Ružić, Ilic and Benac 1
MEASURING POCKET GRAVEL BEACH CHANGES WITH A DIGITAL
CAMERA
Igor Ružić1, Suzana Ilic2 and Čedomir Benac3,
Abstract: This paper examines capabilities of the Structure-from-Motion
(SfM) photogrammetry for monitoring of changes on pocket gravel beaches.
A single digital camera was used to get images from a beach on the Croatian
coast. This provided a cheap and quick image requisition. Images were
processed by an online available SfM photogrammetry software. 3D point
clouds derived from images were of 5 cm accuracy. Georeferencing of 3D
point clouds using permanent ground control points (GCPs) enabled their
comparison and assessment of beach changes. Number of parameters, which
are relevant for examining beach geomorphology and which can potentially
be used for coastal management are derived from the point clouds. Moreover
a series of measurements taken on before and after storms can be useful for
assessing stability and sustainability of these coastal systems.
INTRODUCTION
Most beaches on the Croatian coast are small pocket gravel beaches formed by the
deposition of sediments by torrential flows. These are pebble and gravel beaches usually
not longer than 300 m (Juračić et al. 2009, Benac et al. 2013). Gravel pocket beaches
change fast due to their short response time to energetic events. Changes during these
events ranging in several decimeter to meters in vertical elevation to several meters in
the shoreline position. Past studies (e.g. Carter and Orford 1993; Forbes et al. 1997;
Shulmeister and Kirk 1997; Jennings et al. 1998) focused mainly on long-term
morphological changes and not much is known how these beaches response to changes
in driving conditions on shorter scale , such as waves, tides and storm surges. Beach
changes that take place over temporal scales of days (e.g. storms) up to seasons are also
important for coastal management. However measurements on gravel beaches are
extremely difficult due to movement of sediment of large sizes and shear forces that
1 University of Rijeka, Faculty of Civil Engineering, Radmile Matejčić 3, HR-51000 Rijeka, Croatia.
2 University of Lancaster, Lancaster Environment Centre, LA1 4YQ Lancaster, UK. [email protected]
Email address
3 University of Rijeka, Faculty of Civil Engineering, Radmile Matejčić 3, HR-51000 Rijeka, Croatia.
Ružić, Ilic and Benac 2
these beaches experience. Also rare measurements in single points are not able to
capture changes across and along the beach areas. Advances in new technologies, such
as RTK-GPS surveying and in particular remote sensing provide spatial and temporal
resolutions for capturing short-term changes. These provided more in depth information
about morphological changes on these beaches and mechanisms behind them (e.g.
Curtiss et al. 2009, Ruiz de Alegria-Arzaburu and Masselink 2010, Poate et al. 2013).
Remote sensing enables fast and simple collection of high-accuracy three-dimensional
(3D) spatial data. Remote sensing technologies, which are applied or have a potential to
be applied for studies of gravel beaches are laser scanning (airborne and terrestrial),
photogrammetry, soft-copy photogrammetry, video imaging and Structure-from-Motion
(SfM) photogrammetry.
The latter is becoming increasingly used in geomorphological studies (Westoby et al.
2012, James et al. 2013, Ružić et al. 2013, Fonstad et al. 2013). The SfM
photogrammetry creates 3D point clouds from photographs, taken from a different
location with a single consumer-grade camera. The method works by matching image
texture in different photographs, and determining the 3D geometry under the
assumption that the scheme is static while the photographs were taken. This single
camera based approach has been found to have an advantage over laser scanning in
rapidly changing or spatially limited areas, due to low cost and fast recording (James et
al. 2013; Ružić et al. 2014). It provides fast collection of high-accuracy and high-density
3D point clouds.
Here we examine capabilities of the method in monitoring and assessing changes of a
gravel pocket beach. An ultimate aim is to develop a monitoring approach, which can
be easily applied and used by coastal managers on day-to-day basis.
STUDY SITE AND METHODOLOGY
The study was carried on a small pocket beach in Brseč, located at the western edge
of the Kvarner Bay in the North Adriatic, Croatia (Fig. 1.). The beach is located in a
small cove and surrounded by high limestone cliffs. The length of the Brseč beach is 37
meters, and the width of the beach varies from 0 to 20 meters, depending on
environmental conditions. Beach sediment size is in a range from 2 to 45 mm. Although
rather small, this beach attracts tourists to the village of Brsece and hence makes an
important part of the local economy. Beach is also interesting from geomorphological
perspective as it was completely washed out and recovered back to its original state on
several occasions.
Ružić, Ilic and Benac 3
Fig. 1. Investigation area map
Nine successive surveys were carried out between 4 October and 27 December 2013
(Table 1). During these surveys significant beach morphological changes were recorded
as a consequence of high energetic events and changing wind directions.
Table 1. Conducted survey dates and numbers
Survey
number Date
1 04.10.2013
2 30.10.2013
3 06.11.2013
4 10.11.2013
5 12.11.2013
6 28.11.2013
7 04.12.2013
8 24.12.2013
9 27.12.2013
Images were collected using a single 10-megapixel Ricoh GR Digital IV camera. A
special care was taken to acquire sharp and well-focused images, as blurred and out of
focus images are not suitable for the SfM reconstruction. Surrounding cliffs enabled
photographs to be taken from a wide range of horizontal and vertical angles, which is
beneficial for creating 3D point clouds. The 3D point clouds were generated from
approximately 250 images using Autodesk ReCap online service
Ružić, Ilic and Benac 4
(https://recap360.autodesk.com/). Cloud Compare software
(http://www.danielgm.net/cc/) was used for the 3D point cloud geo-referencing. Images
were georeferenced to Croatian ordinate system (HTRS 96) by using four ground
control points (GCP), which were clearly marked on the rocks with a fluorescent paint
(Fig. 2.) and which coordinates were obtained from surveys with the Real Time
Kinematic GPS (RTK-GPS).
Fig. 1. Left: Ground control point (GCP) layout; Right: GCP marking
Comparison of 3D coordinates from the point cloud and RTK-GPS surveys at 24
randomly chosen ‘verification points’ gave the root mean square error (RMSE) of the
vertical offset of 4.0cm when all points were taken into account. The RMSE was 3.2cm
for points on the beach and 4.9cm for the points on the rocks (Fig. 3.). Only at one point
on the beach (Fig. 3.), vertical offset was 9 cm, which may be a consequence of the poor
GPS signal below the cliff. Figure 3 shows locations of GCPs and verification points. A
verification criterion was set taking into account the RTK-GPS vertical precision of 5
cm. The vertical offset lower than 5cm was considered as good and is marked in green
in figure 3. Those points with the vertical offset higher than 10 cm are considered as
poor (red points) while those points between 5 and 10 cm as reasonably good (yellow
points). The accuracy of the 3D point-cloud is comparable to the accuracy of the RTK-
GPS surveys.
Ružić, Ilic and Benac 5
Fig. 3: GCP and verification points layout; distribution
RESULTS AND DISCUSSION
Figure 4 shows beach shore line changes derived from the images at three different
dates, showing beach response to three typical environmental conditions; calm
conditions and storm conditions with waves generated by SE winds and waves
generated by NE winds.
Fig. 4. Brseč (Klančac) beach rotation are shown on photo-realistic 3D point
cloud; A – after a calm period (4.10.2013); B – after NE storm (12.11.2013.); C
– after SE storm (27.12.2013.); D – Measured beach shorelines shown in A, B
Ružić, Ilic and Benac 6
and C; (In all plots red contour stands for a waterline during calm period,
green contour stands for waterline after NE storm and blue contour stands for
waterline after SE storm)
While this is a micro-tidal beach with low tidal elevation, there is still difference in
sea surface elevation and wave set-up during images acquisition. Hence a contour
elevation of 0.4 m was taken as a reference waterline to describe the shoreline changes.
Fig. 5. Brseč gravel pocket beach recorded cross-shore (A-C) and long-shore
(D-F) profile changes between 4th October and 27th December.
Approximately average beach shoreline location is marked with red color (Fig. 4.A.)
and it stretches uniformly along the beach. Green line (Fig. 4.B.) marks the shoreline
Ružić, Ilic and Benac 7
resulting from waves generated by strong NE wind. This line retreats in the north-
eastern corner of the beach, resulting in slightly rotated shoreline. Red line (Fig. 4.C.) is
affiliated to the shoreline resulting from waves generated by SE wind, which advanced
in the north-eastern corner of the beach and retreats in the south-western corner of the
beach resulting in the opposite rotation. Figure 4.D. shows all three shoreline together.
The shoreline during calm period is approximately an average shoreline. The shoreline
rotation was up to 32° between measurements taken after SE and NE storm. A resulting
shift of shorelines was 7.5m on the north-eastern part and about 8m on the south-
western part.
Cross-shore (A, B, C) and longshore profiles (D, E, F) derived from 3D point clouds are
presented in figure 5. The top panel in figure 5 shows the location of these profiles.
Changes in beach elevation at cross-shore profile A (Fig. 5.A.) were greater than 1m.
The beach extension (width) varied from 0 (27.12.2013.) to 8 meters (12.11.2013).
Changes in beach elevation at cross-shore profile C (Fig. 5.) were greater than 1.2 m.
Beach width varied from 5 m after strong NE wind (04.12.2013.) up to 14 meters after
moderate SE wind (6.11.2013). The least changes were recorded at profile B in the
middle (Fig. 5.). Profile B is in the middle of the beach where a beach pivot point is.
Beach width varied here up to 4 meters and beach height varied up to 0.60 m.
Beach elevation changes are also visible in longshore profile D (Fig. 5.), which were up
to 1 meter. The largest recorded elevation was after moderate SE winds (6.11.2013).
Erosion was recorded after strong NE wind (12.11.2013.).
Beach elevations at long-shore profile E (Figure 5) were ranging from total erosion
(27.12.2013) to 1 m (12.11.2013.) at south-eastern part of the beach. Elevations on the
opposite side of the beach were ranging from 0.40 (4.12.2013) to 1.5 meters
(06.11.2013). The largest changes which were greater than 1 m, were recorded at long-
shore profile F, near the shore line (Fig. 5.).
These surveys revealed that beach erosion occurs quickly. Storms from both wind
dominant directions (SE and NE) have large effect on beach stability. Such behavior
was observed on gravel beaches (Sherman 1991, Ivamy and Kenc 2006, Ruiz de
Alegria-Arzaburu and Masselink 2010) and pocket sandy beaches (Dail et al. 2000;
Dehouck et al. 2009), but this has been proofed for the first time on gravel pocket
beaches. Changes in beach elevations and in particular in beach orientation were
pronounced. Change in shoreline orientation was up to 30° degrees (Fig. 2.). Focus of
further study will be to relate these observed changes to hydrodynamic and sediment
beach processes.
The SfM photogrammetry approach has proved to be a useful method for monitoring
of gravel pocket beaches. It reduces data acquisition significantly. For this particular
site, image acquisition took about 15 minutes, once GCPs were set-up permanently. The
accuracy of the 3D point-cloud spatial data were verified with measurements from the
RTK-GPS survey. Recorded beach height changes are more than ten times greater than
the assessed accuracy. However, further improvement are possible if GCPs surveys are
derived more precisely (Harwin and Lucieer 2012). However, there are limitations as
only changes at emerged part of the beach can be recorded. When assessing volume
changes after the storms, it would be useful to have measurements taken in nearshore
Ružić, Ilic and Benac 8
where sediment is washed to. Despite this the method has a large potential for gravel
beach monitoring application.
CONCLUSION
This paper demonstrated a successful application of the Structure-from-Motion (SfM)
photogrammetry for studying and monitoring gravel pocket beaches. The SfM
photogrammetry provides quick and cheap measurement of high accuracy 3D point
clouds, which can be used to study beach changes. Shoreline changes and changes in
cross-shore and longshore beach profiles, which were derived from georeferenced 3D
point clouds showed rapid changes taking place on a pocket gravel beach due to
changes in wind conditions from SE to NE. It also highlighted a need for permanent
monitoring to assess the beach changes properly.
ACKNOWLEDGEMENTS
This paper is a result of a research activity under the project “Geological hazard in the
Kvarner area” of the University of Rijeka.
REFERENCES
Carter, R.W.G., Orford, J.D. (1993). “The morphodynamics of coarse clastic beaches
and barriers: a short term and loA3:A23ng term perspective” J. Coast. Res. SI 15,
158–179.
Curtiss, G. M., Osborne, P. D., and Horner-Devine, A. R. (2009). Seasonal patterns of
coarse sediment transport on a mixed sand and gravel beach due to vessel wakes,
wind waves, and tidal currents. Marine Geology, 259, 1-4, 73–85.
Benac, Č., Juračić, M., Matičec, D., Ružić, I., and Pikelj, K. (2013). “Fluviokarst and
classical karst: examples from the Dinarics (Krk Island, Northern Adriatic, Croatia)”.
Geomorphology 184:64–73
Dail, H.J., Merrifield, M. and Bevis, M. (2000). “Steep beach morphology changes due
to energetic wave forcing.” Marine Geology, 162(2-4), 443–458.
Dehouck, A., Dupuis, H., and Sénéchal, N. (2009). “Pocket beach dynamics: the
example of four macrotidal beaches, Brittany, France.” Marine Geology 266, 1–17.
Fonstad, M.A., Dietrich, J.T. , Courville, B.C., Jensen J.L., and Carbonneau P.E. (2013)
“Topographic structure from motion: a new development in photogrammetric
measurement.” Earth Surf Process Landforms 38(4):421–430
Forbes, D.L., Taylor, R.B., Orford, J.D., Carter, R.W.G., and Shaw, J. (1997). Gravel-
barrier migration and overstepping. Mar. Geol. 97, 305–313.
Harwin, S. and Lucieer, A. (2012). “Assessing the Accuracy of Georeferenced Point
Clouds Produced via Multi-View Stereopsis from Unmanned Aerial Vehicle, UAV
Imagery.” Remote Sensing, 4, 12, 1573–1599.
Ivamy, M.C., and Kench, P.S. (2006). “Hydrodynamics and morphological adjustment
of a mixed sand and gravel beach, Torere, Bay of Plenty, New Zealand.” Marine
Geology, 228, 1-4, 137–152.
Ružić, Ilic and Benac 9
James, M. R., and Robson, S. (2012). “Straightforward reconstruction of 3D surfaces
and topography with a camera: Accuracy and geoscience application.” Journal of
Geophysical Research, 117, F3, F03017
James, M.R., Ilić, S., and Ružić, I. (2013). “Measuring 3D coastal change with a digital
camera.” Coastal Dynamics 2013. 7th International Conference on Coastal
Dynamics, Arcachon, 893-904.
Jennings, S., Orford, J.D., Canti, M., Devoy, R.J.N., and Straker, V. (1998). “The role
of relative sea level rise and changing sediment supply on Holocene gravel barrier
development: the example of Porlock, Somerset, UK.” Holocene 8, 165-181.
Juračić, M., Benac, Č., Pikelj, K., and Ilic, S. (2009). “Comparison of the vulnerability
of lime-stone (karst) and siliciclastic coasts (example from the Kvarner area, NE
Adriatic, Croatia).” Geomorphology, 107: 90–99.
Poate, T., Masselink, G., Davidson, M., McCall, R., Russell, P., and Turner, I. (2013).
“High frequency in-situ field measurements of morphological response on a fine
gravel beach during energetic wave conditions”. Marine Geology, 342, 1–13.
Ruiz de Alegria-Arzaburu, A., and Masselink, G. (2010). “Storm response and beach
rotation on a gravel beach, Slapton Sands, U.K.” Marine Geology, 278, 1-4, 77–99.
Ružić, I., Marović, I., Vivoda, M., Dugonjić Jonjčević, S., Kalajžić, D., Benac, Č., and
Ožanić, N. (2013). “Application of Structure-from-Motion photogrammetry for
erosion processes monitoring, Moscenicka Draga example.” Proc. of the 4th
Workshop of the Japanese-Croatian Project on “Risk Identification and Land-Use
Planning for Disaster Mitigation of Landslides and Floods in Croatia, University of
Split, Split, pp. 49-50.
Ružić. I., Marović, I., Benac, Č., and Ilić, S. (2014). “Coastal cliff geometry derived
from Structure-from-Motion photogrammetry at Stara Baška, Krk Island, Croatia.”
Geo-Marine Letters, 34: 555-565.
Sherman, D.J. (1991). “Gravel beaches”. Nat. Geogr. Res. 7 (4), 442–452.
Shulmeister, J., and Kirk, R.M. (1997). “Holocene fuvial-coastal in- teractions on a
mixed sand and sand and gravel beach sys- tem, North Canterbury, New Zealand.2
Catena 30, 337-355.
Westoby, M. J., Brasington, J., Glasser, N. F., Hambrey, M. J., and Reynolds, J. M.
(2012). “Structure-from-Motion photogrammetry: A low-cost, effective tool for
geoscience applications.” Geomorphology, 179, 300–314.
Recommended