Tracking Typhoons

24-08-2012Source:China Meteorological News Press

 As stronger and more typhoons come one after another during the peak time of typhoon season, China Meteorological Administration is taking cutting-edge technologies to track and forecast typhoon, with its advanced observing tools, telecommunication system and prediction models.

Observation System 


CMA's operational meteorological satellites.  

A new polar-orbiting satellite (FY-3B) has been put into operation since July 2011. FY-3A and FY-3B are operating in a network that covers both morning and afternoon orbits. They have improved China's meteorological observing and weather forecasting capabilities. The FY-3 series play an important role in monitoring natural disasters and environment, addressing climate change and disaster prevention and reduction.       

A new geostationary satellite FY-2F has been put into trial operation since April this year and was formally delivered for in-orbit operation on August 20. Since June this year, the FY-2F has been switched into intensified observations for 8 times with a frequency of one observation in minimum 5 minutes besides of the regular observation of geostationary FY-2D and FY-2E satellites. Through these higher temporal resolution satellite data, the occurrences and evolution characteristic of TCs can be captured. It played important role in TC monitoring and forecasting.

As of marine observing system, by the end of 2011, the surface observation network in marine composed of 191 shore-based stations, 99 AWSs on islands, 18 buoys, 1 storm surge station, 6 oil drilling platform-based stations and 5 ship stations, which greatly enhance the ability of coastal ocean observing ability.       

In 2011, 5 CINRAD radars were set up in Jiangsu, Fujian, Anhui, Hunan and Henan provinces. They have further enhanced capability for TC monitoring along the China’s southeast coast among others. They have contributed to severe weather prevention and reduction in many provinces. Six boundary wind profilers were set up in Anhui, Fujian, Jiangxi, Shandong and Zhejiang provinces.

Telecommunication System

In last year, CMA completed its new-generation data-broadcasting system-CMACast. The CMACast system is a satellite-based data broadcasting system that complies with DVB-S2 standard, substituting the present satellite data-broadcasting systems (PCVSAT, DVB-S & FENGYUNCast), which have 2400, 700 and 200 users respectively. The CMACast is a major component of the national meteorological data dissemination system, continuously broadcasting the real-time observational data and products crucial for the weather forecasts and related services for more than 2500 users. It is also the most effective way to share the various meteorological data and products with user communities in China. At present, the daily data volume is more than 150 GB, including national and international observations, CMA T639 NWP products, satellite data and products from FY2D/E, FY3A/B, EUMETSAT, etc.   

CMACast is also a component of WMO IGDDS and GEONETCast systems. It substitutes the Regional GEONETCast Network Centre (GNC) of the FENGYUNCast, which was formerly integrated into GEONETCast in 2007 as a contribution of China to GEONETCast. Compared to the FENGYUNCast, CMACast has a higher bandwidth, many new data contents, and improved user and data management. CMACast provides a full range of services within GEONETCast framework, especially for the users in the Asia-Pacific Region.

The Sixteenth World Meteorological Congress (Cg-16) designated Beijing as one of the Global Information System Centre (GISC Beijing). Following the trial operation over the year, the GISC Beijing started its operation as from 15 August 2011. Apart from continued provision of existing GTS data transmission and exchange services, GISC Beijing provides the Data Discovery Access and Retrieve (DAR) service, Data Subscription service and CMA's NWP products and FY satellite products. CMACast Users can also receive meteorological data and products from GISC Beijing.

 The WIS and CMAcast system

Rapid Radar Data Assimilation and Analysis System for Landing Typhoons

A rapid radar data assimilation and analysis system for landing typhoons was established in National Meteorological Center (NMC), CMA in 2011. The system is targeted to landing typhoons based on 3D-VAR data assimilation (ARPS3DVAR) and composite cloud analysis technique developed by Oklahoma University, incorporating inversed Doppler radar data and other data (conventional data & model outputs). Figure 2 gives the 10m wind fields of severe typhoon Nesat (1117) from the radar data assimilation and analysis. The product provides objective analyses for operational TC intensity and high wind forecasts.      


GRAPES Regional Typhoon Prediction Model (GRAPES-TYM)

GRAPES Regional Typhoon Prediction Model (GRAPES-TYM) is a mesoscale typhoon prediction model developed by NMC/CMA based on CMA GRAPES-MESO model. GRAPES-TYM has been put into operation since typhoon season 2011 as a main component of CMA typhoon forecast experiment. This model runs twice (00, 12 UTC) per day. Its 24, 48 and 72 h track forecast errors are 98, 181 and 311km respectively in 2011. Its skill is equivalent to that of the NMC/CMA T213 model.

GRAPES-TYM has improved TC vortex initialization scheme, physical process and dynamic framework. The vortex initialization scheme mainly uses the artificial vortex technique based on the nonlinear equilibrium model and combines with GFDL TC separation technique, its work flow is shown in Fig 2. The main improvements in physical process and dynamic framework include improved surface heat flux calculation by correcting surface roughness parameter under strong wind conditions, improved model TC intensity prediction by introducing heat expansion from continuous equation, and adjustment of physical values in the model by introducing surface pressure from TC vortex initialization scheme.

GRAPES-TYM domain (color shadow shows terrain heights)
GRAPES-TYM work flow in TC vortex initialization scheme

Objective TC Positioning and Intensity Estimation System Based on Geostationary Meteorological Satellites

CMA has established an objective TC positioning and intensifying system based on satellite TC data over Northwest Pacific and the South China Sea. The objective TC estimation system was established by introducing the advanced objective Dvorak technique (AODT) developed by Cooperative Institute for Satellite Meteorological Studies (CIMSS) of the Wisconsin University. It can process FY2C/D/E and MTSAT infrared channel 1 data. The objective TC positioning system was established by using mathematical morphology, image processing and intelligent information processing technique. It now can process images from MTSAT satellite infrared channel 1 from Kochi University, Japan.  

The interface of objective TC intensity estimation system

The interface of objective TC positioning system


Developed by Guangzhou Institute on Tropical Marine Meteorology, the South China Sea Typhoon Model (SCS-TM) is operated by CMA in a stable manner. To address TC track and intensity forecast issues, based on its operational system, improvements were made in the moisture advection scheme, upstream physical processes, model boundary layer and dynamic framework, and a regional assimilation model was set up for the tropical South China Sea, which is a useful tool for the Southeast Asia region. A model for typhoon-induced sea wave was developed with SCS-TM as an atmospheric component. Incorporating TC conditions, the model was successful in forecasting sea waves in the Leizhou Peninsula and Qiongzhou Straits, which agreed well with the observations.    

24-h forecasts 0800 BT 3 October 2011 for wind and effective wave height (a, contours and shades represent effective wave height; and (b, shade shows water rise zone due to TC Nalgae).

Upgraded Typhoon Information Processing System (TIPS)

As an operational forecasting platform, MICAPS typhoon version provides visualized observations, interactive TC forecasts and graphics of TC track, intensity, wind and rainfall. The improved platform frees forecasters from repetitious hard work and makes weather analysis and forecasting more effective. NMC/CMA set up a real-time statistical forecast error analysis system, covering not only TC track and intensity forecast errors but also TC motion speed and direction errors. This system can be used to evaluate real-time forecast and lead to reasonable refinement of forecasting techniques. To retrieve historical TC tracks and synoptic backgrounds, a new-version system has been built up based on CMA best track database. After a similar key area is located, past TC tracks together with a series of synoptic backgrounds can be provided. Another useful function of the system is to generate a sequence under given criterion.

GRAPES-based Regional TC Ocean-Atmosphere-Wave Coupled Model

A regional ocean-atmosphere-wave coupled model has been developed for TC prediction over the Northwest Pacific. In this model, a multi-model coupler is introduced to accommodate the flux transfer and interaction among atmospheric, ocean and wave models. It has 4 major components:  a regional typhoon model (based on GRAPES-TCM), an ocean module (ECOM), a wave module (WAVEWATCH III) and a multi-model coupler (OASIS). To reasonably describe the ocean-atmosphere-wave interactions, the physical processes in the coupled model have been investigated through numerical simulations of a number of typhoon cases. The results show this model runs stably and reproduces observed air-sea-wave processes, which evolve with the changing SST, sea surface flux (e.g., latent heat flux, sensible heat flux), which have impacts on TC intensity variation. Comparisons with model outputs without coupling (GRAPES-TCM) show better performance of the coupled model in simulating TC intensity, which is consistent with observations. The planned improvement of this coupled model will provide key support for TC predictions in China.


The drag coefficient of boundary layer in the trial GRAPES model with higher resolution in Shanghai Typhoon Institute of CMA has been adjusted, and BOGUS data is assimilated with MC-3DVAR. The new typhoon initial state is more consistent, while calculating time is reduced compared with 4DVAR. For typhoon “Muifa”, the 24-h mean track forecast error was reduced from 80 to 73 km (19 samples), and the 48-h mean track forecast error was reduced from 183 km to 147 km (16 samples).   
Editor Zhang Yong