Category Archives: Winter weather

Ice dam in Lake Erie

GOES-16

GOES-16 “Red” Visible (0.64 µm) images, with hourly surface wind barbs plotted in yellow and wind gusts (knots) plotted in cyan [click to play animation]

Thanks to Dave Zaff (NWS Buffalo) for the email alerting us to an ice dam that had formed across the eastern portion of Lake Erie on 19 January 2018 — GOES-16 (GOES-East) “Red” Visible (0.64 µm) images (above) showed that the northeastward drift of ice floes was effectively being blocked by this ice dam feature.

A toggle between 250-meter resolution Terra MODIS True-color and False-color Red-Green-Blue (RGB) images from the MODIS Today site (below) provided a more detailed view of the Lake Erie ice dam and upwind drift ice at 1615 UTC. Snow and ice appear as shades of cyan in the False-color image, in contrast to supercooled water droplet clouds which are shades of white.

Terra MODIS True-color and False-color RGB images [click to enlarge]

Terra MODIS True-color and False-color RGB images; red arrows denote the location of the ice dam [click to enlarge]

The Terra MODIS Visible (0.65 µm) image with an overlay of RTMA surface winds (below) showed the southwesterly flow across the long axis of the lake.

Terra MODIS Visible (0.65 µm) image with surface METAR reports and RTMA surface winds [click to enlarge]

Terra MODIS Visible (0.65 µm) image with surface METAR reports and RTMA surface winds [click to enlarge]

A toggle between 1607 UTC Terra MODIS and 1757 UTC Suomi NPP VIIRS Visible images (below) showed the motion of the lake drift ice during that time period.

Terra MODIS and Suomi NPP VIIRS Visible images, with METAR surface reports [click to enlarge]

Terra MODIS and Suomi NPP VIIRS Visible images, with METAR surface reports [click to enlarge]

Blowing snow in North Dakota and Minnesota

GOES-16

1-minute GOES-16 “Red” Visible (0.64 µm, left) and Near-Infrared “Snow/Ice” (1.61 µm, right) images, with plots of hourly surface wind barbs in cyan and surface weather type in yellow [click to play MP4 animation]

Several inches of new snow followed by strong northerly winds led to widespread blizzard conditions across the Red River Valley of North Dakota and Minnesota on 11 January 2018 (NWS Grand Forks summary). A GOES-16 (GOES-East) Mesoscale Sector had been positioned over the Upper Midwest to monitor the winter storm, providing images at 1-minute intervals — and a comparison of “Red” Visible (0.64 µm) and Near-Infrared “Snow/Ice” (1.61 µm) images (above) showed the development of horizontal convective rolls that are a common feature associated with blowing snow.

Ice floes in Chesapeake Bay

Landsat-8 false-color RGB image [click to enlarge]

Landsat-8 false-color RGB image [click to enlarge]

In the wake of the explosive cyclogenesis off the East Coast of the US on 04 January 2018, very cold air began to spread across much of the eastern half of the Lower 48 states. Focusing on the Hampton Roads area of southeastern Virginia, satellite imagery began to show the formation of ice in the rivers and bays. On 06 January, a 30-meter resolution Landsat-8 false-color Red-Green-Blue (RGB) image viewed using RealEarth (above) revealed some of this ice — in particular, long narrow ice floes (snow and ice appear as shades of cyan) that likely emerged from the Back River (northeast of Hampton) and were drifting northward and southward just off the coast of the Virginia Peninsula.

On the following day (07 January), 250-meter resolution Terra MODIS true-color and false-color RGB images from the MODIS Today site (below) showed that a larger V-shaped ice floe was located just southeast of the Peninsula, with its vertex pointed toward the Hampton Roads Bridge-Tunnel (HRBT). Snow and ice also appear as shades of cyan in the MODIS false-color image.

Terra MODIS true-color and false-color RGB images [click to enlarge]

Terra MODIS true-color and false-color RGB images [click to enlarge]

07 January also happened to be the last full day of imagery to be broadcast by the GOES-13 satellite — a comparison of 1-minute Mesoscale Sector GOES-16 (GOES-East) Visible (0.64 µm) and 15-30 minute interval GOES-13 Visible (0.63 µm) images (below) showed that the V-shaped ice floe continued to drift southwestward toward the HRBT. However, it was difficult to tell whether the ice feature made it over and past the tunnel; even with the improved GOES-16 Visible spatial resolution (0.5 km at satellite sub-point, compared to 1.0 km for GOES-13) and the 1-minute rapid image scans, the ice floe became harder to track during the afternoon hours before high clouds began to overspread the region.

"GOES-16

GOES-16 Visible (0.64 µm, left) and GOES-13 Visible (0.63 µm, right) images, with hourly surface air temperatures (ºF) plotted in yellow [click to play MP4 animation]

However, a close examination of Suomi NPP VIIRS true-color and false-color images at 1826 UTC (below) indicated that some of the ice had indeed moved westward past Fort Monroe (on the far southeastern tip of the Peninsula) and over/past the HRBT.

Suomi NPP VIIRS true-color and false-color RGB images [click to enlarge]

Suomi NPP VIIRS true-color and false-color RGB images [click to enlarge]

On the topic of cold temperatures in southeastern Virginia, a new daily record low of -3 ºF was set at Richmond on the morning of 07 January, and at Norfolk new daily record low and record low maximum temperatures were set (10 ºF and 23 ºF, respectively).

Nighttime views of lake effect snow bands over Lake Superior

Suomi NPP VIIRS Day/Night Band (0.7 µm) images, with morning minimum temperatures at Embarrass, Minnesota [click to enlarge]

Suomi NPP VIIRS Day/Night Band (0.7 µm) images, with morning minimum temperatures at Embarrass, Minnesota [click to enlarge]

Shown above are detailed nighttime views of multiple lake effect snow (LES) bands over Lake Superior, provided by Suomi NPP VIIRS Day/Night Band (0.7 µm) images on 04 January, 05 January and 06 January 2018. These “visible images at night” were possible due to ample illumination by the Moon, which was in the Waning Gibbous phase (at 92% of Full on 04 January, 84% of Full on 05 January and 75% of Full on 06 January). The continued flow of arctic air across the still-unfrozen waters of Lake Superior (and the other unfrozen Great Lakes) was responsible for the formation of these and a variety of other LES bands.

For perspective, the daily morning minimum temperatures at Embarrass, Minnesota are also plotted on the images — on these 3 days Embarrass was the coldest official site in the US (including Alaska).

The VIIRS images were captured by the Space Science and Engineering Center direct broadcast ground station.

Explosive cyclogenesis off the East Coast of the United States

GOES-16 Clean Window (10.3 µm) Imagery, 0102-1337 UTC on 4 January 2018 (Click to animate)

A strong extratropical cyclone that deposited snow in the deep south developed explosively during the early morning hours of 4 January 2018. The GOES-16 Clean Window (10.3 µm) animation, above, from 0102 – 1337 UTC on 4 January, brackets the explosive development: from 993 hPa at 0000 UTC to 968 mb at 0900 UTC, a strengthening that easily meets the “Bomb” criteria set forth by Sanders and Gyakum (1980). The Clean Window animation shows the strong surface circulation with well-defined conveyor belts. Convection develops at the leading edge of the dry slot that is approaching southern New England at the end of the animation. The Low-Level Water Vapor (7.3 µm) animation for the same time, below, suggests very strong descent behind the storm, where brightness temperatures warmer than -10º C (orange in the enhancement used) are widespread.

GOES-16 Low-Level Water Vapor (7.3 µm) Infrared Imagery, 0102-1332 UTC on 4 January 2018 (Click to animate)

This storm can also be viewed using Red-Green-Blue composites (in addition to the single-channel animations shown above). The Airmass RGB, below, combines the Split Water Vapor Difference (6.2 µm – 7.3 µm) as Red, Split Ozone (9.6 µm – 10.3 µm) as Green, and Upper level Water Vapor (6.2 µm) as Blue. (Other storms analyzed with the Airmass RGB can be seen here, here, and here). The strong red signal in the Airmass RGB south of the storm suggests very strong sinking motion.

GOES-16 AirMass RGB Product, 0102-1332 UTC (Click to animate)

ASCAT Scatterometer winds over the system at 0205 UTC showed an elongated surface circulation with multiple observations of winds exceeding 50 knots (in red), and a large region (in yellow) of winds exceeding 35 knots.

GOES-16 ABI Clean Window (10.3 µm) and ASCAT Scatterometer winds, 0205 UTC on 4 January 2018 (Click to enlarge)

GOES-16 ABI Red Visible (0.64 µm) and ASCAT Scatterometer winds, 1520 UTC on 4 January 2018 (Click to enlarge)

The 1520 UTC ASCAT pass, above, sampled half the storm, and hurricane-force winds were indicated.

The snow that was deposited in the Deep South by this storm (also discussed here) persisted through a cold night and was visible in the GOES-16 Visible (0.64 µm) imagery, below. Highly reflective snow can be difficult in a still image to distinguish from clouds — but the Snow/Ice Channel on GOES-16 (1.61 µm) detects energy at a wavelength that is strongly absorbed by ice. Thus, snow (and ice) on the ground (or in clouds), has a different representation. (Here are toggles between the two images, with and without a map). The snow cover over coastal Georgia, South and North Carolina appears dark in the Snow/Ice channel because the snow is absorbing, not reflecting, the 1.61 µm radiation.  It is noteworthy that the 1.61 µm image is especially dark over far southeastern Georgia northeastward along the immediate coastline of South Carolina.  These are regions where freezing rain and sleet fell, versus predominantly snow to the north and west (as also noted here; The National Weather Service in Tallahassee tweeted out an ice/snow accumulation map that also agrees with the 1.61 µm image).  Ice in the cirrus clouds northeast of North Carolina is also apparent in the Snow/Ice 1.61 µm imagery.

GOES-16 Band 2 Visible (0.64 µm) Imagery, 1412 UTC on 4 January 2018 (Click to enlarge)

GOES-16 ABI Band 5 Snow/Ice (1.61 µm) Imagery, 1412 UTC on 4 January 2018 (Click to enlarge)

Suomi NPP overflew the storm shortly after midnight on 4 January; Day Night band visible imagery (courtesy Kathleen Strabala, CIMSS), below, shows a well-developed cyclone covering much of the northeast Atlantic Ocean. Snow cover is apparent over the deep south of the United States.

Suomi NPP Day Night Band Visible (0.7 µm) Imagery, 0614 UTC on 4 January 2018 (Click to enlarge)

(Added, 5 January 2018: This website shows a during-the-day CIMSS True Color Image animation of the storm on 4 January 2018. Animation courtesy Dave Stettner, CIMSS).

Eastern US winter storm

GOES-16

GOES-16 “Red” Visible (0.64 µm, left) and Near-Infrared “Snow/Ice” (1.61 µm, right) images, with plots of hourly surface reports [click to play MP4 animation]

The initial impacts of a large Eastern US winter storm were seen in a comparison of GOES-16 (GOES-East) “Red” Visible (0.64 µm) and Near-Infrared “Snow/Ice” (1.61 µm) images (above) on 03 January 2018 — areas of southeastern Georgia received freezing rain and/or 1-6 inches of snowfall. As clouds began to dissipate, the resulting snow cover appeared bright on the Visible images (since fresh snow is highly reflective at the 0.64 µm wavelength), and darker shades of gray on the Near-Infrared images (since snow and ice are strong absorbers of radiation at the 1.61 µm wavelength). Note the brief appearance of a cloud plume streaming southward from the Hatch Nuclear Power Plant.

Earlier that morning, the Florida Panhandle also received snowfall (text | map), but the lighter accumulations there were insufficient to exhibit a good satellite signature.

In a toggle between Suomi NPP VIIRS true-color and false-color Red-Green-Blue (RGB) images from RealEarth (below), the deeper snow cover in Georgia appears as darker shades of cyan.

Suomi NPP VIIRS true-color and false-color images [click to enlarge]

Suomi NPP VIIRS true-color and false-color images [click to enlarge]

===== 04 January Update =====

Suomi NPP VIIRS Day/Night Band (0.7 µm) and Infrared Window (11.45 µm) images [click to enlarge]

Suomi NPP VIIRS Day/Night Band (0.7 µm) and Infrared Window (11.45 µm) images [click to enlarge]

A toggle between Suomi NPP VIIRS Day/Night Band (0.7 µm) and Infrared Window (11.45 µm) images at 0620 UTC (1:20 AM Eastern time) on 04 January (above; courtesy of William Straka, CIMSS) showed a nighttime view of the rapidly-intensifying storm when it had an estimated minimum central pressure of 972 hPa or 28.70″. Note the signature of snow cover — extending from southeastern Georgia across eastern portions of South Carolina and North Carolina — which is evident on the “visible image at night” Day/Night Band (made possible by ample illumination from the Moon, which was in the Waning Gibbous phase at 92% of Full). A full-resolution version of the Day/Night Band image is available here.

During the following daytime hours, 30-second interval Mesoscale Sector GOES-16 “Red” Visible (0.64 µm) images (below) showed the evolution of the low pressure center of circulation as it continued to rapidly intensify (surface analyses) off the US East Coast.

30-second GOES-16

30-second GOES-16 “Red” Visible (0.64 µm) images [click to play MP4 animation]

GOES-16

GOES-16 “Red” Visible (0.64 µm) images, with hourly surface weather type plotted in yellow [click to play MP4 animation]

A larger-scale view (using 5-minute CONUS sector data) of GOES-16 “Red” Visible (0.64 µm) images with hourly plots of surface weather (above) depicted the widespread precipitation associated with the storm. Similarly, plots of hourly wind gusts (below) portrayed the large wind field of the system. Some of the highest snowfall/ice accumulations and peak wind gusts are listed here and here.

GOES-16

GOES-16 “Red” Visible (0.64 µm) images, with hourly surface wind gusts plotted in yellow [click to play MP4 animation]

In the wake of the departing storm, the tropospheric column over Florida and the southeastern US was dry enough (3.7 mm or 0.15″ at Tallahassee FL and 4.0 mm or 0.16 ” at Charleston SC) to allow the GOES-16 Lower-level (7.3 µm) Water Vapor imagery (below) to detect the thermal contrast of surface land/water boundaries — portions of the coastline and a few of the larger inland lakes were evident.

"GOES-16

(7.3 µm, left), Mid-level (6.9 µm, center) and Upper-level (6.2 µm, right) images [click to play animation]” class=”size-medium” /> GOES-16 Lower-level (7.3 µm, left), Mid-level (6.9 µm, center) and Upper-level (6.2 µm, right) images [click to play animation]

A full-resolution Suomi NPP VIIRS true-color RGB image at 1738 UTC (below) revealed interesting storm features such its very large cloud shield and convection near the circulation center, as well as the swath of snow cover across parts of Georgia, South Carolina and North Carolina.

Suomi NPP VIIRS true-color RGB image [click to enlarge]

Suomi NPP VIIRS true-color RGB image [click to enlarge]

A toggle between the corresponding Suomi NPP VIIRS Visible (0.64 µm) and Snow/Ice RGB images (below) helped to highlight locations which received a significant accrual of ice from freezing rain– these areas show up as a darker shade of red on the Snow/Ice RGB image (along the southeastern edge of the swath of snow cover, which is a lighter shade of red). Notable ice accumulations included 0.50″ at Brunswick and Folkston GA, 0.25″ at Georgetown and Myrtle Beach SC, and 0.25″ at Kure Beach NC.

Suomi NPP VIIRS Visible (0.64 µm) and Snow/Ice RGB images, with surface station identifiers plotted in white [click to enlarge]

Suomi NPP VIIRS Visible (0.64 µm) and Snow/Ice RGB images, with surface station identifiers plotted in white [click to enlarge]

Finally, a 30-meter resolution Landsat-8 false-color RGB image viewed using RealEarth (below) showed the snow-covered Charleston, South Carolina area — areas with less dense trees and vegetation showed a more pronounced snow cover signature (shades of cyan). The Charleston International Airport remained closed, due to snow and ice-covered runways.

Landsat-8 false-color RGB image [click to enlarge]

Landsat-8 false-color RGB image [click to enlarge]

Additional imagery of this explosive cyclogenesis event can be found at this blog post.

Chesapeake Bay effect snow in North Carolina

Suomi NPP VIIRS Day/Night Band (0.7 µm) and Infrared Window (11.45 µm) images at 0710 UTC, with plots of 07 UTC surface reports [click to enlarge]

Suomi NPP VIIRS Day/Night Band (0.7 µm) and Infrared Window (11.45 µm) images at 0710 UTC, with plots of 07 UTC surface reports [click to enlarge]

Suomi NPP VIIRS Day/Night Band (0.7 µm) and Infrared Window (11.45 µm) images (above) showed a well-defined Chesapeake Bay “streamer” cloud  at 0710 UTC or 3:10 AM local time on 01 January 2018. This cloud feature resulted from the flow of unusually-cold air over the relatively warm water of the bay — a process identical to that which produces the more common “lake effect” cloud bands. With the benefit of ample illumination from a Full Moon, the “visible image at night” capability of the Day/Night Band was vividly illustrated (and a VIIRS instrument on the JPSS series of satellites — including the recently-launched NOAA-20 — will provide similar imagery).

During the subsequent daylight hours, 1-minute Mesoscale Sector GOES-16 (GOES-East) “Red” Visible (0.64 µm) images (below) showed the Chesapeake Bay streamer cloud moving southward. Note that this cloud produced light snow as far south as Currituck, North Carolina (KOTX) from 14-15 UTC or 10-11 AM local time. It is possible that some light snow also occurred across a portion of the Eastern Shore of Virginia and the Outer Banks of North Carolina, but verification is not possible due to the scarcity of surface observation sites in those areas.

1-minute GOES-16

1-minute GOES-16 “Red” Visible (0.64 µm) images, with hourly surface reports plotted in yellow [click to play MP4 animation]

Time series plot of surface weather conditions at Currituck, North Carolina [click to enlarge]

Time series plot of surface weather conditions at Currituck, North Carolina [click to enlarge]

Lake Michigan Mesovortex

1-minute GOES-16

1-minute GOES-16 “Red” Visible (0.64 µm) images, with hourly surface reports plotted in yellow [click to play MP4 animation]

GOES-16 (GOES-East) “Red” Visible (0.64 µm) images (above) showed a well-defined mesoscale vortex (or “mesovortex”) moving southward across southern Lake Michigan on 31 December 2017. The default western GOES-16 Mesoscale Sector provided images at 1-minute intervals. This feature was responsible for brief periods of heavy snow at locations such as South Haven, Michigan KLWA (beginning at 1455 UTC), Benton Harbor, Michigan KBEH (beginning at 1625 UTC) and La Porte, Indiana KPPO (from 2055 to 2115 UTC).

Comparisons of POES AVHRR/Terra MODIS/Suomi NPP Infrared (10.8 µm/11.0 µm/11.45 µm) and Visible (0.86 µm/0.65 µm/0.64 µm) images along with an overlay of the corresponding Real-Time Mesoscale Analysis (RTMA) surface winds (below) provided views of the mesovortex at 1522 UTC, 1714 UTC and 1852 UTC, respectively.

POES AVHRR Infrared (10.8 µm) and Visible (0.86 µm) images at 1522 UTC, with 15 UTC RTMA surface winds [click to enlarge]

POES AVHRR Infrared (10.8 µm) and Visible (0.86 µm) images at 1522 UTC, with 15 UTC RTMA surface winds [click to enlarge]

Terra MODIS Infrared (11.0 µm) and Visible (0.65 µm) images at 1714 UTC, with 17 UTC RTMA surface winds [click to enlarge]

Terra MODIS Infrared (11.0 µm) and Visible (0.65 µm) images at 1714 UTC, with 17 UTC RTMA surface winds [click to enlarge]

Suomi NPP Infrared (11.45 µm) and Visible (0.64 µm) images at 1852 UTC, with 19 UTC RTMA surface winds [click to enlarge]

Suomi NPP Infrared (11.45 µm) and Visible (0.64 µm) images at 1852 UTC, with 19 UTC RTMA surface winds [click to enlarge]

During the preceding nighttime hours, a comparison of Suomi NPP VIIRS Infrared (11.45 µm) and Day/Night Band (0.7 µm) images at 0729 UTC along with an overlay of 07 UTC RTMA surface winds (below) showed in spite of patchy thin cirrus clouds over the region, ample illumination from the Moon (which was in the Waxing Gibbous phase, at 96% of Full) enabled a signature of the early stage of mesovortex formation to be seen on the Day/Night Band (DNB) image. Ice crystals within the thin cirrus clouds were responsible for the significant scattering city light signatures on the DNB image.

Suomi NPP VIIRS Infrared (11.45 µm) and Day/Night Band (0.7 µm) images at 0729 UTC, with 07 UTC RTMA surface winds [click to enlarge]

Suomi NPP VIIRS Infrared (11.45 µm) and Day/Night Band (0.7 µm) images at 0729 UTC, with 07 UTC RTMA surface winds [click to enlarge]

As an aside, it is interesting to note that ice could be seen in the nearshore waters of Lake Michigan — both in the western part of the lake, off the coast of Wisconsin and Illinois, and in the eastern part of the lake off the coast of Lower Michigan. The lake ice appeared as darker shades of cyan in the 250-meter resolution Terra MODIS false-color (Band 7-2-1 combination) Red-Green-Blue (RGB) image from the MODIS Today site (below).

Terra MODIS true-color and false-color images over southern Lake Michigan [click to enlarge]

Terra MODIS true-color and false-color images over southern Lake Michigan [click to enlarge]

Mixed-phase stratiform clouds in an arctic air mass

AWIPS screen capture of GOES-16 Cloud Top Phase (top left), Near-Infrared

AWIPS screen capture of GOES-16 Cloud Top Phase product (top left), Near-Infrared “Snow/ice” (1.61 µm, top right), Cloud Phase brightness temperature difference (8.5 – 11.2 µm, bottom left) and “Clean” Infrared Window (10.3 µm, bottom right) images [click to enlarge]

An AWIPS screen capture showing GOES-16 (GOES-East) Cloud Top Phase, Near-Infrared “Snow/ice” (1.61 µm), Cloud Phase brightness temperature difference (8.5 µm11.2 µm) and “Clean” Infrared Window (10.3 µm) images on 28 December 2017 (above) was provided by Dan Baumgardt and Dave Schmidt (NWS La Crosse) — they were inquiring as to the why the 1.61 µm Snow/Ice imagery appeared bright across southern Minnesota (suggesting cloud tops composed primarily of supercooled water droplets), where light snow was being reported at a number of locations. Note that the Cloud Top Phase product also indicated that much of the stratus cloud deck over that same region was either Supercooled (light green) or Mixed (dark green).

An animation of GOES-16 Snow/Ice (1.61 µm) imagery (below) showed that the high reflectance (brighter white) signature of the lower-altitude stratiform cloud deck persisted across southern Minnesota into western Wisconsin and northern Iowa during the daylight hours, along with widespread surface reports of light snow. In contrast, higher-altitude clouds composed predominantly or entirely of ice crystals exhibited a darker gray appearance (since ice crystals, as well as surface snow cover and frozen lakes/rivers, are strong absorbers of radiation at the 1.61 µm wavelength).

GOES-16 Near-Infrared "Snow/Ice" (1.61 µm) images, with hourly surface-observed precipitation type plotted in yellow [click to play MP4 animation]

GOES-16 Near-Infrared “Snow/Ice” (1.61 µm) images, with hourly surface-observed precipitation type plotted in yellow [click to play MP4 animation]

In the corresponding GOES-16 “Clean” Infrared Window (10.3 µm) animation (below), much of the aforementioned lower-altitude stratiform cloud layer exhibited cloud-top infrared brightness temperatures in the -10 to -20 ºC range across far southern Minnesota into northern Iowa, with colder -20 to -30 ºC values seen in the more northern and eastern portion of the stratus cloud.

GOES-16 "Clean" Infrared Window (10.3 µm) images, with hourly surface-observed precipitation type plotted in yellow [click to play MP4 animation]

GOES-16 “Clean” Infrared Window (10.3 µm) images, with hourly surface-observed precipitation type plotted in yellow [click to play MP4 animation]

Plots of rawinsonde data (at 12 UTC on 28 December) from Aberdeen, South Dakota and Chanhassen, Minnesota (below) showed that the temperature profiles within the low-altitude cloud layers were close to isothermal, with air temperatures generally in the -16 to -22 ºC range.

Rawinsonde data from Aberdeen, South Dakota [click to enlarge]

Rawinsonde data from Aberdeen, South Dakota [click to enlarge]

Rawinsonde data from Chanhassen, Minnesota [click to enlarge]

Rawinsonde data from Chanhassen, Minnesota [click to enlarge]

So how could snow be falling from stratus clouds whose tops appeared be be composed of supercooled water droplets? A journal article titled “Vertical Motions in Arctic Mixed-Phase Stratiform Clouds” demonstrated that in-cloud glaciation can and does occur below the supercooled liquid cloud top in an arctic air mass. This example certainly shows that in an arctic air mass, mixed/supercooled cloud above snow or ice cloud is possible, particularly in temperatures between -20 ºC and -30 ºC — and cloud phase classification for operational decisions must sometimes look beyond the examination of single-band satellite imagery (or even derived products such as Cloud Phase).

Thanks to Mike Pavolonis (NOAA/NESDIS/CIMSS) and Jordan Gerth (CIMSS) for their insightful explanations regarding cloud phase — and thanks to the NWS La Crosse staff for bringing this interesting case to our attention!

Record-setting lake effect snow event at Erie, Pennsylvania

1-minute GOES-16 "Clean" Infrared Window (10.3 µm) images, with hourly surface reports plotted in cyan/yellow [click to play MP4 animation]

1-minute GOES-16 “Clean” Infrared Window (10.3 µm) images, with hourly surface reports plotted in cyan/yellow [click to play MP4 animation]

GOES-16 “Clean” Infrared Window (10.3 µm) images centered over Lake Erie (above) showed the evolution of lake effect snow bands on 25 December26 December 2017, which produced very heavy snowfall at locations such as Erie, Pennsylvania (station identifier KERI); a Mesoscale Sector provided images at 1-minute intervals. Some noteworthy snowfall records were set at Erie PA:

(27 December Update: additional lake effect snow at Erie on 27 December brought the final storm total accumulation to 65.1 inches: NWS Cleveland summary. NOHRSC plots showed a maximum snow depth of 49 inches just southwest of downtown Erie; the maximum snow depth at Erie International Airport was 28 inches on 26 December, which was still less than their all-time record snow depth of 39 inches on 21 December 1989)

A sequence of Infrared Window images captured by Terra/Aqua MODIS (11.0 µm) and Suomi NPP VIIRS (11.45 µm) is shown below. The coldest cloud-top infrared brightness temperatures associated with the dominant lake effect snow bands were in the -30 to -35 ºC range (dark blue to pale green color enhancement), similar to what was seen in the GOES-16 Infrared Window imagery.

Infrared Window images from Terra/Aqua MODIS (11.0 µm) and Suomi NPP VIIRS (11.45 µm), with surface reports plotted in yellow [click to enlarge]

Infrared Window images from Terra/Aqua MODIS (11.0 µm) and Suomi NPP VIIRS (11.45 µm), with surface reports plotted in yellow [click to enlarge]

Farther to the northeast, these Lake Erie lake effect bands also produced significant snowfall in far southwestern New York, with 32 inches reported at Perrysburg (located 20 miles west of Dunkirk, station identifier KDKK). In addition, lake effect snow bands over Lake Ontario were responsible for even higher snowfall amounts:


1-minute GOES-16 “Red” Visible (0.64 µm) images (below) showed the lake effect snow bands over Lake Ontario on 26 December.

1-minute GOES-16 "Red" Visible (0.64 µm) images, with hourly surface reports plotted in yellow [click to play MP4 animation]

1-minute GOES-16 “Red” Visible (0.64 µm) images, with hourly surface reports plotted in yellow [click to play MP4 animation]