Flood forecasting

I was in Brisbane because Queensland had recently been in floods and it was anticipating that floods might be coming up. I do research on how to make better short-term (less than 10 day ahead) streamflow forecasts. So the Bureau of Meteorology invited me up to see the flood warning office in full swing.

Here's their building on at 69 Ann St:

It's a professional and modern building, with stunning views from the 21st floor. The flood warning center shares office space with the weather forecasters. Here's one of the meteorologists at work.

I had monitor envy. The guy who does data quality control and monitoring of the rainfall systems gets to work at a station with 6 screens:

Inland queensland had been flooding on and off since Christmas. During the visit there was a weather briefing from the meteorologists, saying that there was near uniform agreement from the weather models that a major storm was going to cross down from the north coast into the interior. About 5 days ahead, the forecast was that it would drop more than 100 mm in a day, which is about 4 inches.

In addition to output from models run on some of the world's most powerful computers, there's still a lot of work that happens drawing weather patterns on paper maps with colored pencils:

Yearly average rainfall for the interior is about 200-300 mm. So, yeah, 100 mm's a huge deal, especially since the ground was already saturated!

On the left is Peter Baddiley, the head of the Queensland flood forecasting group. When the rainfall scenarios were run through the river models, the response was epic floods. However, 5 days ahead there's so much uncertainty, you try and be very cautious to avoid false alarms. Although a general alert wasn't sent out that day, high level government authorities (including an emergency management representative who had a jacket with the agency's acronym on the back) were brought in and briefed on the situation while I was there.

It brought me right back to when I was an operational forecasting in the US. Is it going to happen, is it not going to happen? Can you trust your models? Could it be even worse than the models suggest? The thrill, the drama, the anticipation. At the same time, you have to not let your emotions get in the way; Pardon the gruesome imagery, but it reminds me of this quote from the movie Jaws

"Sometimes that shark he looks right into ya. Right into your eyes. And, you know, the thing about a shark... he's got lifeless eyes. Black eyes. Like a doll's eyes. When he comes at ya, doesn't seem to be living... until he bites ya, and those black eyes roll over white and then... ah then you hear that terrible high-pitched screamin'."

Nature doesn't care about you. Doesn't care what you think is going to happen. You may fear this or that, or maybe last night you saw a movie about something that plants the seed in your mind of a big drought or flood. But the important thing is to be level headed and objective, pay attention to what the evidence is telling you and always be aware of self-delusion.

In the end, the rainfall was ginormous, widespread areas of heavy rainfall.

The flood response has been big too, large areas in the Major Flood category.

It reminded me that the favorite part of my last job as an operational forecaster was being in that situation of knowing that the tools I had at the time were functional but could be improved. I'd go off and do a bit of tinkering and development and then next time around we'd be better armed to know what was really going to happen. When the event finally unfolded like we thought it would, there was a mix of disbelief and ecstasy that it actually worked and we had been able to discover Truth before it happened. So rarely in life do you discover with certainty that your belief is right or wrong, often you can get by with just arguing for a convincing position. I'd like to think having that experience built character.

My father loved gambling but my mother hated it, so at the very least I got to satisfy that betting urge without losing any money!