Abstract
Results of a modeling study of the 24 May 1989 dryline are presented. A nonhydrostatic, two-dimensional version of the Colorado State University Regional Atmospheric Modeling System (CSU-RAMS) is used to deduce the impact of east-west variability of soil moisture and vegetation on convective boundary layer evolution and dryline formation. The effects of the initial moisture and wind fields and the impact of the Coriolis force on the model results are also examined. Model output is compared with special airborne and sounding observations of the 24 May dryline.
Several findings of an earlier observational study of the 24 May dryline are supported in the present study. The modeled drylines are broadly comparable to the observed dryline with respect to the following properties: 1) width, 10 km; 2) strong horizontal moisture and virtual potential temperature gradients, >4 g kg−1/10 km and 2 K/10 km; 3) strong horizontal convergence, updraft, W–E shear of N–S wind component, 8 × 10−4 s−1, 1 m s−1, 10 m s−1/10 km, respectively., 4) bulges of moisture and N–S wind component above the surface dryline location (due to combination of vertical mixing and vertical advection); and 5) elevated moist layer cast of the dryline during late afternoon (due to eastward advection of the moisture bulge).
A rather diffuse dryline begins to form by about noon from the combined effects of strong vertical mixing of moisture and westerly momentum to the west of the dryline. Convective boundary layers (CBLs) of differing depths develop on either side of the dryline, giving rise to horizontal airflow acceleration in low levels via a mesoscale hydrostatic pressure gradient east of the dryline. A vorticity dynamics analysis indicates that a horizontal virtual potential temperature gradient causes strong solenoidal forcing of horizontal vorticity (order 10−5 s−2) and a thermally direct vertical circulation that concentrates at the dryline. Low-level winds east of the dryline decelerate, generating convergence and resulting in the development of an easterly, subgeostrophic airflow component during midafternoon. Strong convergent frontogenesis contributes to rapid development of the horizontal thermal gradients at the dryline during middle to late afternoon. For example, convergent and net frontogenesis of vapor mixing ratio achieves order 10−7 g kg−1 m−1 S−1. Ageostrophic forcing helps maintain a low-level southerly jet east of the dryline during the daytime via the Coriolis term. Virtual temperature differences across the dryline during late afternoon, implying virtual density differences, are consistent with the notion that the dryline propagates relative to the ambient westerly shear according to the theoretical phase speed of a density current.
The formation of the dryline and the evolution of the CBL are sensitive to the east–west profile of sensible heating, which in turn is sensitive to the soil moisture. A west-to-east change of volumetric soil moisture from 0.35 to 0.5 over 50 km (0.15/50 km), resulting in a sensible heat flux gradient in the surface layer of about 100 W m−2/50 km, is sufficiently large to cause a dryline to form. Drylines do not form in test cases where volumetric soil moisture is horizontally homogeneous with values of 0.5 or 0.35. A mesoscale convergence line, characterized by a strong updraft collocated with a weak moisture gradient, forms in the latter case. Horizontal variations of sensible beat flux in the surface layer are accentuated by vegetation contrasts, causing additional sensitivity of CBL and dryline evolution.
