Improving the Use of Hydrologic Probabilistic and Deterministic Information in Decision-Making

Uncertainty is everywhere and understanding how individuals understand and use forecast information to make decisions given varying levels of certainty is crucial for effectively communicating risks and weather hazards. To advance prior research about how various audiences use and understand probabilistic and deterministic hydrologic forecast information, a social science study involving multiple scenario-based focus groups and surveys at four locations (Eureka, CA; Gunnison, CO; Durango, CO; Owego, NY) across the U.S. was conducted with professionals and residents. Focusing on the Hydrologic Ensemble Forecast System, the Advanced Hydrologic Prediction Service, and briefings, this research investigated how users tolerate divergence in probabilistic and deterministic forecasts and how deterministic and probabilistic river level forecasts can be presented simultaneously without causing confusion. This study found that probabilistic forecasts introduce a tremendous amount of new, yet valuable, information but can quickly overwhelm users based on how they are conveyed and communicated. Some were unaware of resources available, or how to find, sort and prioritize among all the data and information. Importantly, when presented with a divergence between deterministic and probabilistic forecasts, most sought out more information while some others reported diminished confidence in the products.Users in all regions expressed a desire to “ground-truth” the accuracy of probabilistic forecasts, understand the drivers of the forecasts, and become more familiar with them. In addition, a prototype probabilistic product that includes a deterministic forecast was tested, and suggestions for communicating probabilistic information through the use of briefing packages is proposed.

Transfer learning applications in hydrologic modeling

<p>Early work in the field of Machine Learning (ML) for hydrologic prediction is showing significant potential. Indeed, it has provided important and measurable advances toward prediction in ungauged basins (PUB). At the same time, it has motivated a new research targeting important ML topics such as uncertainty attribution and physical constrains. It has also brought into question how to best harness the wide variety of climatic and hydrologic data available today. In this work, we present initial results employing transfer learning to combine information about meteorology, streamflow, surface fluxes (FluxNet), and snow (SNOTEL) into a state of the art ML-based hydrologic model. Specifically, we will present early work demonstrating how relatively simple implementations of transfer learning can be used to enhance predictions of streamflow by transferring learning from flux and snow station observations to the watershed scale. Our work is shown to extend recently published results from Kratzert et al. (2018) using the CAMELS data set (Newman et al. 2014) for streamflow prediction in North America.</p><ul><li>Kratzert, F., Klotz, D., Brenner, C., Schulz, K., and Herrnegger, M.: Rainfall–runoff modelling using Long Short-Term Memory (LSTM) networks, Hydrol. Earth Syst. Sci., 22, 6005-6022, https://doi.org/10.5194/hess-22-6005-2018, 2018a.</li> <li>Newman; K. Sampson; M. P. Clark; A. Bock; R. J. Viger; D. Blodgett, 2014. A large-sample watershed-scale hydrometeorological dataset for the contiguous USA. Boulder, CO: UCAR/NCAR. https://dx.doi.org/10.5065/D6MW2F4D</li> </ul>

Qom—A New Hydrologic Prediction Model Enhanced with Multi-Objective Optimization

The efficient calibration of hydrologic models allows experts to evaluate past events in river basins, as well as to describe new scenarios and predict possible future floodings. A difficulty in this context is the need to adjust a large number of parameters in the model to reduce prediction errors. In this work, we address this issue with two complementary contributions. First, we propose a new lumped rainfall-runoff hydrologic model—called Qom—which is featured by a limited set of continuous decision variables associated with soil moisture and direct runoff. Qom allows to separate and quantify the volume of losses and excesses of the rainwater falling in a hydrographic basin, while a Clark’s model is used to determine output hydrograms. Second, we apply a multi-objective optimization approach to find accurate calibrations of the model in a systematic and automatic way. The idea is to formulate the process as a bi-objective optimization problem where the Nash-Sutcliffe Efficiency coefficient and percent bias have to be minimized, and to combine the results found by a set of metaheuristics used to solve it. For validation purposes, we apply our proposal in six hydrographic scenarios, comprising river basins located in Spain, USA, Brazil and Argentina. The proposed approach is shown to minimize prediction errors of simulated streamflows with regards to those observed in these real-world basins.

High-Resolution QPF Uncertainty and Its Implications for Flood Prediction: A Case Study for the Eastern Iowa Flood of 2016

This study addresses the uncertainty of High-Resolution Rapid Refresh (HRRR) quantitative precipitation forecasts (QPFs), which were recently appended to the operational hydrologic forecasting framework. In this study, we examine the uncertainty features of HRRR QPFs for an Iowa flooding event that occurred in September 2016. Our evaluation of HRRR QPFs is based on the conventional approach of QPF verification and the analysis of mean areal precipitation (MAP) with respect to forecast lead time. The QPF verification results show that the precipitation forecast skill of HRRR significantly drops during short lead times and then gradually decreases for further lead times. The MAP analysis also demonstrates that the QPF error sharply increases during short lead times and starts decreasing slightly beyond 4-h lead time. We found that the variability of QPF error measured in terms of MAP decreases as basin scale and lead time become larger and longer, respectively. The effects of QPF uncertainty on hydrologic prediction are quantified through the hillslope-link model (HLM) simulations using hydrologic performance metrics (e.g., Kling–Gupta efficiency). The simulation results agree to some degree with those from the MAP analysis, finding that the performance achieved from the QPF forcing decreases during 1–3-h lead times and starts increasing with 4–6-h lead times. The best performance acquired at the 1-h lead time does not seem acceptable because of the large overestimation of the flood peak, along with an erroneous early peak that is not observed in streamflow observations. This study provides further evidence that HRRR contains a well-known weakness at short lead times, and the QPF uncertainty (e.g., bias) described as a function of forecast lead times should be corrected before its use in hydrologic prediction.