Automation and human expertise in operational river forecasting

Abstract

Increased automation and use of computer-aided decision support systems are attractive options for hydrologic forecasting agencies faced with growing product complexity and institutional resourcing pressures. Although the hydrologic literature has been nearly silent on the roles of expertise and automation in forecasting, other disciplines such as meteorology have had decades of open discussion on the topic. To address the lack of dialogue in hydrology on automation, this article seeks to contextualize relevant findings from similar disciplines, including meteorology, psychology, decision support systems, and interface design. We predict which aspects of operational hydrology have the greatest chance for successfully increasing automation in the near future. Some applications have employed higher levels of automation, notably flash flood forecasting which requires rapid response times, and extended prediction which requires heavy emphasis on uncertainty quantification. Short-range flood forecasting may be more challenging to automate and traditionally has been less automated than other types of forecasts, partly because of existing practices of interfacing with meteorologists and water system operators, and the difficulties in modeling human impacts on the water cycle. Overall, we suggest that the design of computer-aided decision support systems for forecasting systems should consider three factors: (1) processes change under automation and people may require new roles; (2) automation changes the way people behave, sometimes negatively; and (3) people may not have accurate perceptions of the quality of the automated guidance. Seven lessons learned from automation in meteorology are highlighted and translated into a hydrologic forecasting context, leading to a set of recommendations for how to make best use of expertise in increasingly automated systems.

Can big data tame a “naughty” world?

“…Further, these changes in observational and methodological priorities could lead towards the prioritization of reanalysis of open source/public datasets over field science, triggering the “rise of the data scientist” (Levy ) and the death of the environmental expert (Death ). There is no doubt that technicians, volunteers, and citizen scientists, together with industrial and commercial data analysts, can each play valuable and important roles in collection and analysis of big datasets, but we agree with Pagano et al () that the human judgment of the environmental scientist remains critical to meaningful interpretation.…”

“…A second shift invoked by big data involves a movement in environmental scientists' roles towards the analysis of correlations automatically highlighted by algorithms, the aim being to identify patterns that make physical sense and develop theoretical arguments and empirical hypotheses to test these (Peters et al ). Unguided, automated exploration of big datasets is perceived as a productive way to analyze big data and compare multiple variables across space and time (e.g., Death ; Krause et al ; Pagano et al ). However, results from automated techniques can be misleading if they are not interpreted within the context of existing knowledge frameworks and the limitations of the dataset (O'Sullivan and Manson ).…”

An intercomparison of approaches for improving operational seasonal streamflow forecasts

“…Finally, this work is part of a larger project that explores the potential of an automated (i.e., "over-the-loop") forecasting workflow as a viable strategy for operational streamflow prediction that can open the door to potential scientific and technical advances in streamflow forecasting (Pagano et al, 2016). In this context, a critical lesson is that the entire study, in particular the assessment of approach alternatives, depends on the automation of the forecast workflow to enable the generation of hindcasts that are consistent with real-time forecasts.…”

“…The hydroclimatology of the PNW region is affected by a number of large-scale climate teleconnections. The warm (cold) phase of El Niño-Southern Oscillation (ENSO) is typically associated with above (below) average temperatures and below (above) average precipitation during winter (e.g., Redmond and Koch, 1991) and therefore decreased (increased) snowpack (Clark et al, 2001) and spring/summer runoff (e.g., Piechota et al, 1997). The Pacific Decadal Oscillation (PDO; Mantua et al, 1997) -which reflects the dominant mode in decadal variability of sea surface temperatures (SSTs) -has also been found a relevant driver for the hydroclimatology of the PNW (e.g., McCabe and Dettinger, 2002).…”

“…Proactive water management has become critical for many regions in the world that are susceptible to water stress asso-ciated with the intensification of the water cycle. Paradoxically, although we have seen important technological advances, including increased computing power, the broader availability to climate reanalysis, forecasts and reforecasts, and more complex process-based hydrologic models (Pagano et al, 2016), the skill of operational seasonal runoff predictions in the US, termed water supply forecasts (WSFs), has shown little or no improvement over time (e.g., Pagano et al, 2004;Harrison and Bales, 2016). Hence, there is both a scientific and practical need to understand the potential of new datasets, modeling resources, and methods to accelerate progress towards more skillful and reliable operational seasonal streamflow forecasts.…”