Comparing SAR data over Lake Superior to radar

August 5th, 2021 |
SAR winds from RCM-2, 2340 UTC on 5 August 2021 (Click to enlarge)

Synthetic Aperture Radar (SAR) winds are derived from a microwave signal pinged from a satellite; backscatter is converted into winds (given a background field that is typically from a numerical model). The wind structure here is suggestive of a bowing feature to a convective downdraft. How does it compare to radar or satellite imagery? GOES-16 Satellite imagery for the hour bracketing the imagery above is shown below. Convection (weak) is apparent moving east from the tip of Keewenaw peninsula.

GOES-16 “Red” Visible (0.64mm) imagery, 2301-2356 UTC on 5 August 2021 (click to enlarge)

What did radar imagery look like at this time? The imagery below, courtesy Nick Langlieb, the SOO at WFO Marquette, shows radar echoes (base reflectivity) at a different level than the Lake-surface values sampled by SAR, so a direct comparison is a challenge (click here to see Correlation Coefficient at the same time). It’s peculiar that no radar signal is apparent to match the strong SAR winds just offshore from the tip of the Keewenaw peninsula (between 88º and 87.5º W) — although there is a feature oriented north-south a small distance to the west of the peninsular tip; there seems to be a better (but not exact) match with the SAR winds near 87º W.

1.3º – Base Reflectivity, 2340 UTC on 5 August 2021 (click to enlarge)

SAR winds near Tropical Storm Nepartak

July 27th, 2021 |
RCM1 SAR Winds over Tropical Storm Nepartak just east of Honshu, 0828 UTC on 27 July 2021 (Click to enlarge)

RADARSAT Constellation Mission One (RCM1) passed over Nepartak at 0828 UTC on 27 July 2021, and the image above shows the Synthetic Aperture Radar (SAR) winds derived at that time. There is a widespread region of ~40-knot winds (cyan to green in the color enhancement), a bit stronger than the 30-knot winds viewed by MetopA at 1040 UTC (shown here, in this blog post). How certain can a forecaster be of the even-stronger winds that exist in an area near the coast near 37.4ºN, 141.4ºE, and in an arc from 37.5ºN, 142.5ºE to 37.2ºN, 144ºE? There are isolated SAR estimates in that band that are near 60 knots! Are there other data sources to confirm that kind of wind? (See information at the bottom for clarification!)

The image below shows Himawari-8 Clean window imagery at 0829 UTC, just after the image above. Three points with cold cloud tops, suggestive of more vigorous convection, are indicated: (37.6ºN, 141.18ºE); (37.57ºN, 143.01ºE); (37.35ºN, 144.01ºE). The structures in the infrared imagery do match the structures in the SAR winds, but offset a bit to the north in the Himawari-8 imagery, as expected because of the parallax shift: features will be displaced away from the sub-satellite point, with the displacement increasing for higher clouds, and for greater distance from the sub-satellite point (on the Equator at 140.2ºE for Himawari-8). Convective downdrafts could be responsible for the highest winds shown in the SAR analysis.

Himawari-8 Clean Window Infrared (10.41 µm) imagery, 0829 UTC on 27 July 2021 (Click to enlarge)

The strong winds in the image are not observations of strong winds by SAR. Rather, these are most likely the result of reflection off of ice in the atmosphere — ice that is likely to be present around convective towers that have glaciated. (Thanks to Christopher Jackson, GST/NOAA, for this information!)

SAR Winds over Lake Michigan compared with Radar winds

June 28th, 2021 |

RCM3 SAR winds over Lake Michigan, 1201 UTC on 28 June 2021 (Click to enlarge)

RADARSAT Constellation Mission 3 (RCM3) Synthetic Aperture Radar (SAR) wind data over southern Lake Michigan at 1201 UTC on 28 June, above (click to enlarge, taken from this website), shows a small region of strong winds just offshore of southeast Wisconsin and northeastern Illinois.  GOES-16 Band 2 visible imagery (clipped from the CSPP Geosphere website, click here for the direct link to the imagery — valid until mid-July) shows modest convection just off the Wisconsin/Illinois shorelines.

CSPP Geosphere visualization of GOES-16 Visible (Band 2, 0.64) imagery, 1200 UTC on 28 June 2021 (Click to enlarge)

Winds from the Sullivan WI (WFO MKX) radar, below, (courtesy John Gagan, SOO), show a similar structure. The challenge in comparing the SAR winds and the radar winds: The Sullivan radar beam in that location is about 5000 feet above the surface. Storm-relative velocities are about 20 kt.

1201 UTC WFO MKX Radar display showing Base Reflectivity (upper left), Storm Relative Velocities (upper right), ZDR (Lower left) and Base velocities (lower right).  Beam angle of 0.5 degrees. (Click to enlarge)

Strong winds from SAR over the South Pacific

June 28th, 2021 |

GOES-17 ABI Band 13 infrared (10.3 µm) imagery, 0500-0600 UTC on 28 June 2021 (click to animate)

The animation above shows convection over the south Pacific to the east of American Samoa (note the Manu’a Islands just east of 170 W Longitude in the animation). Extensive cloud cover will limit the ability of the GOES-17 instrument to detect low-level cloud motions underneath the deep convection, which motions can be used to infer wind speeds. Other data sources are needed. In addition to scatterometry from the MetOp series of satellites, synthetic aperture radar (SAR) that flies on the RADARSAT Constellation Mission (RCM) satellites (and on Sentinel) (available at this website) can give high-quality estimates of wind speeds. Given the animation of full-disk imagery above (at ten-minute timesteps), taken from AWIPS and therefore sub-sampled down to 6-km horizontal resolution, how fast would you guess wind speeds might be near the surface under the convection. SAR data below shows winds near 25 m/s, i.e., storm-force winds, under the convection.

RCM2 SAR surface winds, 0526 UTC on 28 June 2021 (Click to enlarge)

A full-resolution image of GOES-17 clean window infrared data (10.3 µm) (created using Geo2Grid) is shown below, at the closest time to the observed winds shown above. The deepest cloud tops are close to the strongest surface wind speeds.

GOES-17 ABI Band 13 infrared (10.3 µm) imagery, 0530 UTC on 28 June 2021 (click to enlarge)