Refining the standardized growth change method for pointer year detection: Accounting for statistical bias and estimating the deflection period

Changing climate sensitivity of secondary growth following extreme drought events in forest ecosystems: a global analysis

“…when negative, implies lower than average water availability, not necessarily water deficit [78]. Therefore, we defined climatological droughts as the cooccurrence of a −2 standard deviation in SPEI [79] and negative CWB for the month and SPEI time-scale selected for each site, while for the corresponding ecological responses, we scanned all chronologies for negative anomalies using a recently published pointer year detection method, the bias-adjusted standardized growth change method (BSGC) [80]. The BSGC method relies on probability density functions of annual growth changes and defines abnormal growth changes outside the 95% confidence interval.…”

“…The BSGC method relies on probability density functions of annual growth changes and defines abnormal growth changes outside the 95% confidence interval. This method removes the arbitrary thresholds used by traditional methods to identify pointer years, while additionally resulting in a more precise pointer-year detection rate for both existing and artificially generated tree-ring data [80,81]. The combination of BSGC-determined negative pointer years and climatological droughts constrained our analysis to all droughts that featured extraordinary growth changes, ensuring significantly lower RWI-values compared to only using the climatological thresholds (Wilcoxon rank-sum test, p < 0.001, SI figure 4).…”