TREE-RING DATA BANK of SAIMA Centre for Environmental Sciences in Savonlinna 

Jouko Meriläinen¹ (JM) and Mauri Timonen² (MT),  2004

¹   University of Joensuu Saima Centre for Environmental Sciences, Linnankatu 11, FIN-57200 Savonlinna   

²   Finnish  Forest Research Institute,   Rovaniemi Forest Research Station,   P.O. Box 16 , FIN-96301 Rovaniemi  

The data bank of Saima Centre was established in 1992 to support the joint works of Mr. Markus Lindholm and JM. The cooperation and later work with an increasing group of scientists have produced a good number of publications and generally contributed in many ways to Finnish dendrochronology.

Now is a good time to open the data bank for the needs of all researchers interested in Finnish tree-ring data. However it is a strong wish that Meriläinen and Timonen will be included in the future lists of authors as follows. JM is answerable for the whole project and specialized in questions of environmental changes. MT responds to questions concerning dendrochronological methods.

Mr. Risto Tanninen was responsible for data processing under the guidance of Timonen.  Mr. Kari Tolvanen was highly professional to check the accuracy of all tree-ring measurements. They both have become experienced in the field of dendrochronology during several years.

The data bank distributes data from 42 selected sites. Data can be opened as text files in all Windows systems.

In addition to the data from 42 selected sites, the Saima Centre can provide also other dendrochronological data sets available for joint research. Almost all sites are situated in eastern  Finland.

“Responsible Partners” of each site description include the names of persons who have supported JM´s sampling trips in one way or another. The complete list should  also include  Jouko’s  wife Leena Örnberg!

Raw measurements.  All samples were measured using  CATRAS. Those data sets are available. The other programs applied to produce the data are CONVERT5, COFECHA12K and COFCOR. The KINSYS software is introduced in detail by M. TIMONEN (FINTRLABS).

Measurement file indicates the original measured ring-widths (  TUCSON  format).

Quality Control file is created by COF12K. It contains information e.g. on: number of dated samples, period and the number of years, total number of years of all series, and total number of dated tree-rings. 

Descriptive statisticS contains plenty of information on the samples, including the number of segments of the sample, and the number of flags (a value indicating exceptional or problematic tree-rings) per a segment.   

Grouping of sites is based on living trees, wooden buildings  and subfossil trees from lake sediments and bogs.

Sample identification. Each of the samples is provided with a code (e.g. OMA1200B). The first three letters indicate the position on the map grid. After the set of numbers the letters A, B (or C) show the radius and X the average of the radii.

Cofecha-KINSYS Quality Control Method. The main criteria we pay attention to in Cofecha’s output file analysis are described in Part 5 (Correlation by segment of each series with Master), Part 6 (Potential problems: low correlation, divergent year-to-year changes), and Part 7 ( Descriptive statistics).  The flags play main roles in these sections. The flags are classified into two categories: the A-type indicates a low correlation with Master; the B-type announces about possible shifts. Closely related to these parameters, poor correlations, even minus-signed, appear in the data. A closer look in the Part 6 provides detailed information of the nature of flags and is useful for deciding about future actions that depends on what one is doing, quality checking, dating or chronology building. 

Our data quality assessment starts usually from Cofecha’s Descriptive statistics (Part 7). We have found it useful to check the total number of flags, their proportion from the total number of segments in the data and also their distribution by samples.  An ideal case of high-quality data would be a Cofecha output file with zero or just a few insignificant flags and high correlations (>.40).

In a normal case, however, there are less or more problematic cases to be investigated. A good rule of thumb for high-quality tree-ring width data for Scots pine in  Finland is a threshold value of flags less than 3% calculated from the number of total flags in the data. An exceptional number of flags in an individual sample suggest measurement or dating errors, sometimes also to other reasons.

We have included in our international data set only the samples that we have been able to  verify or judge to be correctly measured and dated. The rest of the data (undated, uncertain, erroneous, strangely formed samples, etc.) are not included, but they are available by request.

If we need to filter data, e.g. for improving climatic signal for chronology building, a special iterative process utilizing Cofecha’s output files will take place. This approach is called the KINSYS-Cofecha Quality Control and Dating. This procedure, developed by MT mainly 1996-1999 in the Advance-10K project, makes it possible to filter the data according to a preset user-defined quality criteria. The quality-criteria is based on threshold values defined by Cofecha’s “Correlation with master” table (statistics in Part 7), number of segments and number of rings in a sample.  Only those samples exceeding all the criteria will be accepted for building a new data set. The qualifying work is done by the program COFCOR. Its output file can be edited, which makes it possible also add or remove samples there.

The next step is to input Cofcor’s output file to a KINSYS program called KINDATA, which creates a new   Tucson  file.  If the idea is just to filter data, the next iterative operation will be a new Cofecha Quality Control run, a new sample selection in Cofcor, and a new data creation in Kindata. In case of chronology building, the method is the same, but new Cofecha dating scans from the undated original sample source data are needed to bring new dated data to the chronology. Also, a new Master series based on the last accepted  data has to be created by Cofecha. 

The greatest financial support: most money for salaries of the technicians, came from the Employment  Office of Savonlinna (Mr. Reima Salo, director). The other important Maecenas was Valtiokonttori, the government office having  paid the pension to JM  already for 11 years!