In the meso-scale observing network, there will be 16 fixed (red X in Fig. 6; 14 boundary layer radars, 2 troposphere radars) and 2 portable wind-profiling radars (green X in Fig. 6). Wind profiler is a five-wave beam radio remote sensing instrument. Based on Bragg scattering theory, each antenna can measure radial velocity through several signal and data processing such as coherent accumulate, FFT transform, spectrum average, etc. Final products such as profile of horizontal wind, vertical velocity and atmospheric structure constant of refractive index can then be obtained.
According to the different working frequency and detecting range, wind profilers are classified into three types: stratosphere wind profiler (46–68MHz), troposphere wind profiler (440–450MHz) and boundary layer wind profiler (1270–1295MHz, 1300–1375MHz). The two latter wind profilers are widely used in China at present.
Based on wind profiler standards of functional specification and data format established by China Meteorology Administration, wind profiler observation data include: real time wind profile data, half-hour averaged wind profile data and one-hour averaged wind profile data. All of these data contain radial velocity, spectrum width, SNR, horizontal wind, vertical velocity and power spectrum density.
2) Radiosonding Stations
Twenty-two upper air sounding stations (blue dots in Fig. 5), built by China Meteorological Administration (CMA) are put in place in the experiment area to conduct operational observation twice per day. Fourteen of them (4 in Guangdong, 3 in Hainan, 6 in Guangxi, and 1 in Hong Kong) will release radiosondes four times a day during intensive observation. High-resolution sounding data from these sites, not just mandatory-level data, are necessary for the study of individual events as well as obtaining accurate initial conditions for modeling.
3) GPS/MET Stations
There are 85 ground-based GPS water vapor observation stations in the large-scale observing network, among which 35 are located in the meso-scale region (black triangles in Figs. 5 and 6). They are designed to measure precipitable water in the atmospheric column at one hour interval. The GPS/MET observed precipitable water will be used to quality control the humidity measurement from the soundings.
4) Air-Borne Observations
Hong Kong Observatory (HKO), in collaboration with the Government Flying Service (GFS) of HKSARG, has conducted several experimental reconnaissance flights for tropical cyclones (TCs) over northern SCS in 2011 and 2012. In-situ measurements of winds, temperature, humidity and pressure have been collected using a data probe installed on a GFS fixed-wing aircraft.
In the latter part of 2014, new aircrafts with dropsonde measurement system and in-situ data probe will be available. Subject to the air traffic conditions and availability of the GFS team in addition to search and rescue (SAR) operation when necessary, air-borne observations could be collected over the HKFIR (Hong Kong Flight Information Region, see Figure xx) for the experiment. Detailed flying route will be decided with GFS before the flight(s) and the dropsondes could be launched at altitudes up to 12 km. The flight measurement data will include positions (latitude, longitude and height), pressure, wind, temperature, and relative humidity.
(2) Surface Observation
Land-based (in situ) observation network consists of national-level automatic weather stations (AWSs) with observers on duty and regional AWSs. There are 366 national-level stations (small black dots in Fig. 5) within the large-scale area; 113 national-level stations and 1,503 regional AWSs are located in the meso-scale region with an average distance of about 10 km. They will provide temperature, sea level pressure, humidity, wind direction/speed, and precipitation observations every hour.
(3) Precipitation System Observation
1) Doppler Weather Radar
There are 22 Doppler weather radars in the large-scale observing network (red triangles in Fig. 5). The 11 radars in the mesoscale region are S band, of which 9 are SA radars, named CINRAD-SA in China and Internationally known as WSR-98D, produced by Beijing METSTAR Radar Co,. Ltd. The specifications of the SINRAD-SA (Table 2) are similar to those of the WSR-88D of US (http://www.roc.noaa.gov/). The volume scan pattern of the radars is VCP21, with 9 elevation angles from 0.5o to 19.5o. The scan period is 6 minutes.
2) Dual Polarimetric Radar
Two C- and one X-band portable dual polarimetric radars (the main technical parameters are listed in Table 3, and the performance parameters in Table 4, respectively) will participate in the field experiment. Dual polarimetric radar has the significant advantage over ordinary Doppler weather radar on quantitative precipitation estimation, drop size distribution inversion, hydrometeor phase identification, etc. (Hu et al., 2010, 2012; Liu et al., 2010；Chu et al. 1997). The reason is that the amplitude, phase, frequency, and polarization state of electromagnetic wave will be changed during scattering and propagation process. In contrast with ordinary Doppler radar, polarization radar has the capability of detecting the echoes in different polarimetric directions, which can alternately or simultaneously transmit and receive the echo signals in horizontal and vertical directions. It can also measure polarimetric parameters, such as the difference of reflectivity between the two directions (differential reflectivity, ), cross- polarization reflectivity (de-polar reflectivity, ), the difference of phase (differential propagation phase shift, ), specific differential propagation phase shift ( ), and correlation coefficient ( ) between the two directions, as well as the echo intensity in horizontal direction (ZH), radial velocity (Vr), and spectrum width (SW) (Seliga and Bringi 1976; Matrosov et al. 2002; Hu et al. 2008, 2010, 2012). From these measurements, microphysical properties of the precipitating systems such as hydrometeor identification could be derived. However, attenuation will be a problem in the heavy rain systems for the C-band and X-band radars, so getting hydrometeor identification will be restricted in a number of instances.
As examples, Fig. 9 shows the plane position indicator (PPI) pictures of ZH (dBZ), ZDR (dB), KDP (0 km-1), and ρHV detected with the C-band dual polarimetric radar, which was developed by State Key Laboratory of Severe Weather (LaSW) in 2008, in the 4.500 elevation at 0943 BST June 6, 2008 in Boluo, Guangdong province. Fig. 10 shows the inversion products, which include the phase identification, raindrop mean diameter (D0, unit: mm), numerical density (N0, unit: 105 mm-1 m-3), and liquid water content (Lw, unit: g m-3). (Fig. 9 and 10 are attached at the end).
3) Micro Rain Radar (MRR)
One MRR made by METEK (http://www.metek.de/) will be used in the experiment, and its main technique variables are described in Table 5. The MRR is a compact 24.1 GHz FM-CW radar for the measurement of profiles of drop size distributions and, derived from this, rain rates, liquid water content and characteristic falling velocity resolved into 30 range gates. MRR can detect very small amounts of precipitation, which isbelow the detecting threshold of conventional rain gauges. Due to the large scattering volume (compared to in situ sensors), statistically stable drop size distributions can be derived within a few seconds. The droplet number concentration in each drop-diameter bin is derived from the backscatter intensity in each corresponding frequency bin, with assumption of the relation between terminal falling velocity and drop size.
4) Raindrop Disdrometer
Four raindrop disdrometers made by OTT ( http://www.ott.com/web/ott_uk.nsf/id/pa_parsivel2_e.html ) will be used in the experiment. Based on operational principle of photoelectric effects, raindrop special density, fall speed and raindrop size can be measured when raindrops fall through the rectangle sample connection (180×30 mm). Raindrop diameter and fall speed are classified into 32 grades. The measuring precision of raindrop size and fall speed and their corresponding grades are: 0.125 mm and 0.1m/s in 1-10 grades, 0.25 mm and 0.2m/s in 11-15 grades, 0.5 mm and 0.4m/s in 16-20 grades, 1mm and 0.8m/s in 21-25 grades, 2mm and 1.6m/s in 26-30 grades, 3mm and3.2m/s in 31-32 grades, respectively. From the raindrop diameter and fall speed, a variety of properties can be deduced: raindrop size distribution, rain intensity, visibility, energy of raindrop motion and precipitation type (such as drizzle, rain, sleet, hail, snow, fog).
5) Millimeter-Wave Radar
The Ka band cloud radar from State Key laboratory of Severe weather (LaSW), Chinese Academy of Meteorological Sciences, and Institute of Heavy Rain, Wuhan (WHIHR) (Table 7) will be used to observe the internal structures of nonprecipitating and weakly precipitating clouds with excellent sensitivity, spatial and temporal resolutions, and accuracy. The cloud and precipitation parameters, such as cloud and precipitation drop size distribution (DSD), air vertical speed and turbulent fluctuation standard deviation (σ) could be retrieved from reflectivity, velocity , spectral width and Doppler spectral density data that are directly measured by the cloud radars (Gossard, 1994; Gossard et al., 1997). The cloud radar of LaSW normally operates in vertically pointing mode while the cloud radar of WHIHR can operate in scan mode. The minimum observation ranges are 500m and 150m for the LaSW and WHIHR radars, respectively.
6) Lightning Location System (LLS)
Meteorological bureaus in Guangdong province, Hongkong, and Macao started to jointly construct the Guangdong-Hong Kong-Macao LLS in 2005. Five sub-stations were built in 2005, with the 6th sub-station added in Sep 2007, and another 11 sub-stations in 2012. As a whole, 7 IMPACT sensors and 10 LS-7000 sensors are employed in the Guangdong-Hong Kong-Macao LLS (Fig. 11). The combined MDF/TOA technology is also used to detect CG lightning stroke information such as longitude and latitude, GPS time, peak current, polarity and reporting sensors, etc. The comparison of triggered lightning and natural lightning observation shows that the detection efficiency and location precision had been obviously improved since the increase of detection sub-stations in 2012. The flash detection efficiency and stroke detection efficiency were about 97% and 91%, respectively; the arithmetic mean location error was about 600 m. However, the peak currents of return strokes were systematically overestimated.
(4) Satellite Observation
In normal mode, the FY-2D and FY-2E geostationary satellites of China each take 28 images in a daily basis. With 20 additional images taken by each satellite when operated in the dual-satellite intensive observation mode, it will provide geostationary satellite data over target area once every 15 minutes. Such data can be obtained once every 6 minutes by FY-2F geostationary satellite during the intensive observation period. The FY-3A and FY-3B polar orbiting satellites will work in the normal operational mode, and data that cover the target area will be available once in the morning and once in the afternoon. More details of satellite products based on the FY satellites are listed in Table 8.
Other satellite datasets, such as those from COSMIC (http://www.cosmic.ucar.edu/) during the field experiments will also be collected.
Chinese Academy of
46 Zhonguancun South Avenue,Beijing,100081,P.R.China