Cyril Julien Morcrette — Met Office - UK
Kwinten Van Weverberg — Met Office - UK
Category
General Topics – Cloud
Description
Time-height plot of the water content (ice + liquid) at the SGP during the MC3E campaign. Shown are the observed water content (top; regridded to the MetUM grid) and the MetUM water water content (bottom). The black horizontal lines at 3 km and 6 km altitude denote the boundaries between the low-/mid-level clouds and between the mid-/high-level clouds respectively. Areas with no data in the observations are shaded grey in all three panels.
Many Global Circulation Models (GCMs) exhibit significant warm biases over the mid-latitude continental regions, such as the Southern Great Plains (SGP) in North America. Typically, this bias is present both in short-term forecasts of a few days, as well as in long multi-year climate simulations. A number of hypotheses have been proposed to explain this bias, ranging from a poor representation of the soil-vegetation-atmosphere heat exchange, to a lack of low-level boundary-layer clouds to misrepresentation of deep convective storms.
This study revolves around the exploitation of multi-instrument synergies at the ARM SGP central facility in Oklahoma to identify what role clouds play in the creation of the warm bias in a number of GCMs. The goal of this study is twofold. Firstly, the availability of a variety of collocated, sub-time step-level observations allows for an alternative approach in model evaluation. Based on the occurrence of clouds at three tropospheric levels, a number of cloud regimes are first identified. It is then established which regimes contribute most to the growth of the surface temperature bias at a time-step level.
This approach was applied to a 6-week simulation using the MetUM GCM, coinciding with the Midlatitude Continental Convective Clouds Experiment (MC3E) Field Campaign during Spring 2011. We could show that the primary source of the excess energy at the surface originates from missed convection in the afternoon and excessive low-level cloudiness during the night. However, a poor representation of the surface energy balance modulates the absorption and release of this excess heat over the course of a diurnal cycle, leading to the fastest growth of the bias during the night and a decrease of the bias during the early daytime.
A second goal of this study is to exploit sub-grid information from the ARM SGP instruments to test and improve assumptions made in the parameterisation of clouds in a GCM. Cloud parameterisations typically make a number of assumptions about the sub-grid variability of moisture and temperature within a single grid-box. It will be shown that the cloud properties and the surface temperature bias at the SGP can be improved using the observed shape and width of the moisture variability during a 20-year climate simulation.
