Environmental Geodesy Lecture 12 (April 18, 2011): Sea Level Changes - The societal challenge - The scientific challenge - The observing system - Summary

“Sea-level hazard” is one of many hazards challenging society: - includes waves/currents/tides, storm surges, tsunamis, saltification; - is modified by slow changes of mean Local Sea Level (LSL). Do we have a reasonable understanding of the hazards and its time- variability? Response is a 'insurance problem'; how much do we want to spend to mitigate a risk, prepare for a hazard that could lead to a disaster, and disaster response.

About 10,000 years ago, Local Sea Level (LSL) became stable after a long phase of rapid variations. Human settlements could move into the coastal zone. Humanity learned about the perils of the sea... Even in times of relatively stable sea level, coast lines changed and humans had to adapt and to move settlements...

Storm surges wiped out whole settlements and humans repeatedly left the coast zones. Tsunamis contributed to large disasters that shocked society and impacted art, philosophy and culture. And humans learned to built coastal protections to avoid moving out of dangerous zones as the only adaptation. But when the coastal protections failed, for example, during the North Sea storm surge in 1953, disaster was unavoidable. Over the last five decades, we have developed amazing engineering skills in coastal protection. We can maintain countries that are for a large part below current sea level. We built new settlements in areas prone to flooding, particularly if the protections fail. Today, we are building magnificent barriers...

But recent disasters have shown limitations of coastal protection: - Hurricane Katrina exemplified the urban disaster. - The 2004 Sumatra tsunami exemplified large-scale hazards. - Venice exercises adaptation to a slowly developing disaster

Annual averages of the global mean sea level (mm). Red curve: reconstructed sea level since 1870 (Church and White, 2006); Blue curve: coastal tide gauge measurements (Holgate and Woodworth, 2004) Black curve: satellite altimetry (Leuliette et al., 2004). Global Sea Level Rise IPCC, AR4

Budget of the global mean sea level change Blue: 1961 to 2003 Brown: 1993 to 2003 Bars represent the 90% error range. Global Sea Level Rise IPCC, AR4

< cm/century < 10 cm/century up to 50 cm/century Basic societal and scientific assumption: Sea level changes very slowly and coast lines are mostly stable. Bindoff et al. (2007) - IPCC AR4

Hansen et al., 2009 Paleo-records tell a different story... Paleo-records of global sea level - demonstrate large changes: roughly 1.5 m/century during warming and peak rates of 4 m/century; - about 6,000 years ago, sea level became very stable; - humans could settle in the coastal zone and benefit from the many advantages of this location; - this stable phase made us (falsely) believe that ‘sea level doesn’t change very much.’

Harrison & Stainforth (2009): “Atmospheric CO 2 concentrations are now higher than they have been for at least the past 650,000 years...” The Earth system may have passed some threshold values. Humanity has re-engineered the planet: The recent past and the presence have only limited value for assessing the range of plausible futures. We can not exclude that sea level becomes unstable again.

- Significant sea level rise would trigger extreme disasters; - There is trade-off between the costs for adaptation today and disasters tomorrow; - Coastal Defense is very expensive and there is a risk of overspending; - Adaptation may require relocation of settlements and infrastructure (airports, highways, pipelines,...) (airports, highways, pipelines,...) The Societal Challenge The UN Development Program, 2008: 332 million People live in low-lying coastal zones Damage caused by a single disaster could exceed $100 billion World Bank, 2008: * 1 m of sea level rise = 2% of GDP in East Asia

What do policy and decision makers ask science for? - Local sea level (LSL) rise projections for the next 100 to 200 years, particularly high end; - reliable uncertainties; - full range of “plausible LSL trajectories” with probability density function (PDF); Meeting the Societal Challenges

Science of Sea Level Changes

The Science Challenge LSL changes are caused by many Earth system processes acting on a wide range of spatial and temporal scales that change the sea surface height or the land surface height or both. Local Sea Level (LSL): vertical distance between sea surface and land surface.

The Science Challenge We understand the forcing processes but we have no Earth system model to predict LSL variations, taking into account, for example: - carbon emissions; - impact of re-engineering on climate; - ice sheets' response to global warming.

Local Sea Level (LSL) = high-frequency part + low-frequency part High-frequency part of the LSL equation: h hf (x,t) = h waves + h tidal + h atmos + h seiches + h tsunami Important for projections of maximum flood levels High-frequency part of the LSL equation: h hf (x,t) = h waves + h tidal + h atmos + h seiches + h tsunami Important for projections of maximum flood levels High-frequency LSL variations are the result of local and regional processes, including waves, tides, storm surges, seiches, tsunamis, earthquakes,... The Science Challenge

Local Sea Level (LSL) = high-frequency part + low-frequency part Low-frequency part of the LSL equation: h lf (x,t) = h steric + h circ + h atmos + h cryosphere + h land-water + h post-mass + h vertical-land + h basin-changes Important for projections of mean LSL changes Low-frequency part of the LSL equation: h lf (x,t) = h steric + h circ + h atmos + h cryosphere + h land-water + h post-mass + h vertical-land + h basin-changes Important for projections of mean LSL changes Low-frequency LSL variations are the result of local, regional, and global processes including temperature and salinity changes, circulation in ocean and atmosphere, past and present mass exchange with cryosphere and terrestrial water storage, geodynamics,... All mass movements on the Earth surface: - change the geoid (gravity field); - displace the ocean bottom and land surface vertically; - redistribute water mass in the oceans. All mass movements on the Earth surface: - change the geoid (gravity field); - displace the ocean bottom and land surface vertically; - redistribute water mass in the oceans. The Science Challenge

Load on ocean areas Loads on land areas LSL change Science of Sea Level Changes

All mass movements on the Earth surface: - change the geoid (gravity field); - displace the ocean bottom and land surface vertically; - redistribute water mass in the oceans. LSL changes caused by redistribution of mass in the global water cycle are the difference between the local changes in geoid and ocean bottom height plus an offset for the mass change in the ocean. We can compute these LSL changes using the integral equation that links the mass redistribution over the whole surface of the Earth (including the oceans) to LSL changes. All mass movements on the Earth surface: - change the geoid (gravity field); - displace the ocean bottom and land surface vertically; - redistribute water mass in the oceans. LSL changes caused by redistribution of mass in the global water cycle are the difference between the local changes in geoid and ocean bottom height plus an offset for the mass change in the ocean. We can compute these LSL changes using the integral equation that links the mass redistribution over the whole surface of the Earth (including the oceans) to LSL changes. Science of Sea Level Changes

The Science Challenge... but we have no Earth system model to predict LSL variations, taking into account, for example: - carbon emissions; - impact of re-engineering on climate; - ice sheets' response to global warming. We understand the forcing processes.... LSL changes are caused by many Earth system processes acting on a wide range of spatial and temporal scales that change the sea surface height or the land surface height or both. Local Sea Level (LSL): vertical distance between sea surface and land surface. LSL = high-frequency part + low-frequency part High-frequency part: result of local and regional processes, including waves, tides, storm surges, seiches, tsunamis, earthquakes,... Low-frequency LSL variations are the result of local, regional, and global processes including temperature and salinity changes, circulation in ocean and atmosphere, past and present mass exchange with cryosphere and terrestrial water storage, geodynamics,... All mass movements on the Earth surface: - change the geoid (gravity field); - displace the ocean bottom and land surface vertically; - redistribute water mass in the oceans.

Science Support for Societal Decisions

“Localizing” global projections Science Support for Societal Decisions

“Best Practice” for LSL Assessments: - Local approach: sum of contributions from “most relevant” processes - Individual treatment of contributions, depending on predictability: extrapolation, model predictions, scenarios

Science Support for Societal Decisions “Best Practice” for LSL Assessments: - Local approach: sum of contributions from “most relevant” processes - Individual treatment of contributions, depending on predictability: extrapolation, model predictions, scenarios Current Situation: - Many assessments in the last 10 years; - Large range of plausible trajectories of future LSL; - Large uncertainties particularly for ice sheet contribution; - Due to new observations/knowledge out-dated within a few years; - Example: accelerated ice mass loss of Greenland and Antarctica discovered in 2009.

Reaction of Decision Makers: “‘We can’t make multi-billon dollar decisions based on the hypothetical’ says Rohit Aggarwala, the city’s director of long-term planning and sustainability.” Wall Street Journal, September 11, 2009, “New York City Braces for Risk of Higher Seas” Science Support for Societal Decisions “Best Practice” for LSL Assessments: - Local approach: sum of contributions from “most relevant” processes - Individual treatment of contributions, depending on predictability: extrapolation, model predictions, scenarios Current Situation: - Many assessments in the last 10 years; - Large range of plausible trajectories of future LSL; - Large uncertainties particularly for ice sheet contribution; - Due to new observations/knowledge out-dated within a few years; - Example: accelerated ice mass loss of Greenland and Antarctica discovered in Sea level rise is an uncertain, slowly developing hazard that has its impact mainly during infrequent extreme events.

How do we prepare for the impacts of LSL rise? Current 'best practice' for LSL assessments does not provide actionable information. Deterministic approach does not work since we lack the capability to predict the future within reasonable uncertainties.

An Alternative Approach LSL rise is an “insurance problem”: Preparation for less likely, extreme LSL rise: - Extreme LSL rise (> 1 m/century) can not be excluded; - comparable to meteorite impact: Very unlikely but disastrous if it happens - “Early warning” when an extreme LSL rise is likely to occur. Knowing the LSL rise hazard: - Global map of LSL rise hazards, including statistics; - Statistics based on long-term paleo-record, not just recent past and model studies. - "Early-warning" for rapid LSL rise (smoke detector) Building resilience to LSL rise hazards: How much do we want to spend on monitoring the hazard and on "early warning"? How much do we want to spend on reduction of vulnerability and risk?

Tide Gauges GRACE Laser Radar Satellite Altimetry GNSS Reflections Zufada & Zavorotny (2004) “Early Warning” - LSL Forecasting - Planning - Adaptation What is needed to enable LSL forecasting: - Maintain/improve the monitoring system (comparable to “Meteorite Watch”); - Community effort to develop the models needed for decadal predictions (in particular, mass changes in ice sheets, ocean and atmospheric circulation, loading); - Start forecasting as soon as possible to get a basis for the assessment of the forecasting capabilities. What is needed to increase resilience with respect to LSL hazards: - transition from deterministic predictions to statistical insurance approach; - understanding the possible trajectories and know the worst case; - focus on what is/can be resilient to large LSL changes; - planning for the worst case; - life cycle analysis for coastal projects taking into account the worst case. Towards Forecasting of Local Sea Level Changes: - Monitor the main reservoirs in the global water cycle; - Develop models that can predict reservoir changes on decadal time scales; - Develop models that can relate reservoir changes to LSL; - aim for five to ten years and more...