Saturday, September 23, 2017

Winter '16 / '17 - Arctic Oscillation (AO) Analog Verification

The analog forecasting technique seeks similarities to the AO state during the multi-month run-up to the coming winter with same AO run-ups of winters past.

The NEWxSFC method ranks analog years by their the sum of square errors (SSE) statistic.
Lower SSE errors ==> stronger analog

Arctic Oscillation (AO) Index Analog Forecast Verification

In the run-up to Winter '16 / '17 ... the leading analog was '12 / '13 followed by '73 / '74 ... '02 / '03 ... '74 / '75 ... and '88 / '89.  A qualitative assessment of the forecast's accuracy would rate the top three analogs as 'poor' ... IOW ... useless.

Winter '12 / '13 NAO oscillated weak negative to strong negative; whereas ... Winter '16 / '17 was essentially a mirror opposite.  The closest analogs ... on the basis of the AO remaining positive throughout the three-month forecast period ... were ranked last ('88 / '89) and next to last ('75 / '76).

Conclusion:  The top three Winter '16 / '17 AO analogs showed little ... if any ... skill.

---
An alternative forecasting technique looks at the AO's sign for any calendar-year's month preceding the pending winter as a potential leading indicator of the AO's sign for D-J-F period.

Results from a chi-square 'test for independence' infers a statistically significant relationship ... at the 95% confidence level and a p-value < 0.05 ... between NOV's AO sign and AO's sign of the upcoming D-J-F period.  IOW ... if NOV's AO is negative (positive) ... then average AO state during the upcoming winter will also be negative (positive); although the classification model is stronger ... i.e., lower false alarm rate ... for the predictor's month with a negative sign than positive.

Conclusion:  If NOV's AO is negative ... chances are good the AO state will average negative during D-J-F period.

---
Period-of-record AO data here.

Friday, September 15, 2017

Winter '16 / '17 - North Atlantic Oscillation (NAO) Analog Verification

The analog forecasting technique seeks similarities to the NAO state in the run-up to the coming winter with NAO run-up states of winters past. 

The NEWxSFC method ranks analog years by their the sum of square errors (SSE) statistic.
Lower SSE errors ==> stronger analog
Constraining the number of analog winters for analysis to five is arbitrary.

North Atlantic Oscillation (NAO) Index Analog Forecast Verification
In the run-up to Winter '16 / '17 ... the leading analog was '08 / '09 followed by '51 / '52 ... '54 / '55 ... '74 / '75 ... and '11 / '12.  A qualitative assessment of the forecast's accuracy would rate the top four analog as 'poor' ... IOW ... useless.

Winter '08 / '09 NAO was negative to neutral trending positive into MAR; whereas Winter '16 / '17 was weakly positive trending moderate.  The closest analog ... on the basis of the NAO remaining positive throughout the three-month forecast period ... was ranked last ('11 / '12).

Conclusion:  Winter '16 / '17 NAO analogs showed little ... if any ... skill.

---
An alternative forecasting technique looks at the NAO's sign for any calendar-year's month preceding the pending winter as a potential leading indicator of the NAO's sign for D-J-F period.

Results from a chi-square 'test for independence' infers a statistically significant relationship ... at the 95% confidence level and a p-value < 0.05 ... between NOV's NAO sign and NAO's sign of the upcoming D-J-F period.  IOW ... if NOV's NAO is negative (positive) ... then the average NAO state during the upcoming winter will also be negative (positive); although the classification model is stronger ... i.e., lower false alarm rate ... for the predictor's month with a positive signs than negative.

Conclusion:  If NOV's NAO is positive ... chances are good the NAO state will average positive during the D-J-F period.

---
Period-of-record NAO data here.