Journal of The American Water Resources Association最新文献

This technical note describes recent efforts to integrate machine learning (ML) models, specifically long short-term memory (LSTM) networks and differentiable parameter learning conceptual hydrological models (δ conceptual models), into the next-generation water resources modeling framework (Nextgen) to enhance future versions of the U.S. National Water Model (NWM). We address three specific methodology gaps of this new modeling framework: (1) assess model performance across many ungauged catchments, (2) diagnostic-based model selection, and (3) regionalization based on catchment attributes. We demonstrate that an LSTM trained on CAMELS catchments can make large-scale predictions with Nextgen across the New England region and match the average flow duration curve observed by stream gauges for streamflow with low exceedance probability (high flows), but diverges from the mean in high exceedance probability (low flows). We demonstrate improvements in peak flow predictions when using δ conceptual model, but results also suggest that performance increases may come at a cost of accurately representing hydrologic states within the conceptual model. We propose a novel approach using ML to predict the most performant mosaic modeling approach and demonstrate improved distributions of efficiency scores throughout the large sample of basins. Our findings advocate for the future development of ML capabilities within Nextgen for advancing operational hydrological modeling.

Interbasin water diversion project has been considered as an effective way to assure water resource system sustainability. In order to assess the impacts of water diversion on sustainability, we propose a framework in terms of reliability, resilience, and vulnerability. The estimated water availability from hydrological models and the projected water demand are input to a water resource allocation model. The water resource allocation model allocates the two available water sources (i.e., the local and the diverted water) in the water-receiving areas. The differences of the allocated water resources between these two water sources are figured out to quantify the impacts of water diversion on water resource system sustainability. The water-receiving area of Bailong River Water Diversion Project, located in Gansu, China, was selected as a case study. The results show that compared to the reference planning years, the runoff in future planning years will be reduced, while their water demands will almost increase under all scenarios. Although the current designed water diversion scheme is effective in increasing resilience, there is still potential for increasing resilience through optimizing the designed scheme. Further, the more unfavorable the water supply and demand conditions are, the larger the space for optimizing the system sustainability. This study can help understand the impacts of water diversion on water resource system sustainability in a changing environment.

Tidal freshwater forests were once extensive across temperate coastlines, but loss and fragmentation have made estimation of their ecosystem functions challenging. We measured water temperature for 2 years in three Sitka spruce tidal forests, a restoration site, and an adjacent emergent marsh on the Columbia River, Washington, United States. We assessed spatial variability of water temperature within sites including the effects of hydrology, differences among bay and tributary tidal forests, and differences between the tidal forests and the mainstem Columbia, the restoration site, and the emergent marsh. The tidal forests nearest to the bay had lower interior water temperatures than their channel confluences by up to 2.5°C (weekly median temperature) and 2.0°C (weekly maximum temperature), with most cooling occurring during the low-flow months of July–September. Tributary sites had maximum temperatures up to 1.9°C cooler than bay sites and 4.2°C cooler than the mainstem. Temperatures in the two bay sites decreased by −0.16°C/100 m and −0.07°C/100 m, on average. The restoration site had the smallest within-site temperature gradient. Differences in maximum temperatures were greatest when tidal range was low, while higher tidal ranges were associated with warmer and more variable site interiors relative to their confluences. These results suggest that water temperatures in these tidal forests can provide temperature refugia for cold water biota including salmon.

Inter-basin water transfers (IBTs) are important components of water balances of basins, and they can have substantial impact on regional water availability. Flow information is often not available at locations with known IBTs, which is a drawback in several published IBT databases. Few, if any, studies examine whether IBT flow behavior can be generalized, and if these behaviors can be predicted at undocumented locations or known IBT locations with no flow information. In this study, we employ a clustering method based on image matching to identify similar classes of flow behavior of IBTs. Machine learning models are used to assess how well IBT flow characteristics (e.g., average flow) associated with these behaviors can be predicted. These evaluations of IBTs are done for two regions in the United States. Three primary classes of IBTs (seasonal, nonseasonal/not mixed, and seasonal/mixed) are identified across the two regions analyzed. The IBT flow characteristics are accurately predicted in the northeast region. In the Colorado region, however, only the flow characteristics related to timing were accurately predicted. These results indicate that the proposed modeling framework can be used to identify generalizable IBT flow characteristics. This framework is shown to predict flow characteristics with a reasonable amount of accuracy to undocumented locations and improves previously published IBT databases by backfilling flow information to locations with a known IBT presence.

Unimpaired flow, also known as “virgin” or “naturalized” flow, is the theoretical runoff that would occur absent human alteration and is a common metric used worldwide to support watershed management, regulation and ecosystem restoration. California water managers use an existing record of unimpaired runoff from 10 watersheds upstream of the San Francisco Estuary to inform a variety of planning and regulatory functions. In this work, we extended the available record by several decades to 1872. This runoff reconstruction, accomplished through a multivariate regression-based modeling approach, used a novel time series methodology that decomposes and combines sparse records measured at varying locations to generate composite temperature and precipitation input for each watershed. Runoff predictions were used to calculate watershed runoff indices and assign water year classifications using current regulatory conventions. The first third of the extended record (1872–1921) has a significantly higher percentage of wet years and a significantly lower percentage of dry and critically dry years compared with the latter two-thirds of the record (1922–2021). Consistent with earlier paleohydrology research, this finding indicates the occurrence of a dramatic decadal-scale hydrologic shift from very wet to very dry conditions during the late 19th and early 20th centuries.

Storm Water Management Model (SWMM) was widely used in hydrological simulation, river network calculation, flood control, and disaster reduction, but it was rarely used in ship lock hydraulics simulation. In this paper, SWMM is used to simulate the water filling and emptying process in ship lock. One-dimensional hydraulic mathematical models are established for the centralized filling and emptying system, simple decentralized filling and emptying system, and complex decentralized filling and emptying system, respectively, and they are applied to Mengli Ship Lock of North River in Guangdong Province, Qiaogong Ship Lock of Hongshui River, and Bajiangkou Ship Lock of Gui River in Guangxi Province of China. Compared with the results from the physical model experiments of ship lock hydraulics, it is shown that SWMM has better simulation effects on the centralized filling and emptying system with short corridor and the simple decentralized filling and emptying system, and their relative errors of filling and emptying time and maximum flow are totally less than 5%. For the complex decentralized filling and emptying system, SWMM has a good simulation effect on the water level change process, but the simulation error of the flow change process is larger relatively, and the relative errors of filling and emptying time and maximum flow are totally less than 10%. In addition, the filling and emptying process simulations are all completed within 5 s for three types of filling and emptying systems. The results show that the mathematical model of ship lock filling and emptying based on SWMM has high operation efficiency and good simulation accuracy and can not only be used as an effective tool for hydraulic calculation of ship lock but also provide mathematical model support for navigation intelligent scheduling.