Established closed-form analytical solutions for using heat as a tracer of vertical groundwater fluxes typically rely on assumptions of steady hydraulic conditions. We introduce a novel analytical approach and associated tool, PyTFLUX, to account for transient changes in vertical groundwater fluxes. The analytical solution uses a Fourier series to represent diurnal surface temperature variability and a differential method to represent vertical flux changes. Optimization techniques are employed to achieve faster convergence and prevent the estimation of unreasonable vertical fluxes. The PyTFLUX script, presented in a Python Jupyter notebook, enables the easy adoption of the new analytical framework. To test the new approach, illustrative transient vertical flux time series were developed for three time-varying groundwater flux scenarios: a step-change, a single sine-wave, and a mixed sine-wave. These profiles were analyzed to infer vertical groundwater flux time series using PyTFLUX and previously published methods implemented in VFLUX2. Results show that PyTFLUX can reproduce temporal variability in groundwater fluxes not typically captured by existing methods. Finally, previously published high-resolution sediment temperature data from the Quashnet River in Massachusetts, USA, were analyzed to demonstrate the efficacy of PyTFLUX in analyzing complex field data. The analysis of field data yielded a vertical flux time series with mean values that agreed with fluxes yielded from other approaches, but the new approach also revealed pronounced temporal flux variability that was obscured by other methods.

Cryogenic vacuum distillation (CVD) is a widely used technique for extracting plant water from stems for isotopic analysis, but concerns about potential isotopic biases have emerged. Here, we leverage the Cavitron centrifugation technique to extract xylem water and compare its isotopic signature to that of CVD-extracted stem water as well as source water. Conducted under field conditions in tropical northern Australia, our study spans seven tree species naturally experiencing a range of water stress levels. Our findings reveal a significant deuterium bias in CVD-extracted bulk stem water when compared to xylem water (median bias \($-14.9\textperthousand$\)), whereas xylem water closely aligned with source water (median offset \($-1.9\textperthousand$\)). We find substantial variations in deuterium bias among the seven tree species (bias ranging from -19.3 to \($-9.1\textperthousand$\)), but intriguingly, CVD-induced biases were unrelated to environmental factors such as relative stem water content and pre-dawn leaf water potential. These results imply that inter-specific differences may be driven by anatomical traits rather than tree hydraulic functioning. Additionally, our data highlight the potential to use a site-specific deuterium offset, based on the isotopic signature of local source water, for correcting CVD-induced biases.