Going Whichever Way the Wind Blows: The Effect of the North Atlantic Oscillation (NAO) on Cava Quality

Is it possible to predict the quality of a wine, for example Cava (Spanish sparkling wine) just by using meteorological data?  Could you predict if a particular Cava would be good quality or poor quality just by knowing certain atmospheric conditions during a growing season, without ever needing to actual open the bottle and taste it?

One group recently examined the effect of the North Atlantic Oscillation and how this climatic event might influence the quality of Cava in a given year.


One thing I remember from the course I took in university on Environmental Climatology is that it’s complicated.  There is way more to atmospheric science that I can even fathom of delving into right now, so after reading this post you’re curious to know more, I’m afraid you might have to consult the Google.

The North Atlantic Oscillation (NAO) is defined as the changes in the difference of atmospheric pressure at sea level between the Icelandic low and the Azores high.  These changes control the strength and direction of the westerly winds as well as the storm tracks across the North Atlantic Ocean.  Additionally, the NAO is associated with temperature and rainfall levels in the North Atlantic, including the region where grapes for Cava wines are grown.

Apparently, North Atlantic Oscillation is also a band from Scotland.  Here is your musical interlude of the day (see embedded video):

Temperature and rainfall levels in the Cava region are very important in terms of physiological development of the grapes, as vine and grape maturation rates and overall quality are directly tied to these and other climatic variables.


In this short study, the researchers set out to answer one question:  what are the possible effects of the NAO on the quality of Spanish Cava?

The Cava Regulatory Council determines the quality of Spanish Cavas by categorizing them as excellent, very good, good, and fair.  Data from the vintage years 1970 through 2008 were available and used in the analysis.  For the purposes of creating a simpler model, the researchers combined wines rated as either excellent or very good into the “top quality” category, while those wines in the good or fair categories were placed in the “good and fair quality” category.

Photo in Public Domain - Source: NASA

Photo in Public Domain – Source: NASA

Monthly NAO values between 1970 and 2008 were taken from an extensive data set managed by the United States National Oceanic and Atmospheric Administration (NOAA).   In addition to a growing season-wide average (March through August), monthly average values and season average values were also determined to see if there were specific months or seasons that were better predictors of Cava quality or not.  Specifically, spring NAO values were averaged for the months of March through May, while summer NAO values were averaged for the months of June through August.

Using logistic regression as well as favorability and probability mathematically models, the researchers could determine if NAO values had anything to do with the probability of a favorable Cava vintage or if it’s all just crazy talk.

Results & Discussion

A few interesting results came out of this study:

1)      A significant negative relationship between NAO for the months of March through August and the probability of having a high/top quality Cava for the given vintage.

A negative NAO is associated with increased rainfall in the Cava region.  This increased rainfall may help keep temperatures at a more mild level and help with a more prolonged maturation process.  A positive NAO, on the hand, is associated with warmer conditions and higher daily maximums in the Cava region (and throughout the entire Iberian Peninsula).  According to other research studies, increased numbers of days with high temperatures can reduce the overall quality of the finished Cava wines.

2)      Monthly and seasonal NAO values were not related to Cava quality.

While the NAO values over the entire growing season (March through August) proved significant in predicting the quality of Cava, month-by-month values and season-by-season values did not.  It appears as though the cumulative behavior of the NAO is more important to determining Cava quality than any one particular month or one particular spring or summer season.  This effect has also been shown in other studies for other organisms, like Albacore tuna.  This effect was more pronounced in seasons with “extreme” NAO values.

3)      Comparing predicted quality with actual quality:  80% of years determined as “favorable” based on NAO values were found to actually produce top quality Cavas.  70% of years determined as “unfavorable” based on NAO values were found to actually produce lower quality Cavas.

Basically, this means that the probability model the researchers used to predict actual quality of Cava was accurate 80% of the time when NAO values indicated that the year would be favorable to high quality Cavas, and accurate 70% of the time when NAO values indicated that the year would most likely yield lower quality Cavas.  When favorability was even higher than 80% for a given year based on the NAO value, the classification predicted for the resulting Cava was 100% on target with the actual quality level of Cava.

In other words, when the model was “really sure” that the NAO conditions were favorable for a top quality Cava year (i.e. greater than 80% favorability), the model predicted the high quality with 100% accuracy.  If the NAO conditions were favorable, but not quite as high as 80% favorability, the predicted high quality level of the Cava was correct 80% of the time.  On the other hand, if NAO conditions were not favorable, the predicted low quality level of the Cava was correct only 70% of the time.

Concluding Thoughts

While 80% and 70% accuracy is pretty good, in my opinion, it shows that NAO values can’t COMPLETELY predict Cava quality at a certain point.  Sure, when favorability for a good year based on NAO values was greater than 80%, there was a 100% accuracy rate in determining overall Cava quality.  However, since favorability fluctuates from year to year, this model using NAO alone can’t be relied on for completely accurate results.

While NAO values appear to have a significant influence on the overall quality of Cava wines, it isn’t the only thing that is going on.  Except when extreme conditions dictate, there are clearly some other environmental or possible winemaking techniques used to help account of the remaining variability in being able to predict overall Cava quality that need to be determined.

In general, however, if the answer were purely yes or no, I would say that yes, the North Atlantic Oscillation does a good job predicting the quality of Cava in a given vintage, though sometimes there are other factors involved, be it other climatic events, environmental conditions, or winemaking techniques, that can’t be predicted by NAO alone.

I would love to see an ongoing study looking at the effect of NAO on Cava quality, considering patterns and behaviors of many atmospheric conditions will fluctuate even more during this time of fluid climate change.

What do you all think of this study?  Is weather awesome, or what?  😉

Primary Source Study:  Real, R., Báez, J.C. 2013. The North Atlantic Oscillation affects the quality of Cava (Spanish sparkling wine). International Journal of Biometeorology 57: 493-496.

Finally, the inspiration for the title of this post: