The UK weather radar network current and future capabilities including the upgrade to dual polarisation.

Transcription

1 The UK weather radar network current and future capabilities including the upgrade to dual polarisation. Dr Jacqueline Sugier, Radar R&D, Observations, Met Office RMetS National Meeting, 20 th March 2013

2 The Importance of Weather Radar Very important: Underpin many Met Office services Used for weather and flood forecasting Provide vital information for emergency response teams It is our only observing system with the spatial and temporal resolution to resolve convective cells. Historically, the UK weather radar network has provided detail information regarding the intensity, the location, distribution and evolution of precipitation.

3 Early days of the UK Radar Network The network declared operational in It was comprised of 5 radars covering England and Wales (2 S- band and 3 C-band single polarisation non-doppler radars). The network was delivering in realtime: 2km resolution single site rainrate data every 5 mins to remote users such as the regional water boards. And 5km resolution composite rainfall map of England and Wales every 15 mins.

4 UK Radar Network Status February 2013 UK radar network now comprised of 15 radars: All the network but 1 are Doppler radars 2 dual polarisation Doppler radars (yellow dots). All C-band With increased telecomms and computing capabilities, raw data is now sent directly to Exeter for central processing (limited processing at site).

5 Outline Radar Network upgrade to dual polarisation UK Weather Radar Network products and capabilities Continuous development

6 Outline Radar Network upgrade to dual polarisation UK Weather Radar Network products and capabilities Continuous development

7 Network Renewal: Top-level Aims Continuity - avoid running into age related failures. (Hameldon Hill radar is 33 years old and has completed ~14M scans!) Use latest technology to increase radar capability ( e.g. dual-polarisation to improve quality control and rainfall rate accuracy).

8 Weather Radar Network Renewal Design Radar systems and software have been developed in partnership with academia and industry to create a bespoke and fully customisable radar system and network

9 OSA radar design overview Central processor Radarnet Met Office HQ Exeter Waveguide switch: Split power H & V or Full power H polarisation Radar cabin C-band Transmitter Aerial mounted Receiver & Digitiser PC for Radar control, GUI & Product Generation

10 Analogue receiver Dual Channel Analogue receiver design - Based on current UK network single channel standard (Met Office /Pascall. UK) Performs downconversion from C-band to intermediate frequency (IF) 30MHz

11 Digitiser VHDL digital down converters are used to convert I & Q to baseband prior transferring them to the Cyclops PC for further signal processing and product generation. Met Office design based on Pentek (US) card: 4 channels A/D 200 MSPS 16 bits analogue to digital converters, controlled by Virtex-6 FPGA 30 channels of digital IO

12 Cyclops-D4 PC system Cyclops-D4 provides: Radar control Signal processing/product generation: PC based C++ software. Produces Z, V, SQI, CI, CPA, N, Z DR, ρ HV, Φ DP, and LDR, at a resolution of 1 deg per 300 / 75 m for long / short pulse respectively. System monitoring FTP/ remote desktop connection

13 Renewal on the ground Replacement of the receiver, antenna, radome and waveguide system, and refurbishment of the pedestal.

14 Re-engineering of the pedestal New motors, gearboxes, servos, supplied by Danaher motion (Germany) Brushless AC Motor with epicylic gearbox Control via CanOpen interface Hengstler 14 bit Synchronous Serial Interface encoder for absolute position information

15 Re-engineering of the pedestal Old motor (~ 25 kg) and gear head (~ 35 kg) New motor (11 kg) and gear head (13 kg)

16 Dual polarisation antenna Reuse the Precision (UK) 3.7m dish New feedhorn design for dual polarisation (Q-Par, UK) Cross polar isolation design goal of - 40dB, Gain 43dB minimum, Sidelobe design goal of -30dB within +/- 10 deg and -35deg outside +/- 10 deg. 1 degree beam width

17 Weather Radar Network Renewal Implementation Doppler capability rolled-out to all but 1 UK radar. Dual polarisation design is running at 2 Ops sites (Chenies NW London; and Thurnham Kent) and 1 R&D site (Wardon Hill Dorset). Upgrade to a 3rd and 4th operational sites are underway (Castor Bay- NI; and Predannack - Cornwall). 5 installations planned for FY 13/14. Completion late FY 15/16

18 Outline Radar Network upgrade to dual polarisation UK Weather Radar Network products and capabilities Continuous development

19 Operational radar products During the 5 mins cycles, the radar perform 10 scans at different elevation. Data is averaged to 1 x 600m polar resolution. Rainfall products derived from radar reflectivity Doppler products derived from the phase shift caused by moving reflecting targets Rainfall products generated every 5 mins at 5km (Op), 1km (Op), and 500m (R&D) resolution. Radial winds are been assimilated into 1.5km NWP model (UKV).

20 Rainfall products Radarnet processing Flagging of Raw Data Single Site Rain Rate Compositing

21 R&D product: Radar refractivity Fabry (1997) demonstrated that refractivity changes can be derived by monitoring change over a time period of the phase of the electromagnetic return signal from stationary targets. Radar refractivity can provide useful information to NWP models regarding the evolution of the humidity field 30-60km radius around each radar sites, which in turn can provide observation of convergence before a storm forms. Station obs agree with radar, Nicol et al 2012b).

22 Differential reflectivity, ZDR provide mean particle shape The bigger the drop the more oblate it becomes. The shape of the raindrop affects how it reflects the vertically and horizontally polarised radar beam. ZDR is the ratio between the received power measured in the horizontal (H) and vertical (V) plan: Raindrop size against ZDR Small drops (1mm) ZDR ~ 0dB ~ 3 mm ZDR ~ 1.5dB Large raindrop > 6mm ZDR ~ 6 db; ZDR = 10* log10( ZH / ZV)

23 Example of Differential Phase, φ DP Chenies 17 Mar 2013 Reflectivity, dbz Differential Phase Shift, degree As rain becomes heavier raindrop become oblate. The horizontally polarised wave will be more affected by more water than the vertically polarized wave. φ DP = φ H - φ V φ DP Indicates the relative delay between the Horizontal and Vertical wave Increase in differential phase related to attenuation of the radar signal.

24 Example of ρ HV data Wardon Hill - 5 Feb 2013 Reflectivity (Z), dbz Co-polar correlation coefficient (ρ HV ) Correlation between the H and V backscattered field. Decorrelation occurs if both orthogonal backscattered field do not vary simultaneously. Good indicates of spurious echoes, bright-band. Good indicator of the quality of the radar system.

25 Dual polarisation quality: ρ HV assessment During light uniform rain events, the correlation between the H and V channels should be close to 1. Small imperfection in the cross channel isolation result in the degradation of the quality dual polarisation parameters, with their precision decreasing as peak ρ HV falls away from 1. Radar Systems EEC Thurnham (C-band, 4.3m antenna) Peak ρ HV 0.97 Vaisala (C-band, 4.3m dish) Colorado State University, CSU-CHILL (S-band, 8.5m antenna) Met Office Wardon Hill (Cband, 3.7m antenna)

26 Example of LDR data Wardon Hill 5 Feb 2013 Reflectivity (Z), dbz Linear depolarisation ratio (LDR), db Measure of the depolarisation of the H wave LDR is sensitive to the particle shape, dielectric constant and orientation of the particle major axis with respect to the plane of the radar polarisation. LDR = 10* log10( Z HV / Z HH )

27 Dual polarisation quality: LDR assessment LDR used to assess the quality of the dual polarisation measurements Lower LDR in rain means better isolation between the H and V channels i.e. better dual polarisation performance Met Office Frequency of occurence, % Linear depolarisation ratio, db Vaisala Frequency of occurence, %

28 Outline Radar Network upgrade to dual polarisation UK Weather Radar Network products and capabilities Continuous development

29 Beyond network renewal Our goal: Deliver optimum radar products to improve short-range forecast particularly of severe weather. Great opportunity: Dual polarisation Signal Processing (Cyclops D4) and receiver was developed inhouse we have complete control over the signal processing. Radar reflectivity Rainfall Radial wind Radar refractivity NWP (DA, STEPS, UKPP, VER) FSD EA, FFC Working with our academic partners, our research focus on: Extracting maximum information from the radar network. Developing new signal processing and central processing techniques for improving the accuracy of the radar products.

30 Spurious echoes identification Problem: A major limitation to assimilating radar products into NWP and hydrological models is the presence of non-precipitation echoes in the data. POD Jan 2012 POD Jan 2013 Average dbz Jan 2013 Improved scheme was introduced last year based on Nicol et al (IAHS Publ 2012). This technique show clear improvement to the data; however some limitations with sea clutter and RLAN interference. Going forward: use dual polarisation parameters to improve this scheme further.

31 Characteristics of weather radars microwave frequencies Higher frequency: more severe Attenuation during critical events Severe attenuation of the radar return is a major problem in intense rainfall, particularly during event likely to cause floods. From Delrieu et al, 2000: Quantification of Path-Integrated Attenuation for X-band and C-band Weather Radar Systems Operating in the Mediterranean Heavy Rainfall Extreme rain

32 Passive emissions Dept. of Meteorology Attenuators are emitters (Prevost 1818). Radar signal Attenuating storms emit at the radar frequency. Emission can be expressed as the total attenuation along the path of the radar beam. Radar receiver is sensitive enough to act as a radiometer and measure emissions (transmitter switched off) Our goal: to bring new emission techniques combined with dual polarisation parameter into operational use for improving the identification and correction of attenuation of the radar return caused by rain and wet radome.

33 Improved QPE accuracy during heavy rainfall Problem: Severe attenuation of the radar reflectivity in heavy rainfall cause severe under estimation of rainfall. KPD is a phase parameter immune to the error related radar calibration or reduction of the reflectivity factor caused by partial beam blockage, attenuation by precipitation or wet radome. KDP is also less sensitive than reflectivity to changes in the DSD. New opportunity: use R=f(KDP) instead of R=f(Z). KDP can be used to identify area of heavy rain and derive rainfall estimate without using the reflectivity.

34 RAINGAIN Objective: to improve fine-scale measurement and prediction of rainfall and to enhance urban pluvial flood prediction. This will enable urban water managers to adequately cope with intense storms, so that the vulnerability of populations and critical infrastructure can be reduced. Sites: The activity includes the implementation and use of advanced radar technologies (dual polarisation C band) over central London, and (X Band) in Leuven, Paris, and Rotterdam. Partners: The project gathers 13 partners in Belgium, France, Netherlands and the UK (local authorities, universities and enterprises).

35 Super-resolution processing Can signal processing techniques refine resolution of the rainfall estimates for urban catchments? Our goal: 100m or better resolution over central London by m data on Invent

36 Gauge-Radar merging techniques Objective: The development and implementation of a high resolution radar-raingauge merged product for use in real-time of quality controlled raingauge data. The scheme is required to run using a 15 minute accumulation time for use in flood forecasting. Radar only Gauge only The project is assessing the influence of the following parameters on the quality of the merged product: Merged Choice of merging method. Gauge density. Degree of rainfall correlation. Polling time

37 Summary The new UK radar system is delivering top class parameters including dual polarization radar refractivity and passive emission to support research activities focusing to improving short-range forecast particularly of severe weather.

38 Questions & answers