1. CCl/CLIVAR/JCOMM ETCCDI Workshop De Bilt 13-15 May 2008 Climate Indices from Marine Data Workshop illustrates that most studies focus on land T and P: need for marine indices Elizabeth Kent - National Oceanography Centre, Southampton Val Swail - Environment Canada, Toronto Scott Woodruff - NOAA Earth System Research Laboratory, Boulder David Parker - Met Office, Exeter

2. FOCI ANTICIPATED FOR MARINE INDICES detection and attribution of climate change detection and attribution of climate change impact on marine industries (fishing, shipping, oil and gas production, tourism) impact on marine industries (fishing, shipping, oil and gas production, tourism) sea-level change sea-level change marine hazards (extreme winds and waves, harmful algal blooms, pollution) marine hazards (extreme winds and waves, harmful algal blooms, pollution) changes in hydrological cycle changes in hydrological cycle changes in ocean circulation changes in ocean circulation changes in sea ice and ice bergs changes in sea ice and ice bergs effects on coastal communities effects on coastal communities ocean acidification ocean acidification

3. MARINE DATA SOURCES AND PROGRAMS ICOADS – ships (from 1662), moored and drifting buoys ICOADS – ships (from 1662), moored and drifting buoys World Ocean Database (WOD) World Ocean Database (WOD) Global Digital Sea Ice Data Bank (GDSIDB) Global Digital Sea Ice Data Bank (GDSIDB) Permanent Service for Mean Sea Level (PSMSL) Permanent Service for Mean Sea Level (PSMSL) Derived data sets – HadISST, HadSLP, HadGOA (www.hadobs.org ) Derived data sets – HadISST, HadSLP, HadGOA (www.hadobs.org )www.hadobs.org Satellite – SST, wind, wave, ice, sea level Satellite – SST, wind, wave, ice, sea level Reanalyses Reanalyses Ship Observations Team Ship Observations Team Data Buoy Cooperation Panel Data Buoy Cooperation Panel Argo Argo Ocean Sites Ocean Sites Global Sea Level Observing System Global Sea Level Observing System International Ocean Carbon Coordination Project International Ocean Carbon Coordination Project Global Temperature and Salinity Profile Program Global Temperature and Salinity Profile Program JCOMMOPS (www.jcommops.org ) JCOMMOPS (www.jcommops.org )www.jcommops.org

4. Integrated Ocean Observing System Ship observationsSea level Moorings Argo Drifting buoy ASAP

5. Annual numbers of marine reports in ICOADS, stratified by platform type for 1936 to 2005 (Woodruff et al. 2008)

6. ICOADS Sampling by Parameter

7. Perspective on Historical Data JCOMM Expert Team on Marine Climatology links with International Comprehensive Ocean- Atmosphere Data Set (ICOADS)JCOMM Expert Team on Marine Climatology links with International Comprehensive Ocean- Atmosphere Data Set (ICOADS) Recovery of more data & metadata: key to improving past climatologies, e.g. Recovery of Logbooks And International Marine data (RECLAIM)Recovery of more data & metadata: key to improving past climatologies, e.g. Recovery of Logbooks And International Marine data (RECLAIM)

8. Improvements to ICOADS Many new data sources added to ICOADS focused on data sparse regions and periods.

9. ICOADS Improvements in 1930s

10. The potential for marine indices OperationalResources required Research required Data required Large scale pressure (e.g. NAO, PNA) Temperature indicesAtlantic Meridional Circulation Max & min temperatures Large scale temperature (e.g ENSO) Marine winds and pressures CurrentsWind gusts Sea Ice parametersWavesPolar lows.Storm surges Ocean heat contentSea levelHydrographic time series (e.g. ICES) Extremes Salinity measuresFisheries information & biology Precipitation Ocean transports and water mass properties Ocean chemistry (e.g. dissolved oxygen) Ph/Ocean Acidification HurricanesDeep convection Clouds, humidity.

11. Temperature trend over 1901-2003

12. 12 Hadley Centre for Climate Prediction and Research Monthly Surface Temperature Sept. 2006 PercentilesAnomalies Tropical Central and EastPacific SST Anomalies, 1850-2005

13. Global Wave Climatology Atlas S. Caires, G. Komen, A. Sterl, V. Swail www.knmi.nl/waveatlas

14. Number of gridpoints Year of changepoint Monthly mean wind speed 1966 Monthly mean sig. wave height Number of gridpoints Year of changepoint 1966 Number of gridpoints of a significant changepoint in the indicated year Wind speed – locations of changepoint in Nov. 1966Sig. wave height – location of changepoint in Nov. 1966 Grid-boxes of significant changepoint are shown in black xx MSC50 Wind and Wave reanalysis 1954-2007

15. 15 Hadley Centre for Climate Prediction and Research Annual sea-ice extent changes, 1973-2006 (updated from IPCC, 2001) Retreating until late1990s. Little retreat 1998-2003 Antarctic sea-ice Not declining since 1976 Arctic sea ice

16. © Crown copyright Background: ocean heat uptake Heat content for Anomaly for the upper 300m See Gregory et al. [2004]

17. © Crown copyright Changes in mean T and isotherm depths Mean 14C isotherm depth Mean T above 14C isotherm 1985-2004 minus 1961-1980  Deepening of isotherms in N. Atlantic associated with change in phase of NAO.  Large areas of slight shoaling and smaller areas of large deepening  Wide-spread warming signal.  Less prone to aliasing from changes in ocean circulation than z-levels.  Greater insight into underlying physical mechanisms

20. Enabling Mechanisms ICOADS - Critical and critically under-resourced Proposed new initiative for value-added ICOADS (QC, bias corrections, etc.) JCOMM Expert Teams Wind Waves and Storm Surges Sea Ice Marine Climatology Task Team on the Marine-meteorological and Oceanographic Summaries (TT-MOCS) Task Team on Delayed-Mode VOS (TT-DMVOS) Engage expertise within the CLIMAR community to assist in the development and production of marine indices (marineclimatology.net) marineclimatology.net Liaise with other groups interested in marine indices such as the AOPC and OOPC

21. Prediction of changes in the ocean is vital for prediction of extremes over the land. Marine focus has been on the creation of high quality gridded data sets with uncertainty estimates Many similar issues to land inhomogeneity sampling Many variables of interest – and multivariate analyses Difficult to calculate high-percentile extremes due to data uncertainty.