Using StorAge Selection functions to quantify ecohydrological controls on the time-variant age of evapotranspiration, soil water, and recharge

Abstract

Abstract. Quantifying ecohydrological controls on soil water availability is essential to understand temporal variations in catchment storage. Soil water is subject to numerous time-variable fluxes (evaporation, root-uptake, and recharge), each with 10 different water ages which in turn affect the age of water in storage. Here, we adapt StorAge Selection (SAS) function theory to investigate water flow in soils and identify soil evaporation and root-water uptake sources from depth. We use this to quantify the effects of soil-vegetation interactions on the inter-relationships between water fluxes, storage, and age. The novel modification of the SAS function framework is tested against empirical data from two contrasting soil-vegetation units in the Scottish Highlands; these are characterised by significant preferential flow, transporting younger water through the soil during 15 high soil moisture conditions. Dominant young water fluxes, along with relatively low rainfall intensities, explain relatively stable soil water ages through time and with depth. Soil evaporation sources were more time-invariant with high preference for near-surface water, independent of soil moisture conditions, and resulting in soil evaporation water ages similar to nearsurface soil waters (mean age: 50 -65 days). Sources of root-water uptake were more variable: preferential near-surface water uptake occurred in wet conditions, with a deeper root-uptake source during dry soil conditions, which resulted in more variable 20 water ages of transpiration (mean age: 56 -79 days). The simple model structure provides a parsimonious means of constraining the water age of multiple fluxes from the upper part of the critical zone during time-varying conditions improving our understanding of vegetation influences on catchment scale water fluxes.

Water ages in the critical zone of long-term experimental sites in northern latitudes

“…Similar to findings by Smith et al (2018) for the heather site in Bruntland Burn, our estimates for A E at that site were highest during periods of limited infiltration (e.g. 10-year return period drought in summer 2013 at Bruntland Burn in Fig.…”

“…Further, observed differences in the isotopic compositions of mobile and bulk soil water in the field were often related to the potential age differences of waters sampled at different mobilities (Landon et al, 1999;Brooks et al, 2010;Geris et al, 2015;Sprenger et al, 2015a;Oerter and Bowen, 2017). Our results and recent simulations by Smith et al (2018) support such interpretations, as the water in the slow flow domain was generally older than the water in the fast flow domain (A Ss > A Sf ). However, since the differences between A Ss and A Sf were variable in time and were often maximized in early spring, such anomalies in water ages are likely to be reflected in the isotopic compositions of the water, with the older water in small pores being less depleted in heavy isotopes (originating partly from autumn precipitation) than the young water in larger soil pores draining recently infiltrated isotopically depleted snowmelt or winter precipitation.…”

“…Such investigations are of special interest in light of ongoing research regarding the consequences of a potential TWW hypothesis on water age estimations based on tracer-aided modelling (Hrachowitz et al, 2016). In particular, our simulations underline that neither is ET flux withdrawal well mixed in its age composition nor is the pool of plant water uptake well mixed, which is increasingly acknowledged in water age studies (Harman, 2015;Smith et al, 2018).…”

“…Investigation of water ages in the ET flux is relatively new (Botter et al, 2011) and age dynamics have usually been assessed for the bulk ET flux (Harman, 2015;Soulsby et al, 2015;van der Velde et al, 2015;van Huijgevoort et al, 2016;Soulsby et al, 2016). While it was shown that tracer-aided modelling using stable isotopes of water benefits from partitioning ET into a fractionating E flux and a non-fractionating T flux (Knighton et al, 2017), separate water age analyses for E and T have been considered only recently (Smith et al, 2018). However, our analyses showed that the two different fluxes can have markedly different travel time dynamics (Fig.…”

“…Further, an assessment of the variability of water ages within different pore spaces (e.g. mobile versus more tightly bound water - Brooks et al, 2010;Good et al, 2015;Sprenger et al, 2018b;Smith et al, 2018) and with soil depth is still needed; this would also provide a test of the common assumption of a well-mixed system in tracer-aided modelling (van der Velde et al, 2015). Also, it has not yet been established how the ages of soil water storage and evaporation and transpiration fluxes are related to the variability of soil storage volumes.…”