Search for: All records

To improve accessibility and community knowledge of applications in the Lidar Radar Open Software Environment (LROSE), a team from the National Science Foundation (NSF) National Center for Atmospheric Research, Colorado State University, and NSF Unidata has developed a lidar and radar meteorology science gateway deployed on the NSF Jetstream2 cloud. Utilizing the “Zero to JupyterHub with Kubernetes” workflow, the science gateway integrates LROSE with other lidar and radar meteorology software packages. This integration allows users to execute applications directly from the JupyterLab terminal, streamlining the creation of datasets for further analysis and visualization within Jupyter notebooks. By combining traditional command-line operations with modern Python-based tools for data analysis and visualization, this gateway provides a robust end-to-end solution that caters to both educational and research needs. The gateway has already facilitated LROSE instructional workshops and classroom exercises. Our work demonstrates the significant potential of merging established scientific computing techniques with advanced Python environments, opening new avenues for computational science education and research. The LROSE team has acquired successive allocations on the NSF Jetstream2 cloud at Indiana University through ACCESS. To develop the LROSE Science Gateway, we employed the “Zero to JupyterHub with Kubernetes” workflow ported to the NSF Jetstream2 cloud, enabling rapid and scalable deployment to accommodate a variable number of users. Authentication is managed through either GitHub OAuth or temporary credentials, depending on the situation. Since LROSE is a collection of C/C++ applications, we configured Docker containers based on the Jupyter Docker Stack to integrate the LROSE software, available via the JupyterLab terminal. These containers also include Conda package manager environments equipped with Python packages like Py-ART, CSU RadarTools, and Metpy for further data analysis. A shared drive accessible to all participants contains instructional datasets for lidar and radar data analysis. Tutorials take the form of Jupyter notebooks for use by individuals, in classroom exercises, or at instructional workshops. Some tutorials are complete with pre-loaded examples to quickly visualize workflows and results. Other tutorials guide students how to run the applications independently. All tutorials are hosted on the LROSE Science Gateway GitHub repository, which is open to contributions from colleagues and community members. Future plans include an "intermediate" level workshop on SAMURAI, one of the multi-Doppler wind applications of the LROSE suite. Additionally, work is currently underway to run GUI applications in the same browser-based JupyterLab environment. GUI applications for radar and lidar data visualization utilize the QT framework and present unique technical challenges. The techniques to accomplish GUI access have immediate applications for other GUI programs, such as NSF Unidata's IDV and their version of the AWIPS CAVE data visualization tools. Lastly, as demand for the resources found on the gateway increases, it becomes increasingly important to efficiently manage the Jetstream2 resources allocated by the ACCESS program. LROSE, NSF Unidata, San Diego Supercomputing Center (SDSC), and Indiana University staff are working together to deploy and evaluate Kubernetes cluster auto-scaling. With auto-scaling, resources will no longer sit idle while awaiting new logins and will instead be provisioned on-demand. 

This paper introduces a lidar and radar meteorology science gateway deployed on the NSF Jetstream2 cloud, designed to enhance educational and research activities in atmospheric science. Utilizing the "Zero to JupyterHub with Kubernetes" workflow, we have created a science gateway that integrates lidar and radar meteorology software packages, notably the Lidar Radar Open Software Environment (LROSE). This integration allows users to execute applications directly from the JupyterLab terminal, streamlining the creation of datasets for further anal- ysis and visualization within Jupyter notebooks. By combining traditional command-line operations with modern Python-based tools for data analysis and visualization, this gateway provides a robust end-to-end solution that caters to both educational and research needs. The gateway has already been vital in facilitating LROSE instructional workshops and will see future enhancements such as GPU acceleration to boost performance. Our work demonstrates the significant potential of merging established scientific computing techniques with advanced Python environments, opening new avenues for computational science education and research. 

Estimates of the surface wind field in a tropical cyclone (TC) are required in real time by operational forecast centers to warn the public about potential impacts to life and property. In‐situ aircraft data must be adjusted from flight level to surface using wind reductions (WRs) since the aircraft cannot fly too low due to safety concerns. Current operational WRs do not capture all the variability in the TC surface wind field. In this study, an observational data set of Stepped Frequency Microwave Radiometer (SFMR) surface wind speeds that are collocated with flight‐level predictors is used to analyze the variability of WRs with respect to aircraft altitude and TC storm motion and intensity. The Surface Winds from Aircraft with a Neural Network (SWANN) model is trained on the observations with a custom loss function that prioritizes accurate prediction of relatively rare high‐wind observations and minimization of variance in the WRs. The model is capable of learning physical relationships that are consistent with theoretical understanding of the TC boundary layer. Radar‐derived wind fields at flight level and independent dropwindsonde in‐situ surface wind measurements are used to validate the SWANN model and show improvement over the current operational procedure. A test case shows that SWANN can produce a realistic asymmetric surface wind field from a radar‐derived flight‐level wind field which has a maximum wind speed similar to the operational intensity, suggesting promise for the method to lead to improved real‐time TC intensity estimation and prediction in the future. 

Abstract Taiwan regularly receives extreme rainfall due to seasonal mei-yu fronts that are modified by Taiwan’s complex topography. One such case occurred between 1 and 3 June 2017 when a mei-yu front contributed to flooding and landslides from over 600 mm of rainfall in 12 h near the Taipei basin, and over 1500 mm of rainfall in 2 days near the Central Mountain Range (CMR). This mei-yu event is simulated using the Weather Research and Forecasting (WRF) Model with halved terrain as a sensitivity test to investigate the orographic mechanisms that modify the intensity, duration, and location of extreme rainfall. The reduction in WRF terrain height produced a decrease in rainfall duration and accumulation in northern Taiwan and a decrease in rainfall duration, intensity, and accumulation over the CMR. The reductions in northern Taiwan are linked to a weaker orographic barrier jet resulting from a lowered terrain height. The reductions in rainfall intensity and duration over the CMR are partially explained by a lack of orographic enhancements to mei-yu frontal convergence near the terrain. A prominent feature missing with the reduced terrain is a redirection of postfrontal westerly winds attributed to orographic deformation, i.e., the redirection of flow due to upstream topography. Orographically deforming winds converge with prefrontal flow to maintain the mei-yu front. In both regions, the decrease in mei-yu front propagation speed is linked to increased rainfall duration. These orographic features will be further explored using observations captured during the 2022 Prediction of Rainfall Extremes Campaign in the Pacific (PRECIP) field campaign. Significance StatementThis study examines the impact of terrain on rainfall intensity, duration, and location. A mei-yu front, an East Asian weather front known for producing heavy, long-lasting rainfall, was simulated for an extreme rain event in Taiwan with mountain heights halved as a sensitivity test. Reducing terrain decreased rainfall duration in northern and central Taiwan. Decreases in rainfall duration for both regions is attributed to increased mei-yu front propagation speed. This increase in northern Taiwan is attributed to a weakened barrier jet, a low-level jet induced by flow blocked by the steep mountains of Taiwan. A unique finding of this work is a change in winds north of the front controlling movement of the front near the mountains in central Taiwan. 

This study analyzes an ensemble of numerical simulations of a heavy rainfall event east of Taiwan on 9 June 2020. Heavy rainfall was produced by quasi-stationary back-building mesoscale convective systems (MCS) associated with a mei-yu front. Global model forecast skill was poor in location and intensity of rainfall. The mesoscale ensemble showed liminal conditions between heavy rainfall or little to no rainfall. The two most accurate and two least accurate ensemble members are selected for analysis via validation against radar-estimated rainfall observations. All members feature moist soundings with low levels of free convection (LFC) and sufficient instability for deep convection. We find that stronger gradients in 100-m θe and θv in the most accurate members associated with a near-surface frontal boundary focus the lifting mechanism for deep, moist convection and enhanced rainfall. As the simulations progress, stronger southerly winds in the least accurate members advect drier mid-level air into the region of interest and shift the near-surface boundary further north and west. Analysis of the verification ensemble mean analysis reveals a strong near-surface frontal boundary similarly positioned as in the most accurate members and dry air aloft more similar to that in the least accurate members, suggesting that the positioning of the frontal boundary is more critical to accurately reproducing rainfall patterns and intensity in this case. The analyses suggest that subtle details in the simulation of frontal boundaries and mesoscale flow structures can lead to bifurcations in producing extreme or almost no rainfall. Implications for improved probabilistic forecasts of heavy rainfall events will be discussed. 

Abstract Polarimetric coastal radar data are used to compare the rainfall characteristics of Hurricanes Harvey (2017) and Florence (2018). Intense rainfall was an infrequent yet important contributor to the total rainfall in Harvey, but its relative contribution varied spatially. The total rainfall over land maximized near the coast over Beaumont, TX, due to intense convection resulting from prolonged onshore flow downshear from the circulation center. Overall, polarimetric radar observations in Harvey show a dominance of high concentrations of small‐to‐medium drops, consistent with prior tropical cyclone studies. The microphysical characteristics were spatially and temporally inhomogeneous however, with larger drops more frequent on 27 August and higher number concentrations more frequent on 28 and 30 August. The polarimetric variables and raindrop characteristics observed during Florence share broad similarities to Harvey, but had reduced variability, fewer observations of stronger reflectivity and differential reflectivity, and a lower frequency of high number concentrations and medium‐sized drops. The radar data indicate Florence had reduced coverage of stronger convection compared to Harvey. We hypothesize that differences in storm motion, intensity decay rates, and vertical wind shear produce the distinct precipitation structures and microphysical differences seen in Harvey and Florence.