A bizarre example of an image on Cyclone Center that is very misleading is the storm CYC1981.  The small white circle in the very center of the storm that at first glance looks like an eye is actually the island Niue in the South Pacific.  The satellite image was taken as the hurricane passed over the island, and the white land boundary lines look like the eyewall.

CYC1981

Unfortunately, with the size of the images on Cyclone Center, it is hard to determine if this storm has an eye, and since having an island in the middle of a storm image is a rare phenomenon, it would be easy to assume that it did.  This image of CYC1981 even stumped most of our science team, so don’t feel bad if it tricked you too.

So how can we tell if a storm is or is not an eye storm?

Many storms look very similar in size and shape to eye hurricanes, but they lack an actual eye.  There are a few things that you can look for, however, to determine for sure whether or not the image you are looking at is an image of an eye storm.

1. Is the center of the storm surrounding the eye cold?  You can tell this by the color of the clouds—shades of red, orange, and grey signify warm clouds while blue and white areas represent cold clouds.
2. Is the eye itself warm?  The eye should be made up of warm clouds, usually grey or pink colored.  White and grey clouds are not one and the same; white clouds are very cold and grey clouds are very warm.

If we applied these three steps to CYC1981, we would find that it does have cold clouds at the center, but there are no warm clouds around where the eye should be and the band of clouds around the storm is very weak.

For more information, visit another recent post: How do I classify this? False eyes.

This post was contributed by Brady Blackburn, an intern with the Cyclone Center team from Asheville High School in Asheville, NC.

This is the strongest part of the storm. Part 4 will conclude the diary and summarize some information from the damage report.

“1605K: Now getting gusts to 100 m.p.h. Everything still holding fairly well. Visibility is less than one eighth of a mile and rain is falling at a rate of about 4 hundredths of an inch per hour. Emergency lights still not working. Believe the starter is out. Pressure still falling. The clouds still the same with a few breaks in the lower stratus with overcast solid above. Ceiling is now about 50 feet.

“1620K: Put in for a call to Clark field and was greeted by a call from Iwo Jima. They want to close down operations there but advised that they stay in operation for at least 24 hours in case of emergency at this station. Gave them the latest information on the storm and our condition. The wind is estimated at 90 knots now and the visibility is almost zero. The ceiling is estimated at 50 feet in precipitation.

“1632K: Corrugated roofing just started tearing off the roof of the terminal, it made a tremendous crash when the section tore off. All of the troops that are using the terminal as a typhoon shelter had looks of apprehension and I can’t say that I blame them. Indications are that the barometer is starting to level out. The crash of sheet metal on the terminal wasn’t the terminal roof apparently but the remains of the freight terminal on the east side of us. Capt. Highley said that it completely collapsed.Visibility has lifted from zero to 3/4 of a mile but the wind still has 110 knot gusts.

“1650K: Report just received from the Rawin crew, who had just left their hut, and they say that all the buildings are going and that sheet metal is flying thick through the air. Looks like the center has passed us as we had a 1 millibar rise in [pressure in] the last 15 minutes, although will wait for one more observation before we draw any definite conclusions. Winds still haven’t decreased although they are beginning to vear into the SE. Maybe its past us.”

Part 4 will contain the conclusion of this diary and a summary of the damage report.

What is rawin? What is the hydrogen shack? What is Antrac? What is a T-6? Well, I’m learning as I transcribe this, too. I’ve provided links to other sites that provide more information, but some things are still a mystery to me. For instance, Antrac appears to be related to air traffic control (that is, what I’ve learned from web searching). But I could be wrong.

Page from Consolidated Report on 1949 Typhoon Allyn

The following is an hour of notes from Typhoon Allyn.

“1437K: Just went outside to check on the T-6′s at the east end of the ramp. They are not in sight. Don’t know whether they have been moved or whether they are out of sight on the low end of the ramp. It is very difficult to stand erect in the open. Andersen AFB is estimating 85 knots. No buildings have blown down but trees are starting to go over. The pole for the Antrac antenna is swaying about six feet at the top. One pole has gone down at the Rawin shack. The banana trees have been uprooted. Rawin reports that the T-6′s were still on the ramp the last time they went by.

“1445K: The corrugated roof on the inflight kitchen quonset is starting to tear off. One section is gone completely and another is starting to rip off. Estimate that the Antrac pole will stay up another hour or hour and a half.

“1455K: Capt. Myers just returned from quarters area, he said that the large hangar doors just went through th PLM hangar as he came by. Heavy rain just began to fall and the visibility has been reduced to less than one eighth of a mile and the wind is now hanging at about ninety knots. The rain is in sheets and horizontal with the ramp. Debris is beginning to fly and the inflight kitchen is still holding its own. The wind tee is just about demolished.

“1510K: The station is being used as a fire control center for this area. The fire marshall has men stationed and is standing by himself for fires that are reprted into this station. The observer will not be able to take psychrometer readings much longer as it is impossible to stand against the wind.

“1515K: Numerous power lines have blown down, and island power is due to go off at anytime. We can’t get the emergency power unit to work.

“1520K: The lights just went out. No emergency power as yet, but repair crews on the way. Estimating the winds at 80 knots and the visibility at three eighths of a mile in torrential rains.

“1530: Pressure is dropping on an average of four millibars per hour and the tendency is still down. The bottom is really falling out of this. Sheet metal is now tearing off of the depot hangars and the Rawin shack is OK; however, the hydrogen shack is going. The doors have been blown off and it shouldn’t last much longer. Debris is traveling along the ramp at about twenty miles per hour.

Before we continue the narrative, one should have an understanding of the time conventions in these posts. The U.S. military uses time zones named by letter. The “Z” time zone is the same as the UTC, the Universal Time Coordinate, which is the time in …. Guam – in the Western Pacific – is in the “K” time zone. So times listed in the report as “K” refer to local time. For instance, 17:00K is 5 p.m. local time. To learn which letter your time zone corresponds to, visit this web page describing time zones. Keep in mind that the following spans a three hour period.

1949 Typhoon Allyn’s path through the Pacific

Page 4: “Running diary of approach of Typhoon Allyn:

“Arrived at Harmon at 1145K on the 17th and set up shop. Things proceeding normally with winds increasing slightly. The nine light indicator at this station won’t last too long from the indications. It is warbling loudly and the count of the flashes at this time is impossible.

“1315K: Wind has now picked up considerably and the reports from North indicate that they are abandoning the station as the radar tower is about to topple over. We are estimating the winds at this time and cross checking with the Navy. The wind at Harmon is now tearing the inflight kitchen apart. The pressure is falling ominously. We are taking readings every 15 minutes and estimating the wind every 5 min. Also, setting up to take readings with the sling psychrometer and trying to keep an accurate count of all the happenings as they occur.

“1322K: A piece of wiod just went down the ramp must have been a 6×6 and 10 feet long.

“1325K: The rawin tower is no longer visible. Winds must be in the vicinity of 70 knots [80 mph] at this time.

“1336K: [A] communication man just came in and said that JMP is going off the air. They are evacuating the station. We are now out of contact with the rest of the Pacific area.

“1355K: Winds are really beginning to blow now. It is whistling thru all the wires, and the poles next to the stations have quite a list to them. Estimating the winds at 60 knots at this time. Wind tee is hanging on by shreds. Communications are out as they have evacuated both the Weather Central and Guam Broadcast stations, so we are out of contact with everyone but Navy. Rain is coming down in sheets and visibility is getting bad. Visibility at 1/2 mile at this time.

“1420K: Low scud 200 feet [high covering] about eight tenths [of the sky] and [completely overcast] stratocumulus clouds just above, moving beter than 50 knots. Antenna pole at the edge of the building now has a 6 foot oscillation.

1989 Typhoon Gay

Time series of the winds associated with Typhoon Gay

This system was interesting in that it is a system that began in the Gulf of Thailand – considered the Pacific Ocean – then moved west into the Indian Ocean, eventually making landfall in India as – potentially – a very strong cyclone. Of course I must qualify that statement because of the differences in the best track data. The graph at right shows the best estimates of the storm’s intensity, in maximum sustained wind speed. The system gained strength near day 2, then crossed into the Bay of Bengal and regained strength. At landfall in India, it was likely between 70 and 150 knots, kind of a large range. Some of these differences in intensity are due to the data available to each agency. Another could be in the interpretati0n of the imagery.

Typhoon Gay during development

A recent talk post from ibclc2 noted the features in Typhoon Gay during its development in the Gulf of Thailand. The organization of the system is beginning to take shape. The convective cell  near B is close to a banding feature (if you were doing a detailed classification). But it is not, since the region between it and the central part of the system is not warm enough (it needs to be red or warmer, see the field guide for more information). The portion near A appears to be an embedded center. But upon further review, there appears a warm spot just north of the darkest blue colors. It could be the beginnings of an eye, but only time will tell … and it does. In the next few images, that small warm spot becomes an eye just prior to making landfall on the Kra Isthmus. So how would you classify it?  Well it’s likely best left as an embedded center with no banding. While there is the hint of an eye, the primary characteristics of an eye (cold cloud surrounding a warm center in a circular fashion) aren’t complete yet.

2005 Hurricane Dennis

Hurricane Dennis in the Caribbean

An image of Dennis recently noted by bretarn showed a large system. Similar to Gay above, the satellite image showed a cold center (A) with a large cold band to the east. The convection near A is showing some circulation, so the center is somewhere below that cold cloud cover. So it is an embedded center. Like the Typhoon above, this is an image just prior to an eye emerging. The next question is what to do with B. It is definitely associated with the system, because it appears to be wrapping around the circulation center near A. The region between A and B is warm, with the warmest color being red. So for a detailed classification, this might be considered a banding feature.

In its own right, Dennis was a very severe system, making landfall in Cuba and in the Gulf on the Florida panhandle. However, its fast movement lessened the impact. It is also less memorable because its Gulf landfall was eclipsed by Hurricane Katrina later in the season. Nonetheless, the name was retired from the North Atlantic hurricane names after the season.

Damage from 1949 Typhoon Allyn

One typhoon from 1949 – Typhoon Allyn – passed quite close to Guam. The forecasters there included in their “Consolidated Report: Typhoon ‘Allyn’ November 14-24, 1949″ a rare addition: a running diary of the typhoon’s approach to Guam. Over the next few days, I’ll be reproducing the diary here for those interested in a first hand description of the landfall conditions. While this doesn’t contain video feeds, or interviews with local residents, it provides insight into the strength of these violent storms and those who had to work through such conditions. This reproduction of that report is dedicated to those who provided forecasts to our armed services and helped protect lives and property during those years of service at Guam and other forecast offices in the Pacific.

Page 4: “During the morning of 17 November, the wind strengthened to 30 knots [34 m.p.h.], and the dark sky and falling barometer gave sure indications of the advent of the storm. Arrangements had been made for Clark [another Air Force Base] to accept forecast responsibility. Also, Haneda had been instructed to perform the functions of the Typhoon Warning Center, for it appeared certain that Guam’s contact with the [outside world] was to be terminated abruptly and for an indefinite period. Acting on instructions to retain responsibility as long as possible, the Center almost overplayed its hand. The plan called for the Center to issue bulletin 13 [bulletins are their term for a warning] by 17:30Z. Clark was to issue bulletin 14; however 15 minutes before the bulletin 13 was ready for transmission, the communication station ceased operations. It was only by special arrangement that this final message was transmitted. By 04:00Z, 17 November, the surface wind was near 50 knots [57 m.p.h.] making the weather station an unhealthy place due to the proximity of a 40 foot radar tower; consequently, all personnel evacuated to typhoon shelters.

“Due to the fact that the weather station at Harmon Air Force Base was located in a typhoon proof structure, it was manned throughout the storm. The following remarks have been taken from a log kept by forecasters who operated that station, beginning 17 November at 11:45K (01:45Z).”

The remainder of this diary chronicles the ensuing 12 hours of the typhoon’s interaction with Guam.

To be continued…

A false eye is a circular feature of warm cloud that at first glance appears to be a genuine tropical cyclone eye (the center of a powerful tropical cyclone).  Since we cannot look at other times during the process to see if the feature persists,  we must look for other clues to determine if the feature really is an eye or not.  The primary thing to look for is the storm structure outside of the suspicious eye.  Does the storm look well organized?  Are there distinct and tightly wound spiral bands?  Are cloud tops very cold or not so much?  Consider the following examples, all examined and discussed in the Cyclone Center Talk feature.

Odette (1985)

 The black circle indicates where an eye could possibly be analyzed.  But look at the cloud patterns outside of the “eye” for confirmation.  Here we see no organized spirals and no circular eyewall (the cold ring the typically surrounds the eye).  The clouds are certainly very cold, which is sometimes an indication of strength; but the overall lack of organization leads me to conclude that the “eye” feature is actually just a gap in the cold clouds and not really an eye at all.  I would probably classify this as a weak spiral band type pattern, but nothing more.

Ami (2003)

The second example is from a very complicated cloud pattern, typically seen in what meteorologists call the “monsoon trough” region.  This is an area where the ocean waters are very warm and atmospheric winds tend to come together in the lower atmosphere, creating a situation that is quite favorable for thunderstorms and sometimes tropical cyclones.

The black circle again indicates a circular area of warm clouds that may be mistaken for an eye.  What I immediately notice is that there are two distinct areas of thunderstorms, labeled “1″ and “2″.  Area 1 is showing some signs of organization, shown by the black lines, which indicate a turning or spiraling of the clouds.  Little organization is seen in area 2, which is essentially a large blob of thunderstorms at this point.  The eye in the middle is actually just a gap in between the 2 systems – there is no organization in clouds around this area.

I classified area 1 as a spiral band pattern.  The center of area 1 is probably very close to the circled area (follow the black lines in).  Since we are only classifying one system at a time in Cyclone Center, I ignored area 2.

Keith (1997)

To contrast the two examples above, lets look at a real eye.  Keith was a very strong tropical cyclone that exhibited a well pronounced eye feature.

At first glance we immediately notice the features of an eye pattern storm: distinct spiral band features, high degree of symmetry, and cold/circular clouds completely surrounding the eye.  Although there are even better examples of eye storms, I would classify this image as a mid-level eye pattern.  The storm intensity is probably in the Category 2 to Category 3 range on the Saffir-Simpson scale.

I hoped that this helps you to become a better Cyclone Center classifier.  Look for more help articles like this on a more regular basis throughout the next few months.

- Chris Hennon is part of the Cyclone Center Science Team and Associate Professor of Atmospheric Sciences at the University of North Carolina at Asheville

Last week members of the CycloneCenter science team attended the meeting of the American Association for the Advancement of Science (AAAS) in Boston, Massachusetts. There we got to engage with members of the science community, as well as the media. As a result, CycloneCenter has appeared in many different articles on websites, including Discovery News, Seattle Times, and even the French Embassy! It’s great to see that the media is focusing on citizen science. Without you, our efforts would not be possible!

In addition, we have recently passed the 150,000 classification mark! We once again thank you for your participation, and want to remind you to keep on classifying!

Speaking of which, don’t forget that we have our own Twitter account, as well as a Facebook page. Make sure to like / follow us, as well as tell your friends and colleagues!

Since our launch, we have received a lot of feedback and questions about the project, through members posting on our talk page, as well as comments on this blog. Thank you very much and keep them coming! We have put together a FAQ that we hope will answer many of your burning questions:

FAQ’s:

How do I pick the right match?

Focus on the shapes of the clouds and their colors. Size and orientation don’t really matter. The key is remembering that tighter spirals and colder colors are signs of stronger storms. It’s a little subjective, and that’s why we’re doing this. Your opinion matters, so give us your best guess.  

If it’s subjective, then what’s the point?

When two opinions about a storm are different, we need to determine which one is best. Cyclone Center has up to 30 volunteers look at each image, and then we’ll combine them statistically. This approach has worked well in other citizen science projects. It’s never been tried with tropical cyclones before, but we are confident that your classifications will help us learn more about these storms.

Hasn’t an expert already looked at these storms?

Yes, but the methods have changed over time and between regions. We have nearly 300,000 satellite images of tropical cyclones for the last 32 years from around the world. You’re helping us classify all of them with one method.  You’re also providing us with valuable new information about the shape of the storm, like whether or not it had an eye. It would take a team of experts years to finish this task, so we’re relying on you to help us do it much faster.

Can a regular person like me really help?

Yes!!  We even have some preliminary data that tells us how great you’re all doing.  The two graphs here show the current estimate of intensity for a particular storm (top), and the estimate that we’ve gotten from your classifications (bottom).  We are just starting this project, so there is more analysis to be done. But it’s clear that your classifications are matching what we expected!  You can find more information on our blog here.

What’s a Cyclone?

The dictionary says a cyclone is a storm that rotates around a center of low atmospheric pressure. That can include tornadoes, winter storms, and the “tropical cyclones” that we’re looking at here. Tropical cyclones are organized systems of thunderstorms that get their energy from warm tropical oceans. They’re often called Hurricanes in the Western Hemisphere, Typhoons in the western North Pacific, or just Cyclones in the rest of the world, but they all mean the same thing.

This doesn’t look like a hurricane. Where’s the eye?

When you think of tropical cyclones, you probably picture the classic circular storm with a clear eye in the middle.  However, eyes only happen in stronger storms, like the ones that make the news. We think about 15% of our images will have those picturesque eyes.  Most of the others will be less organized and harder to classify. In fact, because they’re more difficult, those weaker ones are where we need the most help and you can make the biggest difference!

Are there only these six images for this storm?

Probably not.  Because the images are collected every three hours, and most storms last several days, a storm may have several hundred images that cover its whole life.  In fact, Tropical Cyclone Hondo lasted nearly all of February 2008 and has over 600 images!

You are seeing six images taken from some period within the storm’s life. So in most cases, you’ll see a portion of the storm’s life cycle, but not the entire thing.  

What do I pick when it doesn’t look like anything?

If it looks like there is no organized storm in the image, try looking through the curved band options.  Curved bands are generally the weakest of the storm types, and often appear very disorganized.  However, from time to time there is an error in the satellite location, and there might truly be nothing in the image.  If that’s the case, choose “No Storm” under the “Other” option.

I made a mistake!  How do I go back?!

If you’ve already finished the image, don’t worry! There’s more than one way to classify most of these images. By averaging all your answers together, we’ll still get a pretty good estimate.

If you haven’t completed the classification, you can always go back to the beginning and start that image over.  In the bottom right corner, next to the Detailed Classification checkbox, you should see a small orange arrow.  Clicking it will take you back to the beginning of that image.

I found a neat storm. Can I talk about it or save it for later?

After you’ve finished a set of images, you’ll be shown all 6 of them side by side, and each one will have a “Discuss” button beneath it.  Simply click on that button, and you’ll be taken to our Talk page, where you can post your comments about that image.

Alternatively, you can always go to talk.cyclonecenter.org and see what other people are discussing, or start your own conversation about anything that interests you!

If you’d like to save an image for later, just click the heart next to the “Discuss” button, and the image will be saved to your favorites.  

To get to your favorite images, click the “Profile” button at the top of the page.  You’ll be shown how many images you’ve classified, and here is where you will see the list of all of your favorite images.  From there, you can quickly get to the “Talk” page, where you can discuss that image.

Is there any other way I can help?

Yes!  When it comes to citizen science, strength in numbers is our greatest asset, so recruit some friends!  You can like us on Facebook at facebook.com/cyclonecenter or follow us on Twitter at @CycloneCenter. The more classifications we get, the better the science will be in the end!