Forecasting's Achilles Heel

Today we are experiencing a noticeable forecast failure and one that in some sense is self-inflicted.

Here is Puget Sound country it is going to be a beautiful day...lots of sun and temps rising into the 70s. Look outside or view the latest visible satellite picture.


On the other hand the National Weather Service forecast RELEASED THIS MORNING paints a less optimistic picture.


And yesterday's forecast was even more pessimistic.


The computer forecasts yesterday showed the break between systems (see example) and certainly this morning it was clear.


Why didn't the message about a spectacular break on Saturday get out?

I think there are three main reasons:

1. The National Weather Service forecast cycle is only updated every 6 hr in most cases and there is a lack of emphasis on nowcasting--describing what is happening now and during the next few hours.

2. There is a distinct tendency for the National Weather Service to broadbrush their forecasts--smear out clouds and weather over an extended period and not to put emphasis on breaks in the weather...even when they are pretty obvious.

3. Finally, there is the tendency in the NWS to maintain forecast consistency--staying with the same story--even when new guidance suggests otherwise. This is based on an internal philosophy not to jerk the forecast around as numerical guidance changes.

Personally, I think this all has to change...and in fact this blog is partially a reaction my feelings.

I believe that that providing frequent updates on current and expected weather is a hugely important area for development and that society has much to gain from this direction. For many of us, knowing what is happening and what will happen in the next 6 hrs is hugely important...and has great value for saving property and lives. To be fair, when severe weather is occurring the NWS does do more nowcasting, but I think they need to do so on a more regular basis.

In a day with smartphones, internet-capable cell phones, and computers on the internet everywhere, the ability to deliver real-time weather information exists. New software applications, better computer modeling, and a huge increase in observations will make the information available. We just have to put the package together--and society has much to gain from it.

The nightly weather on the local news is great, but people need weather information all the time...and we have to find a way of delivering it. An idea: every major city could have a nowcasting weather broadcast on the internet, updating the current weather situation every 15 minutes.