The voyage of the Challenger (1872-6)
Meteorological observations from The Challenger Expedition of 1872-6 were available from the book "Report on the scientific results of the voyage of H.M.S. Challenger during the years 1873-76. Narrative - Vol. 2, by C Wyville Thomson and John Murray", a scanned copy of which is provided by the Scripps Institute library. The pressure and temperature observations have been digitised and converted into IMMA records.
- Data as digitised (Excel spreadsheet, Tab delimited text)
- Script to convert to IMMA format (uses this Perl module).
- The logs contain more weather observations than position records, so positions for some obs are generated using linear interpolation by this script.
- IMMA records.
Problems with the digitisation
- It is not clear how ship's time relates to GMT (what time-zone were they keeping on board)? The observation times have been converted to GMT assuming a time zone based on the exact longitude of the observation.
- When the ship is in port or anchored, the log gives a place name, rather than a latitude and longitude. Positions (lat. and long.) have been found for most of the named places, but not all (where are Port Hardy, Api Island, Dobbo Harbour, Ki Doulan, Banda Harbour, Zebu, Port Isabella, Oosima Harbour, Grey Harbour, Tom Bay, Isthmus Harbour and Port Churruca?). Also some places have been given approximate positions - all the 10 named places around Kerguelen Island have been assigned the same position, for instance.
Have these observations been digitised before?
To see if the challenger observations had already been digitised, all observations from ICOADS 2.3 that are close to one of the challenger observations (within 5 degrees of latitude and longitude and 2 days in time) were extracted using this script. These observations were displayed, alongside the new observations, in Google Earth (see sidebar for KML files, see also this video comparing the old and new observations).
This comparison shows that the Challenger observations in the Atlantic sector are not in ICOADS 2.3, but elsewhere things are more complicated. It looks as if the data has been digitised several times - in each case a different subset of the observations for some section of the voyage was done:
- The voyage from Indonesia to South America has been done at least twice, with observation IDs "CHALLENG" (SST only) and 03716 (dry-bulb and sea temperatures only).
- The voyage from Australia to Indonesia has been done with observation ID 03484 (dry-bulb and sea temperatures only).
- The voyage in the Southern ocean has been done, but only between 40 and 50 South (100-120 East). Ship ID 103484: dry-bulb temperature only.
Despite all this previous work, less than half of the voyage has been previously digitised; and even where previous observations are available they are less frequent than they should be, and don't contain all the observed variables (never any pressure observations, for example).