Tree rings and volcanoes : the climate of NW North America

This thesis aimed to refine temperature reconstruction strategies for Northwest North America (NWNA), introducing new applications of the latewood blue intensity parameter (LWBI), and pooling multiple regional datasets to create a new millennial-length temperature reconstruction. Currently, NWNA is underrepresented by long tree records in comparison to Eurasia. Such records are crucial for contextualizing modern warming in the context of past variability. The NWNA region represents the origin of the “Divergence Problem”, a phenomenon referring to the loss/weakening of temperature sensitivity over time. To address this, a strategy was developed, using a network of sites in the southern Yukon, to minimise its impact and optimise the temperature signal. Trees growing within 100m of upper treeline provide the most temporally stable signal, making them the optimal targets for a temperature reconstruction. The successful use of LWBI for developing temperature reconstructions is well-documented, yet its application for dating is less understood. LWBI was utilised to absolutely date the 9th century Mt. Churchill eruption for the first time using dendrochronological techniques, a result not possible using RW. By demonstrating that LWBI can be effectively used for crossdating between species, dated subfossil trees preserved in the volcanic ash placed the eruption between 852-853 CE. The final LWBI reconstruction provided temperature variability for NWNA back to the 3rd century CE. While full period modelling (1901-2014) explained 46% of the temperature variance, the fidelity of the reconstruction reduced significantly prior to 933 CE due to decreased replication and significant underestimation of extreme warm season years was noted. The NWNA BI-based reconstruction demonstrates the most time stable temperature signal compared to other NWNA reconstructions but struggles to capture long term trends. Future research should integrate these data for a multi-parameter reconstruction for a more comprehensive, robust reflection of NWNA temperature variability.