The combined team from the National Center for Atmospheric Research (NCAR) and the University of Miami (UM) produced from this project: 1) improvements to two numerical weather models that are NSF resources; 2) advances in understanding of landfalling hurricanes and how buildings in coastal cities shape the wind in such storms; 3) datasets from physical experiments in wind tunnels and in-situ observations of hurricanes; 4) articles and presentations delivered to international audiences; and 5) education and training of multiple early-career scientists.

The numerical weather models improved as the result of the project are Cloud Model 1 (CM1) and the Weather Research and Forecasting (WRF) Model, both publicly available NSF resources.  The team introduced into CM1 a way of placing obstacles such as buildings at the model's lower boundary, for the first time allowing CM1 to be used to simulate how individual buildings affect the wind from street level to building top.  The team improved the WRF Model in several ways.  First, a previously overlooked bug was uncovered that can lead to large, unrealistic oscillations of atmospheric pressure at the beginning of some hurricane simulations.  Second, the team modified a widely used module used for representing the atmospheric boundary layer, the lower part of the atmosphere most affected by everything on the planet's surface (terrain, plants, buildings, etc.).  Until this project, that module could not be used in combination with two other standard modules that are integral to simulating weather in cities.  The third modification to the WRF Model was to refine the drag coefficient used to calculate how much buildings retard wind in urban areas.

The team used idealized simulations and observations of real hurricanes to reveal that the friction that landcover imposes on a hurricane produces significant changes in the lower parts of storms, including to accelerate how wind converges inward toward the storm center.  How moist the ground is beneath a landfalling hurricane was also found to be important, with higher soil moisture leading to hurricanes that do not weaken as quickly.  When the team performed more complex and realistic simulations, focusing on Hurricane Wilma (2005), the WRF Model was found capable of predicting realistic details in wind at individual locations just above the ground as long as the simulated hurricane's overall track, intensity, and size were fairly accurate.  A separate study of Hurricane Irma (2017) showed that model predictions of some landfalling hurricanes are sensitive to how the aggregate effects of buildings and other structures are depicted in the model, and to the drag coefficient.  As expected, the team found that heights and spacings of buildings in coastal cities dramatically affect how strong and turbulent the wind is at pedestrian level.  This was demonstrated through idealized simulations at a high enough resolution to reveal sudden, small-scale variations in the turbulent wind blowing through street canyons between adjacent buildings.  It is sometimes suggested that geoengineering might partially mitigate the threat hurricanes pose to coastal cities.  Modeling conducted during the project demonstrates that one form of geoengineering, cooling the ocean to reduce the energy available to hurricanes, might seem possible based on theory, but applying this approach in the real world is entirely unrealistic because of the prohibitive resources and effort that would be required. 

In addition to the datasets produced by the modeling described above, the project also yielded datasets from wind tunnels, and provided the team with multiple opportunities to observe actual hurricanes in nature.  The wind-tunnel data come from one set of experiments at the NSF Natural Hazards Engineering Research Infrastructure (NHERI) wind tunnel at the University of Florida (UF), and from another set at a wind tunnel operated by the School of Architecture and Wind Engineering of the Tokyo Polytechnic University, Japan. Both sets of experiments relied on high-resolution sensors that recorded observations of wind around building-like obstacles in the tunnels.  For the hurricane observations, a doctoral student on the team participated in two NOAA P3 "Hurricane Hunter" flights into Hurricane Lorenzo (2019).  Also in 2019, two other members of the team traveled to Wilmington, NC to assist UF in taking wind measurements of Hurricane Dorian in both openly exposed and suburban environments.

The team published eight articles and presented seventeen talks and posters.  Three more articles will be completed after the project's end.  One PhD dissertation was supported.  The team organized and hosted at the 34th Conference on Hurricanes and Tropical Meteorology the oral session "Hurricanes hazards at Landfall" and the poster session "Hurricanes at landfall in the urban environment."  Members of the team were session chairs.

Early-career scientists supported by this project were a postdoctoral scientist at NCAR; a PhD student at UM, who completed his dissertation in 2021; one undergraduate student at UM, and one MS student at UM who is expected to graduate in 2023.

Last Modified: 12/27/2022
Modified by: David S Nolan