11B.3 Evaluating the Representation of the Land-Atmosphere Feedback during the 2017 Solar Eclipse in the High-Resolution Rapid Refresh (HRRR) Model

Thursday, 14 June 2018: 8:30 AM
Ballroom E (Renaissance Oklahoma City Convention Center Hotel)
David D. Turner, NOAA/ESRL, Boulder, CO; and J. B. Olson, J. Kenyon, V. Wulfmeyer, A. Behrendt, A. Choukulkar, T. A. Bonin, W. A. Brewer, R. K. Newsom, and D. Cook

On 21 August 2017, a solar eclipse occurred over the continental United States resulting in a rapid reduction and subsequent increase of solar radiation over a large region of the country. The eclipse’s effect on the land-atmosphere system is documented in unprecedented detail using a unique array of sensors deployed at the ARM Southern Great Plains site in north-central Oklahoma. The observations showed that turbulent fluxes of heat and momentum at the surface responded quickly to the change in solar radiation. The decrease in the sensible heat flux resulted in a decrease in the air temperature below 200 m, and a large decrease in turbulent motions throughout the boundary layer. Furthermore, the turbulent mixing in the boundary layer lagged behind the change in the surface fluxes, and this lag depended on the height above the surface. After the eclipse ended, the turbulent motions increased from the surface upward and the convective boundary layer was reestablished as the sensible heat flux recovered.

The High-Resolution Rapid Refresh (HRRR) model is an operational, hourly-updated numerical weather prediction model that runs at 3 km resolution over the continental United States. The HRRR was modified to include a module to simulate the obscuration of the sun by the moon during the solar eclipse. The evolution of the surface fluxes and the atmospheric response in the HRRR is evaluated using the observations at the SGP site. Initial results demonstrate that the model captured the decrease in the downwelling solar radiation at the surface well, that the response of the 2-m air temperature shows the same decrease in temperature as the observations, and that a low-level jet developed both in the observations and model during the eclipse. However, the comparison also identified areas where the model does not agree well with the observations, such as the depth of the convective boundary layer (CBL) before the eclipse, although the depth of the new CBL that developed after the eclipse was well captured, as well as the turbulent kinetic energy in the nocturnal residual layer that developed during the peak of the eclipse.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner