![Page 1: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/1.jpg)
Seasonal to Interannual Predictability of Arctic Climate
by Ed Blanchard and Cecilia Bitz Atmospheric Sciences
University of Washington
![Page 2: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/2.jpg)
![Page 3: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/3.jpg)
http://nsidc.org/arcticseaicenews/2010/040610.html
![Page 4: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/4.jpg)
![Page 5: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/5.jpg)
Based onSatelliteMotion Fields
Maslanik andFowler CU
http://nsidc.org/arcticseaicenews/2009/040609.html
![Page 6: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/6.jpg)
Part 1:
Analysis of a 30 member ensemble of A1B scenario from CCSM3 T42 resolution
Thanks to H. Teng and C. Deser
![Page 7: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/7.jpg)
Dec
Nov
Oct
Lagged Correlation of Sea Ice Area in Ensemble Mean
Lag (months)
Sep
Sep
Sep
JanFebMarAprMarJunJulAugSepOctNovDec
0 5 10 15 20 25 30 35
![Page 8: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/8.jpg)
Sea ice Area Climatology in 106 km2
Thickness memory
SST memory
Blanchard, Armour, Bitz, and DeWeaver, submitted
![Page 9: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/9.jpg)
Squared one-month lag correlation
J F M A M J J A S O N D
Total ice area is more persistent than extentBoth are highest in mid winter and mid summer
![Page 10: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/10.jpg)
Part 2: “Perfect Model” Studies with CCSM4
Initialized in year ~2000 of a 20th century runstarting May 1, July 1, Sep 1, and Jan 1
**
*
Sea
Ice
Area
(106
km2 ) *
* = start timeslines = 20th
Century runs
![Page 11: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/11.jpg)
****
Arctic Monthly Sea Ice Area Anomaly (106 km2)
* = start timeslines = 20th
Century runs
![Page 12: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/12.jpg)
* = start timeslines = 20th
Century runs
Arctic Monthly Sea Ice Area Anomaly (1013 km3)
****
![Page 13: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/13.jpg)
20 Ensemble members for each start date (80 runs)
CCSM4 1deg resolution
Runs are 2 years long
IC are taken from 20th century run year 2000 with atmosphere perturbed using adjacent days
![Page 14: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/14.jpg)
Growth of the Standard Deviation of Sea Ice Area (106 km2)
ControlJan 1 StartMay 1 StartJuly 1 StartSep 1 Start
error bar is standard error of the standard deviation
σ
![Page 15: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/15.jpg)
ControlMay 1 StartJuly 1 Start
Ratio of Green Line to Red Line
Ratio of AR1 models using lag correlations from the control
σ
σMay/σJuly
σ - Growth of Sea Ice Area
(106 km2)
![Page 16: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/16.jpg)
All Curves are from the ensemble with July 1 I.C.’s
July (lag 0)Aug (lag 1)Sep (lag 2)Oct (lag 3)
Ratio of σpredicted/σcontrol by sector in the Arctic of Sea Ice Area
0 50 100 150 200 250 300 350
1.2
1.0
0.8
0.6
0.4
0.2
0
Longitude
Barents Sea Siberian Shelf Beaufort
![Page 17: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/17.jpg)
σcontrol Standard Deviation of Surface Temperaturein 1995-2005 of 20th century runs
Jul Aug
Sep Oct
![Page 18: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/18.jpg)
σpredicted Standard Deviation of Surface Temperatureacross 20 ensemble members with same July 1 I.C.’s
Jul Aug
Sep Oct
σpredictedσcontrol
σpredictedσcontrol
Aug
Oct
![Page 19: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/19.jpg)
Standard Deviation of Surface Temperaturein October
σpredicted July start σpredicted May start σcontrol
![Page 20: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/20.jpg)
All Curves are from the ensemble with July 1 I.C.’s
July (lag 0)Aug (lag 1)Sep (lag 2)Oct (lag 3)
Ratio of σpredicted/σcontrol by sector of Arctic* Surface Temperature
0 50 100 150 200 250 300 350
1.2
1.0
0.8
0.6
0.4
0.2
0
Longitude
Barents Sea Siberian Shelf Beaufort*Northof 60N
![Page 21: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/21.jpg)
Growth of the Standard Deviation of Sea Ice Volume (1013 m3)
Control
Jan 1 StartMay 1 StartJuly 1 StartSep 1 Start
σ
error bar is standard error of the standard deviation
![Page 22: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/22.jpg)
Growth of the Standard Deviation of Multiyear Ice Area (106 km2)
error bar is standard error of the standard deviation
Control
Jan 1 StartMay 1 StartJuly 1 StartSep 1 Start
![Page 23: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/23.jpg)
Standard Deviation of Ice Area (106 km2)in the 20th Century “Control” for 1995-2005
0.5
0.4
0.3
0.2
0.1
0
![Page 24: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/24.jpg)
Control
Ensemble initialized in May
Standard Deviation of Ice Area (106 km2)
Ensemble initialized in July
![Page 25: Seasonal to Interannual Predictability of Arctic Climate€¦ · 20 Ensemble members for each start date (80 runs) CCSM4 1deg resolution. Runs are 2 years long. IC are taken from](https://reader034.vdocuments.mx/reader034/viewer/2022050520/5fa3a9b85386ec5d48589da3/html5/thumbnails/25.jpg)
Summary
Sea ice month-to-month persistence (decorrelation timescale) of 3-5 months, depending on the reference month
Interesting sea ice reemergence mechanism seen in CCSM3 large ensemble, though it may be too strong
Spring to Fall reemergence is due to SST
Summer to Summer is due to thickness
CCSM4 perfect model studies indicate
Quite good predictability of late summer and Fall sea ice and temperature for initialization in July owing to persistence and reemergence. May offers less good but still significant predictability of late summer and Fall.
most of the memory is carried by the multiyear ice