Daily imagery from the CSPP Geosphere website above, at 0400 UTC from 12-19 September, show haze (highlighted by the blue arrows) moving across the Pacific Ocean. The haze is likely a by-product of the ongoing eruption at Kilauea. The initial burst of haze moved beyond the view of GOES-West by 19 September, but imagery shows other regions of haze are following behind. The haze has moved into the western Pacific, and has affected visibility in the northern Marianas Islands, as noted in the Forecast Discussion from the National Weather Service in Guam, shown below.

Multi-spectral satellite imagery from earlier this morning showed a
narrow band of haze reaching the Marianas along easterly trade-wind
flow pointing back to the Hawaii region. This haze has occasionally
dropped visibility at the Saipan airport earlier this morning, but
appears to have improved slightly in the afternoon hours. It is
likely that this haze originated from the recent Kilauea activity
from September 10th-16th of last week. High uncertainty remains to
exactly for how long this haze might persist, but it is expected to
be perceivable through at least Wednesday morning.

The image below, from the Guam NWS Facebook page, shows Garapan on Saipan.

---

The progression of haze from Hawaii to the Marianas is more easily viewed in a map that includes both Hawaii and Guam. Daily views (all at 0400 UTC as in the animation above) of GOES-18 Band 1 (0.47 µm) imagery show the haze approaching the Marianas by 0400 UTC on 18 September.

The toggle below of GOES-18 and Himawari-9 Band 1 imagery, both at 0400 UTC on 17 September, shows the importance of view angle relative to the Sun in detecting the presence of haze. Haze that is apparent in GOES-18 imagery (near 160^oE) is not apparent in Himawari-9 imagery because of differences in view angle and the location of the Sun relative to the satellite.

_________________________________________

JMA Himawari-9 True Color RGB images created using Geo2Grid (above) helped to highlight the westward transport of hazy volcanic smog (vog) from Hawai`i to the Mariana Islands during the 13-18 September time period. A plot of surface report data from Saipan Island (below) indicated that the surface visibility dropped as low as 5-6 miles (from 2100-2200 UTC on 18 September, or 7-8 AM ChST on 19 September) when the vog arrived. The satellite imagery was certainly convincing, but HYSPLIT model back trajectories also helped to implicate Kilauea vog as the source of haziness observed at Saipan (with subsidence occurring during the period of long-range transport).

---

Thanks to Brandon Aydlett, the Science and Operations Officer at Guam, for the alert on the haze.

View only this post Read Less

The GXI (GeoXO Imager) instrument to be flown on the GeoXO (Geostationary eXtended Observations) satellite, scheduled for launch in the mid-2030s, is intended to replace ABI data from the current set of GOES-R Satellites. One of the novel channels to be added to GXI senses infrared information at 5.15 µm, a wavelength at which absorption by water vapor occurs. As the weighting functions below show (source: this paper by Milller et al.), the wavelength is sensing information from far down in the atmosphere (in clear skies), much closer to the surface than the current water vapor infrared channels (bands 8-10, 6.19 µm, 6.95 µm and 7.34 µm) on the ABI (as shown here for a mid-latitude Summer atmosphere).

The animation below show simulated 5.15 µm and 7.3 µm imagery for a case in 2019. These 5.15 µm and 7.34 µm imagery are simulated using the Pressure layer Fast Algorithm for Atmospheric Transmittances (PFAAST) model applied to a 1-km ECMWF nature run (XNR1K). The spatial resolution is 2 km and imagery is courtesy of Zhenglong Li. Pay especial attention to the gradients in the low-level field (5.15 µm) at around 1215 UTC over the Gulf of Mexico and around 1545 UTC over the lower Mississippi River — regions highlighted by boxes. Note how such features are absent in the 7.3 µm, the water vapor channel on the ABI that senses infrared information from lowest down in the atmosphere. Convection subsequently develops within those boxes along these gradients that are apparent at 5.15 µm before they are at 7.34 µm. A conclusion might be: data from GXI will likely allow a forecaster to determine earlier where convection might subsequently develop.

View only this post Read Less

11 hours of 1-minute Mesoscale Domain Sector GOES-16 (GOES-East) “Red” Visible (0.64 µm) images (above) showed Post-Tropical Cyclone Lee as it approached Atlantic Canada and Maine on 16 September 2023. PTC Lee made landfall in far western Nova Scotia around 2000 UTC, and produced strong winds and heavy rainfall across much of Maine.

A longer animation of 5-minute CONUS Sector GOES-16 Air Mass RGB images (below) portrayed the large size of Lee, which transitioned from a Category 1 Hurricane to a Post-Tropical Cyclone by 0900 UTC. Shades of orange in the RGB imagery indicated that dry air had wrapped into the southern and eastern quadrants of Lee.

Hourly MIMIC Total Precipitable Water (TPW) images (below) showed the moisture associated with Lee as it moved northward across parts of New England and the Canadian Maritimes. Darker shades of red represented TPW values of 2.5 to 3.0 inches.

A plot of rawinsonde data (source) from Caribou, Maine KCAR at 0000 UTC on 17 September (below) indicated that the TPW value of that very moist sounding was 1.51 inches — which was not much less than the record value for all 17 September / 0000 UTC soundings (source) for Caribou, which was 1.63 inches (bottom).

View only this post Read Less

10-minute GOES-18 (GOES-West) “Red” Visible (0.64 µm) + Fire Power derived product (a component of the GOES Fire Detection and Characterization Algorithm FDCA), Shortwave Infrared (3.9 µm), “Clean” Infrared Window (10.3 µm) and Day Land Cloud Fire RGB images (above) showed a wildfire east of Fort Nelson (CYYE) in far northeastern British Columbia that produced 3 consecutive pyrocumulonimbus (pyroCb) cloud pulses late in the day on 15 September 2023. This wildfire burned very hot — 3.9 µm shortwave infrared brightness temperatures reached 137.88ºC (the saturation temperature of GOES-18 ABI Band 7 detectors) at 2350 UTC, with a Fire Power value at that time of 6282.52 MW (below).

A closer look at GOES-18 Shortwave Infrared images (below) showed the rapid east-southeastward run of the pyroCb-producing British Columbia wildfire. For nearly 4 hours, the peak 3.9 µm brightness temperature of that large fire remained at 137.88ºC (from 2350 UTC on 15 September to 0340 UTC on 16 September).

In a longer animation of GOES-18 Infrared Window images covering a larger area (below), a total of 5 distinct pyroCb clouds were produced by the British Columbia wildfire — while farther to the east, another fire in northwestern Alberta (located just south of Rainbow Lake, CWSH) later produced 2 pyroCb clouds. Strong surface winds (with gusts as high as 35 knots at Fort Nelson) associated with the approach and passage of a cold front likely played a role in intensifying wildfire behavior that led to these pyroCb formations.

The first (and largest) of the Alberta pyroCb clouds passed near/over the High Level Airport (CYOJ), where a thunderstorm was reported from 0603-0652 UTC (below). In addition, smoke from the nearby wildfire restricted surface visibility at CYOJ to 2-1/4 miles at one point.

A toggle between VIIRS Infrared Window (11.45 µm) images from NOAA-20 at 0828 UTC and Suomi-NPP at 0919 UTC (below) showed the northward transport of the first (large, elongated) Alberta pyroCb as it crossed over the Alberta / Northwest Territories border just ahead of the approaching cold front. The second (much smaller) Alberta pyroCb was apparent in the earlier NOAA-20 image, just behind the cold front. VIIRS data used to create those 2 images were downloaded and processed by the CIMSS/SSEC Direct Broadcast ground station.

____________________________________________

After sunrise, a cyclonic gyre of dense wildfire smoke (shades of tan to light brown) was seen moving eastward across the Northwest Territories toward Nunavut in True Color RGB images (source) from both GOES-18 (above) and GOES-16 (below). Boundary layer smoke from widespread fires on the previous day became entrained into the circulation of a low pressure system (surface analyses).

A Suomi-NPP VIIRS True Color RGB valid at 1912 UTC — viewed using RealEarth — is shown below.

View only this post Read Less