The impact on simulated bow echoes of changing grid spacing from 3 km to 1 km in WRF
Date
2023-12
Authors
Dodson, Dylan Jacob
Major Professor
Advisor
Gallus, William A
Gonzalex, Alex
Franz, Kristie
Committee Member
Journal Title
Journal ISSN
Volume Title
Publisher
Altmetrics
Abstract
Convection allowing models, such as the Weather Research and Forecasting (WRF) model, have shown some shortcomings when forecasting convective modes. This is especially true for the forecasting of bow echo events. Previous studies have suggested that finer grid spacing, specifically 1 km, do a better job of predicting bow echo events than 3 km grid spacing. However, these studies also suggest that the Morrison microphysical scheme is less affected by the grid spacing change, as it has less of an underprediction of bow echoes at 3 km than present when other schemes are used. The Thompson microphysical scheme, for instance, is more affected, showing more bow echo events at 1 km than at 3 km.
The current study expands upon these prior works by analyzing 10 observed bow echo events in the United States. These 10 events were simulated with the WRF model using 3 km and 1 km grid spacing with both the Morrison and Thompson microphysics schemes. The simulated and observed composite reflectivity were used to classify convective mode. Two skill measures were used to quantify how similar the simulated convective mode was to reality, with one evaluating all modes, and another specifically the bow echo modes. The general morphology skill score showed that the Thompson microphysics scheme was more impacted by the refined grid spacing, while the Morrison scores saw no meaningful difference. However, both the Thompson and Morrison bow echo scores saw improvement when grid spacing was refined, although the Thompson improvement was more significant. Additionally, near storm environments were analyzed to better understand why depictions of bow echoes changed in some cases as grid spacing was changed. A relationship was found between bow echo production and cold pool strength, as well as microphysical mixing ratio tendency fields. More numerous updrafts were also present at the finer grid spacings, allowing for the development of longer, intense lines of convection. It’s suspected that this promoted a positive feedback within the systems simulated on 1-km grids, resulting in longer lived bow echo simulations in more cases. 3-km simulations were found to be more reliant on large scale features, such as the low-level jet and surface boundaries, to help with the production of bow echoes.
Series Number
Journal Issue
Is Version Of
Versions
Series
Academic or Administrative Unit
Type
thesis