Author Archives: Scott Lindstrom

Fog around Puget Sound

GOES-15-based GOES-R IFR Probabilities (Upper Left), GOES-15 Brightness Temperature Difference Product (10.7 µm – 3.9 µm) (Lower Left), MODIS-based GOES-R IFR Probabilities (Upper Right), Suomi-NPP Day/Night Band (Lower Right), all times as indicated (click image to enlarge)

IFR Conditions developed around the Puget Sound during the night of 20 October. How did the GOES-R IFR Probabilities capture this event? The animation above includes imagery from 0500, 0900, 0945, 1115 and 1915 UTC. Higher-resolution polar orbiter data (from MODIS and Suomi/NPP) shows the value of higher-resolution in capturing fog that settles into valleys over southeast British Columbia and western Washington. GOES data are unable to resolve those features.

The Brightness Temperature difference fields have a strong signal over the Pacific Ocean and adjacent coastal areas (IFR Probabilities are high in those regions: both satellite data and Rapid Refresh data are consistent with a high likelihood of fog/low stratus). Over land, the signal is more noisy, perhaps because of differences in land emissivity. (That noise is not present when the sun is up — at that time the brightness temperature difference signal is determined by reflected solar radiation). Where the brightness temperature difference signal is smaller over land, the IFR Probability is also lower. That it is not even smaller suggests that model fields are at or near saturation over land. Note also a strength of the IFR Probability: A consistent signal both day and night. IFR Probabilities are high over Seattle where IFR conditions persist.

As above, but for times with Day/Night band data at night only (click image to enlarge)

The Suomi-NPP Day/Night band can give a good indication of where clouds are present at night when, as occurred last night, the moon is near full. (The Day/Night band does not, however, by itself give any indication of surface visibility) In the example above, the clouds do not change much in the 90 minutes between overpasses. (The slight shift in the apparent location of snow-covered mountains is apparently due to parallax) GOES can just barely resolve the very thin fog features that are so evident in the Suomi/NPP data.

IFR Conditions over southwest Alaska

Strong extratropical storms that move northward into the Gulf of Alaska, or into the Bering Sea, can bring IFR conditions to many parts of Alaska. However, they typically also bring multiple cloud layers that make traditional satellite-only methods of detecting fog and low stratus problematic. In cases like these, a fused product that incorporates model predictions of low-level saturation is helpful in defining just where IFR conditions are most likely.

GOES-15 Brightness Temperature Difference Product (10.7 µm – 3.9 µm), times as indicated (click image to enlarge)

For example, the brightness temperature difference field, above, does not show a strong signal in regions where near-IFR conditions are present. In contrast, the IFR Probability field, below, that incorporates model fields that are influenced by surface features, better highlights the region of IFR conditions. It captures the edge of the fog/low stratus field over SW Alaska, and probabilities are highest in regions where IFR and near-IFR conditions exist. The relatively flat field over land is a typical feature of the IFR Probability when it is determined chiefly by model data. Because satellite data are not included in the predictors, the total probability is somewhat smaller. Where the satellite brightness temperature difference field does have a strong signal is where the IFR Probabilities are highest (over the Bering Sea).

GOES-R IFR Probabilities derived from GOES-15 and Rapid Refresh data Brightness Temperature Difference Product (10.7 µm – 3.9 µm), times as indicated (click image to enlarge)

One shortcoming with IFR Probability is the pixel resolution at high latitudes. MODIS data can also be used to compute IFR Probabilities, and three comparisons between MODIS and GOES values are shown below. Alaska’s high latitudes means not only large GOES pixels, but also fairly frequent coverage from the polar-orbiting Terra and Aqua satellites that hold the MODIS instrument.

Toggle between GOES-R IFR Probabilities derived from GOES-15 and from MODIS satellite data at ~0900 UTC 15 October (click image to enlarge)

As above, but at ~1300 UTC (click image to enlarge)

As above, but at ~1430 UTC (click image to enlarge)

Fog over northern Vermont and New Hampshire

GOES-13-based GOES-R IFR Probabilities (Upper Left), GOES-13 Brightness Temperature Difference Product (10.7 µm – 3.9 µm) (Upper Right), GOES-13-based GOES-R Cloud Thickness (Lower Left), AVHRR Brightness Temperature Difference (10.8 µm – 3.74 µm) (Lower Right), all times as indicated (click image to enlarge)

High Pressure over southern Quebec allowed for light winds over northern New England, and fog and low stratus developed. The animation above shows the benefit of the fused product; when mid- and high-level clouds are present in the satellite field of view, the brightness temperature difference product loses the ability to detect fog and low stratus. Data from the Rapid Refresh Model will be used in these regions to produce an IFR Probability Field. Note how Burlington, VT, for example, reports IFR conditions. The brightness temperature difference product has only a small signal over Burlington, but IFR Probabilities are high.

Resolving fog in the Cumberland Valley

Toggle between GOES-13-based and MODIS-based GOES-R IFR Probabilities at ~0415 UTC on 11 October 2013 (click image to enlarge)

The toggle between the GOES-based and MODIS-based IFR Probabilities, above, shows how high-resolution MODIS data can give an earlier alert to the formation than is possible from GOES. Fog formed in the Cumberland Valley of central Tennessee and southeast Kentucky early in the morning of 11 October. MODIS-based IFR Probabilities over the river valley are peaking around 45% at 0415 UTC (vs. 5% for GOES-based IFR Probabilities). In fact, GOES-based IFR Probabilities do not reach 45% until about 0515 UTC, a full hour later. The toggle below shows the two brightness temperature difference products used at ~0415 UTC to make the IFR Probability fields. MODIS data are better able to resolve the small-scale river valleys where fog is forming earlier.

Toggle between GOES-13-based and MODIS-based Brightness Temperature Differences at ~0415 UTC on 11 October 2013 (click image to enlarge)

The animation below shows the fog quickly burning off in the morning. It has dissipated by 1500 UTC.

GOES-13 Visible Imagery during the morning of 11 October 2013 (click image to enlarge)

Fog Development near Lake Michigan

GOES-13-based GOES-R IFR Probabilities (Upper Left), GOES-13 Brightness Temperature Difference Product (10.7 µm – 3.9 µm) (Upper Right), GOES-13-based GOES-R Cloud Thickness (Lower Left), Suomi/NPP Brightness Temperature Difference (Lower Right), all near 0615 UTC on 10 October (click image to enlarge)

The GOES-R IFR Probability product gave useful advance warning to the development of fog near Lake Michigan’s eastern shore overnight. The image above, from 0615 UTC, shows a flat brightness temperature difference field over the lakeshore counties in Wisconsin and Illinois (values are from -7.1 to -7.3); there are two regions of high values in the IFR Probability field, however: Near Manitowoc WI (values up to 29%) and over southeast WI and northeast IL (values near 20%). So by 0615 UTC on 10 October, IFR Probabilities are suggestive of a nascent fog development.

As above, but for 0702 UTC on 10 October (click image to enlarge)

Forty-five minutes later, at 0702 UTC (above), IFR Probabilities have increased dramatically in eastern WI even as the brightness temperature difference field remains flat. Thus, the Rapid Refresh Data is accurately capturing the development of low-level saturation in the atmosphere, and that is influencing the IFR probability field. In addition, the GOES-R Cloud Thickness field is suggesting that the cloud bank is 500-600 feet thick. The strip of enhanced brightness temperature difference paralleling the Lake Michigan shore in lower Michigan is an artifact of the co-registration error between the 10.7 µm and 3.9 µm band detectors on GOES-13. Between 0656 UTC and 0734 UTC, visibility at Manitowoc, WI (KMTW), dropped from 5 to 3/4 statute miles. The visibility at Burlington WI (KBUU) dropped from 4 to 1 statute miles between 0600 and 0700 UTC, and Waukegan, IL (KUGN) reported a visibility of 1/4 mile at 0552 UTC and 0652 UTC.

As above, but for 0802 UTC on 10 October (click image to enlarge)

At 0802 UTC, the GOES-East brightness temperature difference field shows greater differences over the region of SE Wisconsin where the fog is developing. Accordingly, the IFR probability increases past 80% IFR Probabilities are near 70% in Manitowoc County (and Manitowoc reported 1/2-mile visibility at 0834 UTC). Compare the GOES-East and Suomi/NPP Brightness Temperature Difference Fields; note the lack of a signal in the Suomi/NPP field along the western shore of Lake Michigan, confirming the co-registration error present in GOES-13.

As above, but for 1145 UTC on 10 October (click image to enlarge)

The last pre-sunrise image, 1145 UTC, shows a definite signal of fog/low stratus in both the IFR Probability field and in the Brightness Temperature Difference field. However, the early detection in the IFR Probability field gives a nice head’s up to the forecaster. Note also in this image how the strong signal in the brightness temperature difference field that arises because of the co-registration error can contaminate the IFR Probability field. The Cloud Thickness in this field has been related to dissipation time, as shown in this chart. The maximum thickness of 1000 feet predicts a dissipation time around 1500 UTC. The 1445 and 1515 UTC GOES-13 visible images are shown below.

GOES-13 Visible Imagery at 1445 UTC and 1515 UTC on 10 October (click image to enlarge)

IFR Conditions in the Southeast

GOES-13-based GOES-R IFR Probabilities (Upper Left), GOES-13 Brightness Temperature Difference Product (10.7 µm – 3.9 µm) (Upper Right), GOES-13-based GOES-R Cloud Thickness (Lower Left), Suomi/NPP Brightness Temperature Difference (Lower Right), all times as indicated (click image to enlarge)

High Pressure of the southeast US allowed for clear skies and light winds overnight, and radiation fog developed over coastal portions of eastern Georgia. Because high clouds were present, the traditional method for detecting fog and low stratus, the brightness temperature difference between 10.7 µm and 3.9 µm on GOES could not capture the entire areal extent of the cloud. Fog is initially reported in eastern Georgia where IFR Probabilities are increasing underneath an ice-phase cloud deck that prevents the GOES satellite from seeing the development of low clouds and fog. Accordingly, the IFR probabilities are lower than they would be if satellite data were included. The uniform nature of the field is testament to the use of Rapid Refresh Data to drive the IFR Probability field. Shortly after 0600 UTC, a satellite signal develops over South Carolina as the high clouds shift to the south. When this happens, IFR probabilities increase (and acquire a more pixelated look). Regions under high clouds with IFR conditions persist through the end of the loop, however. It’s important to have a fused data product that allows two different complementary fields to diagnose where IFR conditions are likely. Where the brightness temperature difference cannot be used, Rapid Refresh Data gives vital information. Where brightness temperature difference data can be used, the Rapid Refresh Data can fine-tune things.

As above, but for 1115 UTC (click image to enlarge)

The 1115 UTC image, above, shows the GOES-R cloud thickness just before twilight conditions that accompany sunrise prohibit its computation. The cloud thickness in a radiation fog is related to dissipation time, and the cloud thickness is shown to be quite thin: Thickest values, in blue, are around 800 to 900 feet. This scatterplot relates cloud thickness to dissipation time, and it suggests a dissipation time of 1-2 hours. However, the visible imagery animation, below, shows actual dissipation occurred shortly after 1500 UTC. Note that there is considerable spread in that predictive scatterplot.

GOES-13 Visible Imagery, hourly at 1315, 1415 and 1515 UTC on 4 October (click image to enlarge)

IFR Conditions on the West Coast as a front passes

GOES-15-based GOES-R IFR Probabilities (Upper Left), GOES-15 Brightness Temperature Difference Product (10.7 – 3.9 ) (Upper Right), Color-shaded topographic map (Lower Left), GOES-15 Visible imagery and Suomi/NPP Day/Night Band (Lower Right), all times as indicated (click image to enlarge)

A strong low pressure system — the first strong system of the Fall Season — has made landfall along the Pacific Northwest Coast, and it provides an opportunity to see how GOES-R IFR Probabilities perform with extratropical systems.

Several aspects of the IFR Probability Fields — which are far more coherent than the Brightness Temperature Difference fields — require explanation. There is an increase in the IFR Probability off the coast of Oregon in the 4th image in the loop. This jump — from IFR Probabilities near 55% (orange) to Probabilities near 68% (darker red-orange) — is likely caused by a changed in the Rapid Refresh model output that is suggesting a greater likelihood of low-level saturation. Note that this region in the very next image displays the characteristic signature of the boundary between day-time predictors being used and night-time predictors being used (IFR Probabilities drop from orange yellow — 39%). In the early part of the animation, the IFR Probability field off the Oregon Coast maintains the flat (un-pixelated) look that is characteristic of a region where only Rapid Refresh model output is being used in the computation of IFR Probabilities because high cloud are present. GOES-R Cloud Thickness (of the highest water-phase cloud, not shown) would not be computed in this region, then, for two reasons: ice and mixed-phase clouds are present and it is during twilight conditions.

IFR and near-IFR Conditions are observed along the coast at Newport and North Bend before the frontal passage. IFR Probabilities are high along the coastal range, somewhat reduced in the Willamette Valley, high again in the Cascades, and lower again downstream of the Cascades. Note how the higher IFR Probabilities do take into account the presence of terrain; brightness temperature difference fields use only satellite data. Thus, after the frontal passage, when near-surface winds are west and south-west (that is, upslope), the IFR Probabilities remain high on the windward side of mountain slopes. (They are typically high, for example, at Sexton Summit — over KSXT — where IFR conditions are present until about 1100 UTC)

Suomi-NPP Day/Night band data can sometimes be used to discern regions of cloudiness. However, the Moon Phase is now a waning crescent that has not quite risen at the times shown in the animation above.

Resolution and Valleys

GOES-based GOES-R IFR Probabilities, 0345 UTC 24 September 2013 (click image to enlarge)

Consider the GOES-R IFR Probabilities computed from GOES-East data (and Rapid Refresh data) above. How confident are you that, at 0345 UTC, fog is forming in river valleys of western Pennsylvania? Is the likelihood the same in the southern part of the state (say, along the Monongahela River) as in the northern part of the state (along the Clarion or Allegheny Rivers)? GOES resolution in the infrared channels is 4 km at the sub-satellite point. In Pennsylvania, resolution is degraded to 5 or so kilometers. The knowledge of pixel size should color your interpretation of the GOES-R IFR Probabilities (and of the brightness temperature difference field computed from GOES). The MODIS-based GOES-R IFR Probabilities from 0339 UTC, below, show a ribbon of high probabilities over many of the river valleys of Pennsylvania. This 1-km resolution information is handy at capturing the initial development of fog and low stratus.

MODIS-based GOES-R IFR Probabilities, 0339 UTC 24 September 2013 (click image to enlarge)

Suomi/NPP VIIRS Day/Night band imagery, 0640 and 0820 UTC 24 September 2013 (click image to enlarge)

Day/Night band imagery over Pennsylvania and New York shows the expansion of fog coverage between successive Polar Passes, at 0640 and 0820 UTC. The imagery below shows the corresponding GOES-based GOES-R IFR Probabilities at those two times. The large cloud features over northeast Pennsylvania and the Southern Tier of New York are captured well by the GOES-based fields; the river valley fogs are not captured quite so well because of resolution limitations.

GOES-based GOES-R IFR Probabilities, 0645 and 0815 UTC 24 September 2013 (click image to enlarge)

A MODIS-based IFR probability field, below, far better represents the presence of River Valley fogs at 0746 UTC than the GOES-based IFR Probability Field, bottom, from 0745 UTC. (These times are between the two times in the GOES-R IFR Probability animation above) A good method for monitoring fog would incorporate the fine spatial resolution at the start of the fog event to ascertain which river valleys are starting first to become fog-bound. The good temporal resolution of GOES data is then used to outline the evolution of the event. Periodic Polar Orbiter passes from Terra, Aqua of Suomi/NPP as the fog event is occurring can confirm the GOES-based predictions of the evolution of fog.

MODIS-based GOES-R IFR Probabilities, 0746 UTC 24 September 2013 (click image to enlarge)

GOES-based GOES-R IFR Probabilities, 0745 UTC 24 September 2013 (click image to enlarge)

Day/Night Band Imagery of Fog near Lake Superior

Suomi/NPP VIIRS Day/Night Band Visible Imagery, 0640 and 0820 UTC 23 September 2013 (click image to enlarge)

The orbital geometry of Suomi/NPP is such that one geographic region will be scanned on two successive polar passes, about 90 minutes apart. The likelihood that this will happen increases as you approach the Poles. On the morning of 23 September, 2013, western Lake Superior was thus viewed twice as fog developed. This was also a night shortly after a Full Moon so ample lunar illumination allowed for a distinct view of the evolution of the fog. The 0645 UTC Day/Night band imagery shows what appears to be a much thinner cloud bank over the far northeast Minnesota Arrowhead. By 0820 UTC the cloud bank has a thicker look. Note the corresponding changes in sky/visibility at Thunder Bay (CYQT) and Grand Marais (KCKC).

Suomi/NPP Brightness Temperature Difference (11.35 µm – 3.74 µm) at 0640 and 0820 UTC, 23 September 2013 (click image to enlarge)

The Brightness Temperature Difference product can be used to identify regions of fog/low clouds. Clouds comprised of water droplets have different emissivity properties at short and long infrared wavelengths. That is, clouds do not emit as a blackbody at wavelengths around 3.74 µm; they do emit more like a blackbody at wavelengths near 11.35 µm. Thus, a brightness temperature difference from Suomi/NPP, 11.35 µm – 3.74 µm, will be warm in regions where clouds comprised of water droplets exist. In the example above, note that the brightness temperature difference is much warmer (a maximum of 5.5 K) at 0640 UTC than at 0820 UTC (where the maximum is only 4K). Why is there a difference in the two scenes?

The view at 0640 UTC is along the edge of the Suomi/NPP scan, and therefore the scan traverses a longer stretch of atmosphere, allowing for more signal absorption. Both the 3.74 µm and 11.35 µm I-Bands (the purple lines in the linked-to images) on VIIRS are broad, meaning they sense photons over a relatively large part of the electromagnetic spectrum (compared to the M-bands and to MODIS). Note that the relative response suggests that a longer pathlength through the atmosphere will cause more attenuation for the 11.35 µm channel, meaning a colder temperature. This would likely diminish the difference between the longwave and shortwave IR imagery. The opposite effect is occurring here — the brightness temperature difference is smaller in the 0820 UTC image. Why? The answer can be found in the cirrus shield impinging upon western Lake Superior in the second image. Even though the cirrus is thin, it’s radiative effect is such that the brightness temperature difference decreases.

Loop of Suomi/NPP VIIRS Day/Night Band, VIIRS Brightness Temperature Difference (11.35 µm – 3.74 µm) and GOES-East-based GOES-R IFR Probability from ~0640 and ~0820 UTC on 23 September (click image to enlarge)

How does the GOES-R IFR Probability change in the time between the two Suomi/NPP overpasses? The loop above cycles between the Day/Night Band and the Brightness Temperature Difference (from VIIRS) and the GOES-Based IFR Probability at 0645 and 0815 UTC. Highest IFR Probabilities do exist in regions where VIIRS Day/Night Band imagery and Brightness temperature difference suggest the presence of fog — between Grand Marais and Thunder Bay, and in the river valleys northeast of Lake Superior. (Poor GOES Resolution northeast of Lake Superior hampers a precise identification of where the Valley Fog is). Note that the high values of GOES-IFR probability along the Lake Superior Lakeshore, especially those lakeshores that are oriented north-south, likely stem from the coregistration error between the 3.9 µm and 10.7 µm channels on GOES-13.

GOES-R IFR Probability for three times centered on 0515 UTC 23 September 2013 (click image to enlarge)

The GOES-R IFR Probability shows there is still Stray Light occasionally (in the present case, at 0515 UTC only) that will contaminate the brightness temperature difference signal, and therefore also the GOES-R IFR Probability signal. A longer loop of GOES-R IFR probability, below, shows the slow expansion of IFR Probabilities during the course of the night. It also shows the effect, later in the loop, of cirrus impinging on the western shoreline of Lake Superior. IFR Probabilities drop in regions where Cirrus Clouds preclude the use of satellite data in the determination of IFR Probabilities. In addition, GOES-R Cloud Thickness is not computed in regions where cirrus clouds are present.

GOES-R IFR Probability (Upper Left), GOES-East Brightness Temperature Difference (Upper Right), Suomi/NPP Day/Night Band (Lower Left), GOES-Based GOES-R Cloud Thickness (Lower Right), Times as indicated in imagery (click image to enlarge)

Brightness Temperature Differences over the Rockies

The heritage brightness temperature difference method of detecting fog/low stratus works because clouds that are comprised of liquid water droplets have different emissivity properties at 3.9 µm and at 10.7 µm. Clouds are not black-body emitters at 3.9 µm; they are more closely blackbody emitters at 10.7 µm. Consequently, the 3.9 µm radiance detected by the satellite suggests a cooler emitting blackbody temperature than the 10.7 µm radiance detected by the satellite. The difference between those two temperatures therefore highlights water-based clouds.

Some soils over the western US also have emissivity properties that are a function of wavelength such that the brightness temperature difference product will show a maximum in some regions but not in others. A careless interpretation of the brightness temperature difference signal, then, might lead to an erroneous assumption that fog/low clouds are present in a region of clear skies. The loop above shows the brightness temperature difference field at 0930 UTC on 19 September, and there are many regions with a signal that is consistent with fog/low clouds. The GOES-R IFR Probability algorithm correctly screens out many of these regions because the Rapid Refresh data does not predict low-level saturation. The day-night band image from Suomi/NPP can be used to verify where low clouds are present, and the image shows that most of the western US was clear. The IFR Probability field has false positives in 4 locations: extreme northeastern Arizona, Northwestern Mexico just to the east of the Colorado River, and two patches in the Pacific, one west of northern California, and one west of southern California.

Brightness Temperature Difference signals sometimes show positives over the central and eastern US in cases of extreme drought, as shown here.

GOES-R Fog Product Examples

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