Marine and Petroleum Geology 91 (2018) 723–734

Contents lists available at ScienceDirect

Marine and Petroleum Geology

journal homepage: www.elsevier.com/locate/marpetgeo

Research paper

Drones in carbonate geology: Opportunities and challenges, and application T

in diagenetic dolomite geobody mapping
M.Y.A. Madjida,b, V. Vandeginstea,c,∗, G. Hampsona, C.J. Jordand, A.D. Boothe
a
Qatar Carbonates and Carbon Storage Research Centre and Department of Earth Science and Engineering, Imperial College London, UK
b
Geodynamics Research Group, Institute of Technology Bandung, Indonesia
c
GeoEnergy Research Centre and School of Chemistry, University of Nottingham, University Park, NG7 2RD, Nottingham, UK
d
British Geological Survey, Environment Science Centre, Keyworth, NG12 5GG, Nottingham, UK
e
School of Earth and Environment, University of Leeds, Woodhouse Lane, Leeds LS2 9JT, UK

A R T I C L E I N F O A B S T R A C T

Keywords: Unmanned aerial vehicles (UAVs) or drones have become widely available for use in a broad range of disciplines.
Drone Despite the decreased cost and technological developments of platforms, sensors and software, the use of drones
UAV in carbonate geology has yet to be exploited. Nevertheless, drones offer multiple advantages over traditional
Carbonate field work or high-altitude remote sensing techniques, in that they enable the reconstruction of three dimen-
Dolomite
sional models of inaccessible or unsafe outcrops, and can bridge the spatial scale gap in mapping between
Diagenesis
Heterogeneity
manual field techniques and airborne, high-altitude remote sensing methods. In this contribution, we present our
Reservoir methodology of structure from motion photogrammetry on drone-captured images and highlight opportunities
Geobody and challenges of using drones in carbonate geological studies. Moreover, we apply this method to document the
spatial distribution and dimensions of diagenetic dolomite geobodies in Carboniferous limestone host rock of the
Picos de Europa, northern Spain. The results of our study indicate that dolomite geobodies occur preferentially
near to strike-slip faults, rather than thrust faults or specific types of limestone host rock lithology, in the study
area. The geobody dimensions appear to be related to the strain magnitude of the strike-slip faults. We propose
that these identified links and controls are applicable to structurally-controlled dolomitization on a more general
basis. Our study has demonstrated the potential for using drones in carbonate geological studies.

1. Introduction multiple fields including archaeology (De Reu et al., 2016), natural
hazards (Jordan and Napier, 2016; Mateos et al., 2017; Serban et al.,
In the last decade, unmanned aerial vehicles (UAVs) or drones have 2016; Tamminga et al., 2015), ecology (Anderson and Gaston, 2013;
become widely available for use in a broad range of disciplines (Watts Faye et al., 2016; Flynn and Chapra, 2014; Ouedraogo et al., 2014;
et al., 2012). The success and popularity of these platforms have sig- Ventura et al., 2016; Yang et al., 2014), forestry (Masek et al., 2015;
nificantly decreased their market cost, and have driven technological Padua et al., 2017; Paneque-Galvez et al., 2014; Puliti et al., 2015;
developments of the platforms themselves, their sensors and software Sankey et al., 2017; Tang and Shao, 2015), sedimentary geology
systems. The widespread presence of tried-and-tested rotary and fixed- (Chesley et al., 2017; Nieminski and Graham, 2017), glaciology
wing platforms in the market place, allied to the speed of miniatur- (Bhardwaj et al., 2016; De Michele et al., 2016; Westoby et al., 2015),
isation and development of novel sensors, enable innovative applica- geothermal research (Nishar et al., 2016), atmospheric chemistry
tions. Important advantages over traditional remote sensing systems (Caturegli et al., 2016; Schrod et al., 2017), hydrology (Bandini et al.,
include greater cost effectiveness, higher resolution small budget plat- 2017; Spence and Mengistu, 2016), mining (Jakob et al., 2017; Tong
forms, and the flexibility to acquire data while in the field when con- et al., 2015), environmental science (Hird et al., 2017; Mlambo et al.,
ditions are appropriate. There are various forms of commercially- 2017; Sanders and Masri, 2016), and structural geology (Bemis et al.,
available drones encompassing rotary and fixed-wing, each with asso- 2014).
ciated flight characteristics and limitations such as the ability for ver- Although drones can play an important and innovative role in nu-
tical take-off and landing (VTOL) and the requirement for various sizes merous applications, the use of drones in carbonate geology has not yet
of landing zones. Drones are revolutionizing scientific research in been exploited much. As far as the authors are aware of, there are at the

∗
Corresponding author. School of Chemistry, University of Nottingham, University Park, NG7 2RD, Nottingham, UK.
E-mail address: veerle.vandeginste@nottingham.ac.uk (V. Vandeginste).

https://doi.org/10.1016/j.marpetgeo.2018.02.002
Received 6 October 2017; Received in revised form 17 January 2018; Accepted 2 February 2018
Available online 03 February 2018
0264-8172/ © 2018 Elsevier Ltd. All rights reserved.
M.Y.A. Madjid et al. Marine and Petroleum Geology 91 (2018) 723–734

time of writing this contribution, only two publications that focus on Europa make them an ideal candidate to evaluate the use of drone
the application of drones in sedimentological studies, both doc- technology. We captured photographic images using a small drone to
umenting siliciclastic systems (Chesley et al., 2017; Nieminski and create a high resolution three-dimensional (3D) digital outcrop model
Graham, 2017). Most carbonate geological studies rely on detailed by integrating the digital surface model (DSM) with an assemblage of
description of sedimentological strata and characteristics, diagenetic rectified aerial photographs (orthomosaic, which is a two-dimensional
features and/or structural phenomena. Essential in these studies is planar image containing coordinate position and colour information at
mapping and documenting continuous sections, or correlating (with a each pixel, that has been corrected in terms of different camera per-
high degree of confidence) strata in different outcrops. Accessibility to spectives). We explain this methodology for field mapping, discuss the
outcrops is thus critical. However, ancient carbonate platforms are opportunities and challenges of using drones in carbonate geological
often exposed in the field in very steep cliff faces, such as the Natih applications and interpret the results of our study in the framework of
Formation at Jebel Shams in Oman (Arndt et al., 2014; Homewood carbonate diagenesis and its implications for hydrocarbon reservoirs.
et al., 2008), the Guadalupe Mountains in West Texas and New Mexico This study illustrates the quantitative approach that can be achieved in
(Budd et al., 2013; Kerans et al., 2017), the Jurassic Amellago cliffs in geosciences using an inexpensive method.
Morocco (Bodin et al., 2010; Merino-Tome et al., 2017). The use of
drones can overcome much of this challenge of accessibility to outcrops 2. Geological setting
(Nieminski and Graham, 2017). Although sampling rock material from
cliff faces is not feasible yet with the use of drones, important sedi- The Picos de Europa tectonic unit is situated in the Cantabrian Zone
mentological, diagenetic and structural information can be collected (Fig. 2) which is part of the Iberian Massif. This Massif is the Western
through close-range aerial photography or more advanced imaging European part of the Variscan orogenic belt, which was generated due
techniques. Importantly, this detailed data collection from inaccessible to the continental collision of Gondwana, Laurentia, and peri-Gond-
outcrops enables the possibility of more quantitative geoscientific stu- wana microplates during the late Paleozoic. The shape of the Cantab-
dies, as illustrated in this paper. Applying drones in sedimentary geo- rian zone depicts typical features of a foreland basin developed via a
logical studies provides a low-cost, easy method that helps in the doc- thin-skinned deformation mechanism and fold-and-thrust imbrication
umentation and interpretation of small to intermediate scale features in of a thick sequence of Paleozoic strata (Merino-Tome et al., 2009). The
sedimentary rocks (Chesley et al., 2017). Compared to traditional re- Cantabrian zone is divided into ten tectonic units based on the in-
mote sensing systems based on satellite and aerial photography, pho- tegration of stratigraphic and structural characteristics (Weil et al.,
togrammetry and ground based laser scanning systems, drones offer the 2013).
opportunity to fill the gap in effective spatial range and resolution of From Late Paleozoic to present day, the Cantabrian zone has un-
low-altitude flights with the flexibility of documenting horizontal, in- dergone several tectonic episodes. The main tectonic events include: 1)
clined and vertical surfaces. Compression in the Late Carboniferous with the emplacement of east-
Here, we have applied drone remote sensing to document the spatial west thrusts of the Picos de Europa Province over the Palentine Zone or
distribution of diagenetic dolomite geobodies. The particular im- Pisuerga-Carrión province (Gomez-Fernandez et al., 2000); 2) Exten-
portance of mapping diagenetic geobodies in carbonate rocks lies in the sion in the late Carboniferous-Permian along sub-vertical faults that are
critical aspect of characterising heterogeneity, which can be sedi- oriented N 105°–120° E, and in the north of the Picos de Europa pro-
mentological, structural and diagenetic in origin, and is key to under- vince, along some of the east-west sub-vertical thrusts (Gomez-
stand and predict reservoir quality and performance in subsurface re- Fernandez et al., 2000), Merino-Tome et al. (2009) attributed this ex-
servoirs. Diagenetic alteration in carbonate reservoirs is generally more tensional episode to the formation of wrench and normal faults that cut
extensive than in siliciclastic reservoirs, since carbonate minerals are previous structures due to North Atlantic rifting; 3) The Late Eocene
more reactive than quartz (Ehrenberg and Nadeau, 2005). Since di- Alpine Orogeny with uplift and emergence of the Paleozoic Asturian
agenetic heterogeneities are often subseismic in scale, gathering in- Massif and overlying Mesozoic sediments (Gomez-Fernandez et al.,
formation from outcrop analogues provides important information at a 2000).
critical spatial scale (Vandeginste et al., 2015b). The use of drones in The Carboniferous stratigraphy in the Picos de Europa unit is
addition to manual sampling fills crucial gaps in continuity of doc- characterized by pre-orogenic stratigraphic sequences that consist of
umenting exposed sections by covering inaccessible parts and in high- dolomitized limestone overlain by syn-orogenic stratigraphic sequences
resolution mapping of subseismic-scale heterogeneities by compara- that consist of predominantly shale (Gomez-Fernandez et al., 2000)
tively brief field surveys. The Carboniferous carbonate outcrops of the (Fig. 3). This study focuses on the upper part of pre-orogenic sequences
Picos de Europa National Park in northern Spain (Fig. 1) are well stu- which are known as the Caliza de Montaña Group (Montaña Beds) and
died and are considered reference outcrops for the Carboniferous Picos de Europa Formation. The Montaña Beds consist of two forma-
(Bahamonde et al., 2000; Gomez-Fernandez et al., 2000; Merino-Tome tions, namely the Barcaliente Formation and the Valdeteja Formation.
et al., 2009). The outcrop conditions of the carbonates in the Picos de The Barcaliente Formation is characterized by basinal sequences of

Fig. 1. Location of the study area as indicated by the red polygon. Google Earth maps. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web
version of this article.)

724
M.Y.A. Madjid et al. Marine and Petroleum Geology 91 (2018) 723–734

Fig. 2. Simplified tectonic unit map of the Cantabrian zone. The study area is location within the Picos de Europa unit as highlighted by the red rectangle in the figure (modified after Weil
et al., 2013). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

dolomite geobodies. Detailed outcrop observation of selected geobodies

was also carried out on the ground, to collect additional information on
dolomite geobody texture, colour, and dolomite-limestone contacts
which served as validation for the drone mapping of spatial dolomite
distribution.
In this study, we used a DJI Phantom III Professional drone (quad-
copter) with a mounted 12 Mega Pixel digital camera that can tilt 0 to
90°. Before conducting drone mapping, we selected flight paths and
areas for each flight mission. Before take off, calibrations were made to
ensure the platform was orientated prior to take-off. For each flight
mission, we selected the drone take-off area generally at the top of a hill
and in the centre of the survey area in order to cover the area effec-
tively. For most missions, the drone was set to take aerial photographs
using “autopilot mode” with a camera facing directly downwards for a
hilly terrain. A few surveys were conducted with the camera mounted
45° sideways to enable high-quality capture of data from steep cliff
faces. We selected 70–90% forward and sideways overlap of images. We
carried out more than 80 flight missions, capturing a total of 4606
pictures, and mapped in total an area of about 3.7 km2 in seven and a
half days of drone field work with a capacity of six drone batteries.
There are several software packages available, both free and paid
versions, that can be used to create digital surface models and ortho-
mosaic from the drone captured photographs. In this study, we used
Pix4DMapper Pro software which applies Structure from Motion (SfM)
photogrammetry to process raw images from the drone.
SfM photogrammetry identifies a key feature (key-point) from
overlapping images automatically and reconstructs it within 3D space
as a single point without knowing the camera position and scene geo-
metry (Westoby et al., 2012). The camera position and scene geometry
Fig. 3. Stratigraphic sequence of the Picos de Europa province. This study focuses on the are reconstructed by the automatic identification of the identical key-
Caliza de Montaña (Montaña Beds) and the Picos de Europa Formation as highlighted on points in multiple images (Westoby et al., 2012). These key-points are
the figure (modified after Gomez-Fernandez et al., 2000). traced in every individual image which are then processed iteratively
using non-linear least squares minimisation (Snavely et al., 2008) so
well-bedded, dark, microlaminated and non-fossiliferous lime mud- that the original camera positions and object coordinates (tie-point) can
stones, whereas the Valdeteja Formation consists of 300 m of thick- be estimated (Westoby et al., 2012) (Fig. 4). In addition, we selected a
bedded carbonate breccias of micritic lithoclasts and microlaminated few ground control points for each flight mission. Subsequently, 3D
mudstones separated by thin, well-bedded graded grainstones with models (e.g. point clouds) are created from the images by using multi-
spiculitic alternations (Bahamonde et al., 2000). The Picos de Europa view-stereo algorithms (Chesley et al., 2017; Seitz et al., 2006; Westoby
Formation is characterized by interbedded limestones and shales in the et al., 2012).
lower part and massive bioclastic limestone in the upper part (Gomez- A standard SfM workflow (Chesley et al., 2017; Westoby et al.,
Fernandez et al., 2000). 2012) has typically five steps: 1) Identification of key-points and
creating feature descriptors using Scale Invariant Feature Transform
(SFIT) object recognition system (Snavely et al., 2008); 2) Matching of
3. Methodology
all identical key-points between images; 3) Estimating camera position,
reconstruction of scene geometry and 3D position using triangulation
Fieldwork was conducted in the Picos de Europa national park,
method to create 3D low-density (sparse) point clouds; 4) Point cloud
Northern Spain, where we deployed a drone to capture overlapping
densification; 5) DSM and orthomosaic generation. In our study, due to
aerial photographs in order to document the spatial distribution of

725
M.Y.A. Madjid et al. Marine and Petroleum Geology 91 (2018) 723–734

Fig. 4. Schematic illustration of SfM photogrammetry for

the creation of a 3D model from overlapping aerial photo-
graphs. (A) SfM photogrammetry workflow diagram. (B)
Identical key-points are observed in four overlapping
images captured by the drone's camera and are subse-
quently used for reconstructing camera position and an
object coordinate (tie-point). The background image is an
aggregate of dense point clouds, whereas the four front
pictures are thumbnails of drone captured images. Note that
although thumbnails are presented next to each other and
much smaller than the background image, they show
overlapping pictures of the area presented.

Table 1
Processing quality report of all models.

No. Model Total area Average Ground Median of Key-points Median of Matches Over-lapping DSM and ortho- Absolute Geolocation RMS
name covered (km2) Sampling Distance (per calibrated image) (per calibrated image) images (per each mosaic resolution Errors
(cm) pixel) (cm)
X (m) Y (m) Z (m)

1 full model 3.7 6 3147 1058 >5 25 4 4 23

2 day 2 0.77 6 45653 25843 >5 6 3 4 5
3 day 3 0.99 5 3059 992 >5 20 2 3 2
4 day 4 1 6.5 3556 1350 >5 13 3 4 3
5 day 6 0.56 10 3321 1658 >5 10 6 2 3
6 day 8 0.78 5.5 3014 1154 >5 5 4 2 1
7 day 9 1.5 9 3126 1116 >5 9 3 3 7
8 day 10 0.91 6 3287 1329 >5 6 2 2 4

a huge number of raw images, processing was divided into two levels (Fig. 5), whereas seven smaller DSMs and orthomosaic models were
based on coverage area, such as a full study area and seven smaller created for parts of the study area at a higher resolution of approxi-
areas with higher resolution. The processing quality of each model is mately 10 cm. The high-resolution models were used to identify small-
documented in Table 1. scale dolomite bodies that are challenging to distinguish in the lower
The distinction of dolomite bodies was made primarily based on resolution model, and this information was then transferred to the
colour difference, as discussed in the results section below. The bodies model of the full study area. The geological map of the Careña-Cabrales
were then mapped manually on the orthomosaic using ArcGIS software. area (Garcia et al., 1977) and the DSM were used as a base for identi-
Dolomite mapping was carefully conducted particularly close to areas fication and interpretation of stratigraphic and structural features in
that were extensively covered by vegetation, snow patches, and aban- our model. Two stratigraphic units, three strike-slip faults, and two
doned mine pits. Subsequently, three-dimensional geometrical char- thrust faults are identified in the study area. Furthermore, we created a
acteristics of dolomite bodies such as length, width and height were fault proximity map in the ArcGIS software to assess abundance and
calculated from mapped dolomite polygons in ArcGIS software. The size of dolomite bodies in the proximity of faults in the study area, in
texture and 3D characteristics of the DSM were mainly of importance in the context of our investigation of the potential structural control on
the calculation of the dimensions of the bodies, whereas the distinction dolomitization (Fig. 6).
between dolomite and limestone was made from the orthomosaic,
primarily based on colour. Length and width estimates of dolomite
bodies were generated using the “minimum bounding geometry” 4.2. Mapping of dolomite bodies
toolbox, with the smallest box width as the boundary. This method
creates a rectangular geometry covering the true irregular dolomite Dolomite geobodies can be distinguished from the limestone host
shape and generates length and width values automatically, as well as rock by their texture and colour contrast. Dolomite appears more
the orientation of the long axis of the rectangle. The height of dolomite brownish and blackish in colour, is highly fractured, and has an irre-
bodies was derived from the integration of mapped dolomite polygons gular shape with sharp and discordant contacts with limestone host
with the DSM. All dolomite polygons were assigned to the DSM using rock. In contrast, limestone is commonly characterized by light grey
the “interpolate shape” toolbox, and a height value was created by colour, less fractured appearance and more massive texture (Fig. 7).
subtracting the difference between minimum and maximum elevations The dolomite cuts through limestone beds, and there are a few rem-
(Z value) of the bodies. nants of pre-existing limestone host rocks enclosed by dolomite bodies.
The dolomite - limestone contacts are sharp and irregular (Fig. 7). The
brown dolomite colour versus grey limestone colour (on the pictures
4. Results taken by drone) is the surface or weathering colour in the outcrop,
which can be slightly different than fresh rock surfaces. A colour con-
4.1. Models trast between limestone and dolomite has been described also in study
areas in Iran (Lapponi et al., 2011; Sharp et al., 2010), Italy (Bistacchi
Two levels of models were generated with different resolutions, to et al., 2015; Jacquemyn et al., 2015), Oman (Beckert et al., 2015;
enable smooth handling with consumer-grade computers. A DSM and Vandeginste et al., 2015a), Spain (Dewit et al., 2012; Vandeginste et al.,
orthomosaic of the full study area was created at a resolution of 25 cm 2015b), United Arab Emirates (Fontana et al., 2010), and elsewhere.

726
M.Y.A. Madjid et al. Marine and Petroleum Geology 91 (2018) 723–734

Fig. 5. (A) DSM of the study area with spatial resolution of 25 cm. The altitude ranges from 1350 m to 1984 m. (B) Orthomosaic of the study area with spatial resolution of 25 cm.

The weathering colour is linked to: 1) the chemistry of the rocks, for 4.3. Spatial distribution of dolomite geobodies
example, iron-rich rocks show a rusty brown weathering colour
(Beckert et al., 2018; Vandeginste and John, 2012; Vandeginste et al., The two lithostratigraphic units in the study area contain several
2015a), and 2) the rock texture and mineralogy, fine grained limestone limestone host rock lithologies, including microlaminated lime mud-
with some clay impurities or organic material is generally grey, stones and inter-bedded carbonate breccias and grainstones in the
whereas coarse crystalline dolomite has a light or brownish colour Montaña Beds (Bahamonde et al., 2000), and massive limestones and
(Vandeginste et al., 2005). The limestone host rock in the Picos de interbedded limestones and shales in the Picos de Europa Formation
Europa Formation consists of bioclastic packstones and grainstones (Gomez-Fernandez et al., 2000). There is little apparent difference in
with a fine grained muddy matrix and fragments of sponges, ooids, the distribution of dolomite bodies between the Montaña Beds and the
foraminifera, coralline algae, microbial mats, crinoids and gastropods, Picos de Europa Formation (Fig. 8), although dolomite bodies form a
whereas the dolomite bodies have a fabric-destructive coarse-crystal- greater proportion of the surface area of the Montaña Beds (3%) than of
line texture with dolomite crystals from 0.4 to 3 mm that have growth the Picos de Europa Formation (1%). Although the primary fabrics of
zones identified using cathodoluminescence microscopy (Racine, the limestone host rock lithologies have been extensively modified lo-
2017). The dolomite has higher concentrations of iron, manganese and cally by dolomitization, the distribution of dolomite bodies does not
barium compared to the limestone (Racine, 2017), which influences the pick out stratigraphic units or conform to stratigraphic contacts (e.g.
weathering colour. The microscopic features of the host limestone and the boundary between the Montaña Beds and the Picos de Europa
dolomite in the studied formations have also been described and illu- Formation in Fig. 8, as mapped by Garcia et al., 1977). Instead, dolo-
strated by Bahamonde et al. (2000) and Gomez-Fernandez et al. (2000). mite bodies appear to occur in clusters that tend to be aligned along and
Fluid inclusion data demonstrate that the dolomitizing fluids were lie in close proximity to the surface traces of strike-slip faults, as de-
highly saline, 20–23 wt% NaCl eq., and relatively warm, 100-170 °C scribed below. This dominant link of dolomite bodies with strike-slip
(Racine, 2017). Similarly, Gomez-Fernandez et al. (2000) reports tem- faults is clear on the larger (> meter) scale. At the smaller (decimeter)
peratures of 170–200 °C and salinities of around 15 wt% NaCl eq. for scale, the discordant dolomite-limestone contact is to some extent af-
dolomite in the Picos de Europa area. fected by bed-scale lithological difference, as the irregular contact line
In addition to mapping on the 3D models, we conducted checks in may show small extensions in certain beds, similar to structures dis-
the field by direct observation of dolomite bodies in the outcrops cussed in Vandeginste et al. (2013).
matched to the locations in the models. In our analysis, we grouped the The distribution of the dolomite bodies is slightly more con-
dolomite bodies based on the stratigraphic unit they are hosted in. In centrated close to strike-slip faults in the study area (Fig. 9). Most of the
total, more than 400 dolomite bodies were mapped in the study area, Montaña Beds hosted dolomites are located near the San Carlos Fault,
with a predominant occurrence of the bodies in the Montaña Beds. The whereas almost all Picos de Europa hosted dolomites are located near
percentage surface area of dolomite bodies within the Montaña Beds is either the Middle Fault or the Northern Fault. The number and the total
around 3%, whereas in the Picos de Europa Formation it is around 1%, surface area of dolomite bodies both exhibit an overall decrease away
with the total surface area of the bodies being around 0.07 km2 (Fig. 8). from the San Carlos Fault in a 600 m wide zone. A total of 185 bodies
Dolomite bodies cannot be determined in some areas that are ex- are concentrated within 100 m of the fault. The total surface area of the
tensively covered by vegetation, scree, and abandoned mine pits. The dolomite is more than 31,000 m2.
total surface area of the undetermined zone is around 0.40 km2. For the dolomite bodies that are located near the Middle Fault, the

Fig. 6. (A) Fault proximity map showing various distance intervals (values in m) to investigate the distribution of dolomite bodies as a function of distance from the nearest faults. (B) The
geological map of the study area based on the Careña-Cabrales map by Garcia et al. (1977).

727
M.Y.A. Madjid et al. Marine and Petroleum Geology 91 (2018) 723–734

Fig. 7. Distinction of dolomite bodies from limestone at

different observation scales. Contacts between dolomite and
limestone are marked in yellow. (A) Uninterpreted and (B)
interpreted outcrop photograph of dolomite body and host
limestone, with field compass for scale. The limestone
bedding is indicated by a red dashed line, whereas dolomite
bodies are delineated by a yellow line. The picture also il-
lustrates the colour difference and the more fractured tex-
ture of the brown dolomite versus the more massive nature
of the grey limestone. (C) Uninterpreted and (D) interpreted
aerial photograph consisting of irregularly-shaped dolomite
bodies (delineated by a yellow line) and host limestone.
(For interpretation of the references to colour in this figure
legend, the reader is referred to the Web version of this
article.)

number and the total surface area of dolomite bodies also generally 4.4. Dimensions of dolomite bodies
decrease away from the fault. Within a distance of 100 m from the fault,
there are 54 dolomite bodies with a total surface area of more than The width of dolomite bodies within both the Montaña Beds and
3700 m2. It is interesting to note that there is a large dolomite body Picos de Europa Formation show lognormal distributions (validated by
located about 270 m away from the fault with a surface area of more chi-squared test at 95% confidence level) with mean values of 9 m and
than 800 m2, whereas the total surface area of surrounding bodies does 8 m for dolomites within the Montaña Beds and Picos de Europa
not exceed 300 m2. Formation, respectively (Fig. 10). About 90% of the dolomite bodies
In the area of the Northern Fault, the number of dolomite bodies within the Montaña Beds are less than 20 m wide, whereas 90% of the
also generally decreases with distance from the fault. Although a si- dolomite bodies in the Picos de Europa Formation are less than 15 m
milar trend of decreasing dolomite total surface area with distance from wide. The majority of dolomite bodies both within the Montaña Beds
the fault is observed, the trend is different within a 30 m range of the (about 83%) and the Picos de Europa Formation (about 88%) are
fault. A total of 45 dolomite bodies with a total surface area of more 0–30 m long. The dolomite body length data show lognormal dis-
than 4700 m2 occur within 50 m distance of the fault. tributions (validated by chi-squared test at 95% confidence level) with
Compared to the presence of dolomite bodies close to strike-slip mean values of 20 m and 18 m for the bodies within the Montaña Beds
faults, there seems to be a much smaller number of bodies located close and Picos de Europa Formation, respectively (Fig. 10). Similarly, the
to thrust faults. The distribution of the number of dolomite bodies in the dolomite body height data also have a lognormal distribution (validated
area around the Northern Thrust is quite irregular, whereas the total by chi-squared test at 95% confidence level) with mean values of 14 m
surface area of the geobodies generally increases away from the thrust and 6 m in the Montaña Beds and Picos de Europa Formation, respec-
fault. There are only two dolomite geobodies of area 162 m2 within tively. About 95% of the body height values within the Montaña Beds
150 m of the Southern Thrust, which represents a population that is too are less than 40 m, whereas they are less than 15 m in the Picos de
small to analyse quantitatively. Europa Formation (Fig. 10).

Fig. 8. Detailed interpreted map of dolomite bodies in the

study area. The uninterpreted model is presented in Fig. 5B,
and an example of uninterpreted and interpreted aerial
photograph is presented in Fig. 7C–D. The northwestern
limit of the study area is mainly covered by grass, whereas
the southeastern part of the study area contains regions
covered by scree and abandoned mine pits. The dolomite
bodies cover a slightly larger surface area in the Montaña
Beds than in the Picos de Europa Formation, and they are
slightly more concentrated close to strike-slip faults. The
majority of the bodies are less than 30 m long and less than
15 m wide.

728
M.Y.A. Madjid et al. Marine and Petroleum Geology 91 (2018) 723–734

Fig. 9. The number and total surface area of dolomite bodies as a function of distance from (A) the San Carlos Fault, (B) the Middle Fault, (C) the Northern Fault, and (D) the Northern
Thrust. There are only two dolomite bodies of area 162 m2 within 150 m of the Southern Thrust in the Picos de Europa Formation in the study area.

We identified a weak positive correlation between width and length available at a range of prices, models, and flexibilities in terms of
of the dolomite bodies based on a coefficient of determination of 0.62 compatible sensor systems. However, there are certain aspects that
and 0.41 for the geobodies in the Montaña beds and the Picos de Europa make them in general more or less suitable for the type of geological
Formation, respectively (Fig. 10). The correlation indicates that dolo- work intended. Decision on the type of system to use should align with
mite body width is about one third of the dolomite body height in both the research question to be answered and the terrain that is being
stratigraphic units (Fig. 10). In terms of height-length relationship, the analysed. Fixed-wing drones are more suitable to map relatively large
body height within the Montaña Beds is about 70% of the length areas (3 km2 per day), they generally achieve longer flight times (about
(correlation with coefficient of determination of 0.74), whereas in the 50 min) and are more expensive (£10,000 or more) compared to con-
Picos de Europa Formation, the body height is 33% of the length sumer-grade multi-rotor systems. Fixed-wing drones achieve typically a
(correlation with coefficient of determination of 0.48) (Fig. 10). resolution of about 5 cm per pixel for a flight at about 150 m height
The length axis orientations of dolomite bodies were used for (Campana, 2017) and they are generally capable of flying in higher
structural analysis and plotted on a rose diagram, separately for the two winds than rotary systems. The most common consumer-grade drone
stratigraphic units (Fig. 11). The orientation of dolomite bodies hosted systems are quadcopters (i.e. equipped with four propellers), and cost
within the Montaña Beds show a wider distribution than those in Picos about £2000 to £5000. Larger rotary systems are available that can
de Europa Formation. The dominant orientation of the dolomite bodies carry heavier payloads, such as LiDAR (Light Detection and Ranging),
is N 110 E in the Montaña Beds, and N 130 E in the Picos de Europa but these can cost up to £250,000. Compared to fixed-wing systems, the
Formation. multi-rotor drones need a much smaller area for take off and landing,
which significantly enhances their flexibility in the field, as they can
take off from very irregular rocky surfaces. Their flight time is generally
5. Discussion limited to no more than 20 min due to battery capacity. The multi-rotor
drones have accurate positioning, a good level of automation and a high
5.1. Types of drones and their application in carbonate geology geometric resolution of data acquisition (Campana, 2017) similar to
fixed-wing systems. In contrast to fixed-wing drones, they can be used
There are two main types of drones of interest in carbonate geolo- for imaging vertical cliff faces in great detail (Nieminski and Graham,
gical studies, namely fixed-wing and multi-rotor systems. Both types are

729
M.Y.A. Madjid et al. Marine and Petroleum Geology 91 (2018) 723–734

Fig. 10. Histograms of (A) length, (B) height, and (C) width of the dolomite body's dimensions, and cross plots of dolomite body (D) width versus length and (E) height versus length.

Fig. 11. Rose diagrams of the orientation of the dolomite body long axes, showing a dominant NNW-ESE to NW-SE orientation.

730
M.Y.A. Madjid et al. Marine and Petroleum Geology 91 (2018) 723–734

2017), a critical advantage for carbonate geological field work. Hence, operators who conduct commercial flights. Finally, drone aerial pho-
such drones could be used to map world class cliff outcrops such as the togrammetry fills a real gap in effective spatial scale and resolution
Guadalupe Mountains, the Miocene carbonate cliffs in Mallorca, the between that of manual field mapping or ground-based LiDAR and that
Amellago cliffs in Morocco, etc. of traditional airborne higher-altitude methods.

5.2. Drones and imaging sensors in carbonate geology 5.4. Applying drone aerial photogrammetry for mapping of dolomite
geobodies
Alongside the type of platform (drone), also the type of sensor and
data processing are remote sensing components that should be aligned The application of drone based remote sensing is an innovative way
with the research objectives. Moreover, the type of platform and sensor to map diagenetic geobodies. In only seven days of survey, we mapped
are interlinked, since for example heavy SLR digital cameras cannot be a total area of 3.7 km2 by drone remote sensing in mountainous terrain.
carried by a small, lightweight, consumer-grade drone. In our study, we One could achieve larger areal coverage depending on greater avail-
used a panchromatic optical sensor, capturing the visible portion of the ability of batteries, selection of a less mountainous area requiring less
electromagnetic spectrum, in a 12 Mega Pixel digital camera. There are daily hiking, or the use of a different type of drone (although heavier
also commercially-available multispectral and hyperspectral sensors platforms would be less practical when significant hiking is needed).
that can capture shortwave to thermal infrared and near-ultraviolet Also the extent of the coverage is linked to the targeted resolution.
portions of the electromagnetic spectrum (Verhoeven, 2012; Verhoeven However, in mountainous regions, the maximum resolution will depend
and Schmitt, 2010). In contrast, advances in active source sensors for on the topography and extent of area to be captured in one flight. In our
high resolution 3D data collection mainly focus on laser scanning study, we acquired models with a resolution of 5 cm–25 cm, which is
technologies such as LiDAR, which measures the distance to a target comparable to the resolution of airborne LiDAR-based models, with
based on the travel time of reflected light (Bemis et al., 2014; Hodgetts, typical accuracy of up to 20 cm (Jones et al., 2010). Processing time
2013). Although airborne LiDAR is possible using drones, they are depends strongly on the number of photos, coverage area of study, the
currently expensive, and further developments on these systems are user-defined processing quality, and the performance of computers that
needed to miniaturise them as suitable payloads for low-cost platforms. will be used for creating the DSMs and orthomosaic (i.e. a standard
consumer-grade laptop took about 24 h to process the full model).
5.3. Practicality of using drones in geological surveys Besides the highlighted advantages and opportunities of using a
small drone for mapping geological features, the following two aspects
Obtaining detailed maps of 3D models from surveys of sedimento- are more challenging. First, the presence of vegetation on the surface
logical, diagenetic and structural study sites can be a key component in may compromise slightly the accuracy of the data collected. Depending
carbonate geological research, and provide the fundament for further on the type of vegetation, information on the rock underneath can still
analysis. In the past, such research focused mainly on selected sites, be collected by manual survey or by the use of a laser scanning tech-
documenting features manually, and was labour intensive. nique. Also, point cloud processing software of greater sophistication
Developments in GPS-based surveys and subsequently laser scanning than used in our study (such as CloudCompare) allows manual editing
methods have greatly improved accuracy, precision, and extent of and rendering of 3D point clouds and triangular meshes with advanced
mapping. Laser scanning techniques are, however, still time consuming, processing algorithms, which may be helpful in taking account of ve-
expensive, and restricted by environmental and field conditions (visi- getation cover. LiDAR collects a 3D point location with a single pulse of
bility, morphology, etc). The advance of extraordinarily fast, cost-ef- light that collects ground surface points as long as the pulse of light can
fective and high geometric resolution aerial photogrammetry using pass through the vegetation (Bemis et al., 2014). LiDAR derived 3D
small drones is revolutionary compared to the traditional methods. models have been used in previous studies to characterize dolomite
Quantitative measurements on 3D models constructed from drone bodies, for example in the Latemar Platform, Italy (Jacquemyn et al.,
surveys have greater repeatability than those from outcrops, are col- 2015), and the Zagros Mountains, Iran (Lapponi et al., 2011). However,
lected in only a fraction of the time, and enable volumetric measure- LiDAR systems are at this time still too heavy and too costly to be used
ments (Chesley et al., 2017). Moreover, using the drones for field sur- on small drones. In our study, uncertainty in dolomite presence in a
veys is very flexible and practical. The small drone that we used in our small part of the model is predominantly caused by the occurrence of
study is lightweight and can be easily carried by one person in the field, snow patches or vegetation cover. Nevertheless, the extreme high de-
which makes it ideal in terms of mobility for mapping areas in moun- finition is rarely necessary, and therefore, the real need for LiDAR is
tainous regions, such as the Picos de Europa, which require long daily fairly limited. Second, although volumetric measurements can be made
hikes to gain access. Furthermore, one can deploy a drone at short directly from the model and are helpful for estimating resources in
notice, and at a wide range of times and localities, which presents major analogue subsurface reservoir models (Chesley et al., 2017), the drone
advantages over the hire of a traditional light aircraft or helicopter. based 3D model represents what we observe at the surface, but does not
Legislation on drone use, however, is evolving, and needs some in- provide subsurface information that could be derived from seismic
vestigation when flying the drone, especially as local rules may differ by imaging, for example. In many cases, geoscientists, and in particular
country or district. Restrictions generally apply in terms of flight height, sedimentary geologists, aim for an understanding of the 3D geometry of
no-fly zones, and permission from air traffic control. Drone regulations the geobodies that extend into the subsurface. The fact that we can only
in Spain are strict and prohibit flying a drone more than 100 m high or delineate the rims of the dolomite bodies at the surface remains a
above stadiums, streets, concerts, any political building, beaches, ca- limitation and represents uncertainty in the reconstruction of dolomite
thedrals, etc. One needs a license to use the drone and has to be at least body dimensions, even though the irregularity of the topography pre-
18 years old to be legally able to fly it. To give an example from another sents an advantage in determining 3D dimensions. As a consequence,
country, the Civil Aviation Authority UK Dronecode involves the fol- the reported data on dolomite body dimensions should be interpreted as
lowing: 1) always keep your drone in sight, 2) keep your drone below minimum dimensions.
120 m height, 3) follow the manufacturer's instructions, 4) keep 50 m Despite these challenges, drone based remote sensing has proven to
distance from people and properties and 150 m distance from crowds be very effective for the mapping of dolomite geobodies, and this
and built up areas, 5) legal responsibility lies with the drone operator, method could offer many other opportunities for geological research.
and 6) stay well away from aircrafts, airports and airfields. If one Further improvement in our method could develop semi-automated
wishes to fly the drone closer to buildings and crowds, then permission dolomite body mapping, similar to semi-automatic interpretation de-
from the CAA needs to be obtained, similar to that required by drone veloped for mapping other geological structures (Vasuki et al., 2014). In

731
M.Y.A. Madjid et al. Marine and Petroleum Geology 91 (2018) 723–734

addition, further developments could focus on the application of other fractures, compared to the other faults in the study area. Fault damage
types of sensors such as hyperspectral imaging and the use of texture zones with higher fracture density or larger aperture of fractures
mapping (Beckert et al., 2018; Kurz et al., 2012). Moreover, the DSM probably acted as preferential conduits for dolomitizing fluids and, as a
from drone mapping could also be integrated with data from other consequence, generated larger dolomite bodies. Further evidence of
sources, potentially using near-surface geophysical surveys such as fluid flow along fault zones is derived from the microscopic and fluid
ground-penetrating radar (GPR). GPR offers 3D geometrical re- inclusion data indicating that relatively warm, highly saline fluids
presentations of subsurface geological horizons (Fernandes et al., formed discordant bodies of fabric-destructive coarse dolomite (Gomez-
2015), such as those at the contact of boundaries of dolomite bodies. Fernandez et al., 2000; Racine, 2017; Vandeginste et al., 2015b). Hy-
Combination of the DSM and a 3D GPR dataset may allow identification drothermal dolomitization has also been documented in Carboniferous
of the 3D geometry of dolomite bodies and their degree of connectivity. host limestone in the Bodón unit of Cantabria (Gasparrini et al., 2006;
A previous study of GPR gave promising results, where the method Lapponi et al., 2014), southwest of the study area.
generated images of stratabound dolomite bodies (Tsoflias and Becker,
2008). 5.6. Future developments in drone based remote sensing

5.5. Spatial distribution and dimensions of dolomite geobodies controlled by Despite the major advances in drone based remote sensing cap-
strike-slip faults abilities and flexibility, there is still ample room for improvement.
Technical developments are still needed to extend flight times of the
The detailed map of the distribution of the dolomite bodies in the small platforms, increase their lifting capacity and improve flexibility of
study area shows that these geobodies are not regularly or randomly the software. The most recent small platforms have seen major im-
distributed, but rather occur in a clustering pattern, in particular those provement in advanced collision detection already. Another challenge
hosted in the Montaña Beds. An analysis of dolomite body proximity to is that the small drones are affected by strong winds, which makes them
faults reveals that there is a strong structural control on the spatial significantly weather dependent. In current systems, terrestrial laser
distribution of dolomite bodies. Despite the lack of correlation between scanning still provides higher resolution data and more consistency
the thrust faults and dolomite bodies, the geobodies are generally more compared to the drone-SfM method (Wilkinson et al., 2016). Greater
abundant close to the nearby strike-slip faults. This link reveals, im- diversity and more revolutionary opportunities will result from further
portantly, the predominant control of strike-slip faults on dolomitiza- advances in lightweight sensors that can measure and analyse non-
tion, rather than thrust faults in the study area, and is in agreement visible portions of the electromagnetic spectrum, such as infrared and
with interpretations of the common association of hydrothermal dolo- hyperspectral imaging, and lightweight, cost-effective LiDAR systems.
mite with extensional or strike-slip faults by Davies and Smith (2006).
The results indicate that the geometry of dolomite bodies is prob- 6. Conclusions
ably not strongly controlled by the limestone host rock lithology (ex-
cept for minor control on the irregular dolomite boundary at decimeter The use of drones with SfM photogrammetry offers the opportunity
scale), since the dolomite bodies cut through limestone beds. This in- to build 3D models from geometrically complex or critical rock ex-
terpretation is consistent with other dolomite body geometry studies in posures. The method is particularly powerful to collect data from out-
which it was demonstrated that there is no distinct link between the crops that are inaccessible or unsafe to access by foot, providing the
host rock limestone texture and the extent of the fault-related dolomite possibility to build large quantitative data sets. The limited infra-
body away from the fault into the host rock (Vandeginste et al., 2013). structure requirements and resulting greater portability and decreased
Analysis of comparison between dolomite bodies across regions in- cost enable the reconstruction of photo-based high resolution 3D
dicates that the general dolomite body dimensions probably relate to geospatial data at relatively low cost and expertise, and relatively
the permeability of the host rock limestone at the time of dolomitiza- quickly. Drone aerial photogrammetry is especially useful to bridge a
tion, or to the source and fluid flow driving mechanisms for dolomi- real gap in effective spatial scale and resolution between that of manual
tizing fluids (Vandeginste et al., 2014; Wilson et al., 2007). The di- field mapping or terrestrial laser scanning and that of traditional air-
mensions (width, length, and height) of dolomite bodies in the study borne higher-altitude methods.
area range from a few meters to about 200 m. The length/width ratio of The method explained in this paper illustrates how we can char-
the dolomite bodies is on average about 3, similar to that for dolomite acterize diagenetic geobodies in an effective way at subseismic scale
bodies documented in the southern part of the Picos de Europa unit with appropriate accuracy using high resolution drone captured images
(Vandeginste et al., 2015b). The length orientation of the dolomite in carbonate geological studies. The results of our study indicate that
bodies seems to be strongly controlled by structural features. In the the dolomite geobodies preferentially occur near to strike-slip faults,
study area, the dominant orientation of the dolomite body length axis is rather than thrust faults or types of host rock limestone lithology in the
sub-parallel to the orientation of strike-slip faults (NNW-SSE to NW-SE), study area. Importantly, the dolomite geobody distribution and or-
reinforcing the identified link between the distribution of dolomite ientation can be plausibly related to the strain magnitude of the strike-
bodies and strike-slip faults (preferentially over thrust faults). slip faults. We propose that these links and controls may be applicable
Generally, the dolomite bodies located near the San Carlos fault are more generally to structurally-controlled dolomitization.
larger than those located near the Middle and Northern Fault (Fig. 9). Further developments in drone based remote sensing are still
The difference in length and width of dolomite bodies can be inter- needed to improve flight time, lifting capacity, software flexibility and
preted as the difference in the strain magnitude of nearby strike-slip greater range of lightweight sensors. Advances on these fronts will
fault zones. Fault zones commonly consist of a fault core and an asso- enable further revolutionary opportunities not only in geological re-
ciated damage zone that is mechanically related to the growth of the search, but also in a wide range of other disciplines.
fault zone (Billi et al., 2003; Caine et al., 1996; Putz-Perrier and
Sanderson, 2008). The damage zone is composed of a volume of rock Acknowledgements
affected by fault-related fracturing (Billi et al., 2003). The fact that,
according to the geological map of the study area, the length of San We would like to thank Qatar Petroleum, Shell and the Qatar
Carlos Fault is significantly larger or has a higher strain magnitude than Science and Technology Park for funding fieldwork in the Picos de
the Middle Fault and the Northern Fault, suggests that the San Carlos Europa, Spain. The first author is also very grateful to Indonesia
Fault may have generated an associated fracture network within its Endowment Fund for Education (LPDP) for their financial support
damage zone with a higher fracture density, and a larger aperture of during his study at Imperial College London. We thank Fergus Kennedy

732
M.Y.A. Madjid et al. Marine and Petroleum Geology 91 (2018) 723–734

and Manuel Antonio Diaz Rodriguez for conducting the drone flights investigation of karst guided by comparison with outcrop and unmanned aerial ve-
and help with image processing. The director of the National Park of hicle imagery. J. Appl. Geophys. 112, 268–278.
Flynn, K.F., Chapra, S.C., 2014. Remote sensing of submerged aquatic vegetation in a
Picos de Europa is acknowledged for permission to conduct this re- shallow non-turbid river using an unmanned aerial vehicle. Rem. Sens. 6,
search. C Jordan publishes with permission of the Executive Director of 12815–12836.
the British Geological Survey. We are grateful to the reviewers, Fontana, S., Nader, F.H., Morad, S., Ceriani, A., Al-Aasm, I.S., 2010. Diagenesis of the
khuff formation (Permian-Triassic), northern United Arab Emirates. Arabian J. Geosci
Aurelien Pierre and an anonymous reviewer, for their positive and 3, 351–368.
constructive comments, and to Alex MacNeil for editorial efforts and Garcia, E.M., Marquinez, J., Heredia, N., Navaro, D., Trodriguez-Fernandez, L.R., 1977.
ensuring an effective review process. Mapa Geológico de la Región Careña-Cabrales, scale 1:50,000. Instituto Geológico y
Minero de España.
Gasparrini, M., Bechstadt, T., Boni, M., 2006. Massive hydrothermal dolomites in the
References southwestern Cantabrian Zone (Spain) and their relation to the Late Variscan evo-
lution. Mar. Petrol. Geol. 23, 543–568.
Gomez-Fernandez, F., Both, R.A., Mangas, J., Arribas, A., 2000. Metallogenesis of Zn-Pb
Anderson, K., Gaston, K.J., 2013. Lightweight unmanned aerial vehicles will revolutionize
carbonate-hosted mineralization in the southeastern region of the Picos de Europa
spatial ecology. Front. Ecol. Environ. 11, 138–146.
(central northern Spain) province: geologic, fluid inclusion, and stable isotope stu-
Arndt, M., Virgo, S., Cox, S.F., Urai, J.L., 2014. Changes in fluid pathways in a calcite vein
dies. Econ. Geol. Bull. Soc. Econ. Geol. 95, 19–39.
mesh (Natih Formation, Oman Mountains): insights from stable isotopes. Geofluids
Hird, J.N., Montaghi, A., McDermid, G.J., Kariyeva, J., Moorman, B.J., Nielsen, S.E.,
14, 391–418.
McIntosh, A.C.S., 2017. Use of unmanned aerial vehicles for monitoring recovery of
Bahamonde, J.R., Vera, C., Colmenero, J.R., 2000. A steep-fronted Carboniferous car-
forest vegetation on petroleum well sites. Rem. Sens. 9.
bonate platform: clinoformal geometry and lithofacies (Picos de Europa, NW Spain).
Hodgetts, D., 2013. Laser scanning and digital outcrop geology in the petroleum industry:
Sedimentology 47, 645–664.
a review. Mar. Petrol. Geol. 46, 335–354.
Bandini, F., Jakobsen, J., Olesen, D., Reyna-Gutierrez, J.A., Bauer-Gottwein, P., 2017.
Homewood, P., Razin, P., Grelaud, C., Droste, H., Vahrenkamp, V., Mettraux, M., Mattner,
Measuring water level in rivers and lakes from lightweight Unmanned Aerial
J., 2008. Outcrop sedimentology of the Natih Formation, northern Oman: a field
Vehicles. J. Hydrol. 548, 237–250.
guide to selected outcrops in the adam foothills and Al Jabal al Akhdar areas.
Beckert, J., Vandeginste, V., John, C.M., 2015. Exploring the geological features and
GeoArabia 13, 39–120.
processes that control the shape and internal fabrics of late diagenetic dolomite
Jacquemyn, C., Huysmans, M., Hunt, D., Casini, G., Swennen, R., 2015. Multi-scale three-
bodies (Lower Khuff equivalent - Central Oman Mountains). Mar. Petrol. Geol. 68,
dimensional distribution of fracture- and igneous intrusion-controlled hydrothermal
325–340.
dolomite from digital outcrop model, Latemar platform, Dolomites, northern Italy.
Beckert, J., Vandeginste, V., McKean, T.J., Alroichidi, A., John, C.M., 2018. Ground-based
AAPG Bull. 99, 957–984.
hyperspectral imaging as a tool to identify different carbonate phases in natural cliffs.
Jakob, S., Zimmermann, R., Gloaguen, R., 2017. The need for accurate geometric and
Int. J. Rem. Sens (in press).
radiometric corrections of drone-borne hyperspectral data for mineral exploration:
Bemis, S.P., Micklethwaite, S., Turner, D., James, M.R., Akciz, S., Thiele, S.T., Bangash,
MEPHySToA toolbox for pre-processing drone-borne hyperspectral data. Rem.
H.A., 2014. Ground-based and UAV-Based photogrammetry: a multi-scale, high-re-
Sens. 9.
solution mapping tool for structural geology and paleoseismology. J. Struct. Geol. 69,
Jones, T.G., Coops, N.C., Sharma, T., 2010. Assessing the utility of airborne hyperspectral
163–178.
and LiDAR data for species distribution mapping in the coastal Pacific Northwest,
Bhardwaj, A., Sam, L., Akanksha, Martin-Torres, F.J., Kumar, R., 2016. UAVs as remote
Canada. Remote Sens. Environ. 114, 2841–2852.
sensing platform in glaciology: present applications and future prospects. Remote
Jordan, C.J., Napier, B., 2016. Developing digital fieldwork technologies at the british
Sens. Environ. 175, 196–204.
geological survey. In: Bowman, M., Smyth, H.R., Good, T.R., Passey, S.R., Hirst,
Billi, A., Storti, F., Salvini, F., 2003. Particle size distributions of fault rocks and fault
J.P.P., Jordan, C.J. (Eds.), Value of Outcrop Studies in Reducing Subsurface
transpression: are they related? Terra. Nova 15, 61–66.
Uncertainty and Risk in Hydrocarbon Exploration and Production, pp. 219–229.
Bistacchi, A., Balsamo, F., Storti, F., Mozafari, M., Swennen, R., Solum, J., Tueckmantel,
Kerans, C., Zahm, C., Garcia-Fresca, B., Harris, P., 2017. Guadalupe Mountains, west
C., Taberner, C., 2015. Photogrammetric digital outcrop reconstruction, visualization
Texas and New Mexico: key excursions. AAPG Bull. 101, 465–474.
with textured surfaces, and three-dimensional structural analysis and modeling: in-
Kurz, T.H., Dewit, J., Buckley, S.J., Thurmond, J.B., Hunt, D.W., Swennen, R., 2012.
novative methodologies applied to fault-related dolomitization (Vajont Limestone,
Hyperspectral image analysis of different carbonate lithologies (limestone, karst and
Southern Alps, Italy). Geosphere 11, 2031–2048.
hydrothermal dolomites): the Pozalagua Quarry case study (Cantabria, North-west
Bodin, S., Mattioli, E., Froehlich, S., Marshall, J.D., Boutib, L., Lahsini, S., Redfern, J.,
Spain). Sedimentology 59, 623–645.
2010. Toarcian carbon isotope shifts and nutrient changes from the Northern margin
Lapponi, F., Bechstadt, T., Boni, M., Banks, D.A., Schneider, J., 2014. Hydrothermal
of Gondwana (High Atlas, Morocco, Jurassic): palaeoenvironmental implications.
dolomitization in a complex geodynamic setting (Lower Palaeozoic, northern Spain).
Palaeogeogr. Palaeoclimatol. Palaeoecol. 297, 377–390.
Sedimentology 61, 411–443.
Budd, D.A., Frost III, E.L., Huntington, K.W., Allwardt, P.F., 2013. Syndepositional de-
Lapponi, F., Casini, G., Sharp, I., Blendinger, W., Fernandez, N., Romaire, I., Hunt, D.,
formation features in high-relief carbonate platforms: long-lived conduits for diage-
2011. From outcrop to 3D modelling: a case study of a dolomitized carbonate re-
netic fluids. J. Sediment. Res. 83, 14–38.
servoir, Zagros Mountains, Iran. Petrol. Geosci. 17, 283–307.
Caine, J.S., Evans, J.P., Forster, C.B., 1996. Fault zone architecture and permeability
Masek, J.G., Hayes, D.J., Hughes, M.J., Healey, S.P., Turner, D.P., 2015. The role of re-
structure. Geology 24, 1025–1028.
mote sensing in process-scaling studies of managed forest ecosystems. For. Ecol.
Campana, S., 2017. Drones in archaeology. State-of-the-art and future perspectives.
Manag. 355, 109–123.
Archaeol. Prospect. 24, 275–296.
Mateos, R.M., Azanon, J.M., Roldan, F.J., Notti, D., Perez-Pena, V., Galve, J.P., Perez-
Caturegli, L., Corniglia, M., Gaetani, M., Grossi, N., Magni, S., Migliazzi, M., Angelini, L.,
Garcia, J.L., Colomo, C.M., Gomez-Lopez, J.M., Montserrat, O., Devantery, N., Lamas-
Mazzoncini, M., Silvestri, N., Fontanelli, M., Raffaelli, M., Peruzzi, A., Volterrani, M.,
Fernandez, F., Fernandez-Chacon, F., 2017. The combined use of PSInSAR and UAV
2016. Unmanned aerial vehicle to estimate nitrogen status of turfgrasses. PLoS
photogrammetry techniques for the analysis of the kinematics of a coastal landslide
One 11.
affecting an urban area (SE Spain). Landslides 14, 743–754.
Chesley, J.T., Leier, A.L., White, S., Torres, R., 2017. Using unmanned aerial vehicles and
Merino-Tome, O., Della Porta, G., Pierre, A., Kenter, J.A.M., Durlet, C., Verwer, K., 2017.
structure-from-motion photogrammetry to characterize sedimentary outcrops: an
Intact seismic-scale platforms and ramps in the Lower to Middle Jurassic of Morocco:
example from the Morrison Formation, Utah, USA. Sediment. Geol. 354, 1–8.
implications for stratal anatomy and lithofacies partitioning. AAPG Bull. 101,
Davies, G.R., Smith Jr., L.B., 2006. Structurally controlled hydrothermal dolomite re-
505–513.
servoir facies: an overview. AAPG (Am. Assoc. Pet. Geol.) Bull. 90, 1641–1690.
Merino-Tome, O.A., Bahamonde, J.R., Colmenero, J.R., Heredia, N., Villa, E., Farias, P.,
De Michele, C., Avanzi, F., Passoni, D., Barzaghi, R., Pinto, L., Dosso, P., Ghezzi, A.,
2009. Emplacement of the Cuera and Picos de Europa imbricate system at the core of
Gianatti, R., Della Vedova, G., 2016. Using a fixed-wing UAS to map snow depth
the Iberian-Armorican arc (Cantabrian zone, north Spain): New precisions concerning
distribution: an evaluation at peak accumulation. Cryosphere 10, 511–522.
the timing of arc closure. Geol. Soc. Am. Bull. 121, 729–751.
De Reu, J., Trachet, J., Laloo, P., De Clercq, W., 2016. From low cost UAV survey to high
Mlambo, R., Woodhouse, I.H., Gerard, F., Anderson, K., 2017. Structure from motion
resolution topographic data: developing our understanding of a medieval outport of
(SfM) photogrammetry with drone data: a low cost method for monitoring green-
bruges. Archaeol. Prospect. 23, 335–346.
house gas emissions from forests in developing countries. Forests 8.
Dewit, J., Huysmans, M., Muchez, P., Hunt, D.W., Thurmond, J.B., Verges, J., Saura, E.,
Nieminski, N.M., Graham, S.A., 2017. Modeling stratigraphic architecture using small
Fernandez, N., Romaire, I., Esestime, P., Swennen, R., 2012. Reservoir characteristics
unmanned aerial vehicles and photogrammetry: examples from the Miocene East
of fault-controlled hydrothermal dolomite bodies: ramales Platform case study. In:
Coast Basin, New Zealand. J. Sediment. Res. 87, 126–132.
Garland, J., Neilson, J.E., Laubach, S.E., Whidden, K.J. (Eds.), Advances in Carbonate
Nishar, A., Richards, S., Breen, D., Robertson, J., Breen, B., 2016. Thermal infrared
Exploration and Reservoir Analysis. Geological Society, London (Special Publications,
imaging of geothermal environments and by an unmanned aerial vehicle (UAV): a
London).
case study of the Wairakei - tauhara geothermal field, Taupo, New Zealand. Renew.
Ehrenberg, S.N., Nadeau, P.H., 2005. Sandstone vs. carbonate petroleum reservoirs: a
Energy 86, 1256–1264.
global perspective on porosity-depth and porosity-permeability relationships. AAPG
Ouedraogo, M.M., Degre, A., Debouche, C., Lisein, J., 2014. The evaluation of unmanned
Bull. 89, 435–445.
aerial system-based photogrammetry and terrestrial laser scanning to generate DEMs
Faye, E., Rebaudo, F., Yanez-Cajo, D., Cauvy-Fraunie, S., Dangles, O., 2016. A toolbox for
of agricultural watersheds. Geomorphology 214, 339–355.
studying thermal heterogeneity across spatial scales: from unmanned aerial vehicle
Padua, L., Vanko, J., Hruska, J., Adao, T., Sousa, J.J., Peres, E., Morais, R., 2017. UAS,
imagery tolandscape metrics. Meth. Ecol. Evol. 7, 437–446.
sensors, and data processing in agroforestry: a review towards practical applications.
Fernandes, A.L., Medeiros, W.E., Bezerra, F.H.R., Oliveira, J.G., Cazarin, C.L., 2015. GPR
Int. J. Rem. Sens. 38, 2349–2391.

733
M.Y.A. Madjid et al. Marine and Petroleum Geology 91 (2018) 723–734

Paneque-Galvez, J., McCall, M.K., Napoletano, B.M., Wich, S.A., Koh, L.P., 2014. Small Vandeginste, V., John, C.M., 2012. Influence of climate and dolomite composition on
drones for community-based forest monitoring: an assessment of their feasibility and dedolomitization: insights from a multi-proxy study in the central Oman Mountains.
potential in tropical areas. Forests 5, 1481–1507. J. Sediment. Res. 82, 177–195.
Puliti, S., Orka, H.O., Gobakken, T., Naesset, E., 2015. Inventory of small forest areas Vandeginste, V., John, C.M., Beckert, J., 2015a. Diagenetic geobodies: fracture-controlled
using an unmanned aerial system. Rem. Sens. 7, 9632–9654. burial dolomite in outcrops from northern Oman. SPE Reservoir Eval. Eng. 18, 84–93.
Putz-Perrier, M.W., Sanderson, D.J., 2008. Spatial distribution of brittle strain in layered Vandeginste, V., John, C.M., Cosgrove, J.W., Manning, C., 2014. Dimensions, texture-
sequences. J. Struct. Geol. 30, 50–64. distribution and geochemical heterogeneities of fracture-related dolomite geobodies
Racine, T., 2017. Characterising the dolostone host of an MVT Zn-Pb deposit in the Picos hosted in Ediacaran limestone, northern Oman. AAPG (Am. Assoc. Pet. Geol.) Bull.
de Europa mining district (NW Spain): optical mineralogy, geochemical and fluid 98, 1789–1809.
inclusion studies. Department of Earth Science and Engineering. Imperial College Vandeginste, V., John, C.M., van de Flierdt, T., Cosgrove, J.W., 2013. Linking process,
London, London, pp. 58. dimension, texture and geochemistry in dolomite geobodies: a case study from Wadi
Sanders, K.T., Masri, S.F., 2016. The energy-water agriculture nexus: the past, present and Mistal (northern Oman). AAPG (Am. Assoc. Pet. Geol.) Bull. 97, 1181–1207.
future of holistic resource management via remote sensing technologies. J. Clean. Vandeginste, V., Rafiuddin, N.L., Peskin, A., 2015b. Three-dimensional reconstruction of
Prod. 117, 73–88. diagenetic geobodies for geological carbon dioxide storage. Int. J. Greenh. Gas Contr.
Sankey, T., Donager, J., McVay, J., Sankey, J.B., 2017. UAV lidar and hyperspectral fu- 42, 319–328.
sion for forest monitoring in the southwestern USA. Remote Sens. Environ. 195, Vandeginste, V., Swennen, R., Gleeson, S.A., Ellam, R.M., Osadetz, K., Roure, F., 2005.
30–43. Zebra dolomitization as a result of focused fluid flow in the rocky Mountains fold and
Schrod, J., Weber, D., Drucke, J., Keleshis, C., Pikridas, M., Ebert, M., Cvetkovic, B., thrust belt, Canada. Sedimentology 52, 1067–1095.
Nickovic, S., Marinou, E., Baars, H., Ansmann, A., Vrekoussis, M., Mihalopoulos, N., Vasuki, Y., Holden, E.J., Kovesi, P., Micklethwaite, S., 2014. Semi-automatic mapping of
Sciare, J., Curtius, J., Bingemer, H.G., 2017. Ice nucleating particles over the Eastern geological Structures using UAV-based photogrammetric data: an image analysis
Mediterranean measured by unmanned aircraft systems. Atmos. Chem. Phys. 17, approach. Comput. Geosci. 69, 22–32.
4817–4835. Ventura, D., Bruno, M., Lasinio, G.J., Belluscio, A., Ardizzone, G., 2016. A low-cost drone
Seitz, A.R., Nanez, J.E., Holloway, S.R., Watanabe, T., 2006. Perceptual learning of mo- based application for identifying and mapping of coastal fish nursery grounds. Estuar.
tion leads to faster flicker perception. PLoS One 1. Coast Shelf Sci. 171, 85–98.
Serban, G., Rus, I., Vele, D., Bretcan, P., Alexe, M., Petrea, D., 2016. Flood-prone area Verhoeven, G.J., 2012. Near-infrared aerial crop mark archaeology: from its historical use
delimitation using UAV technology, in the areas hard-to-reach for classic aircrafts: to current digital implementations. J. Archaeol. Meth. Theor 19, 132–160.
case study in the north-east of Apuseni Mountains, Transylvania. Nat. Hazards 82, Verhoeven, G.J., Schmitt, K.D., 2010. An attempt to push back frontiers - digital near-
1817–1832. ultraviolet aerial archaeology. J. Archaeol. Sci. 37, 833–845.
Sharp, I., Gillespie, P., Morsalnezhad, D., Taberner, C., Karpuz, R., Verges, J., Horbury, A., Watts, A.C., Ambrosia, V.G., Hinkley, E.A., 2012. Unmanned aircraft systems in remote
Pickard, N., Garland, J., Hunt, D., 2010. Stratigraphic architecture and fracture- sensing and scientific research: classification and considerations of use. Rem. Sens. 4,
controlled dolomitization of the Cretaceous Khami and Bangestan groups: an outcrop 1671–1692.
case study, Zagros Mountains, Iran. In: Van Buchem, F.S.P., Gerdes, K.D., Esteban, M. Weil, A.B., Gutierrez-Alonso, G., Johnston, S.T., Pastor-Galan, D., 2013. Kinematic con-
(Eds.), Mesozoic and Cenozoic Carbonate Systems of the Mediterranean and the straints on buckling a lithospheric-scale orocline along the northern margin of
Middle East. Stratigraphic and Diagenetic Reference Models, pp. 343–396. Gondwana: a geologic synthesis. Tectonophysics 582, 25–49.
Snavely, N., Seitz, S.M., Szeliski, R., 2008. Modeling the world from Internet photo col- Westoby, M.J., Brasington, J., Glasser, N.F., Hambrey, M.J., Reynolds, J.M., 2012.
lections. Int. J. Comput. Vis. 80, 189–210. 'Structure-from-Motion' photogrammetry: a low-cost, effective tool for geoscience
Spence, C., Mengistu, S., 2016. Deployment of an unmanned aerial system to assist in applications. Geomorphology 179, 300–314.
mapping an intermittent stream. Hydrol. Process. 30, 493–500. Westoby, M.J., Dunning, S.A., Woodward, J., Hein, A.S., Marrero, S.M., Winter, K.,
Tamminga, A.D., Eaton, B.C., Hugenholtz, C.H., 2015. UAS-based remote sensing of flu- Sugden, D.E., 2015. Sedimentological characterization of Antarctic moraines using
vial change following an extreme flood event. Earth Surf. Process. Landforms 40, UAVs and Structure-from-Motion photogrammetry. J. Glaciol. 61, 1088–1102.
1464–1476. Wilkinson, M.W., Jones, R.R., Woods, C.E., Gilment, S.R., McCaffrey, K.J.W., Kokkalas, S.,
Tang, L.N., Shao, G.F., 2015. Drone remote sensing for forestry research and practices. J. Long, J.J., 2016. A comparison of terrestrial laser scanning and structure-from-mo-
For. Res. 26, 791–797. tion photogrammetry as methods for digital outcrop acquisition. Geosphere 12,
Tong, X.H., Liu, X.F., Chen, P., Liu, S.J., Luan, K.F., Li, L.Y., Liu, S., Liu, X.L., Xie, H., Jin, 1865–1880.
Y.M., Hong, Z.H., 2015. Integration of UAV-based photogrammetry and terrestrial Wilson, M.E.J., Evans, M.J., Oxtoby, N.H., Nas, D.S., Donnelly, T., Thirlwall, M., 2007.
laser scanning for the three-dimensional mapping and monitoring of open-pit mine Reservoir quality, textural evolution and origin of fault-associated dolomites. AAPG
areas. Rem. Sens. 7, 6635–6662. (Am. Assoc. Pet. Geol.) Bull. 91, 1247–1272.
Tsoflias, G.P., Becker, M.W., 2008. Ground-penetrating-radar response to fracture-fluid Yang, C.H., Westbrook, J.K., Suh, C.P.C., Martin, D.E., Hoffmann, W.C., Lan, Y.B., Fritz,
salinity: why lower frequencies are favorable for resolving salinity changes. B.K., Goolsby, J.A., 2014. An airborne multispectral imaging system based on two
Geophysics 73, J25–J30. consumer-grade cameras for agricultural remote sensing. Rem. Sens. 6, 5257–5278.

734