Author Archives: Scott Lindstrom

Dense Fog over the Carolinas

GOES-13 Brightness Temperature Difference (3.9 µm – 10.7 µm), GOES-R IFR Probability and GOES-R Cloud Thickness at 1115 UTC on 17 January 2017 (Click the enlarge)

Dense Fog Advisories (National Weather Service Website) and IFR SIGMETs (Aviation Weather website) were issued early in the morning for dense fog over the southeastern United States. The toggle above from 1115 UTC on 17 January shows the Brightness Temperature Difference field (3.9 µm – 10.7 µm), the GOES-R IFR Probability Field, and the GOES-R Cloud Thickness fields associated with this dense fog event. Note the presence of high clouds over northern South Carolina and western North Carolina — the dark region in the Brightness Temperature Difference enhancement — prevents the brightness temperature difference field from highlighting that region of reduced ceilings/visibilities. The GOES-R Cloud Thickness field is not computed under cirrus either, as it relates 3.9 µm emissivity of water-based clouds to cloud thickness (based on a look-up table generated using data from a SODAR off the West Coast of the United States). If cirrus blocks the view, then, neither the Brightness Temperature Difference field nor the GOES-R Cloud Thickness field can give useful information about low clouds.

In contrast, the GOES-R IFR Probability field does give useful information in regions where cirrus clouds (and low clouds/fog) are present — because Rapid Refresh information about the lower troposphere can be used. IFR Probability values will be smaller in those regions because satellite predictors are unavailable, and the Probability incorporates both predictors from satellites and from Rapid Refresh model output — if the satellite predictors are missing because of cirrus, the IFR Probability values will be affected. Despite the smaller values, however, the IFR Probability fields in regions of cirrus are giving useful information for this event.

GOES-R Cloud thickness fields can be used to estimate Fog dissipation using the last GOES-R Cloud Thickness field produced before twilight conditions at sunrise (shown below for this case). (GOES-R Cloud Thickness is not computed during twilight conditions because of rapidly changing 3.9 µm emissivity related to the reflected solar radiation as the sun rises, or as it sets). This scatterplot gives the relationship between thickness and dissipation time after the Cloud Thickness time stamp (1215 UTC in this case). In this case, the thickest fog is near Athens GA; the algorithm predicts that clearing should happen there last, at about 1515 UTC.

GOES-R Cloud Thickness, 1215 UTC on 17 January 2017 (Click to enlarge)

Widespread IFR Conditions over the Plains

GOES-R IFR Probability Fields, hourly from 0115-1315 UTC (Click to enlarge)

A cyclone over the southern Plains, in addition to causing severe weather over Texas on 15 January also generated widespread IFR Conditions over the southern Plains, as shown below in screengrabs from the Aviation Weather Center and from the National Weather Service. An overnight Water Vapor image (here) testifies to the ubiquitous presence of high clouds over the Plains; in such cases with widespread high clouds, low-cloud detection by satellite is a big problem. A strength of the GOES-R IFR Probability field is that it is a fused data product, incorporating both satellite information (not particularly useful for much of the overnight hours on 15-16 January) and Rapid Refresh model data that can be used to discern conditions near the surface. When the Rapid Refresh model suggests saturation is occurring near the surface (in, say, the lowest 1000 feet of the model atmosphere), IFR Probabilities will be large. They won’t be as large as they might be if both satellite and model data suggest low clouds are present, but useful information emerges in the IFR Probability fields, above, where the Rapid Refresh is predicting low-level saturation. IFR Probabilities are large over much of the southern Plains where IFR conditions are observed. This is the region where the color enhancement is orange.

The low pressure system develops such that high clouds diminish over Texas and Oklahoma. When that happens, the IFR Probability fields change in two ways. First, values increase because satellite data and model data can be used as predictors. When only model data can be used, IFR Probability fields will have smaller values. Secondly, the character of the IFR Probability field takes on a more pixelated appearance because the satellite data values will vary from pixel to pixel. In contrast, when only model data can drive the IFR Probability field (for example, over Kansas at the beginning of the animation), the IFR Probability fields vary quite slowly from pixel to pixel in part because of model smoothing.

Screen Capture from Aviation Weather Center (left, showing widespread IFR Conditions) and from Weather.Gov (right, showing Dense Fog Advisories in grey) (Click to enlarge)

The toggle below includes sampling over Abilene, TX (KABI), a station at the edge of the IFR Probability field. IFR Probabilities are relatively constant at ~40% for the two hours shown, but station conditions change from IFR to VFR. IFR Probabilities at Abilene become quite small by 1315 UTC, at the end of the animation above.

GOES-R IFR Probability at 1015 and 1215 UTC on 16 January 2017. Station conditions at KABI are indicated by the sample probe (Click to enlarge)

Stratus over Texas

GOES-13 Visible (0.64 µm) Imagery, 1945 UTC on 6 January 2017 and surface observations of ceilings and visibilities (click to enlarge)

Visible imagery over Texas shows an extensive stratus deck blanketing the southern and eastern portions of the state. Can you tell at a glance — without looking at the observations — if the stratus is extending to the surface? The animation below shows how GOES-R IFR Probabilities describe the scene, with highest IFR Probabilities offshore (where dense fog is observed over the warm water). Higher Probabilities also hug the high terrain of eastern Mexico, where IFR conditions are also reported (at Monclova, ID MMMV, where a 500-foot ceiling and 3-mile visibility is reported). The toggle below cycles through the visible and GOES-R IFR Probability fields and also includes terrain.

GOES-R IFR Probability provides useful situational awareness information during the daytime as well as at night.

GOES-13 Visible (0.64 µm) Imagery, 1945 UTC on 6 January 2017 and surface observations of ceilings and visibilities, and with surface analysis superimposed, as well as GOES-R IFR Probabilities (1945 UTC) and Terrain (click to enlarge)

Persistent IFR Conditions over Kansas

GOES-R IFR Probability fields, 0400-1700 UTC on 2 January 2017 (Click to enlarge)

Late in the morning on Monday 2 January 2017, the weather.gov website shows Dense Fog Advisories persisting in parts of Kansas and Nebraska. The animation above shows GOES-R IFR Probabilities from 0400 through 1700 UTC on the 2nd. High IFR Probabilities show excellent correspondence with very low ceilings and reduced visibilities. Note, for example, the sharp demarcation at 1700 UTC between IFR conditions (and large IFR Probabilities) and VFR/MVFR conditions (and very low IFR Probabilities) over northwestern Kansas/east central Colorado.

In the animation above, large parts of central Kansas have IFR Probabilities that are uniform, and around 52% (indicated by the mustard orange color). The IFR Probability field has this character at that time — and later in the animation over much of eastern Kansas and Missouri/Iowa — because of high clouds that are present. When high clouds prevent the satellite from viewing low clouds, model data in the form of Rapid Refresh estimates of low-level saturation are the driving force behind the IFR Probability value. The animation below of Brightness Temperature Difference fields shows the characteristic dark streak that in the enhancement used represents high clouds at night. High clouds then spread over Missouri and Iowa during the day.

GOES-13 Brightness Temperature Difference fields (3.9 µm – 10.7 µm), 0400-1700 UTC on 2 January 2017 (Click to enlarge)

GOES-R Cloud Thickness can be used in cases of radiation fog to estimate dissipation time, using the value just before sunrise, here, and this scatterplot. Low clouds on 2 January however were synoptically forced so it would be inappropriate to expect the GOES-R Cloud Thickness field to estimate dissipation times correctly.

Dense Fog in Oregon

GOES-R IFR Probabilities computed with GOES-15 and Rapid Refresh Data, hourly from 0300 through 1400 UTC on 21 December 2016 (Click to enlarge)

Dense Fog developed in the Willamette Valley of western Oregon during the early morning hours of 21 December 2016. How did GOES-R IFR Probability fields and GOES-15 Brightness Temperature Difference (3.9 µm – 10.7 µm) Fields diagnose this event that led to the issuance of Dense Fog Advisories? The hourly GOES-R IFR Probability animation, above, shows increasing probabilities in the Willamette Valley, starting around Eugene (KEUG) and spreading northward until high probabilities cover the valley by 1400 UTC. IFR Conditions are first reported near Eugene, then through the entire valley by 1400 UTC. Widespread high IFR Probability values are not present elsewhere over Oregon (although they do exist over Washington State, where IFR conditions were also observed).

The Brightness Temperature Difference field (3.9 µm – 10.7 µm), below, shows a different distribution. Early in the animation a strong signal is apparent over much of northern Oregon, and also over the Pacific Ocean. Rapid Refresh output on low-level saturation is used in the GOES-R IFR probability algorithm to screen out regions where stratus that is detected by the satellite likely is not extending down to the surface — over the ocean, for example (coastal sites do not show IFR Conditions), or over much of eastern Oregon/Washington. Brightness Temperature Difference fields do eventually highlight the presence of fog in the Willamette Valley, but considerable regions outside the valley have a strong return and no indication of IFR conditions.

GOES-15 Brightness Temperature Difference (3.9 µm – 10.7 µm) fields, 0300-1300 on 21 December 2016 (Click to enlarge)

GOES-R IFR Probability fields will frequently (and correctly) screen out regions of strong returns in the Brightness Temperature Difference fields that do not correspond to surface obscuration of visibility and/or low ceilings.

It is possible to alter the Brightness Temperature Difference Colormap, as below (animation courtesy Mike Stavish, SOO at Medford) to better highlight regions of fog in this case. Note in this enhancement the cirrus clouds appear white, rather than dark, as well.

GOES-15 Brightness Temperature Difference (3.9 µm – 10.7 µm) Fields, hourly from 0800 to 1400 UTC on 21 December 2016 (Click to enlarge)

The toggle below shows two color enhancements at 1200 UTC: the default version with many orange pixels, and an altered version that shows fewer pixels, mostly in regions where fog is present. A similar toggle for 0400 UTC is here.

Brightness Temperature Difference (3.9 µm – 10.7 µm) fields at 1200 UTC with two different enhancements. (Click to enlarge)

IFR Conditions across the Deep South

GOES-R IFR Probability computed with GOES-13 and Rapid Refresh Model Data, hourly from 0215 through 1415 UTC on 20 December 2016 (Click to enlarge)

Widespread IFR Conditions developed across Mississippi, Alabama, Georgia and neighboring states on Tuesday morning 20 December, as evidenced by the Aviation Weather Center website screen grab at the bottom. GOES-R IFR Probabilities captured the evolution of the low ceilings and reduced visibilities. In particular, note in the animation above the westward progress of the higher IFR probabilities through Mississippi; IFR conditions develop at, for example, Jackson (KJAN), Greenwood (KGWO) and Oxford (KUOX) as the high probabilities move over the station. Its motion was useful as a forecast tool on this morning.

The GOES-R IFR Probability field has noticeable stripes in it at the end of the animation. This occurs because high clouds have overspread the low stratus/fog. When that happens, satellite data can no longer be used as a predictor in the IFR Probability algorithm because the satellite can no longer view the low clouds; Rapid Refresh data alone are driving the values. The toggle between the brightness temperature difference field (3.9 µm – 10.7 µm) and the IFR Probability field, below, from 1115 UTC on 20 December, shows the effect. Where fog/stratus are present (yellow in the enhancement used for the brightness temperature difference field), IFR Probability values are larger because satellite and model data can be used to compute IFR Probability. Where high clouds are present (dark grey in the brightness temperature difference enhancement), only Rapid Refresh data can be used.

GOES-R IFR Probability and GOES-13 Brightness Temperature Difference (3.9 µm – 10.7 µm) fields, 1115 UTC on 20 December 2016 (Click to enlarge)

Aviation Weather Center website screengrab, 1443 UTC on 20 December 2016 (Click to enlarge)

Fog over South Texas

Toggle between Brightness Temperature Difference (3.9µm – 10.7µm) and GOES-R IFR Probability fields, 2300 UTC on 11 December 2016 (Click to enlarge)

Dense Fog developed over south Texas during the early morning of 12 December 2016 (IFR Sigmet from this website shown here ; Advisories from the weather.gov website shown here). The toggle above shows in the brightness temperature difference field a signature of high clouds — and where those high clouds exist, IFR Probability fields rely on Rapid Refresh Model data to diagnose where IFR conditions might be occurring, or where IFR conditions might develop. The animation of Brightness Temperature Difference fields from 0215 through 1115 UTC, below, shows that the high clouds over south Texas diminished with time: by 0815 UTC only low stratus is present over south Texas. But is that stratus also hugging the ground — that is, is it fog? From the satellite’s perspective, the top of a stratus deck and the top of a fog bank can look very similar.

GOES-13 Brightness Temperature Difference (3.9µm – 10.7µm), 0215 through 1115 UTC on 12 December (Click to enlarge)

GOES-R IFR Probability fields give a more complete estimate about the presence of fog/low stratus because Rapid Refresh data and satellite data are used to diagnose the probability of IFR conditions. If the Rapid Refresh model shows low-level saturation, then the presence of stratus clouds also likely indicates the presence of fog; conversely, if the Rapid Refresh Model does not show low-level saturation, then the presence of stratus cloud need not indicate the presence of fog. IFR Probability fields below, from 0215 through 1115 UTC, start off regions with uniform values where only Rapid Refresh data are used in the algorithm — where high clouds block the satellite view of low clouds/fog. As the high clouds dissipate, the field acquires larger values because there is higher confidence of the presence of clouds (in part because satellite data can be used to observe them). In addition, these larger values have pixel-sized variability because of variability in the satellite observations.

IFR conditions are observed latest over far south Texas — this is also where IFR Probabilities are slowest to reach large values.

Fog and Ice Fog over the southern Plains

GOES-R IFR Probabilities, hourly from 0400 through 1215 UTC on 5 December 2016 (Click to enlarge)

Dense Fog developed over the southern Plains early on Monday 5 December, and the GOES-R IFR Probability fields, above, were a tool that could be used to monitor the evolution of this event. A challenge presented on this date was the widespread cirrus (Here’s the 0700 UTC GOES-13 Water Vapor (6.5 µm) Image, for example) that prevented satellite detection of low clouds. The Brightness Temperature Difference fields, below, at 3-hourly intervals, also show a signature (dark grey/black in the enhancement used) of high clouds, although they are shifting east with time — by 1300 UTC there is a signature (orange/yellow in the enhancement used) of stratus clouds over central and eastern Oklahoma.

The IFR Probability fields, above, have a characteristic flat nature over Arkansas and Missouri, that is, a uniformity to the field, that is typical when model data are driving the probabilities. The more pixelated nature to the fields over Kansas and Oklahoma, especially near the end of the animation, typifies what the fields look like when both satellite data and model data are driving the computation of probabilities. Careful inspection of the fields over Arkansas shows regions — around Fayetteville, for example, around 1000 UTC where IFR Probabilities are too low given the observation at the airport of IFR conditions. This inconsistency gives information on either the small-scale nature of the fog (unlikely in this case) or on the accuracy of the Rapid Refresh model simulation that is contributing to the probabilities. In general, the Rapid Refresh model has accurately captured this event, and therefore the IFR Probabilities are mostly overlapping regions of IFR or near-IFR conditions. The region over southern Illinois that has stratus, and low probabilities of IFR conditions, for example. Adjacent regions have higher IFR Probabilities and lower ceilings and/or reduced visibilities. A screen shot from the National Weather Service, and from the Aviation Weather Center, at about 1300 UTC document the advisories that were issued for this event.

Brightness Temperature Difference fields (3.9 µm – 10.7 µm), 0400, 0700, 1000 and 1300 UTC on 5 December 2016 (Click to enlarge)

IFR Conditions over the High Plains of west Texas

GOES-13 Brightness Temperature Difference (3.9 µm – 10.7 µm, left) and GOES-R IFR Probability Fields (Right), hourly from 0215 through 1415 UTC on 2 December 2016 (Click to enlarge)

Near-IFR and and IFR Conditions developed over the High Plains of Texas on 2 December 2016, and a SIGMET for IFR conditions was issued as shown below.

The animation above shows plentiful cirrus (in the brightness temperature difference enhancement used in the imagery on the left, above, cirrus clouds are dark) over south Texas, with occasional breaks. This makes continual monitoring via satellite of the developing stratus/fog field problematic: the satellite cannot monitor what it is blocked from being observed by intervening cloud layers — in this case cirrus. (Click here for a brightness temperature difference only animation; Click here for an IFR Probability only animation) Because IFR Probability fields include model-based data about saturation in the lower troposphere, in the form of Rapid Refresh model output, a useful and coherent signal can be generated underneath cirrus clouds. The GOES-R IFR Probability signal can be better used for situational awareness and anticipation of the development of the IFR conditions shown below.

In the animation above, note the change between 1315 and 1415 UTC fields — in the Brightness Temperature Difference fields (1315 UTC ; 1415 UTC), this change arises because of increasing amounts of reflected solar 3.9 µm radiation: this causes a sign change in the brightness temperature difference. For IFR Probability fields (1315 UTC ; 1415 UTC), the change occurs because the Predictors used at night (1315 UTC) and during the day (1415 UTC) are different.

1605 UTC screen capture from Aviation Weather Center. Note IFR Sigmet over west Texas (Click to enlarge)

Dense Fog in Georgia and Florida

GOES-R IFR Probability Fields, 0000 – 1300 UTC on 25 November 2016 (Click to enlarge)

Light winds and a long November night allowed radiation fog formation over much of the deep south early on 25 November 2016. (1200 UTC Surface analysis is here). The Aviation Weather Center Website indicated widespread IFR Conditions, below, over the south, with the sigmet suggesting improving visibilities after 1400 UTC.

The animation above shows the evolution of the GOES-R IFR Probability fields from just after sunset to just after sunrise. There is a good spatial match between observed IFR conditions and the developing field. IFR Probability can thus be a good situational awareness tool, identifying regions where IFR Conditions exist, or may be developing presently.

Screenshot of Aviation Weather Center Front Page, 1405 UTC on 25 November (Click to enlarge)

Did the GOES-13 Brightness Temperature Difference Field identify the fields? The animation below, from 0200-1300 UTC, shows a widespread signal that shows no distinguishable correlation with observed IFR conditions. Note also how the rising sun at the end of the animation changes the difference field as more and more reflected solar radiation with a wavelength of 3.9 µm is present. In addition, high clouds that move from the west (starting at 0400 UTC over Louisiana) prevent the satellite from viewing low clouds in regions where IFR conditions exist.

GOES-13 Brightness Temperature Difference field (3.9 µm – 10.7 µm), 0200-1300 UTC on 25 November 2016 (Click to enlarge)

The high clouds that prevent satellite detection of low clouds, as for example at 1100 UTC over parts of Alabama, cause a noticeable change in the IFR Probability fields, as shown in the toggle below. Values over the central part of the Florida panhandle are suppressed, and the field itself has a flatter character (compared to the pixelated field over southern Georgia, for example, where high clouds are not present). Even though high clouds prevent the satellite from providing useful information about low clouds in that region, GOES-R IFR Probability fields can provide useful information because of the fused nature of the product: Rapid Refresh information adds information about low-level saturation there, so IFR Probability values are large. In contrast, over southern Florida — near Tampa, for example, Rapid Refresh data does not show saturation, and IFR Probabiities are minimized even through the satellite data has a strong signal — caused by mid-level stratus. Soundings from Tampa and from Cape Kennedy suggest the saturated layer is around 800 mb.

GOES-R IFR Probability fields and GOES-13 Brightness Temperature Difference Fields, 1100 UTC on 25 November 2016 (Click to enlarge)

GOES-R Cloud Thickness, below, related 3.9 µm emissivity to cloud thickness via a look-up table that was generated using GOES-West Observations of marine stratus and sodar observations of cloud thickness. The last pre-sunrise thickness field, below, is related to dissipation time via this scatterplot. The largest values in the scene below are around 1000 feet, which value suggests a dissipation time of about 3 hours, or at 1445 UTC.

GOES-R Cloud Thickness Field, 1145 UTC on 25 November 2016 (Click to enlarge)

GOES-R Fog Product Examples

Fog detection fusing GOES, Terra/Aqua or Suomi/NPP Satellite data with Model output