One of the most destructive natural phenomena known are powerful tropical cyclones (better known in the US as “hurricanes” when they originate in the Atlantic or eastern Pacific Oceans or in nearby seas). Their powerful winds, rain and associated storm surges can cause huge losses of life and property. While there have been great improvements made in recent decades in our ability to track and predict these destructive storms largely as a result of advances made in satellite remote sensing technology, there is still much room for improvement especially when these storms occur outside of the range of American hurricane-chasing aircraft whose patrols are limited to the Atlantic and northeastern Pacific Oceans.
Limitations of Today’s Methods
Today, the empirically-based Dvorak technique and its subsequent modifications are generally used to categorize the intensity of tropical cyclones in the absence of observations from surface ships or instrumented ocean buoys. Named after American meteorologist Vernon Dvorak who developed and refined the method between 1969 and 1984, the Dvorak technique relies on data from constellations of meteorological satellites in low polar or geosynchronous orbit to estimate the strength of storms based on such attributes as their appearance, temperature and the apparent motion of cloud features over time. But the modified Dvorak technique can experience errors of at least 20% in estimating maximum sustained surface winds. A dangerous Category 3 tropical cyclone could be miscategorized as a weaker Category 1 storm or vice versa.
Image of Typhoon Haiyan taken by NASA;s Earth Observatory as it approached the Philippines on the left. (NASA)
One example of this was Typhoon Haiyan which hit the Philippines on November 8, 2013. It was not recognized that this typhoon was still intensifying as it approached landfall and this resulted in an estimated death toll of about 10,000 partly because the appropriate evacuation orders were never issued. A more recent example was last month’s Hurricane Amanda which unexpectedly intensified but fortunately caused no loss of life or property since it remained far out at sea. A counterexample of a storm that never intensified as expected was tropical Cyclone Phailin which made landfall near Gopalupur, India on October 12, 2013. Expected to intensify to a dangerous Category 5 storm, the government of India evacuated 800,000 people along the vulnerable coast at great expense only to have the storm turn out to be a a much less dangerous low-end Category 4 storm. Obviously better methods are needed to help predict the strength of tropical cyclones.
CyMISS (Cyclone Intensity Measurements from the ISS)
Professor K. Emanuel of the Massachusetts Institute of Technology (MIT) developed the currently accepted theory that the thermodynamics of tropical cyclones are the equivalent of a Carnot engine. The Carnot cycle operates between the warm ocean and the cold stratosphere to power these storms’ powerful surface winds and intense rainfall. Emanuel’s model was subsequently extended with input from Professor Paul C. Joss of MIT and Visidyne, Inc. in Burlington, Massachusetts in the late-1990s as part of an experiment of the RAMOS (Russian-American Observation Satellites) cooperative program with Russia.
Diagram demonstrating how the Carnot cycle powers cyclones. By accurately measuring the altitude and temperature of the clouds of the storm’s eye wall, the strength of the storm can be determined. Click on image to enlarge. (Visidyne)
Cancelled in 2004, the RAMOS program was to use a pair of Russian-built satellites carrying an American and Russian-built instrument suite to make stereo observations of clouds and other atmospheric phenomena at wavelengths ranging from the infrared to the ultraviolet with a nominal 100-meter pixel footprint (see this site’s RAMOS Page for more information). Dr. Joss and other members of the American science team in the RAMOS program had worked out an improved method of estimating the strength of tropical cyclones based on the Carnot engine model using information about the temperature of the cloud tops around the storm’s eye wall (derived from infrared measurements from the satellites) and the altitudes of those clouds (derived from a three dimensional reconstruction using stereo data from the pair of satellites). With cloud top temperature accuracy of ±1° K and altitude determinations of about ±100 meters, the observations by RAMOS coupled with this new Carnot engine model promised to provide estimates of tropical cyclone strength superior to those based on the Dvorak techniques using data from conventional meteorological satellites.
Although the RAMOS program was cancelled, work has continued on this improved method at Visidyne with the hope that we would one day have an opportunity to implement it. During this past year, a team from Visidyne working in cooperation with engineers at Utah State University’s Space Dynamics Laboratory in Logan, Utah (the former prime contractor of the RAMOS program) has developed the proposed CyMISS (Cyclone Intensity Measurements from the ISS) project. Selected by CASIS (Center for the Advancement of Science in Space which manages the US National Laboratory on the International Space Station), we are currently developing a relatively inexpensive demonstration of our technique for an initial daytime-only implementation called CyMISS-D (with “D” for “daytime”). Astronauts on the ISS are currently gathering images of tropical cyclones with hand-held cameras to support development of this project.
Example of a 3D reconstruction of a thunderstorm using pseudo-stereo. Click on image to enlarge. (NASA & A.J. LePage/Visidyne)
CyMISS-D will use a camera operating at visible wavelengths with precisely known pointing mounted on the ISS to obtain a sequence of images of tropical cyclones as the space station passes overhead. Building on earlier work done as part of the RAMOS program, these images will then be analyzed using a new pseudo-stereo technique developed by this author at Visidyne to determine the altitude of the clouds associated with the eye of a tropical cyclone. Instead of using a pair of satellites to provide stereo information about a scene, a single platform is used in the pseudo-stereo technique with the orbital motion providing the baseline required for stereo.
The accuracy of earlier attempts to derive cloud altitudes using pseudo-stereo analysis of satellite images have been limited because of the motion of the clouds being measured – a problem especially severe in powerful cyclonic storms. In order to get the required altitude accuracy of 100 to 200 meters, information from the analysis of a sequence of images acquired by the CyMISS-D camera on ISS will be combined with information on cloud motions derived from a sequence of high frame-rate, visible-band mesoscale images acquired by the ABI (Advanced Baseline Imager) to be carried by the new GOES-R geosynchronous meteorological satellite slated for launch in 2015. For the CyMISS-D demonstrations, infrared data from ABI will also be used to derive cloud top temperatures.
The ABI (Advanced Baseline Imager) to be carried by the new GOES-R satellite. Images from this instrument will be combined with image sequences from CyMISS-D in low Earth orbit to accurately determine the strength of tropical cyclones. (NOAA/NASA)
The uncertainty in the sea level pressure at the center of the storm is expected to be ±5 hPa with the uncertainty in the surface wind speed of ±20 knots – a significant improvement over the existing Dvorak-based techniques. The addition of an infrared imager to our ISS-based instrument package in the future would allow CyMISS to derive its own cloud top temperatures at higher spatial and temporal resolution independently of ABI as well as allow the system to make nighttime observations.
After the concept has been demonstrated on ISS, we currently envision deploying a constellation of four or more Cubesats (or equivalent miniaturized satellites) that will allow storms to be observed every two to five hours. Information derived from CyMISS could then be incorporated into storm forecast models which are updated every six hours and promise to greatly improve forecasting of the strongest tropical cyclones.
General References
P.C. Joss, A.T. Stair, J.G. DeVore and G.E. Bingham, “A New Method for Accurate Remote Measurement of Tropical Cyclone Intensities”, Poster at AGU Science Policy Conference (Washington, DC; June 16-18, 2014), Poster # NH-12, 2014 [Poster]
“The Cyclone Intensity Measurements from the ISS (CyMISS) (TROPICAL CYCLONE)”, NASA ISS Web Site, [Link]
Related Reading
“RAMOS: The Russian-American Observation Satellites”, Drew Ex Machina, June 21, 2014 [Post]
“A Little Piece of the ISS”, Drew Ex Machina, April 5, 2014 [Post]