This study presents a comprehensive assessment of the variability and trends of the precipitation and temperature along with the trends in drought indices over the State of California. The non-parametric Mann–Kendall trend test is applied with a trend-free pre-whitening procedure in trend identification. A dataset containing 120-year (water years 1896–2015) monthly precipitation, average temperature, maximum temperature, minimum temperature and the Palmer Index for seven climatic regions of the state is used for this purpose. The results confirm previous work indicating that no clear trends are observed in precipitation, while a distinct warming trend is evident in temperature over the state. New findings of this study include: (1) in general, the variability of annual, winter (December–February) and spring (March–May) precipitation shows an increasing tendency, implying intensified frequency of the occurrence of dry or wet extremes; (2) on the annual scale and in the summer, statewide meteorological, hydrological and agricultural drought indices all have decreasing trends, indicating the more frequent occurrence of drought events; and (3) among seven regions, the South Coast Drainage region generally has the most significant warming trend, as well as the most significant declining trends in drought indices. Overall, these findings are highly meaningful from both theoretical and practical perspectives, in the context of providing critical information in developing prediction models and guiding water resources management practices, respectively.

California experienced an extraordinary drought from 2012–2015 (which continues into 2016). This study, from an operational perspective, reviewed the development of this drought in a hydroclimatic framework and examined its characteristics at different temporal and spatial scales. Observed and reconstructed operational hydrologic indices and variables widely used in water resources planning and management at statewide and (hydrologic) regional scales were employed for this purpose. Parsimonious metrics typically applied in drought assessment and management practices including the drought monitor category, percent of average, and rank were utilized to facilitate the analysis. The results indicated that the drought was characterized by record low snowpack (statewide four-year accumulated deficit: 280%-of-average), exceptionally low April-July runoff (220%-of-average deficit), and significantly below average reservoir storage (93%-of-average deficit). During the period from 2012–2015, in general, water year 2015 stood out as the driest single year; 2014–2015 was the driest two-year period; and 2013–2015 tended to be the driest three-year period. Contrary to prior studies stating that the 2012–2015 drought was unprecedented, this study illustrated that based on eight out of 28 variables, the 2012–2015 drought was not without precedent in the record period. Spatially, on average, the South Coast Region, the Central Coast Region, the Tulare Region, and the San Joaquin Region generally had the most severe drought conditions. Overall, these findings are highly meaningful for water managers in terms of making better informed adaptive management plans.

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.