Is Hawaii Prepared for the Next Wildfire Event? I Worry That the Answer is No.

Just over a year ago, western Maui was hit by a powerful downslope windstorm that initiated a wildfire that destroyed the historic town of Lahaina.

Lahaina after the August 8-9, 2023 fire

A terrible tragedy that cost the lives of approximately 100 people and resulted in roughly 4 billion dollars of damage.

Hawaiian Electric, local government, and others have taken some steps to prevent the tragedy from reoccurring.  Unfortunately, for the reasons outlined below, I worry there are gaps in their efforts that could lead to a repeat of the terrible disaster of August 2023.

Examining the Meteorology of the Lahaina Event

I do a lot of research on the meteorology of western U.S. wildfires, and particularly wildfires associated with downslope windstorm events, which characterize the 2023 Maui wildfire.  I received a grant from the National Science Foundation to study the meteorology and predictability of the Lahaina event, and after some intensive work with research meteorologist David Ovens, published a paper on the subject.

There were several significamt findings of the paper:

1.  The wildfire was associated with a strong downslope windstorm, which damaged powerlines starting the fire.

2.  The event was highly predictable up to approximately a week ahead using high-resolution model forecasts.

3.  The observational network on Maui and the rest of the islands was very poor.

4.  There were multiple levels of failure by Hawaiian Electric and public authorities.  Furthermore, the National Weather Service never highlighted the particular threat to Lahaina even though its high-resolution model (HRRR) clearly predicted the strong winds (see below)

The NWS HRRR Model

Hawaiian Electric's Attempt to Deal with the Threat

During the last months, Hawaiian Electric announced its response to the wildfire threat.  

First, they are adding several dozen new weather stations around the island, particularly in observation gap areas.  For example, there are now observations near Lahaina, where none had existed before (see current map below).  This is good.  But there are still major gaps in the observational network.  Furthermore, surface observations tell you what is happening NOW, not what will happen in the future.  And to save lives and property, forecasts are critical.

Finally, there is a lack of upper-air observations, which are critical for forecasting (more on this later).

Current Real-Time Surface Weather Observations on Maui

Second, Hawaiian Electric plans to harden and modernize its distribution network, much of which is outdated and below modern standards for robustness to strong winds.  Commendable and high priority, but will take years to complete.

Third, Hawaiian Electric plans to start doing Public Safety Power Shutoffs (PSPSs) when threatening conditions are expected, not unlike what has been done by California utilities.  Its a way of dealing with deficient power distribution network surrounded lots of flammable fuels.

Unfortunately, there are weaknesses in their plans for PSPSs.

According to Hawaiian Electric documents, their decision to initiate a PSPS will first depend on observations (identifying persistent drought conditions, wind gusts of 45 mph and more, relative humidity below 45%).  That is a good start, but reacting to observations is not good enough.  Winds can increase suddenly.  People, businesses, and others need more time to prepare.

Hawaiian Electric states they will use National Weather Service (NWS) public forecasts, but that is not good enough, something highlighted by the relatively poor performance of NWS predictions during the Lahaina wildfire.  During that event, the NWS failed to highlight the great threat to Lahaina.  Instead, the NWS put out a broad-brush red flag warning for the lee side of all the Hawaiian Islands.  Not good enough.

Hawaiian Electric needs to use high-resolution numerical model forecasts for their decision making including PSPS declarations.

There are a variety of ways Hawaiian Electric can secure such model predictions:

1.  Run the models themselves, which is what California utilities  (e.g. PGE) are doing.

2.  Use the NOAA/NWS HRRR-Hawaii model, which is run regularly in "experimental" mode.  This is totally silly by the way....NOAA/NWS needs a hardened, high-resolution operational model for Hawaii.

3. Hawaiian Electric should use the high-resolution regional weather prediction made by the University of Hawaii Manoa.

Hawaiian Electric also needs to have a team of meteorologists constantly monitoring both the observations and high-resolution model forecasts.  I don't thing they have any at this point.

They also should talk to the USDA Forest Service to give Hawaii critical wildfire weather diagnostics enjoyed by the rest of the U.S., such as the valuable HOT-DRY-WIND diagnostic.

Better Observations Aloft of Hawaii

To forecast and diagnose downslope windstorms.... the key driver of catastrophic wildfire on Hawaii, one has to know the structure of air approaching the islands through the depth of the lower atmosphere.   Right now Hawaii has such information only twice day at two radiosonde locations:  Hilo and Lihue, which weather balloons with instruments are launched twice a day (see stars on map).  Not good enough.

Hawaiian Electric, the State of Hawaii, and the Federal Government need more upper-air observations for the islands.  One approach is to add another radiosonde unit, with more frequent vertical soundings.  Or Hawaii could secure some vertical profilers, devices that sense the atmosphere remotely (see below).   The West Coast has many of these devices, why not Hawaii?

We Can Prevent Hawaiian Wildfire Disasters

The disaster in Lahaina last year was preventable.   Weather modeling and observational capabilities have improved so dramatically that we can foresee such events with great clarity and skill, allowing an effective response that will minimize loss of life and damage.  Hawaiian Electric and Hawaiian governmental entities must use these tools to prevent another tragedy.