Spatial dimensions of water quality value in New England river networks.

Households' willingness to pay (WTP) for water quality improvements-representing their economic value-depends on where improvements occur. Households often hold higher values for improvements close to their homes or iconic areas. Are there other areas where improvements might hold high value to individual households, do effects on WTP vary by type of improvement, and can these areas be identified even if they are not anticipated by researchers? To answer these questions, we integrated a water quality model and map-based, interactive choice experiment to estimate households' WTP for water quality improvements throughout a river network covering six New England states. The choice experiment was implemented using a push-to-web survey over a sample of New England households. Voting scenarios used to elicit WTP included interactive geographic information system (GIS) maps that illustrated three water quality measures at various zoom levels across the study domain. We captured data on how respondents maneuvered through these maps prior to answering the value-eliciting questions. Results show that WTP was influenced by regionwide quality improvements and improvements surrounding each respondent's home, as anticipated, but also by improvements in individualized locations identifiable via each respondent's map interactions. These spatial WTP variations only appear for low-quality rivers and are focused around particular areas of New England. The study shows that dynamic map interactions can convey salient information for WTP estimation and that predicting spatial WTP heterogeneity based primarily on home or iconic locations, as typically done, may overlook areas where water quality has high value.

US climate policy yields water quality cobenefits in the Mississippi Basin and Gulf of Mexico.

We utilize a coupled economy-agroecology-hydrology modeling framework to capture the cascading impacts of climate change mitigation policy on agriculture and the resulting water quality cobenefits. We analyze a policy that assigns a range of United States government's social cost of carbon estimates ($51, $76, and $152/ton of CO2-equivalents) to fossil fuel-based CO2 emissions. This policy raises energy costs and, importantly for agriculture, boosts the price of nitrogen fertilizer production. At the highest carbon price, US carbon emissions are reduced by about 50%, and nitrogen fertilizer prices rise by about 90%, leading to an approximate 15% reduction in fertilizer applications for corn production across the Mississippi River Basin. Corn and soybean production declines by about 7%, increasing crop prices by 6%, while nitrate leaching declines by about 10%. Simulated nitrate export to the Gulf of Mexico decreases by 8%, ultimately shrinking the average midsummer area of the Gulf of Mexico hypoxic area by 3% and hypoxic volume by 4%. We also consider the additional benefits of restored wetlands to mitigate nitrogen loading to reduce hypoxia in the Gulf of Mexico and find a targeted wetland restoration scenario approximately doubles the effect of a low to moderate social cost of carbon. Wetland restoration alone exhibited spillover effects that increased nitrate leaching in other parts of the basin which were mitigated with the inclusion of the carbon policy. We conclude that a national climate policy aimed at reducing greenhouse gas emissions in the United States would have important water quality cobenefits.

A hybrid deep learning approach to predict hourly riverine nitrate concentrations using routine monitored data.

With high-frequency data of nitrate (NO3-N) concentrations in waters becoming increasingly important for understanding of watershed system behaviors and ecosystem managements, the accurate and economic acquisition of high-frequency NO3-N concentration data has become a key point. This study attempted to use coupled deep learning neural networks and routine monitored data to predict hourly NO3-N concentrations in a river. The hourly NO3-N concentration at the outlet of the Oyster River watershed in New Hampshire, USA, was predicted through neural networks with a hybrid model architecture coupling the Convolutional Neural Networks and the Long Short-Term Memory model (CNN-LSTM). The routine monitored data (the river depth, water temperature, air temperature, precipitation, specific conductivity, pH and dissolved oxygen concentrations) for model training were collected from a nested high-frequency monitoring network, while the high-frequency NO3-N concentration data obtained at the outlet were not included as inputs. The whole dataset was separated into training, validation, and testing processes according to the ratio of 5:3:2, respectively. The hybrid CNN-LSTM model with different input lengths (1d, 3d, 7d, 15d, 30d) displayed comparable even better performance than other studies with lower frequencies, showing mean values of the Nash-Sutcliffe Efficiency 0.60-0.83. Models with shorter input lengths demonstrated both the higher modeling accuracy and stability. The water level, water temperature and pH values at monitoring sites were main controlling factors for forecasting performances. This study provided a new insight of using deep learning networks with a coupled architecture and routine monitored data for high-frequency riverine NO3-N concentration forecasting and suggestions about strategies about variable and input length selection during preprocessing of input data.

Spatial and temporal heterogeneity of methane ebullition in lowland headwater streams and the impact on sampling design.

Headwater streams are known sources of methane (CH4) to the atmosphere, but their contribution to global scale budgets remains poorly constrained. While efforts have been made to better understand diffusive fluxes of CH4 in streams, much less attention has been paid to ebullitive fluxes. We examine the temporal and spatial heterogeneity of CH4 ebullition from four lowland headwater streams in the temperate northeastern United States over a 2-yr period. Ebullition was observed in all monitored streams with an overall mean rate of 1.00 ± 0.23 mmol CH4 m-2 d-1, ranging from 0.01 to 1.79 to mmol CH4 m-2 d-1 across streams. At biweekly timescales, rates of ebullition tended to increase with temperature. We observed a high degree of spatial heterogeneity in CH4 ebullition within and across streams. Yet, catchment land use was not a simple predictor of this heterogeneity, and instead patches scale variability weakly explained by water depth and sediment organic matter content and quality. Overall, our results support the prevalence of CH4 ebullition from streams and high levels of variability characteristic of this process. Our findings also highlight the need for robust temporal and spatial sampling of ebullition in lotic ecosystems to account for this high level of heterogeneity, where multiple sampling locations and times are necessary to accurately represent the mean rate of flux in a stream. The heterogeneity observed likely indicates a complex set of drivers affect CH4 ebullition from streams which must be considered when upscaling site measurements to larger spatial scales.

Coastal eutrophication as a driver of salt marsh loss.

Salt marshes are highly productive coastal wetlands that provide important ecosystem services such as storm protection for coastal cities, nutrient removal and carbon sequestration. Despite protective measures, however, worldwide losses of these ecosystems have accelerated in recent decades. Here we present data from a nine-year whole-ecosystem nutrient-enrichment experiment. Our study demonstrates that nutrient enrichment, a global problem for coastal ecosystems, can be a driver of salt marsh loss. We show that nutrient levels commonly associated with coastal eutrophication increased above-ground leaf biomass, decreased the dense, below-ground biomass of bank-stabilizing roots, and increased microbial decomposition of organic matter. Alterations in these key ecosystem properties reduced geomorphic stability, resulting in creek-bank collapse with significant areas of creek-bank marsh converted to unvegetated mud. This pattern of marsh loss parallels observations for anthropogenically nutrient-enriched marshes worldwide, with creek-edge and bay-edge marsh evolving into mudflats and wider creeks. Our work suggests that current nutrient loading rates to many coastal ecosystems have overwhelmed the capacity of marshes to remove nitrogen without deleterious effects. Projected increases in nitrogen flux to the coast, related to increased fertilizer use required to feed an expanding human population, may rapidly result in a coastal landscape with less marsh, which would reduce the capacity of coastal regions to provide important ecological and economic services.

Controls of Chloride Loading and Impairment at the River Network Scale in New England.

Chloride contamination of rivers due to nonpoint sources is increasing throughout developed temperate regions due to road salt application in winter. We developed a river-network model of chloride loading to watersheds to estimate road salt application rates and investigated the meteorological factors that control riverine impairment by chloride at concentrations above thresholds protective of aquatic organisms. Chloride loading from road salt was simulated in the Merrimack River watershed in New Hampshire, which has gradients in development density. After calibration to a regional network of stream chloride data, the model captured the distribution of regional discharge and chloride observations with efficiencies of 93 and 75%, respectively. The estimate of road salt application is within uncertainties of inventoried estimates of road salt loading and is 122 to 214% greater than recommended targets. Model predictions of chloride showed seasonal variation in chloride concentrations despite a large groundwater storage pool. Interannual variation of mean summer chloride concentration near the outlet varied up to 18%, and the total river length exceeding impairment thresholds varied 12%. Annual snowfall, which drives road salt loading, correlated with chloride impairment only in headwater streams, whereas concentration variability at the outlet was driven primarily by dilution from clean runoff-draining undeveloped forested areas of the watershed. The role of summer meteorology complicates the protection of freshwater systems from chloride contamination.

Characterizing storm-event nitrate fluxes in a fifth order suburbanizing watershed using in situ sensors.

Land use influences the distribution of nonpoint nitrogen (N) sources in urbanizing watersheds and storm events interact with these heterogeneous sources to expedite N transport to aquatic systems. In situ sensors provide high frequency and continuous measurements that may reflect storm-event N variability more accurately compared to grab samples. We deployed sensors from April to December 2011 in a suburbanizing watershed (479 km2) to characterize storm-event nitrate-N (NO3-N) and conductivity variability. NO3-N concentrations exhibited complex patterns both within and across storms and shifted from overall dilution (source limitation) before summer baseflows to subsequent periods of flushing (transport limitation). In contrast, conductivity generally diluted with increasing runoff. Despite diluted NO3-N concentrations, NO3-N fluxes consistently increased with flow. Sensor flux estimates for the entire deployment period were similar to estimates derived from weekly and monthly grab samples. However, significant differences in flux occurred at monthly time scales, which may have important implications for understanding impacts to temporally sensitive receiving waters. Evidence of both supply (nutrient-poor) and transport (nutrient-rich) limitation patterns during storms is consistent with watersheds undergoing land use transitions. Tracking shifts in these patterns could indicate N accumulation in developing watersheds and help identify mitigation opportunities prior to N impairment.

Nitrous oxide emission from denitrification in stream and river networks.

Nitrous oxide (N(2)O) is a potent greenhouse gas that contributes to climate change and stratospheric ozone destruction. Anthropogenic nitrogen (N) loading to river networks is a potentially important source of N(2)O via microbial denitrification that converts N to N(2)O and dinitrogen (N(2)). The fraction of denitrified N that escapes as N(2)O rather than N(2) (i.e., the N(2)O yield) is an important determinant of how much N(2)O is produced by river networks, but little is known about the N(2)O yield in flowing waters. Here, we present the results of whole-stream (15)N-tracer additions conducted in 72 headwater streams draining multiple land-use types across the United States. We found that stream denitrification produces N(2)O at rates that increase with stream water nitrate (NO(3)(-)) concentrations, but that <1% of denitrified N is converted to N(2)O. Unlike some previous studies, we found no relationship between the N(2)O yield and stream water NO(3)(-). We suggest that increased stream NO(3)(-) loading stimulates denitrification and concomitant N(2)O production, but does not increase the N(2)O yield. In our study, most streams were sources of N(2)O to the atmosphere and the highest emission rates were observed in streams draining urban basins. Using a global river network model, we estimate that microbial N transformations (e.g., denitrification and nitrification) convert at least 0.68 Tg·y(-1) of anthropogenic N inputs to N(2)O in river networks, equivalent to 10% of the global anthropogenic N(2)O emission rate. This estimate of stream and river N(2)O emissions is three times greater than estimated by the Intergovernmental Panel on Climate Change.

Prediction of riverine daily minimum dissolved oxygen concentrations using hybrid deep learning and routine hydrometeorological data.

Dissolved oxygen (DO) depletion is a severe threat to aquatic ecosystems. Hence, using easily available routine hydrometeorological variables without DO as inputs to predict the daily minimum DO concentration in rivers has huge practical significance in the watershed management. The daily minimum DO concentrations at the outlet of the Oyster River watershed in New Hampshire, USA, were predicted by a set of deep learning neural networks using meteorological data and high-frequency water level, water temperature, and specific conductance (CTD) data measured within the watershed. The dependent variable, DO concentration, was measured at the outlet. From April 2013 to March 2018, the dataset was separated into training, validation, and test portions with a ratio of 5:3:3. A Long Short-Term Memory (LSTM) model and a hybrid Convolutional Neural Networks (CNN-LSTM) model were trained and evaluated for predicting the daily minimum DO concentration. The hybrid CNN-LSTM model exhibited the better predictive stability but the comparable accuracy (the mean R2 value = 0.865) compared with the pure LSTM model (the mean R2 value = 0.839). The model performance (both the stability and accuracy) was improved by aggregating the input data frequency from 15 min of raw data to 24 h. Likewise, the modeling performance didn't benefit from including 'forecasted' meteorological data at the predicted time step in the input dataset. This study provided an efficient and low-cost approach to predict the water quality in rivers and streams to realize the scientific watershed management.

Existing wetland conservation programs miss nutrient reduction targets.

Restoring wetlands will reduce nitrogen contamination from excess fertilization but estimates of the efficacy of the strategy vary widely. The intervention is often described as effective for reducing nitrogen export from watersheds to mediate bottom-level hypoxia threatening marine ecosystems. Other research points to the necessity of applying a suite of interventions, including wetland restoration to mitigate meaningful quantities of nitrogen export. Here, we use process-based physical modeling to evaluate the effects of two hypothetical, but plausible large-scale wetland restoration programs intended to reduce nutrient export to the Gulf of Mexico. We show that full adoption of the two programs currently in place can meet as little as 10% to as much as 60% of nutrient reduction targets to reduce the Gulf of Mexico dead zone. These reductions are lower than prior estimates for three reasons. First, net storage of leachate in the subsurface precludes interception and thereby dampens the percent decline in nitrogen export caused by the policy. Unlike previous studies, we first constrained riverine fluxes to match observed fluxes throughout the basin. Second, the locations of many restorable lands are geographically disconnected from heavily fertilized croplands, limiting interception of runoff. Third, daily resolution of the model simulations captured the seasonal and stormflow dynamics that inhibit wetland nutrient removal because peak wetland effectiveness does not coincide with the timing of nutrient inputs. To improve the health of the Gulf of Mexico efforts to eliminate excess nutrient, loading should be implemented beyond the field-margin wetland strategies investigated here.

Superlinear scaling of riverine biogeochemical function with watershed size.

River networks regulate carbon and nutrient exchange between continents, atmosphere, and oceans. However, contributions of riverine processing are poorly constrained at continental scales. Scaling relationships of cumulative biogeochemical function with watershed size (allometric scaling) provide an approach for quantifying the contributions of fluvial networks in the Earth system. Here we show that allometric scaling of cumulative riverine function with watershed area ranges from linear to superlinear, with scaling exponents constrained by network shape, hydrological conditions, and biogeochemical process rates. Allometric scaling is superlinear for processes that are largely independent of substrate concentration (e.g., gross primary production) due to superlinear scaling of river network surface area with watershed area. Allometric scaling for typically substrate-limited processes (e.g., denitrification) is linear in river networks with high biogeochemical activity or low river discharge but becomes increasingly superlinear under lower biogeochemical activity or high discharge, conditions that are widely prevalent in river networks. The frequent occurrence of superlinear scaling indicates that biogeochemical activity in large rivers contributes disproportionately to the function of river networks in the Earth system.

Long-term ecological research and the COVID-19 anthropause: A window to understanding social-ecological disturbance.

The period of disrupted human activity caused by the COVID-19 pandemic, coined the "anthropause," altered the nature of interactions between humans and ecosystems. It is uncertain how the anthropause has changed ecosystem states, functions, and feedback to human systems through shifts in ecosystem services. Here, we used an existing disturbance framework to propose new investigation pathways for coordinated studies of distributed, long-term social-ecological research to capture effects of the anthropause. Although it is still too early to comprehensively evaluate effects due to pandemic-related delays in data availability and ecological response lags, we detail three case studies that show how long-term data can be used to document and interpret changes in air and water quality and wildlife populations and behavior coinciding with the anthropause. These early findings may guide interpretations of effects of the anthropause as it interacts with other ongoing environmental changes in the future, particularly highlighting the importance of long-term data in separating disturbance impacts from natural variation and long-term trends. Effects of this global disturbance have local to global effects on ecosystems with feedback to social systems that may be detectable at spatial scales captured by nationally to globally distributed research networks.

Influences of agricultural land use composition and distribution on nitrogen export from a subtropical watershed in China.

Despite the significant impacts of agricultural land on nonpoint source (NPS) nitrogen (N) pollution, little is known about the influence of the distribution and composition of different agricultural land uses on N export at the watershed scale. We used the Soil and Water Assessment Tool (SWAT) to quantify how agricultural distribution (i.e., the spatial distributions of agricultural land uses) and composition (i.e., the relative percentages of different types of agricultural land uses) influenced N export from a Chinese subtropical watershed, accounting for aquatic N retention by river networks. Nitrogen sources displayed high spatial variability, with 40.7% of the total N (TN) export from the watershed as a whole derived from several subwatersheds that accounted for only 18% of the watershed area. These subwatersheds were all located close to the watershed mouth. Agricultural composition strongly affected inputs to the river network. The percentages of dry agricultural land and rice paddy fields, and the number of cattle together explained 70.5% of the variability of the TN input to the river network among different subwatersheds. Total N loading to the river network was positively correlated with the percentage of dry land in total land areas and the number of cattle within subwatersheds, but negatively with the proportion of paddy fields. Distribution of agricultural land uses also affected N export at the mouth of the watershed. Moreover, N retention in the river network increased with increasing N transport distance from source subwatershed to the watershed mouth. Results provide important information to support improved planning of agricultural land uses at the watershed scale that reduces NPS nutrient pollution.

Extreme rainfall, vulnerability and risk: a continental-scale assessment for South America.

Extreme weather continues to preoccupy society as a formidable public safety concern bearing huge economic costs. While attention has focused on global climate change and how it could intensify key elements of the water cycle such as precipitation and river discharge, it is the conjunction of geophysical and socioeconomic forces that shapes human sensitivity and risks to weather extremes. We demonstrate here the use of high-resolution geophysical and population datasets together with documentary reports of rainfall-induced damage across South America over a multi-decadal, retrospective time domain (1960-2000). We define and map extreme precipitation hazard, exposure, affectedpopulations, vulnerability and risk, and use these variables to analyse the impact of floods as a water security issue. Geospatial experiments uncover major sources of risk from natural climate variability and population growth, with change in climate extremes bearing a minor role. While rural populations display greatest relative sensitivity to extreme rainfall, urban settings show the highest rates of increasing risk. In the coming decades, rapid urbanization will make South American cities the focal point of future climate threats but also an opportunity for reducing vulnerability, protecting lives and sustaining economic development through both traditional and ecosystem-based disaster risk management systems.