Upper-air rawinsondes at station 74455 (Davenport, Iowa, above) show the response at that point to the approach and development of a potent winter storm (See the Airmass RGB animation below). The lowest levels of the atmosphere have saturated in the 12 hours between upper air observations. Model soundings, after the RGB animation, (from the Tropical Tidbits website) show that the GFS predicted the saturating that was observed. What kind of observations (independent of the model) might be available before the 1200 UTC upper air sounding to check to see if the actual atmosphere is behaving as the GFS predicts?

NOAA-20 overflew the Mississippi River valley shortly after 0800 UTC on 12 January 2023. NUCAPS profiles derived from the CrIS and ATMS instruments on NOAA-20 give valuable asynoptic information about the state of the atmosphere. The image below shows a toggle of ABI Band 13 image (without and with the 0.5 NEXRAD radar reflectivity overlain) along with the NUCAPS Sounding availability plot. The Green Points of the NUCAPS Soundings availability are where the infrared and microwave retrievals converged to a solution. Yellow points are where only the microwave retrieval converged. Red points denote retrievals that did not coverge. Of particular note: the slot of warmer brightness temperatures over central Iowa. Note that there are green points associated with that feature (including a green point in southwestern Wisconsin).

What do some of the 0800 UTC profiles show? That is shown in the animation below that includes profiles from green and yellow points over the mid-Mississippi River valley. The retrievals from the green points over Wisconsin and Iowa both show the same kind of thermal structure below 650-700 mb as is predicted in the forecast shown above, and as observed in the 1200 UTC sounding at Davenport. This is also true in some way for the yellow point in extreme northwest Illinois; for the one in in eastern Iowa, and especially the one near Davenport, the effects of falling snow lead to progressively less reliable information in the retrieval. Microwave-only retrievals that converge to a solution are far more likely to converge to a sensible solution if precipitation is not ongoing, so refer to radar imagery when using NUCAPS profiles.

You can use NUCAPS profiles, even in cloudy regions surrounding precipitation, to assess model performance in an ongoing event. In this case, NUCAPS profiles suggested that the atmosphere was evolving in a manner consistent with model predictions. Other kinds of observations that give useful model-independent thermal profiles (ACARS observations, for example) are unlikely to be present at this time of day over Iowa.

View only this post Read Less

In regions without radar, satellite estimates of rainfall intensity are vital in understanding precipitation. Microwave estimates have an advantage in that microwave energy is not strongly affected by clouds, so the internal structure of the cloud that can affect rainfall distribution can be detected. (However, microwave data has poorer spatial resolution than infrared.) Rain Rates shown above are generally less than 1″/hour. Microwave estimates do require a knowledge of the surface emissivity. Over the ocean, that emissivity is fairly well-known. Microwave rainfall estimates above show no rainfall along the coast of southern/eastern New England, an artifact, perhaps, of changes in the surface emissivity estimates that is affecting the Rain Rate algorithm. The GsMap site (from Japan’s JAXA), shows computed rain rate as well, using a variety of satellite sources. The toggle below compares the 0600-0659 and 0700-0759 UTC estimates on 10 January, bracketing the observation above (derived, incidentally, from data downloaded at the Direct Broadcast antenna at CIMSS and processed using CSPP software). The 0700-0759 UTC image does a much better job of matching the 0622 UTC Rain Rate shown above. Values off the east coast of the US are 10-15 mm/hour. The region of missing data along the coast is absent.

NOAA produces CMORPH2 estimates of precipitation as well, available at this website. The image below shows a screen-capture of rain rate at 0630 UTC on 10 January 2024. A zoomed-out view, it nevertheless shows the heaviest precipitation is offshore.

CMORPH2 is also available at RealEarth (link — then enter ‘CMORPH2’ in the Search Box). 1-hour precipitation rates 10-25 mm/hour are shown (25 mm/hr, of course, is 1 inch/hour)

GOES-R data are also used to create Rain Rate estimates, and the estimate from 0620 UTC, below, shows the original GOES-R algorithm output. Close inspection of the field will show a correlation with the cold cloud tops. However, the product algorithm has been modified considerably in the past five years, and output from the new ‘Enterprise’ algorithm (courtesy Bob Kuligowski, NOAA), not yet implemented in AWIPS, is shown at bottom. The Enterprise algorithm shows much less of a relationship with the very cold cloud tops, in agreement with the microwave estimates above.

View only this post Read Less

GOES-18 “Clean” Infrared Window (10.3 µm) images + Total Precipitable Water derived product, from 0000 UTC on 08 January to 0200 UTC on 10 January [click to play animated GIF | MP4]

10-minute Full Disk scan GOES-18 (GOES-West) “Clean” Infrared Window (10.3 µm) + Total Precipitable Water (TPW) images (above) showed the colder clouds associated with rain showers (with isolated embedded thunderstorms) and the clear-sky TPW field in the vicinity of Hawai`i during the 08-09 January 2024 time period. Surface analyses depicted a cold frontal boundary — which became more diffuse with time — that moved southeast across the island chain as a Storm Force Low intensified NW of Hawai`i.

A closer view using 5-minute PACUS Sector GOES-18 Infrared images (below) included 15-minute METAR plots — which showed periodic moderate to heavy rain; Local Storm Reports mentioned some flooding (and wind damage).

GOES-18 “Clean” Infrared Window (10.3 µm) images, from 0001 UTC on 08 January to 0101 UTC on 10 January  [click to play animated GIF | MP4]

Of note was the issuance of a Tornado Warning at 0515 UTC on 09 January for the far western portion of Molokai island (below).

GOES-18 “Clean” Infrared Window (10.3 µm) images from 0506 UTC to 0526 UTC, with the Tornado Warning polygon plotted in red [click to play animated GIF | MP4]

In a comparison of GOES-18 10.3 µm infrared brightness temperature, CLAVR-x Cloud Top Temperature derived product and CLAVR-x Cloud Top Height derived product for a portion of the tornado-warned storm (below), the Cloud Top Temperature value was nearly 3C colder than the single-channel infrared brightness temperature, and the Cloud Top Height was around 30800 ft.

Cursor sample of the GOES-18 Infrared image at 0516 UTC (white) along with the CLAVR-x Cloud Top Temperature (cyan) and Cloud Top Height (green) at 0521 UTC [click to enlarge]

Later in the day, a Severe Thunderstorm Warning was issued at 2044 UTC for the far southern portion of Hawai`i (below).

GOES-18 “Clean” Infrared Window (10.3 µm) images from 2021 UTC to 2111 UTC, with the Severe Thunderstorm Warning polygon plotted in pale yellow [click to play animated GIF | MP4]

In a comparison of GOES-18 10.3 µm infrared brightness temperature, CLAVR-x Cloud Top Temperature derived product and CLAVR-x Cloud Top Height derived product for a portion of the severe-warned storm (below), the Cloud Top Temperature value was about 3.6C colder than the single-channel infrared brightness temperature, and the Cloud Top Height was around 29800 ft.

Cursor sample of the GOES-18 Infrared image (white) along with the CLAVR-x Cloud Top Temperature (cyan) and Cloud Top Height (green) at 2101 UTC [click to enlarge]

Microwave estimates of Rain Rate (derived using the MiRS algorithm within the CSPP software package) are available at this website. The animation below shows seven observations on 8 January 2024, from 07 UTC to 23 UTC. The axis of the heaviest rain was over Oahu starting at 1900 UTC on 8 January.

View only this post Read Less

Blizzard warning were hoisted over the high plains of Kansas, Nebraska, New Mexico, Texas and Oklahoma for a strong storm on 8 January 2024 (image from the weather.gov website). The animation above shows visible imagery overlain with LightningCast probabilities centered over Kansas. Convection occurring over the cold aid will lead to increased snowfall rates, worsening local conditions (as if northerly winds of 30-40 knots weren’t enough!). LightningCast probabilities (that suggest where GLM observations within the next 60 minutes are most likely) give an advance warning of where conditions might be the worst in the short term. Note that Flash Extent Density values are scaled in the imagery from 1-15 vs. the default of 1-260.

LightningCast is available online here. LightningCast is a tool created via machine learning that uses ABI Bands 2, 5, 13 and 15 (0.64 µm, 1.61 µm, 10.3 µm and 12.3 µm, repsectively) to ascertain the lightning probabilities. Band 2, 5 and 13 are also the components of the Day Cloud Phase Distinction RGB, and a 1-hour animation, 1801-1901 UTC, of that field, overlain with LightningCast probability and GLM observations is shown below. It’s hard to tease the information in LightningCast from the Day Cloud Phase distinction RGB.

MIMIC Total Precipitable Water fields, below, for the 24 hours ending 1900 UTC on 8 January 2024, show how the large-scale circulation is drawing moisture northward out of the Gulf of Mexico.

The large-scale circulation is also apparent in the Advected Layer Precipitable Water (ALPW) fields, below. (source) Data for the different layers suggest that moisture transport is most apparent below 500 mb.

As is frequently the case with strong storms over the Southern Plains, strong surface winds on the south side of the storm have lofted dust, as shown in the toggle below between Visible and Dust RGB imagery at 2110 UTC on 8 January 2024 (available here as well). The pink/magenta enhancement in this RGB shows where dust is detected.

View only this post Read Less