Monthly Archives: September 2014

Fog over the Piedmont of Virginia

PiedmongFog_30Sep2014_SNPPBTD

GOES-R IFR Probabilities (Upper Left), GOES-R Cloud Thickness (Lower Left), Brightness Temperature Differences from GOES-13 (Top Right, 10.8 µm – 3.9 µm) and from Suomi-NPP (Bottom Right, 11.35 µm -3.74 µm , all at 0715 UTC 30 September (Click to enlarge)

Radiation fog developed over the Piedmont of Virgina on the morning of 30 September in a region with little pressure gradient. The image above, from 0715 UTC, shows IFR Probability to be very high (from 85-95%) over the Virginia Piedmont. Numerous stations in the region were reporting IFR visibilities and ceilings. Brightness Temperature Difference products are also highlighting the region (Note the advantage in the Suomi NPP Brightness Temperature Difference field that ably captures fine detail related to Ridge/Valley topography over the Appalachians); the advantage of IFR Probability fields is that it includes surface information so that someone using the field can be more certain that the stratus detected by the satellite is in contact with the ground as fog. The animation below shows the evolution of the fields from 0200 through 0700 UTC.

PiedmontFog_30Sept2014_03-07

As above, but with Suomi NPP Day/Night band in lower Right. Hourly imagery from 0200 through 0700 UTC on 30 September 2014 (Click to enlarge)

GOES-R Cloud Thickness (the thickness of the highest water-based cloud) can be used to estimate when fog/low clouds will burn off. The estimate is most accurate for strict radiation fog. GOES-R Cloud Thickness is only computed for water-based clouds during non-twilight times (in other words, it is not computed in the hours surrounding twilight at both sunrise and sunset). The last value of Cloud Thickness before morning twilight (shown below) can be used in concert with this scatterplot to guess when clouds might dissipate. Values over eastern Virginia, near Richmond, exceeding 1100 feet, correspond to a dissipation time 4+ hours after 1100 UTC, or at 1500 UTC. In this case, that value was an underestimate.

PiedmontFog_30Sept2014-43

As above, but for 1100 UTC 30 September 2014

GOES13_VIS_30SEPT2014_12-17

GOES-13 Visible imagery, hourly between 1215 and 1715 UTC 30 September 2014 (Click to enlarge)

Comparing IFR Probabilities and Brightness Temperature Differences over Wisconsin

GOES_MODIS_IFR_PROB_26Sep2014_03-11

GOES-R IFR Probabilities computed from GOES-East (Upper Left), GOES-13 Brightness Temperature Difference (10.7 µm – 3.9 µm) (Upper Right), GOES-R Cloud Thickness computed from GOES-East (Lower Left), GOES-R IFR Probabilities computed from MODIS (Lower Right), hourly from 0300 through 1200 UTC 26 September (Click to enlarge)

Fog developed over southeast Wisconsin during the morning of 26 September 2014. The GOES-R IFR Probability fields did a better job of detecting the hazard than did the traditional method of satellite-based fog detection (brightness temperature difference) in part because of the presence of higher clouds. (Fog/low stratus also generated IFR conditions over northeast Ohio). The animation above also includes a MODIS-based IFR Probability field. In fact, three separate MODIS-based fields were created, one at 0430 UTC, one at 0700 UTC, and one at 0845 UTC, shown below (Note that the MODIS-based IFR toggles with the MODIS Brightness Temperature Difference field). MODIS data confirms the small-scale nature of the fog event over southeast Wisconsin. (Note also how the MODIS-based data at 0700 UTC, below, are able to resolve the small river valleys in Pennsylvania in ways the GOES data cannot.)

Note how the IFR Probability fields ignore regions of mid-level stratus, such as over northeast Ohio along Lake Erie.

MODIS_BTDIFR_26Sept2014_0430

As above, but for 0430 UTC, and with the MODIS Brightness Temperature Difference in the lower right (Click to enlarge)

MODIS_BTDIFR_26Sept2014_0700

As above, but for 0700 UTC, and with the MODIS Brightness Temperature Difference in the lower right (Click to enlarge)

MODIS_BTDIFR_26Sept2014_0845

As above, but for 0845 UTC, and with the MODIS Brightness Temperature Difference in the lower right (Click to enlarge)

Suomi NPP Overflew the region as well. GOES-R IFR Probability algorithms do not yet incorporate Suomi NPP data; when that happens, an early morning snapshot that complements MODIS overpasses will be available.

SNPP_BTD_26Sept2014loop

As up top, but with Suomi NPP Brightness Temperature Difference (11.35 µm- 3.74 µm) in the bottom right, times at 0647 and 0831 UTC 26 September (Click to enlarge)

Cloud Thickness as a Predictor of Fog Dissipation, part II

This post showed examples of Cloud Thickness and how its use as a predictor for dissipation time might be incorrect because of synoptic or mesoscale forcing.

Radiation fog formed in River Valleys of Wisconsin early in the morning on 22 September, and the image below shows the final Cloud Thickness field computed before twilight conditions developed over western Wisconsin (twilight conditions are already occurring over eastern Wisconsin).

CloudThickness_1145UTC_22Sep2014

GOES-R Cloud Thickness (of the highest liquid water cloud layer) just before sunrise, 22 September 2014

Cloud Thickness values near LaCrosse, WI, are around 900 feet; values are closer to 1200 feet over northeast Wisconsin along the St. Croix River. The chart suggests a dissipation time over the southwest part of Wisconsin of around 3 hours, and more than 4 hours over northwestern Wisconsin. The animation below shows that those estimates were accurate.

GOES13_VIS_22Sept_1215-1515

GOES-13 Visible (0.63 µm) Animation over Wisconsin, 22 September, 1215-1515 UTC (Click to animate)

Cloud Thickness as a Predictor of Fog Dissipation

GOES_R_Cloud_Thickness_1137UTC_18Sep2014

GOES-R Cloud Thickness over Wisconsin and surrounding States, 18 September 2014, just before sunrise (Click to enlarge)

GOES-R Cloud Thickness can be used as a predictor for dissipation time of Radiation Fog, using this chart and the thickness (as above) from the last pre-dawn GOES-R Cloud Thickness field (Recall that GOES-R Cloud Thickness is not computed in the few hours of twilight surrounding sunrise or sunset; in the image above, twilight has reached lower Michigan but not yet Wisconsin). However, it’s important to remember that the chart is valid for radiation fog. Other forcings might cause fog to dissipate (or persist).

In the example above, Cloud Thickness values ranges from around 700 over southwest Wisconsin to as much as 1400 over north-central Wisconsin. Most of south-central Wisconsin (cyan) has values around 1200. According to the best-fit line, that suggests a burn-off time of more than 5 hours (although those values are extrapolated; note that no values that large went into the creation of the best-fit line) over WI, except over southwestern WI where a burn-off time of less than 1 hour is predicted. Did that work out?

The animation below shows fog/low stratus moving towards the southwest with time. The cool and damp northeasterly flow from the Great Lakes into Wisconsin (surface map at 1800 UTC on 18 September) suppressed the heating necessary to reduce the relative humidity and foster fog evaporation. Perhaps the fog initially formed as advection fog; however, the northeasterly flow that developed early in the morning on 18 September came from a synoptic set-up that allowed fog to persist longer than the GOES-R Cloud Thickness algorithm suggests. This is not an uncommon occurrence. Clouds did not burn off over south-central WI until after 1800 UTC. During September, delayed burn-off of morning clouds can significantly affect the day-time temperature.

WIFOG_18Sep2014_12-20loop

Half-hourly visible imagery over Wisconsin, 1215-2045 UTC on 18 September (Click to animate)

=============================================================================

Low clouds and fog redeveloped during the morning of the 19th of September as well. This occurred during persistent southerly flow in advance of a low pressure system over the Northern Plains. The hourly animation of IFR Probabilities, below, shows IFR Probabilities developing over the course of the early morning of the 19th between 0315 and 1215 UTC. The animation shows a gradual overspreading of the IFR Probability field with higher clouds moving in from the west. (Here is a toggle between IFR Probability and GOES-13 Brightness Temperature Difference Fields at 1115 UTC; note how smooth the field is over much of WI where only Rapid Refresh model data can be used in the computation of the IFR Probability).

GOESR_IFRPROB_WI_19Sep2014_03-12

GOES-R IFR Probability fields, hourly from 0315-1215 UTC on 19 September (Click to animate)

When high clouds overspread the scene, GOES-R Cloud Thickness is not computed. Thus, the last image before twilight, below, shows Cloud Thickness in only a few locations, but those values over southeast Wisconsin exceed 1200 feet, suggesting a burn-off of around 1615 UTC — 5 hours after this last Cloud Thickness image. In this case, that is an overestimate because the southerly winds over WI promote mixing, and the fog quickly dissipates after sunrise. It’s important to consider the synoptic forcing when you use Cloud Thickness. The last Cloud Thickness field and its use as a predictor for fog dissipation (using this chart) is most useful for radiation fog. The visible imagery animation at the bottom shows that the fog dissipated by 1415 UTC.

GOESR_CTH_WI_19Sept2014-9

GOES-R Cloud Thickness just before Sunrise (1115 UTC on 19 September 2015) (Click to enlarge)

WIFOG_19Sep2014_12-16loop

GOES-13 Visible Imagery, 1215-1615 UTC on 19 September (Click to animate)

Fog over South Dakota

The NWS Office in Sioux Falls tweeted a picture of fog near Kimball, SD, in Brule County (below).

Kimball_SD_FogImage_17Sept2014

Webcam image of Fog along I-90 in central South Dakota near Kimball, after sunrise on 17 September 2014

The toggle below shows the Brightness Temperature Difference Field from GOES-13 (10.7 µm – 3.9 µm) and the GOES-R IFR Probability fields computed using data from GOES-13 and the Rapid Refresh Model. The Brightness Temperature Difference field detects the presence of water-based clouds (yellow and orange in the enhancement used) and works because such clouds have difference emissivity properties at 3.9 µm and 10.7 µm. Temperatures inferred from the 3.9 µm radiation detected are cooler than those temperatures inferred from the 10.7 µm radiation because water-based clouds do not emit 3.9 µm radiation as blackbodies. The Brightness Temperature Difference field gives information about the top of the cloud only, however, and it typically overestimates regions of fog/low stratus. That is the case on the morning of 17 September 2014. For example, IFR Conditions are not reported over much of southeastern South Dakota or western Minnesota along Interstate 90. The IFR Probability algorithm is correctly minimizing the influence of the strong brightness temperature difference signal there because Rapid Refresh Data is not showing boundary-layer saturation. The highest IFR Probabilities in South Dakota are associated with reported IFR Conditions, for example at Chamberlain, SD (west of Kimball SD and also in Brule County) and at Aberdeen in northeastern South Dakota.

GOES13_BTD_IFR_1215UTC_17Sept2014

Stratus and Fog over the northeast and mid-Atlantic States

GOES13_BTD_15Sep2014_02-11UTC

GOES-13 Color-Enhanced Brightness Tempreature Difference Fields (10.7 µm – 3.9 µm), hourly from 0200 through 1100 UTC, 15 September 2014 (Click to enlarge)

Brightness Temperature Difference Fields from GOES-13 show large regions over Pennsylvania and surrounding states during the early morning hours of September 15th. (Note that the image at 0515 UTC, not in the loop above, shows the effect of stray light). If you look at the ceilings and visibilities in the imagery above, you will note that many regions where stratus/fog are indicated by the brightness temperature difference field (over upstate NY, for example), do not in fact show anything near IFR conditions. Always recall that the satellite is seeing the top of the cloud deck; whether or not that cloud extends to the surface is beyond the capability of present satellite systems. (You can infer it sometimes, of course, especially if the signal is confined to a narrow river valley, as occurs in the animation above: The Ohio River along the northern panhandle of West Virginia shows up very well).

GOES_IFR_15Sept2014_02-11UTC

GOES-13 IFR Probability Fields, hourly from 0200 through 1100 UTC, 15 September 2014 (Click to enlarge)

IFR Probability fields for the same time do a better job of highlighting only where reduced ceilings and visibilities are present. For example, the region of stratus over upstate New York is screened out, as well as the region over southern and southeastern Virginia. Probabilities are also quite high over the Ohio River Valley, where river fog is likely occurring. Note that IFR Probabilities over southwestern Indiana at the end of the animation have the characteristic look (a flat field) associated with IFR Probabilities created without the benefit of satellite data.

MODIS data were able to provide a a high-resolution image of this scene in the middle of the night. As with GOES, MODIS identified the large region of stratus over upstate New York and over southeastern Virginia, and the IFR Probabilities correctly screened out those stratus clouds. River Valleys show up distinctly along the Ohio River downriver from Pennsylvania; smaller IFR Probabilities surround the rivers. Sometimes MODIS data can give an early alert to the development of fog; in the present case, when MODIS overflew the region, fog development was at sufficiently large a scale that GOES-13 could also detect it.

MODIS_IFR-BTD_0722UTC_15Sep2014

MODIS Color-Enhanced Brightness Tempreature Difference Fields (11 µm – 3.9 µm) and IFR Probability Fields at 0722 UTC 15 September 2014 (Click to enlarge)

Suomi NPP data also viewed the developing river fogs, both in the day/night band, and in the brightness temperature difference (11.35 µm – 3.74 µm), below. At present, IFR Probabilities are not computed from Suomi NPP satellite data.

SNPP_BTD_DNB_0653UTC_15Sep2014

Suomi NPP Color-Enhanced Brightness Tempreature Difference Fields (11.35 µm – 3.74 µm) and Day Night band visible imagery at 0653 UTC 15 September 2014 (Click to enlarge)

IFR Probability when an extratropical storm passes by

The approach of an extratropical cyclone, such as the Colorado Cyclone in this animation in the upper Midwest on 10 September 2014, will frequently result in areas of IFR or near-IFR conditions. However, the many cloud layers that accompany these baroclinic disturbances will always make difficult the task of identifying (using satellite imagery) regions of low stratus and fog. Consider the animation below of Brightness Temperature Difference (10.7µm – 3.9µm) fields (a traditional method of detecting water-based clouds) over the Upper Midwest on 10 September.

BTD_10Sep2014_02-14UTC

Color-enhanced Brightness Temperature Difference fields (10.7µm – 3.9µm), hourly from 0100 UTC to 1400 UTC on 10 September 2014 (Click to enlarge)

Interpretation of this loop is time-consuming. Not only is there little distinct signal related to observed IFR and near-IFR conditions, but the rising sun (at the end of the animation) causes the Brightness Temperature Difference to flip sign, altering the enhancement. There are regions where the Brightness Temperature Difference field detects water-based clouds that may be associated with fog or stratus, chiefly over the western third of the domain (and especially over the Dakotas) in the later half of the animation.

Compare the animation above to the loop of IFR Probabilities for the same time period below. IFR Probabilities are highest where near-IFR or IFR conditions are present, and the IFR Probability field screens out regions where low stratus (but not fog) is present, such as over the Dakotas at the end of the animation. Regions where IFR Probability fields have a flat character — such as over Wisconsin around sunrise — are where only Rapid Refresh model data (but not satellite data) are used as predictors, and the field does not have pixel-scale variability. Because fewer predictors are used, the magnitude of the IFR Probability is smaller than in regions where both satellite and model data can be used as Predictors. Thus, a flat field (over eastern Wisconsin at the end of the animation, or over Iowa at the beginning) has values that should be interpreted differently from similar values in regions where both satellite and model data can be used in the computation of IFR Probabilities.

IFRPROB_10Sep2014_01-14UTC

GOES-R IFR Probability fields (Computed from GOES-13 and Rapid Refresh Data), hourly from 0100 UTC to 1400 UTC on 10 September 2014 (Click to enlarge)

Interpreting IFR Probability Fields

IFR_PROB_11-3.9_Sat_20140908_0700

Toggle between GOES-R IFR Probabilities and GOES-East Brightness Temperature Differences (10.7 µm – 3.9 µm) over the southeast US, 0700 UTC 8 September 2014 (Click to enlarge)

When multiple cloud layers are present, such as when a cirrus shield overlays a region, the traditional method of detecting fog/low stratus, the brightness temperature difference product, struggles to identify regions of low clouds because the satellite sees only the signature of the high clouds. Low clouds are hidden from view. In such cases, it is vital to incorporate low-level information to identify regions where fog/low stratus might be present. The required low-level information can come from the model fields of the Rapid Refresh Model. The model predictors can be used to generate IFR Probabilities in regions where satellite predictors are unavailable because of the presence of high clouds.

In the toggle above, the Brightness Temperature Difference field shows high clouds over Georgia and the Carolinas. IFR Probabilities in this region are around 50% — relatively low because Cloud Predictors cannot be used in the algorithm. But IFR Conditions are present over North and South Carolina. Tailor your interpretation of the IFR Probability values to account for which predictors are used.

Over Tennessee, IFR Probabilities are much higher. In this region, satellite predictors can be used, and a strong satellite signal is present. IFR Conditions are not widespread, however. Use the IFR Probability field as one tool (but not the only tool) when making nowcasts about the possibilities of fog/low stratus.

MODIS-based and GOES-based IFR Probabilities over the High Plains

GOES_IFR_0430_03Sep2014

GOES-based GOES-R IFR Probabilities over Kansas and surrounding states, 0430 UTC 3 September 2014 (Click to enlarge)

GOES-based IFR Probabilities over Kansas before midnight on 2 September highlight two regions where IFR Conditions might be developing: over western Kansas, near the Colorado border, and over south-central Kansas. These would be two places to monitor most closely over the coming hours. The MODIS-based IFR Probabilities for the same time, below, can be used to refine the interpretation of the GOES fields. IFR probabilities over western Kansas are higher with the MODIS data. IFR Probabilities from MODIS better capture the difference in the field over south-central Kansas as well: there is a more obvious distinction between IFR Probabilities influenced solely by model output (because of the multiple cloud layers associated with the thunderstorm at Hutchinson and Newton) and those controlled by both model and satellite predictors. The strength of GOES-based IFR Probabilities is temporal continuity. How do the fields evolve with time?

MODIS_IFR_0424UTC_03Sep2014

MODIS-based GOES-R IFR Probabilities over Kansas and surrounding states, 0424 UTC 3 September 2014 (Click to enlarge)

The animation below of GOES-based IFR Probabilities shows increasing values over western Kansas (the region drifts northward, as well); by 1045 UTC, at the end of the animation, IFR Probabilities are very high over western and northwestern Kansas, and IFR conditions are observed in the form of both low ceilings and reduced visibilities. This was a case where MODIS data gave an early alert to where GOES-based IFR probabilities might later become high. Fog can start at small scales and then grow in size and MODIS data offers an advantage of higher spatial resolution. A toggle between the MODIS and GOES-based IFR Probabilities at 0836 UTC is at bottom.

GOES_IFR_0500-1045UTC_03Sep2014

GOES-based GOES-R IFR Probabilities over Kansas and surrounding states, times as indicated on 3 September 2014 (Click to enlarge)

MODIS_GOES_IFR_0837UTC_3Sep2014

MODIS- and GOES-based GOES-R IFR Probabilities over Kansas and surrounding states, 0836 UTC on 3 September 2014 (Click to enlarge)

IFR Probabilities over the Texas Panhandle

GOES_MODIS_IFR_SNPP_BTD-DNB_0930UTC_02Sep2014

GOES-R IFR Probabilities (Upper Right), GOES-13 Brightness Temperature Difference (10.7 µm – 3.9 µm) (Upper Left), MODIS IFR Probabilities (Lower Left), Suomi NPP Brightness Temperature Difference (11.35 µm – 3.74 µm) and Day Night Band (Lower Right), all near 0930 UTC 2 September 2014 (Click to enlarge)

GOES-R IFR Probabilities (from GOES and from MODIS) over the Great Plains and southern Rockies indicated one region where IFR conditions were most likely: over the Texas panhandle, where IFR conditions were reported. There is a strong signal in the GOES-based Brightness Temperature Difference field there (and in the Suomi NPP Brightness Temperature Difference field) as well. There is also a Brightness Temperature difference signal in regions where IFR conditions are not occurring; in those locations, stratus is present, or (over the Rockies) emissivity differences in the dry soil are present, both of which conditions will lead to a signal in the brightness temperature difference that is unrelated to surface visibility and ceilings. This is therefore another example showing how incorporation of model data that accurately describes saturation (or near-saturation) in the lowest model layers can help the GOES-R IFR Probability more accurately depict where IFR conditions are present.