Plumes from the Kilauea Volcano in Hawai’i

April 8th, 2008 |

GOES-11 visible images (Animated GIF)

Explosive events from the Kilauea Volcano (located on the Big Island of Hawai’i) began to occur in mid-March of 2008 — these were the first explosive events from that particular volcano since 1927. Activity from Kilauea then continued for several weeks; GOES-11 (GOES-West) 0.63 µm visible imagery from 07 April 2008 (above) showed the hazy signature of a long volcanic plume (composed primarily of steam, but possibly containing small amounts of ash) streaming southwestward from Hawai’i. With the typical northeasterly trade winds that often persist over that region, this was the common scenario seen on many days during late March into early April.

However, the northeasterly trade wind flow regime was interrupted by a surface trough of low pressure on 08 April 2008, and southerly to southeasterly winds began to advect the Kilauea plume to the north and northwest during the day (photo). The volcanic plume at that time contained significantly elevated amounts of sulfur dioxide (SO2), which forced the closure of Hawai’i Volcanoes National Park on 08 April. GOES-11 visible imagery (below) revealed two separate plumes, emanating from the Halema`uma`u and the Pu`u `O`o vents of the Kilauea volcano.

GOES-11 visible images (Animated GIF)

The volcanic SO2 plume on 08 April could be tracked using a GOES-11 sounder brightness temperature difference product (subtracting the 13.4 µm band 5 temperature from the 7.4 µm band 10 temperature) — a small “bubble” of elevated SO2 concentration (brightness temperature difference values of 0º to +5º K, yellow to orange colors) was seen to move slowly northwestward from the Big Island of Hawai’i toward the smaller islands of Maui/Kahoolawe/Lanai/Molokai (below). Unfortunately, GOES-11 sounder data over the Hawai’i region is only available 7 times a day (not once per hour, as it is over the continental US), so the motion of the SO2 feature was more difficult to follow compared to using the more frequent 15-minute visible imagery from the GOES-11 imager.

GOES-11 sounder difference product

The surface visibility at Lahaina / West Maui (station identifier PHJH, below) decreased from 15 miles to 7 miles (with haze reported) as southerly winds blew the volcanic plume and SO2 cloud over the island of Maui on 08 April. On the Big Island of Hawai’i, volcanic fog (sometimes referred to as “vog”) reduced visibility to less than 1 mile at Hilo.

Lahaina/West Maui surface meteorogram

The 7.4 µm band 10 of the GOES sounder is primarily a “water vapor absorption” band, but this particular sounder channel is also sensitive to high SO2 loadings in the atmosphere (as shown by the figure shown below, taken from Ackerman, S. A., A. J. Schreiner, T. J. Schmit, H. M. Woolf, J. Li1, and M. Pavolonis, 2008: Using the GOES Sounder to Monitor Upper-level SO2 from Volcanic Eruptions, submitted to Journal of Geophysical Research). The plot also shows that high SO2 loading could be detected using a channel located within the 8.4-9.0 µm band.

GOES sounder spectral response function plot

The Advanced Baseline Imager (ABI) on the future GOES-R satellite will have a similar 7.3 µm channel (at a 2 km spatial resolution, compared to the 10 km spatial resolution on the current GOES sounder), and with ABI imagery available at more frequent time intervals (images every 5 minutes over the full disk), the detection of these types of volcanic SO2 plumes will be significantly improved in the GOES-R era.

Terra MODIS images at 20:55 UTC on 08 April (below; courtesy of Mat Gunshor, CIMSS) demonstrate the utility of using the 11.0 µm – 8.5 µm brightness temperature difference product to help discriminate between the SO2 plume (darker blue enhancement on the difference product image, moving north from the Big Island of Hawai’i) and the larger steam plume (evident as the hazy area on the visible image, moving westward and northwestward from the island).

MODIS images (Animated GIF)

Precipitable water plume in the Gulf of Mexico

April 7th, 2008 |

GOES-12 10.7µm IR image

An AWIPS image of the GOES-12 10.7 µm “IR window” channel (above) showed a large cluster of convection that developed in the vicinity an inverted surface trough axis over the Gulf of Mexico on 07 April 2008.

GOES-12 6.5 µm water vapor images (Animated GIF)

An animation of AWIPS images of the GOES-12 6.5 µm “water vapor channel” (above) suggested that much of the middle to upper troposphere was quite dry (yellow to orange colors) over the Gulf of Mexico region on that particular day.

POES AMSU total precipitable water (Animated GIF)

However, AWIPS images of the POES AMSU total precipitable water (above) and the DMSP SSM/I total precipitable water (below) revealed that a plume of significant moisture was moving northward into the central Gulf of Mexico during 06-07 April, providing a necessary ingredient for the development of the convection.

DMSP SSM/I total precipitable water (Animated GIF)

A comparison of the GOES water vapor channel image with total precipitable water products from the GOES sounder, POES AMSU, and DMSP SSM/I (below) demonstrates how misleading it would be to simply interpret the GOES vapor image alone and conclude that the entire Gulf of Mexico region was generally “very dry” (aside from the cluster of convection that had developed). The bulk of the precipitable water plume existed at lower levels of the atmosphere, below the layer that was being sensed by the GOES imager water vapor channel (which, according to the GOES water vapor weighting function plot for Brownsville, Texas at 12 UTC on 07 April was centered around 500 hPa).

GOES + POES imagery (Animated GIF)

An animation of hourly composites of the CIMSS MIMIC total precipitable water product (below) showed the evolution of the moist plume as it emerged into the southwestern Gulf of Mexico and then moved northeastward into the central Gulf during the 06-07 April period. The MIMIC product blends microwave total precipitable water data from the DMSP SSM/I and the Aqua AMSR-E polar orbiting instruments, and advects the blended moisture fields using lower-tropospheric mean layer wind derived from the GFS model.

MIMIC total precipitable water (Animated GIF)

Convergence of the Brazil Current and the Malvinas/Falkland Current

April 7th, 2008 |

AVHRR sea surface temperature (Google Maps)

[Hat-tip to Amato Evan at CIMSS for pointing out this very interesting AVHRR sea surface temperature imagery] AVHRR sea surface temperature (SST) data (above, viewed using Google Earth) revealed the striking convergence of the Brazil Current and the Malvinas/Falkland Current off the east coast of South America on 07 April 2008. The Brazil Current transports warm subtropical water (SST values of 22º to 28ºC, yellow to orange colors) southward, while the Malvinas/Falkland Current transports cold Antarctic water (SST values of 6º to 12ºC, cyan to dark blue colors) northward. These two ocean currents are seen to converge several hundred kilometers off the coast of Argentina — the exact location of this Brazil/Malvinas convergence zone changes with the seasons.

An animation of daily AVHRR SST images from 01 to 07 April (below) shows subtle variations in the position of the Brazil and Malvinas ocean currents, as well as interesting eddy structures in the vicinity of the Brazil-Malvinas Confluence. Note the appearance of a well-defined “comma cloud” associated with a strong cyclone off the coast on 06 April (recall that winds flow clockwise around a cyclone in the Southern Hemisphere).

AVHRR sea surface temperature images (Animated GIF)

The colder waters of the Malvinas Current are rich in nutrients which support the growth of marine plant life, which then attracts large numbers of fish to feed — therefore, the commercial fishing industry is very interested in satellite data that accurately depict the location of such cold ocean currents.

Reference: Convergence Zones – Where the Action Is (NASA Goddard Space Flight Center)

Fires in Kansas/Oklahoma, and continued river flooding in the Ohio Valley

April 6th, 2008 |

GOES + MODIS shortwave IR images (Animated GIF)

A comparison of AWIPS images of the GOES-12 3.9µm and the MODIS 3.7µm “shortwave IR” channels (above) revealed that numerous small fires were burning across parts of eastern Kansas and northeastern Oklahoma during the afternoon of 06 April 2008. Note how the 1-km resolution MODIS imagery did a better job of detecting more of the smaller fire “hot spots” (yellow to red pixels) than the corresponding 4-km resolution GOES-12 imagery, and many more hot fires (with brightness temperature values of 50ºC or greater) were seen on the MODIS imagery.

These small fires were the result of agricultural burning to prepare fields for a new round of planting. AWIPS images of the MODIS visible and Normalized Difference Vegetation Index (NDVI) product (below) indicated that much of eastern Kansas and northeastern Oklahoma (where these fires were burning to clear cropland) exhibited much lower NDVI values of 0.2 to 0.3 (pale yellow to beige colors), while higher NDVI values of 0.5 to 0.7 (darker green colors) were seen in surrounding areas (where mature crops or dense trees and other vegetation dominated the landscape).

MODIS visible + NDVI images (Animated GIF)

Farther to the east, the impacts of the ongoing episode of extensive river flooding in the Ohio River Valley region were quite apparent by examining MODIS false color imagery from the SSEC MODIS Today site (below) — water appears dark blue in the false color imagery, allowing swollen rivers and flooded areas to be easily identified. According to the NWS Advanced Hydrologic Prediction Service site, parts of Missouri, Illinois, Indiana, and Kentucky had received 10-20 inches of precipitation in the 30 day period ending on 06 April, which was 200-400% above normal.
MODIS true color + false color images (Animated GIF)