Gloss-Africa website

EXISTING TIDE GAUGE DATA SETS (PSMSL AND UHSLC) AND 'DATA ARCHAEOLOGY'

Even though the current sea level network in Africa is not as extensive as one would like, there is a considerable amount of existing information which can be used for research and education.

Tide Gauge Data Sets

The two main international data sets of sea level information are those of the PSMSL and the University of Hawaii Sea Level Center (UHSLC). The PSMSL is the global databank for long term sea level information from tide gauges and contains over 47000 station-years of monthly and annual mean sea level values. The UHSLC evolved from the former Tropical Ocean - Global Atmosphere (TOGA) databank, was one of the two World Ocean Circulation Experiment (WOCE) databanks, and is now the GLOSS/WOCE Fast Sea Level Centre.

Figures 2and figure 3 show the distribution of data held by the PSMSL and UHSLC (its Joint Archive for Sea Level). Detailed listings of data holdings can be found at http://www.pol.ac.uk/psmsl/pub/nucat.dat and http://uhslc.soest.hawaii.edu/uhslc/data.html) respectively. All data are freely available either via the web or on CD-ROM for people without web access. Note that these two centres are by no means the only relevant ones. For a list of web sites of the other centers, see the PSMSL web page http://www.pol.ac.uk/psmsl/.

Figure 3 also indicates African stations capable of delivering data in 'fast' mode (which at present means within a few weeks) via telephone or satellite or other fast means to the GLOSS Fast Centre at UH. At the present time, the UHSLC maintains a number of near-real time sea level stations on the African continent and nearby islands. The network (as of August 2001) consists of five stations. Handar encoders are installed at Lamu and Mombasa, Kenya, and Zanzibar, Tanzania. Mombasa is presently down but is expected to be serviced by late August 2001. Aquatrak acoustic gauges are operated at Dakar, Senegal and Palmeira, Cape Verde. The latter is down but will be serviced by September 2001.

A considerable amount of tide gauge data are likely to be required in quasi-real time in future for assimilation into and validation of deep ocean and coastal (e.g. storm surge) models. The definition of 'fast' will depend on the application. For example, for storm surge work, 'fast' means within an hour or so. For deep ocean modelling, 'fast' could mean within a couple of days or a week.
Figure 4 shows some representative time series of annual mean sea level from Africa indicating two of the main features of the data set: long time series discontinued (or in difficulty at present) but encouraging efforts to undertake new recordings, especially in East Africa.

Data Archaeology

At the 6th and 7th sessions of the IOC Global Sea Level Observing System (GLOSS) Group of Experts in 1999 and 2001, and at the 16th session of the IOC Committee on International Oceanographic Data and Information Exchange (IODE) in 2000, the topic of data rescue of historic tide gauge data in non-computer form (charts, tabulations etc.) was discussed. These mostly paper-based data sets are of potential great value to the sea level community in a range of applications, of which the most obvious is the extension of existing sea level time series as far back as possible in order to understand more completely the timescales of sea level change. It is essential that, for the priority sites at least (e.g. GCN sites), the charts are digitized to provide hourly (or similar) sea level values in computer form, with the resulting data sets made available to the international community and, of course, to the chart-owners. Chart digitizing has historically been a labour-intensive activity. However, the availability of new scanner-based packages seems to provide the possibility for speeding up the work considerably.

In order to obtain a first estimate of how many such charts exist from GLOSS and other priority sites, and to make an estimate of the total effort required for their digitization, a GLOSS/IODE questionnaire has been circulated to the combined mailing lists. (Additional copies may be obtained via psmsl@pol.ac.uk). We are certain that such historic data exist in Africa (e.g., data sets held at SHOM in France), and that a 'data archaeology' exercise for African sites following on from the questionnaire would a major benefit to both GLOSS and GOOS in the continent.

This site is designed and maintained by the Gloss Africa Webteam. For more in formation contact the Webmaster


	Stations	News	Reports	Equipments	Data and predictions	Contacts	Home
EXISTING TIDE GAUGE DATA SETS (PSMSL AND UHSLC) AND 'DATA ARCHAEOLOGY' Even though the current sea level network in Africa is not as extensive as one would like, there is a considerable amount of existing information which can be used for research and education. Tide Gauge Data Sets The two main international data sets of sea level information are those of the PSMSL and the University of Hawaii Sea Level Center (UHSLC). The PSMSL is the global databank for long term sea level information from tide gauges and contains over 47000 station-years of monthly and annual mean sea level values. The UHSLC evolved from the former Tropical Ocean - Global Atmosphere (TOGA) databank, was one of the two World Ocean Circulation Experiment (WOCE) databanks, and is now the GLOSS/WOCE Fast Sea Level Centre. Figures 2and figure 3 show the distribution of data held by the PSMSL and UHSLC (its Joint Archive for Sea Level). Detailed listings of data holdings can be found at http://www.pol.ac.uk/psmsl/pub/nucat.dat and http://uhslc.soest.hawaii.edu/uhslc/data.html) respectively. All data are freely available either via the web or on CD-ROM for people without web access. Note that these two centres are by no means the only relevant ones. For a list of web sites of the other centers, see the PSMSL web page http://www.pol.ac.uk/psmsl/. Figure 3 also indicates African stations capable of delivering data in 'fast' mode (which at present means within a few weeks) via telephone or satellite or other fast means to the GLOSS Fast Centre at UH. At the present time, the UHSLC maintains a number of near-real time sea level stations on the African continent and nearby islands. The network (as of August 2001) consists of five stations. Handar encoders are installed at Lamu and Mombasa, Kenya, and Zanzibar, Tanzania. Mombasa is presently down but is expected to be serviced by late August 2001. Aquatrak acoustic gauges are operated at Dakar, Senegal and Palmeira, Cape Verde. The latter is down but will be serviced by September 2001. A considerable amount of tide gauge data are likely to be required in quasi-real time in future for assimilation into and validation of deep ocean and coastal (e.g. storm surge) models. The definition of 'fast' will depend on the application. For example, for storm surge work, 'fast' means within an hour or so. For deep ocean modelling, 'fast' could mean within a couple of days or a week. Figure 4 shows some representative time series of annual mean sea level from Africa indicating two of the main features of the data set: long time series discontinued (or in difficulty at present) but encouraging efforts to undertake new recordings, especially in East Africa. Data Archaeology At the 6th and 7th sessions of the IOC Global Sea Level Observing System (GLOSS) Group of Experts in 1999 and 2001, and at the 16th session of the IOC Committee on International Oceanographic Data and Information Exchange (IODE) in 2000, the topic of data rescue of historic tide gauge data in non-computer form (charts, tabulations etc.) was discussed. These mostly paper-based data sets are of potential great value to the sea level community in a range of applications, of which the most obvious is the extension of existing sea level time series as far back as possible in order to understand more completely the timescales of sea level change. It is essential that, for the priority sites at least (e.g. GCN sites), the charts are digitized to provide hourly (or similar) sea level values in computer form, with the resulting data sets made available to the international community and, of course, to the chart-owners. Chart digitizing has historically been a labour-intensive activity. However, the availability of new scanner-based packages seems to provide the possibility for speeding up the work considerably. In order to obtain a first estimate of how many such charts exist from GLOSS and other priority sites, and to make an estimate of the total effort required for their digitization, a GLOSS/IODE questionnaire has been circulated to the combined mailing lists. (Additional copies may be obtained via psmsl@pol.ac.uk). We are certain that such historic data exist in Africa (e.g., data sets held at SHOM in France), and that a 'data archaeology' exercise for African sites following on from the questionnaire would a major benefit to both GLOSS and GOOS in the continent. This site is designed and maintained by the Gloss Africa Webteam. For more in formation contact the Webmaster