1. ESA’s DUE eSurge Project: Improving storm surge modelling with advanced satellite data products. Living Planet Symposium, Edinburgh, 10 th September 2013. Phillip Harwood, CGI, Phillip.Harwood@cgi.comPhillip.Harwood@cgi.com Paolo Cipollini, Helen Snaith, NOC Jacob Høyer, Weiwei Fu, Kristine Madsen, DMI Rory Scarrott, Declan Dunne, CMRC Ad Stoffelen, Tilly Driesenaar, KNMI Craig Donlon, ESA

2. A storm surge is an unexpectedly high sea level caused by an extreme weather event. The flooding resulting from a storm depends on the sum of the mean sea level, the tide, the storm surge, and the wave heights.

3. Some (in)famous storm surges: North Sea floods (1953): 2100 killed in worst natural disaster in modern Europe. Hurricane Katrina (2005): 1833 killed, $81bn damage, severe devastation in New Orleans Cyclone Sidr (2007): One of Bangladesh’s worst disasters, with at least 3500 deaths estimated. Cyclone Nargis (2008): At least 138,000 fatalities in Myanmar’s worst natural disaster. Hurricane Sandy (2012): $65bn damage in the USA alone. Also 11 killed and 15000 homes destroyed in Cuba.

4. It is not just the big storms that count …. … Outer Hebrides, 11-12 January 2005. 2m storm surge 5 lives lost €20m infrastructure damage Loss of electricity Broken transport links Loss of communications

5. However the modeling and forecasting of surges could be improved by use of newer types of satellite products. METOP-A image of Sandy from Eumetsat Satellite data already play a critical role in storm forecasting. Coastal altimetry. High resolution and near-coast winds from scatterometry. These can be used: For validating model results after an event. As inputs to NRT models during an event.

6. Replace in-situ data where infrastructure is lacking Ensemble pruning Post event model reassessment Better input data (esp. wind fields)

7. eSurge is the ESA Storm Surge Demonstration Project, funded through the Data User Element (DUE) programme. Its mission is: To support storm surge systems, their services, engineers and scientists by facilitating the widespread user uptake and application of advanced information products from ESA and other Earth Observation missions.

8. To do this, the project will: Make relevant data freely available through our web portal (www.storm-surge.info). Provide new types of data, especially coastal altimetry. Perform experiments to show the value of the data. Build a storm surge community through a series of workshops and training courses.

9. Provide data and analysis for the specific conditions applying in the Adriatic Sea and Venetian lagoon. This data is available through the main eSurge portal. See www.esurge-venice.eu

10. Making data available is only part of the problem. We also need to convince others to use it, by demonstrating its value.

11. To demonstrate the usefulness of data, DMI have looked at assimilating it in the area of the North/Baltic seas. The first step is to validate the data against tide gauges. With along-track smoothing and spike removal, data can be used within 1-3 km of the coast. ALES generally performs better than CoastAlt retracker. Use of the data closer to the coast requires individual processing of the data.

12. Example of blended water level product: surge component (8/1/2005). A blended tide gauge and altimetry product is used: 2D field of sea surface height (5km, hourly). Tide gauge data used to derive statistical model coefficients. Along track altimetry expended to yield better temporal resolution.

13. Incorporate product into DMI’s surge model: HIROBM-BOOS model used. Hydrostatic 3D circulation model. Forced using regional climate model HIRHAM Compare 2 runs: Free model run Assimilation of blended product every 24hrs using ensemble optimal interpolation (EnOI) method. Look at 5-9 January 2005 (strong storm surge from winter storm Erwin/Gudrun)

16. Assimilating the blended product improves model performance, by comparison to independent tide gauge data. Model-data misfit reduced by 5-10% RMSE reduced by 10-15 cm Next steps: Further work on blending product Use latest coastal altimetry Also include JASON Assimilate this into models

17. The Indian coastline, and especially the Bay of Bengal, is extremely vulnerable to storm surges. Since 6 th May 2013, ESA has been making NRT Cryosat SAR data available around this area. Cryosat Mode Mask 3.4 (Since October 2012 ) Altimetry with Cryosat SAR

18. Cyclone Mahasen Made landfall in Bangladesh on 16 th May 2013. Significant storm surge was forecast, but landfall was at low tide so the worst case flooding was avoided.

20. 10 May 04:24 11 May 16:08 17 May 03:27 16 May 04:17 14 May 04:19 13 May 16:05 15 May 16:03 12 May 04:21 17 May 16:00 17 May 03:27

21. 16 May 04:17 14 May 04:19 15 May 16:03

22. The coastal altimetry processor performs well, all the way to the coastline … … but we do not get a direct image of the surge.

23. Next steps: Be prepared for the next cyclone season. Data will be made available through our website during an event (eSurge Live). Working with modellers to add this data to their models. Include altimetry from other satellites (JASON).

24. Conclusions Altimetry is a powerful tool for improving storm surge modelling and forecasting. Even standard (non-coastal) altimetry is useful for calibrating models and ensemble pruning, but coastal altimetry adds an extra dimension. This has practical benefits: better forecasts, longer warning period, fewer false alarms. Excellent opportunity to demonstrate the practical value of coastal altimetry. There is an issue with sampling, we rarely get measurements where we want them.

25. For more information: Session P-02 Tuesday: Coastal Zones and Wind Improving Storm Surge Hindcast: Assimilation of Satellite Altimetry and Tide Gauge data in the Baltic Sea (W. Fu, J. Høyer, L. Jonasson.) Validation and use of Operational Coastal Satellite Altimetry Observations for Storm Surges (J. Høyer, K. Madsen, P. Cipollini, L. West.) Session P-08 Thursday: Microwaves; Wind, Waves and Ocean Features The ESA eSurge-Venice Project: Altimeter and Scatterometer Data to Improve the Storm Surge Forecasting (S. Zecchetto, F. De Biasio, G. Umgiesser, M. Bajo, D. Bellafiore, S. Vignudelli, A. Papa, C. Donlon.) Scatterometer Wind Services in Europe (A. Stoffelen, A. Verhoef, J. Verspeek, J. Vogelzang, T. Driesenaar, A. Bentamy.)

26. ISMAR, Venice 18-20 November 2013.  To explore how earth observation data can contribute to our understanding of storm surges.  To discuss how the community can best exploit current and future satellite data.  To foster networks and collaborations between researchers, modellers and data providers.

28. www.storm-surge.info