Diurnal dynamics of streamflow in an upland forested micro‐watershed during short precipitation‐free periods is altered by tree sap flow

Diurnal variations in streamflow are becoming acknowledged as a way of analysing how changing climatic conditions and land use affects watersheds but also as a way to understand watersheds as a whole. Yet not many studies from uplands below 900 mm mean annual precipitation zone are available from European countries. During the 2012 growing season, a sampling campaign took place in an upland forested micro‐watershed, Czech Republic (65 ha). Tree sap flow, rainfall and temperature were measured continuously, while streamflow at the discharge point and soil moisture were estimated from short‐term measurements. Short precipitation‐free periods lasting several days were identified for evaluation of trends in diurnal dynamics of both sap flow and streamflow. The results demonstrated that during these periods, the main factor altering streamflow was almost exclusively tree sap flow. A decrease in streamflow was observed during the day and an increase at night. The decline in sap flow after sunset was accompanied by a continuous increase in streamflow throughout the night up to its initial maximum in the morning. The amplitude in diurnal variations reached 18%. The observed time lag between the diurnal variations of sap flow and streamflow was approximately 2 h. Relatively low changes in diurnal dynamics of streamflow pointed out a strong regulatory role of the forest in buffering water discharge from the catchment. Copyright © 2015 John Wiley & Sons, Ltd.     

Distribution of surface imperviousness in small urban catchments predicts runoff peak flows and stream flashiness

Urban growth is a global phenomenon, and the associated impacts on hydrology from land development are expected to increase, especially in peri‐urban catchments. It is well understood that greater peak flows and higher stream flashiness are associated with increased surface imperviousness and storm location. However, the effect of the distribution of impervious areas on runoff peak flow response and stream flashiness of peri‐urban catchments has not been well studied. In this study, a new geometric index, Relative Nearness of Imperviousness to the Catchment Outlet (RNICO), is defined to correlate imperviousness distribution of peri‐urban catchments with runoff peak flows and stream flashiness. Study sites include 21 suburban catchments in New York representing a range of drainage area from 5 to 189 km² and average imperviousness from 10% to 48%. On the basis of RNICO, all development patterns are divided into 3 classes: upstream, centralized, and downstream. Results showed an obvious increase in runoff peak flows and decrease in time to peak when moving from upstream to centralized and downstream urbanization classes. This indicates that RNICO is an effective tool for classifying urban development patterns and for macroscale understanding of the hydrologic behavior of small peri‐urban catchments, despite the complexity of urban drainage systems. We also found that the impact of impervious distribution on runoff peak flows and stream flashiness decreases with catchment scale. For small catchments (A < 40 km²), RNICO was strongly correlated with the average (R² = .95) and maximum (R² = .91) gaged peak flows due to the relatively efficient subsurface routing through stormwater and sewer networks. Furthermore, the Richards–Baker stream flashiness index in small catchments was positively correlated with fractional impervious area (R² = .84) and RNICO (R² = .87). For large catchments (A > 40 km²), the impact of impervious surface distribution on peak flows and stream flashiness was negligible due to the complex drainage network and great variability in travel times. This study emphasizes the need for greater monitoring of discharge in small peri‐urban catchments to support flood prediction at the local scale.     

Landscape influence on small‐scale water temperature variations in a moorland catchment

We monitored temperatures in stream water, groundwater and riparian wetland surface water over 18 months in a 3.2‐km² moorland catchment in the Scottish Highlands. The stream occupies a glaciated valley, aligned east–west. It has three main headwater tributaries with a large north facing catchment, a south facing catchment and the smallest east facing headwater. The lower catchment sampling locations begin after the convalescence of all three headwaters. Much of the stream network is fringed by riparian peatlands. Stream temperatures are mainly regulated by energy exchanges at the air–water interface. However, they are also influenced by inflows from the saturated riparian zone, where surface water source areas are strongly connected with the stream network. Consequently, the spatial distribution of stream temperatures exhibits limited variability. Nevertheless, there are significant summer differences between the headwaters, despite their close proximity to each other. This is consistent with aspect (and incident radiation), given the south and east facing headwaters having higher temperatures. The largest, north‐facing sub‐catchment shows lower summer diurnal temperature variability, suggesting that lower radiation inputs dampen temperature extremes. Whilst stream water temperature regimes in the lower catchment exhibit little change along a 1‐km reach, they are similar to those in the largest headwater; probably reflecting size and comparable catchment aspect and hydrological flow paths. Our results suggest that different parts of the channel network and its connected wetlands have contrasting sensitivity to higher summer temperatures. This may be important in land management strategies designed to mitigate the impacts of projected climatic warming. Copyright © 2015 John Wiley & Sons, Ltd.     

Combined influences of irrigation diversions and associated subsurface return flows on river temperature in a semi-arid region

Irrigation activities alter water distribution and storage in arid and semi-arid regions worldwide. The removal of water from streams can drastically impact instream flows. However, irrigation water conveyance and application onto fields can create surface and subsurface hydrologic connections, or lateral inflows, that return some of this diverted water back to streams. Prior research has shown the impact of surface water diversions from streams on downstream warming that increases stress on aquatic species. However, the combined effects of flow depletion and irrigation-enhanced lateral inflows on stream temperature and river ecosystems remains poorly studied. To further understand these relationships, we combined intensive field monitoring over three irrigation seasons and thermal aerial imagery to identify irrigation-enhanced subsurface lateral inflow locations and evaluate their effects on stream flow and temperature patterns over a 2.5-km highly depleted study reach. Considering variable hydrology, weather, flow diversions, channel geometry and lateral inflows, we found irrigation-enhanced lateral inflows were the likely explanation for buffered longitudinal and diel warming patterns that prevented stressful or lethal thermal conditions for brown trout. These localized temperature effects were more pronounced in drier years, under high diversion rates and during high solar radiation intensity. We also found that lateral inflows corresponded with greater spatial variability of stream temperatures and potential thermal refugia. Study results illustrate the potential ecological consequences of reducing irrigation-enhanced lateral inflows and highlight the importance of hydrologic monitoring in irrigated arid river valleys. The role and preservation of these lateral inflows should be considered in water resources management related to irrigation efficiency and environmental flows.     

Watershed studies at the Hubbard Brook Experimental Forest: Building on a long legacy of research with new approaches and sources of data

The Hubbard Brook Experimental Forest (HBEF) was established in 1955 by the U.S. Department of Agriculture, Forest Service out of concerns about the effects of logging increasing flooding and erosion. To address this issue, within the HBEF hydrological and micrometeorological monitoring was initiated in small watersheds designated for harvesting experiments. The Hubbard Brook Ecosystem Study (HBES) originated in 1963, with the idea of using the small watershed approach to study element fluxes and cycling and the response of forest ecosystems to disturbances, such as forest management practices and air pollution. Early evidence of acid rain was documented at the HBEF and research by scientists at the site helped shape acid rain mitigation policies. New lines of investigation at the HBEF have built on the long legacy of watershed research resulting in a shift from comparing inputs and outputs and quantifying pools and fluxes to a more mechanistic understanding of ecosystem processes within watersheds. For example, hydropedological studies have shed light on linkages between hydrologic flow paths and soil development that provide valuable perspective for managing forests and understanding stream water quality. New high frequency in situ stream chemistry sensors are providing insights about extreme events and diurnal patterns that were indiscernible with traditional weekly sampling. Additionally, tools are being developed for visual and auditory data exploration and discovery by a broad audience. Given the unprecedented environmental change that is occurring, data from the small watersheds at the HBEF are more relevant now than ever and will continue to serve as a basis for sound environmental decision-making.     

Relative contributions of stream concentration,stream discharge and shoreline load to base cation trends in Red Chalk and Harp lakes,south‐central Ontario,Canada

We evaluated the importance and contributions of changes in stream base cation concentration, stream discharge and lake shoreline load to changes in base cation concentrations in two small inland lakes in south‐central Ontario. The shoreline load from ungauged drainage areas was calculated with a mass balance equation. An evaluation method based on the partial Mann–Kendall test quantified the relative contributions of these three explanatory drivers to the observed trends in lake concentration of Ca, K, Mg and Na. Over a 29‐year period (1978–2006) at Red Chalk and Harp lakes, declines in stream concentration and discharge were correlated with the declines in lake concentrations of all base cations (except for Na in Harp Lake), with contributions of 21–81% from stream concentration and 12–58% from discharge, whereas the shoreline load had little contribution to observed trends. The observed unusual increase in Na concentration at Harp Lake was correlated with the increase in stream Na concentration and additional load from road salts, with the shoreline load contributing a substantial 37%. These results may be applicable to numerous inland lakes with similar site conditions, including lakes found on the Canadian Shield. Copyright © 2015 John Wiley & Sons, Ltd.     

Short‐term stream water temperature observations permit rapid assessment of potential climate change impacts

Assessment of potential climate change impacts on stream water temperature (T_s) across large scales remains challenging for resource managers because energy exchange processes between the atmosphere and the stream environment are complex and uncertain, and few long‐term datasets are available to evaluate changes over time. In this study, we demonstrate how simple monthly linear regression models based on short‐term historical T_s observations and readily available interpolated air temperature (T_a) estimates can be used for rapid assessment of historical and future changes in T_s. Models were developed for 61 sites in the southeastern USA using ≥18 months of observations and were validated at sites with longer periods of record. The T_s models were then used to estimate temporal changes in T_s at each site using both historical estimates and future T_a projections. Results suggested that the linear regression models adequately explained the variability in T_s across sites, and the relationships between T_s and T_a remained consistent over 37 years. We estimated that most sites had increases in historical annual mean T_s between 1961 and 2010 (mean of +0.11 °C decade^?1). All 61 sites were projected to experience increases in T_s from 2011 to 2060 under the three climate projections evaluated (mean of +0.41 °C decade^?1). Several of the sites with the largest historical and future T_s changes were located in ecoregions home to temperature‐sensitive fish species. This methodology can be used by resource managers for rapid assessment of potential climate change impacts on stream water temperature. Copyright © 2014 John Wiley & Sons, Ltd.     

Influence of turbidity and aeration on the albedo of mountain streams

Stream surface albedo plays a key role in the energy balance of rivers and streams that are exposed to direct solar radiation. Most physically based analyses and models have incorporated a constant stream albedo between 0.03 and 0.10, based primarily on measurements from low‐gradient streams with low suspended sediment concentrations. However, albedo should vary with solar elevation angle, suspended sediment concentration, aeration, and fraction of direct versus diffuse radiation. The objective of this study was to quantify the dependence of albedo of mountain streams on the controlling factors and to develop a predictive model for use in physically based analysis and modelling of stream temperature, especially for future climate and land‐use scenarios. Stream surface albedo was measured at nine sites with a variety of gradients and suspended sediment characteristics in the southern Coast Mountains of British Columbia, Canada. As expected, albedo of low‐gradient, non‐white water (flatwater) streams increased with solar elevation angle, suspended sediment concentration, and proportion of diffuse to direct solar radiation, ranging between 0.025 during cloudy periods over clear water to 0.25 for turbid water at elevation angles of less than 20°. Albedo was enhanced in steep reaches or at channel steps and cascades where flow was visibly aerated, with a range of 0.09 to 0.33. In clear weather, albedo exhibited notable diurnal variability at flatwater sampling sites. For example, during late summer, surface albedo typically fluctuated between 0.08 and 0.15 on a daily basis at a flatwater site on the highly turbid, glacier‐fed Lillooet River. Multiple regression models explained approximately 60% and 40% of the variance under cross validation for flatwater and white water data subsets, respectively, with corresponding root mean square errors of approximately 0.02 and 0.06.     

Towards quantifying the negative feedback regulation of peatland evaporation to drought

The frequency and intensity of drought is projected to increase within the boreal region under future climatic conditions. Peatlands are widely considered to regulate water loss under drought conditions, increasing surface resistance (r_s) and reducing evaporative losses. This maintains peat moisture content, increasing the resilience of these globally important carbon stores. However, the magnitude and form of this important negative feedback response remains uncertain. To address this, we monitored the response of r_s to drought within four peat cores under controlled meteorological conditions. When the water‐table was dropped to a depth of 0.30 m and the humidity reduced to ≤40%, a step shift in r_s from ~50 s m^‐1 up to 1000 s m^‐1 was observed within burned and unburned peat, which virtually shuts down evaporation, limiting water loss. We show that measured near‐surface tension cannot be used to directly calculate this transition in peat surface resistance. However, empirical relationships that account for strong vertical variations in tension through the near‐surface and/or disequilibrium between pore air and near‐surface pore water pressure provide the potential to incorporate this negative feedback response into peatland ecohydrological models. Further observations are necessary to examine this response under dynamic atmospheric conditions. We suggest that the link between surface temperature and evaporation provides potential to further examine this feedback in either burned peatlands or peatlands with a low vascular vegetation cover. Copyright © 2013 John Wiley & Sons, Ltd.     

Deep learning approaches for improving prediction of daily stream temperature in data-scarce,unmonitored, and dammed basins

Basin-centric long short-term memory (LSTM) network models have recently been shown to be an exceptionally powerful tool for stream temperature (T_s) temporal prediction (training in one period and predicting in another period at the same sites). However, spatial extrapolation is a well-known challenge to modelling T_s and it is uncertain how an LSTM-based daily T_s model will perform in unmonitored or dammed basins. Here we compiled a new benchmark dataset consisting of >400 basins across the contiguous United States in different data availability groups (DAG, meaning the daily sampling frequency) with and without major dams, and studied how to assemble suitable training datasets for predictions in basins with or without temperature monitoring. For prediction in unmonitored basins (PUB), LSTM produced a root-mean-square error (RMSE) of 1.129°C and an R² of 0.983. While these metrics declined from LSTM's temporal prediction performance, they far surpassed traditional models' PUB values, and were competitive with traditional models' temporal prediction on calibrated sites. Even for unmonitored basins with major reservoirs, we obtained a median RMSE of 1.202°C and an R² of 0.984. For temporal prediction, the most suitable training set was the matching DAG that the basin could be grouped into (for example, the 60% DAG was most suitable for a basin with 61% data availability). However, for PUB, a training dataset including all basins with data was consistently preferred. An input-selection ensemble moderately mitigated attribute overfitting. Our results indicate there are influential latent processes not sufficiently described by the inputs (e.g., geology, wetland covers), but temporal fluctuations can still be predicted well, and LSTM appears to be a highly accurate T_s modelling tool even for spatial extrapolation.