Search NASA⌕ Search

SEARCH · Search NASA

Results for “forecast accuracy”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 181 records · Page 10
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Temporal sequence transformer to advance long-term streamflow prediction

Accurate streamflow prediction is crucial for understanding climate change impacts on water resources and for effective management of extreme hydrological events. While Long Short-Term Memory (LSTM) networks have been the dominant data-driven approach for streamflow forecasting, recent advancements in transformer architectures for time series tasks have shown promise in outperforming traditional LSTM models. This study introduces a transformer-based model that integrates historical streamflow data with climatic variables to enhance streamflow prediction accuracy. We evaluated our transformer model against a benchmark LSTM across five diverse basins in the United States. Results demonstrate that the transformer architecture consistently outperforms the LSTM model across all evaluation metrics, highlighting its potential as a more effective tool for hydrological forecasting. This research contributes to the ongoing development of advanced AI techniques for improved water resource management and climate change adaptation strategies.

Singh, Ruhaan [Farragut High School]↗
DOE OSTI2024

CovTransformer: A transformer model for SARS-CoV-2 lineage frequency forecasting

With hundreds of SARS-CoV-2 lineages circulating in the global population, there is an ongoing need for predicting and forecasting lineage frequencies and thus identifying rapidly expanding lineages. Accurate prediction would allow for more focused experimental efforts to understand pathogenicity of future dominating lineages and characterize the extent of their immune escape. Here, we first show that the inherent noise and biases in lineage frequency data make a commonly-used regression-based approach unreliable. To address this weakness, we constructed a machine learning model for SARS-CoV-2 lineage frequency forecasting, called CovTransformer, based on the transformer architecture. We designed our model to navigate challenges such as a limited amount of data with high levels of noise and bias. We first trained and tested the model using data from the UK and the USA, and then tested the generalization ability of the model to many other countries and US states. Remarkably, the trained model makes accurate predictions two months into the future with high levels of accuracy both globally (in 31 countries with high levels of sequencing effort) and at the US-state level. Our model performed substantially better than a widely used forecasting tool, the multinomial regression model implemented in Nextstrain, demonstrating its utility in SARS-CoV-2 monitoring. Assuming a newly emerged lineage is identified and assigned, our test using retrospective data shows that our model is able to identify the dominating lineages 7 weeks in advance on average before they became dominant. Overall, our work demonstrates that transformer models represent a promising approach for SARS-CoV-2 forecasting and pandemic monitoring.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2025

The Lyman- α forest from LBGs: First 3D correlation measurement with DESI and prospects for cosmology

The Lyman-α (Lyα) forest is a key tracer of large-scale structure at redshifts z > 2, traditionally studied using the spectra of luminous but relatively rare quasars. In this work, we explore the viability of using the fainter yet significantly more abundant Lyman Break Galaxies (LBGs) as alternative background sources for Lyα forest studies. We analyze 4,151 Lyα forest skewers extracted from LBG spectra obtained in the DESI pilot surveys conducted in the COSMOS and XMM-LSS fields. From this dataset, we present the first measurement of the Lyα forest auto-correlation function derived exclusively from LBG spectra, probing comoving separations up to 48 h -1 Mpc at an effective redshift of z eff = 2.70. The measured LBG Lyα forest auto-correlation is consistent with that derived from DESI DR2 quasar Lyα forest spectra at a comparable redshift, validating the use of LBGs as reliable background sources for Lyα forest analyses. In addition, we measure the cross-correlation between the LBG Lyα forest and the positions of 13,362 galaxies, demonstrating that this observable serves as a sensitive diagnostic for assessing the precision and accuracy of galaxy redshift estimates, and for identifying and correcting systematic offsets. Finally, using both synthetic LBG spectra and Fisher matrix forecasts, we show that a future wide-area survey covering ∼5,000 deg 2 , targeting 1,000 LBGs per square degree at signal-to-noise levels comparable to our sample, could enable LBG-based Lyα forest baryon acoustic oscillation (BAO) measurements with expected uncertainties of σ α ISO = 0.4% (isotropic) and σ α AP = 1.3% (Alcock-Paczynski). This performance is further enhanced when combining the BAO analysis with a Lyα forest Full Shape (FS) approach, yielding a predicted uncertainty of σ α ISO FS = 0.6%. These results open a new avenue for precision cosmology at high redshift using the Lyα forest in dense LBG samples.

Lyman alpha forest↗
DOE OSTI2025

Toward Drilling the Perfect Geothermal Well: An International Research Coordination Network for Geothermal Drilling Optimization Supported by Deep Machine Learning and Cloud Based Data Aggregation

The EDGE project, supported by the U.S. Department of Energy Geothermal Technologies Office under award DE-EE0008793, established a data-driven framework for improving the efficiency, cost-effectiveness, and reliability of geothermal well drilling. The project focused on developing scalable data infrastructure, advanced machine learning and probabilistic models, and integrated analytics tools to support continuous drilling optimization. A central objective was to reduce geothermal drilling costs by up to seventy percent while minimizing the risk of well failure through predictive diagnostics and adaptive planning. Over the project period, a comprehensive data repository was designed and deployed, incorporating records from over one hundred geothermal wells across varied geological settings. This repository supported both structured and unstructured data and adhered to FAIR data principles, enabling provenance tracking, quality control, and standardized metadata. The project introduced automated ingestion pipelines and a cloud-hosted platform that facilitated access to raw, processed, and derived datasets. This infrastructure served as the foundation for model development and analysis. Machine learning workflows were developed to predict key drilling metrics including rate of penetration, non-productive time, and total drilling costs. Self-organizing maps and dimensionality reduction methods were used to uncover operational patterns and outliers, while supervised learning algorithms such as random forests and deep neural networks were applied to forecast performance outcomes. The models were validated on heterogeneous datasets from both U.S. and Icelandic fields, demonstrating variable but significant predictive accuracy. The results indicated that finer temporal resolution, inclusion of lithological data, and consistency in operational annotations could substantially improve model performance. The project also implemented process mining techniques to reconstruct state-transition models from drilling event logs. These models enabled the identification of deviations from optimal workflows and provided insights into recurring failure modes. Analysis of non-productive time highlighted the impact of equipment failures, geological challenges, and human factors, offering opportunities for targeted mitigation strategies. The EDGE Dashboard was developed as a web-based expert system integrating data visualization, model outputs, and user-driven queries. It provided an accessible interface for operators to explore historical data, evaluate predicted outcomes, and compare drilling scenarios. Initial feedback from project partners suggested that the dashboard could serve as a foundation for more advanced advisory and optimization tools. Overall, the EDGE project demonstrated the feasibility and value of applying modern data science techniques to geothermal drilling. It delivered a set of interoperable tools and models that can support more efficient, lower-risk well development. The findings point toward a viable path for transitioning from advisory analytics to semi-autonomous drilling systems, contingent on continued collaboration, expanded datasets, and field validation. The project results have immediate relevance for drilling operations, data management practices, and future geothermal R&D efforts aimed at achieving reliable, cost-competitive geothermal energy at scale.

15 GEOTHERMAL ENERGY↗
DOE OSTI2023

Attention for Causal Relationship Discovery from Biological Neural Dynamics

This paper explores the potential of the transformer models for learning Granger causality in networks with complex nonlinear dynamics at every node, as in neurobiological and biophysical networks. Our study primarily focuses on a proof-of-concept investigation based on simulated neural dynamics, for which the ground-truth causality is known through the underlying connectivity matrix. For transformer models trained to forecast neuronal population dynamics, we show that the cross-attention module effectively captures the causal relationship among neurons, with an accuracy equal to or superior to that of the most popular Granger causality discovery method. While we acknowledge that real-world neurobiology data will bring further challenges, including dynamic connectivity and unobserved variability, this research offers an encouraging preliminary glimpse into the utility of the transformer model for causal representation learning in neuroscience.

Lu, Ziyu↗
DOE OSTI2024

Reduced‐Order Probabilistic Emulation of Physics‐Based Ring Current Models: Application to RAM‐SCB Particle Flux

Abstract In this work, we address the computational challenge of large‐scale physics‐based simulation models for the ring current. Reduced computational cost allows for significantly faster than real‐time forecasting, enhancing our ability to predict and respond to dynamic changes in the ring current, valuable for space weather monitoring and mitigation efforts. Additionally, it can also be used for a comprehensive investigation of the system. Thus, we aim to create an emulator for the Ring current‐Atmosphere interactions Model with Self‐Consistent magnetic field (RAM‐SCB) particle flux that not only improves efficiency but also facilitates forecasting with reliable estimates of prediction uncertainties. The probabilistic emulator is built upon the methodology developed by Licata and Mehta (2023), https://doi.org/10.1029/2022sw003345 . A novel discrete sampling is used to identify 30 simulation periods over 20 years of solar and geomagnetic activity. Focusing on a subset of particle flux, we use Principal Component Analysis for dimensionality reduction and Long Short‐Term Memory (LSTM) neural networks to perform dynamic modeling. Hyperparameter space was explored extensively resulting in about 5% median symmetric accuracy across all data sets for one‐step dynamic prediction. Using a hierarchical ensemble of LSTMs, we have developed a reduced‐order probabilistic emulator (ROPE) tailored for time‐series forecasting of particle flux in the ring current. This ROPE offers accurate predictions of omnidirectional flux at a single energy with no pitch angle information, providing robust predictions on the test set with an error score below 11% and calibration scores under 8% with bias under 2% providing a significant speed up as compared to the full RAM‐SCB run.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Rock Physics-Based Data Assimilation of Integrated Continuous Active-Source Seismic and Pressure Monitoring Data during Geological Carbon Storage

Summary There has been substantial controversy concerning the role of geological carbon storage (GCS) in sequestering anthropogenic carbon emissions to mitigate climate change and global warming. Arguments center on the inability to monitor a geological storage site precisely and continuously, especially highlighting the associated costs and spatiotemporal trade-offs when using conventional subsurface monitoring techniques (well logs, core samples, chemical tracers, and 4D seismics). Active surveillance of GCS sites is essential for managing and mitigating potential leaks but is also required by regulation. With the goal of enhancing the monitoring capability at GCS sites, we present a rock physics-based joint data assimilation model to study a popular GCS site at Cranfield, Mississippi, USA. Synthetic continuous active-source seismic monitoring (CASSM) data (in the form of Vp and Qp measurements) and wellbore pressure monitoring data are assimilated with an ensemble of reservoir realizations to monitor gas saturation and reservoir pressure changes over a period of 100 years. Synthetic seismic attributes are generated using rock physics models (RPMs) and wellbore pressure monitoring data are extracted from the ground truth. Two assimilation methods, ensemble Kalman filter (EnKF) and ensemble Kalman smoother (EnKS), are tested in an observation system simulation experiment (OSSE) environment to assess the prediction accuracy of the individual and composite observation systems. The joint monitoring system achieves more accurate estimates of gas saturation and pressure, across the time span from start of injection to end of forecast, as compared to a single type of monitoring tool and irrespective of data assimilation algorithm choice. These results indicate that jointly assimilated data from two types of sensors (in this case, crosswell seismic and downhole pressure) may lead to a more risk-reducing monitoring design. One would expect that more data, vis-à-vis inclusion of a new sensor type, will improve the accuracy of any GCS monitoring system. However, from a practical standpoint, one important question is whether such a gain in accuracy is worth the additional cost associated with the new sensor. This paper focuses on quantifying the gain in accuracy, such that a practitioner can answer this question.

Engineering↗
DOE OSTI2024

Improved Subseasonal Forecasting of Extreme Polar Vortices Using Machine Learning

Our research was focused on forecasting the position and shape of the winter stratospheric polar vortex at a subseasonal timescale of 15 days in advance. To achieve this, we employed both statistical and neural network machine learning techniques. The analysis was performed on 42 winter seasons of reanalysis data provided by NASA giving us a total of 6,342 days of data. The state of the polar vortex for determined by using geometric moments to calculate the centroid latitude and the aspect ratio of an ellipse fit onto the vortex. Timeseries for thirty additional precursors were calculated to help improve the predictive capabilities of the algorithm. Feature importance of these precursors was performed using random forest to measure the predictive importance and the ideal number of precursors. Then, using the precursors identified as important, various statistical methods were tested for predictive accuracy with random forest and nearest neighbor performing the best. An echo state network, a type of recurrent neural network that features sparsely connected hidden layer and a reduced number of trainable parameters that allows for rapid training and testing, was also implemented for the forecasting problem. Hyperparameter tuning was performed for each methods using a subset of the training data. The algorithms were trained and tuned on the first 41 years of data, then tested for accuracy on the final year. In general, the centroid latitude of the polar vortex proved easier to predict than the aspect ratio across all algorithms. Random forest outperformed other statistical forecasting algorithms overall but struggled to predict extreme values. Forecasting from echo state network suggested a strong predictive capability past 15 days, but further work is required to fully realize the potential of recurrent neural network approaches.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

A Statistical Evaluation of WRF-LES Trace Gas Dispersion Using Project Prairie Grass Measurements

In recent years, new measurement systems have been deployed to monitor and quantify methane emissions from the natural gas sector. Large-eddy simulation (LES) has complemented measurement campaigns by serving as a controlled environment in which to study plume dynamics and sampling strategies. However, with few comparisons with controlled-release experiments, the accuracy of LES for modeling natural gas emissions is poorly characterized. In this paper, we evaluate LES from the Weather Research and Forecasting (WRF) Model against Project Prairie Grass campaign measurements and surface layer similarity theory. Using WRF-LES, we simulate continuous emissions from 30 near-surface trace gas sources in two stability regimes: strong convection and weak convection. We examine the impact of grid resolutions ranging from 6.25 to 52 m in the horizontal dimension on model results. We evaluate performance in a statistical framework, calculating fractional bias and conducting Welch’s t tests. WRF-LES accurately simulates observed surface concentrations at 100 m and beyond under strong convection; simulated concentrations pass t tests in this region irrespective of grid resolution. However, in weakly convective conditions with strong winds, WRF-LES substantially overpredicts concentrations—the magnitude of fractional bias often exceeds 30%, and all but one t test fails. The good performance of WRF-LES under strong convection correlates with agreement with local free convection theory and a minimal amount of parameterized turbulent kinetic energy. The poor performance under weak convection corresponds to misalignment with Monin–Obukhov similarity theory and a significant amount of parameterized turbulent kinetic energy.

17 WIND ENERGY↗
DOE OSTI2024

High-resolution mapping of nitrogen oxide emissions in large US cities from TROPOMI retrievals of tropospheric nitrogen dioxide columns

Satellite-derived spatiotemporal patterns of nitrogen oxide (NOx) emissions can improve accuracy of emission inventories to better support air quality and climate research and policy studies. In this study, we develop a new method by coupling the chemical transport Model-Independent SATellite-derived Emission estimation Algorithm for Mixed-sources (MISATEAM) with a divergence method to map high-resolution NO x emissions across US cities using TROPOspheric Monitoring Instrument (TROPOMI) tropospheric nitrogen dioxide (NO 2 ) retrievals. The accuracy of the coupled method is validated through application to synthetic NO 2 observations from the NASA-Unified Weather Research and Forecasting (NU-WRF) model, with a horizontal spatial resolution of 4 km × 4 km for 33 large and mid-size US cities. Validation reveals excellent agreement between inferred and NU-WRF-provided emission magnitudes (R= 0.99, normalized mean bias, NMB = —0.01) and a consistent spatial pattern when comparing emissions for individual grid cells (R = 0.88 ± 0.06). We then develop a TROPOMI-based database reporting annual emissions for 39 US cities at a horizontal spatial resolution of 0.05° × 0.05° from 2018 to 2021. This database demonstrates a strong correlation (R= 0.90) with the National Emission Inventory (NEI) but reveals some bias (NMB = —0.24). There are noticeable differences in the spatial patterns of emissions in some cities. Our analysis suggests that uncertainties in TROPOMI-based emissions and potential misallocation of emissions and/or missing sources in bottom-up emission inventories both contribute to these differences.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Are long-term climate projections useful for on-farm adaptation decisions?

The current literature on climate services for farmers predominantly focuses on seasonal forecasts, with an assumption that longer-term climate projections may not be suitable for informing farming decisions. In this paper, we explore whether certain types of long-term climate projections may be useful for some specific types of farming decisions. Through interviews with almond tree crop farmers and farm advisors in California, we examine how farmers perceive the utility and accuracy levels of long-term climate projections and identify the types of projections that they may find useful. The interviews revealed that farmers often perceive long-term climate projections as an extension of weather forecasts, which can lead to their initial skepticism of the utility of such information. However, we also found that when farmers were presented with long-term trends or shifts in crop-specific agroclimatic metrics (such as chill hours or summer heat), they immediately perceived these as valuable for their decision-making. Hence, the manner in which long-term projections are framed, presented, and discussed with farmers can heavily influence their perception of the potential utility of such projections. The iterative conversations as part of the exploratory interview questions, served as a tool for “ joint construction of meaning” of complex and ambiguous terms such as “long-term climate projections,” “long-term decisions” and “uncertainty.” This in-turn supported a joint identification (and understanding) of the types of information that can potentially be useful for on-farm adaptive decisions, where the farmer and the interviewer both improvise and iterate to find the best types of projections that fit specific decision-contexts. Overall, this research identifies both the types of long-term climate information that farmers may consider useful, and the engagement processes that are able to effectively elicit farmers' long-term information needs.

Jagannathan, Kripa↗
DOE OSTI2020

Data-Driven Day-Ahead PV Estimation Using Autoencoder-LSTM and Persistence Model

Inherent variability in photovoltaic (PV) and associated impacts on power systems is a challenging problem for both the PV owners and the grid operators. Existing statistical and machine learning algorithms typically work well for weather conditions similar to historical data. Furthermore, uncertain weather conditions pose a great challenge to the estimation accuracy of the estimation models. With the enhanced integration of intelligent electronic devices and the realization of associated automation in the power grid, renewable energy data is becoming more accessible, which can be utilized by deep learning models and improve the PV power generation estimation accuracy. In this paper, a hybrid deep learning model driven by external weather data is proposed to do day-ahead PV output forecasting at 15-minute-interval. The proposed model is motivated by the recent advancement of Long-Short-Term-Memory (LSTM) networks and AutoEncoder (AE), which estimates uncertainties in sequence while making the prediction for complex weather conditions. Meanwhile, the persistence model (PM) is used to predict continuous sunny weather conditions. The forecasting result is validated with data from multiple locations

42 ENGINEERING↗
DOE OSTI2026

Iterative ML and Experiments for Emerging VOCs

SAND2026-17074O Iterative ML and Experiments for Emerging VOCs is a tool that analyzes and predicts the behaviors of SARS-CoV-2 variants. It processes experimental data on ACE2 (the receptor for the SARS-CoV-2 virus that allows it to infect the cell) and antibody binding using machine learning models, including neural networks, to forecast ACE2 interactions and variant expression. The tool employs transfer learning and global epistasis modeling, integrating public datasets with proprietary data to enhance prediction accuracy. Additionally, it fits concentration-response curves to determine dissociation constants and generates visualizations to support research findings, thereby aiding in the identification of new antibodies for emerging variants of concern. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

Sheffield, Thomas [Sandia National Lab. (SNL-NM), ↗
DOE OSTI2021

Wake Steering Wind Farm Control With Preview Wind Direction Information: Preprint

Wake steering is a wind farm control strategy in which upstream turbines operate with a yaw misalignment to deflect their wakes away from downstream turbines, yielding a net power gain for the wind plant. But the inability of wake steering controllers to perfectly track the wind direction leads to suboptimal performance. In this paper, we propose the use of preview wind direction measurements upstream of the turbine to improve controller performance by anticipating wind direction changes. Further, data from an operational wind plant are used to determine realistic preview measurement accuracy. Using the FLORIS engineering wind farm control tool, we compare the performance of standard and preview-enabled baseline and wake steering control for a two-turbine array during below-rated operation. Assuming perfect preview information, preview-based wake steering increases energy production by the equivalent of 9.5% of the baseline wake losses, compared to a wake loss recovery of 5.6% with standard wake steering. However, when realistic measurement accuracy is included, the preview-based controller provides no advantage over standard control, motivating the need for more sophisticated control and wind direction forecasting strategies.

49 EE - Wind and Water Power Program - Wind (EE-4W↗
DOE OSTI2023

Deep Learning Accelerated History Matching and Forecasting in Geologic CO 2 Sequestration [Slides]

We have successfully developed a predictive workflow for the Geologic CO 2 Sequestration using a deep learning model based on the Fourier Neural Operator. The workflow has high accuracy for predicting pressure & saturation during the injection and post-injection periods and has decent accuracy for predicting water and CO 2 production rates during injection period. It is necessary to train exclusive DNN models for pressure prediction in long term GCS, but for saturation prediction, we can use a single model to predict it.

58 GEOSCIENCES↗
DOE OSTI2025

Super-Resolution for Renewable Energy Resource Data with Wind from Reanalysis Data and Application to Ukraine

With a potentially increasing share of the electricity grid relying on wind to provide generating capacity and energy, there is an expanding global need for historically accurate, spatiotemporally continuous, high-resolution wind data. Conventional downscaling methods for generating these data based on numerical weather prediction have a high computational burden and require extensive tuning for historical accuracy. In this work, we present a novel deep learning-based spatiotemporal downscaling method using generative adversarial networks (GANs) for generating historically accurate high-resolution wind resource data from the European Centre for Medium-Range Weather Forecasting Reanalysis version 5 data (ERA5). In contrast to previous approaches, which used coarsened high-resolution data as low-resolution training data, we use true low-resolution simulation outputs. We show that by training a GAN model with ERA5 as the low-resolution input and Wind Integration National Dataset Toolkit (WTK) data as the high-resolution target, we achieved results comparable in historical accuracy and spatiotemporal variability to conventional dynamical downscaling. This GAN-based downscaling method additionally reduces computational costs over dynamical downscaling by two orders of magnitude. We applied this approach to downscale 30 km, hourly ERA5 data to 2 km, 5 min wind data for January 2000 through December 2023 at multiple hub heights over Ukraine, Moldova, and part of Romania. With WTK coverage limited to North America from 2007–2013, this is a significant spatiotemporal generalization. The geographic extent centered on Ukraine was motivated by stakeholders and energy-planning needs to rebuild the Ukrainian power grid in a decentralized manner. This 24-year data record is the first member of the super-resolution for renewable energy resource data with wind from the reanalysis data dataset (Sup3rWind).

17 WIND ENERGY↗
DOE OSTI2021

Can artificial intelligence and data-driven machine learning models match or even replace process-driven hydrologic models for streamflow simulation?: A case study of four watersheds with different hydro-climatic regions across the CONUS

With recent developments in computational techniques, Data-driven Machine Learning Models (DMLs) have shown great potential in simulating streamflow and capturing the rainfall-runoff relationship in given watersheds, which are traditionally fulfilled by Process-based Hydrologic Models (PHMs). There are debates on whether the DMLs can outperform and possibly replace the classical PHMs for streamflow simulation and river forecasting, but no clear conclusions have been made. This study aims to investigate whether the newer DMLs have any potential in further improving the simulation accuracy of classical PHMs, and vice versa. To do this, we compared a few popular PHMs and DMLs over four watersheds across the Continental US (CONUS) that are associated with different input, climate, and regional conditions. A total of five hydrologic models were chosen, including (1) two classical lumped models, i.e., the Sacramento Soil Moisture Accounting (SAC-SMA) and Xinanjiang (XAJ); (2) one modern distributed model, termed Coupled Routing and Excess Storage (CREST); (3) and two DMLs including an Artificial Neural Networks (ANN) and a deep learning model, termed Long Short Term Memory (LSTM). Our results demonstrated that the DMLs still significantly biased when using the baseline input scenario with the PHMs. However, the DMLs fed with delayed input scenarios had great potential and can reach high simulation accuracy. The DMLs, especially the ANN, outperformed other employed models under the rainfall-runoff relationship in which rainfall dominantly drives. Furthermore, the DMLs also showed better performance in the high-flow regime, while the PHMs had a better performance for the low-flow regime, implying both PHMs and DMLs have their own merits and are worthy of joint development. In general, our study indicated a great potential of using DMLs to simulate streamflow, but further studies are still needed to verify the transferability and scalability of DMLs in large-scale experiments, such as the Distributed Model Intercomparison Projects 1&2 conducted by National Weather Services but to compare modern DMLs and PHMs.

58 GEOSCIENCES↗