WunderBlog Archive » Category 6™
Category 6 has moved! See the latest from Dr. Jeff Masters and Bob Henson here.CycloneCenter.org: Harnessing the Power of Crowd-Sourcing to Improve Hurricane Data
Today's guest post is by Dr. Jim Kossin, a hurricane scientist with NOAA's National Climatic Data Center stationed at the University of Wisconsin/NOAA Cooperative Institute for Meteorological Satellite Studies (CIMSS). I flew with Jim in 1988 with the NOAA Hurricane Hunters into Hurricane Gilbert of 1988, when it was the strongest Atlantic hurricane ever measured--888 mb. Jim was just a graduate student at the time, and has gone on since to write over 70 scientific papers, mostly on hurricane science.
- Jeff Masters
Has climate change made tropical cyclones stronger? This is a common question and it comes up even more often after a particularly devastating landfall event, such as Hurricane Katrina striking the U.S. Gulf coast in 2005 or Typhoon Haiyan moving over the Philippines in 2013. But it turns out that this is also a very difficult question to answer. The difficulty lies in the inconsistency of the available data. The data and their quality can change a lot from one time to the next and from one region of the globe to the next. This inconsistency means that we can never be sure whether a trend is real or just caused by the changes in the data.
Removing the inconsistencies is no easy job. And for data collected 50 to 100 years ago or more, there is relatively little that can be done with a high degree of confidence. Fortunately though, there is hope for improving things in the period of the modern satellite era, which began in the 1970s and really got going in the 1980s. In 1975, Vernon Dvorak, a scientist working at what is now the National Oceanic and Atmospheric Administration (NOAA) published a method that estimates tropical cyclone strength using satellite images.
At its heart, the method, known as the Dvorak Technique, is essentially a “pattern-recognition” routine. Images are assessed for certain shapes and patterns and these are then related to storm strength (Figure 1). The method has been applied all over the world for many years, but it has not been applied consistently and the satellite data has steadily improved in quality over time. The inconsistent application and progressively changing data quality creates a lot of uncertainty in any attempts at global trend analysis.
Figure 1. Satellite images of tropical cyclones of varying strength.
One thing we can do is go back and take all of the available satellite images of tropical cyclones from all of the world’s ocean basins, and “homogenize” them so that they all have about the same quality. We’ve actually already done this. Then we could recruit an expert at applying the Dvorak Technique and have him or her reanalyze all of the images from the past 30 to 40 years. Since it’s the same person doing it, there should be a high degree of consistency. When he or she is finished, we would have a more consistent global record of tropical cyclone intensity and we can see whether any trends show up. But there are about 300,000 images that need to be analyzed…assuming that it would take our expert a few minutes per image, that’s about 12 years working 8 hours a day, 5 days a week with no holidays!
A solution to this problem can be found in the concept of “crowd-sourcing”. In crowd-sourcing projects, large numbers of ordinary people attack an unwieldy problem together. The project can be a scientific one, as ours is, but the “crowd” doesn’t need to be scientists. What the individual “citizen scientist” may lack in formal training and experience can be made up for by having multiple evaluations by a large group. More evaluations means less uncertainty, and this is where the “power of the crowd” comes from. It’s been found, over and over again and in many situations, that a multitude of lay-people can provide more accurate answers than a small handful of experts.
Figure 2. The Cyclone Center home page.
The website cyclonecenter.org provides the platform for this solution (Figure 2). The Cyclone Center project began in September 2012, and since that time more than 20,000 citizen scientists have completed almost 400,000 image evaluations. But we still have a ways to go and we need your help. Please consider contributing to this ongoing and important project by visiting the Cyclone Center website and trying your hand as a citizen scientist. It’s fun to do and you can interact with other citizen scientists along the way. And you’ll know that you are contributing to finding the answer to the important question: are tropical cyclones getting stronger?
(A paper describing the Cyclone Center project and how successful it has been over the past two years has been accepted for publication in the Bulletin of the American Meteorology Society. An Early Online Release is available at http://journals.ametsoc.org/doi/abs/10.1175/BAMS-D-13-00152.1)
Jim Kossin, NOAA National Climatic Center
- Jeff Masters
Has climate change made tropical cyclones stronger? This is a common question and it comes up even more often after a particularly devastating landfall event, such as Hurricane Katrina striking the U.S. Gulf coast in 2005 or Typhoon Haiyan moving over the Philippines in 2013. But it turns out that this is also a very difficult question to answer. The difficulty lies in the inconsistency of the available data. The data and their quality can change a lot from one time to the next and from one region of the globe to the next. This inconsistency means that we can never be sure whether a trend is real or just caused by the changes in the data.
Removing the inconsistencies is no easy job. And for data collected 50 to 100 years ago or more, there is relatively little that can be done with a high degree of confidence. Fortunately though, there is hope for improving things in the period of the modern satellite era, which began in the 1970s and really got going in the 1980s. In 1975, Vernon Dvorak, a scientist working at what is now the National Oceanic and Atmospheric Administration (NOAA) published a method that estimates tropical cyclone strength using satellite images.
At its heart, the method, known as the Dvorak Technique, is essentially a “pattern-recognition” routine. Images are assessed for certain shapes and patterns and these are then related to storm strength (Figure 1). The method has been applied all over the world for many years, but it has not been applied consistently and the satellite data has steadily improved in quality over time. The inconsistent application and progressively changing data quality creates a lot of uncertainty in any attempts at global trend analysis.
Figure 1. Satellite images of tropical cyclones of varying strength.
One thing we can do is go back and take all of the available satellite images of tropical cyclones from all of the world’s ocean basins, and “homogenize” them so that they all have about the same quality. We’ve actually already done this. Then we could recruit an expert at applying the Dvorak Technique and have him or her reanalyze all of the images from the past 30 to 40 years. Since it’s the same person doing it, there should be a high degree of consistency. When he or she is finished, we would have a more consistent global record of tropical cyclone intensity and we can see whether any trends show up. But there are about 300,000 images that need to be analyzed…assuming that it would take our expert a few minutes per image, that’s about 12 years working 8 hours a day, 5 days a week with no holidays!
A solution to this problem can be found in the concept of “crowd-sourcing”. In crowd-sourcing projects, large numbers of ordinary people attack an unwieldy problem together. The project can be a scientific one, as ours is, but the “crowd” doesn’t need to be scientists. What the individual “citizen scientist” may lack in formal training and experience can be made up for by having multiple evaluations by a large group. More evaluations means less uncertainty, and this is where the “power of the crowd” comes from. It’s been found, over and over again and in many situations, that a multitude of lay-people can provide more accurate answers than a small handful of experts.
Figure 2. The Cyclone Center home page.
The website cyclonecenter.org provides the platform for this solution (Figure 2). The Cyclone Center project began in September 2012, and since that time more than 20,000 citizen scientists have completed almost 400,000 image evaluations. But we still have a ways to go and we need your help. Please consider contributing to this ongoing and important project by visiting the Cyclone Center website and trying your hand as a citizen scientist. It’s fun to do and you can interact with other citizen scientists along the way. And you’ll know that you are contributing to finding the answer to the important question: are tropical cyclones getting stronger?
(A paper describing the Cyclone Center project and how successful it has been over the past two years has been accepted for publication in the Bulletin of the American Meteorology Society. An Early Online Release is available at http://journals.ametsoc.org/doi/abs/10.1175/BAMS-D-13-00152.1)
Jim Kossin, NOAA National Climatic Center
The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.