the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Reanalyses of Maskelyne’s Tidal Data at St. Helena in 1761
- National Oceanography Centre, Joseph Proudman Building, 6 Brownlow Street, Liverpool L3 5DA, United Kingdom
- National Oceanography Centre, Joseph Proudman Building, 6 Brownlow Street, Liverpool L3 5DA, United Kingdom
Abstract. The construction of an electronic data set of the tidal measurements made at St. Helena in 1761 by Nevil Maskelyne is described. These data were first analysed by Cartwright (1971, 1972) in papers which have importance within studies of changing tides. However, Cartwright’s data files were never archived for the benefit of other researchers, demonstrating that ‘old data’ at risk can sometimes take the form of electronic rather than paper records. In the present paper, the newly digitised Maskelyne data have been reanalysed by several techniques in order to obtain an updated impression of whether the tide has changed at that location in over two and a half centuries. Our main conclusion, consistent with that of Cartwright, is that the major tidal constituent (M2) has changed little. However, the results of the various techniques demonstrate how difficult it is to obtain reliable conclusions for the smaller constituents.
Philip Leslie Woodworth and John McLauchlan Vassie
Status: open (until 30 Aug 2022)
-
EC1: 'Comment on essd-2022-218', Giuseppe M.R. Manzella, 08 Jul 2022
reply
The paper proposes an analysis of historical data that can help to understand changes in the characteristics of the tides over the centuries (if changes are occurring). The critical analysis of historical data is very well done and the methods used to verify the reliability of the data and their current value are exhaustive.
There are some missing details on 'methodologies and technologies' that may be added to satisfy the reader's curiosity.
In the introduction one could have expected a brief presentation of why tidal characteristics can change (lines 30 - 33) and whether the reasons for the changes are measurable over a period of a few hundred years.
Line 78: … agreed consistently to better than half an inch (12 mm). Considering that the tidal movements are of lesser amplitude than those due to wave movements and other high frequency coastal movements, I wonder that short waves, spray, shower … may have influenced the measurement reading. May be the pole was in a sheltered place. Are there any indications in the original manuscript or in other works by Maskelyne?
Line 112: … times set down are exact to the minute It would be useful to know the type of clock used exactly to the minute and what is the meaning of 'apparent time' in figure 1.
Line 492: … several measurements that Maskelyne flagged as suspect Is it possible to add more details on how Maskelyne flagged ‘suspect’ data?
It is a pity that some documents are available only in English libraries. Their on line availability would have fascinated many readers to the subjects of historical oceanography.
However, the article represents an excellent example of revisiting and processing historical data
-
AC1: 'Reply on EC1', Philip Woodworth, 11 Jul 2022
reply
Many thanks for the kind remarks under EC1 and RC1. We will certainly attend to these in the next version. Just one thing for now - local apparent time is the time one measures with a sun dial as opposed to local mean time which one measures with a clock - the difference, which occurs because the sun is not always on the equator and can be as large as 15 minutes at certain times of the year, is called the 'equation of time' - there is a good Wikipedia page about it. Best wishes.
-
AC1: 'Reply on EC1', Philip Woodworth, 11 Jul 2022
reply
-
RC1: 'Comment on essd-2022-218', Giuseppe M.R. Manzella, 09 Jul 2022
reply
The comments have been sent as Chief Editor (EC1) (answer to them)
I am directly interested to historical oceanography and to the availability of original manuscripts. Most of cited references are in libraries located in UK. It would be interesting to have an open access to the old publications.
-
AC2: 'Reply on RC1', Philip Woodworth, 11 Jul 2022
reply
-
RC2: 'Reply on RC1', Laurent Testut, 02 Aug 2022
reply
This new dataset is undoubtedly of great interest especially in the recent context where ocean tide evolution at the coast but also globally is in discussion in the ocean tidal community. Datasets back to the eighteen century is very rare especially in the Southern Hemisphere making the release of this new old-dataset even more valuable. The sea level data are analysed using 3 different methods which gives confidence on the robustness of the main conclusion of no significant semi-diurnal tide evolution during the past 2 and a half century. This paper also offers nice recipes to help the tidal scientist to process their old short-term datasets using these different methods. Although I mainly agree with the way the data is processed I have some questions listed below :
(1)The dataset used to estimate the modern tidal constituents (TC), is the sub-surface pressure (SSP) recorded at Saint-Helena between oct-1995 and oct-1996. There is a more recent dataset at the same location, with direct observations of sea level from a radar made between jan-2011 and feb-2013. This dataset is available at University of Hawaii Sea Level Center (as the one used by the authors from 95-96). What drives the choice of analysing the SSP dataset instead of the more recent one ? The recent radar data gives at the mm level the same values as the (S2,S1) modified TC used by the authors without the need to modified S1 and S2.
(2) In the second approach, they interpolate the Maskelyne data at 1 minute using a tidal prediction. As they authors quoted themselves, there is a clear danger about this method due to a leak of information from the prediction to the observation. Anyway, to make the tidal prediction they use the TC obtained with another dataset (called STHL2 obtained from SSP between Nov-1993 do Feb-1995). As they quote the STHL2 TC is very close to STHL4.X. I don't understand the reason behind using another set of TC for this second approach ? Using the prediction from the STHL4.X would have give almost the same results no ?
(3) The standard deviation of the 478 historical data (28.41 cm) is higher than in any of the recent 2 months period available (from 1995 to 2013 where maximum value is 27.1 cm and the mean standard deviation of every 2 months period is close to 25.5 cm). Does the authors have any explanation of this slightly larger standard deviation. Can it comes from the sampling of the Maskelyne data ? Can it be a slight scale error on the tide staff build by Maskelyne or a small difference in the conversion between inch (used by Maskelyne) and meter (change in definition between 1761 and the official conversion).
Minor comments :
I recommend to add in the text or in a separate table the list of constituents used in the tidal prediction (for the STHL2 and STHL4.X), this will allow the futur users of the Maskelyne dataset to reproduce the results of this paper or at least to test their own harmonic analysis code.A typo has been found in the Maskelyne data file provided by the authors on line 121 (date 1761-11-16 6:44 PM) a decimal '.' is missing, one should read 11.3 instead of 11 3. This is confirmed in the original Maskelyne report.
-
AC3: 'Reply on RC2', Philip Woodworth, 09 Aug 2022
reply
Many thanks for these remarks which are certainly all reasonable. We shall take them into acccount in the updated version.
-
AC3: 'Reply on RC2', Philip Woodworth, 09 Aug 2022
reply
-
AC2: 'Reply on RC1', Philip Woodworth, 11 Jul 2022
reply
Philip Leslie Woodworth and John McLauchlan Vassie
Data sets
Maskelyne sea level observations from the Island of St. Helena in 1761 Woodworth, P. L. and Vassie, J. M. https://doi.org/10.5285/e0f2b6ea-d11d-3102-e053-6c86abc073ab
Philip Leslie Woodworth and John McLauchlan Vassie
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 174 | 49 | 16 | 239 | 4 | 2 |
- HTML: 174
- PDF: 49
- XML: 16
- Total: 239
- BibTeX: 4
- EndNote: 2
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1