Large-scale moisture budget and land-atmosphere coupling over US Southern Great Plains

Poster PDF

Description

Synoptic weather variability and long-term regional climate are significantly influenced by the interactive processes between land surface and the overlying atmospheric boundary layer (PBL) and clouds. While model simulations showed a strong land-atmosphere (L-A) coupling strength at the Atmospheric Radiation Measurement (ARM) Southern Great Plains (SGP) site, a relatively weak signal is found in observational studies. In the first part of the study, the large-scale atmospheric moisture budget and the L-A coupling at the ARM SGP is examined based on 10-yr warm-season observations. The moisture budget components are computed from ARM continuous forcing data (Xie et al. 2004) and the L-A coupling strength is quantified by linear correlation coefficient between selected pairs of land surface and atmosphere variables within the ‘LoCo Process Chain’ (Santanello et al. 2018). Overall, the result suggests a relatively more important role of surface evaporation on local convective events. The land surface control on the evolution of PBL is dependent on the leaf area index (LAI). With respect to various land covers, afternoon precipitation in grassland areas initiates earlier with larger evaporative fraction while regions covered by forest exhibits a much higher cloud fraction on fair-weather shallow cumulus regime (ShCu). The performance of reanalysis (NARR [North American Regional Reanalysis]) and model simulations (CAPT [cloud-associated parameterizations testbed], RRM [regional-refined model]) is evaluated in the second part, with a focus on the L-A coupling in clear-sky, ShCu, and late-afternoon deep convection regime (Zhang and Klein, 2010). Cases of reanalysis and model simulations are classified into ‘correct’ and ‘wrong’ cases depending on whether it successfully categorizes the day into the same convective regime as identified by ARM observations. Through investigation of the statistical characteristics of ‘correct’ and ‘wrong’ model behaviors, we aim to make diagnosis on model deficiencies and attribute model biases to parameterized processes. This work is performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344. LLNL-ABS-767929.