A presentation by Anthony Beck presented at the workshop "Potential of satellite images and hyper/multi-spectral recording in archaeology"
Poznan – 31st June 2012
1. DART – Archaeological detection
Anthony (Ant) Beck
Twitter: AntArch
Potential of satellite images and hyper/multi-spectral
recording in archaeology
Poznan – 31st June 2012
School of Computing
Faculty of Engineering
2. Overview
•How do we detect stuff
•Why DART
•Going back to first principles
•DART overview
•Platforms
•Knowledge base – impact on deployment
3. Archaeological Prospection
What is the basis for detection
We detect Contrast:
• Between the expression of the remains
and the local 'background' value
Direct Contrast:
• where a measurement, which exhibits a
detectable contrast with its surroundings,
is taken directly from an archaeological
residue.
Proxy Contrast:
• where a measurement, which exhibits a
detectable contrast with its surroundings,
is taken indirectly from an archaeological
residue (for example from a crop mark).
5. Archaeological Prospection
What is the basis for detection
Micro-Topographic variations
Soil Marks
• variation in mineralogy and
moisture properties
Differential Crop Marks
• constraint on root depth and
moisture availability changing
crop stress/vigour
Proxy Thaw Marks
• Exploitation of different thermal
capacities of objects expressed
in the visual component as
thaw marks
Now you see me
dont
8. Archaeological Prospection
Summary
The sensor must have:
• The spatial resolution to resolve the feature
• The spectral resolution to resolve the contrast
• The radiometric resolution to identify the change
• The temporal sensitivity to record the feature when the contrast is
exhibited
The image must be captured at the right time:
• Different features exhibit contrast characteristics at different times
11. Why DART? ‘Things’ are not well understood
Environmental processes
Sensor responses (particularly new
sensors)
Constraining factors (soil, crops etc.)
Bias and spatial variability
Techniques are scaling!
• Geophysics!
IMPACTS ON
• Deployment
• Management
15. Why DART? Traditional AP exemplar
Significant bias in its application
• in the environmental areas where it is
productive (for example clay
environments tend not to be
responsive)
• Surveys don’t tend to be systematic
• Interpretation tends to be more art
than science
16. What do we do about this?
Go back to first principles:
• Understand the phenomena
• Understand the sensor
characteristics
• Understand the relationship
between the sensor and the
phenomena
• Understand the processes better
• Understand when to apply
techniques
17. What do we want to achieve with this?
Increased understanding
which could lead to:
• Improved detection in marginal
conditions
• Increasing the windows of
opportunity for detection
• Being able to detect a broader
range of features
18. What do we do about this? Understand the
phenomena
How does the object generate an
observable contrast to it's local
matrix?
• Physical
• Chemical
• Biological
• etc
Are the contrasts permanent or
transitory?
19. What do we do about this? Understand the
phenomena
If transitory why are they
occurring?
• Is it changes in?
• Soil type
• Land management
• Soil moisture
• Temperature
• Nutrient availability
• Crop type
• Crop growth stage
20. What do we do about this? Understand the
relationship between the sensor and the phenomena
21. What do we do about this? Understand the
relationship between the sensor and the phenomena
Spatial Resolution
22. What do we do about this? Understand the
relationship between the sensor and the phenomena
Radiometric Resolution
Radiometric resolution
determines how finely a system can
represent or distinguish differences of
intensity
23. What do we do about this? Understand the
relationship between the sensor and the phenomena
Temporal Resolution
24. What do we do about this? Understand the
relationship between the sensor and the phenomena
Spectral(?) Resolution
http://www.youtube.com/v/Nh-ZB5bxPhc
25. What do we do about this? Understand the
processes better
So what causes these
localised variations?
• Local conditions structure how any
contrast difference is exhibited:
• Soil type
• Crop type
• Moisture
• Nutrients
• Diurnal temperature variations
26. What do we do about this? Understand the
processes better
Expressed contrast differences
change over time
• Seasonal variations
• crop phenology (growth)
• moisture
• temperature
• nutrients
• Diurnal variations
• sun angle (topographic features)
• temperature variations
27. What do we do about this? Understand the
processes better
Exacerbated by anthropogenic
actions
• Cropping
• Irrigation
• Harrowing
28. What do we do about this? Example from multi or
hyper spectral imaging
31. DART: Ground Observation Benchmarking
Try to understand the periodicity of change
• Requires
• intensive ground observation
• at known sites (and their surroundings)
• In different environmental settings
• under different environmental conditions
32. DART: Ground Observation Benchmarking
Based upon an understanding of:
• Nature of the archaeological residues
• Nature of archaeological material (physical and chemical structure)
• Nature of the surrounding material with which it contrasts
• How proxy material (crop) interacts with archaeology and surrounding
matrix
• Sensor characteristics
• Spatial, spectral, radiometric and temporal
• How these can be applied to detect contrasts
• Environmental characteristics
• Complex natural and cultural variables that can change rapidly over
time
33. DART: Sites
Location
• Diddington, Cambridgeshire
• Harnhill, Gloucestershire
Both with
• contrasting clay and 'well draining'
soils
• an identifiable archaeological
repertoire
• under arable cultivation
Contrasting Macro environmental
characteristics
42. DART: Laboratory Measurements
Plant Biology • Soil and leaf water content
• Rate of germination • Root studies
(emergence)
• Root length and density.
• Growth analysis
• Root – Shoot biomass ratio.
• Number of Leaves
• Total plant biomass
• Number of Tillers
• Biochemical analysis: Protein and
• Stem length chlorophyll analysis.
• Total plant height • Broad spectrum analysis of soil
• Drought experiment (Nutrient content) and C-N ratios of
leaf.
• A - Ci Curve
• Chlorophyll a fluorescence
50. DART: Data so far - Permittivity
TDR - How does it work
• Sends a pulse of EM energy
• Due to changes in impedance, at the start and at the end of the probe,
the pulse is reflected back and the reflections can be identified on the
waveform trace
• The distance between these two reflection points is used to determine
the Dielectric permittivity
• Different soils have different dielectric permittivity
• This needs calibrating before soil moisture can be derived from the
sensors
51. DART: Data so far - Permittivity
Further analysis of permittivity and conductivity against rainfall
Linking the changes to the weather patterns
Comparisons can be made between
• Soils at different depths
• Archaeological and non-archaeological features
• Different soil types at the different locations
Conversion to moisture content is also a priority
55. Spectro-radiometry: Methodology
• Recorded monthly
• Twice monthly at Diddington during the growing season
• Transects across linear features
• Taken in the field where weather conditions permit
• Surface coverage evaluated using near-vertical photography
• Vegetation properties recorded along transect
• Chlorophyll (SPAD)
• Height
56.
57.
58. Diddington transect 1: Spectroradiometry June 2011
0.12
R
e
l 0.1
a
t
i
v 0.08
e
r
0.06
e
f
l
e 0.04
c
t
a
n 0.02
c
e
0
400 500 600 700
Wavelength (nm)
27/06/2011 Archaeology 27/6/2011 Outside archaeology 14/06/2011 Archaeology
14/06/2011 Outside archaeology 08/06/2011 Archaeology 08/06/2011 Outside archaeology
59. Diddington transect 1: Spectroradiometry June 2011
0.4
R
0.35
e
l
a
0.3
t
i
v
0.25
e
r
0.2
e
f
l
0.15
e
c
t
0.1
a
n
c 0.05
e
0
350 450 550 650 750 850 950 1050 1150 1250 1350 1450 1550 1650 1750 1850 1950 2050 2150 2250 2350 2450
Wavelength (nm)
27/06/2011 Archaeology 27/6/2011 Outside archaeology 14/06/2011 Archaeology
14/06/2011 Outside archaeology 08/06/2011 Archaeolgy 08/06/2011 Outside archaeology
61. DART: Plant Biology
Lab experiments conducted in collaboration with Leeds Plant
Biology in 2011 and repeated in 2012
From soils at Quarry Field
Soil structure appears to be the major component influencing
root penetration and plant health
65. Open Data: Server (in the near future)
The full project archive will be available from the server
Raw Data
Processed Data
Web Services
Will also include
TDR data
Weather data
Subsurface temperature data
Soil analyses
spectro-radiometry transects
Crop analyses
Excavation data
In-situ photos ETC.
67. Why are we doing this – it’s the right thing to do
DART is a publically funded project
Publically funded data should provide benefit to the public
68. Why are we doing this – IMPACT/unlocking potential
More people use the data then there is improved impact
Better financial and intellectual return for the investors
69. Why are we doing this – innovation
Reducing barriers to data and knowledge can improve
innovation
70. Why are we doing this – education
To provide baseline exemplar data for teaching and learning
71. Why are we doing this – building our network
Find new ways to exploit our data
Develop contacts
Write more grant applications
72. Discussion
SFM Plant Biology
Pushbroom Phenology
High resolution frame Differential growth parameters
Oblique and UAV Data mining (process from
Topographic measurements)
From SFM Environmental
Full Waveform LiDAR Soils
Detection Temperature
Hyperspectral (including thermal)
Spectral Analysis
Visualization
ERT and tomography
Complex data!
74. Overview
There is no need to take notes:
Slides – http://goo.gl/ZHYaB
Text – http://goo.gl/osQZi or http://goo.gl/M5Eu1
There is every need to ask questions
The slides and text are release under a Creative Commons by
attribution licence.
Notas do Editor
Image re-used under a Creative Commons licence: http://www.flickr.com/photos/catikaoe/183454010/We identify contrast Between the expression of the remains and the local 'background' value In most scenarios direct contrast measurements are preferable as these measurements will have less attenuation.Proxy contrast measurements are extremely useful when the residue under study does not produce a directly discernable contrast or it exists in a regime where direct observation is impossible
Traces can be identified through evidence Clusters of artefacts Chemical and physical residues Proxy biological variations Changes in surface relief
Traces can be identified through evidence Clusters of artefacts Chemical and physical residues Proxy biological variations Changes in surface relief
Image re-used under a Creative Commons licence: http://www.flickr.com/photos/arpentnourricier/2385863532Dependant on localised formation and deformation Environmental conditions Soil moisture Crop Temperature and emmisivity
Image re-used under a Creative Commons licence: http://www.flickr.com/photos/dartproject/6001577156Dependant on localised formation and deformation Land management
Image re-used under a Creative Commons licence: http://www.flickr.com/photos/dartproject/6001577156Dependant on localised formation and deformation Land management
Satellite approaches should be considered in a multi-sensor environment which includes ground survey and excavationThe point is to learn more about the past
Image re-used under a Creative Commons licence: http://www.flickr.com/photos/southernpixel/3480710493/Not really.We have great techniques but some are in danger of becoming redundant
Image re-used under a Creative Commons licence: http://www.flickr.com/photos/san_drino/1454922072/Environmental processesSensor responses (particularly new sensors)Constraining factors (soil, crops etc.)Bias and spatial variabilityIMPACTS ONDeploymentManagement
Image re-used under a Creative Commons licence: http://www.flickr.com/photos/jimmysmith/720356377/Changes in land management may reduce the appearance of the phenomena we seekUsing science to maximise crop return
Image re-used under a Creative Commons licence: http://www.flickr.com/photos/tangyauhoong/4502062656/Actual crop returns controlled so they approximate towardsthe 'norm'NEW, i.e. not observed before, archaeology is contained within the tailsThese outlier values are being removed.The outlier is an exceptional year ;-)
Reliant on specific seasonal and environmental conditions Increasingly extreme conditions are required for the detection of ‘new’ sitesLow understanding of the physical processes at play outside the visual wavelengths
Significant bias in its application in the environmental areas where it is productive (for example clay environments tend not to be responsive) Surveys don’t tend to be systematic Interpretation tends to be more art than science
Image re-used under a creative commons licence: http://www.flickr.com/photos/8203774@N06/2310292882/
Image re-used under a creative commons licence: http://www.flickr.com/photos/8203774@N06/2310292882/
Image re-used under a Creative Commons licence:How does the object generate an observable contrast to it's local matrix?PhysicalChemicalBiologicaletcAre the contrasts permanent or transitory?
Image re-used under a Creative Commons licence:If transitory why are they occurring?Is it changes in?Soil typeLand managementSoil moistureTemperatureNutrient availabilityCrop typeCrop growth stage
Image re-used under a Creative Commons licence:
Image re-used under a Creative Commons licence: DARTSpatial Resolution You need enough to observe the object
Image re-used under a Creative Commons licence: DARTRadiometric Resolution - You need enough to be able to physically detect the expressed differencesdetermines how finely a system can represent or distinguish differences of intensity is usually expressed as a number of levels or a number of bits for example 8 bits or 256 levels The higher the radiometric resolution, the better subtle differences of intensity or reflectivity can be represented Signal to noise ratios can be a problem Example It is difficult to detect small changes in growth if your ruler only measures to the nearect 10cms You need enough to be able to physically detect the expressed differences
Image re-used under a Creative Commons licence: DARTYou need to know when to look for the difference
Spectral Resolution You need to know what part of the spectrum to detect the expressed difference Unsure of the geophysical metaphor for this
Image re-used under a Creative Commons licence: DARTSo what causes these localised variations?Local conditions structure how any contrast difference is exhibited:Soil typeCrop typeMoistureNutrientsDiurnal temperature variations
Image re-used under a Creative Commons licence: DARTExpressed contrast differences change over timeSeasonal variationscrop phenology (growth)moisturetemperaturenutrientsDiurnal variationssun angle (topographic features)temperature variations
Image re-used under a Creative Commons licence: DARTExacerbated by anthropogenic actionsCroppingIrrigationHarrowing
Image re-used under a Creative Commons licence: DARTBut archaeology doesn't tend to produce spectral signatures Rather: produce localised disruptions to a matrix The nature of these disruptions vary and include: Changes to the soil structure Changes to moisture retention capacity Changes in geochemistry Changes in magnetic or acoustic properties Changes to topography At least one of these disruptions will produce a contrast which is detectable The challenge is What sensor to use The sensitivity of the sensor When to deploy the sensor
Try to understand the periodicity of changeRequire intensive ground observation (spectro-radiometry, soil and crop analysis) at known sites (and their surroundings) in a range of different environments under different environmental conditions
Based upon an understanding of:Nature of the archaeological residuesNature of archaeological material (physical and chemical structure)Nature of the surrounding material with which it contrastsHow proxy material (crop) interacts with archaeology and surrounding matrixSensor characteristicsSpatial, spectral, radiometric and temporalHow these can be applied to detect contrastsEnvironmental characteristicsComplex natural and cultural variables that can change rapidly over time
Image reused under a Creative Commons Licence:http://www.flickr.com/photos/kubina/279523019Geotechnical analysesGeochemical analysesPlant Biology
Image reused under a Creative Commons Licence:http://www.flickr.com/photos/kubina/279523019Geotechnical analysesGeochemical analysesPlant Biology
Image re-used under a creative commons licence:http://www.flickr.com/photos/soilscience/5104676427Spectro-radiometrySoilVegetationEvery 2 weeksCrop phenologyHeightGrowth (tillering)Flash res 64Including induced events
ResistivityGround penetrating radarEmbedded Soil Moisture and Temperature probesLogging every hour Weather stationLogging every half hour
Image reused under a Creative Commons Licence:http://www.flickr.com/photos/kubina/279523019Geotechnical analysesGeochemical analysesPlant Biology
Image reused under a Creative Commons Licence:http://www.flickr.com/photos/kubina/279523019Geotechnical analysesGeochemical analysesPlant Biology
Image reused under a Creative Commons Licence:http://www.flickr.com/photos/kubina/279523019Geotechnical analysesGeochemical analysesPlant Biology
TDR - How does it workSends a pulse of EM energyDue to changes in impedance, at the start and at the end of the probe, the pulse is reflected back and the reflections can be identified on the waveform traceThe distance between these two reflection points is used to determine the Dielectric permittivity Different soils have different dielectric permittivityThis needs calibrating before soil moisture can be derived from the sensors
Further analysis of permittivity and conductivity against rainfall Linking the changes to the weather patternsComparisons can be made betweenSoils at different depthsArchaeological and non-archaeological featuresDifferent soil types at the different locations
Conversion to moisture content is also a priorityRequires calibration using different mixing models including:empiricalsemi-empiricalphysical volumetric phenomenological modelsThis will help us to link the changes in geophysical responses to the composition of the soil and predict future responses, as well as supporting investigations into crop stress and vigour.
Conversion to moisture content is also a priorityRequires calibration using different mixing models including:empiricalsemi-empiricalphysical volumetric phenomenological modelsThis will help us to link the changes in geophysical responses to the composition of the soil and predict future responses, as well as supporting investigations into crop stress and vigour.
methodology similar to that employed by Parkyn et al. (2011)Overviewdata pointslie within the ditch featureover the non-archaeological featurefind an average data value for the feature and the surrounding soilThe percentage difference between these two figures can then be considered the amount of contrast within the test area.The higher the percentage, the better the feature is able to be defined.