DendroClim 2002

A statistical tool for analysis of Evolutionary and Moving Intervals in Dendroclimatology

This project involves implementation of bootstrapped correlation and response functions to identify climatic signals in tree rings. The software allows the analysis of a single time period (or interval) as well as of multiple time intervals. This application was developed in Microsoft Foundation Classes (MFC) with Visual C++.

 

(Please e-mail to get your copy of the program. Remember to include your name, affiliation, postal and e-mail address)

 

Add-ins available:

ClimGraph
by Tom Levanic

A small program for Windows that will read and display *.sig files from DendroClim, allowing users to save images as EMF files or print.

Input data needed for this program are monthly climate variables (typically precipitation and temperature), and tree-ring index values for a given time period. After specifying the input files, the user defines window boundaries to run the analysis. Depending on input conditions, the user can run either bootstrapped correlation and response functions on a single time period (interval), or on multiple intervals. The evolutionary and moving intervals essentially iterate the bootstrapped correlation and response functions for different time periods. Each bootstrap estimate is obtained by generating 1000 samples (selected at random with replacement), then running numerical computations on each sample. Computations involve linear correlation, Jacobean rotations for Eigen values, singular value decomposition and solutions of linear systems accompanied by principle component regression. Results are displayed graphically using color-coded symbols, and output files are automatically generated in ASCII format.

Basic program operation (see specific Help files for details):

  1. Upon starting DendroClim, the user chooses a tree-ring sample measurement file along with up to 3 climate variable files for comparison.
  2. The user selects the time window (number of months --- within one calendar year, or spanning two), and the interval (number of years common to all files) to be used in the analysis.
  3. An output file location is chosen by the user.
  4. The type of analysis is determined (a baselength is then set if Evolutionary & Moving Intervals are chosen).
  5. The analysis is run, and results are displayed graphically, along with creating output files (in ASCII format).
 
 

Related publications :

  • 1997. Biondi, F. Evolutionary and moving response functions in dendroclimatology. Dendrochronologia 15: 139-150. (ARTICLE in PDF format - about 3.7 Mb)
  • 2000. Biondi, F. Are climate-tree growth relationships changing in north-central Idaho? Arctic, Antarctic, and Alpine Research 32: 111-116. (ABSTRACT; ARTICLE in PDF format - about 2.3 Mb)
  • 2004. Biondi, F. and K. Waikul. DENDROCLIM2002: A C++ program for statistical calibration of climate signals in tree-ring chronologies. Computers & Geosciences 30: 303-311 (ARTICLE in PDF format - about 0.6 Mb)
 

DendroLab © 1994 - present