Low-Earth Orbit satellite views of Ian as it formed, and comparisons to Geostationary imagery

September 26th, 2022 |

Polar-orbiting satellites have microwave detectors that give important information about the low-level structure of an evolving tropical cyclone. If high clouds are omnipresent, it can be difficult for an analyst to diagnose storm strength with accuracy. Microwave energy penetrates clouds, however, and low-earth orbit (LEO) observations of microwave frequencies can reveal much about a storm’s structure.


24 September: South of Haiti


Consider the imagery below, showing the cluster of thunderstorms associated with then-Tropical Storm Ian south of Haiti. Based on just the still infrared image (admittedly, this would be easier with an animating image!), where would you place the center? Microwave data — 36.5 GHz and 89 GHz data from GCOM-W1 (from the AOML Direct Broadcast site here) suggest a center in between the top large regions of cold cloud tops in the infrared imagery (the 0900 UTC discussion has a center near 14.7oN, 73.5oW). MIMIC Tropical Cyclone imagery (from this link) for Ian on 24 September (here) can help a user determine where the center is as well.

GOES-16 ABI Band 13 Infrared (10.3 µm) imagery, and GCOM-W1 AMSR-2 Microwave imagery (36.5 and 89.0 GHz), 0620 UTC on 24 September 2022 (Click to enlarge)

25 September: Southwest of Jamaica


One day later, imagery from ABI and GCOM-W1 show a better-defined tropical system at 0700 UTC (Here’s the NHC discussion from 0900 UTC, at which time the center was at 14.9oN, 78.8oW). Even from the still ABI image, one could infer a center based on the spiral bands. Microwave information (36.5 and 89.0 GHz) certainly will increase confidence. Indeed, the low-level microwave signal (i.e., from 36.5 GHz) suggests a center very near the 0900 UTC location. The MIMIC TC animation from 0000 UTC 25 September – 0000 UTC 26 September (link) is showing a stronger signal for a center as well.

GOES-16 ABI Band 13 Infrared (10.3 µm) imagery, and GCOM-W1 AMSR-2 Microwave imagery (36.5 and 89.0 GHz), 0700 UTC on 25 September 2022 (Click to enlarge)

26 September: south of Western Cuba


NOAA-20 ATMS imagery (88 GHz) over Ian, 0606 and 0746 UTC on 26 September 2022 (Click to enlarge)

The LEO coverage on 26 September is a great example of why multiple LEO satellites are vital. The early-morning coverage from NOAA-20 is shown above; the gap between the two satellite passes is in an unfortunate spot for monitoring this tropical cyclone! However, Suomi NPP orbits overlap NOAA-20, and on this day Suomi NPP overflew the center of the storm, as shown below. The cadence was NOAA-20 to the east, 45 minutes later Suomi-NPP over the center, 45 minutes later NOAA-20 to the west. Here is an animation of the three passes. Polar monitoring capabilities will receive a big boost when JPSS-2 (slated to become NOAA-21) is launched (tentatively scheduled for 1 November 2022).

Suomi-NPP ATMS Microwave Imagery, 88.0 GHz, 0656 UTC on 26 September 2022 (Click to enlarge)

Ian at 0700 UTC on 26 September, below, is on the cusp of being upgraded to a hurricane (0600 UTC intermediate advisory), and an animation of the Band 13 imagery (a still image is shown below for comparison to the ATMS imagery) shows the center of rotation even though an eye is not present in the infrared (although one in the microwave).

GOES-16 ABI Band 13 Infrared (10.3 µm) imagery, and Suomi-NPP ATMS Microwave imagery (88.0 GHz), ca. 0700 UTC on 26 September 2022 (Click to enlarge)

ATMS and AMSR2 imagery as shown above are created from passive microwave sensors; that is, the sensors are detecting the microwave imagery emitted by the ocean, land, clouds and atmosphere. Other LEO satellites emit energy (“ping”) in the microwave and listen for a return signal. This leads to both scatterometry (not shown, as from the Advanced Scatterometer — ASCAT — instrument on Metop-B and Metop-C — available here) and Synthetic Aperture Radar imagery (available here for tropical cyclones), and shown below. The image below shows infrared and GLM imagery for then-newly upgraded Hurricane Ian (link). Although a distinct eye is still not present in the infrared imagery, SAR wind data defines an obvious region of reduced winds. Maximum SAR winds in this image are just above 70 knots.

GOES-East ABI Band 13 Infrared imagery (10.3 µm), GLM 1-minute aggregate Total Optical Energy (TOE) and RSAT-2 SAR Winds over Ian, 1110 UTC on 26 September 2022 (Click to enlarge)

VIIRS and ATMS imagery of Hurricane Ian on 27 September is here. For the latest information on Hurricane Ian, please refer to the National Hurricane Center. People in southern (and especially southwestern) Florida should be paying very close attention to this storm.

SAR data over Hurricane Fiona

September 19th, 2022 |
Sentinel SAR wind speeds over Hurricane Fiona, 2245 UTC on 19 September 2022 (Click to enlarge)

NOAA/NESDIS STAR is producing occasional SAR imagery over Hurricane Fiona (link). NetCDF files can be imported into AWIPS and displayed, as shown above (color-enhanced with Beaufort Scaling), in an image from late on 19 September when Fiona was just north of Hispaniola. The strongest winds were associated with the northern eyewall of the storm, and an asymmetry in the storm is apparent (The NHC discussion from 2100 UTC on 19 September (here), and the discussion at 0300 UTC on 20 September (here) will give more context). The Radial wind analysis at the SAR link (here) shows the strongest winds in the NW and NE quadrants of the storm, with a peak near 100 knots. The toggle below, from late on 19 September, compares GOES-16 ABI Clean window (10.3 µm) imagery and the Sentinel SAR wind estimates (with a different color enhancement). A parallax shift in the GOES-16 imagery means the two eyes do not overlay.

GOES-16 Band 13 infrared (10.3 µm) imagery and Sentinel SAR windspeeds, 2245 UTC on 19 September 2022 (Click to enlarge)

Update on 23 September: RADARSAT-2 overflew Fiona at 2236 UTC on 22 September. The toggle below shows SAR winds estimated at that time, with peak values in the eyewall near 100 knots (for more information on this SAR pass, including a more complete image over the eye, click here). There is a parallax shift in the high clouds of the storm; Fiona at this time was near 30.5oN, 69.3oW, far from the sub-satellite point at 0oN, 75.2oW: the parallax shift for the high clouds will be away from the sub-satellite point. The time difference in the observations — 5 minutes — should not cause a big shift for a storm moving at 17 knots (i.e., 20 mph) to the north-northeast.

RADARSAT-2 SAR observations over Fiona, 2236 UTC on 22 September 2022 along with GOES-16 Clean Window Band 13 (10.3 µm) infrared imagery at 2230 UTC (Click to enlarge)

Using SAR Winds to center-fix Tropical Cyclone Estelle in the Eastern Pacific

July 20th, 2022 |
OGES-17 Visible (Band 2, 0.64 µm) and Infrared (Band 13, 10.3 µm) at 0150 UTC on 20 July 2022 (click to enlarge)

GOES-17 visible and infrared imagery shows Tropical Storm Estelle over the eastern Pacific to the west-southwest of Baja California. Although there are regions of strong convection, satellite presentation of the storm suggests a modest tropical storm. Center-fixing a storm such as this (with just one image vs. an animation!) is complicated by both parallax (GOES-17 is overhead at 0oN, 137.2oW) and the lack of an easily-discerned eye feature in this system.

Instruments that view surface winds, via Scatterometry or via Synthetic Aperture Radar (SAR), can offer better center fixes. Consider the toggle below of Radarsat-2 SAR winds over Estelle. SAR Winds (available here; SAR winds over Tropical Systems can be found here.) show a lopsided storm, with most of the strong winds on the poleward side of the center. The SAR winds that use 0.5-degree GFS wind data as a first guess show a characteristic hourglass feature near the center that advertises an ambiguity between the first guess winds and the observations. That feature is missing in the SAR wind field that is a product of the 0.25-degree GFS winds, suggesting that the 0.25-degree wind field is more accurate. Both fields show a rather baggy center at or just south of 20oN.

RSAT-2 SAR Winds over Estelle benchmarked by 0.5-degree and 0.25-degree model output, 0150 UTC on 20 July 2022; note the different color scales used in the two images (Click to enlarge)

Scatterometry from MetopB and HY2B between 1730 UTC 19 July and 0230 UTC 20 July, respectively show a storm moving west-northwest, passing north of 20oN just after 0230 UTC.

ASCAT (MetopB) and HY2B Scatterometry winds at 1730 (19 July) and 0230 (20 July), respectively.

Bore feature in SAR winds over Lake Michigan

June 6th, 2022 |
RCM2 Wind Speeds, 1152 UTC on 6 June 2022 (Click to enlarge)

Color-enhanced wind speeds observed from Synthetic Aperture Radar (SAR) data on the second RADARSAT Constellation Mission satellite (RCM-2), shown above in AWIPS (Click here for a similar image online at this website) show parallel lines of enhanced wind speeds, a wind structure suggestive of a bore (click here for many previous blog posts on this phenomena), over southern Lake Michigan. Peak wind values (in yellow over the water) are near 14 m/s, with minima in between the lines showing winds around 5-6 m/s. (Note that SAR wind information over land is invalid). When this kind of event happens under mostly clear skies, parallel lines of clouds (that are perpendicular to the observed wind) develop. In this case, widespread clouds prevented satellite detection of cloud bands. The toggles below show Visible (Band 2, 0.64 µm) and Infrared (Band 13, 10.3 µm) at the time of the SAR observations.

GOES-16 Band 2 visible (0.64 µm) and RCM Winds, 1152 UTC on 6 June 2022 (Click to enlarge)
GOES-16 Band 13 infrared (10.3 µm) and RCM Winds, 1152 UTC on 6 June 2022 (Click to enlarge)

Modest convection (cloud tops are only around -48oC) over central Lake Michigan likely generated the bore. For a bore to propagate, a strong inversion is required. Soundings at Green Bay WI and Gaylord MI likely are unrepresentative of the environment over southern Lake Michigan. The sounding at White Lake Michigan — near Detroit — (from this website) does show a surface inversion, as does the sounding at Davenport IA. The cool late-Spring waters of Lake Michigan will serve to anchor a similar low-level inversion over the Lake; bore features travel along those inversions.