Validating the NASA Goddard single-moment bulk water microphysics scheme using surface, aircraft, and radar observations

Operational forecast centers are transitioning to the use of high resolution models, which incorporate single-moment, bulk water parameterization schemes that simulate the content and size distribution for various hydrometeors. These schemes assume fixed values for properties related to snow crystals, such as their size distribution and density, but new functions can be implemented that allow variability as a function of ambient temperature or position within the cloud column.;Aircraft and surface observations of snow crystals were obtained during the Canadian CloudSat/CALIPSO Validation Project and used to verify the assumptions made within the NASA Goddard single-moment microphysics scheme. The fixed distribution intercept and density assumed by the scheme were incapable of representing observed properties of snow crystal populations, and the terminal velocity-diameter relationship produced a fall speed much lower than was observed among crystals at the surface. In comparisons to radar observations, the NASA Goddard scheme produced a reasonable depiction of ground-based, C-band radar reflectivity distributions, but an assumption of scattering by Mie spheres of uniform snow density was incapable of reproducing the shape or distribution of reflectivity obtained from CloudSat.;Two parameterizations were developed to calculate the snow distribution slope parameter as either a function of temperature, or a column integrated quantity. Snow density is then determined as a function of the modified slope value. Although either approach improved upon the previous use of fixed values, the success of a temperature based approach was limited in regions where the temperature profile was isothermal or inverted. New parameterizations produced a reasonable fit to observed reflectivity distributions from ground-based radar, but the simulations of CloudSat reflectivity requires the use of crystal scattering databases to obtain backscatter characteristics of crystal aggregates. Simulation of reflectivity by Mie spheres is insufficient. These parameterizations show that the NASA Goddard scheme can be modified to produce a better representation of snow crystals at the surface and aloft, in support of improved weather forecasts and satellite simulators.