NUCAPS views Saharan Air over the Atlantic

September 20th, 2018 |

Saharan Air Layer Analysis over the Tropical Atlantic, 0600 UTC on 20 September 2018 (Click to enlarge)

There have been many episodes of Saharan Air over the tropical Atlantic within the past months, and another episode is in progress on 20 September. The Saharan Air Layer (SAL) analysis, above, from the CIMSS Tropical Weather website (Direct Link), shows dry air north and east of the Caribbean. The Clean Window ABI Band 13 (10.3 µm) Full-Disk ABI infrared imagery, below, from 0500 UTC, overlain with NUCAPS sounding points, shows where data were available from that morning overpass of Suomi NPP.

GOES-16 ABI 10.3 µm Infrared Imagery at 0500 UTC along with NUCAPS Sounding Points at approximately the same time (Click to enlarge)

The stepping animation below shows NUCAPS Soundings at a selection of points that starts north of the Saharan Air Layer and ends up within the SAL. The underlying figure is the Dust RGB from AWIPS, an RGB that combines the Split Window Difference (12.3 µm -10.3 µm; Red Component), Split Cloud Top Phase Brightness Temperature Difference (11.2 µm – 8.5 µm; Green Component) and 10.3 µm Infrared Imagery (Blue Component). Typically, regions with dust as might accompany a SAL have a pink tinge. The soundings are annotated to include Total Precipitable Water measurements, and mid-level Relative humidity. NUCAPS soundings identify the region where the SAL is present.

Dust RGB at 0433 UTC north and east of the Caribbean, and NUCAPS Soundings at selected points along a transect (Click to enlarge)

The SAL air continued its movement west during the day on 20 September.  The toggle below shows the Dust RGB, ABI Band 3 (0.86 µm) and the Baseline Aerosol Detection Product (in blue) at about the same time as the afternoon NUCAPS Sounding overpass (from Suomi NPP).  Suomi NPP overflew the eastern half of the SAL air (the overpass from NOAA-20 was more centered on the SAL air approaching the Caribbean, but NOAA-20 NUCAPS soundings are not yet in AWIPS;  they should be by the end of the year).

GOES-16 ABI Dust RGB, “Veggie Band” (Near-Infrared at 0.86 µm), and Baseline Aerosol Detection Product (Blue points), 1615 UTC on 20 September 2018 (Click to enlarge)

NUCAPS Soundings at 3 points (North of the SAL, within the SAL, and south of the SAL), below, show much different thermodynamics within the SAL.

NUCAPS Profiles at ~1600 UTC on 20 September 2018 at three locations as noted (Click to enlarge)

NOAA’s G-IV flew through this outbreak, deploying dropsondes to sample the event. The path of the aircraft (with the dropsonde locations) is here. Sonde #26, below, in the heart of the SAL, is shown below, with a nearby NOAA-20 NUCAPS sounding. (Flight path and Sonde imagery courtesy Chris Barnet, STC/NOAA) Refer to the caption for details.  Recall that the Dropsonde shows values at a point.  The NUCAPS profile is sampling a volume that is approximately a 50-km cylinder!  There is nevertheless excellent agreement.

Dropsonde #26 data (raw data in light grey; values averaged into the 100 NUCAPS vertical layers in black); GFS sounding in magenta. NUCAPS Microwave-only sounding in green; NUCAPS Microwave and infrared retrieval (as might be seen in AWIPS) in Red. Time offset from the Dropsonde is noted (Click to enlarge)

SAL outbreaks cause a significant deterioration in air quality over the Caribbean. The image below, courtesy Ernesto Rodriguez, SOO for the National Weather Service office in San Juan, Puerto Rico, compares Air Quality before and during a SAL outbreak in July, and during the current outbreak.

The view outside of the National Weather Service office in San Juan on 20 September and 13 July 2018 (during SAL outbreaks) and on 12 July 2018 (before a SAL outbreak). Imagery courtesy Ernesto Rodriguez, NWS SJU.

Wildfires in British Columbia

August 17th, 2018 |

GOES-16

GOES-16 “Red” Visible (0.64 µm, left) and Shortwave Infrared (3.9 µm, right) images [click to play MP4 animation]

A 2-panel comparison of GOES-16 (GOES-East) “Red” Visible (0.64 µm) and Shortwave Infrared (3.9 µm) images (above) showed the smoke plumes and thermal anomalies or “hot spots” (darker black to red pixels) associated with a flare-up of wildfires in western British Columbia on 17 August 2018.

A sequence of Shortwave Infrared (3.7 µm) images from Terra / Aqua MODIS and Suomi NPP / NOAA-20 VIIRS (below) revealed the diurnal changes in areal coverage and intensity of the thermal signature of the fires.

Shortwave Infrared (3.7 µm) images from Terra / Aqua MODIS and Suomi NPP / NOAA-20 VIIRS [click to enlarge]

Shortwave Infrared (3.7 µm) images from Terra / Aqua MODIS and Suomi NPP / NOAA-20 VIIRS [click to enlarge]

Toggles between Visible and Shortwave Infrared images from Terra MODIS (1912 UTC), NOAA-20 VIIRS (1950 UTC) ans Suomi NPP VIIRS (2129 UTC) are shown below (note: the NOAA-20 images are incorrectly labeled as Suomi NPP). It is interesting to note the impact that the smoke plume had on the air temperature at Quesnel (CYQZ) — because the smoke layer was optically dense enough (VIIRS True Color image) to significantly reduce incoming solar radiation, the temperature was as much as 14-18ºF (8-10ºC) cooler than Prince George (CYXS) to the north and Williams Lake (CYWL) to the south.

Terra MODIS Visible (0.65 µm) and Shortwave Infrared (3.7 µm) images [click to enlarge]

Terra MODIS Visible (0.65 µm) and Shortwave Infrared (3.7 µm) images at 1912 UTC [click to enlarge]

NOAA-20 VIIRS Visible (0.64 µm) and Shortwave Infrared (3.74 µm) images [click to enlarge]

NOAA-20 VIIRS Visible (0.64 µm) and Shortwave Infrared (3.74 µm) images at 1950 UTC [click to enlarge]

Suomi NPP VIIRS Visible (0.64 µm) and Shortwave Infrared (3.74 µm) images at 2129 UTC [click to enlarge]

Suomi NPP VIIRS Visible (0.64 µm) and Shortwave Infrared (3.74 µm) images at 2129 UTC [click to enlarge]

Farther to the east in Alberta, thick smoke caused very poor air quality in cities like Edmonton and Grande Prairie (photo 1 | photo 2). Daily composites of Suomi NPP VIIRS True Color RGB images from 11 August to 17 August (below) revealed the transport of smoke across British Columbia, Alberta and Saskatchewan.

Daily composites of Suomi NPP VIIRS True Color RGB images (with VIIRS fire detections in red), 11-17 August [click to play MP4 | Animated GIF]

Daily composites of Suomi NPP VIIRS True Color RGB images (with VIIRS fire detections in red), 11-17 August [click to play MP4 | Animated GIF]

A time series of surface reports from Edmonton, Alberta covering the period 14-17 August (below) showed that smoke restricted the surface visibility there to 1.5 miles on 15 August and 17 August.

Time series of surface reports from Edmonton, Alberta during the period 14-17 August [click to enlarge]

Time series of surface reports from Edmonton, Alberta during the period 14-17 August [click to enlarge]

===== 19 August Update =====

* GOES-17 images shown here are preliminary and non-operational *

GOES-17 Near-Infrared

GOES-17 Near-Infrared “Cloud Particle Size” (2.24 µm, left) and Shortwave Infrared (3.9 µm, right) images [click to play 81 Mbyte MP4 animation]

A 2-panel comparison of GOES-17 Near-Infrared “Cloud Particle Size” (2.24 µm) and Shortwave Infrared (3.9 µm) images during the 7-day period of 13-19 August (above) showed the diurnal changes in thermal signatures of the ongoing British Columbia wildfires. The nighttime thermal signatures seen on the 2.24 µm images (brighter white pixels) result from the fact that this spectral band is located close to the peak emitted radiance of very hot features such as active volcanoes or large fires (below).

Plots of Spectral Response Functions for ABI Bands 5, 6 and 7 [click to enlarge]

Plots of Spectral Response Functions for ABI Bands 5, 6 and 7 [click to enlarge]

Wildfire smoke across the Midwestern US

August 11th, 2018 |

GOES-16 Natural Color RGB images [click to play MP4 animation]

GOES-16 Natural Color RGB images, 09-11 August [click to play MP4 animation]

Numerous wildfires burning in southwestern Canada (primarily British Columbia: NOAA HMS fire/smoke product) produced large amounts of smoke, which was subsequently transported eastward across southern Canada and then southward across the Midwestern US during the 09 August11 August 2018 period. GOES-16 (GOES-East) Natural Color Red-Green-Blue (RGB) images from the AOS site (above) showed this smoke, portions of which were optically very thick at times (and were able to cast shadows owing to its significant vertical depth).

On 09 August the smoke was most highly concentrated over the Dakotas, as shown in a comparison of GOES-16 Aerosol Optical Depth (AOD), Smoke Detection, “Blue” Visible (0.47 µm) and “Red” Visible (0.64 µm) images (below). While much of the smoke was likely aloft within the middle troposphere, some had been mixed downward into the boundary layer and was restricting the surface visibility to 3-5 miles at many locations.

Note that the hazy signature of the widespread smoke was a bit more apparent in the 0.47 µm Visible imagery than the 0.64 µm Visible imagery, especially during mid-day when the sun-satellite “forward scattering angle” was at a minimum. The AOD and Smoke Detection derived products use data from Visible and Near-Infrared bands — so it they are only available during daytime hours (and only at solar zenith angles less than 60 degrees). The Smoke Detection product was more effective during times of enhanced forward scattering (early and late in the day) — but it also was susceptible to false alarms due to solar reflectance off water surfaces. Additional information on GOES-R Aerosol Detection Products in AWIPS is available here and here.

GOES-16 Aerosol Optical Depth (top left), Smoke Detection product (top right).

GOES-16 Aerosol Optical Depth (top left), Smoke Detection product (top right). “Blue” Visible (0.47 µm, bottom left) and “Red” Visible (0.64 µm, bottom right) [click to play animation | MP4]

On 10 August, the smoke was most dense across the eastern Dakotas and Minnesota (below) — and once again, surface visibilities were restricted to 3-5 miles at some locations. On this day pilot reports mentioned flight visibility being restricted to 3 miles at altitudes as high as 12,000 feet.

GOES-16 Aerosol Optical Depth (top left), Smoke Detection product (top right). "Blue" Visible (0.47 µm, bottom left) and "Red" Visible (0.64 µm, bottom right) [click to play animation | MP4]

GOES-16 Aerosol Optical Depth (top left), Smoke Detection product (top right). “Blue” Visible (0.47 µm, bottom left) and “Red” Visible (0.64 µm, bottom right) [click to play animation | MP4]

Finally, on 11 August a north-to-south plume of particularly dense smoke drifted southward across Minnesota and Iowa, as seen in a comparison of GOES-16 Aerosol Optical Depth, “Red” Visible (0.64 µm). Near-Infrared “Cirrus” (1.37 µm) and “Clean” Infrared Window (10.3 µm) images (below). In this case the AOD values were quite high (in excess of 3.0 in northwestern Minnesota), beyond the range of values scaled for display in AWIPS — this led to the swath of black “No Data” values where the smoke was most dense. This plume of thick smoke also exhibited a signature in Near-Infrared “Cirrus” images; higher concentrations of airborne particles that are effective scatterers of light at the 1.37 µm wavelength (such as ice crystals, smoke, volcanic ash, or dust) will be detected using this imagery. Note the lack of a well defined signature on the 10.3 µm imagery — smoke is effectively transparent to radiation at these longer infrared wavelengths.

GOES-16 Aerosol Optical Depth (top left), "Red" Visible (0.64 µm, top right). Near-Infrared "Cirrus" (1.37 µm, bottom left) and "Clean" Infrared Window (10.3 µm, bottom right) [click to play animation | MP4]

GOES-16 Aerosol Optical Depth (top left), “Red” Visible (0.64 µm, top right). Near-Infrared “Cirrus” (1.37 µm, bottom left) and “Clean” Infrared Window (10.3 µm, bottom right) [click to play animation | MP4]

On a side note, the north-south plume of dense smoke over southcentral Canada and the Midwest US on 11 August was also very apparent from a distance of 983,269 miles (1,582,418.07 km) — 44 times the distance of the GOES-16 satellite — in EPIC Natural Color imagery from the DSCOVR satellite (below).

DSCOVR EPIC Natural Color images [click to enlarge]

DSCOVR EPIC Natural Color images [click to enlarge]

Smoke from Mendocino Complex fires in California

August 4th, 2018 |

GOES-16

GOES-16 “Red” Visible (0.64 µm, left) and Shortwave Infrared (3.9 µm, right) images, with hourly plots of surface observations [click to play animation | MP4]

GOES-16 (GOES-East) “Red” Visible (0.64 µm) and Shortwave Infrared (3.9 µm) images (above) showed the smoke and thermal anomalies or “hot spots” (red pixels) associated with the Mendocino Complex burning in Northern California on 04 August 2018. Smoke was reducing the surface visibility to 2.5 miles at nearby Sacramento International Airport KSMF and Marysville KMYV. As of 7pm local time on 04 August the Mendocino Complex had burned 229,000 acres.

A 30-meter resolution Landsat-8 False Color Red-Green-Blue (RGB) image viewed using RealEarth (below) showed active burning along the eastern edge of the Ranch Fire (part of the Mendocino Complex) at 1845 UTC. The larger fire was producing a pyrocumulus cloud in addition to the dense smoke plume drifting northeastward.

Landsat-8 False Color image [click to enlarge]

Landsat-8 False Color RGB image [click to enlarge]

GOES-16 Upper-level (6.2 µm), Mid-level (6.9 µm) and Low-level (7.3 µm) Water Vapor images (below) revealed a southwest-to-northeast oriented band of moisture and fast flow associated with a middle to upper-tropospheric jet streak that was moving over the region (300 hPa analyses). “Red” Visible (0.64 µm) images showed the smoke plume drifting rapidly northeastward over California and Nevada, and visible Derived Motion Winds — which are calculated for pressure levels at and below 700 hPa —  tracked the smoke moving as fast as 58 knots at 2337 UTC. This speed was faster than 00 UTC winds at or below 700 hPa on rawinsonde data from either Oakland KOAK or Reno KREV.

GOES-16 Upper-level (6.2 µm, top left), Mid-level (6.9 µm, top right), Low-level (7.3 µm, bottom left) Water Vapor and "Red" Visible with Derived Motion Winds (0.64 µm, bottom right) [click to play MP4 animation]

GOES-16 Upper-level (6.2 µm, top left), Mid-level (6.9 µm, top right), Low-level (7.3 µm, bottom left) Water Vapor images and “Red” Visible (0.64 µm, bottom right) images with Derived Motion Winds [click to play MP4 animation]

===== 07 August Update =====

NOAA-20 VIIRS Day/Night Band (0.7 µm), Near-Infrared (1.61 µm and 2.25 µm) and Shortwave Infrared (3.75 µm) images [click to enlarge]

NOAA-20 VIIRS Day/Night Band (0.7 µm), Near-Infrared (1.61 µm and 2.25 µm) and Shortwave Infrared (3.75 µm) images [click to enlarge]

A comparison of NOAA-20 VIIRS Day/Night Band (0.7 µm), Near-Infrared (1.61 µm and 2.25 µm) and Shortwave Infrared (3.75 µm) images (above; courtesy of William Straka, CIMSS) showed the nighttime glow and thermal signatures of the Mendocino Complex fires on 07 August 2018. As of 8:30am the fire had burned over 290,000 acres, becoming the largest wildfire on record in the state of California.