An investigation of ensemble-based assimilation of satellite altimetry and tide gauge data in storm surge prediction

Abstract

Cyclogenesis and long-fetched winds along the southeastern coast of South America may lead to floods in populated areas, as the Buenos Aires Province, with important economic and social impacts. A numerical model (SMARA) has already been implemented in the region to forecast storm surges. The propagation time of the surge in such extensive and shallow area allows the detection of anomalies based on observations from several hours up to the order of a day prior to the event. Here, we investigate the impact and potential benefit of storm surge level data assimilation into the SMARA model, with the objective of improving the forecast. In the experiments, the surface wind stress from an ensemble prediction system drives a storm surge model ensemble, based on the operational 2-D depth-averaged SMARA model.

Improving the short‐range forecast of storm surges in the southwestern Atlantic continental shelf using 4DEnSRF data assimilation

“…In particular, the total water level envelope (i.e., the SSH anomaly including tides and barometric effects) was used to allow direct comparison with in-situ measurements. Richter et al (2012), Etala et al (2015), and Lago et al (2017) showed that the altimetry data uncertainty in the SWACS depends upon the distance to the shoreline, and ranges between 0.02 and 0.30 m. Therefore, following the results reported by Etala et al (2015), an uncertainty of 0.15 m was assigned to them.…”

“…(2017) showed that the altimetry data uncertainty in the SWACS depends upon the distance to the shoreline, and ranges between 0.02 and 0.30 m. Therefore, following the results reported by Etala et al . (2015), an uncertainty of 0.15 m was assigned to them.…”

“…Time series were detrended, and the averages were subtracted in order to work with water-level anomalies. All the observations were subject to a quality control following the protocol established by IOC (2006) and, as suggested by Dragani et al (2006), Etala et al (2015), and Woodworth (2016), an uncertainty of 0.05 m was assigned to them.…”

“…Remarkably, the only previous exercise in DA in forecasting of the SWACS Etala et al . (2015) was shown to be promising, as the authors demonstrated improvements in the short‐term (6 hr) forecast when using both satellite altimetry and coastal water‐level data (only surge component) along the Argentinean coast.…”

The international workshop on wave hindcasting and forecasting and the coastal hazards symposium

“…Hence the ensemble serves both to forecast the uncertainty as well as improve the central forecast. Etala et al (2015) report advances in storm surge forecasting for the Argentian coast. Using a Local Ensemble Transform Kalman Filter (LETKF), water level from in situ gauges and satellite altimetry is assimilated into a high-resolution 20-member ensemble storm surge model.…”