Meteo 101

Meteo 241

Meteo 361

Meteo 410


Page 1




by Philip Lutzak September 2006


Figure 1. Satellite image of Hurricane Rita at 12:10PM EDT (1610 UTC) on 09-21-2005. Rita was just hitting category 5 status at this time and was about 3 days away from landfall at the Louisiana-Texas border. Florida and Cuba are on the right, with the Yucatan Peninsula visible at bottom and Louisiana at upper left. Courtesy MODIS TERRA.


  If you lived along the upper Gulf Coast in September of 2005, the spectacular satellite image in figure 1 above inspired two feelings: awe and fear. Not long after the infamous Hurricane Katrina, Rita was bound to cause some frayed nerves along the Gulf coast. But what unfolded was in certain ways even more unnerving, because we were about to learn that when people pay attention to an evacuation order, it can cause as much trouble as when they don't. It certainly proved that vulnerable cities still need to learn a lot more, and quickly, about how to properly conduct an evacuation. Over 3 million people evacuated ahead of the storm, a U.S. record. Figure 2 below shows one of the many photos of that awful exodus in Texas as terrified people tried to flee the storm. The story one man recorded of his own personal evacuation study is also very telling about just how tortuous it can be to get out of the way of a hurricane.

  In this report I will examine some of the most important observation tools that hurricane forecasters now have, allowing them to much more accurately measure the strength of these storms. This, in turn, has made forecasters much more believable when alerting public officials, and the public, to the seriousness of any looming hurricane threat. I will concentrate on a few hours of time during the early morning and midday of September 21st, 2005, showing how meteorologists were able to measure Rita's rapid intensification from a category 3 to a category 5 so well. Together with the terrifying memories of Katrina, it sparked a mass exodus on the Texas and Louisiana Gulf coasts.

Figure 2. Residents fleeing Houston, Texas by car as Hurricane Rita approached. Courtesy Houston TranStar.

Figure 3. Extensive damage at Cameron, Louisiana from Rita. Courtesy AP/David J. Phillip & sfgate.com.


  The photo in figure 3 above at Cameron, Louisiana, where over 80% of the city was destroyed, is typical of the severe damage from Rita. It killed 7 people directly, and 113 more through indirect means, causing $10 billion dollars in damage, much of it due to extreme flooding throughout a large portion of southwest Louisiana and southeastern Texas. The storm surge of 15 feet left huge areas under water for weeks, and hampered the ability of the Army Corps of Engineers to empty out areas of New Orleans flooded by Katrina and repair the levees. Over 600,000 homes were damaged or destroyed. The shrimp and cattle industry were severely damaged, and the beautiful Sabine National Wildlife Refuge is still so cluttered with storm debris that it resembles a garbage dump 1 year later (see environmental reference below).




  Hurricanes are really denizens of the sea; they are born there, thrive there, and quickly die out when they move over land. This is because they get their energy from the water vapor directly above the warm tropical oceans that they move over. The warmer the ocean temperature, the more likely they will develop, and the more likely they will maintain and/or increase their intensity. But that is only one of the factors that cause tropical cyclones to develop. To really understand the processes that form these storms, here is an excellent primer on how tropical cyclones form and intensify

  Here is the official track map for Rita from the National Hurricane Center. It started as a tropical depression near the Bahama Islands with winds of 35mph, becoming a hurricane as it moved between the Florida Keys and Cuba on September 20th. From this point, things began to change. It intensified a little more rapidly than expected on the 20th of September, and then the bottom dropped out. During the day on September 21st, 2005, Rita went from a category 3 hurricane with 115 mph winds shortly after midnight to a category 5 with 175 mph winds by the following midnight. Note the two arrows on the hurricane track, marking the central pressure at important moments. In general, the lower the central pressure is, the higher the wind speeds will be. The arrow pointing to near September 22nd shows a reading of 895 mb (millibars). That was the point where Rita reached maximum strength. Below in figure 4 is a chart of the Saffir-Simpson rating scale for hurricanes. We can see that at the time of the September 22nd reading, Rita was a very powerful category 5. As a matter of fact, at that time it became the strongest hurricane ever recorded in the Gulf of Mexico, and the fourth most intense hurricane ever observed in the Atlantic basin. By landfall on the 24th it had lost intensity to a category 3 with a pressure of 937mb, but of course that is a major hurricane capable of producing extensive damage. Here is the National Hurricane Center's (NHC) detailed description of what to expect in storm damage by category.


Maximum Sustained

 Wind Speed (mph) 

Minimum Surface Pressure (mb)

Storm Surge (ft)

 Damage Potential


74-95 mph

980 or >























Figure 4. The Saffir-Simpson hurricane rating scale. Courtesy NHC (National Hurricane Center).

  As mentioned before, the meteorologists who specialize in hurricane forecasting have developed an impressive amount of tools to follow these storms. The following sections show how far meteorologists have come in their ability to collect critical data from these life-threatening storms and produce better forecasts.




THE HURRICANE HUNTERS  While millions were in flight from Hurricane Rita, others were actually flying into it. These are the Hurricane Hunters, the famous branch of airborne hurricane investigators. Using military and NOAA aircraft, they provide meteorologists with critical data from inside and outside of these tropical storms. Among the most important equipment they use is the dropwindsonde, also called dropsonde, which measures air temperature, pressure, humidity and of course, wind speed & direction. These dropsondes are quite small, and encased in a small tube, are dropped from a chute in the plane's floor, taking constant measurements as they fall toward the surface. Together with on-board measurements, these sensors provide invaluable transmissions of data that hurricane meteorologists need to assess what's going on in the eye of the hurricane.    



THE VORTEX DATA MESSAGE  is an encrypted data message sent by the reconnaissance aircraft describing the current weather conditions inside and around the eye of the hurricane.

  During the last flight into Hurricane Rita on the late evening of  September 20th, flight level winds had been measured at 103 knots (118 mph). NHC meteorologists modify this number by subtracting a certain percent (usually 10%) in order to calculate a surface wind. They then combine that calculation with other data such as satellite images to come up with the best estimate of actual surface winds. Their calculations at this time made it a category 3 hurricane. But during the morning of the 21st, further satellite observations showed it was in the midst of a rapid intensification. The next mission went out that afternoon, with the first vortex data message coming in at 1:53 PM EDT proving that Rita was a category 5 hurricane.

  To follow how this message went from raw data to information we can all understand, please examine this page illustrating the decoded Vortex Data Message from Hurricane Rita.






Figure 5. Photo taken inside the eye of Hurricane Rita. Courtesy of  the Hurricane Hunters.

  Although there were only 2 observed category 5 hurricanes in the entire decade of the 90's, now there were 2 in one year. The hurricane continued to strengthen, bottoming out at a central pressure of 895mb shortly after midnight of the 22nd, qualifying it as the strongest hurricane ever recorded in the Gulf of Mexico. One can understand why some people in its path had started to panic. A flight into a category 5 hurricane was about to start a massive flight away from it.




  While we are all aware of weather satellites in space around the world that watch the weather for us, I wonder how many of us know just how much these satellites can "see" now. The amount of data transmitted to earth from satellites has exploded over the last decade. Besides the pictures of clouds we're all familiar with, they are now transmitting huge amounts of data about the earth and its atmosphere back to us by either passively reading different wavelengths of radiation below them, or actively beaming different wavelengths of radiation downward and waiting for the "back-scattering" of this radiation that comes back to them, to find out critical properties of the atmosphere, land or ocean. All of this has helped us make tremendous strides in weather forecasting, and none more than in hurricane research. In this section I will explore a few of the many critical functions that satellites now perform in helping us keep an eye on hurricanes.  



  As mentioned in the previous section, satellites can actually beam down microwaves (short wavelength and high frequency) and wait for the scattering back of these waves to interpret the properties of entities below them. They can detect water vapor, oxygen content, precipitation, sea ice, and even wave heights, to name but a few. This process is called scatterometry, and it has revolutionized the collection of weather data. One very valuable application of this is the monitoring of wave motion on the sea surface below them. Although it is a complicated process, basically the scatterometer mounted on NOAA satellites transmits microwaves down to the sea surface below and then reads the radiation that is scattered back from ocean waves. It can then interpret the motion of these waves to calculate the wind speed. One big advantage of this is that the scatterometer "sees" through clouds, so they are not a hindrance to finding out what the winds are doing under a thick deck of clouds. One drawback is that heavy rains may attenuate, or confuse, the signal coming back to the satellite, so that scatterometry often cannot tell us for certain the wind speed in the parts of a storm system where very heavy precipitation is falling. Since the heaviest rain falls near the center of the storm, we sometimes don't get accurate wind speeds around the eye, but we still get a clear picture of the wind direction, giving us a good fix on the position of the storm center.

  The largest drawback to scatterometers at this time is that they are mounted on Polar orbiting satellites, which circle the earth from pole to pole, gradually covering the entire planet as it spins below them. But as you might have suspected, if you can picture this satellite motion in your mind, this produces pictures of "slices" of the earth's surface, leaving gaps in the picture. Here is a sample picture of a scatterometer's pass over the entire earth. So unfortunately we often get only "partial slices" of a hurricane or storm from these pictures.


WINDS FROM QUIKSCAT  One of the best current scatterometers available for analyzing winds is QuikScat. The QuikScat image below in figure 5 is a rare complete image (not sliced) of Hurricane Rita as it roared through the Gulf of Mexico.


Figure 5. A Quikscat satellite scatterometer image of Hurricane Rita taken at 7:04AM EDT on 09-21-2005. The line symbols with smaller ones extending from them are called "wind barbs", telling wind direction and strength. Courtesy Remote Sensing Systems.

  This image was taken at 7:04 AM EDT, roughly 6 hours before the Hurricane hunters sent their 1:53PM Vortex Data Message discussed in the prior section. The highest wind speeds visible on this Quikscat chart are 75 knots. Note the solid grey dots at the bottom of some of the wind barbs. On Quikscat images, these indicate where heavy rain is obscuring the wind speed. But also note how well it detects wind direction, showing the classic counterclockwise spiraling motion of the wind in towards the center of this tropical cyclone, allowing us to verify the storm center's position and spiral wind structure. Here is a more typical QuikScat partial sampling of Rita, showing one of the numerous "slices" of Rita that was taken about 12 hours after the complete scan shown in figure 5. So meteorologists are especially thrilled when the satellite passes directly over a storm they are interested in. Compare this to figure 1.


SEA SURFACE TEMPERATURES  One fascinating use of satellite technology is the ability to map where the warmest and coolest water temperatures are in any body of water. By using the altimeter data from satellites, the Jet Propulsion Laboratory can map where the water levels are relatively higher or lower. Since higher water levels correspond to warmer water and lower levels to cooler water temperatures, we can now use satellites to map the surface temperatures of any body of water. This is invaluable to tropical meteorologists because they know that areas of very warm water will provide a lot of extra energy to a hurricane. In recent years they have found two areas of unusually warm water in the Gulf of Mexico that have been named the Loop Current and the Eddy Vortex. The primary difference between these areas and other very warm areas of water in the Atlantic Basin is that they are both unusually deep. The warm water extends way below the surface. Why is that significant? Because while hurricanes move over warm bodies of water and draw extra energy from the water vapor over these hot areas, they are also removing that heat energy from the water and can even exhaust it by causing the cooler water below to rise up to the surface. When the cooler water starts to upwell to the surface, the hurricane loses this power source. It has actually "drained the batteries" in a sense. But if the extent of the hot water is very deep, the hurricane will get an inexhaustible supply of energy, and this will usually cause the storm to just keep on strengthening. We now carefully monitor these deep currents from year to year, and watch for rapid intensification when a storm moves over them. Hurricane Rita is a textbook example of a hurricane intensifying rapidly by drawing huge amounts of energy from the Loop Current. Figure 6 shows her long period over the Loop current on September 21st. By the way, category 5 Hurricane Katrina also spent a good deal of time over the Loop Current. Here is an overlay of  Hurricane Katrina's track and the Loop Current.


Figure 6. Sea surface heights in the Gulf of Mexico during the crossing of Hurricane Rita, annotated with the location of Rita's maximum intensity. The Loop Current and Eddy Vortex are areas of very warm water. The red areas are 35 to 60 centimeters (about 13 to 23 inches) higher than the surrounding water. This very warm water provided a lot of extra "fuel" for Rita. This composite map of Gulf water temps used data from four satellites, including the Topex/Poseidon and Jason from NASA. Map courtesy NASA's Ocean Surface Topography.



RAINFALL RATES  Another very important new use of satellite technology is detecting rainfall amounts. TRMM (Tropical Rainfall Measuring Mission), uses both passive sensors of microwaves (TMI) and active microwave sensors (PR or "Precipitation Radar") to determine the vertical structure of clouds and rainfall, so that it can measure the height of cloud tops as well as the amount of rain falling at the surface. So we now have "radars in the sky."  These radars are also mounted on Polar Orbiting Satellites, so we can see the same "slice" problem. But still, the benefits are obvious. Below in figure 7 is an almost complete satellite pass showing the TRMM rainfall rate measurement from Hurricane Rita on 9/21/05 at 5:07AM EDT, not long before the other satellite data we have been examining in this report. Note how heavy Rita's rainfall rates were at the time. They remained so, and together with her enormous tidal surge, caused extreme flooding when she made landfall at the Texas-Louisiana border. In addition to the valuable rainfall rate data, we can also clearly make out her classic spiral rain-bands, her eye, and pinpoint her location. All of these data together were a strong indicator to the NHC that Rita was undergoing rapid intensification. Compare this scan with figure 1 and figure 5.


Figure 7. TRMM satellite measurement of precipitation rates. in millimeters per hour on September 21, 2005 at 5:07AM EDT. Blues and greens are roughly .10 to .40 inches per hour. Yellow to red are 1 to 1.5 inches per hour, and dark reds are 2 or more inches per hour. Courtesy NASA.




  The HRD, or Hurricane Research Division, is the branch of NOAA dedicated to the research and understanding of hurricanes and all tropical weather systems. They are the front line in new thinking and studies of how these storms work. One of the most important tools to analyze the wind structure of a hurricane or tropical storm is the HRD Surface Wind Analysis. Although it is complex in theory and use, the basic idea is simple. They combine all of the data they can get from as many sources as are available, including, but not limited to, ships, buoys, reconnaissance aircraft, coastal platforms (such as oil rigs), and satellite data such as those we have covered in this report. They then combine the data with different weights, or biases, depending on how reliable they are, to get a single unified map of the wind field in a hurricane. All of the data are modified to conform to a surface height of 33 feet, open terrain, and an average 1 minute wind speed. This mapping is put together for as close to real-time use as possible, usually over a 4-6 hour period, so it is a critical tool that forecasters use when making analyses and issuing advisories. This data is very valuable when trying to assess a hurricane's real wind threat to people and property as it comes ashore. It also has developed a very good track record against the wind structure that is really observed at landfall. 

  Figure 8 below shows an HRD Surface Wind Analysis for Hurricane Rita at 1:30 PM EDT on September 21, 2005.

As you can see from the top of the diagram, all tools and observations that were used are indicated. In this case they used GPSSONDEs (the dropsondes discussed earlier), SFMR data (plane mounted radars that measure wind speed and direction), CMAN (automated weather sensing stations along coastal areas; see Coastal Marine Automated Network), SHIP (data from a ship that passed through the storm), MOORED_BUOY (there are many; see this National Buoy Data site), and of course, GOES satellite data. 



Figure 8. The HRD Wind Analysis for Hurricane Rita done from September 21, 2005 at 1:30PM. Courtesy Hurricane Research Division.

  Note the surface wind calculation of 125 knots (144 mph) 13 nautical miles northeast of the center. Also note the almost circular structure in the center around the eye. Farther out from the eye it gets less symmetrical. This Wind Analysis is from around the same time as Vortex Data Message 1753 discussed earlier and 1 hour after the satellite photograph shown in figure 1. It was also roughly 6 hours after the QuikScat analysis shown in figure 5. Examine them for yourself and see how they compare to each other. It is interesting to note that the HRD Wind Analysis is lower in wind speed than what was later found to be at that time, but this makes sense when we realize that Rita was undergoing unusually rapid intensification.





  We can now see the huge arsenal of tools meteorologists have at their disposal to examine and then forecast the strength and movement of hurricanes. I would like to add that, while our ability to forecast the future track of hurricanes has gotten quite accurate, there are still errors, and sometimes they can be fairly large. But the closer it is to landfall, the higher our accuracy goes. Forecasting the strength of a hurricane is more difficult, and the margin of error is much larger than that for the direction. Our case study here, Hurricane Rita, is proof of that. Although they successfully calculated that she would weaken to a cat 3 before landfall, no one forecast Rita to be a category 5 just 24 hours before she did. The NHC is still working hard on reducing these strength estimate shortfalls. But I hope I conveyed one thing well enough here: that when a hurricane is 48 or less hours from landfall, the accuracy is good enough now that people must heed the warnings. 

  The evacuation issue I started with was, of course, not intended to be within the scope of this report, but it seems certain at this point in history that we will be seeing a lot more cooperation between emergency management officials and the weather forecasters who are responsible for telling us how serious the danger is, since dozens died trying to get away from this hurricane. Suffice it to say that Houston's example has caused attempts in many other coastal regions to develop a better evacuation plan (see references.)   

  There are many other issues raised by this hurricane that need to be seriously discussed by all of us. Please see the References section below for a short list of articles on weather and logistics related hurricane issues that could not be covered here.





Storm Reports & Updates

Official NHC Report on Hurricane Rita

Wikipedia Entry on Rita

NOAA - October 2005 Report on Rita

Department of Interior - Hurricane Rita Info and Updates


News Articles

USA TODAY 1-Year Anniversary of Hurricane Rita - Report

Environmental News Service Report on Sabine Wildlife Refuge

WikiNews on Rita Evacuation

New York 1 News report - Mayor Unveils Hurricane Evacuation Plan