Interesting:
Quake hit at 8:04 local time. I'm guessing that's also Vancouver time. Vancouver is GMT -8. Honolulu is GMT -10 if google is correct.
That means the wave would get to Hawaii in 4.5 hours? Could be, seems fast. Am I making a simple arithmetic error or do waves move that fast?
The short answer is "500-600 MPH".
The waves are very tiny in height (only measured in millimeters or maybe inches) but they're thousands of feet, even miles long. There are more details & diagrams here:
Tsunami Characteristics - PDC
Here's the sequence of messages issued by the Pacific Tsunami Warning Center:
http://ptwc.weather.gov/index.php
(This link is only good for the last 30 days of traffic.)
The quake hit at 0304Z (UTC) and set off everyone's seismographs. The PTWC spend a half-hour looking for buoy data and running computer "what if" models, but the quake location & magnitude didn't seem to indicate the formation of a tsunami.
That assessment changed at 0514Z (two hours after the quake) when two DART II buoys started bouncing around-- all of 2.5 inches. (Scroll down this link
NDBC - Deep-ocean Assessment and Reporting of Tsunamis (DART®) Description for the system diagram of a DART buoy.) Added to the evidence of a tide gauge in British Columbia, they called a warning with an arrival of a bit more than three hours later.
This diagram shows the location of various Pacific sensors.
http://www.noaanews.noaa.gov/stories2008/images/buoylocations.jpg
Notice the big gap between the Pacific rim buoys and the island buoys. For most of that three hours, there was no data to update the forecast. Satellites couldn't pick the tsunami wave out of the clutter and shipping wouldn't even notice it. The wave finally passed under a few weather buoys to the north of the islands-- not special-purpose tsunami buoys, but good enough.
My spouse is an oceanographer and she used to be a Navy Emergency Preparedness Liaison Officer working with FEMA and Hawaii Civil Defense, so she talked me through the events that happened after the 0514Z warning. The "problem" was that the quake wasn't big enough to pose an immediate threat, and the wave was still far enough away to allow for a little deliberate thought before action. The PTWC computer model was calling for shore waves of at least a couple of feet, but not necessarily threatening enough for the watch officer's procedures to call for immediately pushing the button to sound the tsunami sirens. Instead they had to discuss the issues (safety vs budget) with CD and make a courtesy call to the governor. (This is the process that they've worked out over the years, although now they might decide to lower the number to let the watch officer hit the button a little sooner.) So the first siren fired up around 0545Z, with just two hours and 45 minutes of warning.
I was sitting at the computer when the sirens went off, and Twitter had the answer faster than the local news media websites could load. (Luckily I chose to search for hashtag #tsunami rather than #hurricane or #nuclearattack, so I got the right answer on the first guess.) Spouse had to channel check her TiVo for a bit before she got to the CD crawler.
I guess the most pain was felt by the traffic police. I don't know if 165 minutes was not enough time for the traffic to clear the streets, or too much time to give all the lookie-lookies the idea to drive down to the beach to see the show. But by 0810Z (T-20) the police were ordering people to abandon their cars in the middle of the traffic jam and start walking smartly inland/uphill. At 0830Z the beach webcam showed a couple people actually wading into the shorebreak to record the arrival on their smartphones.
When a tsunami "arrives" it can be a six-hour event. It can literally take more than an hour for the bigger waves to hit, so it was 1100Z (another 2.5 hours after the tsunami's arrival, 1 AM Honolulu time) before PTWC was ready to cancel the warning. Even then CD waited another hour (until Hilo had settled down) before deciding to declare all clear.
People still died in the traffic jams, but hopefully lives were saved by sounding the alarm. It'll be interesting to read the after-action reports over the next few weeks and see what procedures change.