"Ario 2000" Expedition Final Report
Picos de Europa, Spain
3rd July - 22nd August 2000
Final Report contents
The shaft-bashing guide has been a stalwart feature of OUCC expeditions to the Picos for something like a decade and has served well under Gavin's custody. It is a list of all the features thought possibly likely to lead to greater things, or on which time has been frittered away, or occasionally just thought of as worthy of inclusion in a comprehensive list of karst phenomena. The contents have been gleaned largely from expedition reports and logbooks. It appears on expedition as a neatly printed document which provides something to sit on, to flick through, to swat flies with, or affect to study, but is not too good as a quick and easy shaft-bashing guide. It is now maintained on a computer and has been available over the web as html produced from its current source format, LaTeX. It ought to be a guide as to where new sites may be, where to go to prospect, or where to knock off several of the prospective sites at one go. Already portable and field updateable, it needs maps and flexible means of filtering, indexing and cross-referencing.
Entries are usually given a tag id, sometimes a name, and are grouped by area. They have some form of location and descriptive text which varies from the stark to comic; laconic to obsessive; scientifically objective to impressionistically literary. Often, the location is not clear - bearings may be given to sites that may or may not be identifiable, errors may have been introduced on entry or transcription, or the context of the original entry may have been entirely lost. We needed absolute locations.
In April 2000 I noticed on a trip to Skye that the normally jittery location readings given by my GPS had become stable and accurate. It was soon confirmed that the US Government had provided a gift by turning off selective availability. Suddenly, the use of GPS coordinates to improve the shaft-bashing guide became a viable option. In brief, satellites broadcast their identity and position together with a precise time and a receiver picks up as many as possible of these transmissions (so weak that even your body may block them) and calculates a fix from the timing differences it observes. The reported position is given in one of many different possible coordinate systems, although most of these will be obtained by more or less approximate conversions from a base coordinate system for the receiver in question (probably WGS84).
In the field we had a variety of units available to us. Most readings were taken from my Pioneer and from the Landrover's Garmin. There were occasions when considerable drift was noted and this seemed to coincide with low satellite availability, but generally repeatability appeared to be good and the unit's own estimate of accuracy was usually about 7m. When recorded with or converted to the map datum the positions coincided with dead reckoning and relative locations appeared correct. Altitudes were not recorded with the GPS as these were never very stable, but a cheap barometric altimeter worked well.
In the field Lev and I recorded some key cave entrances but many sites, some scarcely locatable from our current records, remain to be checked. (We also have a few GPS records from previous years but these do not make it clear what datum was used - as will be seen below, this is important, so future users please ensure you record it!).
A few years ago it was not unreasonable to complain about the standard of Spanish maps but the same cannot be said today. The old maps showed latitude and longitude (relative to Madrid) and/or a Lambert grid; detail based on mapping dating back to the Civil War, sometimes showing evidence of hasty interpretation of aerial photo data with valleys recorded as ridges and, in the Picos, lots of fictitious watercourses. Today, the 1:25,000 Mapa Topografico Nacional de Espana maps from the Instituto Geografico Nacional appear to be extremely accurate, clearly drawn and show latitude and longitude (relative to Greenwich) and a UTM grid based on the European datum (date not stated but as the maps show a publication date of 1995, it must be assumed that the datum is that of 1950 rather than 1995).
The choice of datum is important - the difference between the WGS84 and European 1950 datums is calculated (by Lev using the Molodensky transformation; parameters in the WGS reference) to be between 101-109m E and 210-211m N (using constants optimised for EUR-M (Atlantic Europe from Portugal to Finland) and EUR-D (Iberia)). Repeated GPS readings at the Ario camp showed 102m E and 210m N. The detailed mapping of the Top Camp area reported in Proc 13 by Gerhard Niklasch gives figures which are far more accurate than can be delivered by GPS give differences of 107m E and 205m N.
Recording coordinates is one thing, making them useful is entirely another. Technology has come to our assistance this year in the guise of the acronyms XML and SVG.
The data about a site is held in an XML format; essentially this is plain text where every element of data is tagged (using angled brackets and a closing slash) and may have additional named attribute/value pairs. It could (for the sake of argument) look like this: <AREA name="Area 7" > <CAVE NAME="Pozu Jultayu" TAG="2/7" DESCRIPTION="see elsewhere">The main target of the Ario2000 Expedition <LOCATION>Near the top of Jultayu <UTM Zone="30" X="343978" Y="88012"/><ALTITUDE>1848m XSLT allows data to be selected, sorted, and transformed (using a suitable processor) to another format. For example, we might want html for web publishing, pdf for printing, svg for mapping on web or in print, and x3d or vrml for 3D representations.
SVG allows us to produce graphics (including text) from an (XML) text description - the quality is limited only by the display or printer we use; no artistic skill is needed except in deciding what the map should look like in terms of layout, symbols to use, etc. In particular, we can take an XML file containing the location data and get XSLT to generate a description of a corresponding map. If new sites are added to the cave register, a new map can be produced easily. Also, the same data can be displayed in different ways: for checking maybe we need an uncluttered map with just a dot and the site's tag, plus some peaks which might be used for taking bearings; for publication we might want to display different symbols for horizontal and vertical entrances overlaid on a geological map, a topographic map, or a combined topographic and geological map where transparency allows one to be layered over the other.
We are now well on the way to having more than just a better shaft-bashing guide; we have a definitive list of sites with standard entry structure suitable for searching and processing into reports and maps. Together these will remove the need for the obtrusive physical tagging which used to be needed but which the National Park understandably disapproves of, at the same time as removing the duplicate entries that have arisen from repeated investigation without physical tagging.
It is now possible to have a palmtop computer hold and process all our cave details and deliver them in an updateable text or graphic form; it is also possible to have a palmtop computer which incorporates a GPS system. Our new system may mean that getting lost in the Picos is a thing of the past, but its greatest application will be in pointing the way towards new areas to explore. There is still much to be found...