all of Makaraka yard at a ridiculously large scale of 1:500 and reducing this to 1:1000 for most of the yard except for the middle section containing the key infrastructure has not materially impacted on the presentation of the information for this area.

The revised diagrams have not yet been uploaded. This will be performed over the next day or two and announced in the next progress update.

Whilst we will continue to post these progress update reports as the work of completing Volume 5 at Basic level continues, we will no longer be posting significant map content in them. Instead, it is time to revert to posting specifically titled articles about particular sections of the maps such as station yards. To this end we pulled the content for Makaraka Branch out of the last progress update and it is being published in its own right in the first part of a three part article covering Makaraka, Matawhero and Muriwai. This article needed the full revisions of the Makaraka Branch to be completed before it could be produced so we expect this post to appear on the blog tomorrow. This also means the content itself is significantly updated from what was formerly in the previous progress update.

The next stage is to complete the maps of Matawhero Station. We discovered that 1951 aerial coverage of this station was significantly easier to work off than that from 1966 so we added the former to the mosaics, which has necessarily slowed down work a bit, but the maps are now almost complete and reveal this yard was once much bigger than it is today and even had a full bridge at the south end. We also added a 1951 aerial of Park Racecourse to the mosaics for the Makaraka Branch which has also slowed down the revision of the maps for that section. Some additional sidings at Gisborne either side of Stanley Road crossing have been added to the maps for Gisborne Station with the 1986 NZR corridor survey providing the necessary source for tracing. Then Muriwai is the next yard to be completed and that should allow us to finally finish the main PNGL corridor.

After that, to complete Volume 5, we still need to finish some of the yards further down like Dannevirke and Waipukurau, Napier Port, Ngatapa Branch, Gisborne-Opotiki Survey and... maybe that's all of it or maybe there is something else we missed? So there is still quite a bit of detail work. Because it now seems unlikely we can finish all this by the end of this week, we have pushed the deadline out yet again, but we really have to pull all the stops out to have this volume finished by the end of January at the very latest, i.e. by the end of next week. So the pace continues unabated.

Monday, 20 January 2020

Palmerston North Gisborne Line [0N]: Volume 5 Progress Update 14

Well after our announcement earlier in the week of further great progress being expected, things slowed down a lot due to other commitments, the result being the maps are not finished by the end of the week (i.e. today). However we do have all map tiles completed for tracing and are steadily pushing ahead with completion of the volume. We are currently working on the stations around Gisborne and also have Dannevirke and Marakeke to complete. Having these delays occur has cemented our decision not to extend coverage of Napier and Hastings, as it is imperative to complete the production of the maps by the end of next week at the very latest.

One of things we've noticed recently is that a number of older articles we wrote about the PNGL some years back are not actually currently published on the blog. These include a number of historical photos and information about different sections of the rail corridors. To redress this, we plan to recover these articles from the drafts section of the blog and republish them in coming days. This does require completion of certain sections of the maps and this will happen, although probably after all the maps have been completed.