As we are approach the more active part of hurricane season in late August, most of the action so far has been taking place in the Central Pacific with Hurricane Hector passing to the south of the Hawaiian Island chain a couple weeks ago. Now, Hurricane Lane is projected to pass much closer to Hawaii and this time the U.S. National Oceanic and Atmospheric Administration (NOAA) has sent in the “big guns”. In addition to “Gonzo” — NOAA’s Gulfstream jet (NOAA has an agreement that allows them to name their aircraft after Sesame Street characters), the Lockheed Orion P3 “Kermit” is also on duty in Hawaii to fly research missions into Hurricane Lane. Kermit brings an arsenal of scientific sensors including the airborne Doppler radar and the Stepped-Frequency Microwave radiometer (called the “Smurf”) and a supply of GPS dropsondes to launch into the storm.
On Saturday, August 22, 1992, I met with Herb Saffir (coauthor of the Saffir-Simpson Scale) in his Coral Gables office, south of downtown Miami, discussing a manual for post-storm damage investigation. I was also due to be the hurricane scientist member of a panel that Herb was chairing for the American Society of Civil Engineers. As the meeting ended it became apparent that Andrew, which had become a hurricane that morning, was approaching the Bahamas and was not going to recurve northward as hoped. It was coming right at us.
My home was in Coconut Grove, about three miles south of downtown Miami. I called my wife and told her to get to the Publix supermarket since a hurricane warning was imminent and we knew from earlier storms that there would soon be a run on supplies. I headed off to Home Depot for plywood and wondered how I could protect the house and still make my scheduled research reconnaissance flight into the storm on Sunday, working on one of the National Oceanic and Atmospheric Administration’s (NOAA) P3 Hurricane Hunter aircraft. As a hurricane wind specialist, I would be monitoring the wind field and radar displays over a proposed ten-hour mission. That mission was scrubbed when it became apparent that the aircraft would need to be evacuated from their base at Miami International Airport.
Instead, the Air Force Reserves of the 53rd Weather Squadron out of Keesler Air Force Base in Biloxi, Mississippi, flew a C-130 into the storm late Sunday, while I thankfully completed my shutters and we accommodated some friends (and their pets) who lived in a storm surge evacuation zone. Our home was near 20 feet (6 meters) above sea level on the coastal ridge in Coconut Grove, so flooding would not be a problem.
Well, Andrew hit overnight, and by 9 a.m. Monday morning, we seemed to be OK. Lots of trees were down and the power out but the house was intact with just a few broken roof tiles and one cracked window. We walked down to Biscayne Bay to check on the flooding and took some pictures, feeling the excitement of viewing a dramatically changed landscape. The mood changed when we received a call (yes, the phones were working even though there was no power) from another friend who was crying and upset about damage further south. We spent hours picking our way about 10 miles south to find that our guest’s home in “Whispering Pines” was within ground zero of the northern eyewall of Andrew. Their roof covering was peeled, double front doors were blown in, and all their living room furniture had been blasted through a sliding glass door into their pool.
Our friends were in shock and time was short due to an impending curfew so we made our way to the main north-south drag, U.S. Route 1 or Dixie Highway (difficult since all the street signs were blown down) for the drive back to Coconut Grove. We were marveling at the lack of any organized response, when we noticed a white school bus making its way southward on US 1. We couldn’t hold back tears when we saw that the first responders were the City of Charleston Police Department. This was pay back from 1989, when Charleston was hit by Hurricane Hugo and Miami Metro-Dade County helped in the response.
Overnight more than 200,000 were left homeless without power and with few supplies. Fifteen died from blunt trauma or drowning in the storm surge. It took days for significant relief to arrive with Kate Hale, the Miami-Dade County emergency director pleading “Where the hell is the cavalry on this one?” Despite the slow pace of response, it was amazing how the communities of South Florida pulled together to help each other out, with neighbors helping each other and sharing supplies.
It was also amazing that the Air Force crew on that C130 flew all night, monitoring Andrew as it strengthened through landfall and continuing to fly a hazardous pattern over land as the storm progressed inland. For several months’ afterwards, my team at NOAA’s Hurricane Research Division worked to piece together data and reconstruct Andrew’s wind field. We visited sites with incredible damage where lives were lost. The public sent data in and we worked with wind engineering colleagues (Tim Reinhold from Clemson) who borrowed the Virginia Tech wind tunnel to test anemometers similar to the one that measured the peak wind speed in Perrine.
State Attorney General Janet Reno called to check on how strong the winds really were to allay some of the rumors swirling around. The Miami Herald published our wind footprint on the front page in 1993, indicating that the highest winds were north of where originally thought. Our work was finally published as a two-part paper in Weather and Forecasting, which set the stage for objective analysis of hurricane wind fields. Back then the reconnaissance aircraft did not have a way to measure the winds near the surface, so storm intensity was estimated as a fraction of the maximum flight-level winds, resulting in a Category 4 assessment on the Saffir-Simpson scale. After 10 years, analysis of measurements closer to the surface from a new type of instrument, the GPS dropwindsonde, suggested that Andrew may have been a Category 5 storm at landfall. In 2009, research from an even newer instrument (Stepped-Frequency Microwave Radiometer), that remotely senses wind speed from the radiative emission of sea foam, reinforced the Category 5 assessment.
Hurricane Andrew had a profound effect on everyone living in South Florida at the time. It is one of those life milestones from which we measure everything before or after. Miami-Dade County responded with a tough new building code with product testing and enforcement, which influenced the eventual development of a unified Florida Building Code. And it kickstarted the insurance industry into using sophisticated models that could estimate the risk of future Andrews, and performance in emergency management and response was found to influence presidential elections. The rebuilding created an economic boom, but many folks moved away while others moved in transforming rural areas such as Homestead in Miami-Dade, from farm fields into suburbs.
And for me, I went on to develop a research analysis system called H*Wind, now fully within RMS, which now allows us to monitor and analyze the wind field in real time using every piece of data we can get our hands on, such as satellite, dropsondes or portable MET towers placed just ahead of the storm by engineering and atmospheric science students and faculty at the University of Florida and Texas Tech. The HWind fields have become the analysis of record for significant landfall events and a standard for model evaluation with hundreds of citations in peer-reviewed scientific publications.
We work with scientists from all over the world to help develop cutting edge techniques for remote sensing of winds from space, or to provide the best possible forcing for a storm surge or wave model. We have a much better idea of the intensity and extent of the damaging winds now, and also developed new damage scales based on integrated kinetic energy that consider the destructive potential of large storms. Our extensive enhanced archive of historical storms is now helping us to design and evaluate new hurricane forecast products that are destined to find their way into RMS(one). RMS HWind is now the world’s leading provider of tropical cyclone wind field data, with observation-based data products for both real-time and historical wind field analyses in the western North Atlantic, Eastern Pacific and Central Pacific basins.
Andrew occurred during a year forecast to have “below normal” activity. I’m often asked, “What kind of year are we going to have”? My answer? It doesn’t matter… just tell me whether Andrew is coming.
Mark Powell, vice-president – model development, RMS-HWind
Michael Kozar, senior modeler, RMS-HWind
In a Safety Recommendation Report issued by the U.S. National Transportation Safety Board (NTSB) last month, the Board took the unprecedented step of requesting that a fellow federal agency, the National Oceanic and Atmospheric Administration (NOAA), together with the U.S. Coast Guard, act immediately to do more in improving maritime safety. The NTSB Safety Recommendation Report was released as part of its ongoing investigation into the tragic sinking of the merchant vessel El Faro, a U.S. flagged, 790-foot (240 meter) roll-on/roll-off container ship with a cargo of containers and vehicles, which sank with all hands during Hurricane Joaquin in October 2015.
It’s one thing being invited to speak at an industry event in front of dozens of the leading scientists in your field. It’s another to find, with a certain astonishment, that virtually all of them use RMS HWind to validate their scientific work.
Last month, at a U.K. Met Office-hosted workshop, I spoke about the RMS HWind hurricane modeling solutions to a group of high-wind remote-sensing scientists from academic and government agencies from around the world, including:
- European Space Agency (ESA)
- National Aeronautics and Space Administration (NASA)
- French Research Institute for Exploitation of the Sea (IFREMER)
- Royal Netherlands Meteorological Institute (KNMI)
- Met Office (U.K.)
- European Center for Medium-Range Weather Forecasts
- National Space Science Center, Chinese Academy of Sciences
- Institute of Applied Physics of the Russian Academy of Sciences
- National Oceanic and Atmospheric Administration (U.S.)
All of the above agencies are researching how satellite-mounted remote wind sensors can be used, most effectively, to inform on hurricanes and typhoons developing over the ocean.
During the workshops I was delighted to learn that every major remote sensing agency had used the RMS HWind archive of historical storms to validate and calibrate their sensor programs for detecting high winds from space. RMS HWind also provides real-time analysis of hurricanes as they happen with observational data from instruments in the air, in the sea and on land – including aircraft reconnaissance, GPS dropsonde instruments, sea buoys and satellites.
By citing HWind products and research in their peer-reviewed publications, these agencies provide independent endorsements that enhance the scientific credibility of the HWind archives and services, while also giving us a chance to evaluate cutting-edge technology before it becomes operationally available.
There is tremendous value in scientific collaboration and, as such, RMS facilitates the science community’s understanding of hurricanes by providing our academic partners free access to HWind products for their scientific investigations.
Sensors in Space
None of the satellites we discussed at the Met Office workshop actually measure wind directly – rather, they measure a signal which is influenced by the wind. So, for example, the new NASA CYGNSS system measures the reflection of GPS signals off sea’s surface, which is like the reflection of the moon on the surface of a lake. Winds disturb the surface and this scatters the signal.
Another satellite, the Canadian RADARSAT-2, has already been up for a few years and can capture images of the fine scale roughness of the ocean surface. But to collect these images and convert them to a wind speed reading requires a lot of advance planning, followed by lengthy processing.
Which is where RMS HWind comes in. Our 1 km gridded HWind Snapshots make it easy for scientists to overlay their satellite instrument measurements (typically a microwave signal reflected from the sea surface) over our wind analyses. They can do this for several storms of various sizes and intensities to convert the measured signal to wind speed over a range of meteorological and oceanic conditions.
Due for release this winter, the HWind Enhanced Archive of wind hazard metrics will provide a high resolution library of tropical cyclone wind fields for the North Atlantic, the Caribbean, Gulf of Mexico, and the east and central Pacific. In the coming years we’d hope to see the expansion of tropical cyclone wind field monitoring globally.
This expansion could benefit areas of the world with insurance protection gaps hugely. Increased insurance penetration in the Asian and Australian markets, together with new risk transfer products using parametric triggers, could help improve financial resilience to catastrophic tropical cyclones in whole new regions of the world.