https://agupubs.onlinelibrary.wiley.com/doi/epdf/10.1029/2018WR022643

A Transdisciplinary Review of Deep Learning Research and Its Relevance for Water Resources Scientists

The increasing impact of both climatic change and human activities on global river systems necessitates an increasing need to identify and quantify the various drivers and their impacts on fluvial water and sediment discharge. Here we show that mean Yangtze River water discharge of the first decade after the closing of the Three Gorges Dam (TGD) (2003–2012) was 67 km3/yr (7%) lower than that of the previous 50 years (1950–2002) and 126 km3/yr less compared to the relatively wet period of pre-TGD decade (1993–2002). Most (60–70%) of the decline can be attributed to decreased precipitation, the remainder resulting from construction of reservoirs, improved water-soil conservation and increased water consumption. Mean sediment flux decreased by 71% between 1950–1968 and the post-TGD decade, about half of which occurred prior to the pre-TGD decade. Approximately 30% of the total decline and 65% of the decline since 2003 can be attributed to the TGD, 5% and 14% of these declines to precipitation change and the remaining to other dams and soil conservation within the drainage basin. These findings highlight the degree to which changes in riverine water and sediment discharge can be related with multiple environmental and anthropogenic factors.

/pdf/decline-of-yangtze-river-water-and-sediment-discharge-impact-2168xd8138.pdf

Decline of Yangtze River water and sediment discharge: Impact from natural and anthropogenic changes.

Deep learning (DL), a new-generation of artificial neural network research, has transformed industries, daily lives and various scientific disciplines in recent years. DL represents significant progress in the ability of neural networks to automatically engineer problem-relevant features and capture highly complex data distributions. I argue that DL can help address several major new and old challenges facing research in water sciences such as inter-disciplinarity, data discoverability, hydrologic scaling, equifinality, and needs for parameter regionalization. This review paper is intended to provide water resources scientists and hydrologists in particular with a simple technical overview, trans-disciplinary progress update, and a source of inspiration about the relevance of DL to water. The review reveals that various physical and geoscientific disciplines have utilized DL to address data challenges, improve efficiency, and gain scientific insights. DL is especially suited for information extraction from image-like data and sequential data. Techniques and experiences presented in other disciplines are of high relevance to water research. Meanwhile, less noticed is that DL may also serve as a scientific exploratory tool. A new area termed 'AI neuroscience,' where scientists interpret the decision process of deep networks and derive insights, has been born. This budding sub-discipline has demonstrated methods including correlation-based analysis, inversion of network-extracted features, reduced-order approximations by interpretable models, and attribution of network decisions to inputs. Moreover, DL can also use data to condition neurons that mimic problem-specific fundamental organizing units, thus revealing emergent behaviors of these units. Vast opportunities exist for DL to propel advances in water sciences.

X.X. Luo

Papers

Downstream sedimentary and geomorphic impacts of the Three Gorges Dam on the Yangtze River