-
Version 5 of the Advanced Regional Prediction System (ARPS, Xue et al., 2000, 2001, 2003), a non-hydrostatic atmospheric model, is used to perform the simulations of the case. The model domain has 651 and 579 horizontal grid points in X and Y directions with a 1-km grid spacing, which is nested in a 3-km X–Y grid of 483 × 403 horizontal grid points (Fig. 3). A generalized terrain-following coordinate is used in vertical, and 53 stretched levels are defined with the grid spacing increasing from about 20 m near the ground to about 800 m near the model top at about 20 km. The model terrain and land surface characteristics on the 3- and 1-km grids are derived from the 30-arc seconds (approximately 1 kilometer) global elevation and land-use data set from the U.S. Geological Survey.
Figure 3. The 3-km model domain nested with a 1-km subdomain (solid box). The small triangles and circles mark the stations from conventional radiosonde and surface networks, respectively. The small squares mark radar locations of Nanchang (NC), Ji’an (JA), and Shangrao (SR), with large circles indicating their maximum radar ranges. The dashed box denotes the horizontal plotting region in later figures. The terrain elevation is shaded in grayscale with mountainous regions denoted as in Fig. 1a.
The model is initiated at 1200 UTC 27 June 2013 and is integrated for 36 hours until 0000 UTC 29 June 2013. For the 3-km grid, the lateral boundary conditions (LBCs) are generated from GFS global 0.5-degree analysis fields at 6-hour intervals obtained from the Historical Unidata Internet Data Distribution (IDD) Gridded Model Data (National Centers for Environmental Prediction, 2003). The initial condition for the 3-km grid is created using the ARPS three-dimensional variational (3DVAR) data analysis scheme (Gao et al., 2004) based on the GFS global 0.5-degree analysis fields and the conventional radiosonde and surface observations shown in Fig. 3. The nested 1-km grid gets its initial condition from the interpolation of the 3-km analysis and its LBCs from the 3-km forecasts at 10-min intervals. The ARPS is used in its full physics mode, as in Wang and Xue (2012).
In this case, the cold outflow is generally confined within the basin surrounded by the Mufu-Jiuling and Huangshan Mountains on the north side, and several low mountains are scattered among the first back-building zone where stage I convective lines formed. It is not clear whether the low mountains play a role in the formation of stage I convective lines. Neither is it clear whether the Mufu-Jiuling and Huangshan Mountains play a role in altering the strength of the cold outflow within the basin and, thus, the back-building processes associated with the cold outflow interacting with the southerly flow along the leading edge of the mei-yu front during the two stages of convective line formation. To evaluate the two points in question, sensitivity experiments with the mountains parallel to the leading edge of stage I convective lines removed (NoMidMts) and with the Mufu-Jiuling and Huangshan Mountains on the north side of the basin removed (NoNorthMts) were designed, respectively (Table 1).
Abbreviation Description CNTL The control experiment using real terrain. NoMidMts The mountains parallelly near the leading edge of stage I convective lines are removed. NoNorthMts The Mufu-Jiuling and Huangshan Mountains on the north side of the basin are removed. Table 1. Experiment design.
Abbreviation | Description |
CNTL | The control experiment using real terrain. |
NoMidMts | The mountains parallelly near the leading edge of stage I convective lines are removed. |
NoNorthMts | The Mufu-Jiuling and Huangshan Mountains on the north side of the basin are removed. |