Tropical cyclones are one of the deadliest natural disasters in the world that cause significant damage to the environment and infrastructure. The Hurricane Boundary Layer (HBL) plays a major role in hurricane dynamics and its intensification. Most of the existing vertical diffusion parameterizations in the current numerical weather prediction models rely on the Planetary Boundary Layer (PBL) schemes. Previous studies (Momen et al. 2021; Romdhani et al. 2022) showed that there is a significant distinction between turbulence characteristics in HBLs and regular atmospheric boundary layers (ABLs) due to the strong rotational effects of hurricane flows. Nevertheless, such differences are not considered in the current PBL schemes, and they are primarily designed and tested for regular ABLs.
In this talk, we aim to bridge this knowledge gap by conducting real hurricane simulations using the Weather Research and Forecasting (WRF) model. We investigate the role of the PBL height and eddy momentum exchange coefficients in five intensifying hurricanes by probing the parameter space of the problem. Our simulations have shown that the most widely used WRF PBL schemes do not capture the hurricane intensification properly and underestimate their intensity.
We will demonstrate how limiting the amount of the vertical transport of momentum greatly benefits the skill of forecasting in major hurricane simulations. We will also present how changing the height of the PBL significantly impacts the accuracy of the forecasts. By reducing the PBL height, simulated hurricanes become stronger and larger – representing the actual rapid intensification process much more accurately. Not only changes are seen in the predicted wind intensities, but also remarkable impacts are observed in storm size, the radius of maximum wind speed, hurricane track, and minimum sea level pressure. The results of this study provide insights into the role of vertical diffusion parameterizations in hurricane dynamics. Our findings can be used to improve the accuracy of real hurricane forecasts in numerical weather prediction models.
more »
« less
Characterizing atmospheric turbulence over Maunakea through temporal tomography
Early adaptive optics (AO) systems were designed with knowledge of a site’s distribution of Fried parameter (r0) and Greenwood time delay (τ0) values. Recent systems have leveraged additional knowledge of the distribution of turbulence with altitude. We present measurements of the atmosphere above Maunakea, Hawaii and how the temporal properties of the turbulence relate to tomographic reconstructions. We combine archival telemetry collected by ‘imaka—a ground layer AO (GLAO) system on the UH88” telescope—with data from the local weather towers, weather forecasting models, and weather balloon launches, to study how frequently one can map a turbulent layer’s wind vector to its altitude. Finally, we present the initial results of designing a new GLAO control system based off of these results, an approach we have named “temporal tomography.”
more »
« less
- Award ID(s):
- 1910552
- NSF-PAR ID:
- 10373996
- Editor(s):
- Schmidt, Dirk; Schreiber, Laura; Vernet, Elise
- Date Published:
- Journal Name:
- Proc. SPIE 12185, Adaptive Optics Systems VIII
- Volume:
- 12185
- Page Range / eLocation ID:
- 60
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
more » « less
Abstract Future Arctic sea ice loss has a known impact on Arctic amplification (AA) and mean atmospheric circulation. Furthermore, several studies have shown it leads to a decreased variance in temperature over North America. In this study, we analyze results from two fully coupled Community Earth System Model (CESM) Whole Atmosphere Community Climate Model (WACCM4) simulations with sea ice nudged to either the ensemble mean of WACCM historical runs averaged over the 1980–99 period for the control (CTL) or projected RCP8.5 values over the 2080–99 period for the experiment (EXP). Dominant large-scale meteorological patterns (LSMPs) are then identified using self-organizing maps applied to winter daily 500-hPa geopotential height anomalies (
) over North America. We investigate how sea ice loss (EXP − CTL) impacts the frequency of these LSMPs and, through composite analysis, the sensible weather associated with them. We find differences in LSMP frequency but no change in residency time, indicating there is no stagnation of the flow with sea ice loss. Sea ice loss also acts to de-amplify and/or shift the that characterize these LSMPs and their associated anomalies in potential temperature at 850 hPa. Impacts on precipitation anomalies are more localized and consistent with changes in anomalous sea level pressure. With this LSMP framework we provide new mechanistic insights, demonstrating a role for thermodynamic, dynamic, and diabatic processes in sea ice impacts on atmospheric variability. Understanding these processes from a synoptic perspective is critical as some LSMPs play an outsized role in producing the mean response to Arctic sea ice loss. Significance Statement The goal of this study is to understand how future Arctic sea ice loss might impact daily weather patterns over North America. We use a global climate model to produce one set of simulations where sea ice is similar to present conditions and another that represents conditions at the end of the twenty-first century. Daily patterns in large-scale circulation at roughly 5.5 km in altitude are then identified using a machine learning method. We find that sea ice loss tends to de-amplify these patterns and their associated impacts on temperature nearer the surface. Our methodology allows us to probe more deeply into the mechanisms responsible for these changes, which provides a new way to understand how sea ice loss can impact the daily weather we experience.
-
Schmidt, Dirk ; Schreiber, Laura ; Vernet, Elise (Ed.)Adaptive Optics (AO) used in ground based observatories can be strengthened in both design and algorithms by a more detailed understanding of the atmosphere they seek to correct. Nowhere is this more true than on Maunakea, where a clearer profile of the atmosphere informs AO system development from the small separations of Extreme AO (ExAO) to the wide field Ground Layer AO (GLAO). Employing telemetry obtained from the ımaka GLAO demonstrator on the University of Hawaii 2.2-meter telescope, we apply a wind profiling method that identifies turbulent layer velocities through spatial-temporal cross correlations of multiple wavefront sensors (WFSs). We compare the derived layer velocities with nearby wind anemometer data and meteorological model predictions of the upper wind speeds and discuss similarities and differences. The strengths and limitations of this profiling method are evaluated through successful recovery of injected, simulated layers into real telemetry. We detail the profilers’ results, including the percentage of data with viable estimates, on four characteristic ımaka observing runs on open loop telemetry throughout both winter and summer targets. We report on how similar layers are to external measures, the confidence of these results, and the potential for future use of this technique on other multi conjugate AO systems.more » « less
-
more » « less
Abstract Aim Addressing global environmental challenges requires access to biodiversity data across wide spatial, temporal and taxonomic scales. Availability of such data has increased exponentially recently with the proliferation of biodiversity databases. However, heterogeneous coverage, protocols, and standards have hampered integration among these databases. To stimulate the next stage of data integration, here we present a synthesis of major databases, and investigate (a) how the coverage of databases varies across taxonomy, space, and record type; (b) what degree of integration is present among databases; (c) how integration of databases can increase biodiversity knowledge; and (d) the barriers to database integration.
Location Global.
Time period Contemporary.
Major taxa studied Plants and vertebrates.
Methods We reviewed 12 established biodiversity databases that mainly focus on geographic distributions and functional traits at global scale. We synthesized information from these databases to assess the status of their integration and major knowledge gaps and barriers to full integration. We estimated how improved integration can increase the data coverage for terrestrial plants and vertebrates.
Results Every database reviewed had a unique focus of data coverage. Exchanges of biodiversity information were common among databases, although not always clearly documented. Functional trait databases were more isolated than those pertaining to species distributions. Variation and potential incompatibility of taxonomic systems used by different databases posed a major barrier to data integration. We found that integration of distribution databases could lead to increased taxonomic coverage that corresponds to 23 years’ advancement in data accumulation, and improvement in taxonomic coverage could be as high as 22.4% for trait databases.
Main conclusions Rapid increases in biodiversity knowledge can be achieved through the integration of databases, providing the data necessary to address critical environmental challenges. Full integration across databases will require tackling the major impediments to data integration: taxonomic incompatibility, lags in data exchange, barriers to effective data synchronization, and isolation of individual initiatives.
-
Yortsos, Yannis (Ed.)more » « less
Abstract Transfer learning (TL), which enables neural networks (NNs) to generalize out-of-distribution via targeted re-training, is becoming a powerful tool in scientific machine learning (ML) applications such as weather/climate prediction and turbulence modeling. Effective TL requires knowing (1) how to re-train NNs? and (2) what physics are learned during TL? Here, we present novel analyses and a framework addressing (1)–(2) for a broad range of multi-scale, nonlinear, dynamical systems. Our approach combines spectral (e.g. Fourier) analyses of such systems with spectral analyses of convolutional NNs, revealing physical connections between the systems and what the NN learns (a combination of low-, high-, band-pass filters and Gabor filters). Integrating these analyses, we introduce a general framework that identifies the best re-training procedure for a given problem based on physics and NN theory. As test case, we explain the physics of TL in subgrid-scale modeling of several setups of 2D turbulence. Furthermore, these analyses show that in these cases, the shallowest convolution layers are the best to re-train, which is consistent with our physics-guided framework but is against the common wisdom guiding TL in the ML literature. Our work provides a new avenue for optimal and explainable TL, and a step toward fully explainable NNs, for wide-ranging applications in science and engineering, such as climate change modeling.
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1117/12.2628413
