Tropical Cyclone Dorian become a hurricane on 28 August 2019. By the early morning of the 29th, the Dorian remained a category 1 hurricane as it advanced to the northwest, north of Puerto Rico.
GOES-East IR imagery and Sea Surface Temperature derived product with NHC forecast track overlay show the hurricane expected to continue through very warm waters as it approaches the Florida East Coast (Fig 1). Temperatures are between 29C and 30C along the track, or around 85F.
A 1-min mesoscale sector from GOES-East was available over the system. The high resolution imagery aids forecasters in identifying the center of circulation, as well as in tracking thunderstorm activity. Visible imagery with a semi-transparent GLM Flash Extent Density overlay at sunrise on the 29th clearly showed the center of circulation, with areas of thunderstorms, some strong, embedded in the outer bands (Fig 2).
Zoom in view on center of circulation at sunrise Fig 3).
The strong thunderstorm just northeast of the center of circulation exhibited high flash rates (see Fig 2) as well as persistent overshooting tops and even an above anvil cirrus plume, much like we see with strong thunderstorms over the CONUS (Fig 4).
Thunderstorms developed along a SW-NE oriented boundary draped across north-central Oklahoma during the early evening hours of 26 Aug 2019. The storms quickly grew upscale into a Mesoscale Convective System (MCS), and tracked south through central Oklahoma during the night, producing strong/damaging wind gusts and large hail.
One-minute visible imagery from GOES-East clearly shows convergence along a boundary, development of towering cu along the boundary, and eventual convective initiation (Fig 1), all in real-time. Storms quickly reached the equilibrium level and developed overshooting tops (OTs) and above anvil cirrus plumes, indicating particularly robust updrafts. Given the time of day around sunset, the shadows cast on the anvil provide additional insight into the vertical extent of the OTs.
GOES-East IR imagery showed rapid cooling of cloud tops as well as the development of overshooting tops (Fig 2). One method for viewing GOES imagery and GLM fields together in one display is to create a GLM contour color table. Since GLM fields are not gridded data in AWIPS, one cannot easily change it to a contoured field. However, the user can create a color table that implements contours at chosen thresholds instead of a constant fill. This display allows one to view VIS or colored IR imagery and GLM FED data in one display, not needing to make the GLM field semi-transparent. The GLM color table in this example includes three thresholds: 1 Flash/5-min (blue), 25 Flashes/5-min (Cyan), 100 Flashes/5-min (Yellow). For the most part, these solid color contours stand out and do not interfere with the IR colors.
Very deep convection developed over the western Mexico near the Gulf of California during the late afternoon hours of 20 August 2019, lasting into the early evening. This area of convection produced overshooting tops with IR brightness temperatures less than -90C (Fig 1)!
Analyzing a nearby sounding, these storms developed in an extremely moist environment (TPW=2.49″), very strong instability (MLCAPE = 4660 j/kg) and a very high equilibrium level (around 16.5 km). Based on the sounding, the -90C observed IR brightness temperatures likely put the OTs over 17.5 km above the surface!
Corresponding visible imagery of one of the storms is equally impressive. The final frame shows an OT penetrating deep across the equilibrium level casting an obvious shadow on the anvil (Fig 3).
Finally, a VIS/IR comparison of a -90C OT (FIg 4).
A severe thunderstorm developed quickly in the Pueblo, CO area during the early evening of 10 August 2019. GOES-16 imagery and lightning data provided clues to a rapidly intensifying thunderstorm with the potential to be strong to severe. While the environment was not all that supportive of large hail, steep low-level lapse rates and anomalously high Total Precipitable Water (TPW) values (~130% of normal) indicated the potential for strong downdraft wind gusts and heavy rainfall. Indeed, there were reports of heavy rain with this storm, as well as a 76 mph wind gust recorded at KPUB Pueblo ASOS (Fig 1). One-minute satellite imagery was available to forecasters during this event.
Analyzing 1-min visible and IR imagery from GOES-East over a 99-min period, we see several areas of cu initiate in the moist environment just west of Pueblo before organizing into into a single strong updraft (Fig 2 and 3). The updraft quickly reaches the tropopause and spreads out radially. An overshooting top is first apparent around 0000Z, while an above anvil cirrus plume starts to become noticeable after 0015Z.
The sandwich RGB combines the VIS and IR into a single image (Fig 4) into a single high-resolution qualitative product.
Looking at semi-transparent GLM Flash Extent Density (5-min accumulation updating every 1-min) gridded product overlaid on the VIS, we diagnose a rapid increase in total lightning between 2335Z and 2344Z (7 fl/5-min to 31 fl/5-min), or a lightning jump (Fig 5). This jump is an indicator of a strengthening updraft and often a precursor to hazardous weather at the surface. Cloud tops cooled from -45C to -64C during the same 9-min period. Lightning activity remained relatively stable through around 0000Z, while cloud tops continued to cool. Between 2345Z and 0000Z (after the lightning jump and during the continued cloud top cooling), very heavy rainfall (at least ~0.5 inches in 15 min) and gusty winds (estimated around 50 mph) were reported with this storm, in addition to improved presentation in radar imagery.
A second lightning jump occurred between 0002Z and 00009Z when FED increased from 23 fl/5-min to 56 fl/5-min). Cloud top temperatures bottomed out at around -76C at this time as well. During this lightning jump, very strong winds associated with the storm were measured at the surface, including the first severe wind gust of 59 mph at 0008 UTC, the first significant severe gust of 75 mph at 00010 UTC, and maxing out at 76 mph at 0016 UTC. The storm advanced into primarily rural areas thereafter, with lightning activity slowly dropping off, and cloud tops warming. See graph of IR and GLM trends for this storm in Figure 6, made using the AWIPS Tracking Meteogram Tool.