**1. Introduction**

Tropical cyclones (TCs) are the most devastating weather systems. Along with the direct threat from the strong winds, hazards are often brought by the torrential rainfall that causes flood and landslide. Previous studies on TC rainfall either focused on the structure and dynamics of secondary eyewalls [1] and principal rainbands [2] within the inner core or mesoscale effects such as orographically forced ascent (upslope) and coastal front in predecessor rain events

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2018 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

(PREs) [3]. In addition to that of TCs, suggested rainfall genesis mechanisms also include the convergence of moist outflows from previous cellular convection [4], squall line at the advance of moist cold pools [5], unorganized thermodynamic-generated tropical deep convections [6], etc. Among them, cold pools which are areas of evaporatively cooled downdraft air that spread out beneath precipitating cloud [7] have been frequently investigated because of its prominent role in growth of convection. In TCs, although cold pools are less common due to the relative lack of dry air in comparison to that in midlatitudes, they are known to be the primary mechanism for the sustenance of multicell thunderstorms and convective lines [8]. While TC outer rainband formation has been attributed to a variety of processes, including outward propagating inertia-gravity waves [9] and vortex-Rossby waves [10], it seems plausible that the cold pool played a contributory role in the intensification of TC outer rainband. Numerous TC studies [11–13] have utilized coastal buoys, instrumental towers, and aircraft data to document outer rainband and cold pool structure. When observed, cold pools are often associated with outer rainbands or bands adjacent to inward-spiraling dry air intrusions. These studies also noted decreasing storm-relative inflow, decreasing equivalent potential temperatures, and locally enhanced wind speeds in the boundary layer beneath the rainbands. Moreover, while a necessary condition for cold pool formation may be the presence of midlevel dry air, cold pool intensity appears more related to factors other than the degree of midlevel dryness. While affecting the growth of rainfall, cold pools are sometimes associated with cold fronts and cold air damming (CAD) [14]. Atallah and Bosart [15] examined aspects of the precipitation distribution of hurricane Floyd (1999) through synoptic and modeling analyses and found that precipitation ahead of Floyd's track was generally enhanced along the cold front from approximately 12 h before the time of storm passage. CAD occurs most often on the eastern side of approximately north-south-oriented mountains, as cold air moving toward the eastern slopes has insufficient kinetic energy to go over the barrier and is then forced to decelerate. The genesis of cold pool can be associated with extratropical transition (ET) [16–18], which occurs from a warm-core to cold-core cyclone and gains extratropical cyclone characteristics. Snodgrass et al. [19] found that the heaviest convective rainfall occurred in mesoscale arcs around cloud-free areas, reminiscent of outflow boundaries of cold pools created by earlier convection. However, different with the mechanism of cold pool, Tompkins [6] shows that air moistened by evaporating precipitation can be pushed outward by a drier downdraft driven by precipitation and trigger new convection when the surface air becomes sufficiently buoyant.

between helicity and nonlinear energy transition with a helicity budget of the mean and disturbance flows, indicated that the helical mean flow transfers helicity to the convective eddies and both the buoyant effect and mean-eddy exchange are important sources for disturbance. It is further indicated that the disturbances gain helicity from mean flows and the buoyancy effect amplifies it. Fei and Tan [24] reported that weak helicity is favorable for the energy cascading from large scale to convective scale at the early stage of a convective storm. Hendricks et al. [25] and Montgomery et al. [27] argued that vertical hot towers (VHTs), helical by definition because of coincident updrafts and vertical vorticity, were the preferred mode of convection in TCs. Molinari and Vollaro [28] also identified convective cells, generally favored downshear, would be stronger and longer-lived as a result of larger helicity. Although it is widely accepted that intensive helicity favors the genesis of supercell and TC intensification, it is still an open question of the connection between helicity and the evolution of TC heavy rainfall. In this regard, the mechanism for the rapid growth of TC Fung-Wong's heavy rainfall is examined based on the understanding of Fung-Wong's multiscale processes analyzed with helicity.

Numerical Analysis on the Simulated Heavy Rainfall Event of Tropical Cyclone Fung-Wong

http://dx.doi.org/10.5772/intechopen.72264

279

TC Fung-Wong, as the first strong typhoon made landfall over China mainland in 2008, was characterized by torrential rainfall, large size, and wide-range influence. It was developed at 06 UTC, 25 July 2008, over the Northwest Pacific and then intensified into a typhoon at 09 UTC, 26 July. From the morning on 27 July, it started to move northwest and approached the east coast of Taiwan Island. Fung-Wong intensified into a severe typhoon at 12 UTC, 27 July. As shown on the FY2C satellite imagery, the cloud distribution of Fung-Wong was significantly asymmetric before the occurrence of landfall. More clouds were clustered to the southwest of TC. A clear eye formed before it reached Taiwan. At 22 UTC, 27 July, Fung-Wong landed on Hualian, Taiwan, with the maximal wind speed of 45 m s−1 near its center. As Fung-Wong passed over Taiwan, heaviest rainfall of 818 mm was recorded near Tai-Ping Mountain. Fung-Wong landed again on Donghan, Fujian province, at 14 UTC, 28 July, with the estimated maximal wind speed of 33 m s−1 near its center. At that time, NASA's CloudSat satellite's Cloud Profiling Radar showed that the cloud top of Fung-Wong reached more than 15 km with estimated precipitation rate exceeding 30 mm h−1 on 28 Jul. After the second landfall, it was quickly reduced to a severe tropical storm, and its eyewall and spiral structure were significantly vanished. However, during its stay inland for about 52 h, it produced heavy rainfall in Zhejiang and Fujian province, leading to two rivers flooding, with an esti-

To clarify the mechanism for heavy rainfall, it is helpful to carry out high-resolution simulation with mesoscale nonhydrostatic microphysical models. Recent numerical experiments using the Weather Research & Forecasting (WRF) model have shown some promise in forecasting TCs near landfall [29]. In this study, numerical experiments were designed and performed with the multiply-nested WRF model. With an inner resolution of 600 m, it provides the possibility for answering questions on the role of the energy cascading on the rainfall

The methods and data are described in Section 2. The numerical simulations on track, intensity, and rainfall of Fung-Wong are verified in Section 3. The numerical analysis on the multiscale

mechanism of the heavy rainfall is conducted in Section 4. Section 5 is the summary.

mated total loss of CNY 3.37 billion.

development of TCs.

As the development of TC rainfall becomes quite complicated during landfall due to the interaction with midlatitude atmospheric systems and the topographical effect, these complicated processes were usually investigated with advanced atmospheric numerical models based on a series of dynamic and thermodynamic equations. TC rainfall mechanism can then be further understood with the assistance of dynamic or thermodynamic diagnosis. In this study, the mechanism of heavy rainfall associated with the frontal structure and cold pool during the landfall process of TC Fung-Wong (2008) is preliminarily examined based on the numerical simulations and numerical analysis with the special use of multiscale conception of helicity [20, 21]. The original definition of helicity [22] has been frequently employed to investigate the structure of convections and TCs [23–25]. Wu et al. [26], by investigating the relationship between helicity and nonlinear energy transition with a helicity budget of the mean and disturbance flows, indicated that the helical mean flow transfers helicity to the convective eddies and both the buoyant effect and mean-eddy exchange are important sources for disturbance. It is further indicated that the disturbances gain helicity from mean flows and the buoyancy effect amplifies it. Fei and Tan [24] reported that weak helicity is favorable for the energy cascading from large scale to convective scale at the early stage of a convective storm. Hendricks et al. [25] and Montgomery et al. [27] argued that vertical hot towers (VHTs), helical by definition because of coincident updrafts and vertical vorticity, were the preferred mode of convection in TCs. Molinari and Vollaro [28] also identified convective cells, generally favored downshear, would be stronger and longer-lived as a result of larger helicity. Although it is widely accepted that intensive helicity favors the genesis of supercell and TC intensification, it is still an open question of the connection between helicity and the evolution of TC heavy rainfall. In this regard, the mechanism for the rapid growth of TC Fung-Wong's heavy rainfall is examined based on the understanding of Fung-Wong's multiscale processes analyzed with helicity.

(PREs) [3]. In addition to that of TCs, suggested rainfall genesis mechanisms also include the convergence of moist outflows from previous cellular convection [4], squall line at the advance of moist cold pools [5], unorganized thermodynamic-generated tropical deep convections [6], etc. Among them, cold pools which are areas of evaporatively cooled downdraft air that spread out beneath precipitating cloud [7] have been frequently investigated because of its prominent role in growth of convection. In TCs, although cold pools are less common due to the relative lack of dry air in comparison to that in midlatitudes, they are known to be the primary mechanism for the sustenance of multicell thunderstorms and convective lines [8]. While TC outer rainband formation has been attributed to a variety of processes, including outward propagating inertia-gravity waves [9] and vortex-Rossby waves [10], it seems plausible that the cold pool played a contributory role in the intensification of TC outer rainband. Numerous TC studies [11–13] have utilized coastal buoys, instrumental towers, and aircraft data to document outer rainband and cold pool structure. When observed, cold pools are often associated with outer rainbands or bands adjacent to inward-spiraling dry air intrusions. These studies also noted decreasing storm-relative inflow, decreasing equivalent potential temperatures, and locally enhanced wind speeds in the boundary layer beneath the rainbands. Moreover, while a necessary condition for cold pool formation may be the presence of midlevel dry air, cold pool intensity appears more related to factors other than the degree of midlevel dryness. While affecting the growth of rainfall, cold pools are sometimes associated with cold fronts and cold air damming (CAD) [14]. Atallah and Bosart [15] examined aspects of the precipitation distribution of hurricane Floyd (1999) through synoptic and modeling analyses and found that precipitation ahead of Floyd's track was generally enhanced along the cold front from approximately 12 h before the time of storm passage. CAD occurs most often on the eastern side of approximately north-south-oriented mountains, as cold air moving toward the eastern slopes has insufficient kinetic energy to go over the barrier and is then forced to decelerate. The genesis of cold pool can be associated with extratropical transition (ET) [16–18], which occurs from a warm-core to cold-core cyclone and gains extratropical cyclone characteristics. Snodgrass et al. [19] found that the heaviest convective rainfall occurred in mesoscale arcs around cloud-free areas, reminiscent of outflow boundaries of cold pools created by earlier convection. However, different with the mechanism of cold pool, Tompkins [6] shows that air moistened by evaporating precipitation can be pushed outward by a drier downdraft driven by precipitation and trigger new convection when the surface air

278 Finite Element Method - Simulation, Numerical Analysis and Solution Techniques

As the development of TC rainfall becomes quite complicated during landfall due to the interaction with midlatitude atmospheric systems and the topographical effect, these complicated processes were usually investigated with advanced atmospheric numerical models based on a series of dynamic and thermodynamic equations. TC rainfall mechanism can then be further understood with the assistance of dynamic or thermodynamic diagnosis. In this study, the mechanism of heavy rainfall associated with the frontal structure and cold pool during the landfall process of TC Fung-Wong (2008) is preliminarily examined based on the numerical simulations and numerical analysis with the special use of multiscale conception of helicity [20, 21]. The original definition of helicity [22] has been frequently employed to investigate the structure of convections and TCs [23–25]. Wu et al. [26], by investigating the relationship

becomes sufficiently buoyant.

TC Fung-Wong, as the first strong typhoon made landfall over China mainland in 2008, was characterized by torrential rainfall, large size, and wide-range influence. It was developed at 06 UTC, 25 July 2008, over the Northwest Pacific and then intensified into a typhoon at 09 UTC, 26 July. From the morning on 27 July, it started to move northwest and approached the east coast of Taiwan Island. Fung-Wong intensified into a severe typhoon at 12 UTC, 27 July. As shown on the FY2C satellite imagery, the cloud distribution of Fung-Wong was significantly asymmetric before the occurrence of landfall. More clouds were clustered to the southwest of TC. A clear eye formed before it reached Taiwan. At 22 UTC, 27 July, Fung-Wong landed on Hualian, Taiwan, with the maximal wind speed of 45 m s−1 near its center. As Fung-Wong passed over Taiwan, heaviest rainfall of 818 mm was recorded near Tai-Ping Mountain. Fung-Wong landed again on Donghan, Fujian province, at 14 UTC, 28 July, with the estimated maximal wind speed of 33 m s−1 near its center. At that time, NASA's CloudSat satellite's Cloud Profiling Radar showed that the cloud top of Fung-Wong reached more than 15 km with estimated precipitation rate exceeding 30 mm h−1 on 28 Jul. After the second landfall, it was quickly reduced to a severe tropical storm, and its eyewall and spiral structure were significantly vanished. However, during its stay inland for about 52 h, it produced heavy rainfall in Zhejiang and Fujian province, leading to two rivers flooding, with an estimated total loss of CNY 3.37 billion.

To clarify the mechanism for heavy rainfall, it is helpful to carry out high-resolution simulation with mesoscale nonhydrostatic microphysical models. Recent numerical experiments using the Weather Research & Forecasting (WRF) model have shown some promise in forecasting TCs near landfall [29]. In this study, numerical experiments were designed and performed with the multiply-nested WRF model. With an inner resolution of 600 m, it provides the possibility for answering questions on the role of the energy cascading on the rainfall development of TCs.

The methods and data are described in Section 2. The numerical simulations on track, intensity, and rainfall of Fung-Wong are verified in Section 3. The numerical analysis on the multiscale mechanism of the heavy rainfall is conducted in Section 4. Section 5 is the summary.
