Next Meeting

H G S home page

Humberside Geologist on-line

Contact us

Humberside Geologist no. 15

The Provenance of the Glacial Tills on the Holderness Coast, East Yorkshire, UK.

By Luke Beaumont

Abstract:

This is a study of the particle size distribution, lithology, till colour and other morphological features of the Late Devensian deposits (Skipsea and Withernsea Tills) on the Holderness Coast. The aims were to determine the provenance of these tills whilst ultimately considering what ice routes and mechanisms would deposit them.

It is proposed that there is a significant enough difference within the Skipsea Till to separate it into two different till types (referred to in this paper as Skipsea A and Skipsea B Tills). This is based on distinct lithologies and particle size distribution. The provenance of the Skipsea A and Withernsea Tills were determined to be the western Lake District with the ice route passing through: the Vale of Eden, the Stainmore Gap (Pennines) and through the Tees Valley. The provenance of the Skipsea B Till was determined to be southern Scotland and Northumberland with the ice route being the North Sea ice lobe surges onto the Holderness Coast.

Introduction.

This project was undertaken as a final year undergraduate research dissertation at the University of Brighton. The fieldwork was carried out over the summer of 2009.

Fieldwork Site Descriptions

The selected sample sites were the till beds at the bottom of the exposed cliff face along the Holderness Coast. Ranging 33.932km from the first sample on Barmston Sands (NGR TA1704959253) to sample thirty five at Waxholme (NGR TA3263330094) just under two kilometres from Withernsea The sample sites were approximately every 500m from the previous sample site when appropriate. The National Grid Reference and a description of the colour of each sample is given in table 1.

Sampling Methodology

To collect till samples it is important not to contaminate the samples as this will lead to anomalous results. To achieve this, the face of the till was scraped away using a clean trowel to ensure the sample would not be contaminated by external factors. A 1 Kg sample was removed straight from the till face and placed into a sampling bag where it was immediately labelled.

Description of soil colour

To quantify soil colour the Munsell soil colour chart was used. The colours observed are listed with the list of sample number (Table 1). Samples 1 to 27 are ascribed to the Skipsea Till by colour and the reddish ones to the Withernsea Till.

Table 1

Sample No.	Site Name	Grid Reference	Munsell	Colour Description
1	Barmston Sands	TA1704959253	10YR/4/2	Dark Greyish Brown
2	Barmston Main Drain	TA1716458489	10YR/3/2	Very Dark Greyish Brown
3	Ulrome Sands	TA1730557952	10YR/4/2	Dark Greyish Brown
4	Galleon Beach	TA1765556730	10YR/4/2	Dark Greyish Brown
5	Tranmere Cliff	TA1789355973	10YR/4/2	Dark Greyish Brown
6	Skipsea Sands/Cliff	TA1818655175	10YR/3/2	Very Dark Greyish Brown
7	Skipsea Sands/Cliff	TA1862953916	10YR/3/2	Very Dark Greyish Brown
8	Skirlington Sands/Cliff	TA1899952819	10YR/4/2	Dark Greyish Brown
9	Atwick Sands/Cliff	TA1953451336	10YR/4/2	Dark Greyish Brown
10	Atwick Sands/Cliff	TA1974650790	10YR/5/2	Greyish Brown
11	Atwick Sands/Cliff	TA1993150305	10YR/5/2	Greyish Brown
12	Hornsea North Cliff	TA2008849858	10YR/5/2	Greyish Brown
13	Hornsea North Cliff	TA2032749385	10YR/5/2	Greyish Brown
14	Hornsea South Cliff	TA2119346983	10YR/5/2	Greyish Brown
15	Hornsea South Cliff	TA2130946516	10YR/5/2	Greyish Brown
16	Rolston Sands/Cliff	TA2151746101	10YR/4/1	Dark Grey
17	Rolston Sands/Cliff	TA2177045628	10YR/4/1	Dark Grey
18	Rolston Sands/Cliff	TA2199745227	10YR/4/1	Dark Grey
19	Mappleton Sands/Cliff	TA2227744778	10YR/4/1	Dark Grey
20	Mappleton Sands/Cliff	TA2251344382	10YR/4/1	Dark Grey
21	Mappleton Sands/Cliff	TA2282343805	10YR/4/2	Dark Greyish Brown
22	Garton Moat Farm	TA2815436118	10YR/5/2	Greyish Brown
23	Garton Moat Farm	TA2844335738	10YR/5/2	Greyish Brown
24	Garton Bracken Hill	TA2870335310	10YR/5/2	Greyish Brown
25	Garton Bracken Hill	TA2898234908	10YR/6/2	Light Brownish Grey
26	Tunstall Pastures	TA2933234428	10YR/6/2	Light Brownish Grey
27	Tunstall Pastures	TA2972633887	10YR/5/2	Greyish Brown
28	Monkwith	TA3003033488	5YR/4/3	Reddish Brown
29	Monkwith	TA3031833115	5YR/4/2	Dark Reddish Grey
30	Tunstall	TA3064832716	5YR/4/2	Dark Reddish Grey
31	Tunstall	TA3097732314	5YR/5/2	Reddish Grey
32	Tunstall 'Submarine Forest'	TA3169731349	5YR/5/2	Reddish Grey
33	Tunstall 'Submarine Forest'	TA3232330491	5YR/5/2	Reddish Grey
34	Waxholme	TA3263330094	5YR/5/2	Reddish Grey
35	Withernsea North Cliff	TA3297029712	5YR/5/2	Reddish Grey

Particle size analysis

The method used for particle size analysis in this project is the laser diffraction technique using the Mastersizer 2000 instrument. This measures particles between 0.02µm and 2000µm. For the 35 samples taken this instrument automatically goes through three readings and then presents an average reading per sample. The principle of this technique is that as the particles pass through a laser beam they scatter the light at an angle relative to their size. Larger particles disperse the light at a lower angle and smaller particles disperse light at a higher angle (Ma et al., 2000).

Table 2 shows the percentages of clay, silt and sand contained in the thirty five till samples. From the table it is clear that the percentage of silt is the most dominant and the trend generally increases from between sample one and sample thirty five. There is an overall increase of 18.29%. The lowest silt percentage is found in sample 8 with a value of 48.32% and the highest percentage is found in sample 29 with a value of 69.88%. From the table there appears to be three sections of similar results in the percentage of silt. These are between till samples 1 and 13 where the average silt content increases by 9.99% for in-between samples 14 and 27. Then this increases again by 3.98% for the remaining samples. The second overall largest particle size is the percentage of sand. The general trend is that the percentage of sand decreases across the 35 till samples. There is a 21.93% decrease in-between the first and last sample. The highest sand percentage is till sample 8 at 42.39% and the lowest is sample 21 at 14.95%. The percentage of sand particles appears to have only two groups of similar results. These are between samples 1 and 13 where the average sand content then decreases by 15.38% from till sample 14 onwards. The overall least common particle size is clay. The overall trend of the percentage of clay is that it generally increases from samples 1 to 27 before slightly decreasing for the remaining samples. The highest percentage of clay is in sample 19 with a value of 18.9% and the lowest percentage is in till sample 1 with a value of 8.24%. Similarly to the silt, clay has three groups of similar results. These are between samples 1 and 13 where the average clay content increases by 5.35% between till samples 14 to 27. Then from sample 28 onwards this average decreases slightly by 3.87%. Across the 3 particle sizes there appears to be general similarities for till samples 1 to 13, 14 to 27 and 28 to 35. Lastly there is only one clear anomaly found in sample 20 where the percentages of clay and silt are lower than expected and the percentage of sand is higher.

For till samples 1 to12 most of the distributions show a bi-modal distribution where the percentage of silt is most dominant as it covers the largest area but also shows how sand had relatively higher percentages from samples 1 to 12. The only exception to this is sample 8 which shows a slight tri-modal distribution where the percentages of larger sand particles have slightly increased. This is reflected in table 2 as it shows that sample 8 had a higher percentage of sand relative to the average of the first 12 samples.

There is a distinct difference from till sample 14 onwards as the percentage of sand suddenly decreases. Moreover from sample 14 to 28 there is a slight increase in the percentage of clay particles. However this is not as apparent due to the very high percentage of silt and would have been difficult to identify from the distribution graphs if the particle size percentage data from table 2 was not present (see figure 1 for examples). Moreover, from sample 28 onward it is shown that there is a slight decrease in the clay percentages. This coincides with the final overall increase in silt percentages and displays the most pronounced uni-modal distribution for till sample 28 to 35.

An anomaly noted in sample 20 is also evident from the distribution graphs as the sudden increase in sand and decrease in silt percentages are displayed as slightly bi-modal which is noticeably different from the other graphs.

Table 2: A table showing the percentages of clay, silt and sand found in the 35 till samples.

Sample No.	% Clay	% Silt	% Sand
1	8.24	51.03	40.74
2	11.34	62.59	26.06
3	12.99	60.88	26.13
4	10.05	49.13	40.82
5	10.65	52.31	37.05
6	9.13	52.93	37.94
7	8.40	53.26	38.34
8	9.29	48.32	42.39
9	9.80	52.05	38.15
10	10.92	54.10	34.97
11	12.06	59.79	28.15
12	14.68	56.18	29.13
13	11.41	57.77	30.81
14	17.02	64.46	18.52
15	17.82	67.08	15.09
16	16.40	67.36	16.25
17	18.01	65.01	16.98
18	13.88	68.05	18.07
19	18.90	66.81	14.29
20	13.21	55.43	31.36
21	17.91	67.15	14.95
22	15.11	63.70	21.19
23	13.87	64.18	21.95
24	16.77	63.79	19.44
25	15.65	66.62	17.73
26	14.73	65.14	20.13
27	15.25	60.10	24.65
28	11.78	68.22	20.00
29	13.41	69.88	16.71
30	12.83	66.11	21.06
31	12.36	67.98	19.66
32	10.71	68.66	20.63
33	11.08	69.05	19.87
34	13.29	69.63	17.08
35	11.87	69.32	18.81

Till sample 1

Till sample 14

Till sample 20

Till sample 28

(Figure 1. some typical particle size distribution graphs from the research)

Figure 2.Triangular plot of the results.

X-ray diffraction

X-ray diffraction (XRD) is the most widely used method in determining clay mineralogy. The analysis was carried out on ‘whole rock’ subsamples of the till that were less than 2mm in size. The till sample is prepared by drying the sample in an oven to remove moisture; crushing the sample into a fine powder using a mortar and pestle to randomize the orientation of the clay minerals. The sample is then compacted into a holder ensuring that there is a flat upper surface to achieve a randomized distribution of the clay minerals. Care is taken to clean all the preparation equipment each time for all 35 samples to ensure no cross-contamination.

Unfortunately, the results for XRD were both insignificant and sporadic giving no indication of any relationships within the tills or between any of the till types. Moreover, the results showed no resemblance to the mineralogical results from past research by Madgett and Catt (1978). This was probably because Madgett and Catt had anaylsed the clay grade material (less than 2um) rather than ‘whole rock’.

Clast Roundness

The clast roundness is defined by Hubbard and Glasser (2005) as the small-scale directional change of a clasts surface. For this project 100 clasts were collected at each of the thirty five sample sites giving a total of 3500 clasts.

Apart from the first few sample sites where there is a relatively higher percentage of angular shaped clasts there appears to be no trend found in the 35 samples where on average the percentage of sub-rounded clasts is the most dominant with a value of 49.4% with sub-angular being second most dominant on average 10% less than sub-rounded. Moreover figure 14 shows that there is very little change found in the 35 samples with the percentages for sub-rounded and sub-angular fluctuating only slightly from the average with only few major peaks for sub-rounded clasts at sample sites: 28 (62%), 21 (61%), 26 (60%) and 35 (60%). These results suggest that the mode of deposition does not change significantly across the whole sampling area and that all tills are likely to be sub-glacial in nature.

Figure 3: A line graph showing the number of clasts for each roundness type across the 35 sample sites.

Clast Lithological Analysis *

The purpose of collecting samples for clast lithological analysis is to obtain a sufficient amount to make the results statistically significant. The same 3500 clasts collected for the clast roundness study were used in this analysis (i.e. 100 from each site). This allowed for a sufficient comparison between sample sites across a wide ranging distance. Moreover it provides a better representation of clast types as it will allow a greater range in those collected. As the sample sites were cliff exposures this provides ideal conditions for sampling as clasts are readily available. Furthermore due to the rapidly eroding cliff in Holderness providing a clean face it means that less time is spent on cutting a new cliff face to access valid samples. The sizes of the clasts collected for the project were between 40 to80mm. This size range was chosen because it provides large enough clasts for easy identification. This is because with a smaller size range it would become too difficult to break a new rock face to help identify the rock type. In this project systematic sampling was applied whereby a 1m² area was marked out on the exposure and starting from the top working to the bottom collecting all clast types that were in the size range until the required amount of samples were reached. If a 1m² area was insufficient in obtaining the required amount this area would be extended by a further 1m² until a sufficient amount was collected (Bridgland, 1986).

The aim of lithological analysis is to correctly identify and classify the individual clasts in the samples. This therefore requires some basic geological knowledge. However, it was not possible to identify each individual clast type within a sample area due to the large range existing. Therefore the clasts were separated into general categories to allow more manageable data analysis (Gale and Hoare, 1992). Eight basic rock categories were chosen for this project based on early observations of prominent clast types. These categories are: Chalk, Jurassic Sandstone, Jurassic Limestone, Flint, Carboniferous Limestone, Undifferentiated Red Sandstone, Quartz and Igneous/Metamorphic. Methods used to separate clast types included breaking the rock to provide a fresh un-weathered face and the use of hydrochloric acid to identify if the clast contains carbonates (Bridgland, 1986). The results are presented in Table 3.

There appear to be similar trends in the percentages of clasts present at sample sites in comparison to other test results such as particle size analysis. It is shown that some of the clast types display differences in either three distinct sections or differences at either the first sample sites or the last sample sites. For example the average number of Chalk samples starts relatively low at around 6% before rapidly increasing by 34% between sample sites 12 to 28. This number then rapidly decreases to an average of 10%. The reverse of the Chalk results are the average numbers of igneous and metamorphic clasts that start very high with a value of 33% before dropping by an average of 17% between sample sites 12 to 27 then increasing to an average of 32%. The numbers of Jurassic sandstones also follow this trend where the averages are relatively higher between sample sites1 to 14 and 31 to 35 in comparison to the middle group of sample sites. Moreover whilst all the clast types do not follow this exact trend they do show some distinct differences at either the northern or southern sample sites. For example between sample sites 1 to 10 the average number of flint is three times greater than sample site 11 onwards, showing a distinct difference at the beginning. On the other hand Carboniferous Limestone shows an average increase in number of 50% from sample sites 29 to 35 in comparison to the sample sites 1 to 28. However, not all the clast types follow these trends. This is because the percentages of Jurassic Limestone, undifferentiated red sandstone and quartz show no real differences throughout the sample sites and all remain relatively low showing no differences in the northern, middle or southern groups of the sample sites.

We can see that there appears to be a distinct difference around sample site 12 where Jurassic sandstone, flint and igneous and metamorphic rocks start to show decreasing percentages whilst Chalk increases suddenly for a middle section where Chalk remains the most dominant clast type between sample sites 12 to 28. Then from approximately sample site 29 onwards it is shown that Chalk decreases suddenly back down to very low percentages and the percentages of Jurassic sandstone, Carboniferous Limestone and igneous and metamorphic clast types begin to increase until the last sample site. The results for Jurassic limestone, undifferentiated red sandstone and quartz are more irregular than the other sections which help show that there were no real trends with these clast types in these samples.

Table 3: Shows the percentages of the different clast types at found in the 35 sample sites.

Sample Site	Chalk	Jurassic sandstone	Jurassic limestone	Flint	Carboniferous Limestone	Undifferentiated red sandstone	Quartz	Igneous / metamorphic
1	1	11	10	17	12	0	2	45
2	3	15	5	10	14	5	2	39
3	3	13	7	14	11	5	2	37
4	4	11	2	9	15	7	8	34
5	6	14	8	10	19	3	5	24
6	7	14	5	9	16	9	2	26
7	12	15	5	19	7	5	5	30
8	4	13	5	7	16	2	0	40
9	6	13	3	13	10	8	0	35
10	10	13	7	12	7	10	9	25
11	14	12	5	6	10	6	7	29
12	49	12	1	3	5	2	4	17
13	32	17	3	5	5	5	2	21
14	35	14	4	3	15	4	0	13
15	43	5	5	3	10	5	2	14
16	39	7	2	4	11	2	5	21
17	45	2	8	4	12	0	8	19
18	38	4	8	2	17	3	7	19
19	47	11	4	4	9	5	4	14
20	47	8	2	3	11	0	5	15
21	37	9	5	3	11	5	5	17
22	36	8	2	5	15	4	7	15
23	32	3	6	3	13	4	9	28
24	36	7	6	4	13	3	2	19
25	35	9	5	3	12	4	5	21
26	45	5	1	1	18	3	4	15
27	52	4	4	3	11	0	5	15
28	36	8	3	3	11	6	4	27
29	11	10	7	3	16	12	4	25
30	11	7	4	3	19	7	3	35
31	9	15	3	3	18	6	6	31
32	7	12	4	7	21	2	6	35
33	8	11	4	6	17	6	4	30
34	10	12	5	5	16	6	3	37
35	11	10	3	4	18	6	5	33

Chalk shows that relatively higher percentages between samples 12 to 28 were significantly different to the samples at the beginning and end of the sampling area. Jurassic sandstone was proven to have had a significantly smaller middle section. Carboniferous Limestone has significantly larger percentages for sample sites 29 to 35. Flint proved to have significantly greater percentages in sample sites 1 to 10. Finally, igneous and metamorphic rocks were proven to have significantly smaller percentages of rocks for sample sites 12 to 27.

Discussion *

Skipsea A Till lithology

The dominant lithologies of the Skipsea A Till are: igneous/metamorphic (33.1%), Jurassic clasts (28%) and Carboniferous Limestone (12%). The high percentage of Jurassic clasts indicates that some of the till is locally sourced from the North York Moors. This could have been transported by at least two possible routes. Firstly, from the Tees Valley ice stream which could have picked up the Jurassic rocks as it moved around the North York Moors after being diverted by the North Sea ice lobe. This suggested source could also satisfy the other dominant lithologies due to the origin of this ice stream being the Lake District which contains Carboniferous Limestone and various igneous rocks (such as granite and tuff) and south-western Scotland which also contains some igneous clasts. Another potential source would be the North Sea ice lobe that flowed down the east coast. This would explain the higher igneous/metamorphic and Carboniferous Limestone percentage due to the provenance of this ice stream being Northumberland and south-eastern Scotland where there are abundant sources of these clast types. This ice stream could have transported the Jurassic clasts as it moved past the North York Moors. However, this does not explain the low percentage of Chalk as it would then have to move onto the Chalk bedrock before deposition at the Skipsea A Till site. However the low Chalk percentage can be explained with the provenance of the ice flow being from the Tees Valley. This is because the Tees Valley ice stream could have overridden the North Sea ice lobe that was surging onto the Holderness coast meaning that it could not actively erode the underlying chalk be drock as much before depositing the sediment.

Skipsea B Till lithology

The dominant lithologies of the Skipsea B Till are: Chalk (40%), igneous/metamorphic (18%) and Carboniferous Limestone (12%). The most probable origin for this till is that it was part of the North Sea ice lobe. This is because the large amount of Chalk is most likely being brought onto the Holderness coast by the north-east to south-west ice flow as it moved down the east coast. Other than Chalk there are still relatively moderate percentages of both Igneous/Metamorphic rocks and Carboniferous Limestone whose percentages could have been diluted by the very high percentages of Chalk. These would also indicate that this till could have come from the North Sea ice lobe as far-travelled clasts from Northumberland and Scotland. It is possible to discount that this till could have been deposited by the Tees Valley Ice stream as it seems unlikely that such a high percentage of chalk could have been transported in this relatively short distance from the North York Moors to the Holderness coast in comparison to the greater distance the North Sea ice lobe would have travelled through the Chalk bedrock extensions out into the North Sea.

Withernsea Till lithology

The dominant lithologies of the Withernsea Till are: igneous/metamorphic (32%), Jurassic (19.5%) and Carboniferous Limestone (17.9%). The most probable provenance of this till, based on the lithology and analysis of geological maps, is similar to Skipsea A Till, whereby, the high percentage of the dominant lithologies is due to the ice flow from the Tees Valley and that the low chalk percentage can be attributed to the overriding of this ice stream, over existing ice sheets, limiting the percentage of Chalk transported.

Conclusions

The main conclusions that can be gained from this project are:

· It was possible to differentiate the samples studied into three Tills – Skipsea A (samples 1-11), Skipsea B (samples 12-27) and Withernsea (samples 28-35). The Skipsea Tills exhibit lateral variation.

· That based on a significant difference between particle size distribution and lithological analysis it is suggested that the traditional Skipsea Till can be divided into two separate tills (Skipsea A and Skipsea B). If these tills can be differentiated from one another then it is possible that they have a different provenance that will ultimately indicate the ice route.

· Skipsea A Till and the Withernsea Till have a provenance based on the Tees Valley ice stream. The high percentage of Igneous and Carboniferous Limestone can be explained by the origin of this ice stream being in the Lake District. The high percentage of Jurassic clasts can be explained by this ice stream’s route past the North York Moors before being diverted southwards towards the Holderness Coast by the North Sea ice lobe and the low percentage of Chalk clasts may be because this ice stream overrode the ice surging in from the North Sea meaning that it was unable to erode the Chalk bedrock.

· The overall provenance of Skipsea B Till is that it was most likely deposited by the North Sea ice lobe. This is because they have similar clast orientations and share clast provenance, such as Igneous clasts from the Cheviots or Scotland and Carboniferous Limestone from Northumberland. Therefore the ice route that deposited Skipsea B Till is most likely from a north-east to south-west ice surging that flowed onshore to the Holderness coast from the North Sea ice lobe.

Acknowledgments

I would like to thank my dissertation supervisor, Dr Colin Whiteman, for all the help and support in my project from the initial fieldwork to the final stages. I am also grateful to all other University of Brighton staff, my lecturers and lab technicians, who were always available to offer support in all apsects of my time at University.

Lastly, I would like to thank my parents Jill and Peter Beaumont and my brother Edward Beaumont for all their help during the fieldwork.

References

Bridglandd, D.R. (1986) Discussion of procedures and recommendations. In:: Bridglandd, D.R. (Ed) Clast Lithological Analysis.. Quaternary Research Associationn.Cambridge. UK..

Gale, S.J. and Hoare, P.G. (1992) Quaternary Sediments: Petrographic Methods for the Study of Unlithified Rocks. International Books Distributors. Dehra Dun. India.

Hubbard, B. and Glasser, N. (2005) Field Techniques in Glaciology and Glacial Geomorphology. Wiley. Chichester. UK.

Ma,Z. Merkus, H.G. de Smet, J.G.A.E. Heffels, C. Scarlett, B. (2000) New developments in particle characterisations by laser diffraction: size and shape. Powder Technology, 111 (1-2): 66-78.

Madgett, P. A. and Catt, J. A. (1978) Petrography, stratigraphy and weathering of Late Pleistocene tills in East Yorkshire, Lincolnshire and North Norfolk. Proceedings of the Yorkshire Geological Society, 42: 55-108.

* Editors' Comment - Whilst the methodologies used to process the samples are scientifically sound, the editors believe that the groupings used for the clast identification are quite broad. For instance the Old Red Sandstone and New Red Sandstone are in the same group. The grey, black and red flints are not differentiated. These rocks have different original sources and are likely to have travelled on different glacial routes. There is no mention of Scandinavian erratics, though they may have been a bias in the 4 to 8 cm clast fraction. This could affect the conclusions about ice stream directions.

If you enjoy reading the article please conside joining the Society or sending a donation.

HOME