Search

Changes in demand, various hydrological inputs, and environmental stressors are among the issues that reservoir managers and policymakers face on a regular basis. These concerns have sparked interest in applying different techniques to determine reservoir operation policy decisions. As the resolution of the analysis increases, it becomes more difficult to effectively represent a real-world system using traditional methods such as dynamic programming and stochastic dynamic programming for determining the best reservoir operation policy. One of the challenges is the "curse of dimensionality," which means the number of samples needed to estimate an arbitrary function with a given level of accuracy grows exponentially with respect to the number of input variables (i.e., dimensionality) of the function. Deep reinforcement learning (DRL) is an intelligent approach to overcome the curses of stochastic optimization problems for reservoir operation policy decisions. To our knowledge, this study is the first attempt that examines various novel DRL continuous-action policy gradient methods, including deep deterministic policy gradients, twin delayed DDPG (TD3), and two different versions of Soft Actor-Critic (SAC18 and SAC19) for optimizing reservoir operation policy. In this study, multiple DRL techniques were implemented to find an optimal operation policy for Folsom Reservoir in California. The reservoir system supplies agricultural, municipal, hydropower, and environmental flow demands and flood control operations to the City of Sacramento. Analysis suggests that the TD3 and SAC are robust to meet the Folsom Reservoir's demands and optimize reservoir operation policies. Experiments on continuous-action spaces of reservoir policy decisions demonstrated that the DRL techniques can efficiently learn strategic policies in spaces and can overcome the curse of dimensionality and modeling. [ABSTRACT FROM AUTHOR]

Multi-year droughts are ever-present and transformational features of California's Mediterranean climate and can fundamentally affect the water quality and the ecosystem responses of the San Francisco Estuary (estuary) and the Sacramento-San Joaquin Delta (Delta). This study assessed data collected by long-term monitoring programs over the past 46 water years (1975-2021) to evaluate how water quality in the estuary changes during multi-year droughts. Data were aggregated by region (South-Central Delta, North Delta, confluence, Suisun Bay, and Suisun Marsh) and season, then differences between multi-year drought periods, multiyear wet periods, and neutral periods were compared using generalized linear models. We found that multi-year drought periods altered multiple physical and chemical parameters in the estuary, increasing water temperature, salinity, water clarity, and nutrient levels. This trend was consistent across regions and seasons, with few exceptions. Increases in these parameters during drought periods were likely caused by reduced Delta inflows that intensified in each successive dry year because of reduced precipitation and managed estuarine inflows and outflows. Drought periods did not substantially affect tidal velocities within the estuary, which remained mostly consistent across wet and drought periods. Trends in chlorophyll concentrations during drought periods were more nuanced with higher concentrations occurring in the South-Central Delta region and during the winter and spring. Together, these results characterized drought in the estuary as warm, clear, high in nutrients, with patchy phytoplankton blooms (as indexed by chlorophyll), all of which have implications for higher trophic levels. Considering that droughts are expected to increase in frequency and intensity in California with climate change, understanding the effects of multi-year droughts on the water quality conditions of the estuary can help inform water management decisions. [ABSTRACT FROM AUTHOR]

The diffuse attenuation coefficient of photosynthetically active radiation (KdPAR) is commonly used to predict light attenuation in aquatic productivity models, but obtaining measurements of PAR to compute KdPAR is difficult. In situ calculations of KdPAR require multiple measurements of PAR through the water column, and these measurements are infeasible for real-time recording. Instead, predictive models using surface-water measurements may be used. Traditional KdPAR models are based on open-ocean habitats and rely on chlorophyll-- as a proxy measurement for phytoplankton abundance--as the main predictive parameter. However, elevated suspended sediments and dissolved organic materials may also affect KdPAR values of inland water bodies and estuaries. In this study, we leverage KdPAR calculations derived from in situ light measurements collected along with surface-water-quality parameters across the Sacramento-San Joaquin River Delta in California, USA (the Delta). Sampling occurred between January of 2013 and May of 2014. We also explored regional and seasonal effects, but these did not clearly affect the model. Ultimately, the best-performing model included surface-level turbidity only (R² = 0.91). The simplicity of the model facilitates use of KdPAR estimates for a variety of purposes throughout the Delta, including euphotic depth calculations, and as inputs to primary-productivity and habitat-suitability models. We demonstrate the model's usability with two open-sources data sets (one spatially dense, and one temporally dense), and estimate KdPAR, euphotic depth, and primary productivity within the Delta. We provide calculations for each estimation, allowing users to easily adopt these models and apply them to their own data or with open-sourced data, which are abundant. [ABSTRACT FROM AUTHOR]

Salinity management in estuarine systems is crucial for developing effective water-management strategies to maintain compliance and understand the impact of salt intrusion on water quality and availability. Understanding the temporal and spatial variations of salinity is a keystone of salinity-management practices. Process-based numerical models have been traditionally used to estimate the variations in salinity in estuarine environments. Advances in data-driven models (e.g., deep learning models) make them effective and efficient alternatives to process-based models. However, a discernible research gap exists in applying these advanced techniques to salinity modeling. The current study seeks to address this gap by exploring the innovative use of deep learning with data augmentation and transfer learning in salinity modeling, exemplified at 23 key salinity locations in the Sacramento–San Joaquin Delta which is the hub of the water-supply system of California. Historical, simulated (via a hydrodynamics and water quality model), and perturbed (to create a range of hydroclimatic and operational scenarios for data-augmentation purposes) flow, and salinity data are used to train a baseline multi-layer perceptron (MLP) and a deep learning Residual Long-Short-Term Memory (Res-LSTM) network. Four other deep learning models including LSTM, Residual Network (ResNet), Gated Recurrent Unit (GRU), and Residual GRU (Res-GRU) are also examined. Results indicate that models pre-trained using augmented data demonstrate improved performance over models trained from scratch using only historical data (e.g., median Nash–Sutcliffe efficiency increased from around 0.5 to above 0.9). Moreover, the five deep learning models further boost the salinity estimation performance in comparison with the baseline MLP model, though the performance of the latter is acceptable. The models trained using augmented data are then (a) used to develop a web-based Salinity Dashboard (Dashboard) tool that allows the users (including those with no machine learning background) to quickly screen multiple management scenarios by altering inputs and visualizing the resulting salinity simulations interactively, and (b) transferred and adapted to estimate observed salinity. The study shows that transfer learning results more accurately replicate the observations compared to their counterparts from models trained from scratch without knowledge learned and transferred from augmented data (e.g., median Nash–Sutcliffe efficiency increased from around 0.4 to above 0.9). Overall, the study illustrates that deep learning models, particularly when pre-trained using augmented data, are promising supplements to existing process-based models in estuarine salinity modeling, while the Dashboard enables user engagement with those pre-trained models to inform decision-making efficiently and effectively. [ABSTRACT FROM AUTHOR]

Managed aquifer recharge (MAR) can provide long‐term storage of excess surface water for later use. While decades of research have focused on the physical processes of MAR and identifying suitable MAR locations, very little research has been done on how to consider competing factors and tradeoffs in siting MAR facilities. This study proposes the use of a simulation‐optimization (SO) framework to map out a cost‐effectiveness frontier for MAR by combining an evolutionary algorithm with two objective functions that seek to maximize groundwater storage gains while minimizing MAR cost. We present the theoretical framework along with a real‐world application to California's Central Valley. The result of the SO framework is a Pareto front that allows identifying suitable MAR locations for different levels of groundwater storage gain and associated MAR project costs, so stakeholders can evaluate different choices based on cost, benefits, and tradeoffs of MAR sites. Application of the SO framework to the Central Valley shows groundwater can be recharged from high‐magnitude (95th percentile) flows at a marginal cost of $57 to $110 million per km3. If the 10 percent largest flows are recharged the total groundwater storage gain would double and the marginal costs would drop to between $30 and $50 million per km3. If recharge water is sourced from outside local basins (e.g., the Sacramento‐San Joaquin Delta), groundwater storage gain is approximately 25%–80% greater than can be achieved by recharging local flows, but the total cost is about 10%–15% higher because of additional lift cost. Key Points: Combining a simulation‐optimization framework with a groundwater model can identify suitable managed aquifer recharge locationsRecharging water from local rivers within the same basin is the most cost‐efficient managed aquifer recharge approach per unit of groundwater storage gainDiverting more water may not yield the most efficient groundwater storage gain due to higher transportation cost [ABSTRACT FROM AUTHOR]

Little is known about how farms and markets are connected. Identifying critical gaps and central hubs in food systems is of importance in addressing a variety of concerns, such as navigating rapid shifts in marketing practices as seen during the COVID-19 pandemic and related food shortages. The constellation of growers and markets can also reinforce opportunities to shift growing and eating policies and practices with attention to addressing racial and income inequities in food system ownership and access. With this research, we compare network methods for measuring centrality and sociospatial orientations in food systems using two of America's most high-producing agricultural counties. Though the counties are adjacent, we demonstrate that their community food systems have little overlap in contributing farms and markets. Our findings show that the community food system for Yolo County is tightly interwoven with Bay Area restaurants and farmers' markets. The adjacent county, Sacramento, branded itself as America's Farm-to-Fork capital in 2012 and possesses network hubs focused more on grocery stores and restaurants. In both counties, the most central actors differ and have been involved with the community food system for decades. Such findings have implications beyond the case studies, and we conclude with considerations for how our methods could be standardized in the national agricultural census. [ABSTRACT FROM AUTHOR]

Water resources management in estuarine environments for water supply and environmental protection typically requires estimates of salinity for various flow and operational conditions. This study develops and applies two novel deep learning (DL) models, a residual long short-term memory (Res-LSTM) network, and a residual gated recurrent unit (Res-GRU) model, in estimating the spatial and temporal variations of salinity. Four other machine learning (ML) models, previously developed and reported, consisting of multi-layer perceptron (MLP), residual network (ResNet), LSTM, and GRU are utilized as the baseline models to benchmark the performance of the two novel models. All six models are applied at 23 study locations in the Sacramento–San Joaquin Delta (Delta), the hub of California's water supply system. Model input features include observed or calculated tidal stage (water level), flow and salinity at model upstream boundaries, salinity control gate operations, crop consumptive use, and pumping for the period of 2001–2019. Meanwhile, field observations of salinity at the study locations during the same period are also utilized for the development of the predictive use of the models. Results indicate that the proposed DL models generally outperform the baseline models in simulating and predicting salinity on both daily and hourly scales at the study locations. The absolute bias is generally less than 5%. The correlation coefficients and Nash–Sutcliffe efficiency values are close to 1. Particularly, Res-LSTM has slightly superior performance over Res-GRU. Moreover, the study investigates the overfitting issues of both the DL and baseline models. The investigation indicates that overfitting is not notable. Finally, the study compares the performance of Res-LSTM against that of an operational process-based salinity model. It is shown Res-LSTM outperforms the process-based model consistently across all study locations. Overall, the study demonstrates the feasibility of DL-based models in supplementing the existing operational models in providing accurate and real-time estimates of salinity to inform water management decision making. [ABSTRACT FROM AUTHOR]

Background: Breast cancer is the most common cancer among women in the U.S. and the leading cause of cancer death among Hispanics/Latinas (H/L). H/L are less likely than Non-H/L White (NHW) women to be diagnosed in the early stages of this disease. Approximately 5-10% of breast cancer can be attributed to inherited genetic mutations in high penetrance genes such as BRCA1/2. Women with pathogenic variants in these genes have a 40-80% lifetime risk of breast cancer. Past studies have shown that genetic counseling can help women and their families make informed decisions about genetic testing and early cancer detection or risk-reduction strategies. However, H/L are 3.9-4.8 times less likely to undergo genetic testing than NHW women. We developed a program to outreach and educate the H/L community about hereditary breast cancer, targeting monolingual Spanish-speaking individuals in California. Through this program, we have assessed cancer screening behavior and identified women who might benefit from genetic counseling in a population that is usually excluded from cancer research and care. Materials and Methods: The "Tu Historia Cuenta" program is a promotores-based virtual outreach and education program including the cities of San Francisco, Sacramento, and Los Angeles. Participants responded to three surveys: a demographic survey, a breast cancer family history survey, and a feedback survey. Survey responses were described for participants and compared by area where the program took place using chi-square, Fisher exact tests, and t tests. Multinomial logistic regression models were used for multivariate analyses. Results and Conclusion: We enrolled 1042 women, 892 completed the cancer family history survey and 62 (7%) provided responses compatible with referral to genetic counseling. We identified 272 women (42.8% ages 40 to 74 years) who were due for mammograms, 250 women (24.7% ages 25 to 65 years) due for Papanicolaou test, and 189 women (71.6% ages 50+) due for colorectal cancer screening. These results highlight the need of additional support for programs that spread awareness about cancer risk and facilitate access to resources, specifically within the H/L community. [ABSTRACT FROM AUTHOR]

Sedimentation is an important aspect of water resources management with many implications. Often, process‐based methods are employed to predict and assess the amount of sediment in water, but there are still challenges because the mechanisms that govern sediment transport are not yet fully understood. Furthermore, complex domains make model calibration difficult. Thus, as a complementary tool, a machine‐learning model was developed in the present study to emulate an existing process‐based model in simulating suspended sediment concentration (SSC). It employs the long short‐term memory (LSTM) networks, which are a type of artificial neural networks (ANNs) designed for supervised learning of a sequence of data (e.g., time series). The model was applied to the Sacramento‐San Joaquin Delta (the Delta) of California, USA, which is characterized by an interconnected system of sloughs, waterways, and a tidal outlet. The model training was performed with historical records of flow, stage and SSC at various locations within the Delta. The study period was 2010 through 2016, but the training period (i.e., range of observed data used to train the model) was varied to assess the model's sensitivity to the inputs and to determine the optimum model setup. Comparison between the model‐estimated SSC and the observation at 12 key locations within the Delta showed that the estimation accuracy of the LSTM model during the study period is comparable or superior to that of the Delta Simulation Model II‐General Transport Model (DSM2‐GTM), a process‐based operational hydrodynamics and water quality model for the Delta. In terms of the ratio of the root‐mean‐square error to the standard deviation (RSR), LSTM models generally showed higher predictability than DSM2‐GTM in all test cases investigated, with the lowest (most desirable) and highest (least desirable) LSTM‐based RSR being 0.21 and 1.14, respectively. In comparison, the lowest and highest RSR values with DSM2‐GTM were 0.26 and 3.70, respectively. The median LSTM‐based RSR of all study locations is around 0.7 while its DSM2‐GTM counterpart is about 1.0. LSTM models also yielded remarkably higher (more desirable) Nash‐Sutcliffe Efficiency values. Moreover, visual inspection found that LSTM models better captured the timing and magnitude of peaks as well as the temporal variations in the SSC time series. The LSTM model's performance was further analysed with hydro‐meteorological data (precipitation and wind speed) incorporated in training. While precipitation led to some improvement, the wind speed was found to have a negligible effect overall. All in all, the study findings suggest that the LSTM model has the potential to supplement the operational process‐based model in guiding water resources planning and management practices. [ABSTRACT FROM AUTHOR]

The levels of total salinity and its ion constituents in estuarine environments are important indicators of the overall suitability of water for environmental, agricultural, and urban use. These constituents include Total Dissolved Solids (TDS), dissolved chloride (Cl−), dissolved sulfate (SO42−), dissolved sodium (Na+), dissolved calcium (Ca2+), dissolved magnesium (Mg2+), dissolved nitrate (NO3), dissolved potassium (K), dissolved bromide (Br−), dissolved boron (B), Alkalinity, and hardness, among others. In practice, salinity is typically measured indirectly as electrical conductance (EC) via automatic sensors while the concentration of each constituent is often measured from discrete water samples (i.e., grab samples) and thus is available much less frequently than salinity. Quadratic regression equations are generally developed between salinity (as the predictor, represented by EC) and individual constituents (as the predictand) based on grab sample data. The regression models are then applied to estimate the concentrations of individual constituents given EC during the period when grab samples are not available. The current study develops four types of machine learning models: the Generalized Additive Model, Regression Trees, Random Forest, and Artificial Neural Networks, to emulate conventional regression equations in salinity-constituent conversion. A case study in the South Sacramento-San Joaquin Delta of California is presented to illustrate the development and application of these models and to compare their performance with that of the benchmark regression models. The results indicate that machine learning models can provide comparable or superior simulations to the regression models. Among the four machine models examined, the Random Forest model tends to yield the best results. The study further discusses the scientific and practical implications of the models proposed, their limitations, as well as future work to further improve their performance and applicability. [ABSTRACT FROM AUTHOR]

We study supply response to irrigation water salinity—an important and ubiquitous environmental problem facing agriculture around the world. The geographical setting is the Sacramento–San Joaquin River Delta, a main water hub in California, where salinity is a significant part of recent policy debates over water management and infrastructure. We use highly granular farm‐level panel data on Delta farming activity to estimate two sets of response parameters using (i) standard (fixed coefficients) logit model, and (ii) a mixed (random coefficient) logit model. The mixed logit results provide evidence for heterogeneity in supply response across farmers. To put these findings in a meaningful economic context, we use the results from both models to simulate estimates for aggregate acreage responses under two alternative salinity scenarios. We implement a method to simulate individual specific responses together with their confidence intervals, which provides additional insight into the composition of farmer heterogeneity. Based on these post‐regression simulation results, we find the mixed logit model predicts an elasticity of aggregate acreage for salt‐sensitive (hence higher value) crops that is an order of magnitude larger compared to that of the standard fixed coefficients logit method. This result sheds light on category of costs likely to increase in response to rising salinity in the Delta region. [ABSTRACT FROM AUTHOR]

Advances in machine-learning techniques can serve practical water management needs such as salinity level estimation. This study explores machine learning, particularly deep-learning techniques in developing computer emulators for a commonly used process model, the Delta Simulation Model II (DSM2), used for salinity estimation in California's Sacramento-San Joaquin Delta (Delta). We apply historical daily input data to DSM2 and corresponding salinity simulations at 28 study locations from 1990 to 2019 to train two machine-learning models: a multi-layer perceptron (MLP) and Long-Short-Term Memory (LSTM) networks in a multi-task learning framework. We assess sensitivity of both networks to the amount of antecedent input information (memory) and training data to determine appropriate memory size and training data length. We evaluate network performance according to several statistical metrics as well as visual inspection. The study further investigates two additional networks, the Gated Recurrent Unit (GRU) and Residual Network (ResNet) in salinity modeling, and compares their efficacy against MLP and LSTM. Our results demonstrate strong performance of the four neural network models over the study period, achieving absolute bias below 4%, plus near-perfect correlation coefficients and Nash–Sutcliffe efficiency coefficients. The high complexity LSTM shows slight performance edge. We further show that deeper and wider versions of MLP and LSTM yield only marginal benefit over their baseline counterparts. We also examined issues related to potential overfitting by the proposed models, training data selection strategies, and analytical and practical implications. Overall, this new study indicates that machine-learning-based emulators can efficiently emulate DSM2 in salinity simulation. They exhibit strong potential to supplement DSM2 in salinity modeling and help guide water resource planning and management practices for the Delta region. [ABSTRACT FROM AUTHOR]

In 1978, California adopted building codes designed to reduce the energy used for temperature control. Using a rich data set of hourly electricity consumption for 158,112 houses in Sacramento, we estimate that the average house built just after 1978 uses 8% to 13% less electricity for cooling than a similar house built just before 1978. Comparing the estimated savings to the policy's projected cost, our results suggest the policy passes a cost-benefit test. In settings where market failures prevent energy costs from being completely passed through to home prices, building codes can serve as a cost- effective tool for improving energy efficiency. [ABSTRACT FROM AUTHOR]

A new technique was used to directly measure O 3 response to changes in precursor NO x and volatile organic compound (VOC) concentrations in the atmosphere using three identical Teflon smog chambers equipped with UV lights. One chamber served as the baseline measurement for O 3 formation, one chamber added NO x , and one chamber added surrogate VOCs (ethylene, m -xylene, n -hexane). Comparing the O 3 formation between chambers over a 3-hour UV cycle provides a direct measurement of O 3 sensitivity to precursor concentrations. Measurements made with this system at Sacramento, California, between April–December 2020 revealed that the atmospheric chemical regime followed a seasonal cycle. O 3 formation was VOC-limited (NO x -rich) during the early spring, transitioned to NO x -limited during the summer due to increased concentrations of ambient VOCs with high O 3 formation potential, and then returned to VOC-limited (NO x -rich) during the fall season as the concentrations of ambient VOCs decreased and NO x increased. This seasonal pattern of O 3 sensitivity is consistent with the cycle of biogenic emissions in California. The direct chamber O 3 sensitivity measurements matched semi-direct measurements of HCHO/NO2 ratios from the TROPOspheric Monitoring Instrument (TROPOMI) aboard the Sentinel-5 Precursor (Sentinel-5P) satellite. Furthermore, the satellite observations showed that the same seasonal cycle in O 3 sensitivity occurred over most of the entire state of California, with only the urban cores of the very large cities remaining VOC-limited across all seasons. The O 3 -nonattainment days (MDA8 O 3>70 ppb) have O 3 sensitivity in the NO x -limited regime, suggesting that a NO x emissions control strategy would be most effective at reducing these peak O 3 concentrations. In contrast, a large portion of the days with MDA8 O 3 concentrations below 55 ppb were in the VOC-limited regime, suggesting that an emissions control strategy focusing on NO x reduction would increase O 3 concentrations. This challenging situation suggests that emissions control programs that focus on NO x reductions will immediately lower peak O 3 concentrations but slightly increase intermediate O 3 concentrations until NO x levels fall far enough to re-enter the NO x -limited regime. The spatial pattern of increasing and decreasing O 3 concentrations in response to a NO x emissions control strategy should be carefully mapped in order to fully understand the public health implications. [ABSTRACT FROM AUTHOR]

Inventorying nutrient and trace element sources in the Sacramento-San Joaquin Delta (the Delta) is critical to understanding how changes--including alterations to point source inputs such as upgrades to the Sacramento Regional Wastewater Treatment Plant (SRWTP) and landscape-scale changes related to wetland restoration--may alter the Delta's water quality. While island drains are a ubiquitous feature of the Delta, limited data exist to evaluate island drainage mass fluxes in this system. To better constrain inputs from island drains, we measured monthly discharge along with nutrient and trace element concentrations in island drainage on three Delta islands and surrounding rivers from June 2017 to September 2018. These data were used to calculate island-level fluxes and then upscaled to estimate Delta-wide contributions from island drains. Based on these results, we present (1) new estimates of gross and net nutrient and trace element fluxes from Delta island drains, and (2) concomitant N stable isotope data to improve our understanding of island N cycling. Over 60% of nearly all island drainage gross nutrient and trace element loads occurred in winter and spring. Upscaled island drainage net annual total nitrogen (TN), total dissolved nitrogen (TDN), and NH4 + loads comprised an estimated 9%, 7%, and 4%, respectively, of annual inputs to this system in 2018, before the SRWTP upgrade. Under a post-upgrade scenario, we estimated net annual island drainage TDN contributions to increase to 11% and NH4 + contributions to 45% of total Delta inputs as the SRWTP NH4 + load diminished to near zero. Our results suggest that island drainage is a measurable N source that has likely become increasingly important now that the SRWTP upgrade is complete. With over 200 potential active outfalls, these inputs may affect aquatic biogeochemical cycling in many regions of the Delta, especially in areas with long residence times. [ABSTRACT FROM AUTHOR]

Off-channel areas are one of the most impacted aquatic habitats by humans globally, as extensive agricultural and urban development has limited them to roughly 10% of historical extent. This is also true for California's Sacramento River Valley, where historically frequent widespread inundation has been reduced to a few off-channel water bodies along the mid-Sacramento River. This remaining shallow-water habitat provides crucial ecological benefits to multiple avian and fish species, but especially to floodplain-adapted species such as Chinook salmon (Oncorhynchus tshawytscha). Characterizing spatiotemporal off-channel dynamics, including inundation extent and residence time, is fundamental to better understanding the intrinsic value of such habitats and their potential to support recovery actions. Remote sensing techniques have been increasingly used to map surface water at regional and local scales, with improved resolutions. As such, this study maps off-channel inundation areas and describes their temporal dynamics by analyzing pixel-based time- series of multiple water indices, modified Normalized Difference Water Index (mNDWI) and the Automated Water Extraction Index (AWEI), generated from LandSat-8 and Sentinel-2 data between 2013–2021. Quantified off-channel area was similar with each water index and method used, but improved performance was associated with Sentinel-2 products and AWEI index to identify wetted areas under lower mainstem discharges. Results indicate an uneven distribution of off-channel habitat in the study area, with limited inundated areas in upstream reaches (<16% of total off-channel area for greater flows). In addition, much less habitat exists for flows under 400 m3/s, an important migration cue for endangered winter-run Chinook salmon, limiting juvenile access to areas with enhanced rearing conditions. Off-channel habitat residence times averaged between 7 and 16 days, primarily defined by the rate of receding flows, with rapid flow recession providing marginal off-channel habitat. This study shows reasonable performance of moderate resolution LandSat-8 and Sentinel-2 remote sensing imagery to characterize shallow-water inundated habitat in higher-order rivers, and as a method to inform restoration and native fish recovery efforts. [ABSTRACT FROM AUTHOR]

Steelhead (Oncorhynchus mykiss expressing an anadromous life history) in the Sacramento and San Joaquin rivers and their tributaries in California's Central Valley (CCV) belong to a Distinct Population Segment (DPS) that is listed as threatened under the US Endangered Species Act. Although contemporary management and recovery plans include numerous planned and ongoing efforts seeking to aid in DPS recovery--such as gravel augmentation, manipulation of spring flows, and restoration of rearing and spawning habitat--a paucity of data precludes the possibility of evaluating the effect of these actions on populations of Steelhead in CCV streams. Knowledge gaps relating to historic and current abundance, population-specific ratios of resident and anadromous life-history expression, and the influence of hatchery-reared fish remain largely unaddressed. This is partly a result of aspects of Steelhead biology that make them difficult to monitor, including the multitude of factors that contribute to the expression of anadromy, polymorphic populations, and migration periods that coincide with challenging field conditions. However, these gaps in understanding are also partly the result of an institutional focus on Chinook Salmon (Oncorhynchus tshawytscha) and a pervasive notion that actions benefiting Chinook populations will also benefit Steelhead populations. To evaluate these gaps and to suggest approaches for assessing DPS recovery actions, we review available data and existing monitoring efforts, and consider the actions necessary to inform the development of targeted O. mykiss monitoring programs. Current management and recovery goals focus on abundance estimates of Steelhead only, yet current monitoring is insufficient for reliable estimates. We argue that a reallocation of monitoring resources to better understand the interaction between resident O. mykiss and Steelhead would provide better data to estimate the vital rates needed to evaluate the effects of recovery actions. [ABSTRACT FROM AUTHOR]

Domain-specific architectures of artificial neural networks (ANNs) have been developed to estimate salinity levels for planning at key monitoring stations in the Sacramento-San Joaquin Delta, California. In this work, we propose three major enhancements to existing ANN architectures for purposes of training time reduction, estimation error reduction, and better feature extraction. Specifically, we design a novel multitask ANN architecture with shared hidden layers for joint salinity estimation at multiple stations, achieving a reduction of 90% training and inference time. As another major structural redesign, we replace predetermined preprocessing on input data by a trainable convolution layer. We further enhance the multitask ANN design and training for salinity forecasting. Test results indicate that these enhancements substantially improve the efficiency and expand the capacity of the current salinity modeling ANNs in the Delta. Our enhanced ANN design methodologies have the potential for incorporation into the current modeling practice and provide more robust and timely information to guide water resource planning and management in the Delta. [ABSTRACT FROM AUTHOR]

This work aims to assess potential changes in the mean and extreme precipitation and temperature across the Sacramento–San Joaquin Delta (Delta) in California in the 21st century. The study employs operative climate model projections from the Coupled Model Inter-comparison Project Phase 5 (CMIP5). Specifically, 64 individual downscaled daily projections (1/16 degree, approximately 6 by 6 km) on precipitation and temperature from 32 Global Circulation Models (GCMs) under two emission scenarios (RCP 4.5 and RCP 8.5) from 2020–2099 are utilized for the analysis. The results indicate increasing warming (in mean, minimum, and maximum temperature) further into the future under both emission scenarios. Warming also exhibits a strong seasonality, with winters expecting lower and summers expecting higher increases in temperature. In contrast, for mean annual total precipitation, there is no consistent wetter or drier signal. On average, the changes in annual total precipitation are minimal. However, dry season precipitation is projected to decline. The study also shows that the number of wet days is projected to decrease while the number of very wet (daily precipitation over 10 mm) and extremely wet (daily precipitation over 20 mm) days is projected to increase. Moreover, the study illustrates that only about half of the changes in total annual precipitation are projected to come from changes in the wettest 10% of wet days. In contrast, a majority of changes in variance of the annual precipitation comes from changes in variance of the wettest 10% of the wet days. This suggests that fluctuations in large storms are projected to dictate the variability of precipitation in the Delta. Additionally, a general upward trend in dry conditions measured by the Standardized Precipitation-Evapotranspiration Index is expected during the projection period. The trending signal is stronger at multi-year temporal scales (one to four years) and under the higher emission scenario. These change patterns are generally similar across three sub-regions of the Delta (i.e., North, South, and West), even though some changes in the South Delta are the most pronounced. This study further discusses challenges posed by these changes to the Delta's water supply and ecosystems, along with the Delta's resiliency and potential ways to address these challenges. [ABSTRACT FROM AUTHOR]