Weather prediction has become far more skillful in recent decades, with nearly all major regional weather/climate threats accurately predicted days in advance.
Unfortunately, several major cities have not taken advantage of the profoundly more skillful forecasts, resulting in unnecessarily severe loss of life and property.
For example, Los Angeles experienced immense damage and many deaths from the 2025 wildfires, the result of a well-forecast Santa Ana wind event.
Los Angeles, January 2025
Thousands died in New Orleans from Hurricane Katrina, whose track was skillfully forecast several days before.Seattle was caught unprepared for a well-predicted snowstorm in 2008, leading to a crippled city and the mayor being kicked out of office. In 2022, the region was not prepared for a well-forecast freezing rainstorm.
I just published a paper on the topic: Excellent
Forecasts but Poor Outcomes: Why Can’t We Do Better?
Seattle and the vicinity are vulnerable to well-forecast environmental threats, such as:
- Windstorms associated with Pacific cyclones
- Downslope windstorms descending the Cascades
- Snowstorms
- Ice storms
- Wildfire smoke
- Coldwaves
- Heatwaves
- Heavy rain and flooding
To deal with all these threats, Seattle should create an ENVIRONMENTAL WARNING SYSTEM.
A system that would monitor the potential for such threats, provide online forecasts of their impacts, and push critical, site-specific information out to smartphones and other devices.
Here is a sample of what I am thinking of:
You could monitor on your mobile device or computer, or request a notification when a meteorological threat looms. You could receive warnings and recommendations for personal actions to save lives or property.
Since your device knows your location, information and warnings could be specific to your position.
Some of you may know that the UW has already built components of such a system. For example, there is Seattle SnowWatch:
And Seattle WindWatch:
To build a comprehensive environmental monitoring and forecasting system for Seattle, King County, or other locations is quite possible and could save many lives and prevent property damage.
If heavy rains were expected at your location, you could attend to drains and gutters, avoid sleeping in the basement, and evacuate potentially flooded areas. Cities could bring in staff to clear drains.
For cold waves, homeless folks could be required to move inside (there have been too many deaths in Seattle from exposure in the past several years). For strong wind events, folks could avoid sleeping when trees could fall and prepare for outages. City Light could pre-position crews and resources.
You get the idea.
With highly skillful and site-specific warnings, society could make better decisions and greatly reduce the impacts of strong local weather events. The number of people injured or killed by severe weather could be greatly reduced with the proper application of the amazing forecast skill available today.
Thinking of some of the things that went wrong in the examples you site, I wonder if the people who should take action on the recommendations would actually follow through, considering they made poor planning decisions. For Katrina, the Hurricane missed New Orleans and went east and they had very good dykes on Lake Pontchartrain and the Mississippi River. However the canals connected the river to the lake were not adequate due to poor maintenance and in many failed by collapse rather than over-topping. Additionally, the poor constructed canal dykes had no mechanism to be closed off should they fail. That a civil engineering failure.
ReplyDeleteIn California in 2025 the Palisades fire likely erupted from a residual embers that were not completely extinguished a week earlier due to environmental concerns that the firefighters might trample some small plant growing in the region. The other fire near Altadena was caused by induced current that led to sparking on an un-used line. Did anyone anticipate that? However, the lines in use were due to be shut down when winds exceeded 60 MPH but the fire started when winds were blowing in the high 50's. Who set the threshold.
Forecasts have become much better but the limits or gaps in protection from infrastructure needs to be better understood and the gaps need to be corrected. There also has to be a better balance between fire protection and environmental protections that can lead to catastrophic results.
Great Idea. But, everybody wants government services, but mention user fees or higher taxes, and politicians and the public say no. Social Security and Medicare (socialist/Communist programs) are the best example of this push and pull.
ReplyDeleteMeanwhile, as often happens, our first "storm" of the season last week wimped out! I got a disappointing 0.17 inches of rain in Bothell... was supposed to be ~0.9 inches according to prediction.
ReplyDeleteThe NWS gave warnings regarding extreme fire danger several days before the catastrophic Old Trails fire engulfed entire neighborhoods but County emergency response officials waited three hours before providing any warning under the mistaken assumption it couldn't jump the Spokane River...which it did right into a very dense suburban neighborhood, at which time they issued a huge 'leave now' warning with no earlier warnings...prompting several hours of mass gridlock.
ReplyDeleteCliff - an app like this webpage ? https://trumedicines.github.io/wa-warning-system/
ReplyDeleteI do not know how to write to you outside of this blog. Anyway. I am looking at the UW Forecasts for this Saturday 9/12/2026. Today is 9/9/2026. So basically, it is 3 -4 days out for forecasting and should have about 70% accuracy??? UW WRF-GFS shows NO RAIN on Saturday for Seattle. BUT the NWS is showing 50% for Seattle for the same day. In my opinion, that's a big difference between the two organizations. Why the difference?
ReplyDeleteI wonder how much people would believe warnings coming from such systems. Not that they are always wrong, but forecasting in the Seattle area has a bad reputation -- witness the many predicted events that turn out to be nothing at all. This is probably improving, but there's an established reputation that will need to be conquered. Maybe this is true in many places?
ReplyDeleteThe established reputation will need to be addressed for systems like Professor Mass is suggesting here to work to their full potential. I believe that they have strong potential. But they can only work if the public is willing to trust them.
Deleteinteresting post and you have me intrigued. and I'm in software. I am going to see what it would take to build such an app open source. was interested to know if you could recommend good API or automated sources for this data, if you do reach out. would be an interesting project to make a POC.
ReplyDelete