Resolution and Fog Detection in the mountains of the Appalachia

GOES-16 data posted on this page are preliminary, non-operational and are undergoing testing.

Clear skies and light winds allowed for the development of Radiation Fog over parts of West Virginia and surrounding states early on Tuesday 17 October 2017, and IFR conditions were widespread (Source).

GOES-16 and GOES-13 both monitored this event, and toggles between the two brightness temperature difference fields, from 0315, 0415 and 0615 on 17 October, are shown below.  The better spatial resolution (and better precision) of the GOES-16 ABI is apparent in the imagery:  At 0415 UTC, GOES-16 is detecting river fog in many valleys in West Virginia and Kentucky, and is better resolving the detected fog over the Allegheny River in northwest Pennsylvania.  By 0615 UTC, the fog is also detected by GOES-13 — meaning it has increased in size so that the GOES-13 pixel size can detect it.  GOES-16 had a lead time of about 1-2 hours on GOES-13 in this case!

GOES-13 and GOES-16 Brightness Temperature Difference Fields, 0315 UTC on 17 October (Click to enlarge)

GOES-13 and GOES-16 Brightness Temperature Difference Fields, 0415 UTC on 17 October (Click to enlarge)

GOES-13 and GOES-16 Brightness Temperature Difference Fields, 0615 UTC on 17 October (Click to enlarge)

At present, GOES-R IFR Probabilities are created using GOES-13 data (or GOES-15 data for the western United States).  As might be expected, the GOES-R IFR Probabilities also lagged the GOES-16 Brightness Temperature Difference field in alerting to fog.  The short animation below shows values hourly from 0315-0615.  Very small IFR Probabilities are present at 0315 and 0415 — because, in part, the GOES-13 Imager cannot resolve the developing Fog/Stratus (and the Rapid Refresh model lacks the horizontal resolution to reproduce what is happening in narrow river valleys).  By 0515, and 0615 especially, GOES-R IFR probabilities are suggesting that a region of fog is likely.

One other thing to note:  There is a small signal in the brightness temperature difference field at 0315 UTC, above, in northwest Ohio in both GOES-13 and GOES-16 imagery.  There is no corresponding signal in the IFR Probability field.  The conclusion is that this is mid-level stratus that is not affecting surface visibility or ceilings.

GOES-R IFR Probabilities, hourly from 0315-0615 UTC on 17 October 2017 (Click to enlarge)

GOES-R IFR Probabilities, supplied via an LDM feed to National Weather Service Offices, will use GOES-13 (or GOES-15) imagery until GOES-16 is operational at 75.2 degrees West Longitude.  When GOES-16 is operational, GOES-R IFR Probabilities will use GOES-16 data, and that data will flow through the LDM.  In the meantime, GOES-16 IFR Probabilities can be viewed at this site.

The GOES-16 Version of GOES-R IFR Probability

GOES-R IFR Probability fields computed using GOES-16 Data and Rapid Refresh Model Output, 1042 UTC on 3 October 2017 (Click to enlarge)

GOES-16 data posted on this page are preliminary, non-operational and are undergoing testing

The suite of GOES-R IFR Probability products are now being computed using GOES-16 data, with a 5-minute temporal cadence over CONUS.  Products include IFR Probability (above), Low IFR Probability, MVFR Probability and Cloud Height.  These products are only available for now at http://cimss.ssec.wisc.edu/geocat ;  however, work is progressing to have the fields available through an LDM feed.  The GOES-R IFR Probability fields computed using Legacy GOES (GOES-13 and GOES-15) continue to flow to the National Weather Service via the LDM.

IFR Conditions in Pennsylvania and Oregon

GOES-16 Brightness Temperature Difference (10.3 µm – 3.9 µm) at 0912 UTC on 2 October 2017 over the Mid-Atlantic States (click to enlarge)

GOES-16 data posted on this page are preliminary, non-operational and are undergoing testing.

The images above show the GOES-16 Brightness Temperature Difference at the same time at two places over the United States: The mid-Atlantic States (above) and Oregon and surrounding States (below).  The ‘Fog’ Product, as this Brightness Temperature Difference is commonly called, in reality identifies only clouds that are made up of water droplets — that is, stratus.  A cloud made up of water droplets emits 10.3 µm radiation nearly as a blackbody does. Thus, the computation of Brightness Temperature — which computation assumes a blackbody emission — results is a temperature close to that which might be observed.  In contrast, those water droplets do not emit 3.9 µm radiation as a blackbody would.  Thus, the amount of radiation detected by the satellite is smaller than would be detected if blackbody emissions were occurring, and the computation of blackbody temperature therefore yields a colder temperature, and the brightness temperature difference field, above, will show clouds made up of water droplets as positive, or cyan in the enhancement above.

The River Valleys of the northeast show a very strong signal that suggests Radiation Fog is developing over the relatively warm waters in the Valleys.  The Delaware, Hudson, Mohawk, Connecticut, Susquehanna, Allegheny, Monongahela, and others — all show a signature that one would associate with fog.  A signal is also apparent from southern New Jersey southwestward through the Piedmont of North Carolina.  Would you expect there to be fog there as well, given the signal?

The State of Oregon at the same time shows a very strong signal in the ‘Fog’ Product.  A clue that this might be only stratus, and not visibility-restricting fog, lies in the structure of the clouds — they do not seem to be constrained by topographic features as is common with fog.

GOES-16 Brightness Temperature Difference (10.3 µm – 3.9 µm) at 0912 UTC on 2 October 2017 over Oregon and adjacent States (click to enlarge)

GOES-R IFR Probabilities are computed using Legacy GOES (GOES-13 and GOES-15) and Rapid Refresh model information; Preliminary IFR Probability fields computed with GOES-16 data are available here.  These GOES-16 fields should be available via LDM Request when GOES-16 becomes operational as GOES-East.

GOES-R IFR Probability Fields use both the Brightness Temperature Difference field (10.7 µm – 3.9 µm) from heritage GOES instruments and information about low-level saturation from Rapid Refresh Model output.  The horizontal resolution on GOES-13 and GOES-15 is coarser than on GOES-16 (4 kilometers at the sub-satellite point vs. 2 kilometers), so small river valleys will not be resolved.  (It is also difficult for the Rapid Refresh model to resolve small valleys).

GOES-R IFR Probability fields at 0915 UTC, along with 0900 UTC surface observations of ceilings and visibility (Click to enlarge)

The IFR Probability Fields, above, show some signal over the river valleys of the northeast; that signal is mostly satellite-based, but the poor resolution of GOES-13 means that fog/stratus in the river valleys is not well-resolved. Still, a seasoned forecaster could likely interpret the small signals that are developing to mean fog is in the Valleys.  (And restrictions to ceilings and visibilities are certainly reported in the river valleys of the Mid-Atlantic and Northeast)   IFR Probabilities are also noticeable over southeast Virginia, although widespread surface observations showing IFR Conditions are not present.  (Such observations are somewhat more common near sunrise, at 1130 UTC).

IFR Probabilities are much less widespread over Oregon, with most of the signal over western Oregon related to the topography.  In this example, IFR Probabilities are ably screening out regions where elevated stratus is creating a strong signal for the satellite in the Brightness Temperature Difference field.

Dense Fog over Missouri and over Alabama

GOES-R IFR Probability Fields, Hourly from 0215-1315 UTC on 19 September 2017 (Click to enlarge)

GOES-R IFR Probabilities are computed using Legacy GOES (GOES-13 and GOES-15) and Rapid Refresh model information; GOES-16 data will be incorporated into the IFR Probability algorithm in late 2017

Dense fog developed over Missouri on Tuesday 19 September and Dense Fog Advisories were issued. The animation above shows the hourly development of GOES-R IFR Probability fields; values increased from northern Missouri to southern Missouri as dense fog developed, first north of I-70, then south into the rest of the state. The morning of 19 September was mostly devoid of mid-level and high-level clouds over Missouri (exception: west-central Missouri starting after 0900 UTC), and that kind of night means that traditional methods of fog detection work well. The brightness temperature difference field between the shortwave Infrared and the Longwave Infrared (3.9 µm and 10.3 µm on GOES-16, 3.9 µm and 10.7 µm on GOES-13) shows the fog development.

Note that the IFR Probability field, above, does not show fog dissipating around sunrise. That’s in contrast to the Brightness Temperature Difference field below. As the sun rises, the amount of solar radiation at 3.9 µm that is reflected off the clouds increases; this changes the brightness temperature difference from positive (cyan in the color enhancement shown) to negative (grey or black in the enhancement shown).

GOES-16 has better spatial resolution than GOES-13; thus, the small valley fogs that can develop in the rugged (ish) terrain of southern Missouri are resolved in GOES-16, but not in GOES-13. When GOES-R IFR Probability is created using GOES-16 data (slated to begin in late 2017), the resolution improvements in GOES-16 will migrate to IFR Probability fields.

GOES-16 data posted on this page are preliminary, non-operational and are undergoing testing.

GOES-16 Brightness Temperature Difference (10.3 µm – 3.9 µm), hourly from 0412 to 1112 UTC on 19 September (Click to enlarge)


==========================================================
On 20 September, Dense Fog developed over Tennessee and Alabama, leading to the issuance of Dense Fog Advisories. The GOES-R IFR Probability field, below, shows good agreement between high probabilities and reduced ceilings/visibilities.

GOES-R IFR Probability Fields, Hourly from 0415-1315 UTC on 20 September 2017 (Click to enlarge)

As on the 19th over Missouri, top, this was a night with relatively few middle- and upper-level cloud decks. On such nights, the GOES-16 Brightness Temperature Difference field can capably identify regions of stratus (it’s up to a human to decide if the stratus deck extends to the surface; on this night, much of the stratus did). The 2-hour animation of Brightness Temperature Difference, below, highlights two particular strengths of GOES-16: Better spatial resolution that allows small valleys to be sampled correctly, and good temporal resolution (every 5 minutes vs. every 15 minutes for GOES-13) that allows superior monitoring of the cloud evolution with time. Note that the rising sun is eroding the GOES-16 Brightness Temperature Difference signal by the end of the animation below.

GOES-16 Brightness Temperature Difference (10.3 µm – 3.9 µm), hourly from 1002 to 1202 UTC on 20 September (Click to enlarge)

The Brightness Temperature Difference field (10.3 µm – 3.9 µm) is a key component to the Nighttime Microphysics Red/Green/Blue Composite. As the toggle below shows, the Brightness Temperature Difference field overwhelmingly controls the region identified by the RGB as one with a potential for fog.

Fog over southwest Lower Michigan

GOES-R IFR Probability Fields, 0330-1430 UTC on 14 September 2017 (Click to animate)

GOES-R IFR Probabilities are computed using Legacy GOES (GOES-13 and GOES-15) and Rapid Refresh model information; GOES-16 data will be incorporated into the IFR Probability algorithm in late 2017

GOES-R IFR Probability fields (From this site, but also available via LDM feed in AWIPS), above, show the development of dense for over Lower Michigan, leading to the issuance of advisories. IFR Probabilities on this morning remained fairly low over the relatively warm late Summer waters of Lake Michigan. There is also a noteworthy gap in Fog of unknown origin from Grand Rapids Michigan southwestward to Lake Michigan as fog forms on either side of that line. This is especially evident around 0845 UTC; the gap subsequently fills in as fog becomes more widespread.

GOES-16 data posted on this page are preliminary, non-operational and are undergoing testing

GOES-16 also viewed the fog field, from the top of the cloud deck, as shown below in an animation (courtesy of Nathan Jeruzal the National Weather Service Grand Rapids Office) of the Brightess Temperature Difference field (10.3  µm – 3.9  µm) for two hours before sunrise on 14 September 2017. A still image at 1117 UTC suggests that fog over Grand Rapids’ city limits was not widespread, perhaps due to slight warming in the city due to an Urban Heat Island that would reduce the relative humidity.

At the end of the animation, there is a consistent change in signal over eastern Michigan, from cyan indicating low clouds/stratus to grey, indicating no cloud, because of increasing amounts of reflected solar radiation at 3.9 as the Sun rises. The GOES-R IFR Probability field animation at top suggests that the fog persists through sunrise.

GOES-16 Brightness Temperature Difference fields (10.3 µm – 3.9 µm), 0947-1142 UTC on 14 September 2017;  City outlines are denoted in Yellow (Click to enlarge)

Dense Fog over the central Mississippi River Valley

GOES-16 Brightness Temperature Difference field (10.3 µm – 3.9 µm) from 0417 to 1357 UTC on 28 August 2017 (Click to animate)

GOES-16 data posted on this page are preliminary, non-operational and are undergoing testing.

GOES-16 Brightness Temperature Difference fields (10.3 µm – 3.9 µm), above, show the development of stratus clouds (made up of water droplets) over the Plains during the morning of 28 August 2017.  The Brightness Temperature for 10.3 µm is warmer than that for 3.9 µm during the night because cloud water droplets do not emit 3.9 µm radiation as a blackbody but those same cloud water droplets do emit 10.3 µm radiation more nearly as a blackbody would.   The conversion from sensed radiation to brightness temperature does assume blackbody emissions;  thus, the 3.9 µm brightness temperature is cooler where clouds made up of small water droplets exist.  The animation above shows stratus clouds developing over Missouri and adjacent states.  Dense Fog Advisories were issued near sunrise for much of the region (see image at bottom of this blog post) and IFR Conditions were widespread.

The animation above shows a positive signal over the western High Plains from Kansas northward to North Dakota. (Click here for the view at 1132 UTC on 28 August).

GOES-R IFR Probabilities are computed using Legacy GOES (GOES-13 and GOES-15) and Rapid Refresh model information; GOES-16 data will be incorporated into the IFR Probability algorithm in late 2017

How did GOES-R IFR Probability fields capture this event? The animation showing the fields every 30 minutes from 0215 through 1345 UTC on 28 August 2017, below, shows the development of High Probabilities in the region where Dense Fog was observed. There is a signal along the western High Plains, but it has low Probability; a conclusion might be that thin stratus has developed but that the Rapid Refresh model does not suggest that widespread low-level saturation is occurring. As the sun rises, the signal over the western High Plains disappears. Click here for a toggle between the GOES-R IFR Probability and the GOES-16 Brightness Temperature Difference field at 1115 UTC.

GOES-R IFR Probability fields, 0215-1345 UTC, 28 August 2017 (Click to enlarge)

Screen Capture of http://www.weather.gov at 1300 UTC on 28 August 2017 (Click to enlarge)

Dense Fog over Missouri

GOES-R IFR Probability Fields, and surface reports of Ceilings and Visibilities, hourly from 0315 to 1315 UTC on 15 August 2017 (Click to enlarge)

GOES-R IFR Probabilities are computed using Legacy GOES (GOES-13 and GOES-15) and Rapid Refresh model information; GOES-16 data will be incorporated into the IFR Probability algorithm in late 2017

IFR Conditions and Dense Fog developed over southeastern Missouri during the early morning on 15 August 2017, leading to the issuance of Dense Fog Advisories. GOES-R IFR Probabilities, above, showed increasing values as the ceilings lowered and visibilities dropped.  The IFR Probability fields over southern Missouri and northern Arkansas (and Kansas and Oklahoma) have the characteristic uniformity that arises when Rapid Refresh data alone are used to drive the IFR Probability values.  In these regions, high clouds (associated with convection over Arkansas) are blocking the satellite view of lower clouds.

Such high clouds will make it difficult for a satellite-only product to identify the regions of clouds.  For example, the Brightness Temperature Difference field below (10.3 µm – 3.9 µm) from GOES-16, color-enhanced so that low clouds are green and that cirrus (at night) are purple) shows widespread low cloudiness at the start of the animation, including some obvious river fog over Missouri (river valleys that are not well-resolved with GOES-13), but developing convection over Arkansas eventually prevents the view of low clouds.

GOES-16 data posted on this page are preliminary, non-operational and are undergoing testing

GOES-16 Brightness Temperature Difference field (10.3 µm – 3.9 µm), hourly from 0312 to 1312 UTC on 15 August 2017 (Click to enlarge)

Because the Brightness Temperature Difference field (helpfully called the ‘Fog’ Channel Difference in AWIPS) is challenged by high clouds in viewing the low clouds, RGB Products that use the brightness temperature difference field are also similarly impeded by high clouds.  Consider, for example, the stations in southern Missouri shrouded by the cirrus shield.  IFR Conditions are occurring there and a strong signal of that appears in the IFR Probability fields.

GOES-R IFR Probability computed with GOES-13 and Rapid Refresh Data, GOES-16 Fog (10.3 µm – 3.9 µm) Brightness Temperature Difference and GOES-16 Advanced Nighttime Microphysics RGB, all near 1115 UTC on 15 August 2017 (Click to enlarge)

Dense Fog over the Texas High Plains

GOES-R IFR Probability fields, hourly from 0215-1115 UTC on 2 August 2017 (Click to enlarge)

GOES-R IFR Probabilities are computed using Legacy GOES (GOES-13 and GOES-15) and Rapid Refresh model information; GOES-16 data will be incorporated into the IFR Probability algorithm in late 2017

The National Weather Service in Lubbock issued Dense Fog Advisories (below) for parts of their CWA early in the morning on 2 August 2017.  GOES-R IFR Probability fields, above, show a slow increase in values over west Texas during the night of 1-2 August 2017, as visibilities drop and ceilings lower in the region.  This followed a band of showers that moved through the area around sunset on 1 August (Click here for a visible image from 0017 UTC on 2 August, from this site).  Highest IFR Probability values at the end of the animation generally overlay the Dense Fog Advisory.  As a situational awareness tool for the developing fog/low stratus, IFR Probability performed well.

 

GOES-16 data posted on this page are preliminary, non-operational and are undergoing testing

GOES-16 Brightness Temperature Difference field (10.3 µm – 3.9 µm) at 1117 UTC on 2 August 2017 (Click to enlarge)

The GOES-R IFR Probability fields above mostly show the small-scale variability (i.e., pixelation) that is common when both (legacy) GOES data and Rapid Refresh Data are used to produce a probability that IFR conditions will be present.  Some exceptions:  southeastern New Mexico at the end of animation (1115 UTC);  the yellow and orange region there overlain by mid-level or high clouds that prevent a satellite view of the low clouds.  The GOES-16 Brightness Temperature Difference (10.3 µm – 3.9 µm) field at 1117 UTC shows a signal of high clouds there (cyan / blue / purple enhancement showing negative values that typify thin cirrus in the Brightness Temperature Difference field at night).  The Green values in the color enhancement are positive values and correspond to stratus (composed of water droplets) clouds.  Because the Brightness Temperature Difference field shows a signal, the Advanced Nighttime Microphysics RGB will also have a signal for fog (the whitish/cyan color), as shown below.

GOES-16’s better temporal and spatial resolution allow for more accurate monitoring of the development of small-scale features.  However, the shortcomings of using a Brightness Temperature Difference from satellite to monitor fog development should not be forgotten:  In regions of cirrus, satellite views of low stratus and fog are blocked.  In addition, over Texas and the rest of the High Plains, upslope flow can generate stratus over the central Plains that becomes fog over the High Plains as the terrain rises into the clouds.  The top of the stratus cloud and the fog bank in such a case can look very similar from satellite.

Advanced Microphysics RGB Composite at 1117 UTC on 2 August 2017 (Click to enlarge)

Below is a toggle between the 1115 UTC IFR Probability field, the GOES16 Brightness Temperature Difference Field, and the GOES16 Advanced Microphysics RGB Composite.

GOES-R IFR Probability fields computed with legacy GOES data and Rapid Refresh model output, GOES-16 Brightness Temperature Difference (10.3 µm – 3.9 µm) field and GOES-16 Advanced Microphysics RGB, all near 1115 UTC on 2 August 2017 (Click to enlarge)

 

Resolution: GOES-R IFR Probability Fields and GOES-16 Data

GOES-R IFR Probabilities computed with GOES-13 and Rapid Refresh Data, Hourly from 0215-1115 UTC on 31 July 2017 (Click to enlarge)

GOES-R IFR Probabilities are computed using Legacy GOES (GOES-13 and GOES-15) and Rapid Refresh model information; GOES-16 data will be incorporated into the IFR Probability algorithm in late 2017

The animation above shows the evolution of GOES-R IFR Probability fields over West Virginia early on 31 July 2017, when IFR and Low IFR Conditions developed over much of the state. In addition to elevated probabilities over West Virginia, probabilities increased over eastern Virginia as well, where IFR conditions were not reported. The IFR probabilities over eastern Virginia diminished rapidly at sunrise, as indicated at the end of the animation.

Much of the fog on 31 July 2017 over West Virginia was valley fog. Legacy GOES (GOES-13 and GOES-15) has nominal 4-km resolution at the sub-satellite point, and this resolution can be insufficient to resolve the narrow valleys of the Appalachian Mountains.

GOES-16 data posted on this page are preliminary, non-operational and are undergoing testing

The GOES-16 Animation below shows the 10.3 µm – 3.9 µm Brightness Temperature Difference field for approximately the same time as above. The superior spatial resolution of GOES-16 is evident: tendrils of low clouds/fog are apparent in the animation that until sunrise highlights in green the clouds composed of water droplets (such as fog and stratus). A similar animation of the Nighttime Microphysics RGB Composite (here) similarly highlights stratus (as a whitish color) in the narrow river valleys.

GOES-16 Brightness Temperature Difference Fields (10.3 µm – 3.9 µm), hourly from 0312 – 1112 UTC on 31 July 2017 (Click to enlarge)

This Toggle between the GOES-R IFR Probability and the GOES-16 Brightness Temperature Difference field at 1015 UTC suggests how the IFR Probability Fields will better handle small valley fogs when GOES-16 data are used in the algorithm.

Why Fused Data is better than Satellite Data alone in detecting IFR Conditions: Pennsylvania Example

GOES-R IFR Probability Fields, 0200-1100 UTC on 11 July 2017 (Click to enlarge)

GOES-16 data posted on this page are (still!) preliminary, non-operational data and are undergoing testing

GOES-R IFR Probabilities are computed using Legacy GOES (GOES-13 and GOES-15) and Rapid Refresh model information; GOES-16 data will be incorporated into the IFR Probability algorithm in late 2017

Low ceilings and reduced visibilities developed in and around thunderstorms from Ohio and Michigan into southwestern Ontario, upstate New York and northern Pennsylvania during the morning of 11 July 2017. (Reduced visibilities/lowered ceilings persisted past 15 UTC as shown in this image from here). Regions of IFR conditions over northwestern Pennsylvania are in regions of higher IFR Probability after 0500 UTC (over northwest Pennsylvania) and after 0700 UTC (over much of northern Pennsylvania) that have the characteristic flat field look that comes from having only Rapid Refresh Model output drive the Probability (because high clouds, as might occur downwind of Convection, prevent the satellite from seeing low clouds). GOES-R IFR Probability fields also retain a signal through sunrise, as shown in the this toggle between 1000 UTC and 1100 UTC, two times on either side of the terminator.

Because high clouds inhibit the view of low stratus (and potential fog), products that rely on solely satellite data become ineffectual as a situational awareness tool. Consider the toggle below from 0800 UTC of GOES-R IFR Probabilities, the GOES-16 “Fog Product” Brightness Temperature Difference (10.3 µm – 3.9 µm) and the Advanced Nightime Microphysics RGB (that uses the Brightness Temperature Difference Product as one of its components). IFR Conditions over NW Pennsylvania are not diagnosed by the GOES-16 products that rely on the 10.3 µm – 3.9 µm Brightness Temperature Difference field. Where there is a clear field of view, all three products can highlight IFR conditions (in/around London Ontario, for example). Regions of low stratus — but not fog — are also highlighted over parts of upstate New York by the GOES-16 products. (Similar toggles at 0915 and 1100 UTC are available; note that the signal from GOES-16 becomes weak at 1102 UTC because increasing amounts of solar reflectance change the sign of the 10.3 µm – 3.9 µm Brightness Temperature Difference.

These examples typify why GOES-R IFR Probability fields typically have better statistics as far as IFR detection is concerned: Model data fills in regions where high clouds are present, and model data screens out regions where stratus is highlighted by a Brightness Temperature Difference field, but where fog does not exist.

GOES-R IFR Probabilities (computed using GOES-13 and Rapid Refresh Data), GOES-16 “Fog Product” Brightness Temperature Difference (10.3 µm – 3.9 µm) and the Advanced Nighttime Microphysics RGB, all around 0800 UTC on 11 July 2017 (Click to enlarge)