Comparing Approaches for Reconstructing Groundwater Levels in the Mountainous Regions of Interior British Columbia, Canada, Using Tree Ring Widths

Stephanie C Hunter,Karen E Kohfeld,Diana M Allen

doi:10.3390/atmos11121374

Abstract

Observed groundwater level records are relatively short (<100 years), limiting long-term studies of groundwater variability that could provide valuable insight into climate change effects. This study uses tree ring data from the International Tree Ring Database (ITRDB) and groundwater level data from 22 provincial observation wells to evaluate different approaches for reconstructing groundwater levels from tree ring widths in the mountainous southern interior of British Columbia, Canada. The twenty-eight reconstruction models consider the selection of observation wells (e.g., regional average groundwater level vs. wells classified by recharge mechanism) and the search area for potential tree ring records (climate footprint vs. North American Ecoregions). Results show that if the climate footprint is used, reconstructions are statistically valid if the wells are grouped according to recharge mechanism, with streamflow-driven and high-elevation recharge-driven wells (both snowmelt-dominated) producing valid models. Of all the ecoregions considered, only the Coast Mountain Ecoregion models are statistically valid for both the regional average groundwater level and high-elevation recharge-driven systems. No model is statistically valid for low-elevation recharge-driven systems (rainfall-dominated). The longest models extend the groundwater level record to the year 1500, with the highest confidence in the later portions of the reconstructions going back to the year 1800.

Highlights

Mountains are a key supplier of water for many regions around the world, producing invaluable freshwater resources for both human use and natural ecosystems
Of the 42 groundwater well-tree ring record pairs that show significant monthly correlations, only 5 pairs produced negative correlations, and they all occurred when Cana 150 and Cana 152 were used in the pairings, with these negative correlations occurring earlier in the year than the majority of the other pairings
The most consistent pattern among the correlations was a positive correlation between depth to groundwater levels and current-year tree ring width, which corresponds to the timing of peak groundwater levels in the streamflow-driven systems, and the timing of lowest groundwater levels for the low-elevation recharge-driven systems (Figure 4, Supplementary Information, Table S3)

Summary

Introduction

Mountains are a key supplier of water for many regions around the world, producing invaluable freshwater resources for both human use and natural ecosystems. A global assessment of the significance of mountains for water resources showed that 44% of global mountain areas played an important or essential role in the local water supply [1]. These mountain water resources are some of the most vulnerable to climate change [2,3]. It is important to understand how climate variability affects the groundwater system. The response of groundwater to climate variability over time is necessary to put current observations into a long-term

Objectives

Methods

Results

Discussion

Conclusion