June 24, 2019 – On the morning of October 7, 2018, forecasters at the National Hurricane Center (NHC) in Miami, Florida, began issuing advisories on the newly formed tropical depression Michael. Armed with an impressive observation network comprised of aircraft, radars, buoys, and state-of-the-art sensing technologies to monitor the system’s position, intensity and size, the NHC’s forecast track had Tyndall Air Force Base directly in its crosshairs.
Thanks to these advisories, base commanders had sufficient time to evacuate all personnel and many aircraft to safe locations. Unfortunately, they would soon learn just how precise these forecasts were. After rapidly intensifying, Hurricane Michael went on to become one of the most intense and destructive tropical cyclones to ever make landfall in the United States, significantly damaging or destroying nearly every structure at Tyndall and leaving behind more than $1 billion in losses to remaining aircraft unable to make it out ahead of Michael’s fury. Sadly, this is not the first time a tropcial cyclone has struck the Department of Defense. Category 5 Hurricane Andrew obliterated Homestead Air Force Base in 1992 and led to over $100 million worth of base reconstruction. The infamous Typhoon Cobra in 1944, nicknamed “Halsey’s Typhoon,” resulted in the loss of nearly 800 Sailors, three destroyers, and more than 100 aircraft.
For joint forces from U.S. Indo-Pacific Command (USINDOPACOM), U.S. Central Command (USCENTCOM) and U.S. Africa Command (USAFRICOM) operating within and along the Pacific and Indian Oceans where more than 70 percent of the world’s tropical cyclone activity occurs each year, the constant threat of another Michael or Andrew-like storm is very real. However, contrary to the continental United States (CONUS), tropical cyclones in this part of the world can traverse vast stretches of open ocean virtually unobserved except by satellites. The last routine U.S. aerial reconnaissance flights outside CONUS ceased in 1987 due to budget cuts. Since that time, nearly 200 tropical cyclones have passed within 150 miles of Okinawa and Guam alone.
It is the responsibility of the Joint Typhoon Warning Center (JTWC) in Pearl Harbor, Hawaii to provide forecasts, warnings, and decision support to keep U.S. personnel safe, protect assets, and to leverage battlespace environment awareness for tactical planning.
“Persistent surveillance over 55 million square miles to identify an obscure rotation that has the potential to develop into a typhoon is a constant challenge,” according to Cmdr. Robin Cherrett, commanding officer of the JTWC. Novel new uses for a pair of research satellites designed to monitor soil moisture and ocean salinity, launched by the National Aeronautics and Space Administration (NASA) and the European Space Agency (ESA), are providing JTWC forecasters new tools in the fight against Mother Nature.
Providing warfighters and decision makers with the longest lead-times and most accurate forecasts possible of when, where, and how a tropical cyclone may form and evolve over the next five days begins with an accurate environmental analysis. The analysis forms the basis of the entire forecast; therefore, accuracy of the initial starting conditions is essential. “It is very difficult to say where a system is going to go if we can’t even accurately assess its current state,” says JTWC’s Director, Mr. Bob Falvey. However, having to rely almost entirely on satellite data to perform this analysis presents some difficult hurdles to overcome.
Much like in CONUS, geostationary satellites monitor the day-to-day weather patterns occurring over most of the globe. However, these satellites are limited in what they can observe. High clouds typically associated with tropical cyclones can often shield underlying structures, preventing analysts from observing critical low level features during development. During periods of darkness, analysts must rely on infrared imagery which is even more limiting. “When a well-defined super-typhoon is heading towards Kadena Air Base, it is very easy to identify the eye and determine the system’s precise location; however, when we’re looking at a disorganized mass of clouds approaching Guam and trying to decide if and when we need to start issuing warnings that may result in multi-million-dollar sortie decisions – this becomes a much more difficult problem,” according to Lt. Cmdr. Brian Howell, prior JTWC operations officer.
Such a scenario played out this July when Tropical Depression Maria formed southeast of Guam. Without the benefit of aerial reconnaissance to help locate the developing center, early position estimates were scattered about over hundreds of square miles. This resulted in large swings in the JTWC forecast track, and significant intensity errors. JTWC still relies heavily on a decades-old cloud pattern matching process, called the Dvorak Technique, to estimate tropical cyclone intensity. “Although it can’t account for all situations, the technique has proven to be quite reliable – but you must have an accurate estimate of the center position,” notes Capt. Sean Zoufaly, USAF, JTWC Satellite Operations Flight Commander. In this case, the lack of timely and precise data to monitor Maria’s development into a 63 mph tropical storm prior to landfall partially contributed to a reduction in theater strike force readiness.
Despite the usefulness of the Dvorak Technique, it cannot provide details about the extent of a tropical cyclone’s wind field including gale-force and destructive winds, known collectively as wind radii. These wind radii estimates are key factors for decision makers when planning to divert or sortie ships and aircraft, or to set base tropical cyclone conditions of readiness (TCCOR). To analyze tropical cyclone position and wind structure, forecasters rely heavily on polar orbiting microwave satellites, in particular two types of sensors called radiometers and scatterometers. These sensors operate at frequencies that are largely able to see through high clouds and all but the most intense rain and hail to reveal details about a cyclone’s intense inner core.
Scatterometers, specially designed RADARs, measure energy reflecting off “capillary waves” or ripples found on the ocean surface to provide estimates of two-dimensional wind vectors (i.e., both wind speed and direction). Without aerial reconnaissance, scatterometry data is generally regarded as the highest quality data available at JTWC, often providing the only clear indication that a tropical cyclone has formed. “When we look at the value that scatterometer data adds to JTWC forecast confidence and accuracy, and how that translates to high-dollar decisions impacting resource protection and mission readiness, these sensors have proven their worth over and over again,” notes Falvey.
Unfortunately, due to their high cost to launch and operate, microwave satellite passes are becoming increasingly rare as new platforms are either not being funded to replace those that have expired or are using cheaper, less capable technology than what is currently being flown. For instance, despite the unique importance of scatterometer data, plans for future U.S. scatterometer sensors will largely rely on so-called “passive” versus “active” detection, which is not effective in areas of extreme precipitation, like that within a tropical cyclone.
“We leverage research and foreign-partner data to the maximum extent possible; however, in today’s information-assured environment, even that is becoming more challenging,” notes Cherrett. Even the best microwave sensor still has its own drawbacks. At most, a polar-orbiting satellite will pass over a tropical cyclone twice per day. “In many cases, the portion of the Earth’s surface observed as the satellite passes overhead is so narrow that it misses the tropical cyclone completely”, states Capt. Zoufaly. Additionally, even the highest quality scatterometers currently flying cannot measure wind speeds higher than hurricane/typhoon force (75 mph).
To address these limitations, JTWC is always looking for the next new sensor to help fill the growing gap in space-based monitoring. This brings us back to soil moisture and ocean salinity satellites.
The ESA launched the Soil Moisture and Ocean Salinity mission (SMOS) satellite in November 2009 and NASA followed with their own Soil Moisture Active Passive (SMAP) mission which launched on January 31, 2015. The primary goal of these missions are to study the Earth’s water cycle, including monitoring ocean surface salinity and drought conditions, predicting floods, assisting crop productivity, and improving measurements of evaporation from land surfaces to improve numerical weather prediction models. Although not originally considered part of their respective science missions, SMOS and SMAP project investigators, Nicolas Reul (The Institut Français de recherche pour l’exploitation de la mer, France) and Thomas Meissner (Remote Sensing Systems, USA) began examining measurements of over-ocean wind speeds produced by these sensors. SMOS and SMAP radiometers have relatively coarse horizontal resolution compared to current scatterometers (approximately 40 km versus 25 km), however, these active sensors operate within a low microwave frequency known as the “L-band” which is not affected by heavy precipitation and does not suffer from the same 75 mph measurement limitation of scatterometers operating in other bands. After extensively validating the accuracy of SMOS and SMAP data, their reports indicating these sensors could indeed measure winds in even very extreme conditions caught the attention of federal TC researchers.
Researchers Buck Sampson at the Naval Research Lab (NRL), and Dr. John Knaff at the National Oceanic and Atmospheric Administration’s National Environmental Satellite, Data, and Information Service (NOAA/NESDIS) had been collaborating to develop new techniques to measure and forecast TC winds and structure for several years. However, the growing gaps in available satellite data and the absence of wind speed measurements above the maximum scatterometer sensitivity threshold was a constant challenge. When given the opportunity to evaluate the new SMOS and SMAP data, the researchers jumped at the chance. The new data were quickly integrated into the Navy’s Automated Tropical Cyclone Forecast System (ATCF) and new visualization tools were implemented by a team at NRL, allowing JTWC forecasters to view the information. Every time a satellite pass would go over a tropical cyclone, Sampson and Knaff would scrutinize, evaluate, and discuss the data with JTWC Lead Scientist, Brian Strahl. ”The preliminary results looked very promising. Not only were the data proving useful for our storm size estimates, but they seemed to have inner core wind structure that we weren’t seeing in our other estimates”, said Sampson.
“The 34 knot data were consistently matching up with other estimates, and we were routinely seeing 65 and even 100 knot (75 and 115 mph, respectively) estimates coming out of these sensors”, added Knaff.
Lt. Cmdr. Howell recounts one instance shortly after Tropical Storm Kai-Tak formed to the east of the Philippines in December of 2017. Just prior to making landfall, Kai-Tak appeared to rapidly organize. The Dvorak intensity estimate, based on infrared satellite imagery, indicated maximum winds were between 30 to 35 knots, although there were no recent scatterometer data to support these wind estimates. “Conveniently, SMAP passed overhead at the time and the sensor indicated that an isolated area of winds may have been nearly 20 knots higher than these estimates, and the extent of wind field was already quite large. Our ability to view these data in ATCF, and the early confidence we had in it, allowed us to quickly revise our forecast and alert afloat units operating nearby,” says Howell. “Any time a cyclone forms near land we think about troops that may be vulnerable to storm surge and flooding – we want to give these forces as much lead time as possible”, added Cherrett.
As the investigative process continued, Reul and Meissner were instrumental in improving the latency and availability of the SMAP and SMOS data to meet the strict timeline of the JTWC forecast cycle. The collaboration also resulted in the development of new automated algorithms to estimate tropical cyclone wind structure, further streamlining the analysis process. “Having SMAP and SMOS routinely available in our wind and structure guidance tools is definitely moving the needle in terms of our accuracy and reliability and demonstrates the need for additional active L-band sensors in the future”, according to Strahl. Collectively, forecast improvement efforts over the past several years are contributing to new JTWC accuracy records for tropical cyclone forecast intensity and size in 2018. Although the threat of tropical cyclones will always remain, these advancements continue to improve critical decisions impacting mission readiness and resource protection so warfighters will be ready for the next big storm.
For Reul and Meissner, discovering the utility of their data to observe and forecast tropical cyclones, and seeing how JTWC is putting that data to use protecting lives and property has been a win-win. Susanne Mecklenburg , ESA’s SMOS mission manager, adds “As the SMOS mission is now entering its tenth year of service, we’re constantly seeking to demonstrate new ways that unique science missions like SMOS can be used to improve our daily lives and hence increase our return on investment. Although we are not an operational mission, we are happy to see that the JTWC can use these data to help fill space-based sensing gaps and improve their overall mission success”. At JTWC, the hunt for the next new innovation continues. Whether it’s new models, observations, or satellites like SMOS and SMAP, building partnerships across communities and agencies, both domestic and internationally, is a vital component of ensuring U.S. forces are prepared for the threat of tropical cyclones.