Utilising Radar and Satellite Based Nowcasting Tools for Aviation Purposes in South Africa Erik Becker Morné Gijben, Mary-Jane Bopape, Stephanie Landman South African Weather Service: Nowcasting and Very Short Range Forecasting
Overview Aviation Research Demonstration Project (AvRDP) Phase I: Evaluate nowcasting capabilities for airport Phase II: Translate nowcasts product into impact based products 1. Current Nowcasting Capabilities 2. Currently in Development 3. Nowcasting Products for ORTIA 4. Way Forward Impact Based Forecasting
Introduction OR Tambo International Airport (ORTIA): 26 08 21 S and 28 14 46 E 20km east-north-east of the city of Johannesburg and 40km south of Pretoria, situated almost on the plateau of South Africa with two runways, both above 5500ft (1694m) above MSL. The surrounding area consists of hilly country sloping gradually from south-west to north-east. The highest terrain within a radius of 40km is 1902m to the south of the airport. Most of the areas directly around the airport are built up areas.
Introduction Airport local climatology: Fog: radiation fog/advection from SE, mostly early am (2-3 days per month) Low clouds/poor visibility, early am Smog: temperature inversion, winter months, early am Precipitation: thunderstorms, summer, pm, early evening Hail: 5 days per year, Oct-Dec Rain: max in Dec/Jan (600-800 mm per year) Frost: winter, 30 days per year Wind direction: Mostly NW wind Wind speed: 3.3-4.6 m/s
SAWS runs the 2013 version of the Nowcasting Satellite Application Facility (based in Spain) software. Input data from: MSG satellite UM regional model (4km) lightning data from LDN Products operational at SAWS: Cloud Type Cloud Top Temperature, Convective Rainfall Rate Rapidly Developing Thunderstorms (RDT) Currently busy upgrading to 2016 version. Nowcasting SAF
Rapidly Developing Thunderstorms (RDT) Information courtesy Morne Gijben
Example 21 Mar 2015 Information courtesy Morne Gijben
Lightning Detection Network In 2005 SAWS installed a Lightning Detection Network (SALDN) A network of VAISALA sensors were installed Sensors measure cloud-to-ground (CG) lightning 25 sensors Detection efficiency of >90% Location Accuracy of <500m Information courtesy Morne Gijben
SAWS Radar Network TITAN tracking
Community- Short-range Warning of Intense Rainstorms in Localized Systems (Com-SWIRLS) Serious need to improve 0-6 hour lead-time forecasts. Com-SWIRLS made available by the HKO. Produces extrapolated nowcasts from CAPPI reflectivity data (Optical Flow, no growth and decay) SWIRLS installed over: Irene Radar domain 200km range. 400x400 pixels (1kmx1km) 9hr extrapolation
Com-SWIRLS Verification Verification Case Days: 2015-12-21 2016-07-26 2016-10-26 2016-11-02 Extrapolated forecast slower than observation field in all cases. Including growth and decay of thunderstorms might be worth looking into. Thunder storm initiation could also be considered. Two hour extrapolation operational by end of 2017.
Mesoscale Modelling Convective scale model (300 m) using the UK Met Office Unified Model ORTIA nested domain 300 m resolution Vertical resolution of 70 levels Lead-time 36-hours but initialized at 00, 06, 12 and 18 UTC Red (1,5 km) & Purple (300 m) 300x300 (shaded) & 600x600 (non-shaded) Black circle ORTIA aerodrome
Unified Model (300m) Information courtesy Stephanie Landman
WRFDA WRF & WRFDA Radar data to be assimilated (dbz and VEL) Irene only Radar dbz data QC to be in place (R4A) Investigate other obs data to be assimilated (SAT, SYNOP, METAR, etc.) Initialized at every hour on the hour (Rapid Update) Lead-time to between 6 and 18 hours (operational lead-time depended on resources) 300x300 nested grid to ~500m resolution Driving data: 25 km GFS 3DVAR (memory limitations) Investigation will determine what would be possible with SAWS s limited resources.
Developing Forecast Products (AWC guidance) Impact weather at ORTIA: Visibility (Fog, Precipitation, etc.) Convective Thunderstorms (Hail, Lightning, etc.) Turbulence and/or Wind Shear Product development (WORKING ON VARIOUS IDEAS) Convection forecasting Visibility on convective sub-kilometre scale UM Wind Shear & Turbulence Lightning Aerodrome lightning product Plans are to get the products working on what is available at SAWS at the moment and improve and update as new developments become operational
Example: Lightning Proximity Product No Lightning lightning within within 5-15 3-5 3km 15km radius radius Information courtesy Morne Gijben
Example: Storm Tracking Product Impact buffer zones around airport (4, 15, 30 and 50km) are defined. Storm forecast based on 2 hour leadtime using current storm cell tracked speed and direction. (TITAN) Red, Orange, Yellow and Green buffer zones defined around storm track based on the direction of movement. 0, 10, 20, 30, 45, 60, 90, 120 min forecast table is contracted from analysis results.
Example: Storm Tracking Product
Example: Storm Tracking Product
Example: Storm Tracking Product
Way Forward - Summary Goal is to work towards seamless 0-6 hour forecast (even 0-36 hour) Continue to develop and improve on the 0-6 hour forecast lead-time. Implement SWIRLS operationally Continue working towards UM and WRF high-res models. Include WRFDA for radar data assimilation to improve spin-up time. Research to continue on improving forecasts and development of impact based products. Data being collected for more objective evaluation approach. Include more ancillary data. i.e. waypoint location, flight routes, approach and departure paths, etc. Include other Met information as it becomes operational. (SWIRLS, NWP, etc.)
Way Forward - Summary Work towards translating nowcasting products into impact products. Look into impact weather at the airport. Determine what would be minimal, minor, significant and severe for different ATM entities. AWC and ATM entities have to establish new and build on current connections/relationships.
Thank You Questions? erik.becker@weathersa.co.za