Interrogating Tropical Cold Pools with DYNAMO Observations and Modeling

The shallow-to-deep convection transition accompanying Madden-Julian Oscillation (MJO) initiations encompasses small-scale (temporal scales < 1 day and spatial scales < 100 km2 ) physical processes such as cold pools driven by evaporation of precipitation, which, until the DYNAMO experiment, remained poorly articulated. We hypothesize that cold pools are instrumental to the shallow-to-deep convection transition, and that the cold pools from less-organized, shallower convection (precipitating cloud tops between 3 to 10 km) are particularly instrumental to the MJO initiation, through organizing the transfer of moisture from near the surface to above the marine boundary layer. We propose an integrated observational and modeling strategy to articulate the role of all cold pools with the MJO cycle, with an emphasis on MJO initiation and shallow-to-deep convection transitions. Cold pool metrics (e.g., cold pool depth, strength and size as a function of rainfall rate) will be assessed for a range of observed and simulated cold pools and integrated with existing literature to assess sampling from the full convective spectrum.

The observational analysis will highlight highly-resolved measures of lower-atmosphere processes (e.g., moisture, vertical and horizontal velocities) from scanning Doppler lidar, and wind profiler, both on the R/V Revelle, with surface flux measurements connecting the cold pools with boundary layer thermodynamic changes. Dual-wavelength scanning radar and scanning microwave radiometer data gathered on Gan Island provide other cloud-humidity measurements. R/V Revelle flux measurements will form the basis for ocean flux estimates near Gan Island, and precipitation radar data and satellite images from both locations will be used to assess the degree of cloud mesoscale organization.

Specific case studies will be modeled using nested-Weather Research and Forecasting-Large Eddy Simulations (WRF-LES) at U of Miami, and the SAM model at U of Utah, the latter as a cloud-resolving model with a new turbulence closure (SAM-PDF). The nested-WRF simulation will possess both a highly-resolved (50 m spatial resolution) inner domain and incorporate realistic large-scale forcings through its outer domain. The modeling analysis assesses the cause and-effect convective triggering mechanisms to complement the observations. More idealized simulations will investigate the sensitivities of shallow cloud mesoscale organization to moisture and stability, zonal wind, and wind shear, and model parameters as well as to model representations of the microphysics. The SAM-PDF model will be run for the full four months of the DYNAMO IOP. Comparisons of the two models enhance robustness of results and techniques.

This proposal fits squarely with NOAA’s goals in “Understanding and Improving Prediction of Tropical Convection using Results from the DYNAMO Field Campaign”, which in turn directly contributes to NOAA’s Next-Generation Strategic Plan. Small(er) scale processes, i.e. cold pools, contribute to two of the DYNAMO organizing hypotheses for explaining the low prediction skill of MJO initiations within large-scale models. This project thereby contributes to improving climate model representations of cumulus clouds and cold pools, key uncertainties with implications for climate change predictions. A multi-disciplinary, multi-institutional team of investigators combining people at different career stages brings a fresh look to an old problem.