GPS SURVEY - WINCHESTER CENTURIATION

Clark & NEHHAS Routes

Refs to NEHHAS Journal Vol 3, Collected Reports on the Roman Road

Two Centuriation or Roman Estate Roads (RRs) have been surveyed, which may be 50 Roman Acutus apart or 1768.6m (Report 10). They are part of a larger grid. The northerly Clark route was suggest after the War by the archaeologist A J Clark, and the Southern or NEHHAS route by others and surveyed by NEHHAS in 1990s.

Features are selected which can be identified on the ground, including the side of the Road or any ditch, for GPS determination of its OS GR. These include air photo streaks, any part of which goes over identifiable ground features, including plotting points used. These will be plotted out on printed graph paper roll of suitable scale.

The two surveyed Roman Roads will also be plotted beside each other with the bulk of the intervening Northerlies omitted for ease of plotting and measurement of the variation of distance between them. If n is the Northerly omitted distance, then the omitted distance between RRs is d = nCos29, where 29 is the direction of the RR in degrees north of OS Grid east. The actual distance between them D = d + dd, dd being measured from the plots which can be done more accurately than trying to meassure distance apart over some 1769m. Clark appears more consistant in its direction than NEHHAS, and its average angle should be used.

To test the straightness of the two RRs it may be necessary to add Northerlies to plotting them, of .556667m for every meter going West. What is required is to keep the plots on the same roll of graph paper.

Where aggers or terraces occur GR determinations will generally be made at our "No. 2" or "No. 3" position. No.1 = start of Build-up, No. 2 = Level off for Roadway, No. 3 = Base of any Cut-Back, No. 4 = end of any Cut-Back & natural slope of ground resumes.

 

CLARK

Into the Centuriation, Report 12 (and 11)

1. Starting point is taken as the Zig Zag, southern limb (Reports 11 & 12) in Chawton Park Wood, where it seems the trail of features from the Bighton Boundary (Reports 13 & 14) point to. Only the southern edge of this terrace is available to measure (the northern cut back is interfered with by a modern terrace). Determinations made at its two ends.{No permission required}

1a Cut-back visible within air photo streaks Feature 4 (Report 12), and possibly that marked on Survey Map just to NE. Southern edge of RR. {No permission required}

2. Excavation site RD07-9, Dark Air Photo Streak bounded by Light Lines - at time of writing northern one is a band of flints. Measure N & S edges, and the other features depending what is found. [No further permission needed. Roe Down Farm is prepared to house the stationary GPS Unit during the GPS survey - but its not at centre of survey]

3. This dark air photo streak goes over white drive of Roe Downs House. The bulge in the drive is approx. RR centre. [Permission given]

3a Possibly on Feature 6, subject to magnetometary being done there [Permission given]

West of Four Marks, Report 13

(4) Then some distance to next definite feature, partly due to developments through Medstead Common. Prominent light air photo line, sometimes very wide, passes north of Stancomb Farm. But plot of these do not agree with the large agger reported by A J Clark - which itself is not on the line of other features - but can be identified at its two ends on rights of way and placing GPS there. These might indicate is some complex map error results in Clark's agger being on the line or not. [Permission not needed, but may inform Stancomb Farm who gave permission for the survey]

5. The air photo streaks line up with a substantial terrace south of Stancomb Broad Lane. Measure N & S edges at Stancomb Broad Lane and by parish boundary. [Permission probably required]

Centuriation through Ropley, Report 14

6. Air photo streaks continue continuously, becoming light line bounded by dark lines to the Bighton Boundary, where on the Triangular Green (Brighton Dean Lane) the northern edge of the agger can be seen. Measure No. 2s at western and eastern edges of Green. May have to cut down grass in 2 or more places to detect Build-up. [Permission not required].

7. While the Bighton Boundary is a prominent feature, often a double boundary approx. width of RR, which A J Clark thought was an agger, the air photo streaks come into the boundary at its N edge (on the Green in 6) and leave it 4000' to the west at its southern edge. While this is only 10' - 15' difference, it throws out the line by several hundred feet in Chawton Park Wood - missing the Zig Zag. Also the Bighton Boundary is not completely straight - especially at the two ends (where air photos come in and leave). The straightest part (near 2000' west of the Green in 6) is aligned with the two air photo streaks i.e. the double boundaries are on the lines of the two dark air photo streaks (the air photo lines are 50' - 60' apart, the double boundaries are 20' - 30' apart). These co-incidences may suggest these double boundaries may here be approx. following edges of RR - so measure GRs at intersection with Bighton Lane. [No permission needed]

Centuriation Bishops Sutton, Report 15

8. According to the RR Survey map, the Bighton Boundary oscillates W of 7 either side of the line, first to the S, then to N - the point where the first field boundary meets it is near RR centre - or given the likely double boundary here, the Northern Boundary. [Permission may be needed]

9. The Footpath crossing Bighton Boundary W of 8 is near the line of the Northern Dark air photo streak to the west - use the southern boundary (assuming a double boundary here). [No permission needed]

?10. The Bighton Boundary westward veers away to the N and becomes the Zig Zag - but is not on the line and can't be used. The optical Build-up at Feature 8 cuts into the Dark Line, and may be considered. [Permission required]

11. According to air Photo plotting "2 dark lines 42' apart, centre coming into junction NW field boundary and railway". [Permission required]

Alresford (Report 16 - not yet in Vol 3 - NL 107 Summer 2008)

12. On S side of railway becomes a broad light air photo line 60' wide - suggests scattering of metalling - and absent over narrow fields to Spring Gardens (S of Alresford - now built up). Plotting point for light air photo streak W of Alresford map "centre house, east boundary junction with road (N of Spring Gardens, junction New Road). This is RR centre, GPS here. [Permission not needed]

Titchbourne Map (Report 17 - not yet in Vol 3, NL 110 Summer 2009)

13. Light air photo streak continues to Zig Zag with weak optical effects without obvious ground features to relate, over river to Zig Zag - to visible change in slope by FP and field boundary - last GPS - RR centre. [Permission not needed]

14. Line of Clark goes to where the A31 east out of Winchester changes direction, though it is thought the RR does not follow this line, GPS at the mile stone and bench mark - may now be in lay-by [Permission not required}

 

NEHHAS OR SOUTHERN ROUTE

Into the Centuriation (Report 12)

1 Plotting Point centre of air Photo streak DEFGH is field boundary intersection (corner) on edge of Old Chalk Pit (F- E), not shown on modern 1:50K. [Permission may not be obtainable, even though there is a FP to east, its probably not in this field)]

There is a plotting point on the FP from proportions of the boundary cut by the streak, 87' from North field boundary (boundary now not there in field, but can seen in next field). Not a direct observation of Streak and ground feature. [Permission not needed]

2. Plotting Point Air Photo Streak corner of field by D, southern edge of RR. Major Plotting Point for Air Photo Streaks to Feature 1 Report 14. {Permission unlikely to be obtained for this field, but other side of fence is a house and garden, which we can call on.}

3. Line of Air Photo Streak and features in Mike Smith's garden (Feature 8), and garden Feature 9 cuts cross roads Blackberry Lane between Features 8 & 9, NE verge centre RR. [Permission not required]

Centuriation W of Four Marks (Report 13)

4. Air Photo Streak runs into lane S of Manor Farm, North Street. Best estimate is that N verge is RR centre. [Permission not required]

5. Air Photo Streak runs into the A31, and runs for 700' along modern A31. Best point is first field boundary running at right angles to line (hatched) for RR centre. Modern 1:25K map shows modern A31 widened up to this point - take the corner of the narrower A31 boundary. [No permission needed]

Centuriation through Ropley (Report 14)

6. Feature 1 Report 14 is a major plotting point back to No. 2 above, and to Feature 4 Report 15, Scrubbs Lane terrace. Take the kink in FP which is on the corner of the field boundary (now a housing development), where the southern air Photo Dark line runs into, and where the thin Light Air Photo Streak to the east also runs into - assumed RR centre. [No permission required]

Centuriation around Bishops Sutton (Report 15)

7. This Survey Map contains the major Ground feature 4 - a Mixed Terrace - with Air Photo Streaks leading to it on both ends.

At Feature 3, Kink in path on the line, centre Light Air Photo Streak passes through easterly corner - GPS here - RR centre. [No Permission needed]

8. Major Plotting Point, Mixed Terrace, Scrubbs Lane, Feature 4. Try and identify No2s and No 3s on bottom (E) on top (W) - 4 GPS positions. [No permission needed]

Report 15 + Centuriation through Alresford (will be Report 16)(NL No. 107 Summer 08)

9. OS Air Photo 013 71 378 shows 2 stunted hedges with Air Photo Streak going through them. From visits, these hedges may not be there. But if one is taken at its centre as RR centre (they are c 100' wide on master survey map). Otherwise use the recorded Survey Master distances (though translation from Air Photo on OS map will be less accurate with also map errors). There may be geophysics done to locate. [Permission needed from Scrubbs Farm]

Centuriation through Alresford Report 16 (NL 107 Summer 08)

10. Parish Boundary and agger Terrace in Dark Copse. No. 2s and No. 3s at E & W ends, and where there is a change of direction visible on the ground. 6 GPS determinations. [Permission needed from Scrubbs Farm]

11. Light Air Photo Streak carries on W from 10 and runs into field boundary leading to modern road by river. Estimated RR centre is 320' - 470' from W end of boundary. [Permission needed if renewal of previous permission required]

 

SCALE

OS Grid forms on origin in the sea off Lands End, and Easterlies are measured in Km East, and Northerlies in Km North. In the area we are interested, the notation is 460000 for Easterlies, and 140000 for Northerlies. Only the Bold figures appear on the maps as grid lines except at the corners. These mean 460Km East and 140Km North of the origin.

If we plot out the GPS OS Grid readings at a similar scale as the 6":1 Mile published maps, say 1:10K , then
4600 -> 1 cms on plot represents 100m on the ground
46000 -> 1mm on plot represents 10m on the ground
460000 -> .1mm on plot represents 1m on the ground

 

It can be debated if a larger plotting scale is needed. But having regard to the number of readings we have, the length of the plot we will have (1.5m), it is thought not a larger scale. The GPS limit of accuracy is assumed to be a metre plus or minus .5m? which may be comparable to the limits of plotting. Computer plotting might be considered - but in my experience the output scales are poor.

Removing 2Km of Northerlies from Clark

If the round figure of 2Km is subtracted from the Northerlies of Clark, 002000, the two routes will be plotted 1768.606m - 2000Cos29 = 1768.606 - 1749.2 = 19.4m apart. Where the two routes are 50 Actus apart (1768.606m), at 1:10k scale they will only be .194mm apart. This is also about the width expected of the road, and less than the side ditches, and a smaller Northerly reduction will be needed.

About 100m separation is needed to keep the 2 route apart. If 1900m, 001900, is deducted from Clark Northerlies, then the 2 routes would be plotted
1768.606 - 1900Cos29 = 1768.606 - 1661.74 = 106.87m
It may be better to do a first plot with 2000m deduction in Northerlies for Clark, where the Northerly Grid lines can be shown for both routes on the plots, and less error possibilities. Where the separation is insufficient a plot with larger deduction can be made to make separation measurements.

Northerly deductions to test straightness

The addition of .556667m for every m going west (or deduction if start from west and go east) will need to be applied to every reading, and may be done by computer. A BASIC programme could be written, or MS Excel will probably serve. We have a continuous paper printer - though it is liable to leave a gap for each page, but this can be allowed for. The absolute scale is less important here, and the printed scale will enable deviations to be estimated.