Category Archives: Multiple Cloud Layers

Dense Fog Event Over the Upper Midwest U.S.

Leave a reply
False color image of  a winter storm system moving across the upper Midwest U.S. on 11 January 2013.

A strong winter storm system moved over the upper Midwest on 11 January 2013 bringing heavy snow to the Dakotas. Rain and warm air moved in over the mostly snow covered areas of eastern Nebraska, Minnesota, Iowa, Wisconsin and northern Illinois resulting in a large area of locally dense fog. GOES-R IFR probabilities were used to monitor the large-scale event as it moved over the Upper Midwest U.S.

Due to a large amount of overlaying clouds a satellite only product such as the traditionally-used 11-3.9 micron Brightness Temperature Difference (BTD) only sees the top cloud layer and therefore can not detect fog/low stratus (FLS) beneath. The GOES-R IFR probabilities, however, combine both satellite and mesoscale NWP model data to create a blended product that can estimate the probability that IFR conditions are present even where overlaying clouds obscure all or part of the scene. This animation of the weather system shows that the traditional 11-3.9 micron BTD product only detects a small portion of the fog event, confirmed by the surface observations of ceiling and visibility. The GOES-R FLS product provides relatively high probabilities (>50%) that IFR conditions are present over the entire extent of the fog event with significantly high probabilities (>90%) when satellite data is useful.

GOES-R IFR probabilities (Upper Left), GOES-R Cloud Thickness (Upper Right), GOES-East Brightness Temperature Difference (10.7µm – 3.9µm) (Lower Left) and GOES-East Visible Satellite Imagery (Lower Right) on 11 January 2013. Surface observations of ceiling (100’s of ft) and visibility (miles) are shown in blue.

In the scene above the large change in probability seen in the GOES-R IFR product is a direct result of where the overlaying clouds obstruct the view of the lower clouds from the satellite. Where the satellite is able to see the low level clouds both satellite and model information are combined to determine the probability that IFR conditions are present. For this scene a strong satellite component and strong model component result in extremely high probabilities (>90%) that provide high confidence that IFR conditions are present over parts of the region. In areas where the overlaying clouds obstruct the satellite view only model data is used to determine the IFR probabilities. Without a strong satellite component the resulting probabilities are lower, however, they are still relatively high (>50%) and when used in conjunction with surface observations also provide high confidence that IFR conditions are present. The IFR probabilities show the extent of the fog over central Nebraska, southern Wisconsin and Illinois better than the traditional BTD product with very little false detection (high probabilities where surface obs do not indicate IFR conditions). Looking over the Dakotas and Nebraska the GOES-R IFR probabilities closely match the surface observations with relatively high probabilities over all of Nebraska and the eastern Dakotas where IFR conditions are present and very low probabilities in central N. Dakota and western S. Dakota where surface obs indicate VFR conditions.

GOES-R IFR probabilities (Upper Left), GOES-R Cloud Thickness (Upper Right), GOES-East Brightness Temperature Difference (10.7µm – 3.9µm) (Lower Left) and GOES-East Visible Satellite Imagery (Lower Right) on 11 January 2013. Surface observations of ceiling (100’s of ft) and visibility (miles) are shown in blue.

The traditional BTD product is mostly a nighttime only product as solar contamination in the 3.9 micron channel during the day makes it much more difficult to use. As daylight approaches the scene from 11 January 2013 the traditional BTD product appears to drop out. The GOES-R IFR product has no such issues and works smoothly through the night-to-day transition with consistently high probabilities accurately showing the full extent of the area of fog that continued through the rest of the afternoon.

Evolution of Fog/Low Stratus over Florida

Leave a reply
GOES-R IFR Probabilities (Upper Left), GOES-East Brightness Temperature Difference (10.7 µm – 3.9 µm) (Upper Right), GOES-R Cloud Thickness (Lower Left), GOES-East 6.5 µm imagery (Lower Right), from ~0000 UTC on 3 January 2013

GOES-R IFR Probabilities captured the evolution of IFR (and Low IFR) conditions over and around the Florida peninsula from late on 2 January through morning on 3 January 2013.  Advection fog over the chilly coastal waters of the eastern Gulf of Mexico stayed mainly offshore (although Sarasota at 00 UTC reports IFR conditions) and is captured well by the GOES-R product.  This is a region underneath high cirrus and as such, the traditional brightness temperature product is blind to the existence of low clouds there.

Over the course of the night, fog and low stratus developed over land, and the GOES-R IFR probability product captured that evolution as well (below, hourly imagery).  Again, there are regions where the brightness temperature difference product is not useable because of multiple cloud layers, and the Rapid Refresh Model output is controlling the IFR Probabilities — these are regions where the IFR probability field is very smooth and typically exhibits lower probability values even though IFR conditions may be observed (For example, at Gainesville and Jacksonville at 0600 UTC).  By morning, visibilities were under 1/4 mile over much of the central Florida Peninsula (For example, Orlando).

As above, but hourly imagery from 0000 UTC through 1400 UTC on 3 January 2013.

The 3/4-full moon allows for plenty of illumination for the Day/Night band on VIIRS, which is flying on Suomi/NPP.  The 0700 UTC imagery, below, demonstrates the difficulty of using the DNB at night to detect fog — city lights that shine through low clouds.  Fog is detected in rural regions, but where city lights exist, the signal is difficult to extract.

GOES-R IFR Probabilities (Upper Left), GOES-East Brightness Temperature Difference (10.7 µm – 3.9 µm) (Upper Right), GOES-R Cloud Thickness (Lower Left), Suomi/NPP VIIRS Day/Night Band 0.7 µm imagery (Lower Right), from ~0700 UTC on 3 January 2013

The visible imagery at 1500 UTC, below, shows the horizontal extent of the stratus deck through central Florida.  The region matches well with the IFR Probabilities because visible imagery during the day is used as a cloud-clearing mechanism in the GOES-R algorithms.  Note also how the reflected 3.9 µm solar radiation during the day renders the brightness temperature difference product ineffectual.

GOES-R IFR Probabilities (Upper Left), GOES-East Brightness Temperature Difference (10.7 µm – 3.9 µm) (Upper Right), GOES-R Cloud Thickness (Lower Left), GOES-East visible (0.63 µm) imagery (Lower Right), from ~1500 UTC on 3 January 2013

Interpreting IFR Probabilities when multiple Cloud Layers exist

Leave a reply
GOES-R IFR Probabilities computed from GOES-East and the Rapid Refresh, 1400 UTC 26 December 2012, and surface observations of ceilings and visibilities.

Large winter-time extratropical storms generate multiple cloud layers, and those many layers make difficult the detection of IFR conditions at the surface, caused by fog and low stratus.  The winter storm moving up the East Coast on December 26th, 2012, is typical of the type of storm that causes these conditions (although its abnormal production of tornadoes is very atypical).  In the figure above, IFR probabilities are high over eastern North Carolina, and those high probabilities are generated solely from Model Data.  Satellite information from GOES-East there is not useful because of multiple cloud layers.  There are regions over western North Carolina — where the GOES-R IFR probability field is less smooth and more pixelated — where GOES data indicate low-level clouds only and both satellite data and model data contribute to the IFR Probability fields.

GOES-R IFR Probabilities (Upper Left), GOES-East Brightness Temperature Difference Field (10.7 µm – 3.9 µm ) (Upper Right), GOES-R Cloud Thickness (Lower Left), GOES-East Visible Imagery (0.62 µm) (Lower Right)

The hourly animation of GOES-R IFR probabilities, Traditional Brightness Temperature Difference, GOES-R Cloud Thickness and Visible Imagery, above, is instructive on several points.  The Traditional Brightness Temperature Difference, because of the abundant high clouds (likely associated with the warm conveyor belt of the extratropical storm), yields little information about low-level conditions.  Note also how Cloud Thickness field is computed in one region:  the region where multiple cloud layers do not exist, and where twilight conditions are not present.  There are also differences in the GOES-R IFR Probability field between day and night that reflect the different predictors (and different predictor weights) that are used during those two times.

IFR Conditions during a Weak weather event

Leave a reply
GOES-East Water Vapor Imagery (6.7 ) from 2202 UTC on 18 December 2012

A weak weather system moved through the upper Midwest on December 18th, laying down a light strip of snow over Iowa and southern Wisconsin.  The Water Vapor imagery, above, shows the small scale vorticity center that helped to force the precipitation band moving over southern Lake Michigan after most of the snow in Wisconsin had tapered off.

The storm left behind abundant low-level moisture, and Fog/Low Stratus that caused IFR conditions were common.  The animation of the GOES-R IFR Probability product, below, shows high probabilities in the region where light snow and drizzle persisted.  Several aspects of this image require comment.

GOES-R IFR probabiltiies computed from GOES-East, and surface observations of ceilings and visibility, hourly from 2100 UTC 18 December to 0200 UTC 19 December.

The image at 2200 UTC shows the boundary between day-time predictor use and night-time predictor use.  This boundary runs southeast to northwest over Iowa.  IFR probabilities are somewhat lower where the night-time predictors are used initially, but patches of higher IFR probability do occur.  The IFR Probability product does distinguish well between the fog and low stratus that does restrict visibility and the elevated stratus over Illinois that is obvious in the traditional brightness temperature difference product, below.  Note especially how Chicago’s O’Hare airport is not reporting IFR conditions, nor under a region where GOES-R IFR probabilities are high, but it is in a region where the traditional brightness temperature product has a strong signal.

Toggle between GOES-R IFR Probability field and GOES-East Brightness Temperature Difference Field, 0200 UTC 19 December 2012

IFR conditions under high clouds in the Northeast

Leave a reply
GOES-R IFR Probabilities computed using GOES-East (Upper Left), GOES-East Brightness Temperature Difference (10.7 µm- 3.9 µm, The ‘traditional’ fog detection product) (Upper Right), GOES-R Cloud Thickness computed using GOES-East (Lower Left), GOES-East Window Channel (10.7 µm) Brightness Temperature (Lower Right)

When high clouds overspread an area, the traditional brightness temperature difference product cannot be used to highlight areas of fog and low stratus because radiation emissions are originating from high clouds, not from the water-based low clouds.  In this example from Monday morning, 17 December 2012, IFR conditions, causing airport flight delays, are commons from Washington DC to New York, and the GOES-R IFR probability product highlights the area where IFR (and near-IFR) conditions prevail.  Modest values (around 50%) occur where the satellite predictors do not provide a fog/low stratus signal;  however, the Rapid Refresh model data does show high probability of fog and low stratus.  Where the Satellite does contribute to the product (that is, in Pennsylvania north and east of the high cloud deck), IFR probabilities are very high.

Note also that the GOES-R Cloud Thickness product (bottom left), is computed only for the highest water-based cloud (in non-twilight conditions).  It is therefore not shown under the cirrus canopy, over southern New Jersey, Delaware, and Chesapeake Bay.

Fog Detection under multiple cloud layers

Leave a reply
GOES-R IFR Probabilities computed from GOES-East data (Upper left), GOES-East traditional brightness temperature difference product (Upper right), GOES-East 10.7 µm imagery (Lower left), Surface observations of ceiling/visibility (Lower right)

Weather over Florida on 12 December 2012 gives another example of the importance of fused data products in computing IFR probabilities/detecting low clouds and fog.  When mulitple cloud layers are present, as today over Florida, the traditional brightness temperature difference product (10.7  µm – 3.9 µm) can struggle to produce a useful signal.  Adding information about the lowest part of the atmosphere from the Rapid Refresh model, however, allows a coherent prediction of IFR probabilities to be made.  Note that the stations with the most significant reductions in ceilings/visibility do align with the stripe of higher IFR probability over the northern part of the Florida peninsula.

The impact of higher clouds

Leave a reply
GOES-R IFR Probabilities computed from GOES-East (Upper Left), GOES-East Brightness Temperature Difference (10.7 µm – 3.9 µm) (Upper Right), GOES-R Cloud Thickness (Lower Left), GOES-East 10.7 µm Imagery (Lower Right) at 0700 UTC on 6 December 2012.

Upper-level clouds, such as those apparent in both the 10.7 µm imagery and the brightness temperature difference imagery, above, lower right and upper right, respectively, have an impact on the GOES-R IFR Probability and GOES-R Cloud Thickness products.  The most obvious impact is in the Cloud Thickness product (lower left), which product is not computed in regions where high clouds are present.  The GOES-R IFR Probabilities are computed underneath high clouds, using mostly Rapid Refresh model data to determine the probability of fog and low stratus.  However, because cloud predictors are not used, IFR probabilities are somewhat lower.  In addition, the character of the field is flatter, reflecting the smoother fields that are present in the model.  This is especially obvious over northeast Louisiana and extreme east Texas in the IFR Probability image above.  Thus, the heritage product, the brightness temperature difference, gives no information from southwest Louisiana northward into southern Arkansas, but the fused GOES-R IFR Probabilities do suggest enhanced possibilities of IFR conditions in regions where reduced visibilities are reported:  central and northeast Louisiana and east Texas.  IFR Probabilities are much lower over southwest Louisiana where IFR conditions are not reported.

As above, but for 1245 UTC on 6 December 2012.

By 1245 UTC, the high clouds have lifted northeast and dissipated somewhat, so the heritage brightness temperature difference product gives information over the entire lower Mississippi River valley and over east Texas, where IFR conditions were widespread underneath very high GOES-R IFR Probabilities.  The GOES-R Cloud Thickness product indicates cloud thicknesses up to near 1200 feet, suggesting a burn-off time for radiation fog of around 5 hours.

Day/Night Band (from Suomi/NPP) imagery over Louisiana and East Texas, 0733 UTC on 6 December 2012. 

 The Day/Night band image derived from data from Suomi/NPP from 0733 UTC on 6 December, above, shows the higher and lower clouds over Texas and Louisiana.  The clouds between Houston and Dallas/Ft. Worth are low clouds, but the high clouds over Louisiana inhibit the detection of low clouds.  In addition, although the Day/Night band can give a good outline of where the clouds are, it does not show where the visibility restrictions consistent with IFR conditions are.

Visible Imagery from 2000 UTC

Visible imagery from 2000 UTC, above, shows that, although the fog has lifted, it has not burned off in 5 hours as predicted by the thickness/burn-off time relationship (here).  This may be related to the very low sun angle in December.

(Added:  This image loop shows how thin cirrus can show up in the brightness temperature difference product at night (this example is from Suomi/NPP even when fog-bound valleys are plainly evident in the Day/Night band at the same time!)

IFR Probabilities during a Big Storm

Leave a reply
GOES-R IFR Probabilities computed using GOES-West data (upper left), GOES-West traditional brightness temperature difference (10.7 µm – 3.9 µm ) (upper right), surface observations of visibility and ceiling (AGL) (lower left), Blended Total Precipitable Water Product (lower right)

Large extratropical storms are often accompanied by regions of IFR conditions, but the multiple layers of clouds that are produced by the storms make fog/low stratus detection by traditional means — the brightness temperature difference between 10.7µm and 3.9µm microns (at night) — a difficult prospect.  The GOES-R Fog/Low Stratus (FLS) product that fuses satellite data with model (Rapid Refresh) data allows for estimates of IFR probabilities.  The imagery above also includes precipitable water estimates from Sounder, GPS and microwave imagery (bottom right), highlighting the tropospheric river of moisture that is impinging on the West Coast.

Highest IFR probabilities are occurring in several regions in the animation above.  They occur over the Los Angeles basin, for example, where IFR conditions are reported at several airports (San Nicolas and Los Angeles airports, for example).  Reduced visibilities are also occurring in the Sierras — Blue Canyon (at 1500 m above sea level) reports IFR conditions — and IFR probabilities are higher along the spine of the mountains.  IFR probabilities are also higher along the northern California coast, and stations like Ukiah are reporting occasional IFR conditions.  Stations in the central Valley, and along the central coast, are in a region of lower IFR probabilities, and IFR conditions are comparatively rarer there.

IFR Conditions under Cirrus in Florida

Leave a reply
GOES-R IFR Probabilities from GOES-East (Upper Right), GOES-East traditional ‘fog product’ (Brightness Temperature Difference 10.7 µm – 3.9 µm), GOES-R Cloud Thickness, GOES-East Water Vapor (6.5 µm) imagery

Fog developed over central Florida overnight underneath a thin cirrus (as indicated by both the water vapor and brightness temperature difference imagery).  Cirrus clouds prevent the traditional brightness temperature difference field from identifying low fog/stratus because the high ice clouds are detected rather than the developing low-level water clouds.  This is a case, then, when a fused product gives needed surface information to help diagnose the development of fog and low stratus.  The brightness temperature difference product, the traditional method to detect fog and low stratus, is giving no information where dense fog is forming.

On this date, the development and expansion of the higher IFR probabilities over central Florida neatly matches the development of IFR conditions at the observing stations.  Probabilities are not high because the satellite predictors are not contributing to the algorithm.  It is important when interpreting the IFR probabilities to be aware of the presence of high clouds that will influence IFR probability values.  Where the cirrus clouds are not present, notably over northeast Florida, IFR probabilities are much higher because satellite predictors there are contributing to the final probability.

This case also shows that the Cloud Thickness is only computed where the highest clouds detected is a water-based cloud.  Underneath the cirrus shield, except for a few regions where there are apparently holes, cloud thickness is not computed.

GOES-R IFR Probabilities and Day/Night Band from VIIRS on Suomi/NPP, 0715 UTC 5 Nov

The toggle above flips between the Day-Night band from VIIRS on Suomi/NPP and the GOES-R IFR probability at the same time.  The thin cirrus shield is readily apparent, and the regions of fog are also visible in the Day/Night band over north-central Florida and over coastal South Carolina.

A MODIS-based IFR Probability (shown below) was also created at 0715 UTC, and it shows a pattern similar to that above.  The pixelated part of the image corresponds to where satellite data are being used.  The region with lower values, and a flatter field, was created using only model predictors and, as noted above, is characterized by lower probabilities.  The highest probabilities are in regions where both satellite and model predictors are very confident that IFR conditions are present.

MODIS-based GOES-R IFR Probabilities, Monday 5 Nov 2012, 0714 UTC

Marine Stratus, Day 2

Leave a reply
GOES-R IFR Probability (Upper Left), GOES-R Cloud Thickness (Upper Right), GOES-East Brightness Temperature Difference (10.8 µm – 3.9 µm) (Lower Left), GOES-East Visible Imagery (Lower Right) at 0800 UTC 26 October 2012

The marine stratus over the mid-atlantic on Oct 25th (Link) has continued to spread southwestward into North Carolina, and higher values in the IFR probability fields continue to overlap nicely with observed IFR conditions in the Piedmont of North Carolina.  The image above at 0800 UTC, and below at 1100 UTC, show deteriorating flight conditions as the high probabilities ooze southwestward.  Note that the traditional brightness temperature difference field does show a signature of the cirrus outflow from Hurricane Sandy off the coast of the Carolinas, in the extreme southeast part of the images, both at 0800 UTC and 1100 UTC.

GOES-R IFR Probability (Upper Left), GOES-R Cloud Thickness (Upper Right), GOES-East Brightness Temperature Difference (10.8 µm – 3.9 µm) Lower Left), GOES-East Visible Imagery (Lower Right) at 1100 UTC 26 October 2012
GOES-R IFR Probability (Upper Left), GOES-R Cloud Thickness (Upper Right), GOES-East Brightness Temperature Difference (10.8 µm – 3.9 µm) Lower Left), GOES-East Visible Imagery (Lower Right) at 1600 UTC 26 October 2012

By 1600 UTC, the cirrus canopy from Sandy has overspread most of eastern North Carolina.  This makes satellite data less useful in computing IFR probabilities, and Model Data comparatively more important, and the character of the IFR probability field does change, with a smoother field overall.  At 1600 UTC the probabilities have decreased; the temporal evolution of the field (not shown) is still giving important information, especially in concert with the surface visibility/ceiling observations.  The decrease in probabilities is gradual as the Sun slowly works to warm the boundary layer.