GOES-14: improved Image Navigation and Registration (INR)

November 30th, 2009
GOES-11, GOES-12, and GOES-14 visible images

GOES-11, GOES-12, and GOES-14 visible images

GOES-14 (launched on 27 June 2009) began its NOAA Science Test on 30 November 2009 — and McIDAS images of the GOES-11, GOES-12, and GOES-14 visible channel data (above) showed an area of melting snow cover across the Texas / New Mexico border region. This area (southwest of Amarillo, Texas) had received 1-3 inches of new snowfall on 29 November, which then rapidly melted under full sun on the following day. Note the improvement in Image Navigation and Registration (INR) on the GOES-14 images — there is much less image-to-image wobble compared to the GOES-11 (GOES West) and GOES-12 (GOES East) images. The 3 sets of images are displayed in the native projection of their respective satellites.

Better INR performance is critical to improving the accuracy of such satellite-derived products as the 1-hourly MADIS atmospheric motion vectors, seen overlaid on an AWIPS image of the MODIS 11.0 µm IR channel data (below).

MODIS IR image with MADIS 1-hour atmospheric motion vectors

MODIS IR image with MADIS 1-hour atmospheric motion vectors

A comparison of the MODIS visible and 2.1 µm near-IR “snow/ice channel” images (below) confirms that this feature was indeed snow — snow and ice are strong absorbers at the 2.1 µm wavelength, and appear as very dark features on the snow/ice image.

MODIS visible and near-IR snow/ice channel images

MODIS visible and near-IR "snow/ice" channel images

Note how the MODIS Land Surface Temperature (LST) product (below) was showing LST values in the upper 30s to low 40s F (green colors) over the snow-covered areas, in contrast to much warmer LST values in the middle 50s to low 60s F (yellow colors) over the surrounding bare ground. The surface air temperatures were also being held down into the upper 30s to low 40s F over the melting snow cover.

MODIS Land Surface Temperature product

MODIS Land Surface Temperature product

Farewell to GOES-10

November 30th, 2009

FinalGOES10IRloop

Shortly after midnight (EST) on December 1st, GOES-10 was decomissioned and boosted to a disposal orbit (approximately 300 km above the operational orbit). It was shut off because it lacks fuel for the required maneuvers to keep it on station.

GOES-K was launched 25 April 1997, with a life expectancy of five years. A solar array problem shortly after launch in May of 1997 was nearly fatal to the spacecraft; however, a yaw-flip maneuver (that is, flying the spacecraft upside-down) proved successful and GOES-10 has successfully served data nearly continuously since then. The first visible image from GOES is here. Early examples of Sounder and Imager are also available. For more examples of GOES-10 imagery, click here. GOES-10 served as GOES-West from 27 July 1998 (replacing GOES-9) until 21 July 2006 (when it was replaced by GOES-11). GOES-10 then moved from 135 West Longitude to 60 W Longitude, arriving on station in December 2006 to provide near-continuous data over South America (More information on GOES-10 is available here).

As a Geostationary satellite focused on South America, GOES-10 provided valuable information about the Air France Flight Crash over the Atlantic Ocean, volcanic eruptions over South America. In addition, as it moved from 135 W Longitude to 60 W Longitude, it was in Super Rapid-Scan Operation mode — that is, imagery was collected every minute over limited regions — to give insight into various meteorological phenomena. (For more links to GOES-10 imagery, click the GOES-10 category, or click here).

With the termination of GOES-10 operations, routine satellite observation of South America will fall to GOES-12, the operational GOES-East satellite. However, the operational demands on GOES-East preclude the high temporal observations that GOES-10 provided. For example, much of South America now has routine 15-minute coverage; GOES-East will provide only half-hourly coverage. This image loop shows the motion of a smoke plume — at 15-minute intervals — near the Tocantins River just south to the Amazon Delta. A similar loop from GOES-East is here. Reduced temporal resolution introduces greater error to both cloud-tracked features (derived winds) and fires detected.

Similar views from different vantage points can be important. Consider, for example, the twin views of northeast Brazil in the 4-micron band from GOES-10 and GOES-12.

GOES1012

Both platforms observe the fires in the Amazon River delta in the upper left part of the images. Note, however, that only GOES-East shows a very warm Lake behind Sobradinho Dam on the Sao Francisco River. Indeed, the 3.9-micron sensor has saturated on GOES-East (over the Equator at 75 W), but GOES-10 (over the Equator at 60 W) shows very little signature. This is an excellent example of Sun Glint in the 3.9 micron channel. Solar 3.9-micron radiation reflected from the lake is saturating the instrument on GOES-East. GOES-10, farther east, can look at the same region and not see the Sun Glint.

In contrast to GOES-East and GOES-West data, data from GOES-10 have been remapped before distribution since it arrived at 60 West back in late 2006. The remapping is necessary because the satellite inclination was large; indeed, it was more than 4 degrees on 25 November 2009.

Update: The Final Imager images from GOES-10: 0.65 microns; 3.9 microns; 6.8 microns; 10.7 microns; 12.0 microns; Infrared Channels in a loop.

Current plans are for GOES-13 to replace GOES-12 as GOES-East in April of 2010. Subsequently, GOES-12 will move to 60 W and resume GOES-10′s duties.

Standing waves to the lee of the Sierra Nevada

November 29th, 2009
GOES-11 6.7 µm water vapor images

GOES-11 6.7 µm water vapor images

Under normal conditions of westerly flow aloft, one might expect to see occasional standing waves to the east of the Sierra Nevada mountain range; however, due to the presence of a strong cut-off low over the southwestern US, the winds aloft over Nevada and California were from the northeast on 29 November 2009 — and AWIPS images of the 8-km resolution GOES-11 6.7 µm water vapor channel (above) showed a signature of mountain waves to the west of the crest of the Sierra Nevada. This type of lee wave signature on water vapor imagery indicates the potential for clear air turbulence in the proximity of the waves — however, there were no pilot reports of turbulence noted in the immediate area of the lee wave signature (possibly due to the time of day, when air traffic is generally at a minimum).

A pair of 1-km resolution MODIS 6.7 µm water vapor images at 06:20 and 10:38 UTC (below) showed the advantage of higher spatial resolution for detecting such mesoscale signatures.

MODIS 6.7 µm water vapor images

MODIS 6.7 µm water vapor images

A comparison of the 8-km resolution GOES-11 6.7 µm water vapor, 4-km resolution GOES-14 6.5 µm water vapor, and 4-km resolution GOES-12 6.5 µm water vapor images (below) further demonstrated the effects of varying spatial resolution as well as varying satellite viewing angle in resolving the lee wave signatures to the west of the Sierra Nevada. GOES-14 (positioned at 105º West longitude) had the best viewing angle of the region, and its 4-km resolution water vapor channel did a better job of depicting both the areal coverage and the temporal duration of the lee wave structure — especially compared to GOES-11 (positioned at 135º West longitude) with its 8-km resolution water vapor channel. Using GOES-14 imagery, the onset of the lee wave structure was easier to see, and the duration of the lee wave event was also longer. In addition, even though the viewing angle from GOES-12 (positioned at 75º West longitude) was very large — about 65 degrees — the 4-km resolution water vapor channel still managed to show a fairly good signature of the lee waves.

GOES-11, GOES-14, and GOES-12 water vapor images

GOES-11, GOES-14, and GOES-12 water vapor images

Note that the water vapor images also suggested the formation of a downwind “cloud banner” or “cloud crest” after about 06 UTC. The 4-km resolution MODIS Cloud Phase product (below) showed a growing ice phase cloud feature (salmon color enhancement) over central California between 06:20 and 10:38 UTC.

MODIS Cloud Phase product

MODIS Cloud Phase product

However, the 4-km resolution MODIS Cloud Top Temperature product (below) only indicated cloud top temperature values as cold as -20 to -22º C (green color enhancement) within the glaciated cloud banner feature over central California.

MODIS Cloud Top Temperature product

MODIS Cloud Top Temperature product

Another view of the central California cloud banner feature using the 1-km resolution AVHRR Cloud Type product at 09:39 UTC (below) indicated that it was composed of cirrus clouds (yellow enhancement), with supercooled water droplet clouds (cyan color enhancement) immediately upwind over the Sierra Nevada.

AVHRR Cloud Type product

AVHRR Cloud Type product

Furthermore, the corresponding 1-km resolution AVHRR Cloud Top Temperature product (below) indicated significantly colder cloud top temperature values of -60º to -70º C (blue to white colors) within the central California cloud banner feature.

AVHRR Cloud Top Temperature product

AVHRR Cloud Top Temperature product

Finally, it is interesting to note that the GOES-11 10.7 µm IR image (below, upper left panel) showed absolutely no signature of the lee cloud banner — due to the thin nature of this glaciated cloud feature, a great deal of radiation from the warmer land surface below was “bleeding up” though the ice cloud and was masking its presence on IR imagery. The GOES-11 6.7 µm water vapor image, however, did show a better signature of the presence of the cloud banner feature (lower left panel).

GOES IR, GOES water vapor, AVHRR Cloud Top Temperature, and AVHRR Cloud Top Height

GOES IR, GOES water vapor, AVHRR Cloud Top Temperature, and AVHRR Cloud Top Height

GOES-12 Sounder Anomalies

November 25th, 2009

channel15loop

Measurements from the GOES-12 Sounder instrument have shown increased noise over the past weeks. The noise in the signal does not occur with any persistence, but it can be very noticeable, as shown in the loop above of gif imagery taken from the UW CIMSS Derived Products Page. The images above are for Channel 15 (4.4 microns, a wavelength used to investigate the upper atmosphere), a channel on the GOES-12 sounder that has shown considerable noise since launch. Note the marked increase in noise, however, for the 2300 UTC image in the loop.

In addition, increased noise is also affecting channels 13 and 14 (4.57 and 4.53 microns, respectively) and channels 16-18 (4.13, 3.98 and 3.76 microns, respectively). Compare the noise in the images from 25 November at 1800 UTC (significant, noticeable noise in Channels 13-17) and at 1700 UTC (Noise noticeable only in the usually noisy Channel 15).

These noisy satellite observations do impact derived products such as Precipitable Water cloud mask: The 1800 UTC product that uses the noisy data from 1800 UTC shows the speckled result of noise over the southern Plains; the 1700 UTC observations that use the cleaner 1700 UTC data, do not contain such speckles.) The affected channels are used to determine the cloud mask. When there is amplified noise — especially if it results in very cold temperatures that are inferred to be high clouds — then a faulty cloud mask is a result. This is especially true at night when the Channel 18 brightness temperature is compared to the Channel Channel 8, and a cloud is inferred if there is a significant difference between the two. See, for example, this image from 01 UTC on 24 November. The speckling in the cloudtop pressure over Texas results from subtle noise signals in the 3.76-micron band (Channel 18). If these sounder data are being used to quantify the presence of clouds, the increasing noise in the shortwave infrared channels may be problematic.

GOES-14, located above the Equator near 105 W, is currently undergoing science testing. A comparison of Sounder band 15 from GOES-12 to the same band on GOES-14 shows the remarkably cleaner signal from GOES-14.

GOES-12 is scheduled to remain the operational GOES-EAST through March of 2010. It will be replaced by GOES-13.