GOES-13 Sounder Anomalies

September 16th, 2013 |
GOES-13 10.7 µm IR channel images (click unage to play animation)

GOES Sounder DPI Lifted Index from 1300 UTC 15 September

The GOES-13 Sounder has been experiencing an increased number of unexpected scan line lengths. This was originally detected on 1 July 2013 and it has become more common since 12 September 2013. The anomaly is manifest in the data by occasional pixel gaps in all channels of the Sounder images. Some images show only a single gap (and therefore just one missing pixel); other images have several gaps. In the image above, the missing values are the black pixels just off the North Carolina coast (for example), or near the Mississippi River in southwestern Mississippi. Note that missing data only is present in the GOES-13 part of the domain. The GOES-15 signal is clean. The root cause of this error is under investigation. From NESDIS: “No telemetry violations have occurred, and all Sounder filter wheel telemetry data, including Filter wheel currents and period monitors, are within expected values.”

The missing pixels are also present in the real-time Sounder data available at CIMSS at this link. An example from September 15th is shown below. Or, click here for a composite (GOES-13/GOES-15) single-band image.

GOES-13 10.7 µm IR channel images (click unage to play animation)

GOES Sounder DPI Lifted Index from 1300 UTC 15 September

A quick analysis at two times (0246 and 1446 UTC) suggests that prior to 28 August, errors per image were limited to 20 or so pixels. Between 28 August and 9 September, fewer than 100 pixels were affected each hour. Since 10 September, the number of pixels affected in each image has increased one some days to more than 200. There are nearly 64000 pixels in each sounder image, so the number of bad points remains a small percentage of the total.

Cold cloud top temperatures associated with Hurricane Ingrid

September 15th, 2013 |
GOES-13 10.7 µm IR channel images (click unage to play animation)

GOES-13 10.7 µm IR channel images (click unage to play animation)

Hurricane Ingrid had been going though a period of slow intensification (ADT plot) off the Gulf Coast of Mexico during the overnight hours on 14 September15 September 2013. McIDAS images of GOES-13 10.7 µm IR channel data (above; click image to play animation) showed a few convective bursts within the central dense overcast portion of the storm, with cloud-top IR brightness temperatures as cold as -91º C (yellow color enhancement) at 04:45 UTC. These are unusually cold IR temperatures to be sensed by the coarse 4-km resolution of the IR detectors on the GOES Imager instrument.

An AWIPS image comparison of 375-meter resolution (projected onto a 1-km AWIPS grid) Suomi NPP VIIRS 0.7 µm Day/Night Band (DNB) and 11.45 µm IR channel data at 07:30 UTC (below) revealed even colder cloud-top IR brightness temperatures of -95º C (darker violet color enhancement). Another feature of interest was the concentric packet of gravity waves propagating northward away from the intense overshooting tops. These cloud-top gravity waves were best seen on the DNB image, which highlights the “visible image at night” capability of this VIIRS channel data.The intese overshooting tops (produced by the rapid pulsing nature of the convective bursts seen on the GOES imagery above) were likely the source of these gravity waves.

Suomi NPP VIIRS 11.45 µm IR channel and 0.7 µm Day/Night Band images

Suomi NPP VIIRS 11.45 µm IR channel and 0.7 µm Day/Night Band images

The bright streak seen on the VIIRS DNB image was due to cloud illumination by intense lightning activity at the time the instrument was scanning that particular location. An overlay of cloud-to-ground (CG) lightning strikes for the 15-minute period ending around the time of the VIIRS overpass (below) did show a small cluster of positive polarity (red) CG strikes centered near the region of highest overshooting tops — but the bright area detected by the DNB sensor was likely due to in-cloud or cloud-to-cloud lightning.

Suomi NPP VIIRS 0.7 µm Day/Night Band image (and cloud-to-ground lightning strikes)

Suomi NPP VIIRS 0.7 µm Day/Night Band image (and cloud-to-ground lightning strikes)

Also evident on the VIIRS DNB image were southward-propagating mesospheric airglow waves, likely generated by the intense overshooting tops of the strong cluster of convection located to the southeast of Ingrid. Note that these mesospheric airglow waves had no signature in the corresponding 11.45 µm IR image (in contrast to the aforementioned cloud-top gravity waves north of the center of Ingrid, which could be seen on the IR image due to their effect on the topography of the convective cloud top).

Flooding rains in the Front Range of Colorado

September 12th, 2013 |
GOES-15 and GOES-13 6.5 µm water vapor images (click image to play animation)

GOES-15 and GOES-13 6.5 µm water vapor images (click image to play animation)

Heavy rain fell over much of the Front Range of Colorado overnight, with Boulder receiving an all-time 24-hour total of 9.08 inches (Weather Underground blog). Numerous CoCoRAHS locations in Boulder County reported 24-hour rainfall amounts in excess of 6″, in a region where monthly precipitation records are close to 6″. This event could be classified as a 500- to 1000-year event (as shown at NOAA’s Precipitation Data Frequency Server).

A composite of GOES-15 and GOES-13 6.5 µm water vapor imagery (above) showed the large-scale environment that supported heavy rains over Colorado. A quasi-stationary upper-level cyclonic circulation was over the western United States; that circulation was drawing moisture northward from off the Pacific coast of Mexico (some of this moisture may have been associated with tropical cyclone Lorena, which dissipated off Baja California last week). In addition, a lower-tropospheric anticyclonic circulation pushing into the Midwest was shunting moisture westward towards the Front Range (03 UTC surface analysis). Several circulations, therefore, were helping to move moisture-rich air towards eastern Colorado.

Percent of Normal Total Precipitable Water

Percent of Normal Total Precipitable Water

Moisture content of the air can be defined by total precipitable water (TPW). For example, the percent of normal as determined from the blended TPW product, above, showed a thin ribbon of values approaching 200% of normal banked up against the Front Range. An animation of GOES Sounder Derived Product Imagery (DPI) TPW (here) showed values (in between small breaks in the persistent cloudiness) exceeding 30 mm or 1.2 inches — in fact, the 1.33 inches of TPW derived from the 12 UTC Denver rawinsonde data set a new record for the month of September at that site (Weather Underground blog). The sounder data also demonstrates how the Front Range capably prevented moisture from moving from eastern to western Colorado: TPW values are smaller over central and western Colorado.

GOES-15 and GOES-13 10.7 µm IR images (click image to play animation)

GOES-15 and GOES-13 10.7 µm IR images (click image to play animation)

GOES-15 (GOES-West) and GOES-13 (GOES-East) Imager 10.7 µm IR channel data, above, show that the convection that produced the flooding rains was relatively unimpressive on satellite imagery: cloud tops were not particularly cold. This suggests that warm-cloud processes were very efficient in producing the extraordinary rains in this event.One notable convective burst was seen to develop over the Boulder area (station identifier KBJC) after about 03 UTC on 12 September (10 PM local time on 11 September).

MODIS and GOES-13 Longwave IR (~10.7 µm) images at ~0440 UTC

MODIS and GOES-13 Longwave IR (~10.7 µm) images at ~0440 UTC

MODIS and GOES-13 Longwave IR (~10.7 µm) images at ~0900 UTC

MODIS and GOES-13 Longwave IR (~10.7 µm) images at ~0900 UTC

The relatively small-scale nature of this event suggests that high-resolution polar-orbiter data might be useful in describing the environment surrounding the storms. MODIS, for example, provided high-resolution imagery twice overnight, and comparisons between the MODIS and concomitant GOES imagery is shown above. MODIS data does have somewhat cooler cloud-top IR brightness temperatures (-54 to -55 C on MODIS, vs -48 to -50 C on GOES). The strongest convection is over the region that reported very heavy rains. Note that the MODIS imagery does not suffer from the parallax shift inherent in GOES imagery.

Suomi/NPP VIIRS Day/Night Band imagery at 0826 UTC

Suomi/NPP VIIRS Day/Night Band imagery at 0826 UTC

Suomi/NPP VIIRS data was also available over Colorado during the overnight hours. The Day/Night Band imagery at 0826 UTC or 2:26 AM local time, above, showed a region of clouds over the Front Range of Colorado (city lights nevertheless do shine through the clouds). There is little indication of widespread intense lightning (characterized in Day/Night Band imagery as a smear of light along a scan line) — in fact, only a single negative cloud-to-ground strike was seen in the Denver area at that time.

Heavy rains over the Front Range of Colorado are typically accompanied by easterly (upslope) flow. The image below shows METAR data plotted over the Suomi/NPP VIIRS 11.45 µm IR data. Easterly flow is common at many of the reporting sites.

Suomi/NPP VIIRS 11.45 µm IR imagery at 0826 UTC

Suomi/NPP VIIRS 11.45 µm IR imagery at 0826 UTC

Two 12-day animations from GOES-13 (immediately below) and GOES-15 (below GOES-13) show the large-scale environment before, during and after the heavy rains.

GOES-13 6.5 µm water vapor images, 5-16 September 2013 (click image to play animation)

GOES-13 6.5 µm water vapor images, 5-16 September 2013 (click image to play animation)

GOES-15 6.5 µm water vapor images, 5-16 September 2013 (click image to play animation)

GOES-15 6.5 µm water vapor images, 5-16 September 2013 (click image to play animation)

Weekly Precipitation ending 16 September 2013

Weekly Precipitation ending 16 September 2013

Weekly storm totals from this event (from this NOAA site) show widespread heavy rains over both Colorado and New Mexico. A large area exceeded 600% of normal for the week!

Northward advection of stratus along the Oregon and Washington coasts

September 11th, 2013 |
Suomi NPP VIIRS IR brightness temperature difference "fog/stratus product"

Suomi NPP VIIRS IR brightness temperature difference “fog/stratus product”

An AWIPS image of the Suomi NPP VIIRS 11.45 – 3.74 µm IR brightness temperature difference “fog/stratus product” at 10:26 UTC or 2:26 AM local time on 11 September 2013 (above) revealed a relatively narrow tongue of stratus cloud (darker orange to red color enhancement) which was beginning to move northward along the nearshore coastal waters of Oregon. The northward stratus advection was being driven by the presence of an elongated trough of low pressure off the West Coast of the US, with some embedded closed low circulations along the trough axis.

During the subsequent daylight hours, McIDAS images of GOES-15 0.63 µm visible channel data (below; click image to play animation) showed the continued northward and northwestward spread of the plume of stratus cloud along and just off the coast of Washington, with some inland intrusions of marine stratus noted later in the day. One of the aforementioned closed cyclonic circulations (which was not well-analyzed by the GFS40 model surface winds on the VIIRS image above) could be seen within the marine boundary layer stratus off the coast of Oregon. Other features of interest included the dispersion of smoke plumes from 2 small wildfires that were burning in southwestern Oregon. Meteorological fun fact: on this day, for the Lower 48 states both the daily lowest temperature (22º F at Silver Lake) and the daily highest temperature (102º F at Medford) occurred in the state of Oregon!

GOES-15 0.63 µm visible channel images (click image to play animation)

GOES-15 0.63 µm visible channel images (click image to play animation)