Michael Steele Polar Science Center APL University of

Michael Steele Polar Science Center / APL University of Washington Oct 21, 2010 AOMIP @ WHOI The seasonal cycle of Mixed Layer Depth in the Arctic Seas in observations and models Michael Steele Mark Wensnahan Jinlun Zhang Polar Science Center Applied Physics Laboratory University of Washington & NCAR, NOC, NPS, AWI…

Michael. Steele Polar. Science. Center/ APL Universityofof. Washington Oct 21, 2010 AOMIP @ WHOI Phytoplankton Growth = Primary Production depends on… lead light deep MLD: light deficit? shallow MLD: nutrient deficit? nutrients Skyllingstad & Denbo, 1998

Michael. Steele Polar. Science. Center/ APL Universityofof. Washington Def’n: Mixed Layer Depth (m) = 0. 05 MLD 0. 125 0. 2 0. 3 z 0 -10 m Oct 21, 2010 AOMIP @ WHOI 1) Max z “gradient method” …less common: • requires smoothing • more complicated • etc 2) 0 -10 m + “increment method” Sigma-theta (kg/m 3) …more common: • easier!

Michael. Steele Polar. Science. Center/ APL Universityofof. Washington Def’n: Mixed Layer Depth (m) = 0. 05 MLD 0. 125 0. 2 0. 3 z 0 -10 m Oct 21, 2010 AOMIP @ WHOI 1) Max z “gradient method” …less common: • requires smoothing • more complicated • etc 2) 0 -10 m + “increment method” Sigma-theta (kg/m 3) …more common: • easier!

Michael Steele Oct 21, 2010 AOMIP @ WHOI Polar Science Center / APL University of Washington Part 1: World Ocean Database’ 05 (Observed Profiles)

Michael Steele Polar Science Center / APL University of Washington Profile data: obs. by month and 3 -year bins Oct 21, 2010 AOMIP @ WHOI 1980 s & 90 s • CTD • Bottles • Profiling floats • Buoys 1960 s J F M A M J J A S O N D Month 50 250 450 650 850 1050 1250 1450 1650 # obs

Michael Steele Polar Science Center / APL University of Washington Profile data: obs. by month and 3 -year bins Oct 21, 2010 AOMIP @ WHOI 1980 s & 90 s • CTD • Bottles • Profiling floats • Buoys late spring summer J F M A M J J A S O N D Month 50 250 450 650 850 1050 1250 1450 1650 # obs 1960 s

Michael Steele Polar Science Center / APL University of Washington 185, 020 Good profiles Oct 21, 2010 AOMIP @ WHOI # profiles per 100 km EASE grid cell 81, 288 # profiles Rejected profiles 1 5 15 50 200 1000

Michael Steele Polar Science Center / APL University of Washington Reasons for rejection <4 pts in profile unstable profile No data below 10 m 1 5 15 50 200 1000 # rejected No data above 10 m Oct 21, 2010 AOMIP @ WHOI

November 10 September few data in Arctic Ocean June 1000 March 75 Monthly MLD (m) 25 Oct 21, 2010 AOMIP @ WHOI 150 Polar Science Center / APL University of Washington 400 Michael Steele

Michael Steele Polar Science Center / APL University of Washington Regional Means Russia Alaska Oct 21, 2010 AOMIP @ WHOI

Michael Steele Polar Science Center / APL University of Washington Central Arctic Ocean St. Dev / MLD (m) 120 % total 1 4 7 10 13 16% 80 40 Oct 21, 2010 AOMIP @ WHOI 2000 cute lil’ sine wave 1990 0 1980 0. 6 1970 0. 4 1960 ~25% 1950 0. 2 0 J F M A M J J A S O N D Month

Michael Steele Polar Science Center / APL University of Washington Greenland-Iceland-Norwegian Seas St. Dev / MLD (m) 120 Oct 21, 2010 AOMIP @ WHOI % total 1 4 7 10 13 16% 80 2000 40 1990 0 1980 0. 6 1970 0. 4 ~50% ! 1960 1950 0. 2 0 J F M A M J J A S O N D Month

Michael Steele Oct 21, 2010 AOMIP @ WHOI Polar Science Center / APL University of Washington Part 2: PIOMAS Model & …some CCSM

Michael Steele Polar Science Center / APL University of Washington PIOMAS: daily vs. monthly MLD 2000 -2009 output 5 MLD (m) Oct 21, 2010 AOMIP @ WHOI 0 MLD -5 MLD using T, S month minus MLD using T, S day -10 -15 -20 0 MLD -0. 1 -0. 2 MLDmon , MLDday -0. 3 -0. 4 Monthly mean output: -0. 5 J F M A M J J A S O Month N D underestimates MLD by < ~15 m or < ~50%

Michael Steele 0 Oct 21, 2010 AOMIP @ WHOI WOD vs. PIOMAS vs. CCSM Polar Science Center / APL University of Washington Models: monthly means over > 50 y Region 10: central Arctic Ocean MLD (m) 20 WOD 40 CCSM In most regions, models have a larger seasonal cycle amplitude relative to WOD. 60 PIOMAS 80 J F M A M J J Month A S O N D

Michael Steele 0 Models: monthly means over > 50 y Region 10: central Arctic Ocean ? 20 MLD (m) Oct 21, 2010 AOMIP @ WHOI WOD vs. PIOMAS vs. CCSM Polar Science Center / APL University of Washington WOD 40 CCSM In most regions, models have a larger seasonal cycle amplitude relative to WOD. 60 Salt rejection in leads? PIOMAS refs: Smith, Nguyen, Jin 80 J F M A M J J Month A S O N D

Michael Steele Oct 21, 2010 AOMIP @ WHOI WOD vs. PIOMAS vs. CCSM Polar Science Center / APL University of Washington monthly means 0 Region 7: GIN Sea PIOMAS MLD (m) 100 200 CCSM WOD 300 In Nordic Seas, (? !? ) 400 J F M A M J J Month A S O N D

Michael Steele Oct 21, 2010 AOMIP @ WHOI PIOMAS vs. WOD vs. CCSM Polar Science Center / APL University of Washington monthly means 0 Region 7: GIN Sea PIOMAS MLD (m) 100 ~z. E 200 CCSM WOD 300 In Nordic Seas, (? !? ) 300 m! … but it’s all below the euphotic depth z. E 400 J F M A M J J Month A S O N D

Michael Steele Multi-year PIOMAS Polar Science Center / APL University of Washington Oct 21, 2010 AOMIP @ WHOI monthly means 0 10 MLD (m) 20 30 40 50 60 Region 1: Beaufort Sea 1947 1957 1967 1977 1987 1997 2007

Michael Steele Oct 21, 2010 AOMIP @ WHOI Polar Science Center / APL University of Washington Summary 1. MLD stdev / MLD = 25 -50% 2. Monthly model output underestimates MLD <~50% 3. Not shown: MLD stratification below the MLD To do list 1. Optimal increment? 2. More models 3. Interannual variability

Michael Steele Polar Science Center / APL University of Washington ESS + Laptev Seas Oct 21, 2010 AOMIP @ WHOI

Michael Steele Polar Science Center / APL University of Washington Beaufort Sea Oct 21, 2010 AOMIP @ WHOI

Michael Steele Polar Science Center / APL University of Washington 0 PIOMAS (monthly model) vs. WOD (obs. ) Region 10: central Arctic Ocean MLD (m) 80 40 60 80 J F M A M J J Month A S O N D Oct 21, 2010 AOMIP @ WHOI

Michael Steele Polar Science Center / APL University of Washington PIOMAS (monthly model) vs. WOD (obs. ) Region 7: GIN Sea Oct 21, 2010 AOMIP @ WHOI

Michael Steele Polar Science Center / APL University of Washington Data Sets 1. WOD’ 05 ® Original profile data: require T & S (early 1900 s-present) poor winter, central Arctic Ocean coverage 2. PIOMAS ® Jinlun Zhang’s ice-ocean model (daily 2000 -2009) (monthly 1948 -2009) ® excellent coverage everywhere! 3. CCSM & other models… (monthly…) Oct 21, 2010 AOMIP @ WHOI