Date of Award

Summer 8-2015

Degree Type

Masters Thesis

Degree Name

Master of Science (MS)

Department

Geography and Geology

Committee Chair

Dr. Grant L. Harley

Committee Chair Department

Geography and Geology

Committee Member 2

Dr. Carl A. Reese

Committee Member 2 Department

Geography and Geology

Committee Member 3

Dr. Bandana Kar

Committee Member 3 Department

Geography and Geology

Abstract

This thesis research used techniques of dendrochronology to investigate the efficacy of using multiple co-occurring species (MCOS) in a climate reconstruction model compared to a single species (SS) in four old-growth forests in Indiana: Pioneer Mothers Memorial Forest (PM), Donaldson Woods (DW), Hoot Woods (HW), and Lilly Dickey Woods (LD). The objectives of this study were to [1] evaluate the climate response of all chronologies (n = 19; 7 species) to determine the most appropriate climate variable for reconstruction and [2] determine if the MCOS model outperforms the SS model at each individual study site. Model comparison was conducted with r2, adj. r2, standardized residuals, root-mean-square error (RMSE), F statistic, and Akaike Information Criterion (AIC). Summer (June–August; JJA) Palmer Drought Severity Index (PDSI) was the best predicated climate variable, thus two separate models (SS and MCOS) were created at each site for reconstruction. The MCOS outperformed the SS at each site. During the instrumental period (1895–2000), the MCOS at PM, DW, HW and LD explained 50%, 49%, 36%, and 50% of the variance in JJA PDSI, respectively; whereas explained variance of the SS was 40%, 45%, 33%, and 47%. Further, adj. r2, standardized residuals, RMSE, and AIC all suggest that using the MCOS method to reconstruct drought outperforms the SS method. Future tree-ring based climate reconstructions should consider using the MCOS model because it allows reconstructions to go further back in time and produces more accurate estimates of climate conditions.

Share

COinS