Does Model Calibration Reduce Uncertainty in Climate Projections?

Uncertainty in climate projections is large as shown by the likely uncertainty ranges in Equilibrium Climate Sensitivity (ECS) of 2.5-4K and in the Transient Climate Response (TCR) of 1.4-2.2K. Uncertainty in model projections could arise from the way in which unresolved processes are represented, the parameter values used, or the targets for model calibration. We show that, in two climate model ensembles which were objectively calibrated to minimise differences from observed large scale atmospheric climatology, uncertainties in ECS and TCR are about two to six times smaller than in the CMIP5 or CMIP6 multi-model ensemble. We also find that projected uncertainties in surface temperature, precipitation and annual extremes are relatively small. Residual uncertainty largely arises from unconstrained sea-ice feedbacks. The 20+ year old HadAM3 standard model configuration simulates observed hemispheric scale observations and pre-industrial surface temperatures about as well as the median CMIP5 and CMIP6 ensembles while the optimised configurations simulates these better than almost all the CMIP5 and CMIP6 models. Hemispheric scale observations and pre-industrial temperatures are not systematically better simulated in CMIP6 than in CMIP5 though the CMIP6 ensemble seems to better simulate patterns of large-scale observations than the CMIP5 ensemble and the optimised HadAM3 configurations. Our results suggest that most CMIP models could be improved in their simulation of large scale observations by systematic calibration. However, the uncertainty in climate projections (for a given scenario) likely largely arises from the choice of parametrisation schemes for unresolved processes (“structural uncertainty”), with different tuning targets another possible contributor.

Convolution Neural Network for the Prediction of Cochlodinium polykrikoides Bloom in the South Sea of Korea

In this study, the occurrence of Cochlodinium polykrikoides bloom was predicted based on spatial information. The South Sea of Korea (SSK), where C. polykrikoides bloom occurs every year, was divided into three concentrated areas. For each domain, the optimal model configuration was determined by designing a verification experiment with 1–3 convolutional neural network (CNN) layers and 50–300 training times. Finally, we predicted the occurrence of C. polykrikoides bloom based on 3 CNN layers and 300 training times that showed the best results. The experimental results for the three areas showed that the average pixel accuracy was 96.22%, mean accuracy was 91.55%, mean IU was 81.5%, and frequency weighted IU was 84.57%, all of which showed above 80% prediction accuracy, indicating the achievement of appropriate performance. Our results show that the occurrence of C. polykrikoides bloom can be derived from atmosphere and ocean forecast information.

RavenR v2.1.4: an open source R package to support flexible hydrologic modelling

In recent decades, advances in the flexibility and complexity of hydrologic models has enhanced their utility in scientific studies and practice alike. However, the increasing complexity of these tools leads to a number of challenges, including steep learning curves for new users and in the reproducibility of modelling studies. Here, we present the RavenR package, an R package that leverages the power of scripting to both enhance the usability of the Raven hydrologic modelling framework and provide complimentary analyses that are useful for modellers. The RavenR package contains functions that may be useful in each step of the model-building process, particularly for preparing input files and analyzing model outputs, and these tools may be useful even for non-Raven users. The utility of the RavenR package is demonstrated with the presentation of six use cases for a model of the Liard River basin in Canada. These use cases provide examples of visually reviewing the model configuration, preparing input files for observation and forcing data, simplifying the model discretization, performing reality checks on the model output, and evaluating the performance of the model. All of the use cases are fully reproducible, with additional reproducible examples of RavenR functions included with the package distribution itself. It is anticipated that the RavenR package will continue to evolve with the Raven project, and will provide a useful tool to new and experienced users of Raven alike.

High-Definition Hurricanes: Improving Forecasts with Storm-Following Nests

To forecast tropical cyclone (TC) intensity and structure changes with fidelity, numerical weather prediction models must be “high definition”, i.e., horizontal grid spacing ≤ 3 km, so that they permit clouds and convection and resolve sharp gradients of momentum and moisture in the eyewall and rainbands. However, resolutions in operational global models remain too coarse to accurately predict these structures that are critical to TC intensity. Storm-following nests are a solution to this problem because they are computationally efficient at fine resolutions, providing a practical approach to improve TC intensity forecasts. Under the Hurricane Forecast Improvement Program, the operational Hurricane Weather Research and Forecasting (HWRF) system was developed to include telescopic, storm-following nests for a single TC per model integration. Subsequently, HWRF evolved into a state-of-the-art tool for TC predictions around the globe, although its single-storm nesting approach does not adequately simulate TC-TC interactions as they are observed. Basin-scale HWRF (HWRF-B) was developed later with a multi-storm nesting approach to improve the simulation of TC-TC interactions by producing high-resolution forecasts for multiple TCs simultaneously. In this study, the multi-storm nesting approach in HWRF-B was compared with a single-storm nesting approach using an otherwise identical model configuration. The multi-storm approach demonstrated TC intensity forecast improvements, including more realistic TC-TC interactions. Storm-following nests developed in HWRF and HWRF-B will be foundational to NOAA’s next-generation hurricane application in the Unified Forecast System.

Are Compton-thin AGNs Globally Compton Thin?

We select eight nearby active galactic nuclei (AGNs) which, based on previous work, appear to be Compton-thin in the line of sight. We model with mytorus their broadband X-ray spectra from 20 individual observations with Suzaku, accounting self-consistently for Fe Kα line emission, as well as direct and scattered continuum from matter with finite column density and solar Fe abundance. Our model configuration allows us to measure the global, out of the line of sight, equivalent hydrogen column density separately from that in the line of sight. For 5 out of 20 observations (in 3 AGNs) we find that the global column density is in fact ≳1.5 × 1024 cm−2, consistent with the distant scattering matter being Compton-thick. For a fourth AGN, two out of five observations are also consistent with being Compton-thick, although with large errors. Some of these AGNs have been reported to host relativistically broadened Fe Kα emission. Based on our modeling, the Fe Kα emission line is not resolved in all but two Suzaku observations, and the data can be fitted well with models that only include a narrow Fe Kα emission line.

Optimal Performance and Modeling of Wireless Technology Enabling Smart Electric Metering Systems Including Microgrids

This work is focused on the performance analysis and optimal routing of wireless technology for intelligent energy metering, considering the inclusion of micro grids. For the study, a geo-referenced scenario has been taken into account, which will form the structure of a graph to be solved using heuristic-based algorithms. In the first instance, the candidate site of the world geography to perform the case study is established, followed by deploying infrastructure devices and determining variables and parameters. Then, the model configuration is programmed, taking into account that a set of nodes and vertices is established for proper routing, resulting in a preliminary wireless network topology. Finally, from a set of restrictions, a determination of users connected to the concentrator and optimal routing is performed. This procedure is treated as a coverage set problem. Consequently, to establish the network parameters, two restrictions are specifically considered, capacity and range; thus, can be determined the best technology to adapt to the location. Finally, a verification of the resulting network topologies and the performance of the infrastructure is done by simulating the wireless network. With the model created, scenarios are tested, and it is verified that the optimization model demonstrates its effectiveness.

Tropical Cyclone Intensification Simulated in the Ooyama-type Three-layer Model with a Multilevel Boundary Layer

The first successful simulation of tropical cyclone (TC) intensification was achieved with a three-layer model, often named the Ooyama-type three-layer model, which consists of a slab boundary layer and two shallow water layers above. Later studies showed that the use of a slab boundary layer would produce unrealistic boundary layer wind structure and too strong eyewall updraft at the top of TC boundary layer and thus simulate unrealistically rapid intensification compared to the use of a height-parameterized boundary layer. To fully consider the highly height-dependent boundary layer dynamics in the Ooyama-type three-layer model, this study replaced the slab boundary layer with a multilevel boundary layer in the Ooyama-type model and used it to conduct simulations of TC intensification and also compared the simulation with that from the model version with a slab boundary layer. Results show that compared with the simulation with a slab boundary layer, the use of a multilevel boundary layer can greatly improve simulations of the boundary-layer wind structure and the strength and radial location of eyewall updraft, and thus more realistic intensification rate due to better treatments of the surface layer processes and the nonlinear advection terms in the boundary layer. Sensitivity of the simulated TCs to the model configuration and to both horizontal and vertical mixing lengths, sea surface temperature, the Coriolis parameter, and the initial TC vortex structure are also examined. The results demonstrate that this new model can reproduce various sensitivities comparable to those found in previous studies using fully physics models.

An Informal Introduction to Numerical Weather Models with Low-Cost Hardware

Weather, climate, and other Earth system models are growing in complexity as computing resources and technologies continue to evolve with time. Thus, models are and will remain a vital tool for scientific research. Exposure and education on the workings of such models can generate interest towards atmospheric science, and it can increase scientific literacy amongst the general public. Additionally, studies have suggested that early exposure to these models can affect the career trajectory of students. However, gaining exposure and experience remains difficult outside of internships, research settings, and other professional endeavors. Some of these barriers can include hardware and computing costs, curriculum structure, and access to instructors. As a means of addressing these barriers, the goal of this work is to utilize low-cost hardware and abstract away some of the complexities of running a numerical weather model without sacrificing fidelity. The approach is to create a graphical user interface (GUI) where users can quickly configure the model, run it, and analyze the output without knowledge of model configuration, system architecture, or navigation via a command line interface. The Pi-WRF application is packaged such that users can download and run the model within a matter of minutes. The application is designed to promote informal learning through hands-on experience. It is targeted towards lower secondary level students, but it can scale across grade levels, and it can be adapted for general audiences.