**3. Case studies: shoreline changes on the coasts of northern France**

The second part of the chapter presents examples of studies carried out on the northern coast of France in the surrounding of Dunkirk. Topographic surveys were carried out using differential GNNS, LiDAR, and airborne photogrammetry. These examples illustrate the benefits of changing from two- to three-dimensional analysis and improving spatial resolution.

#### **3.1 Toward a more accurate coastline detection**

The shoreline is the indicator most often used to define, map the position of the shore, and study its evolution [48]. There are several definitions of shoreline [8]. The definition may vary according to the coastal environments studied, for example, the boundary between water and sand in a microtidal environment, the base of the top of a cliff, the berm of a pebble barrier, the toe of coastal dunes, and the boundary between sand and beach top vegetation. Along the same coast, the position of these different shoreline indicators does not always coincide: for example, the limit between sand and vegetation may be located several meters away landward of the dune toe (**Figure 7**). The definition of the coastline is therefore an essential prerequisite in diachronic analyses of the shoreline.

Identifying the shoreline in two dimensions on aerial photographs is sometimes difficult. On aerial photographs, whether historical or recent, the coastline is not always clearly distinguishable. **Figure 7** shows two aerial photographs shot at two different dates (1957 and 2015), at the same scale. On the 1957 photo of poor quality, the break in slope is hardly detectable where the foredune slope is gentle. In GIS analyses, an error margin in detection must be taken into account (±*x* pixels,

**101**

**Figure 7.**

*Recent Advances in Coastal Survey Techniques: From GNSS to LiDAR and Digital…*

thus ±*y* m, depending on the spatial resolution of the digitized photos) [49]. Stereoscopic observation of paper photos helps to identify the shoreline position, but the transfer of these observations to a computer is not convenient. Computer and photogrammetric processing of the images improves analysis (see below). The shoreline can be measured in the field by GNSS. This requires a clear definition of what is considered as a shoreline and a good experience in field investigation. In the following examples along the coasts of northern France, the dune toe was selected. The shoreline is easily detected when the dune toe is marked by a sharp change in slope gradient (e.g., in Zuydcoote in **Figure 7**), but not when the slope between the upper beach and the dune is mild and regular (e.g., in Bray-Dunes). The measurement accuracy of GNSS is high, but in these cases, an error

*The complex detection of the shoreline on aerial photographs and in the field.*

margin has to be attributed to the operator's interpretation of the terrain.

The shoreline can be defined by an altitude level, for example, the one reached by a distinct tide level such as the mean high spring tide. This approach is used by the French National Hydrographic and Oceanographic Service (SHOM) to define the official coastline in France, called "Histolitt," which corresponds to the highest astronomical tide (HAT). Along the coastline east of Dunkirk, for example, the HAT corresponds to an altitude of 3.787 m (French elevation datum), which has been mapped on a recent DSM obtained from an aerial photogrammetric survey with a high spatial resolution (5 cm/pixel) carried out in September 2019. The "Histolitt" shoreline has also been superimposed on the 2019 DSM showing an offset of about 11 m between the two shorelines (**Figure 8**) even if the shoreline position was defined by the same altitudinal level (3.787 m) in both cases. This difference in position between the two shorelines can be explained by the fact that the "Histolitt" shoreline was determined from a DTM with a vertical accuracy of

*DOI: http://dx.doi.org/10.5772/intechopen.91964*

*Recent Advances in Coastal Survey Techniques: From GNSS to LiDAR and Digital… DOI: http://dx.doi.org/10.5772/intechopen.91964*

**Figure 7.** *The complex detection of the shoreline on aerial photographs and in the field.*

thus ±*y* m, depending on the spatial resolution of the digitized photos) [49]. Stereoscopic observation of paper photos helps to identify the shoreline position, but the transfer of these observations to a computer is not convenient. Computer and photogrammetric processing of the images improves analysis (see below).

The shoreline can be measured in the field by GNSS. This requires a clear definition of what is considered as a shoreline and a good experience in field investigation. In the following examples along the coasts of northern France, the dune toe was selected. The shoreline is easily detected when the dune toe is marked by a sharp change in slope gradient (e.g., in Zuydcoote in **Figure 7**), but not when the slope between the upper beach and the dune is mild and regular (e.g., in Bray-Dunes). The measurement accuracy of GNSS is high, but in these cases, an error margin has to be attributed to the operator's interpretation of the terrain.

The shoreline can be defined by an altitude level, for example, the one reached by a distinct tide level such as the mean high spring tide. This approach is used by the French National Hydrographic and Oceanographic Service (SHOM) to define the official coastline in France, called "Histolitt," which corresponds to the highest astronomical tide (HAT). Along the coastline east of Dunkirk, for example, the HAT corresponds to an altitude of 3.787 m (French elevation datum), which has been mapped on a recent DSM obtained from an aerial photogrammetric survey with a high spatial resolution (5 cm/pixel) carried out in September 2019. The "Histolitt" shoreline has also been superimposed on the 2019 DSM showing an offset of about 11 m between the two shorelines (**Figure 8**) even if the shoreline position was defined by the same altitudinal level (3.787 m) in both cases. This difference in position between the two shorelines can be explained by the fact that the "Histolitt" shoreline was determined from a DTM with a vertical accuracy of

*Spatial Variability in Environmental Science - Patterns, Processes, and Analyses*

possible to highlight otherwise undetectable detail changes at lower acquisition frequency and/or with coarser spatial resolution surveys (e.g., aeolian sand accu-

**3. Case studies: shoreline changes on the coasts of northern France**

The second part of the chapter presents examples of studies carried out on the northern coast of France in the surrounding of Dunkirk. Topographic surveys were carried out using differential GNNS, LiDAR, and airborne photogrammetry. These examples illustrate the benefits of changing from two- to three-dimensional

The shoreline is the indicator most often used to define, map the position of the shore, and study its evolution [48]. There are several definitions of shoreline [8]. The definition may vary according to the coastal environments studied, for example, the boundary between water and sand in a microtidal environment, the base of the top of a cliff, the berm of a pebble barrier, the toe of coastal dunes, and the boundary between sand and beach top vegetation. Along the same coast, the position of these different shoreline indicators does not always coincide: for example, the limit between sand and vegetation may be located several meters away landward of the dune toe (**Figure 7**). The definition of the coastline is therefore an essential

Identifying the shoreline in two dimensions on aerial photographs is sometimes difficult. On aerial photographs, whether historical or recent, the coastline is not always clearly distinguishable. **Figure 7** shows two aerial photographs shot at two different dates (1957 and 2015), at the same scale. On the 1957 photo of poor quality, the break in slope is hardly detectable where the foredune slope is gentle. In GIS analyses, an error margin in detection must be taken into account (±*x* pixels,

mulations, see Section 2 of this chapter).

*Spatial and temporal scales of application of the three techniques.*

**Figure 6.**

analysis and improving spatial resolution.

**3.1 Toward a more accurate coastline detection**

prerequisite in diachronic analyses of the shoreline.

**100**

**Figure 8.** *Example of shoreline detection on a DTM/DSM and comparison with the shoreline measured by GNSS in the field.*

50 cm calculated by photogrammetry with aerial photographs of 2004. The recent fine-scale mapping of the upper beach and coastal dunes shows that the "Histolitt" shoreline is now outdated since its position does not correspond to the present day morphology anymore. Nevertheless, even if the updated shoreline position using the location of the HAT level on the recent DSM is more consistent with the actual present day shoreline, its position still lays several meters seaward of the dune toe (**Figure 8**) that is the geomorphological expression of the upper limit of action of the highest water levels, including tides and meteorological surges related to wind and changes in atmospheric pressure [38].

The shoreline may also be mapped in a semi-automatic way by identifying in a DTM for a neat variation of the slope gradient at the dune toe [38] using the following approach:

$$\left\| \begin{vmatrix} \dot{\mathbf{g}} \\ \ddot{\mathbf{g}} \end{vmatrix} \right\| = \sqrt{\left(\frac{\partial \mathbf{z}}{\partial \mathbf{x}}\right)^2 + \left(\frac{\partial \mathbf{z}}{\partial \mathbf{y}}\right)^2} \tag{2}$$

**103**

**Figure 9.**

*(adapted from Ref. [38]).*

*Recent Advances in Coastal Survey Techniques: From GNSS to LiDAR and Digital…*

(e.g., paths). Zones with steeper slopes are not displayed on the map; therefore, the actual topography remains visible. The detection of this boundary is validated by comparison with the coastline surveyed in October 2019 with a GNNS: the differences range from 0 to 2.5 m landward, which are due to beach and dune toe erosion between September and October 2019, as a result of high water levels that had

**Figure 9** illustrates a diachronic study using shorelines extracted from DTMs at different dates (2008, 2011, 2012, and 2014) on an 8 km long stretch of coast between Dunkirk and the Belgian border. The analysis of the shoreline evolution from 2008 to 2014 indicates a clear contrast between the sectors west and east of Bray-Dunes. However, the meteorological and marine conditions (winds and water levels) were the same throughout the study area [50]. The first sector is characterized by a general but moderate progradation (a few meters) from 2008 to 2011 and 2012 and from 2012 to 2014 by a mean erosion of 6 m, up to 10 m in some sectors. The second sector east of Bray-Dunes underwent accretion over all the study period with a mean seaward migration of the shoreline of 10 m: it is especially noticeable from 2008 to 2011 and 2012 and is more moderate for the last period from 2012 to

Switching from 2D to 3D analysis enables to detect the shoreline position more accurately. 3D analysis using DTMs or DSMs also allows volume calculation and mapping of topographic changes. On sandy coasts, this approach can be used for calculating sediment volume changes and sediment budgets and for mapping accretionary and eroding sectors that are potentially at risk [51]. Below are two examples

*Evolution of the shoreline in Dunkirk and the Belgian border from 2008 to 2014 with transect location map* 

*DOI: http://dx.doi.org/10.5772/intechopen.91964*

eroded the beach and the foot of the dune.

2014 where some sectors even experienced mild erosion.

**3.2 Calculation of altimetric and volumetric changes**

at two different scales on the coast east of Dunkirk.

where *g* is the slope gradient, *z* is the altitude, and *x* and *y* are the planar coordinates of each pixel in the DTM. **Figure 8** shows a slope gradient map calculated from a DSM computed by photogrammetry in September 2019 in Zuydcoote. The figure also exhibits two cross-sectional profiles: one is a topographic profile, and the other is a slope gradient profile in degrees. In this example, a slope gradient of 22° corresponding to the dune toe has been selected. On the map, areas with slopes between 0 and 22° range from light to dark grey. They are mainly located on the beach; there are, however, some areas with low gradients in the coastal dunes

*Recent Advances in Coastal Survey Techniques: From GNSS to LiDAR and Digital… DOI: http://dx.doi.org/10.5772/intechopen.91964*

(e.g., paths). Zones with steeper slopes are not displayed on the map; therefore, the actual topography remains visible. The detection of this boundary is validated by comparison with the coastline surveyed in October 2019 with a GNNS: the differences range from 0 to 2.5 m landward, which are due to beach and dune toe erosion between September and October 2019, as a result of high water levels that had eroded the beach and the foot of the dune.

**Figure 9** illustrates a diachronic study using shorelines extracted from DTMs at different dates (2008, 2011, 2012, and 2014) on an 8 km long stretch of coast between Dunkirk and the Belgian border. The analysis of the shoreline evolution from 2008 to 2014 indicates a clear contrast between the sectors west and east of Bray-Dunes. However, the meteorological and marine conditions (winds and water levels) were the same throughout the study area [50]. The first sector is characterized by a general but moderate progradation (a few meters) from 2008 to 2011 and 2012 and from 2012 to 2014 by a mean erosion of 6 m, up to 10 m in some sectors. The second sector east of Bray-Dunes underwent accretion over all the study period with a mean seaward migration of the shoreline of 10 m: it is especially noticeable from 2008 to 2011 and 2012 and is more moderate for the last period from 2012 to 2014 where some sectors even experienced mild erosion.

## **3.2 Calculation of altimetric and volumetric changes**

Switching from 2D to 3D analysis enables to detect the shoreline position more accurately. 3D analysis using DTMs or DSMs also allows volume calculation and mapping of topographic changes. On sandy coasts, this approach can be used for calculating sediment volume changes and sediment budgets and for mapping accretionary and eroding sectors that are potentially at risk [51]. Below are two examples at two different scales on the coast east of Dunkirk.

#### **Figure 9.** *Evolution of the shoreline in Dunkirk and the Belgian border from 2008 to 2014 with transect location map (adapted from Ref. [38]).*

*Spatial Variability in Environmental Science - Patterns, Processes, and Analyses*

50 cm calculated by photogrammetry with aerial photographs of 2004. The recent fine-scale mapping of the upper beach and coastal dunes shows that the "Histolitt" shoreline is now outdated since its position does not correspond to the present day morphology anymore. Nevertheless, even if the updated shoreline position using the location of the HAT level on the recent DSM is more consistent with the actual present day shoreline, its position still lays several meters seaward of the dune toe (**Figure 8**) that is the geomorphological expression of the upper limit of action of the highest water levels, including tides and meteorological surges related to wind

*Example of shoreline detection on a DTM/DSM and comparison with the shoreline measured by GNSS in the* 

The shoreline may also be mapped in a semi-automatic way by identifying in a DTM for a neat variation of the slope gradient at the dune toe [38] using the follow-

> \_∂*z* <sup>∂</sup>*x*) 2 + ( \_∂*z* ∂*y*) 2

where *g* is the slope gradient, *z* is the altitude, and *x* and *y* are the planar coordinates of each pixel in the DTM. **Figure 8** shows a slope gradient map calculated from a DSM computed by photogrammetry in September 2019 in Zuydcoote. The figure also exhibits two cross-sectional profiles: one is a topographic profile, and the other is a slope gradient profile in degrees. In this example, a slope gradient of 22° corresponding to the dune toe has been selected. On the map, areas with slopes between 0 and 22° range from light to dark grey. They are mainly located on the beach; there are, however, some areas with low gradients in the coastal dunes

\_\_\_\_\_\_\_\_\_\_\_\_\_ (

(2)


and changes in atmospheric pressure [38].

**102**

ing approach:

**Figure 8.**

*field.*
