Not sure about the cricket analogues here, but they are using a different approach , using a model to predict the winter pattern. Here is an extract from a paper cited in the recent Nature article

'The forecasts used here were produced using the Met Office Global Seasonal forecast System 5 (GloSea5). The climate model at the core of this forecast system is Hadley Centre Global Environmental Model version 3 with atmospheric resolution of 0.83° longitude by 0.55° latitude, 85 quasi-horizontal atmospheric levels, and an upper boundary at 85 km near the mesopause. The ocean resolution is 0.25° globally in both latitude and longitude with 75 quasi-horizontal levels. This resolution is necessary to reduce key biases in the ocean and atmosphere and give a realistic winter blocking climatology in the model [Scaife et al., 2011]. A 24-member ensemble of forecasts was run for each winter in the period 1993 to 2012 with lagged start dates centered on 1 November (25 October, 1 November, and 9 November) and eight members initialized on each of the three start dates. Members from the same start date differ only by stochastic physics [Arribas et al., 2011]. Initial atmospheric and land surface data were taken from ECMWF Re-Analysis (ERA)-Interim observational reanalyses, and initial conditions for the global ocean and sea ice concentration were from the Forecasting Ocean Assimilation Model (FOAM) system [Blockley et al., 2013]. This configuration allows very skillful predictions of various slowly varying components of the climate system to be made for the coming winter (Table S1)'

In other words they are saying that they have a much better method of predicting NAO now certainly for the upcoming winter and indeed as they have now shown a year ahead.

Originally Posted by: TomC 

Not sure about the cricket analogues here, but they are using a different approach , using a model to predict the winter pattern. Here is an extract from a paper cited in the recent Nature article

'The forecasts used here were produced using the Met Office Global Seasonal forecast System 5 (GloSea5). The climate model at the core of this forecast system is Hadley Centre Global Environmental Model version 3 with atmospheric resolution of 0.83° longitude by 0.55° latitude, 85 quasi-horizontal atmospheric levels, and an upper boundary at 85 km near the mesopause. The ocean resolution is 0.25° globally in both latitude and longitude with 75 quasi-horizontal levels. This resolution is necessary to reduce key biases in the ocean and atmosphere and give a realistic winter blocking climatology in the model [Scaife et al., 2011]. A 24-member ensemble of forecasts was run for each winter in the period 1993 to 2012 with lagged start dates centered on 1 November (25 October, 1 November, and 9 November) and eight members initialized on each of the three start dates. Members from the same start date differ only by stochastic physics [Arribas et al., 2011]. Initial atmospheric and land surface data were taken from ECMWF Re-Analysis (ERA)-Interim observational reanalyses, and initial conditions for the global ocean and sea ice concentration were from the Forecasting Ocean Assimilation Model (FOAM) system [Blockley et al., 2013]. This configuration allows very skillful predictions of various slowly varying components of the climate system to be made for the coming winter (Table S1)'

In other words they are saying that they have a much better method of predicting NAO now certainly for the upcoming winter and indeed as they have now shown a year ahead.

Originally Posted by: TomC 

That's all well and good but I still don't see how this improves on the May SST NAO correlation which was being discussed 17 years ago. Am I missing something here? 

Originally Posted by: Brian Gaze 

I would say 2 things:

1. They aren't just predicting NAO but the other factors, temperature, rainfall etc relevant to winter weather
2. These prediction are more accurate than from May SST and then inferred from the NAO forecast.

Originally Posted by: TomC 

How can the meto predict a whole season based soley on the NAO,when clearly there are many other factors that influenze the uk`s weather??

Also coming out and saying the start of winter will be colder and drier than 2015 is stating the bleedin` obvious

Originally Posted by: springsunshine 

How can the meto predict a whole season based soley on the NAO,when clearly there are many other factors that influenze the uk`s weather??

Also coming out and saying the start of winter will be colder and drier than 2015 is stating the bleedin` obvious

Originally Posted by: springsunshine 

Seeing the Met office saying that they  can predict the NAO type a month ahead set me examining the frequency and persistence od NAO winter types.The data set in monthly NAO's from 1826 to 2015 a total of 190 years.

The first  observation is that NA0+winters (DJF) occur in 68%of the years.Another way of saying this is that NAO- winters occur only once in three years sowing the strength the default NAO + position.No wonder cold lovers get frustrated!

The second observation is that there are different levels of frequency in the three winter months.The % for NAO % are Dec 56%,Jan,68% and Feb 66%.This is somewhat surprising result because it shows December as the most blocked month.There  is thought to be a singularity of a mid December Anticyclonic spell so maybe that is showing through in the December figures.

The more blocked nature of the month of December and its position as the first month of winter led me to look at what followed the December NAO types.

NAO + Decembers were followed by NAO + 76% NAO+ Jans and 73%  NAO + febs these being small amount greater than  expected frequencies. NAO -Decembers were followed by 60% NAO + Jans and 61% Febs slightly lower than would be expected.

However here is the fascinating figure .If we look at the entire winter NAO (DJF) than NAO + Decembers are 85% likely result in 85% NAO + for DJF..For NAO- Decembers the figure for DJF is NAO + is just 45% with NAO at 55% forming the majority.These results show strong variation from expected levels.

Could it be that the December NAO sign is the bell weather for warm/cold winters? Given that we can track the NAO through the month ten we have an early indicator of what December may turn out to be

Originally Posted by: roger63 

Seeing the Met office saying that they  can predict the NAO type a month ahead set me examining the frequency and persistence od NAO winter types.The data set in monthly NAO's from 1826 to 2015 a total of 190 years.

The first  observation is that NA0+winters (DJF) occur in 68%of the years.Another way of saying this is that NAO- winters occur only once in three years sowing the strength the default NAO + position.No wonder cold lovers get frustrated!

The second observation is that there are different levels of frequency in the three winter months.The % for NAO % are Dec 56%,Jan,68% and Feb 66%.This is somewhat surprising result because it shows December as the most blocked month.There  is thought to be a singularity of a mid December Anticyclonic spell so maybe that is showing through in the December figures.

The more blocked nature of the month of December and its position as the first month of winter led me to look at what followed the December NAO types.

NAO + Decembers were followed by NAO + 76% NAO+ Jans and 73%  NAO + febs these being small amount greater than  expected frequencies. NAO -Decembers were followed by 60% NAO + Jans and 61% Febs slightly lower than would be expected.

However here is the fascinating figure .If we look at the entire winter NAO (DJF) than NAO + Decembers are 85% likely result in 85% NAO + for DJF..For NAO- Decembers the figure for DJF is NAO + is just 45% with NAO at 55% forming the majority.These results show strong variation from expected levels.

Could it be that the December NAO sign is the bell weather for warm/cold winters? Given that we can track the NAO through the month ten we have an early indicator of what December may turn out to be

Originally Posted by: roger63 

Here is the original Nature Geoscience paper un-paywalled:

http://sci-hub.bz/10.1038/ngeo2824

There's no claim of "62% accuracy" that I can see. That might be the result of the hack at the Telegraph thinking an r-value of 0.62 means 62% accuracy. (it's an expression of the scatter between actual values and modelled values.) To be fair to the hack, rather too many climate/weather "scientists" in recent years have been silent when values with a specific statistical meaning have been misinterpreted to mean percentages, especially when the misinterpretation suggests their models are going to predict the future, when in fact they're simply modelling the past (as here).

Incidentally, the r-value is only 0.62 for the first winter, falling to a rather less impressive 0.42 for the much-vaunted 12-months ahead (an r-value of less than 0.5 suggests only a small relationship between model and reality).

If you want to eyeball how well the modelling in this case works in practice, have a look at the top two diagrams on page 2 of the paper. Notice how in the one on the right (the NAO 12 months ahead) their models backfit quite well until the most recent couple of years. Whatever caused the startlingly positive NAO and mild winter last year, this model missed it completely and predicted the opposite 12 months out.

Originally Posted by: some faraway beach 

Here is the original Nature Geoscience paper un-paywalled:

http://sci-hub.bz/10.1038/ngeo2824

There's no claim of "62% accuracy" that I can see. That might be the result of the hack at the Telegraph thinking an r-value of 0.62 means 62% accuracy. (it's an expression of the scatter between actual values and modelled values.) To be fair to the hack, rather too many climate/weather "scientists" in recent years have been silent when values with a specific statistical meaning have been misinterpreted to mean percentages, especially when the misinterpretation suggests their models are going to predict the future, when in fact they're simply modelling the past (as here).

Incidentally, the r-value is only 0.62 for the first winter, falling to a rather less impressive 0.42 for the much-vaunted 12-months ahead (an r-value of less than 0.5 suggests only a small relationship between model and reality).

If you want to eyeball how well the modelling in this case works in practice, have a look at the top two diagrams on page 2 of the paper. Notice how in the one on the right (the NAO 12 months ahead) their models backfit quite well until the most recent couple of years. Whatever caused the startlingly positive NAO and mild winter last year, this model missed it completely and predicted the opposite 12 months out.

Originally Posted by: some faraway beach