Crowd Sourcing Tropical Systems Using the Dvorak Technique

CycloneCenter uses the Dvorak Technique to classify systems.  For more information, please visit:

As the first blog post of 2013, I wanted to share a link I recently discovered that incorporates trends in data, crowd sourcing, weather and science.  Regardless of your experience, age or background, this post is for you – you can have a part in determining the strength of a tropical system.

Have you ever wondered how scientists classify tropical systems?  Usually that involves analyzing numerous observations from land or ocean systems, flying aircraft into systems, taking measurements  through upper air balloons, etc.  However one tool that they use is a technique known as the Dvorak Technique, named after Vernon Dvorak.  Dvorak discovered that the intensity of a tropical cyclone could be skillfully approximated by the cloud patterns on a single satellite image and he developed and improved his method (now called the “Dvorak Technique”) in the 1970s and early 1980s.

The technique consists of a set of 10 steps, which can be simplified to produce the answers to four important questions:

  1. Where is the center of the system?
  2. What type of cloud pattern best describes the system?
  3. How organized or intense is the cloud pattern?
  4. Does the system look stronger or weaker than 24 hours ago?

CycloneCenter is a web-based interface that enables the public to help analyze the intensities of past tropical cyclones around the globe.  Each user is asked to compare satellite images to determine which is stronger or weaker.  Interested volunteers will be shown one of nearly 300,000 satellite images. They will answer questions about that image as part of a simplified technique for estimating the maximum surface wind speed of tropical cyclones.

Here are just a few items from the frequently asked questions – available on their blog site:

This is hard.  What’s the right answer?

There is no “right” answer, so go with your gut!  This is by far the most common thing we hear, and you’re probably doing a much better job than you think you are.  Remember, the technique is very subjective. Two experts can look at the exact same image and come to different conclusions.  One of the main reasons we’re doing this is to get a wider range of opinions.  Your opinion is just as valid as anyone else’s, so just give us your best guess.  

If there’s no “right” answer, then what’s the point?

If two experts disagree about a storm, then we don’t know which one to trust. However, Cyclone Center has up to 30 volunteers look at each image, so essentially we are taking a vote on the cyclone classification. This statistical approach has been proven to work well in other citizen science projects, and though it’s never been tried with cyclones before, we are confident that put together, your classifications will be as good as, possibly better than that of one expert.

Hasn’t an expert already looked at these storms?

In most cases, yes. But the experts don’t always agree. In fact, sometimes there are big differences among the experts. That’s why we need all of you to help us narrow it down. We’ll take all of your answers, compare them with the experts, and hopefully come up with an even better estimate of how strong these storms were.

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).  There is more analysis to be done, but it’s clear that your classifications are matching what we expected!  If you’d like, you can read the full analysis on the blog here.

Cyclone Center

– CycloneCenter Blog


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