Geography IA

Student Code: kck854

Research Question:

Are the coastal defences used to protect the Start Bay coastline appropriate?

Word Count: 2,445

1
Contents:

Introduction………………………………………………………………………………………3

Hypothesis……………………………………………………………...………………………..6

Methodology……………………………………………………………………………………..8

Results & Analysis………………………………………………………………………..……10

Conclusion……………………………………………………………………………………...20

Evaluation………………………………………………………………………………………21

Bibliography…………………………………………………………………………………....23

2
Introduction:

Start Bay is an area of coastline in south-eastern Devon that ranges from Start-Point to

Blackpool Sands. This study is conducted in Slapton Sands, on this coastline. Figures 1-3 show

its exact location within Devon, and zooming in on Start Bay.

Figure 1 - Map of Devon

(alamy.com)

Figure 3 - Map of Slapton Figure 1 - Map of Start Bay

(ResearchGate.net) (dartmarinayachtharbour.com)

This coastline has been under erosional pressure for centuries, with first evidence of this going

back to the end of the 18 th century, when the village Strete Undercliffe was taken by the sea, as

well as Hallsands in 1916, shown in figure 4 below. This makes it an area of great geographical

interest and an excellent site to study coastal erosion and methods employed to hold the line.

Figure 4 – Before and After of 1916 storm in Hallsands

(news.bbc.co.uk)

3
 Slapton Ley

Longshore drift is taking place in this area: the movement of ‘The Line’ (Bar Beach)

sediment along the coast due to wave motion. Evidence of this

can be observed in Slapton Sands, where a bar has formed due

to this over 10,000 years ago, creating a freshwater lagoon

called Slapton Ley. This is shown in Figure 5.

Figure 5 – Slapton Bar (CoastandCountry.co.uk)

The erosion in this area represents a significant problem for the

local population. An important road exists along the bar, the

A379, but now the eroding process is threatening to wear

down the bar and eventually reach the road, along with other

villages along the coast. Local stakeholders (governmental

organisations, local authorities, residents) have installed

defences to stall the erosional process and protect

infrastructure. Other stakeholders include local business

owners, tourists, and commuters through the A379. Figure 6

shows the distribution of urban land in Start Bay, which will

be where these stakeholders are concentrated, therefore will

be the most crucial areas to protect from coastal erosion

using defences, which poses a threat to coastal

Figure 6 – Hand drawn map of Start Bay (student
infrastructure. During my time here, I will be investigating drawn)

and assessing the ‘appropriateness’ of these defences. ‘appropriate’ means that the coastal

defences used are in the optimal places to protect infrastructure on the shore, and that the

money spent and size of the coastal defence is suitable to the area of coast it is protecting.

4
Hypothesis:

The day prior to data collection, a pilot study was conducted at Torcross beach. I deduced that

longshore drift was taking place, moving sediment northeast. Furthermore, the infrastructure

directly exposed to the shore were entirely commercial and residential, protected by a sea wall

and rip rap.

From this, I formed two hypotheses that I could test in my study.

1 - “As land value increases, the spending on coastal defences increases”. In case of a

storm, the government will aim to limit financial damage, as it will be easier to recover from

these damages. Figure 7 show a contrast between the flood defences used for urban area in

Beesands, and the minimal ones used for agricultural land. This hypothesis helps answer the

research question because if it is confirmed, it means coastal defences are targeted at more

important areas of the coast and can therefore be considered appropriate.

Figure 7 – Coastal defenses in Beesands vs

Agricultural Land (student taken)

5
2 - “Smaller and more rounded pebbles will be found in the North of Start Bay in

comparison to the South of Start Bay”. Longshore drift is moving sediment along the beach

northward. This means that sediment carried to the North of Start Bay will have spent longer

transported by the waves than sediment in the South, so it will be more worn down by attrition.

This will help answer the research question because it confirms the occurrence of longshore

drift, and if it is taking place this means that southern areas of Start Bay will have less sediment

and therefore be more vulnerable to coastal erosion, so will need more intense coastal

defences.

6
Methodology:

To investigate my hypotheses, I will use a stratified sampling technique, selecting specific

sampling sites and evaluating the different samples to make a conclusion. The sites are shown

in the map below:

Figure 8 – Map of Sample Sites (student made – earth.google.com)

My first hypothesis can be investigated

using a cost-benefit analysis. To calculate the cost benefit of the sampling sites, the total value

of exposed properties and the cost of the coastal defence per 100 years need to be found.

Collecting the raw data for this firstly involves identifying the number of small and large houses,

and businesses along the shore of each site, as well as the coastal defence used.

For the total value of the properties, estimates can be used for the value of each type of

establishment: £300,000 for small houses, £550,000 for large houses and £750,000 for

7
businesses. As for other irregular establishments, an estimate can be made using Internet

research. These are then summed up to give one value for each site.

For the cost of the coastal defence per 100 years, using reliable resources provided during my

investigation1, this can be found by dividing 100 by the estimated lifespan of the defence, and

multiplying this by its cost, including repair works.

Sands
Slapton

Torcross

Sunnydale

Beesands

Hallsands
North
Location

Defence Scheme
Date Built
Cost

Estimated life span

Repair works
B:

B:

B:
B:
L:
S:

L:
S:

L:
S:

L:
S:
B:
L:
S:

Number of properties
(small, large, business/ you
make up your own
categories)
Total value of properties
Cost for defence per 100
years (inc. value of repair
work)
Cost benefit = total cost of
properties divided by Cost
of defence per 100 years
Cost benefit is found by dividing total value of properties by the total cost of defence per 100

years. All this data will be stored in the table below:

1
Masselink, G. and Buscombe, D. (2008). Shifting Gravel: A case study of Slapton
Sands. GeographyReview. Sep.

8
For my second hypothesis, I can use coarse sediment analysis to measure and compare the

size and shape of sediment. Data points are taken 16 times along Slapton Sands, with 200m

between each. At each site, 10 pebbles are chosen at random, and using callipers, their lengths

are accurately measured (Figure 9 below) and an average is found for each data point.

Figure 9 – Using calipers to determine the length Figure 10 – Power’s Scale of Roundness (ResearchGate.net)
of a pebble (student taken)

For shape, I visual

analyse the sediment

collected Table 1 – Empty Cost-Benefit Analysis Table (student made) using a well-

known

Figure 11 – Pebble Sampling (student taken)

classification: Power’s Scale of Roundness (Figure 10). Each individual pebble will be classified,

to show the distribution in each data point. I can then use this data to determine whether the

trend stated in my hypothesis is present or not. Below is a photo that clearly demonstrates the

procedure:

9
Repeated 16 times along
the beach, with 200m
between each sampling
site (photo distance not
accurate).

10 pebbles are selected

randomly at the first
change in gradient of the
beach.

Results & Analysis:

The results obtained from the cost benefit analysis are in Appendix 1. They are shown in a

scatter graph in Figure 11 below.

Figure 11 – Scatter Graph of CBA (student made)

10
This data suggests that there could be a positive correlation between the value of the properties

and the cost for defence per 100 years. It is easy to see that in Beesands and Torcross, the most

valuable areas (£14,570,000 and £12,210,000 respectively), there is higher cost for defence per

100 years (£3,000,000 and £2,50,000 respectively). These are the more urban areas, with the

most residents, and therefore the government has the most interest in protecting these areas. The

dotted trend line shows a positive correlation. The only anomaly to this is the data for Slapton

Sands.

However, being able to visually see an increase in both variables does not necessarily confirm the

correlation. To mathematically verify the positive correlation, I can use Spearman’s Rank. Using

the following equation, I can obtain a Spearman’s rank correlation coefficient for my data:

Figure 12 – Spearman’s Rank Formula (statisticshowto.com)

This gives a coefficient of 0.9, which means there is a strong correlation between the value of a

certain area and the defensive coastal measures taken against erosion.

The success of the statistical test allows my first hypothesis, “As land value increases, the

spending on coastal defences increases”, to be confirmed for the Start Bay area. Furthermore,

the correlation could be even stronger: the value of Slapton Sands only accounts for the

monetary value of the A379, when it could be higher if considering its intrinsic value for

commuters making use of it. It saves about 8 hours of travel by crossing it. If it were destroyed,

11
it can be easily imagined that the financial damage will extend much further than simply

repairing the road. This explains the anomaly previously mentioned.

Linking these results to the research question, this means authorities are spending more to

protect areas of high values, suggesting that the defences are appropriate.

As for sediment analysis, this map below shows the average pebble length taken across 10

samples in each of the locations (data available in appendix 2):

It

can be seen as a general trend that the pebbles get smaller further North, which is in

Figure 13 – Map of Pebble Length Data (arcgis.com)

accordance with the second hypothesis.

12
Appendix 3 shows the data obtained for the shape of the pebbles. Theoretically, it would be

expected that the total pebbles sampled should be split equally across all six categories of

Power’s Scale of Roundness.

However, the data deviates considerably from this. Here is the data presented in a bar chart:

Observed Roundness Frequencies

43 43

34

18

12
10

very angular angular sub-angular sub-rounded rounded well rounded

Figure 13 – Bar Chart of Pebble Roundness Totals

It is already visible that the

results deviate significantly from what was expected. To mathematically test the extent to which

the data deviates, a Chi squared (X2) goodness of fit test can be conducted. The formula for this

is:

Figure 13 – Chi Squared Goodness of Fit Test (scribbr.com)

13
Which gives X2 = 43.33, and a critical value of 22.362.

Considering the X2 value is much larger than the critical value, it must be recognised that the

pebble data is split randomly across the 6 categories.

Overall, the second hypothesis, “Smaller and more rounded pebbles will be found in the North

of Start Bay in comparison to the South of Start Bay”, can only be partially accepted. The data

for length of pebbles, although slightly inconsistent, does show that on average, the length of

pebbles decreases as one progresses further Northward of Start Bay. However, the data for

shape of the pebbles shows heavy deviation from what I had theoretically expected it to be.

Linking back to the research question, if pebbles in the South of Start Bay are larger on

average, this will cause more erosional pressure, meaning that using coastal defences more

extensively in this area would be appropriate.

Longshore drift is therefore definitely occurring northward along Start Bay. However, there must

be an external factor that has affected the pebble shape distribution to be different from what

was expected.

This is most logically attributed to human engineering. My findings have led me to an

understanding that beach nourishment had most likely taken place in Start Bay, as a method of

soft engineering against coastal erosion. Authorities have transported pebbles across the beach

to areas with less protection, which would help slow down waves as they crash against the

coast. This would most definitely explain why the pebble shape distribution is far from what was

expected.

14
Conclusion:

To conclude, I have found that only one of my hypotheses was successfully confirmed, to a

certain extent.

The first hypothesis, at first sight, seemed very plausible. When expressed with numerical data

and using Spearman’s Rank Test to statistically assess the correlation, the coefficient came out

to be 0.9, where 1 is a perfect correlation. Therefore, this fully confirms the theory that, “As land

value increases, the spending on coastal defences increases”.

The second hypothesis assessed the trend in length and shape along beach. When expressed

on a map, it was shown that the length of the pebbles decreases Northward of Start Bay. As for

the shape, using Power’s Scale of Roundness, the data deviated quite considerably from the

expected data, which was found using a Chi Squared Goodness of Fit Test. This was attributed

to human intervention. Therefore, it is very plausible that if beach engineering hadn’t taken

place, this hypothesis could have been confirmed, and that the thought process behind this (to

do with longshore drift) was correct.

The results obtained from my research also enable me to answer the research question: the

coastal defences used in Start Bay are indeed appropriate. The cost-benefit analysis has

showed that more extensive coastal defences are used in areas of higher value. The sediment

analysis has shown that pebbles in the South of Start Bay tend to be generally bigger, which

15
means that it is appropriate that more expensive coastal defences are used in this area, as

bigger pebbles hurled by the waves at the coastline will cause more significant erosion, and this

study of Start Bay shows that this is the case.

Evaluation:

While the methods utilised in my investigation were valid and applicable, there are perhaps

some aspects that could be changed for better results.

For the cost benefit analysis, it could be argued that my method of visually assessing the size of

houses is slightly unreliable. Rather than assuming based on personal judgement, it may have

been better to establish a criterion. For instance, classifying their size using the number of

windows could prove to be much more reliable. Furthermore, adding a medium category would

be beneficial as there is a large difference in value between small and large houses, and would

allow for more accurate total values of each site.

Additionally, having only 5 datapoints is not favourable to test a correlation. While Spearman’s

Rank can still be used, the correlation found is much more likely to be correct if more datapoints

are used. For this, it may have been better to investigate more sites further North up Start Bay.

The data presentation using a scatter graph was appropriate and useful to visually see the

correlation, but again due to the low number of data points, the trend line obtained is not very

reliable. However, this method allowed me to single out the anomaly, therefore it has been

successful.

For the sediment analysis, the unexpected results perhaps suggest that there is more that could

be improved with methodology. Firstly, I could extend the sampling area for the pebbles to the

16
entirety of Start Bay. While this was not done in my investigation because it would be time

consuming, it would have improved the accuracy of my results. If sampling took place over a

longer distance, a trend in the shape and length of pebbles would have been much more

apparent, as logically, the change will be much less significant when observed over a short

distance.

Increasing the number of sediment samples that the averages were taken over is always

guaranteed to improve the accuracy of my results and eliminate anomalies. However, this is

again more time-consuming and may not have been possible with the time given.

Presenting the lengths of pebble samples on a map was the right data presentation method to

use because I was able to demonstrate how the length changes along the coast. An alternative

method could have been to plot the data on a graph, but this would not have been as clear

without the use of a map. Finally, the bar chart for the roundness of the pebbles was able to

express the frequency for each category very well.

17
 18
Estimated life span
Defence Scheme

Repair works
Date Built
Location

Cost
North Rip-rap 1991 £100,000 25 Rip-rap
Hallsands and road
repair in
2014 (cost
£20,000,
life span 5
Beesands Wave- 1992 £3 million 100 years None
return sea
Appendix 1: Cost Benefit Analysis Results Table

wall and
rock
armour
(rip-rap)
Sunnydale Gabions 1998 £30,000 20 None
(private)
Torcross Wave- 1980 £2 million 80 Building
return sea repair in
wall and 2014 (cost
rock £500,000,

Appendices:
armour life span
Slapton X X X X 20
2014,
Sands 2018
A379
 Parking
B: 6
L: 8
S: 11
B: 0
L: 3
S: 0
Toilets
Church,
Big Parking Lot,
B: 4
L: 8
S: 22

B: 0
L: 3
S: 11
A379 Road

Parking Lot
Number of properties
(small, large,
business/ you make
up your own

Public
categories)
£12,210,000

£1,650,000

£14,570,000

£4,960,000
£4,000,000

Total value of
properties
£2,500,000

£150,000

£3,000,000

£400,000
back by 10m)
(moving road
£2,500,000

Cost for defence per

100 years (inc. value of
repair work)
Cost benefit = total
4.9

11

4.9

12.4
1.6

cost of properties
divided by Cost of
defence per 100 years

Appendix 2: Average Pebble Length

Latitude Longitud Average Pebble

e Length
50.2702 - 3.13
3.65106
50.2714 - 2.85
1 3.65068
50.2727 - 2.93
3 3.65027
50.2734 - 1.03
3 3.65026
50.275 -3.649 5.41
50.2767 - 1.84
1 3.64881
50.2777 - 2.9
9 3.64874
50.2795 - 1.88

19
 9 3.64777
50.2804 - 3.62
3.64768
50.2824 - 2.55
5 3.64658
50.2839 - 3.44
7 3.64593
50.2854 - 1.92
9 3.64505
50.2868 -3.6443 1.21
7
50.2882 -3.6437 1.83
50.2897 - 3.11
9 3.64274
50.2917 - 0.97
9 3.64123

Appendix 3: Pebble Roundness Data

power's scale of roundness totals

very angula sub- sub- rounde well

  angular r angular rounded d rounded

expectation

s 26.67 26.67 26.67 26.67 26.67 26.67

results 12 10 18 34 43 43

20
Bibliography:

BBC News (2010). In pictures: Hallsands then and now. news.bbc.co.uk. [online] 29 Jan.

Available at:

http://news.bbc.co.uk/local/devon/hi/people_and_places/history/newsid_8487000/8487912.stm

[Accessed 18 Sep. 2022].

Field Studies Council (n.d.). Method for Coastal Management. [online] Field Studies Council.

Available at: https://www.field-studies-council.org/resources/16-18-geography/coasts/coastal-

management/method/ [Accessed 18 Sep. 2022].

Gupta, A. (2021). Spearman’s Rank Correlation: The Definitive Guide To Understand |

Simplilearn. [online] Simplilearn.com. Available at:

https://www.simplilearn.com/tutorials/statistics-tutorial/spearmans-rank-correlation [Accessed 22

21
Sep. 2022].

Makepeace, M. (2020). Slapton Sands | South Devon Beach Guide. [online] Coast & Country

Cottages. Available at: https://www.coastandcountry.co.uk/blog/slapton-sands-south-devon-

beach-bible [Accessed 20 Sep. 2022].

Scribbr (2022). Chi-Square Goodness of Fit Test | Formula, Guide & Examples. [online] Scribbr.

Available at: https://www.scribbr.com/statistics/chi-square-goodness-of-fit/ [Accessed 23 Sep.

2022].

Masselink, G. and Buscombe, D. (2008). Shifting Gravel: A case study of Slapton

Sands. GeographyReview. Sep.

Trudgill, S. (2009). ‘You can’t resist the sea’: evolving attitudes and responses to coastal erosion
at Slapton, South Devon. Department of Geography, University of Cambridge.

Watson, J. (2011). Coastal Erosion at Slapton. GeographyReview. Nov.

22