Monitoring the movement of St. Catherine's Point Lighthouse

Monitoring the Movement of St. Catherine's Point Lighthouse Using the Ordnance Survey GPS Network by Jared L. Ware.

Jared Ware, an engineer officer in the United States Army, is a student at the Royal School of Military Survey in Hermitage, England. He is studying for an MSc in Defence Geographic Information from Cranfield University. His research project was conducted for the Ordnance Survey of Great Britain.

In the autumn of 2000, the active network station at Saint Catherine's Point (SCP1) on the Isle of Wight began to gradually shift south towards the English Channel from its original coordinate position. However, no movement was evident in any other GPS active network station based upon the GPS data processed at Ordnance Survey. The active network consists of 30 continuously operating GPS reference stations located throughout Great Britain. SCP1 is one of a network of 9 stations owned and operated by the General Lighthouse Authority which contribute GPS data to the Ordnance Survey's active network. These stations are permanently installed, precisely coordinated active GPS stations. The extensive rainfall in the autumn and winter of 2000 was raised as a factor contributing to the perceived movement, potentially having affected the soil properties at the base of the lighthouse. Given this situation SCP1 was taken out of the active network until an investigation into its coordinate shift could be conducted. The lighthouse and the GPS antenna are shown below in Figure 1.

Figure 1: St. Catherine's Point Lighthouse (SCP1) with GPS Antenna

The investigation looked into the GPS, rainfall, and soil information available during the time interval in which the lighthouse shifted south. The aim of this project was to investigate the sources of data from St. Catherine's Point to determine the magnitude of movement of the lighthouse. An additional aim was to determine if the lighthouse had stopped moving so that SCP1 could be returned to operation in the active GPS network with its new coordinate position. The objective was to provide information that may help develop potential solutions to mitigate the effects of the lighthouse's movement during periods of extensive rainfall. The focus of the project was to determine the sources of error (if any) in the network, provide information that determines (1) the extent to which the lighthouse has shifted south and (2) if the lighthouse has ceased moving, and to establish new coordinates for the SCP1 station based upon coordinate corrections. The results of the investigation were based upon the GPS data provided by the Ordnance Survey and the data gathered in the field by the author.

It was determined that between September 2000 and March 2001 the lighthouse moved in a southern direction and produced GPS residuals outside of the accepted 95% confidence level. There is a strong correlation between the amount of rainfall experienced at Saint Catherine's Point and the rate of change of movement south of the lighthouse. Analysis indicated St. Catherine's Lighthouse moved approximately 10 centimetres south over a seven months period. It was also determined that SCP1 ceased moving in the second quarter of 2001. A future cumulative rainfall total equivalent to that of the autumn of 2000 coupled with a short duration of intense precipitation could potentially produce an additional southward shift of up to 2 centimetres.

The GPS data from the lighthouse revealed that a significant southern shift in the lighthouse began in October 2000 and gradually continued until late March 2001. This can be seen in Figure 2 below, where the movements are determined in centimetres. Note that no further movement has occurred since April 2001.

Figure 2: SCP1's Southern Shift (Outlined in Red)

The comparison of epochs were essential to determine whether the shifts at St. Catherine's Point were due to network deformation, or whether the lighthouse actually shifted due to another factor. It was necessary to compare the stations in the network at times of normal precipitation, no precipitation, and during period of heavy rainfall to determine if rainfall had an overarching effect on the entire network;

The base epoch, 2000.335, was chosen because this was a period when the residuals in the network were small and there was little rainfall occurring during this period. Epoch 2000.779 was chosen because it was the day on which the greatest amount of rainfall occurred at St. Catherine's Point. Epochs 2001.089 and 2001.122 were chosen because this was a period during the first two weeks of February where the lighthouse experienced almost 1 centimetre of movement to the south. Epoch 20001.251 was chosen because the cumulative rainfall total decreased significantly, and the overall movement of the lighthouse appears to have subsided. Epochs on the first day of each month were chosen in an effort to match monthly rainfall totals with the movement of the lighthouse.

St. Catherine's Point experienced 940.8 millimetres of rainfall from 1 January 2000 to 31 December 2000. From 1 May 2000 to 1 April 2001 the area received 1067.7 millimetres of rainfall, which is over 200 millimetres more rainfall than experienced in a normal year. This indicates that St. Catherine's Point was affected by a significant amount of rainfall, which contributed to the shift in the lighthouse with respect to the soil shear at its base. The lighthouse began to shift southwards immediately after the largest daily rainfall total of the year, which was approximately 38 millimetres in October 2000. Similarly, the greater the daily rainfall totals for amounts over 10 millimetres, the steeper the rate of change of the shift of the lighthouse. Figure 3 shows the shift data (in centimetres to the left) superimposed with the rainfall data (in millimetres to the right).

Figure 3: SCP1's Southern Shift and Rainfall Data Superimposed by Epoch

The analysis of the overall network indicates that the residuals fall within the accepted range of 3s at a 99% confidence level. The entire network achieved millimetric precision throughout the entire data period analysed. In general, the eastings and northings for each station tend to be twice as precise as the up component. The only station in the network to experience a significant coordinate shift is SCP1, and this is mainly in the north component. Note that additional sources of error such as receiver, atmospheric, and multipath were investigated and determined not to have adversely impacted SCP1 or any other station in the national active network.

In comparing various epochs it is evident that SCP1 has moved due to soil shear and not because of deformations in the network. When Epoch 2001.335 is compared with Epoch 2001.497, the change between the adjusted coordinate values for SCP1 is approximately 2 centimetres in east, 9 centimetres in north, and 4 centimetres in up. When comparing both epochs for the active station at Ordnance Survey Headquarters in Southampton, there is no coordinate change at all. The average coordinate change for any one station in the network is 3 millimetres in east, 3.5 millimetres in north, and 1 centimetre in up. The active network was determined to have a strong geometry and does not suffer from horizontal or vertical deformation.

The station at St. Catherine's Point has experienced over 9 centimetres of shift southwards since October 2000. The station appears to have stopped moving, and will not move until another significant period of rainfall occurs. This equates to approximately 40mm of rainfall in a four-day period based upon a trend analysis model developed during the research of St. Catherine's Point. It was determined that short periods of intense rainfall are directly proportional to the rate of change of movement of the lighthouse. This finding was consistent with other groundwater level and horizontal deformation studies conducted in similar environments to that of St. Catherine's Point. It was determined that no structural damage occurred within the structure to cause movement, and no twisting or bending had occurred with the GPS antenna to cause movement. This was confirmed after a visual inspection of the antenna in September 2001, checking maintenance inspection records at Ordnance Survey, and analysing the statistical results of the GPS data.

Figure 4: Southern Base of St. Catherine's Point Lighthouse