How Strong Was Yasi?-Update 11 Feb

Ken Stewart, February 2011

Update Friday Feb 11 8.12pm:  On ABC news tonight, a report from James Cook University, whose scientists have established Yasi’s speed, at

“The CTS team leader in the field, Dr Geoff Boughton, said that the analysis of damage to simple structures throughout the region indicated that the wind speeds on the ground in Tropical Cyclone Yasi were less than those expected in a Category 5 event.

Using techniques developed and refined in previous cyclones, the CTS team has formed a preliminary view that the maximum wind speed in Cyclone Yasi was about the same as that in Tropical Cyclone Larry.

“We estimate that the gust wind speeds in some of the most affected areas were about 220 km/h” Dr Boughton said.”

In fact that’s at the low end of my estimate!

On Wednesday 02 February 2011, TC Yasi struck Far North Queensland.  Despite the enormous devastation in the Cardwell-Mission Beach-Tully area, thanks to Emergency Services Queensland, Anna Bligh, and the Bureau of Meteorology (BOM), people were well prepared, evacuated when it was thought necessary, and only one life was lost.

However, many people have noticed that Yasi was nowhere near as strong as expected.

The Bureau of Meteorology estimated that windspeeds would reach 290kph near the centre, and at 8.00pm warned these could be expected between Cairns and Ingham.  This estimate was made from analysing radar and satellite data.  But at 4.33p.m. BOM’s technical advice estimated winds at the centre as being 205kph (my emphasis):

Technical brief issued by BOM at 6:33 UTC.

The latest imagery shows Yasi as very symmetric showing a well defined eye with surrounding deep convection. Dvorak intensity: Eye pattern white surround [6.0] with OW/W [0.5] eye adjustment giving DT=6.5, adj. MET=6.0. FT/CI=6.5. Max winds estimated at 110 knots [205 km/h] [AMSU/SATCON estimates at 125kn [1min]].
Surface observations indicate hurricane force winds extend about 50 nm [95 km] to the southwest of the track.

As well, they had access to data from Willis Island which was right in Yasi’s path and which transmitted data until just before the eyewall passed over it.  Yet they still maintained Yasi at Cat 5, probably as a precaution.    They received unjustified criticism over the Toowoomba and Lockyer floods so government and BOM were taking no chances.   What most people don’t realize is that cyclone category is not an exact statement but a warning.  From the BOM website:

The severity of a tropical cyclone is described in terms of categories ranging from 1 to 5 related to the zone of maximum winds. An estimate of cyclone severity is included in all tropical advices. Remember that the Warning Service is not designed to give an exact statement of conditions at individual locations but will give a general idea of the expected worst conditions.

And while no more precise estimate of wind speed can be made until completion of engineering analysis of the damaged areas, news reports are still universally reporting winds of 290kph.  In the meantime, here is an alternative estimate.

Although the maximum recorded windgust at Lucinda was 185kph, this was 50km south of the centre and about 35km from the eyewall, and a look at the data shows the wind changing direction every few minutes, possibly an effect of the nearby topography.  These speeds are probably not reliable.

I have used data from Willis Island to estimate windspeeds in the lead  up to the crossing at Mission Beach.

Willis Island was directly in the path of Yasi’s eye.  It transmitted data right from  the edge of the eyewall.  This is its final radar image (courtesy of Watts Up With That):

Here’s the graph of windspeed:

and the table of observations:Notice the wind suddenly changed from south at 6.00 a.m  to north-west at 6.15 a.m.  with no drop in speed (102kph sustained).  I can think of 3 explanations.

  1. Possibly the north face of the eyewall passed over it.  However, the radar image at 8.00 shows the island right in the firing line, and still some distance away.  The cyclone would have had to change course.
  2. Alternatively, the eye passed over for about 15 minutes and the calm was not recorded.  This is odd.  We were told the calm eye would take an hour to pass over.  Besides, the radar image shows the eye still to arrive.  The wind increased to 141kph, gusting to 185kph, at 8.10a.m. The anemometer then ceased operations, although the barometer recorded decreasing pressure (to 938hPa at 9.00) then increasing to 940.1 at 10.00a.m.  The continued drop in pressure means the wind might have increased even further by perhaps 20 to 30kph, making the top gust at around 215kph.
  3. Perhaps the wind vane was damaged at 6.15?  Then for two hours the cyclone travelled 60-70km closer and the wind increased by 40kph sustained and 46kph gusts.  This is most likely.

The 8.00 a.m. radar image shows Willis island was approximately 15-20km from the eyewall and 30-35km from the centre.  Maximum speed is reached at the eyewall.  I used the data to make a plot of the data including an interpolation for 5.45 a.m., and  extrapolating to 8.30 a.m., the approximate time of the eyewall’s arrival.

As you can see I used 3 estimates for the 8.30 windspeed: the BOM 4.33 p.m. estimate of 205kph; the 8.12 p.m. cyclone advice estimate of 290kph, and 231kph, the AMSU/SATCON estimate and just above the borderline Cat 3/Cat 4 speed.

I believe that the estimate of 290kph cannot be supported- it appears too high.  Further, this is not supported by BOM’s 4.33 p.m. technical brief.  However it appears that 205kph could be too low for 8.30pm at Willis Island.  230kph appears to be closer.

The cyclone was travelling at about 30kph at this time, so every hour is approximately 30km.  Using 9.00 a.m. as the time of the cyclone centre passing just a few km to the south of Willis Island with a QNH (barometric pressure) of 938hPa, I constructed a plot of observed windspeeds at approximate distances from the centre.

The horizontal scale is distance from the centre in kilometres.

Therefore, a defensible estimate for windspeed in the Cardwell-Mission Beach area would be in the range 220-240kph.  This indicates it was a low Cat 4, not Cat 5.

Yasi was indeed an enormous system in area covered by cloud, the largest we’ve seen in the satellite era.  The zone of maximum destruction (and winds) extended from roughly Silkwood to Cardwell, a distance of about 60km.  The storm was roughly the same strength as Cyclone Larry but took longer to pass.  Remarkably, it was still classed as a cyclone at Julia Creek, the furthest inland a cyclone has been recorded.

Contrary to many alarming reports, it was not the deepest cyclone (<926hPa, Mourilyan Mill, 1918) nor the highest storm surge (Bathurst Bay >10m, 1899 or Mission Beach 3.6m 1918) nor the the most rain (907mm in 24 hours at Crohamhurst, 1893) nor the deadliest (307 known fatalities, Bathurst Bay, 1899).

Which is no consolation for the residents of Cardwell, Tully, and Mission Beach.


24 Responses to “How Strong Was Yasi?-Update 11 Feb”

  1. val majkus Says:

    thanks Ken; I’ve llnked to Warwick Hughes and Dr Marohasy’s blogs

  2. Marc Hendrickx Says:

    Pity BOM has so few instruments up there. i can’t think why not!

  3. Len van Burgel Says:


    One important correction.

    The Bureau technical brief refers to sustained (10 minute) winds of 110 kn or 205 km/h and not gusts. The estimate is not different to the AMSU/Satcon estimated wind of 125kn because that is a 1 minute wind. Converted to a 10 minute wind used in Australia it is also 110kn.

    Sustained winds of 110 kn (205 km/h) can have associated gusts to 290 km/h. This fits in with the reported Dvorak CI (current Intensity) of 6.5 and a category 5 cyclone. A category 5 cyclone has average maximum wind of over 200km/h with gusts over 279km/h

    So you see you are not comparing like with like.

    The BOM estimate of 205 km/h is a mean 10 minute wind speed from the technical bulletin and represents 290 km/h gusts.
    The 231 km/h AMSU/SATCON is a 1 minute mean wind which is equivalent to a 205 km/h 10 minute wind (and 290 km/h) gusts.
    The cyclone advice of 290 km/h is an estimate of gusts.

    There is no contradiction here.

    It is a common mistake to confuse mean winds with gusts. For example gale warnings are issued for winds (mean winds) in excess of 34 knots. Gusts are 40% greater.

    Unfortunately as far as the public and media are concerned the mean wind and gusts are used indiscriminately and interchangeably. So gales of 65 km/h with gusts to 90 km/h always seem to be reported by the media as winds to 90km/h.

    For your interest JTWC at one stage estimated Yasi to have one minute wind gusts to 165 kn or 305 km/h. The equivalent 10 minute wind is 270 km/hr. However one must also add the movement of the cyclone, so that in the forward quadrant it is 25/30km/hr higher.

    • kenskingdom Says:

      Len, points noted, and thank you. However this raises further serious contradictions with observations. Was the Willis island data completely unreliable hours before it was hit? I haven’t time right now but hope to have an update posted this evening showing QNH pressures and a comparison between estimated sustained wind speed and gusts you have detailed and observations. It doesn’t look right to me.

  4. Len van Burgel Says:


    Be careful. The evidence is overwhelmingly strong that the wind estimate was right, at least when it crossed Willis Island. I have no comment about the later coastal crossing.

    AMSU/SATCON is a method of calculating maximum wind speeds from satellite derived mid-tropospheric temperature gradients. It is a well developed method and more objective than the time honoured Dvorak scheme. Both Dvorak and AMSU/SATCON agree with the estimate of intensity which leads to 110kn (205km/h) wind or 290 km/h gust.

    My maths is rusty but you have kindly put in your article enough to be able to calculate the wind speed theoretically.

    The Cyclostrophic wind equation will give you wind velocity based on pressure gradient and radius from cyclone centre.

    V (squared) = R (radius) X pressure gradient

    The reports show a pressure fall of about 26 hPa between 7.37 and 8.37. Between 8.10 and 8.37 (when it was near the lowest pressure value) the pressure fell 13.6 hPa which is an hourly rate of 30hPa. If you plug in pressure fall of 26 hPa/h and assume a 17 kn (30km/h) movement towards Willis Island, then roughly the pressure gradient is 26 hPa/30 km (0.087 Pa/m). According to the radar at 8 am the cyclone was 45 km (R= 45000m) from Willis Island and the Island was close to the core of maximum winds.

    Hoping my unit conversion is Ok, I get a theoretical wind speed of 62.5 m/sec or 225 km/hr. To get the surface 10m wind from the theoretical wind, a conversion factor needs to be applied. The value is variously quoted as 0.7 to 0.9. Taking 0.8 as a factor the 10m wind is 50 m/sec or 180 km/h mean wind and 255 km/h gusts.

    For a back of the envelope calculation, this is reasonably close to the prediction, also taking into account a 30km/h cyclone movement adding to the actual wind in the forward quadrant. In fact if the cyclone moved just south of Willis Island a subtraction of 30 km/hr from the theoretical 180km/h mean wind places it quite close to the highest observed mean wind at Willis of 141 km/h.

    If the cyclone did not move directly towards Willis then the pressure gradient is actually higher and the theoretical wind will be higher as well.

    Maybe someone could confirm my calculations and methodology.
    Unless my sums are wrong, this adds weight to the Dvorak and AMSU/SATCON assessments.

    • kenskingdom Says:

      Hi Len. Thank you, and your maths seems pretty good! And OK, my assessment rested on the estimate of maximum winds being the gust speed instead of the mean speed. If that is the case then my estimate is incorrect, and the mean winds went from 141 to 180km/hr in about 20-30 minutes- scary! But how is the 290km/hr gust speed derived? I would agree from your explanation that 255km/hr gusts is reasonable.

      • kenskingdom Says:

        Len: On the other hand, using a rule of thumb 25% extra for wind gusts, using your calculation of 180km/hr sustained wind, this would make wind gusts reach 225km/hr wouldn’t it? Which is what I originally estimated.

  5. Len van Burgel Says:

    Thanks, Ken.

    The gusts are determined by applying a gust factor which is normally 1.41. So the BOM”s 205 km/h mean wind gave a predicted gust of 205 X 1.41 = 290 km/h although warnings may use different figures if it is deemed applicable.

    The actual gust factor may be less and it could be a lot more, but then usually for other reasons. Readers may be interested that the world record highest wind gust occurred in Tropical Cyclone Olivia on 10 April 1996 at Barrow Island, WA. Gusts were recorded to 407 km/h.
    In this case the gust factor was 2.75. It was thought it was due to a meso-vortex in the eye wall. Scary!

    • kenskingdom Says:

      This is becoming very hypothetical, and we await final estimates from ground assessment. However, averaging gust increase over sustained from 6:00 to 8:10 gives 1.32, so 1.32 X your calculation of 180 = 237kph. 205 x 1.32 gives 270kph. However I note that JTWC’s final report before landfall gave 127/140 knots : 235/260kph, which is just 10.6%; and in same report 135/165 knots : 250/305kph (22%) and a 60 nautical mile eye. This last estimate of speed and eye size does seem quite high. Time will tell. Until I’m proven wrong, I’ll stick to Yasi as Cat 4.

  6. Mission Beach Survivor Says:

    having survived both Yasi & Larry at Mission Beach
    I can asure you ALL ……..that Yasi’s winds were
    constant 200kph+ for 45m and …….
    gusting 300kph+ for 45m before the eye crossed

    we were on top of a 60m hill, on the coast, the solid concrete house was shaking, like a vibrating roller was outside, and the noise ……….

    Larry was a picnic compared to Yasi……..
    220kph max BULLS**T!!!!

    • kenskingdom Says:

      Not being there, I cannot comment on your experience, except that you have my sympathy. Ului was bad enough for me, and I’m sure I wouldn’t want to be in anything like Yasi – or Larry. We were ready to get out if it looked like changing direction.
      However, how did you measure the wind speed? The ground assessment by JCU engineers said 220km/hr, and there were no other anemometers in the area unfortunately.

  7. Mission Beach Survivor Says:

    Damage from Yasi was far worst…….
    many 10yr old houses deroofed not just the oldies
    (google pics of Dunk Island just 4km off Mission)
    many private readings of max wind speeds recorded around 300k, but ignored.
    Mission had 220kph gusts in Larry, sustained in Yasi was same as Larry’s max +-

    People werent that scared in Larry…….however eveyone was scared sh++tless and in genuine fear that they would not make it through!!

    Fact, just about every house (25,000 odd) has sustained damage (1000 severe), lost gutterings & water entry, windows & roller doors.

    JCU engineers seem to under estimate the max & constant wind speeds, because all the houses havent failed (aka Tracy)……. so the houses designed wind speed hasn’t been exceeded…….so the winds are less than the designed wind speed….?????

    A house designed for 150kph wont fly apart at 155kph ……it may with stand 225kph+, …..some will shed bits ……& a single bad or weak tie down = no roof

    The BOM projected wind speeds have been deduced from hundreds of cyclones & hurricanes world wide where there has been comprehensive data ( working on bent signs is fine, but they say speed was 55ms L to 80ms H ) so 220 L kph could be 300kph H …..These JCU boffins need to put up wind instruments in the path of cyclones

    IF Yasi had hit Cairns or Townsville it wold have been a disaster of New Orleans / Galvston magnitude 100’s of deaths 20,000 destroyed houses etc

  8. Mission Beach Survivor Says:
    (lots of info here)

    “At the time of writing there are no verified observations of the maximum wind gusts near the cyclone centre. However a barograph at the Tully Sugar Mill recorded a minimum pressure of 929 hPa as the eye passed over suggesting wind gusts of about 285 km/h were possible. This is supported by measurements (subject to verification) from instrumentation operated by the Queensland Government (Department of Environment and Resource Management) at Clump Point (near Mission Beach) which recorded a minimum pressure of 930hPa.”
    Wind gust can be 40% stronger than constant

    When Willis Island failed it was just entering the strongest winds 185k
    so we went from gusting 185k ……very quicly
    to at 25-30 km to eye a
    constant 200kph+ for 45m and …….
    gusting 300kph+ for 45m before the eye crossed
    then dead calm & stars ……..
    then hell again from the North

    I was watching the BOM radar on the laptop live from 11pm till the calm

  9. Mission Beach Survivor Says:

    Lots of cool detailed info here on hurricanes

  10. Predate Says:


    Great analysis,

    Just curious – Do you have the complete data for willis island i.e from jan 31. I am assuming the data above is a screenshot from the australian meterological website.


  11. dave Says:

    Doesn’t the maximum sustained windspeed occur at radius R = r*ln((Pn-Po)/(P-Po)). So wouldn’t the maximum sustained wind speed occur at approx 46,500m radius. Which is twice the radius of the maximum pressur gradient dP/dr ?

  12. dave Says:

    So did the anemometer at willis island fail at the strongest winds (around 0730 hrs) or did the winds reach at 0810 hrs when it actually failed? Cause at 0800 it was 30-35 km from center, so maximum wind would have passed around 0730 hrs.

    • kenskingdom Says:

      No, I think it failed at 8.10. If the max windspeed was at 7.30 then Yasi was nowhere near as strong, yet ground estimates of speed put max speed at 220km/hr.

  13. Gerry Says:

    My friend had a barometer and it went to it’s lowest setting which was 920 it was pure ignorance that the qld government didn’t have proper monitoring equipment at Tully or mission beach we probably won’t get any cyclone shelters either we had winifred and Larry and they never shook our huge cement building with steel reinforcement this monster was shaking it fir ten to twenty mins before the eye and ten to twenty minutes after the eye how much energy us needed to do that? Absolutely huge remember if the wind is moving at x and system is moving at x you have to add the two together to get total I think we had gusts past 400 kman hour. As I said a shame we don’t have proper equipment in Tully our simple barometers went to their lowest setting s which was 920 near my Tyson on the foothill no doubt in the future we will get a chance to measure another monster hopefully not this century though

  14. Richard Kleeman Says:

    I have come across this site a bit late but I agree with most that Yasi was not a Cat 5 cyclone and think it was only three. My reasons are the centre pressure was never low enough (big Cat 5 Hurricanes in the US have had centres below 900 mbs down to 888mbs), damage done at Innisfail, only 25 n.m. from the centre is not consistent with Cat 5. On the morning before it crossed the coast a TV interview with a bulk carrier reported 90 kilometre per hour winds and they were 100 klm from the eye. The sea conditions I saw were consistent with that speed and that is the quadrant you would expect to have the highest winds in. The more I see of the Weather Bureau the less I take notice of their forecasts as I don’t think they even know what a severe thunderstorm is. A friend who was a former meteorologist with the Bureau is of the opinion it was only Cat 3 looking at Willis Island Data and in a recent inquiry to them by me on this subject they referred it south saying it was beyond the scope of the Queensland Office. Really. I am still waiting for a reply. I have actually flown through the eye of cyclone Ted as it was transitioning from Cat 3 to Cat 4 and knowing what the winds were on the ground at Gove I think Yasi was only a 3. As far as thunderstorms go on the 9th of March they were forecasting severe storms through the Lockyer Valley. I could see from my home in Brisbane that the tops were only about 25,000′ and collapsing and their radar returns were only showing moderate returns with no cells. The aviation forecast for that time showed ocassional tops to 25,000 on the ranges with variable winds up to 10,000′, hardly the conditions for a severe thunderstorm warning with heavy hail etc. I don’t think they look outside anymore and everthing is done on modelling – that’s why they got Yasi wrong.

    • kenskingdom Says:

      Thanks for your insights Richard. Don’t hold your breath waiting for a reply from BOM. While I have no faith at all in the Bureau’s Climate department (see numerous other posts here) the local blokes do a pretty good job when they’re allowed to, can give a good summary of likely conditions daily and for 3 or 4 days ahead. A bit like the local ABC vs the corporation. It’s when they get into the business of predicting severity of weather and numerical amounts and don’t also say what the margin for error is, that the wheels fall off. The trouble is, while people like you and me have a lifetime of experience comparing forecasts with actual weather events (you more than me from aviation), many people, and the media, take forecasts literally. This is becoming dangerous. While the discussion over Yasi’s strength is no comfort to the people of Mission Beach/ Tully and possibly can be seen as insensitive, facts about such a potentially dangerous event need to be aired and faced. James Cook Uni engineers have twice reported their findings based on ground level damage that maximum speeds were around 220-240 km/hr. That is, borderline Cat3/Cat 4.
      Thanks for your input.

      • kenskingdom Says:

        Wait there’s more.
        BOM has some wonderful things on its website. I was regularly monitoring observations at every station from Cairns to Hay point as Yasi approached the coast that day, and hours before landfall it was obvious there was no connection between their cyclone advices and what the observations were showing. Winds were nowhere near as strong as they said they were going to be, or were, or had been. So as well as not looking out the window, they possibly don’t even look at their own observations. Models say it should be Cat 5, so data must be wrong.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: