GEOS 595e: Journal Club
Spring 2007 Reading Focus Questions

Week 4: Drought reconstruction from tree rings (Mar 28th)
submitted by Troy K.

Introduction:

This week we explore the development of using tree rings to create gridded continental scale drought reconstructions. This work has been spurred by both the explosion in the number of tree ring chronologies in North America, and increases in the depth of time covered by these chronologies in some parts of the continent. This would not be possible without extensive sharing of primary data throughout the dendrochronological community and a central database through which this data can be accessed. Because spatial variation is of vital interest in these works they provide an interesting contrast to the single point streamflow reconstructions we discussed earlier in the semester, especially in terms of the methodological considerations.

Readings:

• Meko, D.M., E.R. Cook, D.W. Stahle, C.W. Stockton and M.K. Hughes. 1993 Spatial patterns of tree-growth anomalies in the United States and Canada. Journal of Climate 6:1773-1786.
• Cook, E.R., D.M. Meko, D.W. Stahle and M.K. Hughes. 1999 Drought reconstructions for the continental United States. Journal of Climate 12:1145-1162.
• Cook E.R., C.A. Woodhouse, C.M Eakin, D.M Meko and D.W. Stahle. Long-term aridity changes in the western United Sates. Science 306(5):1015-1018.

Primary Reading Questions:

1. How do the 1993 and 1999 papers take different approaches to reconstructing drought? Are there advantages to mapping growth variations as opposed a reconstructed climate variable? Is the 1993 work a necessary precursor to the 1999 work?
2. What other methods for spatialized climate reconstructions are available? What advantages do the authors cite for PPR relative to other methods? Do you agree?
3. Why PDSI? What is it, and are there other drought/moisture indices available?
4. How do the three papers differ in the screening process for chronologies to be used in the analysis? Is one method superior? Is quality of each chronology checked through programs such as COFECHA? Why might this be important when using ITRDB data?
5. In the 1999 paper the authors devote considerable space to addressing poor performing areas in their verification statistics. What are these problems and how do they explain theses seeming deficiencies? Are you convinced by their reasoning?  In what other ways could they have tested the validity of their reconstruction?  What further verification do they add in the 2004 paper?

1. Why employ (in all three papers) distance limits on chronologies available for each grid point reconstruction? How does this differ (if at all) from the streamflow reconstruction criteria for chronology inclusion from our earlier discussion?
2. In the 1999 paper how do they justify or explain why they reconstruct summer conditions only? How might reconstruction of other seasons or the entire water year affect calibration and verification of the regression models and the results?
3. In the 1993 paper the authors discuss differences in species make-up of the chronologies at some length. Why is this not addressed in the 1999 and 2004 papers? Should it be (especially in the 1999 paper)?
4. In all 3 papers what considerations are made for decreasing sample size through time in the tree ring chronologies? What other measures could be taken to ensure adequate sample depth in each chronology? Given the length of reconstructions in the 2004 paper should sample size issues have been given more discussion (considering the journal format as well)?
5. From the 2004 paper, what are the advantages and disadvantages to using a simple measure of drought extensiveness such as the Drought Area Index (DAI)?