Trends of Stable Isotopes in Precipitation Over Switzerland and Their Relation to Recent Climate Variability
In Switzerland regular stable isotope measurements in precipitation started in 1970. The sampling network ranges from 290 to 3580 m a.s.l. and covers the main important hydro-meteorological regions. In terms of precipitation meteorology a general precipitation amount -altitude relationship does not exist because of the shielding of inner Alpine valleys and enhanced precipitation along the Alpine foothills. The stable isotope- altitude effect varies over a range 0.1 to 1 per mil/100 m depending on the shielding and/or seasonal distribution of precipitation. Recent warming has resulted in winters with varying solid/liquid precipitation amounts and a higher frequency of heavy precipitation events during summer. The monthly delta 18O time- series follow the observed stepwise temperature increase since 1970 on a NW-SE transect across the Alps. The delta 18O/temperature relation is season-dependent and generally lesser correlated during winter. However, during that time the delta values follow closely the NAO index. Since the North Atlantic Oscillation influences both precipitation and temperature via air mass variability, this might be the reason for the poor correlation. On a monthly basis over the whole year the relation varies between 0.3 and 0.6 per mil/°C, mainly depending on how uniformly precipitation is distributed.
Isotope climatology of Canada: Insights from the first decade of CNIP operation (1997- 2007)
The Canadian Network for Isotopes in Precipitation (CNIP) was initiated as a joint venture between university
and government researchers to provide the spatial and temporal data necessary to examine the sensitivity of
isotope fields to changes in circulation patterns. The network consists of 17 stations distributed across
Canada, spanning almost 40° of latitude and 70° of longitude collecting weighted monthly
precipitation samples. This marks the first time that both the southern and northern regions of the country have
been simultaneously sampled. Sampling of the southern stations was initiated in 1997 to supplement an
existing informal arctic network (now formally incorporated in CNIP) resulting in a 10-year dataset for the entire
country. The sensitivity of isotope-climate signals to modes of interannual variability such as El Nino-Southern
Oscillation, North Atlantic Oscillation, Arctic Oscillation, and the Pacific North American pattern is of interest
because they are a primary cause of interannual climate variability. The ten-year dataset includes some
intriguing results that reveal the complexity and dynamic nature of isotope-climate relations in Canadian
precipitation. Time-series of the CNIP dataset include anomalously low d-excess values during the 1997-98 El
Nino at southern stations located across the entire country, suggesting a wide-reaching influence of this mode
of climate variability on precipitation labeling. Correlations between and isotope time-series and various
circulation indices show strong regional controls on isotope-climate relations.
A GCM partitioning of controls on the stable isotope composition of precipitation
Idealized Rayleigh models of depletion have been used to describe many characteristics of stable water isotope variation. They are complicated, however, by many 'non- Rayleigh' factors: mixing along transport pathways, re-evaporation of cloud liquid water and precipitation, and re-evaporation from the land surface. The effects of these processes have been observed in case studies, and described with simple models. But to what extent do they contribute to isotopic variability in general, and subsequently, to climatological corollaries such as the temperature effect? In this study, we conducted a series of 50-year runs with the NASA GISS ModelE general circulation model to estimate the strength of different controls on the precipitation isotopes. ModelE is equipped with isotope tracers, with the fractionation between heavy and light isotopes modeled through all stages of the hydrologic cycle, from evaporation over the ocean and land surfaces, to condensation and re-evaporation in clouds. The effect of disabling fractionation during Rayleigh condensation generally resulted in a net isotopic enrichment, which grew stronger with decreasing temperature. The absence of post-condensation fractionation generally had a smaller enriching effect under colder conditions, and also had a net depletion effect at lower latitudes over the ocean. We also examined the relative strength of these processes on the temperature effect, finding significant geographic and seasonal differences in the dominant control. With all fractionation processes enabled, we found a distinct correlation between local temperature and isotopic anomalies which was strongest during boreal winter, consistent with many previous studies. Our experimental runs showed, however, that this temperature effect is in fact the superposition of two different processes. Over Alaska, northern Canada, Greenland and Siberia, the temperature effect is attributable to Rayleigh processes, whereas over the southeastern US and Europe, the temperature effect was due to non-Rayleigh processes. In all regions except for Greenland, any summer temperature effect is the result of post-condensation processes. Improved understanding of the mechanistic controls on precipitation isotopic variation will hopefully contribute to improved interpretation of stable isotopes as climate proxies and diagnostics of the hydrologic cycle.
Mesoscale Model Validation using Stable Water Isotopes: The isoWATFLOOD Model
A methodology to improve mesoscale model validation is developed by calibrating simulations of both water and isotope mass simultaneously. The isoWATFLOOD model simulates changes in oxygen-18 of streamflow and hydrological processes contributing to streamflow. The added constraint of simulated to measured delta oxygen-18 in streamflow lowers the models degrees of freedom and generates more physically-based model parameterizations. Modelled results are shown to effectively reduce and constrain errors associated with equifinality in streamflow generation, providing a practical new approach for the assessment of mesoscale modelling. The WATFLOOD model is a conceptually-based distributed hydrological model used for simulating streamflow on mesoscale watersheds. Given the model's intended application to mesoscale hydrology, it remains crucial to ensure conceptualizations are physically representative of the hydrologic cycle and the natural environment. Building upon the existing flowpath-separation module within WATFLOOD, the capability to simulate changes in oxygen-18 through each component of the hydrological cycle is introduced. Masses of heavy-isotope are computed for compartmental storages; compartmental flows transfer flux-weighted portions of isotope mass between storages; and mass outflows from each compartment simultaneously combine to form the resultant channel flow composition. Heavy-isotope compositions are enriched when storages undergo evaporation resulting from the loss of isotopically-depleted vapour described by the well-known Craig & Gordon isotopic fractionation model. The isoWATFLOOD model is forced by oxygen-18 in rain, oxygen-18 in snow, and relative humidity; and requires no additional parameterizations of WATFLOOD. The first mesoscale, continuous simulations of changes in oxygen-18 in streamflow are presented for the remote Fort Simpson basin in Northwest Territories, Canada and for the largely populated Grand River Basin in south western Ontario using the EnSim post-processor software. These simulations shed light on watershed 'hot spots' and the dominant hydrological controls and responses inherent to various regions and dominant hydrological controls.
Towards Simulating Mesoscale Climate Change using an RCM-driven Hydrological Model: Coupling REMOiso and isoWATFLOOD in the Mackenzie River Basin.
This research endeavors to present the first-ever coupling of an isotopic regional climate model (RCM) with an isotopically-enabled hydrological model. This work seeks to improve the reliability of mesoscale (large domain) hydrological prediction while simultaneously assessing mesoscale hydrological variability induced by climate change. The development and testing of a coupled atmospheric-hydrological modeling system is first undertaken and applied to the Mackenzie River Basin (MRB), located in north-western Canada. The WATFLOOD hydrological model, a distributed, partially physically-based model, is directly coupled to the REMO RCM and used to validate RCM-forced streamflow simulations in the MRB. Once validated, the coupled modeling system is used to assess climate-driven changes to streamflow, enabling a direct assessment of the continuous link between climate and land-surface hydrology. Challenges inherent to mesoscale modeling in the MRB surround the validation of model results and the reliability of predictions. Simulation of stable water isotopic (SWI) signals in addition to water quantity can, however, provide a second, more sensitive validation tool for the coupled modeling system. The introduction of the isoWATFLOOD model, an isotopically-enabled version of WATFLOOD, has created the opportunity to simulate both quantity and isotopic composition of streamflow and runoff generation processes. However, the accurate simulation of isotopic variability in streamflow requires the continuous definition of isotopic forcing from climate, which is provided by the isotopically-enabled REMO model, REMOiso. This isotopically-enabled modelling system can therefore be applied in the MRB to simulate isotopic variability of streamflow on a continual basis, providing a second tool for model calibration and validation. Water mass and isotopes simulated in tandem can be used to additionally constrain model simulations to measured isotopic signals in streamflow, facilitating model calibration, parameterization, and validation. The application of isotope modeling is ideal for cold-region river basins such as the MRB where streamflow measurements are typically sporadic and seasonal, and cannot be relied on to validate model simulations. This research seeks to improve the reliability of mesoscale hydrological prediction and will simultaneously enable the assessment of mesoscale hydrological variability as it relates to changes in climate. This work has implications for water use and management within large, remote, and often ungauged Canadian rivers. The proposed integrated modeling methodology facilitates studies of both climate variability and surface hydrology in remote, mesoscale regions.
Isotope mass balance techniques for determining site-specific estimates of water yield: Results from regional acid sensitivity surveys across western Canada
A stable isotope mass balance technique for estimating water yield is applied to a dataset of deuterium and oxygen-18 data from 730 lakes collected as part of provincial surveys across western Canada. These surveys were initiated to constrain water yield estimates to lakes as part of a dynamic critical loadings assessment. The approach uses readily obtainable physical and climatological data combined with analysis of evaporative isotopic enrichment of deuterium and oxygen-18 in lake water as a quantitative tracer of throughflow, lake residency and runoff. The lakes are located in Manitoba, Saskatchewan, Alberta and British Columbia spanning large gradients in climatological parameters and representing very diverse range of landscape types. The results illustrate regional differences on the controls on water yield and lake hydrology but also provide an indication of added value to be gained by incorporating such site-specific measurements within water quality networks. Implications for the role of different terrain types on the regional runoff are discussed.
Stable water isotope surveys of the Mackenzie River between Great Slave Lake and the Mackenzie River Delta
We systematically report oxygen and hydrogen isotope compositions of the main channel and tributaries in the Mackenzie River system, northern Canada. The isotopic survey of main channel exhibits large geographic variations, while isotopic compositions of tributary reflect their catchment characteristics. The tributaries/main- stem isotopic relation in d18O-d2H space indicates a combination of evaporation and mixing processes (between heavy isotopes depleted runoff from Mackenzie Mountains and isotopically enriched water from Great Slave Lake catchment), which dominate the river hydrology in Mackenzie River system. A zone of poor mixing is generally identified along the steep reach of the river below the confluence of Liard River and the main channel. Detailed investigations on time-series of isotopic composition in three key hydrometric stations along the main channel further suggest that snowmelt derived runoff is important to the river system in normal to high flow conditions. However, we also identify isotopic evidence that extremely high-flow events appear to contain a significant proportion of water derived from wetland surface storage. The contribution from wetland storage produces a distinct reversal comparing to the snowmelt-driven isotope-discharge pattern and posts potential impact on chemical and riverine/estuarine biotic processes at times of extremely high runoff.
Landscape-scale assessment of multi-year lakewater balances in the Slave River Delta, NWT
We apply a coupled-isotope tracer model to quantify end-of-thaw-season (fall) lakewater balances in the Slave River Delta (SRD), Canada, during 2003-05, which effectively differentiates the relative importance of hydrological processes across this complex northern freshwater landscape. The model incorporates Great Slave Lake evaporated vapour to the ambient atmospheric vapour pool and is thus tailored to the hydroclimatic setting of the delta, which experiences onshore winds. Results, expressed as evaporation-to-inflow ratios (E/I) for 41 delta lakes, reflect the role of spring break-up flooding and local hydrological setting. Fall E/I ratios for lakes where water balances are dominated by exchange with the Slave River or Great Slave Lake are low (0.06 to 0.53) and do not vary substantially during the three-year monitoring period. E/I ratios for flood-dominated lakes in the active delta are moderate (0.26 to 0.98) and have low inter-annual variability, even in the absence of spring flooding. This suggests that annual flooding during the spring break-up period is not necessary to maintain positive (E/I < 1) water balances in flood-dominated lakes, but multiple years without flooding would clearly lead to greater cumulative evaporation. Fall E/I ratios are generally higher and more variable in evaporation-dominated lakes in the relict delta (0.42 to > 1), although sub-sectors with mature spruce forest produce greater snowmelt runoff that balances open-water vapour loss. Our results indicate that spring inputs (river flooding and snowmelt runoff) are key components of the hydrological evolution of SRD lakes during the open-water season, and distinguish regions of the delta where expected trajectories in river discharge and climate change will likely cause lake level drawdown. Such findings have particular relevance for informed ecosystem management in the Peace-Athabasca-Slave watershed, where unprecedented industrial development is imposing substantial additional pressure on freshwater resources.
Vapour-Phase Processes Control Liquid-Phase Isotope Profiles in Unsaturated Sphagnum Moss
Seminal work in the early 1980s clearly established the basis for predicting patterns of heavy-isotope enrichment of pore waters in soils undergoing evaporation. A key feature of the process under steady-state conditions is the development of stable, convex-upward profiles whose shape is controlled by the balance between downward-diffusing heavy isotopologues concentrated by evaporative enrichment at the surface and the upward capillary flow of bulk water that maintains the evaporative flux. We conducted an analogous experiment to probe evaporation processes within 20-cm columns of unsaturated, living and dead (but undecomposed) Sphagnum moss evaporating under controlled conditions, while maintaining a constant water table. The experiment provided striking evidence of the importance of vapour-liquid mass and isotope exchange in the air-filled pores of the Sphagnum columns, as evidenced by the rapid development of hydrologic and isotopic steady-state within hours, rather than days, i.e., an order of magnitude faster than possible by liquid-phase processes alone. This is consistent with the notion that vapour-phase processes effectively "short-circuit" mass and isotope fluxes within the Sphagnum columns, as proposed also in recent characterizations of water dynamics in transpiring leaves. Additionally, advection-diffusion modelling of our results supports independent estimates of the effective liquid-phase diffusivities of the respective heavy water isotopologues, 2.380 x 10-5 cm2 s-1 for 1H1H18O and 2.415 x 10-5 cm2 s-1 for 1H2H16O, which are in notably good agreement with the "default" values that are typically assumed in soil and plant water studies.