CLIMAR III S104 ©STFC 2008

Using ship tracking methods to assistin bias adjusting marine observations.

Julian Hill, Simon Maskell andMathew Cole.

CLIMAR III S104 ©STFC 2008 2

Introduction•Brief introduction to climatechange•Why Marine Air Temperature(MAT)?•How can tracking help us?•Example:

– Voyage of the Punjab Senator– Calculated track

•Conclusions/further work

CLIMAR III S104 ©STFC 2008 3

Historic Climate Change

Source: Juckes et al., CP, 2007

1000 1200 1400 1600 1800 2000

TemperatureAnomaly (K)‏

0.6

0.3

0.0

-0.3

-0.6

Rapid rise inmodern era

CLIMAR III S104 ©STFC 2008 4

Why is MAT important?

In Situ SST

Satellite SST(North Atlantic) ‏

MAT1860 1900 1940 1980 2002

0.6

0.4

0.2

0.0

-0.2

-0.4

-0.6

-0.8

Anomaly(K)‏

SST – Sea Surface Temperature

CLIMAR III S104 ©STFC 2008 5

What are the observingplatforms?

•Ships•Drifting buoys•Moored buoys•Platforms

All of these havetheir own biasesand uncertainties

CLIMAR III S104 ©STFC 2008 6

Where are the observations?

Source: ICOADS (SST), Data for June 1998

CLIMAR III S104 ©STFC 2008 7

How can we trackunidentified observing

platforms?•Plaforms: ships, buoys, platforms.•Tracking method developed byQinetiQ.

– Uses Kinematics of the target.– Fast & configurable– Able to track 100,000

simultaneous targets

CLIMAR III S104 ©STFC 2008 8

How will tracks help?

•The shape of the track will help identify theclass of platform•Ship's physical parameters are very unlikelyto change during a voyage.

– Approximate unknown metadata•Possible to use voyage based biasadjustments.

CLIMAR III S104 ©STFC 2008 9

Voyage based biasadjustments

•Observations grouped by voyage– Voyages broken down into regions

(5º lat by 15º long) ‏•Climatology (MOHMAT)‏•Anomalies are compared to annualSHIP anoms. in region (MOHMAT)‏•Based on obs. at 12:00 Local Noon.

CLIMAR III S104 ©STFC 2008 10

322824201612840-4

Climatology for June

Source: mohmat43n http://www.hadobs.orgReference: Rayner et al., 2003, JGR

CLIMAR III S104 ©STFC 2008 11

Anomalies for 19982.0

1.2

0.4

-0.4

-1.2

-2.0Source: MOHMAT43N http://www.hadobs.orgReference: Rayner et al., 2003, JGR

6030

0-30-60

-180 -90 0 90 180

CLIMAR III S104 ©STFC 2008 12

Voyage of the PunjabSenator

CLIMAR III S104 ©STFC 2008 13

Along voyage MAT and SST

0 5 10 15 20 25 30 35Time (UTC day)‏

35302520151050

Temp (ºC) ‏SST

MAT

Bias adjusted MAT

CLIMAR III S104 ©STFC 2008 14

A track.

CLIMAR III S104 ©STFC 2008 15

31302928272625

Along track MAT and SST

SST

MAT

Bias adjustedMAT

Temp (ºC) ‏

27 28 29 30 31Time (UTC day)‏

CLIMAR III S104 ©STFC 2008 16

What could we use the data for?•Point comparisons with satelliteobservations.

•Form part of an estimate of a globalhigh resolution field.

Data fusion (Kalman Filter/Smoother) ‏•Used by multi-variate reanalysisschemes.

produces a self consistent best estimateof the past climate by using climatemodels and variational data assimilation

CLIMAR III S104 ©STFC 2008 17

Conclusions

•Tracking is possible (old news).•Voyage based bias adjustments can beused on tracks.•Further development

– Improve tracker to better understandobserving network.

– Explore other track/voyage based biasadjustments

– Expand period of study (to whole ICOADS?)

CLIMAR III S104 ©STFC 2008 18

Questions?