The rudders are on, provisioning is done, the task list has gotten smaller (is any task list ever really complete??) and there’s nothing keeping us from going offshore to New Caledonia—except weather. Since we sailed north from the Hauraki Gulf almost three weeks ago, it’s been an almost unbroken string of bad weather in New Zealand, with lots of rain and days of wind consistently in the 30s, gusting into the 40s, with only small breaks.
In these three weeks, there’s been one good weather window to head north. A couple boats took advantage of it, but most like us, weren’t quite ready. Plus, most insurance policies won’t allow you to go north until May 1, after the official end of cyclone season. So there are a fair amount of boats here in Whangarei and just north of us in Opua all awaiting the same fair weather window.
When we were on the Great Lakes, we often used Windy and Wind Finder to assess conditions for our trips up, down, and across Lake Ontario. When we left the lake to head down the East Coast, we started paying for a subscription to Predict Wind. It has become our trusty old friend and rarely a day goes by that we’re not checking the model runs, especially right now when we’re waiting with bated breath for any hint of a good window.
Generally, we check the new forecast models available for our area—ECMWF, GFS, Spire, UKMO, PWE, and PWG—while we have our morning tea. The model runs are released every 12 hours, so we check again for the newest ones before bed. (If you want to read more about the different forecast models, this a really handy explainer from the PW website.)
The most accurate forecasting is around 24-48 hours out. But we can look longer-term—up to 10 days out for the ECMWF and GFS and 7 days for the Spire, UKMO, PWE, and PWG. Even though the forecasting that far out isn’t dependably accurate, we’re still looking at it twice a day to get a sense of how the forecast is trending, what’s developing further west of us, and watching for the models to (hopefully) start coming into agreement with each other the closer we get. We’ll also periodically check the validation rankings for each model, which tells us which ones have been most accurate for wind speed and direction over the last day, week, or month.
One noted exception to long-range forecasting accuracy was this past February, with Cyclone Gabrielle’s approach toward New Zealand. Ten days out, it started showing up on the models, with a track straight down through the Auckland area. Seven days out, all the models started to agree with each other, with only slight variations on where the eye would go. That was about the time we headed to a dock in a protected harbor, still hoping it would veer away, but knowing that if we didn’t reserve dock space soon, we would be out of luck. A week later, what became post-tropical cyclone Gabrielle came roaring through almost exactly along the trajectory the models had predicted 10 days earlier.
As of this morning’s model runs, there’s a possible weather window for our passage about eight to nine days out, departing next Saturday. Over the last two weeks, we’ve watched several possible weather windows appear on the long-term forecasts, only to disappear a couple of days later, so we’re not getting too excited—yet. But hopefully, this one will stick around!