Searching for a Cyclone - the work behind the scenes of the Melbourne Osaka Race
The ORCV Race Management Team for the 2018 Melbourne to Osaka Double-Handed Yacht Race includes 6 Duty Race Directors and 2 Fleet Weather Officers - Robin Hewitt & Andrew Roberts. Robin and Andrew are on a Daily Roster and maintain a constant watch on weather conditions for the fleet. They prepare detailed analysis of weather outlook at least twice daily (more often if situation dictates) to Duty Race Directors for review. If forecast conditions exceed pre-defined wind speed and sea state thresholds, Duty Race Directors liaise with Fleet Weather Offices then escalation of Weather Event Warnings to the fleet via SMS (sat phone), Email (sat phone) or HF marine radio during the twice daily mandatory position report communications with the fleet (or more often as required). The fleet also source thier own weather forecast information via various methods.
Fleet Weather Officers and Duty Race Directors have been very busy during this race with 3 of 4 starts delayed and several weather events, including Tropical Cyclone Iris which resulted is racing being suspended. The Fleet Weather Officers commenced tracking a deepening Tropical Low which was initially upgraded/classified as Category 1 Tropical Cyclone Iris by the Fiji Meteorological Service, then subsequently downgraded, then upgraded again by Australian Bureau of Meteorology to a Category 3 Cyclone, a few weeks later.
To follow is a report from Robin Hewitt and some of the extensive research on Tropical Cyclone activity, which was conducted long before this year's race and some detailed analysis on Tropical Cyclone Iris that led to the suspension of racing.
Searching for a Cyclone, by Fleet Weather Officer Robin Hewitt
There have been instances of tropical revolving storms occurring in the history of the Melbourne to Osaka Race, in each case as typhoons in the Northern Hemisphere. First race of 1987 saw some yachts anchoring in shelter for up to four days, while the 1991 race had super typhoon Walt formed near Guam, necessitating some course selection. In the 2003 race some yachts were obliged to shelter in Ulul lagoon (Onoun Island, Southwest corner of the Namonuito Atoll, Federated States of Micronesia) as a typhoon neared.
In the race management team for the 2018 event, Robin Hewitt and Andrew Roberts have shared the task of providing warnings of extreme events. What are the possibilities? Tropical Revolving storms in the area of interest are known as Cyclones in the southern hemisphere and Typhoons in the northern hemisphere. In addition, the spring season in Japan is subject to the storms sweeping in from icy Siberian land masses. The polar jets can intensify these Lows to bring severe weather to competitors as they near Japan. Similarly, cold fronts and deep Lows can escape from Antarctica but in March are unusual, more likely is a cyclone over northern WA crossing land and re-appearing near Perth before joining a front thence heading towards or below Tasmania.
So how to prepare to manage a race through the tropics? Methodology is standard-collect data-analyse-prepare hypothesis. First of all, the race is timed for the least climatic possibility of tropical revolving storms. The period of April and either side has the lowest incidence historically. So firstly a table of occurrences was prepared for the times of interest. Then they are compared to the theoretical model for possible variations. Formation does not occur between 5ᴼN & 5ᴼS and mostly do not form between 10ᴼN & 10ᴼS.
This is because they need the Coriolis effect to create spin. They require a vertically ascending motion involving moisture as high humidity over water of about 27ᴼC and utilize latent heat release as cooling water vapor turns to liquid and as the still ascending water turns to ice. If winds aloft can interrupt the vertical ascension then they cease development. They are however, steered by gradient winds and the easterlies prevalent in the equatorial region, generally cause them to initially move west and curve towards the poles until reaching more westerly gradient winds. Most weather agencies use a colour coding system of classification known as the Saffir-Simpson scale shown here.
Saffir–Simpson scale
|
Category
|
Wind speeds
|
Five
|
≥70 m/s, ≥137 knots, ≥157 mph, ≥252 km/h
|
Four
|
58–70 m/s, 113–136 knots, 130–156 mph, 209–251 km/h
|
Three
|
50–58 m/s, 96–112 knots, 111–129 mph, 178–208 km/h
|
Two
|
43–49 m/s, 83–95 knots, 96–110 mph, 154–177 km/h
|
One
|
33–42 m/s, 64–82 knots, 74–95 mph, 119–153 km/h
|
Related classifications
|
Tropical Storm (TS)
|
18–32 m/s, 34–63 knots, 39–73 mph, 63–118 km/h
|
Tropical Depression (TD)
|
≤17 m/s, ≤33 knots, ≤38 mph, ≤62 km/h
|
What is evident from studying historical events is that if they touch land and then go offshore again, they rapidly intensify. The previous track examples illustrate the theoretical model. Following are some variants which illustrate the erratic possibilities.
Historial Tropical Cyclones / Typhoons
April
|
May
|
June
|
1966: Hester
1967: Violet
1981: Gerald, Holly
1986: TD, Ken
1990: Lewis
1991: Vanessa
1995: Chuck
1996: Ann, 03W
1997: Isa, Jimmy
1999: Jacob, Kate, Leo
2001: Barok
2002: 04W
2003: Kujira
2004: Sudai
2005: Sonca
2007: Kong-Rey
2008: Neoguri
2009: Crising, Kujiva, TD
2010: TD
2011: 01W, Amang
2012: Pakhar, TD, TD
2014: Peipah, TD, Maysak 2015: Haishen
2017: Crisng
|
1966: Bising, Irma, Judy
1967: Wida
1986: TD, Lola Mac
1990: Marian, TD, TD, TD
1991: Walt
1995: TD, TD, TD
1996: Bart, Cam
1997: Kelly, Levi, Marie
1998: TD
1999: Maggie, TD
2000: Damey, Longwang, TD, Konsing, 04W
2001: TD, Cimaron
2002: Hagbs, TD, Dagui
Baticot, TD, Charsom, Linfo
2003: Nangka
2004: Nida, 05W, Omas
2005: Crising, Nesat
2006: Chanchu
2007: Yuta
2008: Rammasan, Matmo, Halong, Nakri, TD
2009: Chan-han
2011: Aere, Songda, TD
2012: Sanvu, Mawar
2014: Tapah
2015: Noul, Dolphin
2016: 01W
2017: Muifa, TD
|
1966: Kit
1967: Anta, Bille
1981: Ike, TD, June, TD, Kelly, Lynn
1986: TD, Nancy, TD, Owen
1990: 04W, Nathan, Otelia, Percy, TD
1991: Yunga, TD, TD
1995: Deanna, Eli, TD, TD
1996: TD
1997: Nestor, Opal, Peter, TDx3
1999: Gening
2000: TD, Kirogi
2001: Dami, Chebi, Durian, Utor,
2002: TD Naguri
2003: Sanglefor
2004: Conson, Chanthu,Daninu, Mindulle, TingTing
2005: Nesat
2006: Jelawat, Ewinjar
2008: TD, Fengshen
2009: Linfa, Nangka
2011: Sarka, TD, Haima, Meari
2012: Guchol,Talin
2013: Yagi, TD, Leepi, Betinca, Rumbia
2014: Mitag, Hagbis
2015: Kajira
2016: TD, Ambo
2017: Soudefor
|
Highlight-Typhoons/Tropical Cyclones not applicable the Melbourne to Osaka race area
So how can we look for possible concerns? Lows on the Monsoonal Trough are one prospect. The Monsoonal Trough is an area of low pressure to the north of Australia. Hot air rising is low pressure and this can have a similar effect to the trades as air rushes in to replace the rising hot air, only this time it is the northern cooler air from Indonesia and Asia-Monsoon. Also on the front of the Monsoonal air rush with thunderstorms and rain are the conditions suitable for a Low to become something else. Ditto along the Intertropical Convergence Zone (ITCZ) or Doldrums.

And then there are known hotspots. Eg Guam, Koror, Palau and from 5-10ᴼS or N. Procedure is to use the wonderful 7 or 10 day forecasts that are now available from several organisations. Perhaps best known are the European Centre for Medium-Range Weather Forecasts (ECMWF) prediction model from Europe, the Global Forecast System (GFS) model from the USA’s NOAA, and the BoM Australian Community Climate and Earth-System Simulator (ACCESS) model locally. These models are mathematical predictions based on fluid dynamics using complex algorithms and observation data in giant computers. The Australian local forecasters have sub-routines and experience to ‘adjust’ model outcomes. Next we move ahead in time to see what the predictions are for pressure systems. If a Low continues to deepen in the models, it is a surefire warning to watch closely and see if the pressure predictions are holding true. Because the atmospheric tide in the tropics can vary pressure by 3-5 millibar when closer to the equator, it is necessary to check each day’s prediction at the same times. The atmospheric tides are low at 4am and 4pm, high at 10am and 10pm. There are rules of thumb for rate of pressure drop/rise and wind strengths. The weather bureaus generally issue warnings much closer to certainty and share predictions for worldwide publishing. Some sites publish worldwide warning and illustration products as a collective facility. Relatively slow vessels need more warning.
Also with the advent of satellite technology, there are patterns that can give rise for further investigation or for keeping an eye on. A group of thunderstorms becomes organised in a tropical Low and ripe for further development. Satellites use both visible and infrared satellite imagery to track the motion and cloud patterns of tropical revolving storms and infrared to monitor cloud-top temperatures-thus how high they reach (strength). The weather bureaus have trained observers, and of course, supercomputers to analyse data.
 |
 |
 |
Armed with areas of interest, it then becomes a matter of observing as many prediction models as possible along with actuals which brings the next problem. Each of the prediction modelling programs have a different mathematical basis and the further ahead one looks, the greater the differences are. One model may show, for example, a cyclone developing five days into the future while another model shows nothing unusual. As time goes on all the models will come together to match the actual or close future. If all models agree for a specific time, one has a high degree of confidence, but if they do not???? This requires a degree of judgement and perhaps if using a routing function, one should allow for contingency plans to be able to modify the route. The three pictures (above and left) are all for the same date and time from 3 different forecast models: 1. ECMWF, 2. GFS, 3. ACCESS. |
Now let us look at an actual situation. The remnants of Cyclone Iris is wandering around the vicinity of the M2O route to the ‘gate’ between Bouganville and New Ireland, behaving erratically and not conforming to any rules. The average route is N along longitude 155ᴼE showing areas of 30-40 knots plus gusts of 40%?
 |
Gale strength but probably choose downwind. On this date the early starters Morning Star, The Edge and Bartolome are spread from Roughly 24ᴼS to well off Cairns. The main start are spread out between Gabo and off Canberra moving steadily north using inshore and offshore strategies. This BoM map gives some options but further prognosis shows the Low driftingto NW along the Qld coast.
Add a further concern - will the Low of 1005 deepen and form to cut off escape further to the east of remnants of Iris in it’s current form as a Tropical Low? Shown in the gradient wind analysis. Note the analysis is always given in UTC time.
|
Issued at 5:08 am EST on Sunday 1 April 2018 for the period until midnight EST Tuesday 3 April 2018.
At 5:00am EST Sunday, ex-Tropical Cyclone Iris was located near Latitude 15.1 degrees South, Longitude 149.5 degrees East, about 450 km East-Northeast of Cairns. The Low is currently moving Northwest at about 5 km/h.
The system is likely to remain slow moving for much of today before adopting a more Southwesterly track and moving closer to the tropical or central Queensland coast on Monday and Tuesday.
|
 |
The system is expected to remain a subtropical Low until Tuesday, when the system moves into a more favourable environment and the chance of further development increases to moderate.
 |
The satellite picture am on 1st April shows tropical Low Iris and further east is the Low which eventually develops into TC Josie. If TC Josie follows the rules it will move west towards the coast and the advancing main fleet will be sandwiched. Iris is erratic and could go anywhere, if it nears or crosses the coast it will either dissipate inland or alternatively reform offshore and strengthen-will it? Nexba and Allegro are now near Brisbane while Morning Star, The Edge and Bartolome are through and moving into the ‘No TC’ latitudes. |
On the morning of 2/4/18, Nexba and Allegro are E of 156ᴼ and of latitudes of approx. 25ᴼS off from Fraser Island moving Northwards in excess of 8 kts. By evening warning is given Iris has reformed as a cat 1 cyclone and a projected track issued. Models are not agreeing so the field is wide open.

Below are the comparisons taken with model forecasts side by side. Note on the GFS (right) map a new Low is forming between the two systems.
BoM TC Iris projected track superimposed on a the BlueWaterTracks yacht tracker:

A forward position estimate for most of the fleet shows they could not avoid problem areas and further TC Iris projected track advice (pictured below) forecasts development to Category 3 possible. Check conditions on the Australian Cyclone chart for Category 2 & 3. Unlike the recent Volvo race, which had a wide area of ocean to play in, we have a sandwich due to location of the Great Barrier Reef, TC Iris and TC Josie which would have significantly limited the options for the fleet. All in an area of strong currents.

Mon 2/04/2018 4:30 PM
Notice to Competitors No. 5
Safety Warning
Due to significant concerns about the forecast track and potential strength of the reformed cyclone Iris off the coast of Townsville, we require that you make appropriate plans to ensure your safety.
In accordance with SI 3.3 we instruct that you choose one of the following options Option 1 Upon reaching latitude 22S remain east of longitude 155E until the race committee advises otherwise Option 2.
Irrespective of your longitude, do not sail further north than Latitude 23S until the race committee advises otherwise. Competitors should choose a position that suits them tactically (Latitude and Longitude) and record a GPS time (AEST) should they suspend racing, communicating that position and time to the Race Director. Communication skeds must continue at all times. You may choose an anchorage or a mainland marina at your discretion and should you require assistance to find a safe haven, please contact the Race Director.
In order to protect the integrity of the race, competitors may elect to replace any fuel used to travel to/from safe haven and any provisions used while racing is suspended, but must not take on any additional provisions or seek any external assistance. Competitors must log any provisioning (incl quantities) and include this as part of their written post race declaration.
Our intention is to continue racing as long as it is safe to do so. Competitors may need to suspend racing after which time you will need to return to your recorded suspended position and record a GPS time (AEST) for recommending racing, again communicating that position and time to the Race Director. Results of the race will be based on elapsed racing time.
We have not taken this decision lightly, but our clear priority is competitor safety. Should you feel these instructions are injurious to your competitive status, redress remains an option.
Should you choose a mainland port, there may be some customs implications which we are looking into.
Can you please immediately acknowledge receipt of this Email.
Simon Dryden
Principal Race Director
Fears Tropical Cyclone Iris may intensify to category three overnight
3 April 2018 — 6:25am
Tropical Cyclone Iris is expected to intensify into a category three by Wednesday as it continues to move along the Queensland coast.
Forecasters predicted the weather system would strengthen into the early hours of Wednesday morning, bringing destructive winds and heavy rain to parts of the state.

Example report using satellite imagery
B. TROPICAL DISTURBANCE SUMMARY:
(1) AN AREA OF CONVECTION (INVEST 94P) HAS PERSISTED NEAR 12.2S 178.2E, APPROXIMATELY 335 NM NORTH OF NADI, FIJI. ANIMATED MULTISPECTRAL SATELLITE IMAGERY (MSI) AND A 180339Z SSMIS 91GHZ
MICROWAVE IMAGE REVEAL A CONSOLIDATING LOW LEVEL CIRCULATION CENTER (LLCC) WITH FRAGMENTED, FORMATIVE BANDING BEGINNING TO WRAP INTO THE LLCC. THE SURROUNDING ENVIRONMENT IS FAVORABLE FOR DEVELOPMENT WITH MODERATE VERTICAL WIND SHEAR (15-20KTS), GOOD EQUATORWARD OUTFLOW, ENHANCED POLEWARD OUTFLOW IMPROVING DUE TO THE PROXIMITY OF THE WESTERLIES, AND WARM SEA SURFACE TEMPERATURES (28-29 CELSIUS).
GLOBAL MODELS DEPICT THE SYSTEM TRACKING POLEWARD HOWEVER, THE MODELS HAVE YET TO INDICATE ANY SIGNIFICANT TROPICAL CYCLONE DEVELOPMENT. MAXIMUM SUSTAINED SURFACE WINDS ARE ESTIMATED AT 20 TO 25 KNOTS. MINIMUM SEA LEVEL PRESSURE IS ESTIMATED TO BE NEAR 1004 MB. THE POTENTIAL FOR THE DEVELOPMENT OF A SIGNIFICANT TROPICAL CYCLONE WITHIN THE NEXT 24 OURS IS MEDIUM.
TC Iris overlay on Google Earth:

The race continues. Our search has shifted to the Northern Hemisphere area known as NW Pacific.
The Incident Management Team protocols and training instituted in ORCV by Rik Head has again been proved invaluable in conducting races like this year’s M2O.
If you are interested in becoming more familiar with understanding the terminology, tools and concepts in this article, the ORCV weather courses delve much deeper on this material.
Robin Hewitt with thanks to Andrew Roberts and Neville Rose.
References:
Australian Bureau of Meteorology (BoM) http://www.bom.gov.au/
Blue Water Tracks https://race.bluewatertracks.com/2018-osaka-race
Fiji Meteorological Service http://www.met.gov.fj/
Joint Thyphoon Warning Center (JTWC) from United States Naval Observatory http://www.usno.navy.mil/NOOC/nmfc-ph/RSS/jtwc/jtwc.html
PredictWind https://www.predictwind.com/
Weather Zone http://www.weatherzone.com.au/
WikiPedia https://en.wikipedia.org
Windy https://www.windy.com