VIrtual Institute for Satellite Integration Training

Training Sessions

The VISIT Program

VISIT People


Links / Tutorials


Integrated Sensor Training

Cooperative Institute for Research in the Atmosphere

Cooperative Institute for Meteorological Satellite Studies

Operational Support FacilityOperations Training Branch

National Severe Storms Laboratory

Last Updated by:
Scott Bachmeier

Instructor Notes for the "Enhanced-V" VISITview Training Lesson

These are the Instructor Notes for the "Enhanced-V" VISITview lesson, for Lesson Version ev_12sep00.exe (12.September.2000). The latest version of this lesson can be found at

or at

Instructions for installing this VISITview lesson on a Windows computer can be found at

To run this version locally (for on-site training), use the batch file visitlocal.bat, which is located in the same directory as the rest of the Enhanced-V files. This is a single-site method, so nothing you do (page changes, drawing, text annotation, etc) will be seen by any other sites running a VISITview lesson. Alternately, you can view this VISITview lesson via the Web, using a modern version of a Web browser that supports Java (Netscape Communicator or Microsoft Internet Explorer, versions 4.x and higher). The URL is:

The Web version of the lesson has the unique Group name of Web-Enhanced-V, so anything you do there will not have an impact on any "live" VISITview training sessions (which have the Group Name Enhanced-V-12sep00).

For Help on VISITview commands and functionality, use the keyboard shortcut ALT-?. Questions about these instructor notes can be addressed to


* Introduce yourself (and any other staff who might be assisting the training effort)

* Introduce VISIT and the IST PDS training concept


* Stress the 2 main goals of this lesson


* 2 main theories were advanced during the 1980's: Dynamics ("flow around a barrier" model) and Microphysics (variation in the horizontal distribution of Ice Water Content (IWC) within the anvil layer

* Lower right panel: this is a GOES 10.7 micrometer IR image showing a classic Enhanced-V signature; note how the cold arms of the "V" (-55 to -65 C, enhanced orange to red) spread out to the east and northeast; between the cold arms of the "V" is the so-called "warm wake" region (-45 to -55 C, enhanced green to yellow)

* Lower left panel: some studies suggest that the coldest cloud top temperatures (near the "vertex" of the "V") is often at or near the storm summit, and the "warm wake" (the interior portion of the "V") might be due to adiabatic compression via subsidence of stratospheric environmental flow just downwind of the summit (some investigators like Fujuta [1974] found a distinct "cloud trench" near the summit, which supports such a subsiding flow pattern)

* Click on the "Toggle" button to show a slide from the IST PDS Web-based training lesson Three Classes of Storm Top Signatures in Infrared Satellite Data (alternatively, you can click on the "Show URL" botton to launch a Web browser that will open this Web page). This slide shows a schematic based on the work of Adler and Mack (1986) -- in this particular "class" of storms, the coldest point on the trajectory path is also the highest point. This is the type of storm we will encounter when we examine the outbreak of tornadoes across Minnesota in this lesson.

* Click on the "Toggle" button to return to the original slide

* Upper right panel: other studies suggest that the cold "V" pattern and the associated "warm wake" result more from the horizontal redistribution of ice crystal sizes and concentrations; IWC is highest in the cold "arms" of the "V", while IWC is lower in the middle of the "V"; due to emissivity differences, the satellite radiometer can "see through" thinner portions of the anvil where IWC is lower, with the resulting pixel brightness temperature being warmer)

* Dynamically speaking, it seems to be important to have fairly strong STORM-RELATIVE ANVIL-LEVEL flow in the region near the storm top in order to have a storm exhibit an Enhanced-V signature

* The Enhanced-V signature seems to be assoctiated with strong divergence in the storm top outflow region; this implies that the storm updraft is well-organized and quasi-steady-state.

* The Enhanced-V often develops shortly after tropopause penetration [Heymsfield et al., 1983]


* This is a loop of VIS-5D output showing air parcel trajectories from the University of Wisconsin Non-hydrostatic Modeling System (UW-NMS) model, for a 1981 CCOPE supercell storm in eastern Montana; this a side view of the storm, looking North -- the orange area is the updraft core (greater than 5 m s-1), and the blue surface is a theta-e surface just above the anvil region (at 13.85 km) that is perturbed by the updraft. The trajectories originate as low-level inflow from the northeast, then ascend within the updraft core before spreading out within the anvil layer region.


* This is another view of the same model trajectories shown on the previous page; this perspctive is looking upward from the ground, beneath the updraft region. It can be seen that the updraft trajectories fan outward in a broad "V" pattern once they reach the anvil region aloft. This divergent flow within the anvil region affects the distribution of ice crystal content as discussed earlier.


* This study (published in Monthly Weather Review, 1983, and also in the Federal Meteorological Handbook #6) was based upon a 4-month period (April through July) in 1979

* The Enhanced-V signature *must* be associated with convection that is *growing* (cloud top temperatures are getting *colder*, and/or areal coverage of cold anvil is *expanding*)

* This signature is often associated with strong convection that will soon produce damaging winds, large hail, or tornadoes; the mean lead time (Enhanced-V signature to reports of severe weather on the ground) was about a half hour

* Mean "V" persistence was about an hour; mention one of McCann's "EV Rules": once a storm exhibits a "V", it should be considered "severe" as long as the storm continues to grow (even if it subsequently loses the "V" signature as it is growing)

* On the low Probability of Detection (POD) bullet, stress that this study utilized lower resolution imagery (7-km IR resolution, at 30-minute time intervals) than is currently available from the latest generation of GOES-I/M satellites (4-km IR resolution, 15-minute time itervals); also, the grayscale "MB" enhancement curve can be improved upon by the use of color enhancements that help to highlight the subtle features and variations of Enhanced-V signatures.

* Low False Alarm Ratio (FAR) - 0.04 to 0.19 in the McCann study


* One way to illustrate how an improvement in satellite technology can lead to better Enhanced-V detection is this 2-panel image comparison:

* Top panel: GOES-9 10.7 micrometer IR (4-kilometer resolution) from the Jarrell TX F5 tornado event

* Bottom panel: NOAA-14 10.8 micrometer IR (1-kilometer resolution); note that the improved spatial resolution leads to better detection of the cold and warm portions of the Enhanced-V signature.

* Note the finer detail in cloud top temperature structure on the AVHRR image: coldest pixels are near -80 C (dark blue enhancement); these coldest pixels are 10-20 C colder than the ambient anvil temperatures (-60 to -70 C, cyan to green enhancement); the lack of very cold (-70 to -80 C) brightness temperatures on the 4-km GOES IR due is to field-of-view effects (typical overshooting cloud turrets are around 5 km in diameter, which is at or just below the effective detector IFOV at mid-latitudes); satellite brightness temperatures may be 10-15 C warmer than actual cloud parcel temperatures [Alder et al., JCAM, 1983]

* This also serves as a reminder to utilize satellite data from the NOAA polar orbiters when available, to supplement GOES data

...OK, now that we've discussed some Enhanced-V background, let's begin to examine the 29 March 1998 Minnesota tornado outbreak...

PAGE 8 -- GOES-8 10.7 IR LOOP

* This is a loop of GOES-8 10.7 micrometer IR at 15-minute intervals during the afternoon hours of the Minnesota tornado outbreak (covering about a 5-hour period from 19:02 to 23:45 UTC).

* This particular enhancement uses red to black to denote the coldest cloud top temperatures (-60 to -70 C)

* At this point, mention the various VISITview looping and drawing controls, if the student(s) are not familiar with VISITview

* ask student(s) to identify (by using the Pointer, or by drawing) any convective storms which exhibit an Enhanced-V signature

* 3 main clusters of convection: 1 in southwestern MN, and 2 in eastern MN

* Eastern storms were hailers (2.75 inch in WI, 4.50 inch in MN), while the line in southwestern MN produced large hail and several tornadoes; why the difference in the nature of the convection? instability? shear?

...Let's examine the synoptic situation of 29 March to help explain the different nature of the convective storms which affected parts of the upper Mississippi Valley region...


* Fairly self-explanitory...


* GOES-8 visible imagery at 15-minute intervals, late morning to mid-afternoon hours (from 16:02 to 20:45 UTC); hourly surface reports are overlaid on the imagery

* Ask student(s) to draw any significant boundaries or features that they see (suggest stopping loop at last image, 20:45 UTC)

* Features should include: Low near KFSD; cold front from near KFSD southwestward into central Nebraska / primary warm front stretching eastward cross southern Minnesota; dry line across eastern Nebraska; a subtle W-E trough/windshift line (secondary warm front?) across extreme northern Iowa

* Things to point out: Warm, moist air mass (81/64 F in central Iowa, 75/64 F in NW Iowa); Dew points increase from low-mid 50's to low-mid 60's F across southern Minnesota; Very atypical air mass for this region at the end of March; Persistent SE winds in southern Minnesota, which help to enhance surface moisture convergence near the main warm front

* As an aside, you can mention the cold air in northern South Dakota and northern Minnesota; 2 days after this event, 6-12 inches of snow fell across parts of Minnesota that were affected by this tornado outbreak!


* GOES-8 6.7 micrometer IR ("water vapor) imagery at 15 minute intervals (from 16:45 to 20:45 UTC)

* Ask student(s) to point out any significant features they see on the loop of water vapor imagery

* Note the warming/drying associated with the approach of an upper level jet streak from NW Kansas into central and eastern Nebraska; also note the subtle appearance of thin cirrus filaments, which are often evident near the axis of such upper level jet streaks


* GOES-8 6.7 micrometer IR ("water vapor") at 17:45 UTC; 6-hour forecast of Eta 250 hPa wind streamlines and isotachs overlaid

* Draw jet streak axis within closed 55 m s-1 speed contour

* Click Forward VCR Button ( > ) to select the same water vapor image with Eta 250 hPa wind divergence overlaid; note the +4 x 10^5 s-1 contour of divergence over Nebraska/South Dakota/Iowa

* the approach of this upper level jet streak likely enhanced synoptic-scale upward vertical velocity and deep layer shear across southern Minnesota (where we noted an important surface boundary -- the primary warm front)


* This is a loop of hourly GOES-8 Sounder Lifted Index stability (from 11:45 to 20:45 UTC)

* negative LI's (instability) denoted by blue/yellow/red enhancement

* Note the trend of rapid destabilization of the air mass across eastern Nebrasks and western Iowa after about 14 UTC (LI's of -6 to -8 C); the *trends* in LI are more important that actual values with such a sounder derived product, but destabilization is indeed very evident across the region.

* 850 mb wind streamlines from Eta model overlaid at 12 and 18 UTC; these streamlines are representaive of the mean boundary layer flow that day, and indicate that this unstable air mass was being advected northward into southern Minnesota


* This is a loop of hourly GOES-8 Sounder CAPE (from 11:45 to 20:45 UTC)

* As we noted on the loop of sounder LI, we see the trend of rapid destabilization of the air mass over parts of Nebraska and Iowa after about 14 UTC (CAPE values of 2000-3000 J kg-1, enhanced yellow to red)


* This is a loop of hourly GOES-8 Sounder Total Precipitable Water (from 11:45 to 20:45 UTC)

* Note ample moisture (PWs in the upper 20's to around 30 mm / 1.00 to 1.25 in, dark blue to yellow enhancement) across eastern Nebraska and all of Iowa; these PW values represent 200-265% of normal for this region and time period; again, this moist air mass was being advected northward into southern Minnesota (as indicated by the 850 mb wind streamlines, which were representative of the mean boundary layer flow)

* Also note the dry line across eastern Nebraska; the gradient of PW values appears to increase along the dry line (darker blue enhancement) after about 19:00 UTC.

...So, on the synoptic scale, one common denominator among the various convective systems developing during the afternoon hours across Missesota and Wisconsin was the advection of a moist and unstable air mass into the region where an important surface boundary existed; Now let's examine the role of vertical wind shear across the region...


* 2 profiler sites, conveniently located across our region of interest

* Wood Lake (WDL) in southwestern Minnesota, and Blue River (BLR) in southwestern Wisconsin

* Recall (perhaps by roughly sketching) the position of the surface warm front across southern Minnesota (the front lied roughly between these 2 profiler sites, with WDL to the north of the front and BLR to the south)


* Note that time increases to the Left on this time series, which is from 13 UTC on 29 MArch to 03 UTC on 30 March (there is a 6-hour gap between the final two vertical wind profiles)

* At Wood Lake in southwestern Minnesota, note the pronounced veering with height of the winds within the boundary layer (the lowest 2 km)

* Also note the increase of wind speeds in the upper troposphere after 19 UTC, which signals the approach of the upper level jet streak that we saw on the loop of GOES water vapor imagery (Circle the region of higher wind speeds around the 10 km level)

* Click the Forward VCR Button ( > ) to select the Blue River Wisconsin profiler data; note that the low-level veering with height is not as pronounced, and the deep layer shear is not enhanced by stronger winds aloft over Wisconsin (the upper level jet streak was still well off to the west, over Minnesota)

* Toggle between the two profiler plots to show the differences as you discuss these points...


* At 12 UTC (yellow plot): low-level veering with height; strong SW flow aloft; Equilibrium Level (EL, near -50 to -55 C) just below the Tropopause (near -60 C); Keep these temperatures in mind for when we examine the Enhanced-V signature on IR satellite imagery

* Click on Forward VCR Button ( > ) to select the special 18 UTC Omaha raob; note the classic "Loaded Gun" profile, and the approach of stronger mid-level winds (80 knots at 500 mb).


* Bullets are fairly self-explanitory...

* Given that the different convective clusters across western and eastern Minnesota were tapping a similarly moist/unstable air mass along the low-level focus of the surface warm front, SHEAR was likely the most important factor influencing the difference in storm character (hailers in the east vs. tornadic supercells in the west)

...Now let's take a closer look at the tornadic supercells across southwestern Minnesota, using satellite and radar...


* Click to place the large red Cursor in southwestern Minnesota right away, to align the main IR image with the smaller image portal in the upper left corner

* GOES-8 10.7 micrometer IR at 15-minute intervals (from 20:15 to 22:02 UTC), as the convection was producing large hail and tornadoes across southwestern Minnesota

* Stop loop at 21:32 UTC image, and discuss the portal window in the upper left -- this has GOES-8 10.7 micrometer brightness temperature contours; Select a dark color such as Black to draw the arms of the "V" within the cold contour, and draw a small "W" over the Warm Wake region (between the cold arms of the "V")

* Apply the color enhancement "IR - convection" from the "Choose an enhancement" pull-down menu; compare the appearance of the Enhanced-V signature on the image and the portal

* Apply the "modified EC" enhancement, to show how the V signature might appear slightly different; Now select the "MB curve" to show the lack of fine detail on this "coarse" enhancement (mention that this is very similar to the grayscale "MB" curve used for the McCann study; this enhancement is only colorized on the cold end of the MB scale, which is important for cloud top features such as the Enhanced-V)

* Step back one image (using the "<" VCR button) to show that the Enhanced-V signature was first evident on the 21:15 UTC GOES-8 image; no "V" was evident on 20:45 image (mention no 21:00 image due to a full disk scan at that time); the first reported tornado from this convective system was at 21:23 UTC, which was not long after the 21:15 GOES-8 image -- not much lead time! Stress the importance of calling Rapid Scan Operations (RSO) imagery to try and catch the Enhanced-V signature as early as possible


* There was no RSO called on 29 March 1998, but there was 5 to 10 minute interval imagery from GOES-10 (the GOES-10 satellite was operated in "RSO mode" 24 hours a day during its Science Test in March-April 1998)

* Main image: GOES-10 10.7 micrometer IR, with the corresponding GOES-10 visible imagery in the smaller portal in the upper left; time interval is 20:35 to 22:05 UTC

* The IR enhancement used here uses blue to white to denote the coldest cloud top temperatures (-55 to -70 C)

* Stop image at 20:45: note the appearance of subtle shadowing on the GOES-10 visible image, which helps to locate the overshooting top (storm summit); in this case, the overshooting tops appear to be collocated with the coldest IR brightness temperatures (this is not always the case -- an upcoming VISIT Web-based session on "Storm Top Dynamics" discusses the various classes of storms, based upon the entraining parcel model work by Alder and Mack, 1986)

* Step ahead to 20:55 image: first hint of an Enhanced-V signature; however, should see the "V" on at least 2 consecutive images to be confident

* Step ahead to 21:00 image: now Enhanced-V signature is even more evident, so now we need to view this storm with suspicion; with the first tornado reported at 21:23 UTC, the appearance of the "V" on the 21:00 image (compared to the 21:15 GOES-8 image) gives a greater potential lead time (23 minutes, vs. 8 minutes) for a storm that is about to reach severe levels (Note: the first tornado warning based upon WSR-88D was issued from Sioux Falls at 21:09 UTC).

* Also note that the cloud top temperature differences (TDIFF) between the coldest and warmest portion of the Enhanced-V signature generally ranged from 8 to 12 degrees C, and the horizontal separation between the cold and warm pixels (DIST) ranged from 10 to 44 km


* Coldest temperatures from the Enhanced-V signature regions are plotted in Blue for both GOES-10 (solid line) and GOES-8 (dashed line)

* Warmest temperatures from the Enhanced-V regions (the "warm wake") are plotted in red for each satellite

* Storm Top Height (as defined by the height of the 30 dBz contour) from the Sioux Falls WSR-88D is plotted in violet

* Note that the temperatures within the Warm Wake appear to stabilize around -50 to -55 C, which was the temperature of the Equilibrium Level from the Omaha rawinsonde; the coldest temperatures within the "V" seem to oscillate about the -60 to -64 C level, which was about the tropopause temperature from the Omaha raob (click on the "Toggle" button to show the Omaha raob again, if desired)

* The Enhanced-V signature was first evident at 21:00, and persisted during the entire period shown here

* Also note that there was a period of warming followed by a period of cooling of cloud top temperatures just prior to the two tornadic periods (F0-F1-F2 from 21:23 to 21:45, and F3-F4 from 21:53 to 22:48 UTC); this agrees with a study by Fujita that indicated a warming of cloud top temperatures just prior to the storm becoming tornadic


* Click the main red Cursor in southwestern Minnesota, to align the main image with the smaller image portal on the left

* GOES-10 10.7 IR, from 20:45 to 21:55 UTC; location and size of large hail (A200 = 2.00 inch) and tornadoes (Fujita scale intensity are also plotted on these images; in the main IR images, the Black county outline locations have been corrected for parallax (assuming a 12 km / 39 kft mean storm top height); in the smaller portal images on the left, the blue county outline locations have NOT been corrected for parallax.

* Note how the parallax correction can move the location of the vertex of the "V" almost half a county (the far SW county in Minnesota is about 20 miles across) from its apparent "non-parallax-corrected" location; Parallax is important effect to keep in mind when comparing the location of storm top features to the location of severe weather on the ground.

* Note that AWIPS does not yet incorporate a parallax adjustment capability; parallax capability is scheduled to be added to AWIPS build 5.1.

...Now let's explore how the Enhanced-V satellite signature relates to various radar signatures...


* Click the main red Cursor in southwestern Minnesota, to align the main image with the 2 portal images on the right side

* On the left is a loop of Tilt 1 (0.5 degree) base reflectivity from the Sioux Falls SD WSR-88D, for the time period 20:48 to 21:48 UTC; the corresponding GOES-10 visible (with 5-minute NLDN CG strikes) and 10.7 micrometer IR imagery are in the smaller image portals along the right

* This presents a simultaneous view of the bottom and top portions of the convective systems; note that the blue map outlines in the GOES-10 imagery have NOT been corrected for parallax (the satellite imagery has been remapped into radar coordinates, which prevents us from applying a parallax correction)

* Note the hook echo on the 21:18 UTC radar, and the flanking line off to the southwest; on satellite, anvil cloud from the storm over far NW Iowa masks the flanking line, making it difficult to detect on the visible imagery

* The hook echo is still very obvious on the 21:23 UTC base reflectivity, as is the flanking line; 21:23 was the time of the initial F2 tornado; the vertex of the Enhanced-V appears to be significantly farther off to the northeast, but remember the parallax correction, which shifts the location of the vertex of the "V" back to the south and west, much closer to the location of the initial F2 tornado report.

* Step ahead through the loop, noting that the hook echo and flanking line become less obvious as the radar beam scans higher in the storm with increasing distance; some weak F0/F1 spin-ups appear to be focused along the interface of the rear flank downdraft and the southeasterly inflow region; At 21:58, the hook echo is again well defined, as the first F3 was reported; this F3 reached F4 intensity as it tracked along a 67 mile path to the NE. Again, note the location of the "V" with respect to the hook echo and the F3 report.


* Click the large red Cursor near the middle of this 4-panel figure (at the red * above the "Place Cursor Here" text) to align the panels in a favorable location

* This 4-panel loop covers about 1 hour, from 21:03 to 21:58 UTC, and shows Sioux Falls WSR-88D plus GOES satellite imagery; the upper left panel is the low-level base reflectivity (tilt 1); the upper right panel shows a mid-level tilt chosen to show either a Weak Echo Region or a Bounded Weak Echo Region; the lower left panel is an upper-level tilt chosen to show an echo overhang; the lower-right panel shows the corresponding GOES-10 10.7 micrometer IR image with large hail and tornado reports.

* If desired, you can hold down the SHIFT key and any mouse button as you roam around the images such that the small red pointers within each of the 4 panels are aligned at the location of interest (all 4 small red pointers remain co-located at the same pixel in each panel)

* Note the locations of the various low/middle/upper level features on radar, and the degree of "forward tilt" with height of the hook echo, WER/BWER, and echo overhang (images 8 through 10, 21:39 through 21:49 UTC, best show a low-level hook echo, a mid-level WER/BWER, and upper-level echo overhang); also note that the location ofthe satellite Enhanced-V signature corresponds fairly well with the location of the upper-level echo overhang (remember the parallax correction)


* Click the large red Cursor near the middle of this 4-panel figure (at the red * above the "Place Cursor Here" text) to align the panels in a favorable location

* This 4-panel loop covers about 1 hour, from 21:03 to 21:58 UTC, and shows Sioux Falls WSR-88D plus GOES satellite imagery; the upper left panel is the low-level base reflectivity (tilt 1); the upper right panel shows a mid-level tilt chosen to show either a Weak Echo Region or a Bounded Weak Echo Region; the lower left panel is GOES-10 visible with 5-minute NLDN CG strikes; the lower-right panel shows the corresponding GOES-10 10.7 micrometer IR image with large hail and tornado reports.

* If desired, you can hold down the SHIFT key and any mouse button as you roam around the images such that the small red pointers within each of the 4 panels are aligned at the location of interest (all 4 small red pointers remain co-located at the same pixel in each panel)

* Some trends observed in the NLDN data: (1) the total number of C-G strikes doubled prior to the second round of stronger tornadoes; (2) strikes were positive -dominated early, but there was a transition to negative-dominated C-G strikes as the total number of strikes increased; (3) there was a sharp decrease in the number of C-G strikes as the F3-F4 tornado began; (4) the C-G strikes were mainly clustered around the main mesocyclone region of the storm (not in the flanking line region), near the vertex of the Enhanced-V region on satellite


* Can be read verbatim if desired

* On the bottom bullet, mention that VISITview lessons have been developed which address the topics of AWIPS Enhancements and Rapid Scan Operations (RSO) issues -- see the VISIT Teletraining signup Web page at


* A1: An Enhaced-V signature implies that the updraft of a convective storm is well-organized and has reached a quasi-steady state, and has also penetrated the tropopause

* A2: An environment with little STORM-RELATIVE ANVIL-LEVEL flow would probably not exhibit an Enhanced-V signature; a strong upper-level jet is not required (as was seen in this Minnesota tornado case) - during the Jarrell TX F5 tornado, there was light flow in the upper troposphere, but the STORM-RELATIVE ANVIL-LEVEL flow was strong due to rapid storm propagation; therefore, a persistent Enhanced-V was seen during the Jarrell event

* A3: Certainly; some severe storms will not exhibit an Enhanced-V signature; however, there may be some situations where an Enhanced-V signature is contaminated by anvil debris from upstream convection.

* A4: Even though radar is the primary warning decision tool, an Enhanced-V could provide additional verification that a storm might become or has become severe; in the event of a radar outage (or less-than-optimal radar coverage, such as with beam blockage), this satellite signature could help to "fill in the gap"

* Thank the participants, and request feedback via email...


If time allows, there is additional material included at the end of this VISITview lesson:

PAGE 29 -- Additional material...

PAGE 30 -- KFSD 88D Storm Relative Motion (SRM)

GOES imagery animationss from other Enhanced-V cases:

PAGE 31 -- 03 May 1999 Oklahoma tornado outbreak

PAGE 32 -- 13 April 1999 convection over the Texas panhandle region

PAGE 33 -- 09 April 1998 Alabama tornado outbreak

PAGE 34 -- 27 May 1997 Jarrell Texas tornado event

PAGE 35 -- 14 July 2000 Pine Lake, Alberta tornado

PAGE 36 -- 25 May 2000 GOES/MODIS IR comparison (TN/AR region)