This paper is derived from a keynote talk given at the Google's 2020 Flood Forecasting Meets Machine Learning Workshop. Recent experiments applying deep learning to rainfall-runoff simulation indicate that there is significantly more information in large-scale hydrological data sets than hydrologists have been able to translate into theory or models. While there is a growing interest in machine learning in the hydrological sciences community, in many ways, our community still holds deeply subjective and nonevidence-based preferences for models based on a certain type of "process understanding" that has historically not translated into accurate theory, models, or predictions. This commentary is a call to action for the hydrology community to focus on developing a quantitative understanding of where and when hydrological process understanding is valuable in a modeling discipline increasingly dominated by machine learning. We offer some potential perspectives and preliminary examples about how this might be accomplished.

We present a lattice QCD calculation of the parameters of the ρ meson decay. The study is carried out on spatially asymmetric boxes using nHYP-smeared clover fermions with two mass-degenerate quark flavors. Our calculations are carried out at a pion mass mπ = 304(2) MeV on the set of lattices V = 24 2 × η24 × 48 with η = 1.0, 1.25, and 2.0 with lattice spacing a = 0.1255(7) fm. The resonance mass mρ = 827(3)(5) MeV and coupling constant gρππ = 6.67(42) are calculated using the P-wave scattering phase shifts. We construct a 2×2 correlation matrix to extract the energy of the scattering states and compute the phase shifts using the finite volume formula. By varying the degree of asymmetry, we are able to compute a set of phase shifts that are evenly distributed throughout the spectral region where the ρ decays.

We suggest that there is a potential danger to the hydrological sciences community in not recognizing how transformative machine learning will be for the future of hydrological modeling. Given the recent success of machine learning applied to modeling problems, it is unclear what the role of hydrological theory might be in the future. We suggest that a central challenge in hydrology right now should be to clearly delineate where and when hydrological theory adds value to prediction systems. Lessons learned from the history of hydrological modeling motivate several clear next steps toward integrating machine learning into hydrological modeling workflows.

Graphical Processing Units (GPUs) are more and more frequently used for lattice QCD calculations. Lattice studies often require computing the quark propagators for several masses. These systems can be solved using multi-shift inverters but these algorithms are memory intensive which limits the size of the problem that can be solved using GPUs. In this paper, we show how to efficiently use a memory-lean single-mass inverter to solve multi-mass problems. We focus on the BiCGstab algorithm for Wilson fermions and show that the single-mass inverter not only requires less memory but also outperforms the multi-shift variant by a factor of two.Comment: 27 pages, 6 figures, 3 Table

Although a wide variety of quantum computers are currently being developed, actual computational results have been largely restricted to contrived, artificial tasks. Finding ways to apply quantum computers to useful, real-world computational tasks remains an active research area. Here we describe our mapping of the protein design problem to the D-Wave quantum annealer. We present a system whereby Rosetta, a state-of-the-art protein design software suite, interfaces with the D-Wave quantum processing unit to find amino acid side chain identities and conformations to stabilize a fixed protein backbone. Our approach, which we call the QPacker , uses a large side-chain rotamer library and the full Rosetta energy function, and in no way reduces the design task to a simpler format. We demonstrate that quantum-annealer-based design can be applied to complex real-world design tasks, producing designed molecules comparable to those produced by widely adopted classical design approaches. We also show through large-scale classical folding simulations that the results produced on the quantum annealer can inform wet-lab experiments. For design tasks that scale exponentially on classical computers, the QPacker achieves nearly constant runtime performance, independent of the complexity of the task, up to the limits of the quantum computer's size.

Abstract-Lattice QCD calculations were one of the first applications to show the potential of GPUs in the area of high performance computing. Our interest is to find ways to effectively use GPUs for lattice calculations using the overlap operator. The large memory footprint of these codes requires the use of multiple GPUs in parallel. In this paper we show the methods we used to implement this operator efficiently. We run our codes both on a GPU cluster and a CPU cluster with similar interconnects. We find that to match performance the CPU cluster requires 20-30 times more CPU cores than GPUs.

A hybrid machine learning and process-based-modeling (PBM) approach is proposed and evaluated at a handful of AmeriFlux sites to simulate the top-layer soil moisture state. The Hybrid-PBM (HPBM) employed here uses the Noah land-surface model integrated with Gaussian Processes. It is designed to correct the model only in climatological situations similar to the training data else it reverts to the PBM. In this way, our approach avoids bad predictions in scenarios where similar training data is not available and incorporates our physical understanding of the system. Here we assume an autoregressive model and obtain out-of-sample results with upwards of a 3-fold reduction in the RMSE using a one-year leave-one-out cross-validation at each of the selected sites. A path is outlined for using hybrid modeling to build global land-surface models with the potential to significantly outperform the current state-of-the-art.

No abstract