Frank2 wrote:I agree that hiding truth is not right, but even today if you want to hear a snicker from an operational meteorologist, talk to them about the 360-384 Hour GFS. They still consider it fiction based on climatology, so why make forecasts based on what might happen 3 or 4 months in advance? What is the right or wrong percentage of a 10-day or even a 5-day forecast?
It'd be interesting to know the accuracy percentage of all CSU seasonal forecasts since Dr. Gray began issuing them in the mid-1980's.
In weather forecasting a person can only be 50% wrong, but the public is so overwhelmed at this point, why torment them with breaking news of a disasterous season when they have so much to deal with already? This is what the forecasters I worked for feared - needless worry.
In tough moments I often refer to the Apollo 13 incident, when a controller told Gene Kranz that the spacecraft was shallowing, and his question, "Can they do anything about it?" and hearing they could not, his response, "Then they don't need to know." and the same is true of a disasterous seasonal prediction - not forecast, but prediction.
Frank
There seems to be an attempt to equate seasonal forecasts months in advance with weather forecasts a day or three in advance. They are not the same thing. Seasonal forecasts aim to predict average conditions over a season, the definition of season beng dependant on the parameter being forecast. Some seasonal parameters have strong correlations with large scale monthly mean conditions which are either lagged, or can be predicted in advance. This is a whole different ball game to trying to predict how much rain will fall next weekend which is dependant on highly varying meteorological parameters from the planetary scale down to the microscale, which require numerical solution of the Navier Stokes equations, and which cannot all be fully simulated in any numerical model, leading to the inherant problem of small perturtions in the initial conditions can lead to large deviations in the forecast at even short lead times.
The skill of the CSU forecasts is somewhere on Phil's seasonal forecast documents and briefly on the TSR forecast documents. There is skill from about June onward which climbs through July and August. TSR acknowledge the uncertainty in their seasonal forecasts, which is why a probabilstic forecast and probability of exceedance charts are also included. You should also consider that a forecast does not need to be perfect, it just needs to be useful, and usefullness means it helps people to make better decisions.
Saying the public shouldn't be told about a forecast that is bad news because it might upset them is ridiculous. If you hide from upcoming bad events, they don't magically go away, all that happens is they hit you unprepared, which does you no favours.
I don't even have a clue what you mean when you say "a person can only be 50% wrong".
