Publications

Changes in long-term, montane actual evapotranspiration (ET) in response to climate change could impact future water supplies and forest species composition. For scenarios of atmospheric warming, predicted changes in long-term ET tend to differ between studies using space-for-time substitution (STS) models and integrated watershed models, and the influence of spatially varying factors on these differences is unclear. To examine this, we compared warming-induced (+2 to +6 °C) changes in ET simulated by an STS model and an integrated watershed model across zones of elevation, substrate available water capacity, and slope in the snow-influenced upper San Joaquin River watershed, Sierra Nevada, USA. We used the Soil Water and Assessment Tool (SWAT) for the watershed modeling and a Budyko-type relationship for the STS modeling. Spatially averaged increases in ET from the STS model increasingly surpassed those from the SWAT model in the higher elevation zones of the watershed, resulting in 2.3–2.6 times greater values from the STS model at the watershed scale. In sparse, deep colluvium or glacial soils on gentle slopes, the SWAT model produced ET increases exceeding those from the STS model. However, watershed areas associated with these conditions were too localized for SWAT to produce spatially averaged ET-gains comparable to the STS model. The SWAT model results nevertheless demonstrate that such soils on high-elevation, gentle slopes will form ET “hot spots” exhibiting disproportionately large increases in ET, and concomitant reductions in runoff yield, in response to warming. Predicted ET responses to warming from STS models and integrated watershed models may, in general, substantially differ (e.g., factor of 2–3) for snow-influenced watersheds exhibiting an elevational gradient in substrate water holding capacity and slope. Long-term water supplies in these settings may therefore be more resilient to warming than STS model predictions would suggest.

Delineating pollutant reactive transport pathways that connect local land use patterns to surface water is an important goal. This work illustrates high-resolution river mapping of salinity or specific conductance (SC) and nitrate ( ) as a potential part of achieving this goal. We observed longitudinal river SC and nitrate distributions using high-resolution synoptic in situ sensing along the lower Merced River (38 river km) in Central California (USA) from 2010 to 2012. We calibrated a distributed groundwater-surface water (GW-SW) discharge model for a conservative solute using 13 synoptic SC sampling events at flows ranging from 1.3 to 31.6 m3 s-1. Nitrogen loads ranged from 0.3 to 1.6 kg N d-1 and were greater following an extended high flow period during a wet winter. Applying the distributed GW-SW discharge estimates to a simplistic reactive nitrate transport model, the model reproduced observed river nitrate distribution well (RRMSE = 5–21%), with dimensionless watershed-averaged nitrate removal (kt) ranging from 0 to 0.43. Estimates were uncertain due to GW nitrate data variability, but the resulting range was consistent with prior removal estimates. At the segment scale, estimated GW-SW nitrate loading ranged from 0 to 17 g s-1 km-1. Local loading peaked near the middle of the study reach, a location that coincides with a shallow clay lens and with confined animal feed operations in close proximity to the river. Overall, the results demonstrate the potential for high-resolution synoptic monitoring to support GW-SW modeling efforts aimed at understanding and managing nonpoint source pollution.

Summary The role of hydrogeology in mediating long-term changes in mountain streamflow, resulting from reduced snowfall in a potentially warmer climate, is currently not well understood. We explore this by simulating changes in stream discharge and evapotranspiration from a mid-elevation, 1-km2 catchment in the southern Sierra Nevada of California (USA) in response to reduced snowfall under warmer conditions, for a plausible range in subsurface hydrologic properties. Simulations are performed using a numerical watershed model, the Penn State Integrated Hydrologic Model (PIHM), constrained by observations from a meteorological station, stream gauge, and eddy covariance tower. We predict that the fraction of precipitation occurring as snowfall would decrease from approximately 47% at current conditions to 25%, 12%, and 5% for air temperature changes of +2, +4, and +6°C. For each of these warming scenarios, changes in mean annual discharge and evapotranspiration simulated by the different plausible soil models show large ranges relative to averages, with coefficients of variation ranging from -3 to 3 depending on warming scenario. With warming and reduced snowfall, substrates with greater storage capacity show less soil moisture limitation on evapotranspiration during the late spring and summer, resulting in greater reductions in annual stream discharge. These findings indicate that the hydrologic response of mountain catchments to atmospheric warming and reduced snowfall may substantially vary across elevations with differing soil and regolith properties, a relationship not typically accounted for in approaches relying on space-for-time substitution. An additional implication of our results is that model simulations of annual stream discharge in response to snowfall-to-rainfall transitions may be relatively uncertain for study areas where subsurface properties are not well constrained.

Calibration of watershed models to the shape of the base flow recession curve is a way to capture the important relationship between groundwater discharge and subsurface water storage in a catchment. In some montane Mediterranean regions, such as the midelevation Providence Creek catchment in the southern Sierra Nevada of California (USA), nearly all base flow recession occurs after snowmelt, and during this time evapotranspiration (ET) usually exceeds base flow. We assess the accuracy to which watershed models can be calibrated to ET-dominated base flow recession in Providence Creek, both in terms of fitting a discharge time-series and realistically capturing the observed discharge-storage relationship for the catchment. Model parameters estimated from calibrations to ET-dominated recession are compared to parameters estimated from reference calibrations to base flow recession with ET-effects removed (“potential recession”). We employ the Penn State Integrated Hydrologic Model (PIHM) for simulations of base flow and ET, and methods that are otherwise general in nature. In models calibrated to ET-dominated recession, simulation errors in ET and the targeted relationship for recession (-dQ/dt versus Q) contribute substantially (up to 57% and 46%, respectively) to overestimates in the discharge-storage differential, defined as d(lnQ)/dS, relative to that derived from water flux observations. These errors result in overestimates of deep-subsurface hydraulic conductivity in models calibrated to ET-dominated recession, by up to an order of magnitude, relative to reference calibrations to potential recession. These results illustrate a potential opportunity for improving model representation of discharge-storage dynamics by calibrating to the shape of base flow recession after removing the complicating effects of ET.

Measuring CO2 concentrations and fluxes is key to the evaluation of terrestrial ecosystem carbon dynamics. Both the high cost and low portability of currently available sensors and field instruments are constraints to achieving adequate spatial and temporal coverage in characterizing ecosystem CO2 fluxes and point concentrations. We used commercially available, low-cost and low-power non-dispersive infrared (NDIR) CO2 sensors to develop: (i) a soil CO2 efflux system (K-33 ELG sensor, 0–1% or 10 000 ppm(v)) and (ii) a point CO2 concentration system (K-33 BLG sensor, 0 to 30% or 300 000 ppm(v)). We first calibrated the sensors against benchmark instruments (LI-COR LI-6400 efflux system and Vaisala GMP343 probe). The K-33 ELG sensor tracked the LI-6400 well during a steady reduction from \~ 4000 ppm to background CO2 levels (RMSE = 176 ppm). The K-33 BLG point sensor were less favourable (RMSE = 424 ppm) because of its broad range of detection, but were suitable for proof-of-concept testing at elevated CO2 levels (\textgreater 1500 ppm). In field tests of soil CO2 efflux on locations with and without leaf litter, the K-33 efflux system yielded mean surface efflux rate values significantly lower (by \~ 27%) than those obtained with a LI-COR LI-6400 for the bare (P = 0·006) and litter (P = 0·002). We explain the systematic difference in terms of flux chamber geometry and potential leakage from the prototypical system. In a test on leaf cutter ant nest vents, the K-33 BLG point system yielded comparable spatial and temporal patterns and slightly higher (\~ 10–15%) CO2 concentrations in comparison with a Vaisala GMP343 probe. The results provide proof-of-concept for the use of two low-cost, portable CO2 sensing systems to enable terrestrial ecologists to substantially improve the characterization of CO2 fluxes and concentrations in heterogeneous environments.

High Resolution Synoptic Salinity Mapping To Identify Groundwater–Surface Water Discharges in Lowland Rivers

Variability of river properties such as temperature, velocity, dissolved oxygen (DO), and light at small scales (centimeters to meters) can play an important role in the local exchanges of energy and mass. We hypothesize that significant transverse cross-sectional DO variation is observable within a river. Such variation may influence conventional single-station metabolic rate (primary production and respiration) estimates with respect to DO probe location, and reveal important connections between physical and biogeochemical processes and their drivers in rivers. Using a mobile sensor system, we measured river properties across a bend in the lower Merced River in Central California under stationary flow conditions in April and September. Cross-sectional temperature, DO, and chlorophyll-a concentrations exhibited modest but significant gradients, which varied in magnitude and direction on a diel basis. The spatiotemporal variation was consistent with reach geomorphology and incident light patterns. Gross primary production (GPP), community respiration (CR24), and net ecosystem production (NEP) rates estimates derived from local DO and temperature time series varied by 3–10% over the river cross section, with greater variation in late summer. The presence of transverse metabolic rate gradients in this relatively simple reach implies the existence of substantial gradients in more complex river regimes, such as those spanning distinctively different microhabitats, transient storage zones, and related distributed biogeochemical zones.