A persistent Kona low coupled with strong water vapor transport from an atmospheric river has brought devastating floods to the state of Hawaii in late March 2026. Rain totals over the past ten days have exceeded 40 inches in some locations, prompting evacuation orders for thousands. Over two hundred people have been rescued from the rising flood waters.

The CIMSS MIMIC-TPW2 product gives a good overview of the situation. Here’s it possible to see how the cyclone circulation northeast of Hawaii is drawing the moisture-laden water northward from the near-equatorial region. As that air reaches the Hawaiian archipelago, the orography of the islands contributes to vertical lifting and enhances the rainfall.

Synthetic Aperture Winds from the Sentinel-1 satellite on 21 March show the largely southerly flow running into the island of Maui, where the influence of the terrain on rainfall is going to be significant; similar effects are happening on the other islands.

The Japan Aerospace Exploration Agency (JAXA) Global Change Observing Mission (GCOM) satellite is one that is monitored by CIMSS’s and NOAA’s network of ground-based direct broadcast antennas. With an antenna in Honolulu, CIMSS is able to produce low-latency products from polar orbiting satellites, including from GCOM’s Advanced Microwave Scanning Radiometer (AMSR2). Here’s an animation of the 36.5 GHz microwave brightness temperature, which is sensitive to deep precipitation-producing clouds, from 15 to 21 March 2026. This and other direct broadcast products from Hawaii produced by CIMSS can be seen here. It’s clear that for every day for nearly a week, persistent strong rain has been present over Hawaii.

As of the morning of Sunday 22 March, the National Weather Service anticipates the low continuing to move off to the northeast with atmospheric flow over Hawaii returning to a more typical trade wind pattern. Due to the devastation wrought by this event, the islands and their people will take much longer to return to normal.

View only this post Read Less

5-minute CONUS Sector GOES-18 (GOES-West) Infrared images viewed using RealEarth (above) displayed deep convection (with intermittent GLM-indicated lightning activity) that moved across the Hawaiian island of O’ahu on 20 March 2026 — producing heavy rainfall and flash flooding.

The first Flash Flood Warning for O’ahu on that day was issued at 0857 UTC, covering the northern portion of the island (below).

By 2041 UTC, Flash Flood Warnings covered the entire island (below).

As of 1749 UTC on 20 March, the highest rainfall amounts on O’ahu included 14.80″ at Waialua (on the north coast), 12.56″ at Shofield Barracks (on Wheeler Army Airfield, METAR identifier PHHI) and 12.21″ at Ka’a’awa (along the northeast coast). A plot of gauge height at the Wahiawa Reservoir Spillway near PHHI (below) showed the rapid rise that occurred after a period of heavy rainfall, peaking past minor flood stage at 85.29 ft (at 1840 UTC).

This particular episode of heavy rain and flash flooding was driven by the presence of a Kona low southwest and west of Hawai’i, as seen in surface analyses from the Ocean Prediction Center (below).

Southwest flow to the east of the Kona low was helping to transport rich tropical moisture across Hawai’i — as depicted by the MIMIC Total Precipitable Water product (below).

The tropical moisture being advected across Hawai’i was reflected in the Total Precipitable Water (PW) value of 2.01″ sampled at Lihue (the closest rawinsonde site to O’ahu) at 0000 UTC on 20 March (below).

According to a plot of the climatology of Total Precipitable Water values for all Lihue soundings at 0000 UTC on 20 March (below), the 2.01″ sampled by the Lihue sounding seen above exceeded the previous record daily maximum of 1.85″.

View only this post Read Less

2.5-minute Target Sector JMA Himawari-9 AHI Infrared images (above) showed Cyclone Narelle as it was rapidly intensifying over the Coral Sea on 18 March 2026 — becoming a Category 2 storm by 1200 UTC, a Category 3 storm by 1800 UTC and a Category 4 storm by 0000 UTC (ADT | SATCON). The coldest cloud-top infrared brightness temperatures were in the -90 to -95C range.

Himawari-9 Infrared / Water Vapor Difference images (below) exhibited large difference values — highlighting cloud tops that were significantly overshooting the tropopause.

Himawari-9 Infrared images with an overlay of Deep-Layer Wind Shear (below) indicated that Narelle was moving through an environment of modest shear — which, along with warm Sea Surface Temperatures favored further intensification.

Comparing DMSP SSMIS Microwave images at 0622 UTC (above) and 2004 UTC (below), the eyewall had become more consolidated and the eye diameter contracted a bit during that time period (signatures of an intensifying tropical cyclone).

===== 19 March Update =====

A Radar Constellation Mission (RCM-1) Synthetic Aperture Radar (SAR) wind speed image at 0847 UTC on 19 March (above) depicted a maximum of 130 kts in the SW quadrant of Narelle.

A nocturnal NOAA-21 (mislabeled by AWIPS as NPP) VIIRS Day/Night Band image (above) revealed concentric mesospheric airglow waves (reference) propagating outward from Category 4 Cyclone Narelle as the storm was just east of Australia’s York Peninsula.

2.5-minute Himawari-9 Infrared images (above) suggest that Category 4 Cyclone Narelle made landfall along Australia’s York Peninsula just after 2100 UTC on 19 March. During the hours preceding landfall, pulses of overshooting tops exhibited infrared brightness temperatures of -90C and colder (yellow pixels embedded within darker purple regions).

Even though the eye of Narelle was becoming increasingly cloud-filled toward landfall in Himawari-9 Infrared imagery, a microwave image from DMSP-17 SSMIS at 1950 UTC (above) still displayed a distinct eye signature.

View only this post Read Less

Just north of Libya on the afternoon of 17 March 2026, a large cyclone can be seen forcing a significant amount of dust from the Sahara to the Mediterranean. This true-color view from EUMETSAT’s Meteosat Third Generation (MTG) Flexible Combined Imager (FCI) depicts this situation beautifully. Since MTG-FCI has three visible channels, it is able to produce true color products with fewer assumptions than the GOES-R series of satellites has to rely upon in order to produce similar-looking imagery. This image is from 1230 UTC and is provided courtesy of EUMETSAT’s Eumetview.

Of particular interest is the large dust band that has been pulled from Libya and Egypt into the air over the sea. The Dust RGB can be used to denote where dust is prevalent. The following image is for the same time. Areas in magenta are associated with dust, and it’s obvious how widespread the dust is.

An interesting question is: what kind of storm is this, exactly? A medicane (a portmanteau of “Mediterranean” and “Hurricane”) is a certain type of cyclone that exhibits some characteristics of a tropical storm, including a warm core and an eye, that separates it from a more traditional mid-latitude cyclone. These are able to use the relatively warm waters of the Mediterranean to develop like a traditional tropical system would, but generally at a much lower intensity. It is March, of course, so the sea surface temperatures are still cool. The 15-16 C temperatures are on the low end of what has been observed with medicanes, but is still possible as much of the dynamical driving is done by upper-level cold air. The ASCAT Winds show a maximum wind speed approaching 40 knots, though it is not clear if the center of the storm is truly calm or if there’s just an observational gap there.

Looking at the animation, there is definitely an eye-like structure at the center of the storm, which is not seen in a standard midlatitude cyclone due to the low-level convergence of those systems.

The 700 hPa temperature field from the ECMWF model from 1200 UTC on the 17th also seems to indicate a hint of a warm core to the system, as can be seen in the bottom center of the graphic below.

Together, these factors have caused some European meteorological centers to officially label this a Medicane. Depending on which agency you listen to, it’s either called Samuel (France, Andorra) or Jolina (Italy).

View only this post Read Less