Study of the impact on how urban areas are represented in computer forecast models published in Geophysical Research Letters – Hurricane Research Division

Urban flooding during tropical cyclone landfall can lead to major impacts, and understanding and forecasting it is an important and emerging research area. Two recent Atlantic hurricanes, Florence in 2018 and Harvey in 2017, are forecast using a state-of-the-art computer model with two different ways to include urban areas. The model can predict rainfall more accurately using a simple representation of the urban surfaces than a more complex representation in these two cases.

Hurricanes Harvey (2017) and Florence (2018) both caused extensive flooding in urban areas when they made landfall. To improve forecasts of similar events, two different ways to represent cities in computer models were tested in the Weather Research and Forecasting (WRF) model, and these forecasts were compared with those that included no specific representation of urban areas outlined by the thin black lines (see Figure).  The Urban Canopy Model (UCM) provides a simple way to represent the surface, whereas the Building Environment Parameterization (BEP) model is more complex. The three model simulations shown in the figure show how the rainfall forecast changes using the different ways to include the urban areas in the model.

■ Important Conclusions:

• The two ways to represent the urban areas in the computer model made better rainfall forecasts compared to not representing the urban areas at all.
• The simple representation of the urban areas made better rainfall forecasts in urban areas than the more complex method.

The study can be found at https://doi.org/10.1029/2023GL104078. For more information, contact aoml.communications@noaa.gov.