Waffles in Ashwell?

Anyone new to this blog or geophysics in archaeology is recommended to read the material on the “Geophysical survey in archaeology” page.

Gil Burleigh asked us if we could survey a field in Ashwell. The field is quite small at just 1ha. Some nice finds had come from this field including a nice scabbard chape (Fig. 1) in the 1970s. The field was ploughed for a short period in the 1980s, and North Herts Museums undertook a fieldwalking survey in 1986 which retrieved pottery from many periods with concentrations of Roman, Medieval and post-Medieval material and a thinner scatter of Iron Age and Anglo-Saxon pottery.

Figure 1: scabbard chape © North Herts Museums.

Members of North Herts Archaeological Society excavated three small test trenches over the weekend of the 16/17th March and found archaeological features in two of them. Soil test trenches has also revealed archaeological features in test pits SA01 and SA02 (although they were misinterpreted as just deeper topsoil! Figure 2).

Figure 2: Location of the NHAS test pits (in red) and the soil test pits (SA01 to SA05). © CAGG.

As the site is pasture over chalk, it seemed likely we would get a good result.  Magnetometry was the obvious technique to try (Fig. 3).

Figure 3: Nigel “compensates” the magnetometer. Photo © Mike Smith.

We managed to complete the whole field in a day with the mag, which was excellent progress.  I was exciting to see what we had detected and processed the data that evening (Fig. 4).

Figure 4: the mag survey © CAGG.

My initial reaction was “ugh”.  Why was the data so horribly noisy, like some sort of giant waffle maker? Although we can see some features, especially on the western edge, we cannot see all that much going on, and nothing around where the trenches had revealed archaeological features. Can we explain this?

Peter Alley flew the site with his UAV.  The photos from his flight can be used in two ways: they can be stitched together to produce an orthomosaic (basically an aerial photograph corrected for distortions) or to map the topography.  The orthomosaic is fun, but doesn’t help us solve our problem (Fig. 5).

Figure 5: Orthomosaic from the aerial survey © Peter Alley.

Mapping the topography shows the basic form of the site, basically a flat field on a slope (Fig. 6)!

Figure 6: the topographic survey derived from the UAV images. © Peter Alley.

If we use a hill-shade, however, we can see the micro-topography of the field.  Lighting the field from the SW creates Figure 7.

Figure 7: topography lit from the south west. © Peter Alley.

We can see from Figure 7 that even though the field has not been ploughed since the late 1980s, the plough scars remain running up and down the slope.  If we move the light to shine from the NW we get Figure 8.

Figure 8: topography lit from the NW. © Peter Alley.

Now we can see plough scars running across the slope!  This is, in part, the origin of the noisy data: the cross cutting ploughing has left a surprisingly uneven surface, like the proverbial giant waffle iron.  We should assume this is mirrored by scars in the surface of the underlying chalk and archaeology. One further possible cause of the problem are thirty years worth of ant hills (Fig. 9).

Figure 9: ant hills at Ashwell. © Kris Lockyear.

These aspects of the site, in part, explain the noisy looking data but do not completely explain the lack of contrast between the fills of the features and the chalk.  Gil reported that the metal dectectorists were also having problems with large numbers of responses in this field, so the mystery isn’t completely solved.  A plane did crash in this field during World War Two, but I cannot see how that has created this problem unless the whole field went up in flames.  Ideally, we would take some samples and test the magnetic susceptibility of them.

There are, however, some features in the data, despite the noise.  I have outlined the ones I can see in Figure 10.

Figure 10: magnetometry survey results with interpretation shown in yellow (cf. Fig. 4). © CAGG.

Two of the possible features run down and across the slope in a manner similar to the plough scars, and we must be cautious in their interpretation as a result.  The corner that is protruding from the western field boundary is much more interesting.  It looks like we may have clipped one side of an enclosure. (We seem to specialise in enclosures that don’t enclose things.)  It looks like it probably runs under Ashwell House next door.  In the middle is a “dark blob”.  Initially, I thought this was likely to be metal as it is so close to the edge of the field where the fencing creates a strong response.  Checking the values in TerraSurveyor, the range of values towards the east of the blob is from -2nT to +12nT.  If one goes right to the edge where one can see the impact of the fence on either side, the values jump to about -4nT to +26nT. Contrary to my initial thought, this might well be an archaeological feature in the enclosure.

On the Saturday we completed a 40x40m block of Earth Resistance survey in the NE corner of the field.  Not a great deal showed!  We decided to persist on the Sunday so that we could include the area of the “enclosure”.  We had some unusual help (Fig. 11).

Figure 11: Unusual help with the Earth Resistance survey. © Kris Lockyear.

As has become our standard method, we used the Earth Resistance meter on the 1+2 setting.  In other words, every time we stick the mobile probes in the ground the machine takes three readings.  One uses the two outer probes which are a meter apart to measure down to about a meter or so.  The other two readings use the inner probe to take two side-by-side readings with a 0.5m probe separation, looking about 50–70cm into the ground (Fig. 12). Given the topsoil is only about 30cm deep, this should be fine.

Figure 12: the Earth Resistance survey underway. Photo © Gil Burleigh.

The results, were, underwhelming… (Figs 13 and 14).

Figure 13: Earth Resistance survey with 0.5m mobile probe spacing.

Figure 14: the Earth Resistance survey, 1.0m mobile probe spacing.

I can only conclude that the difference in moisture retention between the features and the chalk was minimal.  Peter Alley made the excellent suggestion that the banding we can see is possibly related to the layers within the chalk geology.

We have one last source of data to examine.  In 1986 Keith Fitzpatrick-Matthews undertook a field walking survey here for North Herts Museum Service.  Obviously, he didn’t have the luxury of a GPS to log his finds or lay in his grid.  We can, however, roughly position the results on the map with a bit of careful editing in a drawing package.  Figures 15–17 show the three most common types of pottery: Romano-British, Medieval and post-Medieval.

Figure 15: Romano-British pottery distribution © North Herts Museum Service.

Figure 16: Medieval pottery distribution © North Herts Museum Service.

Figure 17: Post-Medieval pottery distribution © North Herts Museum Service.

The post-Med distribution doesn’t seem to fit our survey results.  Both the Roman and Medieval ones, however, are temptingly close to the feature on the western side.  By making the distributions transparent we can see how these match (Figs. 18 and 19).

Figure 18: Romano-British pottery overlain on the mag data © North Herts Museum Service and CAGG.

Figure 19: Medieval pottery overlain on the mag data © North Herts Museum Service and CAGG.

Both the Romano-British and Medieval pottery distributions lie just to the north of the “enclosure”, down slope from it. This is what one would expect if the material collected on the surfaces derives from the fills of the features. We can only be sure of the date of this feature by further excavation.

Although this survey has not produced the beautifully clear results we have had from other sites such as Kelshall, it has shown some interesting features.  It is also a good example of the value of combining different data sets, in this case field walking, aerial photogrammetry and magnetometry.

Many thanks to everyone who helped with the survey and fetched and carried equipment.  Also, thanks for Gil for suggesting we have a look at this site and David Short for allowing us access to the site.  As 19 images seems an odd number to finish on, I thought I would sign off with another photograph of one of David’s magnificent sheep (Figure 20)!

Figure 20: A sheep. Photo © Kris Lockyear.

Advertisements

A chilly day at Little Hadham

Anyone new to this blog or geophysics in archaeology is recommended to read the material on the “Geophysical survey in archaeology” page.

Firstly, apologies to those waiting for the results of the GPR survey in the churchyard at Ashwell undertaken just before Christmas.  I need to do some more data processing to see if we have some graves or not.

Meanwhile, the site at Little Hadham has been one to which we have returned on-and-off almost since the beginning of the project.  (Use the drop-down box on the right of this page to see previous posts on this site.)  The site clearly extended beyond the area we had surveyed so far, and so CAGG once more teamed-up with members of the Braughing Archaeological Group (itself part of the East Herts Archaeological Society).  We could only access one of the fields, the one we first surveyed in April 2014.  Our aim was to extend the survey to the edges of the field, and to examine one group of features we had detected with the mag previously with an earth resistance survey.

The morning was cold and crisp with a hard frost.  The thermometer in Jim’s car registered minus 3 Celsius.  Brrrrrr…  Unfortunately, the frost melted quite quickly and muddy wheels on the mag became a problem (Fig. 1)!

Figure 1: Nigel wheels the mag across the field.

The mag team completed six grids, five of which were partials.  Given the amount of time spent trowelling the mud off the wheels, this was a good haul.  We have four partial grids left to complete the field.  The results are given in Figure 2.

Figure 2: the mag survey.

The areas completed yesterday were the block to the far east of the survey area, and the incomplete strip of partials on the southern edge.  In the new area we can see the ditch previously detected (shown by the red arrow) carrying on across the site.  It is fainter in Millfield to the west of the road, but is still evident.  Near the hedge is another clear line of a ditch (marked with the yellow arrow).  This one is worryingly straight and almost parallel to the  field boundary.  It maybe more modern than some of the other features.  The curvy, more irregular ditches (shown with the blue arrow) may be something like a farmstead with boundary ditches.  Picking apart the phasing of all these features is going to be difficult and would require some targeted excavation.

As we had a good sized team we also undertook some Earth Resistance survey (Figure 3).  We targeted one of the possible farmstead enclosures.

Figure 3: Katie Burgess and Peter Baigent (BAG) using the RM85 Earth Resistance meter.

The team completed five 20x20m grids at a 0.5m reading spacing.  The results are shown in Fig. 4, and the underlying mag in Fig. 5.

Figure 4: the Earth Resistance survey results.

Figure 5: the mag results in the area of the Earth Resistance survey.

I had been hoping — rather optimistically — that the resistance survey might pick-up some structures.  Sadly, it did not.  There are, however, correspondences between the mag and res results.  The ditch with the right-angle corner in the mag survey shows well, if slightly more diffuse, in the resistance data.  Similarly, the long curving ditch also shows well. I have indicated one end of it with a blue arrow in Figure 6.

Figure 6: Resistance survey results with arrows.

More curious, however, is the change from low to high resistance along a straight line indicated with a red arrow in Figure 6.  This corresponds exactly with the diagonal line in the mag data which cuts east-nor-east west-sou-west across the D-shaped enclosure. I’m at a loss to know what this represents.  It maybe a reflection of the various cut features in the underlying geology.

At the end of the day we were treated to a beautiful moonrise and sunset.  Not quite the blood wolf moon seen some 12 hours later (when I was tucked-up and asleep in bed!).

Figure 7: Moonrise. Shame about the electric cables!

Many thanks to everyone who turned-out on a freezing but beautiful day.  This site continues to repay our attention, and it worth the effort.  We should try and survey some of it in the summer, however!

Neat and tidy

Anyone new to this blog or geophysics in archaeology is recommended to read the material on the “Geophysical survey in archaeology” page.

Due to being rained off on our last day, a small team of us decided to go out and finish off some things on Bank Holiday Monday. Many thanks to Pauline, Judith, Ruth, Dave and Jim for turning out to do “just one more grid.”  I think it must be a geophysicists ailment that we always would like to be able to just a little bit more…

The mag team completed an impressive ten grids including two awkward partials.  Figure 1 shows the entire survey at the end of the 2018 season.

Figure 1: the mag survey after day 19.

The team have managed to add 19 ha to the survey in the last month.  Figure 2 shows the southern area that we have been surveying this week.  (This field is, confusingly, called “Prae Wood”.)

Figure 2: the southern area (Prae Wood) after day 19.

The team have picked-up an area of intense ferrous noise.  This looks like a small historic period site.  We will have to check out some old maps to see if we can work out what that might be.  The one hiccup in a brilliant last day of work is a single line of data where the sensor froze.  It is very annoying and I’ll have to find some way of fudging that until next summer!

The Earth Resistance survey had one last little block left to make the plot look all neat and tidy.  Many thanks to Pauline and Judith for helping me fill that in (Figures 3 and 4)!

Figure 3: Kris, Judith and Pauline (out of shot) extended the resistance survey. Image © Mike Smith.

Figure 4: the main block of Earth Resistance data collected 2016–2018.

The data collected shows some faint indications of buildings in that corner (Figure 5).

Figure 5: the northern area of the res survey. The NW corner was completed on day 19.

Although my trick of spreading the remote probes wide apart has worked on the whole, this year there is a bit of an edge.  This is because we started with a block in the SW corner, worked eastwards, and then when we had got to the corner, worked back along the hedge line westwards.  Between when we started this block and yesterday we have had in excess of 100mm of rain (or about 4 inches in old money) so it isn’t surprising this shows in the results.

We have now cleared away all the pegs and flags, packed-up the machines and left Gorhambury for another year.  It is a beautiful place to work and we are very grateful to Lord and Lady Verulam and their family for allowing us to extend the survey, to those who work the estate and put up with us getting in the way, and to the estate managers, especially Stuart Gray. Thanks to the Institute of Archaeology, UCL, for lending us the dGPS and the res meter, and SEAHA for the loan of the GPR.  I hope everyone involved thinks the results are worth all the effort. Most of all I would also like to thank all the volunteers who came this year, whether you only managed a day or two, or you came for the whole season.  You are what makes this project so much fun!

 

“The way I see it, if you want the rainbow, you gotta put up with the rain.”

Anyone new to this blog or geophysics in archaeology is recommended to read the material on the “Geophysical survey in archaeology” page.

As I start this entry of the blog, the rain is splashing against my windows as was predicted by the Met Office. Although we might question Dolly Parton’s grammar, the sentiment seems true enough.  Yesterday, however, was a superb day with all three techniques collecting data across the site.

After yesterday’s excellent results, the GPR crew had great expectations.  The only problem was a tree in the way under which the shepherdess had put hay when the grass in the field was dead from lack of rain.  Unfortunately, sheep mean sheep droppings (Figures 1 and 2).

Figure 1: Mike on sheep poo removal duty.

Figure 2: Check out those wheels!

Luckily for everyone concerned, I think the effort was worth it (see Figure 3)!

Figure 3: GPR time slices from Day 18.

I could misquote Dolly along the lines of putting-up with sheep poo if you want excellent GPR results but I might be pushing my luck…  The many buildings are quite obvious in this data set.

Figure 4 shows this grid in context of the other GPR grids in this area.

Figure 4: GPR results including the day 18 data (SW corner).

We have added a very large number of new buildings to the map of Verulamium.  As I was only just starting with GPR data when we started collecting it in 2015, the processing keeps changing a bit from block to block. One of my jobs is to start from scratch and reprocess the whole thing so that the maps are consistent.  Should keep me busy for a while.  Figure 5 is a crude mosaic of images just showing the entire area surveyed so far.

Figure 5: crude mosaic of GPR time slices at the end of the 2018 season.

This represents 19ha of GPR data collected at 0.5m transect intervals.  Just pushing the machine along the lines, not including getting to the block, setting-up, moving strings etc. is 380km.  It also means 380km of radargrams!  No wonder the data takes-up 33gb of my hard disk and consists of over 70,000 files.

The mag team completed nine 40x40m grid squares which is 1.44 hectares.  Excellent progress!

Figure 6: the mag team in the southern field.

Figure 7 shows the whole of the 2018 survey (along with a big chunk of Verulamium).

Figure 7: the mag survey after day 18.

Even though we have been using the machine for some years now, and it does have its frustrations, when all is going well we can really cover some ground.  The season was planned for 20 days: we lost 3 days to rain, and most of a day to testing the mag at the start.  Despite this, the team have managed to collect 17.7 hectares of mag data.  Without actually getting to the grids and back (which is quite a bit of walking in itself), the team have pushed the cart 88.5km over the past four weeks.

Figure 8 shows the southern area in more detail.

Figure 8: the southern area of mag data after day 18.

The blue arrows in Figure 8 indicate the lines of old field boundaries.  These can be seen on old maps such as the 1699 parish map.  The yellow arrows mark ferrous objects.  Some are very big, but there are a scatter of smaller ones too.  Last, but definitely not least, there are a few magnetic features which may be archaeological, such as pits.  I have picked a few out with red arrows.  Although they look small at this scale, they are probably 1m to 2m across, a quite respectable size for a pit.

Although large mainly  blank areas are disappointing to collect, they are important nonetheless. The immediate environs of Verulamium are extremely rich, archaeologically. The field lies:

• 360m W of the busy area of buildings recorded by the GPR discussed above;
• 600m NE of the major Iron Age settlement at Prae Wood;
• 600m N of the fields at Windridge Farm where metal detecting rallies have taken place;
• 500m NW of the major cemetery at King Harry Lane;
• 1,100m SE of Gorhambury Roman villa;
• 1,000m NE of the new villa found at Windridge Farm.

Also, the Fosse, which is preserved in the woodland along the NE edge of the field, is a really very impressive earthwork.   We just seem to have hit an empty bit of landscape between all these sites!

The res survey now covers some 6.58ha, that is about 263,200 earth resistance readings.  Not into the millions like the mag and GPR, but this is res after all!  Figure 9 shows the entire survey.

Figure 10: the entire Earth Resistance survey after day 18.

At this scale the roads show very nicely as do some of the more substantial buildings.  Figure 11 is the area surveyed in 2018.

Figure 11: Res survey after day 18.

Given that the fields were baked hard and the grass was dead at the start of the season, I am pleased we managed any Earth Resistance survey at all this season.  The team yesterday put-up with my geophysics OCD and completed right into the corner by the theatre. We then doubled-back and started filling-in between the top of the survey block and the drive.  We have picked-up some parts of buildings seen in grids to the south, but in general along the edge the deep colluvium, as shown by the sunken nature of the drive, is to some extent masking the archaeology.

Many thanks to everyone on the team who made the 2018 season such a success.  A especially big thanks to those who helped move the equipment about including Ellen, Mike, Jim and Ruth.

For those who haven’t been involved but would like to join future surveys, do get in touch.  We are a friendly group, and provide on-the-job training.

And finally… (as they used to say on the news)

 

A busy day

Anyone new to this blog or geophysics in archaeology is recommended to read the material on the “Geophysical survey in archaeology” page.

We had a large team today and as a result we managed ten mag grids, two and a bit GPR grids and seven earth resistance grids. Good job everybody!

First to the mag.  The team extended their survey in the field to the south of Mobbs Hole.  Figure 1 shows the overall survey and Figure 2 zooms in on this field.

Figure 1: the 2018 magnetometry survey.

Figure 2: the survey in the field to the south after day 2.

I have annotated Figure 2.  The red arrows indicate the line of the ditch of the Fosse.  It is salutary to note that a feature as big as the Fosse barely shows in the mag data.  Clearly the upper fills of the ditch are largely the same soil as the surrounding topsoil.  We can normally see pits and ditches on archaeological sites because they are filled with more organic, and thus more magnetic, soils, the result of nearby human occupation.  The green arrow shows a “blob” of higher magnetic readings. The rather diffuse edges to this feature make me suspicious that this might be a “tree throw”, i.e., where a tree has blown down.  The yellow arrow marks two strongly magnetic parallel lines.  At first I thought these might be something metallic but checking their actual values shows they vary from -10 to +29 nT.  Certainly strong, but unlikely to be metal.  The blue indicates something which is definitely metal; it has values of -1543 to +680nT!  The dark pink arrows indicate a faint line, possibly an old fence line.

The res team consisting of Deborah, Tim, Julia and Anne completed seven squares.  Figure 3 shows the whole survey from 2016–2018.

Figure 3: the earth resistance survey 2016–2018.

We have now covered 6.3ha.  For a resistance survey at 0.5m spacing between readings, that is pretty impressive.  Res has always been a poor third to mag and GPR in this survey.  We didn’t get started until a year after the other techniques when UCL purchased a new RM85, and we have had problems with weather.  Hopefully we can fill in the top corner on Saturday.

Figure 4 shows a detail of the area completed this year.

Figure 4: the northern area completed so far this season.

The street shows very clearly in Figure 4 running SW-NE, and slightly more faintly we can see the buildings either side. One problem to tackle in processing data is that very high areas, like the road, can make the more subtle stuff harder to see.  If we “clip” the image to bring-up the details of the buildings, the road area becomes one big black blob!  One way to get around this is to use a high-pass filter.  Figure 5 shows the same area with the high-pass filter applied.

Figure 5: the 2018 survey area after the application of a high-pass filter.

As you can see, the buildings show much more clearly but the road much less so.  Especially with resistance data, it is worth looking at several versions of the data processing to get the most detail from the survey.

The GPR crew finished off the grid from yesterday and did another 40x80m block.  Figures 6 and 7 are the time slices from the two days.

Figure 6: time slices from day 16 of the GPR survey.

Figure 7: time slices from day 17 of the GPR survey.

As you can easy see, we have some sweet buildings showing.  Figure 8 is a rough composite of the sites in this area.

Figure 8: composite of slices in the area of the day 16 and 17 survey blocks.

I need to do some cleaning-up of the various blocks as they were processed at different times and with different software packages, but in general you can see the mass of buildings crowding along this section of road.  Very nice!

Signing off now so we can go and start day 18.  This may be our last day as the weather forecast for Sunday is dire…

 

Seven red kites, two fire engines and a microlight

Anyone new to this blog or geophysics in archaeology is recommended to read the material on the “Geophysical survey in archaeology” page.

It was an eventful day. At lunch seven red kites descended on some tasty tit-bit not far from where we were sitting, and in the afternoon two fire engines drove up the drive and we were “buzzed” by someone in a microlight. None of this has anything to do with the geophysics, however!

The mag team completed the last two grids in Mobbs Hole (for now), and have started on the field to the south.  The first six grids were all wheel-spinning partials too.  They have, however, only one partial left and then there are eleven whole grids laid-out and waiting.  Partials are not the Foerster’s strong point.  The lack of an “end line” function means hours are wasted spinning the wheel to fool the odometer into thinking we have completed the line.  Open fields, however, are its strength and the team will be glad to be out in the wilds again.  Figures 1 and 2 show the results from Mobbs Hole.

Figure 1: the Mobbs Hole survey in its entirety.

Figure 2: the southern area completed today and the start of the next field.

The GPR team had a partial around the water trough this afternoon and so they didn’t quite complete their usual 80x40m block (I knew I should have kept quiet yesterday).  The next two figures are nine time slices of the western and eastern halves of the block.

Figure 3: time slices from the day 16 GPR data, western block.

Figure 4: Day 16 GPR data, eastern block.

The western block seems to be yet-more blobby stuff, although with some very strong reflections.  The western block, however, has some clearly recognisable Roman-style corridor houses.  Yay! Finally some buildings we can recognise!

The last two images show slice 6 in context, firstly on the mag data, and then the mag data with an outline of the location of the GPR blocks.

Figure 5: GPR data from day 16, slice 6.

Figure 6: mag data with the location of the Day 16 GPR data indicated by the red box.

The huge black and white feature in the middle of the mag plot (Figure 6) is the water trough. As you can see, some of the walls of the buildings show in the mag data, but are much clearer in the GPR data.  Some only show in the GPR.  I know I am beginning to sound like a stuck record, but that is the strength of multi-method survey.

Tomorrow is our antepenultimate day (I had to get that in once again), so fingers crossed for dry weather.

Many thanks to the whole team for their wonderful effort and commitment.

The end is nigh?

Anyone new to this blog or geophysics in archaeology is recommended to read the material on the “Geophysical survey in archaeology” page.

In this case, two ends: we have just started the final week of the 2018 survey season and the mag team are within two partials of completing as much as we can of Mobbs Hole and moving into the field to the south.  First to the mag.

After the annoying plethora of frozen sensors, the mag team spent a good proportion of their day re-doing duff squares.  It was worth it, however, as today’s data looks fine (Figure 1).

Figure 1: the mag survey in Mobbs Hole at the end of Day 15.

Although we can be pleased with the area we have covered, surprisingly little apart from the Fosse itself and related features show.  We must keep in mind, as Isobel Thompson reminded me this morning, that “even such negative evidence is information”.  Negative information may be important, but at the end of a long day’s survey some tasty looking buildings would be nice.  Figure 2 shows one possibility, although we may be grasping at straws!

Figure 2: a possible building in Mobbs Hole?

The Earth Resistance survey takes fourth place in priority after surveying in pegs, mag and GPR.  Anne and I did, however, manage to extend the main block of res data by another three grids.  Figure 3 shows the results.

Figure 3: the Earth Resistance survey after day 15.

As you can see, we have picked-up some more of the building to the east, but also part of Street 25 running SW–NE.  There is quite a break in the line of the street which is curious.  Figure 4 shows the GPR data in this area.

Figure 4: the GPR data in the area of the res survey. The red box marks the outline of the 2018 survey after day 15.

It is useful to note that some parts of the buildings show more clearly in the res data, and some in the GPR thus making the extra effort of doing res as well worth while.  The GPR data also shows a break in the road.  Figure 5 shows the mag data.

Figure 5: the mag data. The red box shows the 2018 res survey area after day 15, and the blue line the course of the aqueduct.

Note how the buildings that show clearly in the res/GPR barely show in the mag data, but how the “burnt building” (assuming my interpretation is correct) only shows in the mag data.  Multiple techniques rule, OK?  I have roughly marked the line of the aqueduct in Figure 5.  Let’s now look at how that maps back onto the res data (Figure 6).

Figure 6: the Earth Resistance data with the line of the aqueduct indicated.

Not only does the aqueduct kink around the two buildings as we noted in an earlier post, but it goes through the break in the road.  I guess there could be a wooden bridge (which we would not detect) or maybe a culvert where the roof has collapsed or has been robbed. Fascinating stuff.

The GPR crew in their machine-like fashion completed yet another 80x40m block.  Figure 7 shows six time slices.

Figure 7: GPR survey, day 15, six time slices.

Most of the action, so to speak, is in the NE corner.  There is a particularly clear corner in the fourth time slice indicated with a red arrow (Figure 7, top-right slice).  This might be a surviving floor. There also appears to be a long linear negative feature, as shown in the fifth time slice by three red arrows.  Figures 8 and 9 show slices 4 and 5 in context with the day 14 data.

Figure 8: GPR data from days 14 and 15, slice 4.

Figure 9: GPR data from days 14 and 15, slice 5.

Three things caught my eye.  The squarish “floor” which crossed over the boundary between the two days data, the sub-circular white “blob” which also lies across the boundary, and the long linear low-reflection feature (shown in white) which runs diagonally SW–NE across the lower half. I traced the square and the blob and had a look at the mag data (Figure 10, click on it to see full-sized).

Figure 10: the mag data with the “square” and the “blob” outlined.

The white blob corresponds with a faint “blob” of higher readings in the mag data.  On its own, I would have been tempted to ignore this, but it does look like a feature about 6m across.  The square is harder to assess.  There are magnetic features parallel to it and close by.  We are probably looking at parts of a building.  I had a quick look at the radargrams and the square high-reflectance feature in the GPR data looks like a solid layer, probably a floor.  I also noticed the long linear ditch-like feature running across the mag data, so I traced that and went back to the GPR data (Figure 11).

Figure 11: GPR data with the linear feature seen in the mag data highlighted.

The linear feature in the mag data fits the linear feature in the GPR data perfectly.  Lovely result.

It was a busy day surveying today, and so I didn’t have time to goof off and take photos of people or the views.  Maybe tomorrow!

Thanks to everyone who helped today.