High-magnitude flooding across Britain since AD 1750

Macdonald, Neil; Sangster, Heather

doi:https://doi.org/10.5194/hess-21-1631-2017

Articles | Volume 21, issue 3

https://doi.org/10.5194/hess-21-1631-2017

© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License.

Special issue:

Floods and their changes in historical times – a European...

https://doi.org/10.5194/hess-21-1631-2017

© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 21, issue 3

Research article

|

20 Mar 2017

Research article |

| 20 Mar 2017

High-magnitude flooding across Britain since AD 1750

Neil Macdonald and Heather Sangster

Download

Final revised paper (published on 20 Mar 2017)
Preprint (discussion started on 08 Sep 2014)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC C3955: 'Comments to the paper', Anonymous Referee #1, 21 Sep 2014
- AC C5211: 'Comments in relation to reviewer 1', Neil Macdonald, 19 Nov 2014
RC C4166: 'Hess-2014-330', Anonymous Referee #2, 01 Oct 2014
- AC C5216: 'Response to Anonymous Reviewer 2', Neil Macdonald, 19 Nov 2014
RC C4513: 'Paper review', Anonymous Referee #3, 16 Oct 2014
- AC C6106: 'Response to Anonymous Reviewer 3', Neil Macdonald, 16 Jan 2015
RC C4957: 'Paper review', Anonymous Referee #4, 06 Nov 2014
- AC C6104: 'Response to Anonymous Reviewer 4', Neil Macdonald, 16 Jan 2015

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

ED: Reconsider after major revisions (05 Feb 2015) by Andrea Kiss

1) Due to the very low data density of the period prior to 1750 (also raised as basic source of problems by the reviewers), insufficient in quantity and quality for the type of analysis the author carried out (and the conclusions drawn), I suggest to leave out this part, and concentrate the paper (and the analysis) on the period after 1750 (but keeping the spatial dimension).
2) Please, provide a basic overview on the most important anthropogenic changes in the catchment areas and the potential human impact on the studied rivers. It would be also important to know more (e.g. some paragraphs) about the source materials and how the indices were constructed (also provide more information how the national indices were made).
3) Concerning the suggestions of Reviewer No 1, the author accepted them and suggested further solutions - these solutions and amendments will sufficiently improve the quality of the paper.
4) As suggested by Reviewer No. 2, please, restrict the analysis to the British situation. A number of useful further suggestions of Reviewer No 2 were also fully accepted (and further suggested to be revised) by the author.
5) Following further recommendations of Reviewer No. 3, please, consider to make more emphasis on revealing quantitative relationship between the studied flood series and the factors compared (i.e. if correlations are mentioned than we presume statistically defined relationship etc).
6) Following the suggestions of Reviewer No 4, I further suggest the author to provide a systematic and more specific description/overview of the methods applied in the paper. Moreover, Reviewer No. 4 had specific questions and useful/reasonable suggestions (e.g. concerning the applied statistics) that should be taken into consideration (as the author himself also suggested) in the next version of the paper (e.g. defining flood rich period).

The paper will be reconsidered (i.e. reviewed) after major revisions.

Hide

AR by Neil Macdonald on behalf of the Authors (20 Jul 2016) Author's response Manuscript

ED: Referee Nomination & Report Request started (25 Jul 2016) by Andrea Kiss

RR by Anonymous Referee #4 (06 Aug 2016)

RR by Anonymous Referee #3 (01 Oct 2016)

Suggestions for revision or reasons for rejection

The corrections that have been made in the revised manuscript, largely consist of the addition of detail on a per catchment basis. The author insists that this paper merely attempts to use a broad-brush approach, but in a field – historical floods on the British Isles over the last 800 years – where very little has been published so far, a ‘rapid communication’ approach is premature. My main concern with the paper remains that the conclusions drawn regarding forcing mechanisms on the millennial scale are simply not supported by the data. The quality and validity of the pre-1750 data is insufficient, as described at length in my previous review. For example, medieval and early modern data are displayed in Fig. 2 for six out of twelve rivers, but for two out of these six flood histories, the rivers Tay and Trent, no contemporary data are provided at all in the original papers (Macdonald, Werritty et al. 2006; Werritty, Paine et al. 2006; Macdonald 2013); nor in the information presented here. The floods pre-dating c. 1750 for these rivers are taken entirely from ‘compilations, nineteenth or early twentieth century water board publications or nineteenth century local histories, types of work that are frequently unaware of historians’ methodology and therefore abound in errors concerning dating, doubling of events, overlooking of events and misallocating evidence’; information given in such works is a starting point for research but requires verification. For the problem with the flood histories of the rivers Tyne and Tweed (the comparatively well documented floods around 1300 are now included in the Tweed series, however, so are the sixteenth and seventeenth-century floods referenced only in non-contemporary works) see the previous review. If historical methods are not employed in such a verification process, then the error-risk remains very high.
As an illustration for the confusion that ensues when original contemporary sources are not consulted, the early flood history of the River Severn given in lines 494-505 is a good example. Three early floods from ‘reliable data’ are noted: the ‘Duke of Buckingham’s Great Water’ in October 1484, a flood in 1258 and a series of unspecified inundations 1377-81. The source given is the ‘Annals of Tewkesbury’ edited in Luard, H.R., Annales Monastici, vol. 1, London 1864. The inundations for 1258 are indeed referenced in those annals, and the year is also correct. However, the annals end in 1263 and neither the floods around 1380 nor the high waters in 1484 can be found in this text. The reference to these two late medieval floods is actually taken from Fosbrooke, T.D., An original history of the City of Gloucester, London 1819. (cited incorrectly as Fosbroke 1819 in line 505). Fosbrooke’s ‘History of Gloucester’ is exactly one of the generally unreliable antiquarian works of the earlier nineteenth century, that always need verification. Fosbrooke might be an honourable exception, his reference to the flood 1258 is after all correctly dated (p. 42). When mentioning the inundations and further problems in the years 1377-81, Fosbrooke even cites a charter (p. 175), which raises confidence and would make verification possible. Fosbrooke also tells the story about the ‘Duke of Buckingham’s Great Water’ in October 1484, except that he cites this event for the year ‘1483’, not ‘1484’ as in this paper. Indeed, the flood and the associated rebellion against Richard III. took place in 1483 and in the November of that year the relevant Duke of Buckingham was executed for his failed attempt to overthrow the King. The change from the correct ‘1483’ of Fosbrooke into the ‘1484’ in the paper is an adequate example of medieval and post-medieval transcription errors. A dating error by one year seems negligible at first glance, but in a study concerned with the drivers of floods it is essential to have precise dating, for being able to connect the floods to possible large-scale forcings such as volcanic eruptions, that over England certainly have the power to turn a summer half year rainy and cause flooding. The paper refers to the floods 1258, c. 1380 and 1484 [sic] as ‘clearly documented in reliable sources’ (line 501) and so distinguishes them from a set of floods for which ‘original sources have not been located’ (line 504-505). For this second set of floods ‘such [threshold] estimates for these events are not included’. However, no floods of either set are actually shown in Fig. 2, does this imply that none of them qualifies as ‘events that exceed the 0.9 percentile (based on the instrumental record; river discharges are given as’ cumecs)?
To alleviate the lack of recorded floods c. 1300 to 1500 and to show that this is not a lack of sources or search for records, but a result of natural forcing, a connection between peat water levels and river flooding is implied (line 795-797). The peat record does indeed not display any wetness in those 150 years, nor does it display much else until c. 1800. What the peat record does definitely not show is any connection to solar magnetism. There is no reaction in the peat record to the increased levels of solar magnetism around 1375 or 1600 or to the sustained increase after c. 1750. During the high level of twentieth-century solar magnetic activity, when flooding increased, the peat water levels were low. Since the peat reflects neither the high level of flooding in the recent decades nor the rather high level of solar magnetism, it is not clear how it could be connected to these factors in the Middle Ages and serve as an independent indicator of flooding on the British Isles for this period. The argument brought forward in the first review, that the lack of recorded floods in the late Middle Ages, a time when the historiographic activity in England was dwindling and the population was reaching a sad nadir, is less to do with the actual occurrence of floods than with the cultural, demographic and socio-economic environment, remains valid.
More context is now provided for the catchment areas and the changes over time. This background is generally helpful, although it is mostly limited to the recent centuries. However, whereas for London and the Thames the construction of the Old London Bridge (medieval) in 1209 is correctly described and the ‘weir beneath the arches’ is referenced, an analysis of the impact of the massive Old London Bridge and the weir on river flow is missing. In fact the bridge caused a noticeable restriction on flow and the pooling of water at the bridge made the freezing of the Thames more likely and therefore had an influence on the frequency of Frost Fairs. The restriction of flow, however, would also likely raise the risk of river floods and increase their severity. Old London Bridge was taken down in 1831 and maybe the noticeable sequence of severe floods in the 90 years preceding this removal (8), which is not matched in the subsequent decades, owes something to the influence of Old London Bridge on river flow.
Simply dismissing flood seasonality as an avenue for further investigation neglects the fact that the seasonality should be an intrinsic component of any analysis that attempts to link flood occurrence to large-scale forcing mechanisms. The conclusions relating to solar forcing can not be made without a thorough appraisal of the seasonality of flood events.
In summary the data simply do not support an 800-year comparative analysis of flood events across British catchments; data density and reliability are only sufficient from c. 1750 onwards. The conclusion ‘The principal finding of this work is that of the strong correlation between flood-rich phases and solar magnetic activity, indicating a clear driver for flooding patters across Britain’ (line 963-964) is not supported by the data in the period before 1750, but also not afterwards. The solar magnetic activity maximum around 1375 is not reflect in the flood data, and the picking up of solar magnetism from c. 1720 onwards is only matched with a national flood rich period in the 1770s (this seems to be based largely on the ‘Eastern and Southern British regions, both containing the Thames data). The flood peak around 1600, a core period of the Little Ice Age, might or might not be connected to solar magnetism, but it is certainly connected to increased record keeping starting in the late sixteenth century and probably growing until the onset of Civil War.
Finally incoherent sentences, ambiguous sentence structures, as well as grammar and spelling mistakes abound in the manuscript. Examples of the spelling mistakes are: line 481 ‘the instillation of the weir’, line 497 ‘Tewksbury’, line 578 ‘whist’, line 964 ‘flooding patters’. For the grammar and sentence structure problems the section on London and the Thames serves as an indication (line 544-582): ‘By the late seventeenth century the city of London was starting to develop its quays and docks along the banks and as such confine the river, as evident in Morgan’s map of the Whole of London (1682); by John Rocques map of 1746, it is evident that the channel is increasingly confined, particularly around London Bridge. The map of Bacon (1868), clearly illustrates the development of the Embankment reach with further constriction of the river and extensive development and expansion of the city of London both up and downstream of the bridge area, which influenced the channel hydraulics, with the constriction of the channel resulted in channel deepening. [...] An earlier event in 1555 appears to have resulted from heavy rains, but also possible tides, though unclear. [...]the 1774 flood mark appears to the earliest, on the wall at Radnor Gardens, Twickenham, with G.B Laffan giving the 1774 level as being 0.85m higher than that of 1894, though recognising a tidal influence was present, though at Windsor 1894 was considered higher (Symons and Chatterton, 1895), as such a notional discharge of 850 m3s-1, is used. [...] Channel changes, river modification and uncertainties involved in estimating discharges makes the ranking of events challenging, as such these are notional magnitudes based on apparent ranking of events for the area around Kingston upon Thames and should only be used as indicative.’

Hide

ED: Reconsider after major revisions (01 Oct 2016) by Andrea Kiss

AR by Anna Wenzel on behalf of the Authors (27 Jan 2017) Author's response Manuscript

ED: Publish subject to technical corrections (13 Feb 2017) by Andrea Kiss

AR by Neil Macdonald on behalf of the Authors (14 Feb 2017) Author's response Manuscript

Short summary

We use historical records to extend current understanding of flood risk, examining past spatial and temporal variability and ask are the perceived high-magnitude flood events witnessed in recent years really unprecedented? We identify that there are statistically significant relationships between the British flood index and climatic drivers, whereby the largest floods often transcend single catchments affecting regions and that the current flood-rich period is not unprecedented.