Hydrogeologic and climate drivers of water isotopes in fractured rock: A word of caution for the use of groundwater isoscapes in humid continental settings

•Temporal δ2H sampling from multi-level wells is recommended in fractured rock.•Bedrock with thin soil needs long-term sampling for representative groundwater means.•Climate is the primary driver of groundwater δ2H, hydrogeology is secondary.•Winter precipitation (rain and snow) plays a critical role in recharge in this region. Time-discrete groundwater isoscapes are commonly used as a proxy for the weighted mean isotopic value of precipitation and to infer recharge processes. In fractured rock environments, however, the complete removal of temporal variation may not occur in the unsaturated zone, and borehole artefacts may complicate isoscape interpretation when sampling open boreholes. The study presented herein investigates the benefits of adding temporal depth-discrete δ2H measurements to isoscapes in fractured rock. Concurrent objectives include determining whether time-discrete groundwater samples are representative of long-term averages of isotopes in precipitation and identifying the primary drivers of groundwater δ2H in these settings. In central and eastern Ontario, Canada, a stable isotope dataset consisting of 175 time-discrete, open-hole bedrock wells was statistically compared to a temporal dataset with 25 multi-level well intervals. Both univariate and multivariate statistical analyses were conducted between δ2H and a wide range of climate and hydrogeologic parameters. Results from the analyses indicate winter rain, average annual snowfall and groundwater age are key variables driving groundwater deuterium values, and hydrogeologic variables such as overburden thickness, watershed divides, and bedrock geology play a secondary role. This suggests that changes to winter precipitation may result in greater changes to groundwater recharge compared to other seasonal precipitation, which has implications for water resources under climate change. Findings also indicate that caution is required when interpreting time-discrete groundwater isoscapes in fractured rock, particularly in areas of thin soil. If stable isotopes of groundwater are temporally variable, long-term sampling is required to determine a representative groundwater mean. Incorporating temporal sampling of multi-level wells with groundwater isoscapes can provide additional insight about recharge timing that time-discrete, open-hole samples alone do not. This study emphasizes the need for temporal depth-discrete isotope measurements in humid continental fractured rock environments for