Lei Yang1 Wei Wei Li1 Dongxiao Wang1 and Yongping Li2

Energy2green Wind And Solar Power System

Wind Energy DIY Guide

Get Instant Access

Key Laboratory of Tropical Marine Environmental Dynamics, The South China Sea Institute of Oceanology, Chinese Academy of Science, Guangzhou, 2Shanghai Typhoon Institute, China Meteorological Administration, Shanghai,

China

1. Introduction

Using reanalysis and satellite data sets, numerical simulation and statistical methods are applied for investigating tropical cyclone (TC) of two ocean basins: the South China Sea (SCS) and Bay of Bengal (BOB). Influenced by Asian monsoon, TCs' feature in these two ocean basins differ from the one of other open oceans. In this chapter, a unique TC case passing through SCS as well as TCs track characteristics in BOB are examined. The Fifth Pennsylvania State University and National Center for Atmospheric Research Mesoscale Model (MM5) is utilized to study the precipitation and wind speed during Typhoon Chanchu (2006) in SCS. Five model experiments with different physical parameterizations and sea surface temperature (SST) distributions are carried out in the study. Simulations are evaluated using satellite observations. It is found that the control experiment that is configured with the Blakadar boundary scheme, Resiner2 moisture, the Betts-Miller cumulus scheme and daily updated SST has the most reasonable precipitation. The MRF boundary scheme tends to simulate a dryer boundary layer and stronger vertical mixing, which can greatly reduce the intensity of tropical cyclone (TC), resulting in smaller maximum wind speed but larger range of medium wind speed (25-30 m/s). Constant SST through the TC cycle provides more energy from ocean surface, which could cause a significant increase in TC's intensity and therefore result in the largest overestimation on rainfall and maximum wind speed. Longitudinally-uniform SST distribution before the rapid intensification could reduce TC's intensity and heat fluxes, which can partially compensate for the overestimation of precipitation in the control experiment. Based on a mixture quadratic regression model and the best track dataset 1980-2008, six distinct clusters of TC tracks in the Bay of Bengal (BOB) were identified. For better capturing the background controlling factors, reclassifying is carried out by treating all westward tracks as one type. Thereafter, three track types are obtained, northeastward (fewest in amount), westward (most in amount), and northward. Seasonal variability indicates that northward track type should be divided by two stages (April-May and October-January). After examining the background circulation of each track type, it is found that except for the northeastward one, anticyclonic circulation located in Indo-China Peninsular as well as trough activity in the region of India-Burma played important roles on modulating local wind. These systems assisted steering TC passing through this region. In addition, TCs happened in April and May (pre-summer-monsoon), generally were prone to be affected by the cross-equatorial-flow and move northward.

Results will be given in the following two sections: "South China Sea: Wind and precipitation pattern during typhoon Chanchu (2006) - Comparison between a mesoscale model and remote sensing" and "Bay of Bengal: Cluster Analysis of Tropical Cyclone Tracks".

2. South China Sea: wind and precipitation pattern during Typhoon Chanchu (2006) - Comparison between a mesoscale model and remote sensing

2.1 Introduction

Tropical cyclones (TCs) are one of the most deadly nature hazards to the coastal areas, causing large amounts of casualties and property losses. The South China Sea (SCS) is the largest semi-enclosed marginal sea (~ 3.5x106 km2) in the Northwest Pacific Ocean (WNP), the region of most frequent TC formation in the earth. About 13.2% of TCs in the WNP originates from the SCS (Chen and Ding, 1979). Certain amounts of TCs generated in the WNP also enter the SCS with large inter-annual and decadal variability (Goh and Chan, 2009).

However, compared to the TCs over other ocean basins, there are relatively fewer studies in the literature on TC formation and development over the SCS. The numbers of TCs entering the SCS from the WNP are found below normal during El Nino events, but above normal during La Nina events (Goh and Chan, 2009). It is more complicated for the interannual and decadal variability of the TCs formed in the SCS. The significant atmospheric characteristic in the SCS is that it experiences winter and summer monsoon every year (Liang, 1991). The onset of summer monsoon in the SCS usually starts during mid-May and maintains until September. It is found that during summer (winter) monsoon period, TCs are mainly formed in the northern (southern) part of the SCS (e.g., Wang et al., 2007). The TC activity in the SCS correlates well with the sea surface temperature (SST) and outgoing longwave radiation (OLR) variation (Lee et al., 2006). The formation region of the TCs corresponds with the area where relative vorticity of surface wind (RVSW) is positive, i.e., almost no TCs formed in the negative RVSW area (Lee et al., 2006; Wang et al., 2007). Lee et al. (2006) examined 20 TCs in the SCS during 1972-2002 (May-June), 11 of which are associated with weak baroclinic environment of a mei-yu front and rest of which are more barotopic and possibly intensifies into a stronger TC. The Columnar water vapor, columnar liquid water and the total latent heat release derived from the Special Sensor Microwave/Imager (SSM/I) are found to have significant different in the developing and non-developing tropical disturbances in the SCS (Wang et al., 2008).

The studies of TCs in the SCS, in terms of model simulations and evaluations, are even less. Numerical model simulation can help us to better understand the dynamics of a TC process and therefore help to improve TC forecast skill. Observational data, such as sounding data, Argo floats, satellite data and others have been widely used to improve model simulation (e.g. Soden et al., 2001; Zhao et al., 2005; Chou et al., 2008; Langland et al., 2009; Rakesh et al., 2009). Besides incorporating observations in the simulation to improve forecast skill, evaluation or validation of a model simulation is also possible by comparing the simulations with observational data (e.g. Li et al., 2008; Nolan et al., 2009; Zou et al., 2009). Several simulation studies have been done for the TCs in the SCS. Typhoon Leo (1999) was successfully simulated using two nested domains in relatively coarse resolution (54 km and 18 km) (Lau et al., 2003). The Fifth Pennsylvania State University and National Center for Atmospheric Research Mesoscale Model (MM5) was applied to simulate several characteristics of Typhoon Fitow (2001), including land falling, center position, and precipitation (Li et al., 2004). SST effects on the simulation of the genesis of Typhoon Durian (2001) were investigated using the Weather Research and Forecasting (WRF) model (Wang et al., 2010). MM5 incorporating 4D variational data assimilation system with a full-physics adjoint model was found to greatly improve typhoon forecast in track, intensity, and landfall position (Zhao et al., 2005) .

In this study, we mainly discuss the simulated precipitation and wind pattern during the case of Typhoon Chanchu (2006) using MM5. Previous studies on Typhoon Chanchu (2006) in the SCS mainly focused on analyzing its track and intensity using National Centers for Environmental Prediction (NCEP)/NCAR reanalysis data (Luo et al., 2008). Except for the impact of meteorological factors, ocean effect such as sea surface temperature (SST) on the TC's development has also been studied (Jiang et al., 2008). It is found that considering typhoon-induced cooling in the model simulation can greatly improve the simulation skill of the TC's intensity. Sensitive studies with different SST configurations in Weather Research and Forecasting (WRF) showed that SST variation could cause change in TC's intensity and slightly affect its track as well (Liu et al., 2009).

2.2 Overview of Chanchu (2006)

Typhoon Chanchu (2006) is the first tropical storm in 2006, with a maximum sustained wind speed reaching 46 m/s near its centre during the rapid intensification (RI) on 15 May 2006 in the SCS. It is the most intense typhoon on the Hong Kong Observatory (HKO) record to enter the SCS in May. The TC was originated at the south Philippine Sea on 8 May 2006, when it was declared as a tropical depression. It later strengthened into a typhoon and struck the Philippines twice. It entered the SCS on 13 May 2006 and rapidly intensified into Category 4 (in Saffir-Simpson tropical cyclone scale) on 14 May 2006. During its passage over the SCS, it moved to NW first and turned sharply to NNE while it made its final landfall near Shantou in eastern Guangdong Province, China.

2.3 Model

MM5 is a non-hydrostatic, primitive equation model with a terrain-following sigma-coordinate (Grell et al. 1995). Model simulations of a TC process can vary with many factors, such as model physical parameters, ocean conditions, topography, among others. A total of five experiments are performed to test model sensitivities to some of the above-mentioned factors. All the model runs are initiated at 00 UTC 12 May when the storm was about to enter the SCS. The model version has a horizontal resolution of 15 km and 5 km for Domain 1 and Domain 2, respectively, and 27 vertical levels. Blackdar PBL, Resiner2 moisture and the Betts-Miller cumulus scheme are used in the CTL. Other four experiments are named as PBL_MRF (using MRF PBL), CUM_G (using the Grell cumulus scheme), SST_C (constant SST distribution during TC process, i.e., ignoring the TC induced ocean cooling) and SST_U (longitudinally-uniform SST distribution in the central SCS before the RI). NCEP Final Analysis (FNL) 6-hourly data are used for the initial and boundary conditions. SST is updated daily with real-time, global SST (RTG_SST) developed at the NCEP/Marine

Modeling and Analysis Branch (MMAB) to represent storm-induced SST changes during the life-cycle of a typhoon. The data has a spatial resolution of 0.5°. Although it is more desirable to use a coupled ocean model to represent storm-induced SST changes with a high resolution, our simulations should have smaller SST errors than those from a coupled model due to technique issues. Regular sounding and synoptic observations were used to refine the NCEP FNL analysis data before the simulations start. Wind bogus is inserted for better results of initial vortex (Zou and Xu 2000). Corresponding winds and locations are obtained from the Joint Typhoon Warning Center (JTWC) of the U.S.

2.4 Data

2.4.1 TRMM

Three-hourly merged rain rate from the Tropical Rainfall Measuring Mission (TRMM) satellite's Precipitation Radar (PR) is used in this study to evaluate the simulations. The PR is a three-dimensional space-borne precipitation radar that, at its nadir point, has horizontal and vertical resolutions of 4.3 by 4.3 km and 250 m, respectively. The PR retrieves reflectivities at a frequency of 13.8 GHz from the surface to 20 km above the Earth ellipsoid. It has a 215-km swath width with a minimum detectable signal of nearly 17dBZ. The 3-hourly rainfall data has a horizontal resolution of 0.25° by 0.25°, and extends from January 1998 to June 2006. The 3-hourly rain rate is then used to calculate 24-houlyr accumulated rainfall in this study.

2.4.2 QuikSCAT

The SeaWinds scatterometer on the QuikSCAT is a microwave radar launched and operated by the U.S. National Aeronautics and Space Administration (NASA) from July 1999 to November 2009. The QuikSCAT measures the back scatter radiation in the Ku-band (~13.402 GHz). The return signal power is proportional to surface stress over the ocean, and the signal can be related to surface wind vector at 10 m height assuming neutral stratification. In addition to wind vector data, the times (minutes of UTC) of measurements are also provided. The data are regridded to 0.25° by 0.25° (latitude by longitude) map for ascending and descending passes. In order to make a better comparison, we choose the model hours that are the closest to the time when the satellite passes the SCS.

2.5 Comparisons between model-simulated and TRMM-derived precipitations

The spatial patterns of 24-hour-accumulated precipitation are shown for three days: 14, 15, and 17 May 2006, which represent before, during and after the RI, respectively. Figure 1 shows 24-hour-accumulated precipitation during the three days simulated by the CTL and derived from TRMM measurements. CTL simulates overall larger rainfall than TRMM, especially in the area surrounding the storm center. The difference in main rainfall area is obviously due to the faster translation speed simulated by CTL, i.e., more northward in CTL than in TRMM. In addition, the rainfall patterns simulated by CTL have more compact structure and clearer spiral rainbands. The maximum accumulated rainfall in CTL is clearly larger than the TRMM measurement, with the largest difference during the RI period. Large rainfall area in the model simulation is mainly located at the center before and after the RI (not shown), but at west or northwest of the rainfall area during the RI period (Fig. 2). After the RI, the TC moves closer to the coast; the formation of maximum rainfall should be related to coastal-enhanced mechanism, which seems much stronger in the model

Fig. 1. Comparison of 24-hour-accumulated precipitation derived from CTL (top) and TRMM PR (bottom) during the three days.
Fig. 2. Comparison of 24-hour-accumulated precipitation derived from TRMM PR and the five model simulations on 15 May 2006 (the RI period).
Fig. 3. Differences in simulated 24-hour-accumulated precipitation during the three days: sensitive run minus CTL.

simulation. TRMM gives larger range of the rainfall but weaker strength. The differences of the large rainfall area location between the sensitive runs and CTL can be clearly seen in Fig. 3, in which rainfall simulated by the four sensitive runs are compared with CTL. The differences between the four sensitivity runs and CTL for the three days are illustrated in Fig. 3. Though having a dryer PBL, PBL_MRF tends to simulate a deeper PBL due to excessive vertical mixing outside of the eye wall (Hong and Pan 1996), which could lead to the overall heavy precipitation in PBL_MRF run (Fig. 3). CUM_G allows more grid-resolved precipitation (Grell 1995). SST_C simulates the largest precipitation (especially during the RI, Fig. 3). In comparison to the CTL, PBL_MRF and CUM_G tend to simulate more precipitation outside the eye wall, especially in the area southwest of the center during the RI. SST_C simulates overall higher precipitation. After the RI, large precipitation area moves to the east or southeast of the center in CUM_G and SST_C, while it still shows large precipitation at the west in PBL_MRF. SST_U has the smallest difference with CTL.

2.6 Comparisons between model-simulated and QuikSCAT-derived wind fields

QuickSCAT-derived and model-simulated wind fields during the three days (phases) are given in Figs. 4 and 5, respectively. QuikSCAT passes over the SCS twice a day (descending and ascending passes), with different spatial coverage of the SCS. We choose the pass when QuikSCAT has the best coverage of the TC process for that day and compare with the simulations at the closest hour. Figure 4 clearly shows the evolution of the TC in the SCS. The wind derived from QuikSCAT is stronger at the inner core than that simulated by the model, especially before and during the RI period. Before the RI, the wind patterns from the five experiments are similar, with slight difference in maximum wind speed. During the RI, PBL_MRF simulates relatively smaller maximum wind speed than the other model runs, but has larger range of medium wind speed (20-25 m/s). SST_U has the weakest wind, including maximum and medium wind speeds. After the RI, maximum wind speed increases significantly in SST_C. Compared to PBL_MRF and CUM_G, the three experiments of CTL, SST_C, and SST_U simulate a more compact storm structure during the RI.

Was this article helpful?

0 0
Renewable Energy 101

Renewable Energy 101

Renewable energy is energy that is generated from sunlight, rain, tides, geothermal heat and wind. These sources are naturally and constantly replenished, which is why they are deemed as renewable. The usage of renewable energy sources is very important when considering the sustainability of the existing energy usage of the world. While there is currently an abundance of non-renewable energy sources, such as nuclear fuels, these energy sources are depleting. In addition to being a non-renewable supply, the non-renewable energy sources release emissions into the air, which has an adverse effect on the environment.

Get My Free Ebook


Post a comment