Probing the core of tropical cyclones

On 15 December 2016 a constellation of eight microsatellites successfully deployed into a low-inclination circular orbit after hitching a ride on a Pegasus XL rocket. The Cyclone Global Navigation Satellite System (CYGNSS), the first orbital mission funded under NASA’s Earth Venture program, peers through Earth’s atmosphere, clouds, and precipitation to gain a unique perspective on tropical cyclones.

CYGNSS is especially valuable to forecasters because it specializes in acquiring a notoriously difficult yet extremely important measurement: surface wind speed near the core of the cyclone. CYGNSS is not plagued by the limitations of previous spaceborne instruments, which have struggled to observe Earth’s surface through the intense precipitation surrounding storms’ centers. Prelaunch research suggests that CYGNSS will become an important contributor to tropical cyclone forecasts. Observations from CYGNSS will also help diagnose cyclones’ destructive potential in terms of storm surge and winds—important information to communicate to those in the path of a storm.

CYGNSS takes advantage of the existing infrastructure of the GPS satellites already bathing Earth in their navigation signals. Using a technique called Global Navigation Satellite System reflectometry, the instruments onboard the spacecraft receive GPS signals—operating at around a 19 cm wavelength—that have diffusely scattered off Earth’s surface. When signals from GPS satellites hit Earth’s surface, they are reflected in many different directions. Luckily for CYGNSS, the properties of the ocean surface are related to the local near-surface wind speed. As wind speed increases, the ocean surface becomes rougher, changing how the signals reflect back to the CYGNSS spacecraft.

The all-weather observation of surface wind speed is critical. Other remote-sensing observation platforms are plagued by contamination in areas of intense rain, and it just so happens that the inner core also has the heaviest rain. As a result, scientists using previous spaceborne instruments have had difficulty estimating surface wind speed in the core. By using signals that are much larger in wavelength than the size of typical drops of rain, which are usually no more than a few millimeters across, the signal-receiving capabilities of CYGNSS—as well as those of your smartphone’s built-in GPS—work in all weather conditions . Observations from CYGNSS will provide a unique set of surface wind speed data covering the entire extent of the tropical cyclone, including the core.

The Storm Intersection Forecast Tool (SIFT) was developed to predict when CYGNSS coverage will overlap with storms of interest. Each CYGNSS satellite can measure four points on the ground at once, referred to as specular points, and SIFT shows the constellation’s coverage of those points over time. The specular points are determined geometrically by the relative locations of a CYGNSS and GPS satellite pair. SIFT keeps track of where all eight CYGNSS satellites are and computes the potential specular point locations of all 240 pairs.

The video below, produced using SIFT, simulates what the CYGNSS constellation would have observed had it been in orbit over Hurricane Matthew last year. Forecasts for Matthew could have benefited from the availability of CYGNSS data, as the storm unexpectedly intensified from a category 1 to a category 5 over the course of 24 hours.

In the video, Matthew’s current-forecasted position and size according to the National Hurricane Center’s forecast are shown by the bright red circle. The size of the storm is defined by using an estimate of the maximum extent of a 34-knot (63 km/h) wind speed, with additional buffer for forecast uncertainty. The future-forecasted position and size are denoted in pink. Colored triangles in the video correspond to the location of one of the satellites. Lines plotted in the video denote predicted specular points where CYGNSS measurements would occur.

The video shows that valuable data would be available in the bright red circle, where the highest wind speeds are expected for any tropical cyclone. Some members of the CYGNSS science team are working to assimilate those data into weather models in an effort to improve the performance of tropical cyclone intensity forecasts. Additionally, those data can be used to diagnose the destructive potential of a storm.

One thing to notice in the video is that CYGNSS is observing the storm in collections of tracks across the surface. The goal of my research with principal investigator Christopher Ruf of the University of Michigan is to determine how to use the data to estimate key tropical cyclone metrics despite variable sampling and incomplete coverage between tracks of CYGNSS observations.

Although CYGNSS data can be used to estimate intensity, or maximum surface wind speed, the process is not as basic as using the maximum estimated surface wind speed provided by CYGNSS over a storm. Such a simple approach would be biased by measurement noise and rendered useless if the satellites never tracked over the specific area of maximum-strength winds. Instead, Ruf and I developed a methodology that interpolates within the gaps between tracks, which results in a more reliable estimate of intensity as well as other wind-speed metrics.

Our methodology relies on a wind model that characterizes the relationship between the surface wind speed and the radial distance from the storm center using only a few parameters. Although the parameters in the model are initially unknown, we can solve for them using a newly developed algorithm that requires just two main inputs: CYGNSS wind speed observations over the storm and the location of the storm center. Prelaunch simulations predict that the methodology, when applied to CYGNSS observations, will provide estimates of intensity, radius of maximum wind speed , integrated kinetic energy , and wind radii .

Although this is a fantastic result, a lot of work remains. Future efforts will include calibration and validation of these methods using the first set of tropical cyclone data from CYGNSS, which we are just beginning to analyze.

Mary Morris is a Caltech postdoctoral fellow at the NASA Jet Propulsion Laboratory in Pasadena, California. She earned her PhD from the University of Michigan in December 2016. Her research interests lie at the intersection of microwave remote sensing and atmospheric science.