Besancon dendro database: Difference between revisions
(link corr etc) |
No edit summary |
||
| Line 5: | Line 5: | ||
expected most narrow ring width value, the resulting "ring width file" may be used for crossdating. | expected most narrow ring width value, the resulting "ring width file" may be used for crossdating. | ||
Be careful when selecting the new smallest value. A too small value may fool your matching algorithms (see | Be careful when selecting the new smallest value. A too small value may fool your matching algorithms (see | ||
Lars-Åke Larsson: ''[http://www.cybis.se/forfun/dendro/foolbynarrow/index.htm How to get fooled by your normalization method and some too narrow ring widths]''. | Lars-Åke Larsson: ''[http://www.cybis.se/forfun/dendro/foolbynarrow/index.htm How to get fooled by your normalization method and some too narrow ring widths]''). | ||
Also published ring-width-look-alike files from Besancon may be stored as only positive values though the smallest value is set to "1". Such files will easily [[Normalization#A_weakness_with_the_Hollstein_and_Baillie.2FPilcher_transformations|fool the Hollstein or Baillie/Pilcher algorithms]]. | Also published ring-width-look-alike files from Besancon may be stored as only positive values though the smallest value is set to "1". Such files will easily [[Normalization#A_weakness_with_the_Hollstein_and_Baillie.2FPilcher_transformations|fool the Hollstein or Baillie/Pilcher algorithms]]. | ||
Revision as of 21:31, 18 December 2009
At the "Laboratoire de Chrono-Ecologie, Université de Franche-Comté" there are a few published dendro data files stored in Besancon .txt format.
Some files are stored in a special normalized format with curves looking almost as ring width curves though with both positive and negative values running above and below a zero line. If the values are offset to make the most negative value positive and also set at a new value corresponding to an expected most narrow ring width value, the resulting "ring width file" may be used for crossdating. Be careful when selecting the new smallest value. A too small value may fool your matching algorithms (see Lars-Åke Larsson: How to get fooled by your normalization method and some too narrow ring widths).
Also published ring-width-look-alike files from Besancon may be stored as only positive values though the smallest value is set to "1". Such files will easily fool the Hollstein or Baillie/Pilcher algorithms.