Lag times and flow pathways of New Zealand's river catchments: Developing robust national scale metrics using stable isotopes

Sub-surface flow pathways and transit times of water to rivers are vital catchment characteristics that determine how climate change and land use affect surface water quality and runoff amounts. These catchment characteristics also determine the appropriateness of catchment scale management decisions to control water quality and runoff.

Conservative hydrologic tracers remain reliable, accurate tools to partition river flows among flow pathways, and calculate transit times. For example, hydrogen (H) and oxygen (O) stable isotopes provide the data for robust calculation of the young (less than a few months old) fraction of river flow. H and O isotopes also have the potential to be integrated into the next generation of 'isotope-enabled' hydrological models, which are designed to provide accurate flow-source partitioning and flow estimates outside the range of historical climate conditions.

Currently, the use of H and O stable isotopes as hydrologic tracers in large catchment scale hydrology across New Zealand is hindered by the requirement for extensive, non-routine sampling. To lower this hurdle, we have prepared national databases of precipitation and river water isotope data, and developed national-scale, time varying isotope models (isoscapes).      

Here, I describe our development of precipitation and river isoscapes for New Zealand, and initial calculations of young water fractions across New Zealand rivers.

Database development used a combination of regional government, research and citizen science collections. Our databases now include regular long-term (>18 months) sampling from around 100 rivers, and over 100 precipitation sites nationally.

Using these databases, we have developed isoscapes using a range of statistical modelling techniques.  Sinusoidal precipitation isoscape results suggest that strong seasonal cycles of precipitation stable isotope values in some areas of New Zealand (but not others) are conducive to calculation of young water fractions for rivers and may require consideration for interpreting sources of recharge to groundwater and river water. Our machine-learning precipitation isoscape captures much of the non-seasonal temporal variation that dominates in windward areas of New Zealand. These results have wider implications for the application of stable isotopes as hydrological tracers in regions with mixed marine- and continental-type climates.

Results indicate that precipitation isoscapes can now be combined with regular river sampling to provide robust comparisons of young water fractions at a regional level.