Archaeological Aerial Thermography and Near Infrared Photography
In the late spring of 2021 our Facebook pages were deleted and our activities no longer posted routinely online.
Captured with a Flir One 160x120 pixel thermal imager
John and the late Rosie
Wells on Cairnpapple
Flir PathFindIR 320x240 pixel imager
Archaeological Aerial Thermography
an influential paper was published which set
the scene for overlooking plant transpiration in favour of
the more easily modeled physics appropriate to bare soil.
However, Kay (McManus) Smith carried out a detailed study over vegetation
imaging therefore provides a useful complementary
prospection tool for archaeological and geological
applications for surfaces covered by vegetation' - 2004).
In 2006, a practical, daytime approach was described by Ulrich Kiesow for sites covered with vegetation. As a physicist, thermal capacity, conductivity and inertia etc. were familiar concepts to me but the images produced over vegetation in this latter paper were inspirational. I corresponded with Uli before buying my first thermal imager.
One of Ulrich Kiesow's aerial thermograms from 2011 with visible spectrum for comparison
The role of water in the thermal imaging of both
vegetation and bare soil captured my imagination. My
considerations are independent of classical cropmarks and
parching, as they can be captured and processed digitally at a much
higher resolution. The ones in Uli's paper are there as
proof of concept.
Transpiration, by day and less by night, is a
cooling factor which can influence the thermal imaging of
vegetation and thereby the delineation of archaeological
residues in the ground as a result of variations in
moisture content at root level. The magnitude of
transpiration is further modified by the air and soil
temperature, sunlight, relative humidity and wind speed. However, the
ground has also to be considered, not only in relation to
transpiration but as an entity in its own right.
(thermal infrared at night - original video).
Such thermal amplification* is not only dependent
on the thermal properties of the ground but it is also
dependent on the quantity and temperature of the rainfall
and its distribution over time.
So, with so many balls to juggle, when do I prefer to fly my imager? Over vegetation (which for me is usually grass), during the growing season, I head out on a sunny afternoon, when the relative humidity is not high, and the wind speed is about 8-20mph. Light rain earlier in the day, or heavier rain in the days before, can be helpful but the vegetation should be dry at the time of imaging:
Stratford Court Playing Fields following a light shower of rain (TIR right) earlier in the day, Stroud, Gloucestershire.
(visible and thermal infrared during the afternoon - the goal was on wheels!)
Prolonged heavy rain resets the landscape for differential drying and differential transpiration to begin. Otherwise I fly at dusk, or in the early hours of darkness, with both wet and dry areas on the ground when flying over vegetation-free sites.
Our current thermal and near infrared kite kit
A Flir One (160x120 pixel) thermal imager attached to a phone and a Sony Nex-5N camera (on a Picavet suspension plus a GentLED-AUTO interval trigger) with the hot mirror removed but with no replacement window. The phone is deployed on the selfie stick which is used fully extended and has lockable sections.
The variables involved in thermally delineating archaeological residues can be numerous, complex and interdependent. However, in practice, I have made some simple observations over the years here in the UK.
1. Choose your thermal imager wisely. When reviewing published work, look at the images that have been produced by different products. Consider the number of pixels and angle of view in relation to your intended maximum height of flight, but remember that image quality does not depend on the number of pixels alone. I always shoot video with the temperature 'span' unlocked. Try to keep the temperature range as small as possible by avoiding temperature extremes within the field of view.
2. If flying over grass (or other vegetation) in the daytime, choose a time when the UV/sunlight/temperature levels are peaking and the Relative Humidity is around a minimum for the day. A compromise usually has to be made on timing. If it is a breezy, washing on the washing line day, then it is a good day to fly. The grass should not be wet, but previous rainfall can be helpful if there has been sufficient time for differential drying of the ground.
3. If flying over grass (or other vegetation) after sunset, choose a time after the steepest drop in temperature. Superficial features will show first. Timing for optimal delineation can be critical with the imager pushed to the limits of its sensitivity if the timing is not the most appropriate.
4. If flying over bare earth, or over a vegetation-free industrial site, you can take the approach outlined in (3) or you can fly at anytime after rainfall during the period of drying. If a road is beginning to dry, then that can be a good indication that you should be out at the site. Timing is usually less critical after sunset as the drying process is often slower.
Parchmarks are often seen on parkland, pastures and playing fields etc. They occur when there is insufficient moisture in the soil (often showing first above underlying foundations) to maintain the viability of the overlying grass. As the grass becomes straw coloured it is contrasted by the surrounding healthy green grass. On well draining sandy and stony soils, the inverse of this can be seen where grass growing above a ditch remains viable longer than the surrounding grass because of a better access to moisture. Archaeological residues outlined in such a way can be starkly emphasised in visual spectrum imaging by manipulating the colour channels and input levels with appropriate software. I use the free, Photoshop-like, GNU Image Manipulation Program (GIMP) for all my image processing.
Grass under such stress looses its ability to reflect light in the near infrared. This can occur before the parchmarks are clearly visible to the naked eye. Consequently, visualising parching and pre-parching in the near infrared is a cost-effective, high-resolution way of recording such features. Also, as the grass looses its ability to reflect in the near infrared, it begins to warm relative to the healthy grass, which maintains its viability through reflecting unnecessary wavelengths and by the added cooling of transpiration. Consequently, this divergence in temperature can be imaged thermally but at a much lower resolution than in the usually more appropriate near infrared. Where thermal imaging comes into it own, in the daytime, is in capturing differential transpiration in uniformly healthy grass resulting from variations in moisture at root level, remembering that there can also be a ground surface component to the thermogram, especially during drying after rainfall.
considerations apply to cropmarks.
Here is a tabulated overview which I will update with time:
Visible/thermal spectrum comparison of a wheat field with bronze age burials
by James Pullen using a Mavic 2 Advanced drone
chosen this long barrow as an example as it is thermally
complex. It is of variable height and width with two
hollows, one enclosed and one transverse. The covering of
grass varies from dense to sparse with areas of exposed
soil and embedded limestone. So, with this feature, all
thermal variables come into play both physical and
Heidi Walker stood on the tump flying the kite with the selfie stick plus phone attached to the line
24th March 2019
Near infrared (720nm filter) in sunlight with John Wells in the foreground flying the kite (West to the left)
26th March 2019
and near infrared parts of the spectrum are good for
producing high contrast shadow relief images in sunlight.
However, when looking for differences in reflected near
infrared light (e.g. as a result of stress-induced
absorption in grass and crops etc.) it is best to capture
images in overcast conditions.
a NIR review see here
Vertical near infrared (850nm filter) in overcast conditions
Note the (excavation?) feature cutting into the tump in all the above images (not the transection)
17th July 2019
Thermal infrared in sunlight
The soil was still moist in places because of the cool days and near freezing nights. The air temperature was 13°C and the relative humidity ~55%. There was a very gentle breeze of about 10mph varying from the north to from the west, 3.30 to 4.30pm.
26th March 2019
Thermal infrared in darkness
image is rotated anticlockwise ~50°
compared with the one above.The
linear feature on the left is the path at the back
of the long barrow in the above images. The
temperature was 8°C, having dropped from 14°C at
2pm. The wind was gusting between ~8-22mph causing
considerable instability. The relative humidity
was ~79%, between 8-45pm and
9-45pm. The central right-angled feature was
stable (not flickering).
30th March 2019
Cirencester Roman Amphitheatre in Darkness
is a location covered in grass with the
amphitheatre in darkness. We have been
thermally visualising features at night which
are covered with grass since 2011. Such
observations are often at the limit of
sensitivity of the thermal imager with the
features often flickering in and out of view,
as the field of view slightly changes. The
Flir One is operated with the temperature span
unlocked. Ground-based imaging like this is
easier if the phone is mounted in a 2D headset.
Below is the same view in the visible spectrum.
Cairnpapple at Night
As with the amphitheatre, with the soil covered in grass, are we seeing these minimal differences as a result of a physical or biological phenomenon? Normally transpiration does not take place in darkness, although there are many exceptions. It seems unlikely that the effect is due to thermally visible interstitial ground, but not impossible, and the wind eliminates any localised heating patterns above the surface of the ground. In some instances, the variation in the dampness of the grass may be a factor, especially in ditches and hollows.
Selected Reports, Comments, Notes and Presentations.
Delineation of parch marks in the near infrared (720nm) by former fellow WLAT Trustee Jim Knowles.
(Later ones at Fort William and Dunkeld are not listed)
and John Wells, Infra-red and kite aerial photography,
Scotland's Rural Past Conference, Dunkeld,
Cairnpapple in August 2009
Rosie and John Wells, Visible and near infra-red kite aerial photography of Ogilface sites in West Lothian, Community Archaeology in Scotland (Conference at Queen Margaret University), East Lothian Council and Archaeology Scotland, 82-85, May 2009.
The site of Ogilface Castle, Armadale (720nm NIR filter), looking northwards, November 2007.
Rosie and John Wells, Aerial Photography by Kite, The Archaeologist, 50-51, No74, Winter 2009.
Rosie and John Wells, The History of Torphichen Preceptory and the Order of the Knights Hospitaller in Scotland, Newsletter of the St John Historical Society, Clerkenwell, 7-14, June 2010.
John Wells, Rosie Wells and Jim Knowles, Kite Aerial Photography - Back to basics, Remote Sensing and Photogrammetry Society, ArchSIG Newsletter, 8-9, Spring, 2011.
John Wells, Rosie Wells and Jim Knowles, Gormyre Hill: Geophysical and kite aerial photographic survey and 3D modelling, Discovery and Excavation in Scotland, Journal of Archaeology Scotland, 12, 184, 2011.
Gormyre Hill after snowfall ^North
John Wells, Rosie Wells, Cade Wells and Jim
Aerial Thermography, International Society for Archaeological
Newsletter 29, 9-10, November
Our first thermal infrared kite rig in 2011
Flir PathFindIR 360 x 240 pixel, auto-ranging thermal IR camera unit, incorporating a window heater (6w/2w on/off), 12v/4800mAh, lithium-ion battery, 12v/60w voltage stabiliser and PVR. The original digital recorder was replaced with a PVR (23 September 2011). This rig required very steady winds.
Cade and John Wells
John Wells and Rosie Wells, Castles From The Air, The Castles Study Group Bulletin, 14, 20-21 (Summer) 2012.
The site of Ogilface Castle in sunlight, Armadale (720nm NIR filter, ^NNW).
Ogilface Castle with high contrast overcast NIR (720nm) with features not visible in the visible spectrum.
1. Eastern half near-vertical (2008) 2. South-East quadrant oblique (2008)
3. Central vertical (2010) 4. Western end vertical (2010)
The site of Ogilface Castle, Armadale (thermal infrared, ^E), low-level oblique
Jim Knowles, Geophysical Survey Report, Blackness, Castle field, Earth Resistance Survey, Report No WLAG 001 (Revised), March 2013.
Jim Knowles, Gormyre Camp, A Multidisciplinary Approach, Survey and Excavation Report, Report No WLAG 002, March 2013.
John Wells, West Lothian Archaeological Trust - Kite Aerial Photography, ArcLand Conference k2u2, Dublin, May, 2013.
West Lothian Archaeological Trust, Low Level Aerial Photography, A technique for everyone, Archaeology Scotland Magazine, Issue 16, 15-17, Spring 2013.
John Wells, Kite Aerial Photography: A low level aerial photographic technique for everyone. A one-day workshop of the R.S.P.Soc. UAV Special Interest Group at the University of Worcester, July 2013. (Keynote Presentation)
John Wells, DART Heritage Remote Sensing Horizon Scanning Workshop, Leeds University, September, 2013.
Stott (Leeds Uni.), John Wells (West Lothian Archeological Trust),
Graham Ferrier (Hull Uni.),
Ant Beck (Leeds Uni.), Tom Smith (Imperial Coll.) and Rosie Wells (W.L.A.Trust) behind the camera.
Scaffolding for mounting the thermal imagers
The two thermal imagers as mounted
A Testo 875 imager for stills and calibration (left) and for 24 hour continuous video a Flir PathFindIR
CC-BY DART Project 2012 (David Stott)
A sample Testo 875 thermal image with two heat markers
CC-BY DART Project 2012 (David Stott and Ant Beck)
The location of one of the experimental trenches dug the previous year (NIR 720nm filter)
John Wells, Jim Knowles and Rosie Wells, Kite Aerial Photography, Independent Archaeology, Newsletter of the Council for Independent Archaeology, ISSN 2046-3227, 76, 10-14, November 2013. (Colour reprint of article from Newsletter No 71)
John Wells, The Scottish National Aerial Photography Scheme (SNAPS)*, News Article, History Scotland. Vol 14, No2, 14, March/April, 2014.
Rosie Wells with her main KAP kit on the left and other cameras (and wall posters) shortly before she died.
*SNAPS was set up in memory of Rosie. The thermal rig is in the middle of the table (2012).
John Wells, Jim Knowles, Ron Dingwall and Cade Wells, The Scottish National Aerial Photography Scheme (SNAPS), Rathcroghan, Archaeology Above and Below 2014, Tulsk, Ireland, April 2014. (Abstract)
John Wells, Kite Aerial Photography in the Near and Thermal Infra-Red, Rathcroghan, Archaeology Above and Below 2015, Tulsk, Ireland. April 2015. (Abstract)
Jim Knowles, Kite Aerial Photography: How to Get Started, Its Uses and Safety, Rathcroghan, Archaeology Above and Below 2015, Tulsk, Ireland. April 2015. (Abstract)
John Wells, The Scottish National Aerial Photography Scheme 2013-2016, Rathcroghan, Archaeology Above and Below 2016, Tulsk, Ireland. April 2016. (Abstract)
Jim Knowles, Kite Aerial Photography as used in Professional Archaeology, Rathcroghan, Archaeology Above and Below 2016, Tulsk, Ireland. April 2016.
John Wells, An Introduction to Archaeological Kite Aerial Thermography, Rathcroghan, Archaeology Above and Below 2017, Tulsk, Ireland. April 2017. (Abstract)
Jim Knowles, West Lothian: Above and below, a different perspective, Rathcroghan, Archaeology Above and Below 2017, Tulsk, Ireland. April 2017. (Abstract)
Notes on Linkedin
The Cheapest Way to Begin Kite Aerial Photography (KAP) and Thermography (KAT), Linkedin, 2 August 2017.
Kite Aerial Thermography and Rainfall for Delineating Archaeological Residues, Linkedin, 6 August 2017.
Aerial Thermal Imaging of Archaeological Residues is Not All About Thermal Inertia, Linkedin, 12 December 2018.
We Need to Change Our Mindset When it Comes to Archaeological Thermography, Linkedin, 12 July 2019.
Amplification in Relation to the Delineation of Surface and
Subsurface Archaeological Residues, Linkedin,
26 July 2020.
Handheld and Headset-Based Visualisation of Archaeological Residues Using a 160x120 Pixel Flir One Imager, Linkedin 5 August 2020.
Parchmarks and Cropmarks in the Near and Thermal Infrared, Linkedin 15 August 2020.
Aerial and Ground-Based Thermography: A Framework for
Considering the Theory, Linkedin 24 August 2020.
Ultraviolet Aerial Photography
Sutton Hoo excavations in 1985 (visible spectrum)
© Dr Christopher Brooke, Department of History, University of Nottingham.
Sutton Hoo excavations in 1985 (false-colour ultraviolet 340-400 nm)
Christopher applied this approach to buildings and more recently he has been using thermal imaging (PDF download 2018).
© Dr Christopher Brooke, Department of History, University of Nottingham.
ARCHAEOLOGY IS BOTH HOT AND COOL
Archaeology is both hot and cool
canvas follows heavy rain
simply fail to see