A numerical study of back-building process in a quasistationary rainband with extreme rainfall over northern Taiwan during 11–12 June 2012

Chung-Chieh Wang,Bing-Kui Chiou,Ching-Hwang Liu,Hung-Chi Kuo,George Tai-Jen Chen

doi:10.5194/acp-16-12359-2016

Abstract

Abstract. During 11–12 June 2012, quasistationary linear mesoscale convective systems (MCSs) developed near northern Taiwan and produced extreme rainfall up to 510 mm and severe flooding in Taipei. In the midst of background forcing of low-level convergence, the back-building (BB) process in these MCSs contributed to the extreme rainfall and thus is investigated using a cloud-resolving model in the case study here. Specifically, as the cold pool mechanism is not responsible for the triggering of new BB cells in this subtropical event during the meiyu season, we seek answers to the question why the location about 15–30 km upstream from the old cell is still often more favorable for new cell initiation than other places in the MCS. With a horizontal grid size of 1.5 km, the linear MCS and the BB process in this case are successfully reproduced, and the latter is found to be influenced more by the thermodynamic and less by dynamic effects based on a detailed analysis of convective-scale pressure perturbations. During initiation in a background with convective instability and near-surface convergence, new cells are associated with positive (negative) buoyancy below (above) due to latent heating (adiabatic cooling), which represents a gradual destabilization. At the beginning, the new development is close to the old convection, which provides stronger warming below and additional cooling at mid-levels from evaporation of condensates in the downdraft at the rear flank, thus yielding a more rapid destabilization. This enhanced upward decrease in buoyancy at low levels eventually creates an upward perturbation pressure gradient force to drive further development along with the positive buoyancy itself. After the new cell has gained sufficient strength, the old cell's rear-flank downdraft also acts to separate the new cell to about 20 km upstream. Therefore, the advantages of the location in the BB process can be explained even without the lifting at the leading edge of the cold outflow.

Highlights

During 11–12 June 2012, both TL/AS and BB mesoscale convective systems (MCSs) developed in succession near northern Taiwan and produced extreme rainfall up to 510 mm overnight, in Taipei City and the surrounding metropolitan area
During 11–12 June 2012 in the meiyu season, both TL/AS and BB MCSs developed in succession near northern Taiwan and together produced extreme rainfall up to 510 mm overnight in the Taipei metropolitan area, causing serious flooding in many densely populated regions
Observations show that BB behavior occurred in these MCSs, especially in the second, E–W-aligned quasistationary linear MCS during 18:00–24:00 UTC, June (02:00–08:00 LST, June), and was a contributing factor to the extreme rainfall and related hazards in Taipei

Summary

Introduction

As a common type of mesoscale convective systems (MCSs) with a lifespan around 3–12 h, organized rainbands such as squall lines are capable of producing persistent precipitation at high intensity, compared to ordinary, isolated, or scattered convection (e.g., Carbone, 1982; Bluestein and Jain, 1985; Rotunno et al, 1988; Browning, 1990; Houze et al, 1990; Chen and Chou, 1993; LeMone et al, 1998; Parker and Johnson, 2000; Doswell, 2001; Johnson and Mapes, 2001; Sun and Lee, 2002; Weisman and Rotunno, 2004; Meng et al, 2013) When such rainbands are slow moving and the embedded deep convective cells travel at small angles almost parallel to the line, multiple cells can pass through the same locations in succession to rapidly increase rainfall accumulation and the potential for flash floods (e.g., Maddox et al, 1979; Doswell et al, 1996; Brooks and Stensrud, 2000; Parker and Johnson, 2004). While some MCSs may possess characteristics of both types (Schumacher et al, 2011; Peters and Schumacher, 2015), the BB systems are typically smaller and more localized and more difficult to predict (e.g., Schumacher and Johnson, 2005)

Results

Discussion

Conclusion