Impact of Data Assimilation on Simulations of Continental Shallow Cumulus near the ARM Southern Great Plains Site during HI-SCALE

In this study, the impact of atmospheric data assimilation (DA) on modeling of continental shallow cumulus (ShCu) clouds is explored for a case on 30 August 2016, as part of the Holistic Interactions of Shallow Clouds, Aerosols, and Land-Ecosystems (HI-SCALE) field campaign which was conducted near the U.S. Department of Energy’s (DOE) Atmospheric Radiation Measurement (ARM) Southern Great Plains (SGP) site. Coupled with the Weather Research and Forecasting (WRF) model, the Gridpoint Statistical Interpolation (GSI) system serves for optimization of model states with inclusion of valid observations. Two sets of experiments are conducted that demonstrate the sensitivity of data assimilation techniques as well as the impact of additional ARM SGP observations on both analyses and subsequent forecasts.

Comparing simulations with and without assimilation, analyses of the DA experiments exhibit better agreement with observations in atmospheric temperature, moisture, and wind. Minor discrepancies are found between radiosonde profiles and sensitivity experiments that use three different assimilation techniques (3DVar, 3DEnVar, and 4DEnVar), while additional assimilation of ARM profile-type observations (radiosonde, radar wind profiler, and raman lidar) further reduce the biases of atmospheric profiles near the ARM SGP site. Comprehensive examination of simulated ShCu is performed by using liquid water path (LWP) retrieved from GOES-13 satellite measurements and other sources. In addition to horizontal distribution of LWP, significant modifications are found in terms of the cloud initiation timing, fraction, and depth with the aid of DA process. Intercomparison between modeled and observed evolutions of atmospheric profiles and surface properties suggests thermodynamic profile (especially moisture at 2 to 4 km) modulates the initiation and vertical extent of ShCu, whereas the inhomogeneity of surface temperature (variance) plays a role in formation of cloud populations but is underpredicted in the model.