16 April 2017: 1.37 micron band (“Cirrus band”) features other than cirrus clouds

The GOES-16 data posted on this page are preliminary, non-operational data and are undergoing testing.  Users bear all responsibility for inspecting the data prior to use and for the manner in which the data are utilized.

By Dan Bikos, Louie Grasso, and Ed Szoke

For this blog entry, we are going to focus in on the state of Durango in Mexico during the mid-day hours of 16 April 2017.  Conditions during that time were warm and very dry:


The sky cover was mostly clear throughout the period of interest (mid-day hours):


A topographic map of the region reveals that the elevation (given below in thousands of feet) is quite high:



If we analyze the GOES-16 1.38 micron (“Cirrus band”):


There are features that are moving that are approximately oriented southwest-northeast (ignoring the cirrus clouds later in the loop in the northern regions and also the low-level cumulus to the south).  These features are not clouds since we did not see them in the visible channel shown above.

Let’s look at the GOES-16 7.34 micron (“low-level water vapor”) band:


Features similar to those that were observed in the 1.38 micron band appear at 7.34 microns.  The 1.38 micron band can be displayed with a different color table to increase the contrast, thus bringing more clarity to the features that we observe:


Recall at this wavelength, 1.38 microns, water vapor is the primary absorber.  If there is sufficient moisture to absorb incoming radiation, cirrus clouds show up rather clearly due to the large contrast between bright cirrus clouds and a dark background, hence the band being named the “Cirrus band”.  In the case discussed here, moisture is limited, particularly at higher elevations where we see the southwest-northeast oriented banded  features.  In fact, here is a comparison of the corresponding features at 1.38 and 7.34 micron band.


We note that each feature labeled above has the following characteristics:

1) 1.38 micron band darker corresponds to 7.34 micron band cooler brightness temperature and

2) 1.38 micron band lighter corresponds to 7.34 micron band warmer brightness temperature.


In this relatively dry, higher elevation environment, the surface is not completely obscured by the intervening (and highly absorbing) atmospheric water vapor when viewed at 1.38 microns.  In this near-infrared band, regions that are darker correspond to more column-integrated water vapor (and a lower surface reflectance contribution), while regions that are brighter correspond to less column-integrated water vapor (and a  higher surface reflectance contribution).

The alignment of these features most likely associated with water vapor are oriented with the terrain:


Note that the lower valleys at locations 5 and 6 can be seen as darker regions at 1.38 microns (recall, more water vapor is associated with darker regions at 1.38 microns).

Can we rule out that these features are associated with dust or smoke?  This will now be investigated.

The split window difference product (11.2 micron minus 12.3 micron band difference):


would have negative values (brown in this color table) if lofted dust was present, since the values are positive, we can rule out lofted dust.

For assessing smoke, we look at the GOES-16 0.47 micron (“Blue”) visible band:


There are no obvious smoke plumes during this time period.  However, if we look later in the afternoon when fires tend to be more pronounced and show up more clearly due to  favorable scattering associated with the view angle:


We do observe a few smoke plumes.  However, the orientation of the smoke plumes does not match with what we observed in the 1.38 or 7.34 micron bands and does not cover such a large area in bands that are oriented with the terrain.

In conclusion, the GOES-16 1.38 micron (“Cirrus”) band can observe features other than cirrus clouds and plumes of water vapor may be observed under specific circumstances.

Posted in GOES R | Leave a comment

1-minute applications for severe thunderstorms from 15-16 April 2017

The GOES-16 data posted on this page are preliminary, non-operational data and are undergoing testing.  Users bear all responsibility for inspecting the data prior to use and for the manner in which the data are utilized.

During the afternoon hours of 15 April, one of the GOES-16 mesoscale sectors captured severe thunderstorms in the Iowa / Nebraska region:


Often times we like to point out the use of 1-minute imagery for the evolution of severe thunderstorms.  However, the imagery also has utility to identify regions where the potential for thunderstorm development is suppressed which can be an important operational application.  In this particular case, notice the highlighted yellow regions at 2209 UTC:


Focus on those regions in the animation above.  Note that thunderstorm development is  suppressed in these regions.  The 1-minute imagery actually helps you identify these regions with more clarity than current GOES due to the increased temporal and spatial resolution.  We might not always be able to understand why suppression is occurring.  In this case, just considering the surface observations:


for the southern region of interest, although there is cloud cover, the temperature and moisture does not appear unfavorable for thunderstorm development, however the imagery shows that any attempts at convective initiation are being suppressed by something in the environment.  Meanwhile, in the northern region of interest there is nothing apparent in the METARs that would suggest suppression, however GOES-16 imagery clearly indicates that something in the environment is unfavorable for convective development.  In operations, GOES-16 imagery could be integrated with model analyses, surface observations and other observational data to try to understand why the imagery shows what it does.

On the following day, another round of severe thunderstorms took place this time in Oklahoma and the Texas panhandle.  We’ll start with a loop of the IR band at 10.35 microns at 5 minute temporal resolution:


We observe a number of thunderstorms across the scene;  the storm in southern Oklahoma has an enhanced-V signature as well as multiple gravity waves and is backbuilding southwestward for a while.  We see other thunderstorms in the scene as well, however lets focus in on the eastern Texas panhandle activity.  We see a westward moving boundary with low-clouds to its east and clear to the west.  The northern part of this boundary leads to a triple point with an outflow boundary oriented east-west.  It’s interesting to note that convective initiation occurs first south of the triple point and then later new convection develops at the triple point.  The initial storm that developed along the westward moving boundary south of the triple point develops an enhanced-V signature and in fact severe hail was observed with this storm.  Also note that this storm is moving southwestward in time as it backbuilds along the boundary against the mean west-southwest flow aloft shown in the 0000 UTC Amarillo, TX sounding:



Are there are any indications in the GOES-16 imagery to explain why the storm of interest along the boundary developed?

A 4 panel animation of the visible along with the 3 water vapor bands:


The visible band indicates what might be a gravity wave advancing east-southeast across the northeastern portion of the Texas panhandle (also shown in a subtle sense in the earlier IR loop).  The storm seems to initiate coincident with the passage of this feature across the boundary.  We include the 3 water vapor channels in an attempt to further identify this feature or any other feature that would not appear in the visible band.  The most obvious feature (at least in the mid-level WV band) appears later in the loop and is a north-south oriented brightness temperature gradient moving eastward, perhaps this is an approaching shortwave.  However, this comes after convective initiation for the storm noted above.  For this case, it’s difficult to definitely attribute convective initiation to any signature from the water vapor imagery, although a 4 panel like this may help in other cases.

Shortly thereafter, this region was in a mesoscale sector with 1-minute imagery, here is a loop with METARs included:


One point to note is that this was not a classic dryline with very dry westerlies behind it, rather the winds here are easterly and the moisture gradient is more subtle.  We do eventually see a storm initiate along the triple point discussed earlier north of the storm of interest.

Posted in Convection, GOES R, Hail, Severe Weather | Leave a comment

Snow on the ground as depicted in the GOES-16 1.6 micron band

The GOES-16 data posted on this page are preliminary, non-operational data and are undergoing testing.  Users bear all responsibility for inspecting the data prior to use and for the manner in which the data are utilized.

The GOES-16 visible (0.64 micron) loop shows what appears to be pretty straightforward – recent snow cover melting during daytime heating in southern Colorado:


Now let’s look at the GOES-16 1.6 micron loop:


At this wavelength, snow and ice surfaces are strongly absorbing, thus darker in this imagery.  However, all darker regions are not equally dark, see the highlighted region here:


Note the darker, almost black regions within the highlighted yellow ovals.

On the previous day, the highlighted regions did receive rain mixed with the snow following all snow that accumulated earlier in the day.  This likely led to a more slushy type of snow in the darker regions highlighted above.

It turns out, at this wavelength, the size of the ice particles on the surface matters.  Where there was rain on snow, the ice particles are larger relative to regions that received all snow.  As the melting began on 5 April, the snow cover in the darker regions highlighted were relatively wetter / “slushier” compared to other regions that received just snow.

For a much more detailed explanation of this effect in the 1.6 micron band, see this blog by Curtis Seaman:


A potential application of this principle regarding the 1.6 micron band is identification of hail swaths left behind thunderstorms (assuming clear skies).  The hail swath should appear relatively dark in this channel, similar to the wet / slushy snow signature shown above.


Posted in Convection, GOES R, Hail | Leave a comment

Low-level moisture as observed from the GOES-16 shortwave difference product

The GOES-16 data posted on this page are preliminary, non-operational data and are undergoing testing.  Users bear all responsibility for inspecting the data prior to use and for the manner in which the data are utilized.

The split window difference product is the difference between GOES-16 bands 13 and 15, that is 10.35 minus 12.3 microns.  The high spatial resolution of GOES-16 allows for detection of small-scale gradients in water vapor.  Lindsey et al. (2014) demonstrates the utility in this product for identification of local deepening of low-level moisture in a cloud-free environment.  The following schematic illustrates the concept that larger positive values of SWD correspond to deepening low-level moisture (in the absence of clouds):



In this example with simulated data, a cross section is taken across a dryline in Texas in clear sky conditions.  The cross section on the right of specific humidity (shaded) illustrates the local deepening (i.e., moisture pooling) along the dryline.  Split window difference values are maximized in the vicinity of the localized deeper moisture.

For an example, lets look at the GOES-16 visible (0.64 micron) animation:


Note the line of cumulus that develops oriented approximately east-west from Denver eastwards to just south of the Kansas / Nebraska border.  Before the line of cumulus develops, skies were clear.  Now let’s consider the split window difference product during the same time period:


In this color table, positive brightness temperature difference values are green, followed by yellow then red for increasing magnitudes.  The low-level convergence boundary where the cumulus later develops is now readily identified.  Also, the magnitude of the brightness temperature difference increases (becomes more positive, towards the red values) prior to the development of the cumulus.  This is the signal of deepening moisture along the convergence boundary.  Supporting evidence from METARs at 1800 confirm the presence of the low-level convergence boundary with higher dewpoints north of the east-west segment of the boundary:


This product can be utilized to identify regions (in clear sky conditions) where localized moisture deepening occurs prior to the development of convective clouds (and potential convective initiation).  One caveat to this signal is that we tend to see this where moisture is relatively shallow, not in regions of deep low-level moisture.  More research is needed to understand the limits of how deep the moisture must be to not see this signal in the split window difference product.

Note that in this example, the split window difference product is band 14 (11.2 micron) minus band 15 (12.3 micron).  It should now be band 13 (10.35 microns) minus band 15 (12.3 microns), this is a small change that would not affect your interpretation of the imagery but it should make the signal a bit stronger (easier to identify).

Another example may be seen in Himawari over Bangladesh / eastern India, where the dryline commonly develops in April.  First, here is the visible loop:


Initially, skies are clear which is an important prerequisite to seeing this signal in the split window difference product.  Later, thunderstorms develop.

The split window difference product:


Shows increasing values of brightness temperature difference (orange/red in this color table) in clear skies before cumulus develops, followed by thunderstorms.  Again, the signal is increasing depth of moisture along the convergence line (slowly eastward moving dryline in this case).  Moisture depth in the pre-storm environment can be assessed with the nearest sounding from Calcutta:



like the Colorado case, the moisture depth is shallow.  This seems to be one of the caveats for being able to identify this signal in the split window difference product.


Lindsey, D.T., L. Grasso, J.F. Dostalek, and J. Kerkmann, 2014: Use of the GOES-R Split-Window Difference to Diagnose Deepening Low-Level Water Vapor. J. Appl. Meteor. Climatol., 53, 2005–2016, doi: 10.1175/JAMC-D-14-0010.1.

Posted in Convection, GOES R, Severe Weather | Leave a comment