CLAMP Updates

(23/06/10) The physiognomy files Physg3arc and Physg3brc have been re-ordered alphabetically to facilitate relating these files to the original scoresheets. They are now designated Physg3arcAZ and Physg3brcAZ. Correspondingly the meteorological files have also been re-ordered and are designated Met3arAZ, Met3brAZ, GRIDMet3arAZ and GRIDMet3brAZ. Following on from that new results spreadsheets have been introduced and are called Res3arcAZ, Res3brcAZ for use when the meteorological data observed at the CLAMP calibration sites are used, and Res3arcGRIDAZ and Res3brcGRIDAZ when gridded meterological data are used. To avoid confusion the previous files, arranged in decending order of mean annual temperature, have been withdrawn from the Datasets, Gridded Data and results Spreadsheets. Bundled with these changes are further minor corrections to the calibration files.

Some of you have reported that sometimes the results of your analyses of the calibration data sets do not exactly match the results shown in the results spreadsheets. This caused some puzzlement until Frédéric Jacques reported to us that WCanoImp.exe, used to import data to the CANOCO package, only copies what is on screen in terms of the numbers of significant figures and not, as was assumed, the exact figure embedded in the spreadsheets. Thus, if you were using scores rounded to the nearest whole number for display purposes then WCanoImp.exe would only import integers into the analysis. This results in different biplot scores to those used in the results spreadsheets. To overcome this we recommend everyone displays physiognomic data to three decimal places when using WCanoImp.exe. Meteorological files should be displayed at 1 decimal place. To assume greater precision would be unwarranted.

To determine which is the more appropriate calibration set to use for a particular fossil site is usually done by running Physg3arcA-Z first and seeing where the fossil site lies within physiognomic space. If the fossil site lies within the space occupied by the Physg3brcA-Z samples then it is appropriate to use the Physg3brcA-Z calibration. In practice this may not be a straightforward exercise, particularly if the fossil site resides close to the so called "alpine nest" of cold sites. To help choose which dataset is most appropriate for calibration Vasilis Teodoridis of Charles University, Prague, has developed a spreadsheet. This can be downloaded by clicking here. Note that the spreadsheet is in .xlsx format. If you have any questions please contact Vasilis.

Coming Soon - to streamline the process we are developing a more integrated CLAMP analytical process using the open source package "R". Initial tests yield identical results to analyses run in CANOCO. This will avoid issues such as that associated with the use of WCanoImp. New calibration data sets from South Africa, India and China will be available after the testing process is complete. We hope to have these innovations publicly available towards the end of 2010.

(01/05/09) CLAMP teaching resources are now available for download. For more information and to access the downloads go to the Teaching Materials link in the Downloads section of the Site Index panel or click here.

(29/04/09) New CLAMP meteorological calibration files are now available that utilise gridded data as used in climate modelling. This allows CLAMP to generate climate predictions that are more directly comparable to the data sets used to evaluate climate models. It also allows additional modern vegetation sites to be added to the CLAMP calibration sets even though the sites may not be near climate recording stations. To find out more go to the Gridded Data link in the Site Index panel or click here.

(09/04/09) Additional scoresheets for the calibration training sets collected and scored by JAW that previously only existed in hard copy have now been transferred to the electronic automated scoresheets. These can be downloaded from the list accessed by clicking here. In transferring the data some small arithmetic errors were detected and these have now been corrected.

(04/02/09) The instructions for scoring leaves have now been improved following feedback from the presentation of CLAMP training classes in India and China. The new instructions are available as a downloadable .pdf file. Similarly there is now a .pdf file available from the CLAMP analysis page with step-by-step instructions for carrying out a CLAMP analysis using CANOCO.

(06/03/2006) The locations of the calibration (training set) sites can now be viewed in Google Earth. To find out how to do this click here or go to the Google Earth CLAMP link in the Site Index Downloads panel to the left of this page.The scoresheets for many sites in the calibration training sets (Physg3brcAZ and Physg3brcAZ) are now available. They can be accessed by clicking here, or by going to the Calibration Scoresheets link on the Site Index panel. Note that this facility is still under development and not all site scoresheets are currently available.

(25/08/2005) In Spicer et al., 2004, Journal of Geology, v.112, p.685-702 (click here to download the .pdf file) the 173 dataset was expanded by the inclusion of 20 sites from across Russia. This file has been designated Physg3cr.xls and is accompanied by the matching meteorological file Met3cr.xls. Click on the file names to download the files. Note, however, that these files should not yet be used for routine analysis and for that reason the automated results spreadsheet is not given. Also added is an example of morphotyping using a Coniacian fossil flora from the Northern Pekulney Range, northeastern Russia (Craggs, 2005). This is presented as an example of how a leaf assemblage can be divided into morphotypes (species) on the basis of not only leaf characters used in CLAMP, but also venation features. To view this classification click here.

(10/10/2001) On October 10th 2001 Quicktime movies were updated to reflect the changes above and to give improved visualisation (e.g. depth queuing) and 3D rendering. The movies were generated using CrystalMaker 5.0. Howard Schorn (UC Berkeley) has kindly made available additional leaf overlays for determining leaf shape and length-to-width characteristics. To download the leaf shape overlay as a .pdf file click here. For the leaf length-to-width overlay click here.