Dendrochronological assessment of drought severity indices for Panola Mountain Research Watershed, Georgia, U.S.A.

Thursday, 17 December 2015
Poster Hall (Moscone South)
Anna McKee and Brent T Aulenbach, USGS Georgia Water Science Center Norcross, Norcross, GA, United States
Quantifying the relation between drought severity and tree growth is important to predict future growth rates as climate change effects the frequency and severity of future droughts. Two commonly used metrics of drought severity are the Standardized Precipitation Index (SPI) and the Palmer Drought Severity Index (PDSI). These indices are often calculated from proximal weather station data and therefore may not be very accurate at the local watershed scale. The accuracy of these commonly used measures of drought severity was compared to a recently developed, locally calibrated model of water limitation based on the difference between potential and actual evapotranspiration (ETDIFF). Relative accuracies of the drought indices were assessed on the strength of correlations with a 20-year tree-ring index chronology (1986-2006) developed from 22 loblolly pine (Pinus taeda) trees in water-limited landscape positions at the Panola Mountain Research Watershed (PMRW), a 41-hectare forested watershed located in north-central Georgia. We used SPI and PDSI index values from the weather station located at the Atlanta Airport, approximately 36 kilometers from PMRW. ETDIFF was calculated based on precipitation, temperature, runoff, and solar radiation data collected at PMRW. Annual index values for all three drought indices were calculated as the mean value over the growing season (May to September). All three indices had significant Pearson correlations with the tree-ring index (p = 0.044, 0.007, 0.002 for SPI, PDSI, and ETDIFF, respectively). The ETDIFF method had the strongest correlation (R2 = 0.40) compared to SPI and PDSI results (R2 = 0.19 and 0.32, respectively). Results suggest SPI and PDSI provided a general measure of drought conditions, however, the locally calibrated model of water limitation appears to measure drought severity more accurately. Future studies on the ecological effects of drought may benefit from adopting ETDIFF as a measure of drought severity.