Copernicus Collaborative Exchange - Altimetry data for tsunami verification


Deep Blue Globe (DBG) was supported by the EUMETSAT Copernicus Collaboration Exchange scheme to work with Altimetry experts at Plymouth Marine Laboratory (PML) on validating and improving the concept of tsunami verification using satellite altimetry data.

Last Updated

14, December 2020

OneWave is a system designed by DBG to verify the existence of a tsunami on open ocean areas by using space assets. The main data source currently used is radar data, provided by altimetry missions. The pilot project in development with LAPAN, the Indonesian Space Agency, is taking advantage of the exceptional features of the missions Sentinel-3A and Sentinel-3B.

The main goals were to:

  • Understand the altimetry data processing workflow and the specificity of Sentinel-3 satellite processing.
  • Comprehend the Sea Level Anomaly (SLA) computation from Level 0 to Level 2.
  • Cross-validate Deep Blue Globe's assumption of the SLA usage for tsunami detection with PML altimetry experts.

Current tsunami warning systems are based on the estimation of numerical simulations. However, the numerical simulations are not accurate enough in the confirmation of the tsunami wave, and some false alerts have been issued in the past. This provoked the incorporation of tidal gauge and buoy measurements to confirm the generation of a tsunami, and provide additional useful information, but there are also issues with this. Not all countries are covered by these instruments and, even when they exist, they are usually too close to the coast to provide warnings very far in advance. In addition, they require expensive maintenance and some emergency services, such as those of Indonesia, have stated the loss of quite a few of these instruments, leading to a faulty Tsunami Warning System.

OneWave will merge the measurement of the signature of a tsunami visible in the sea level observed from space, with the information provided by the existing Tsunami Warning System to verify the occurrence of the event. Access to the satellite data could be implemented using near real-time capabilities, preventing the authorities from alerting the population when the tsunami waves are still travelling through the ocean. DBG, with OneWave, is proposing the first implementation of the Copernicus Collaborative Ground Segment with access to data through local stations.

OneWave High Level System Architecture
Figure 1: OneWave high level system architecture

The full OneWave data process chain is mapped on the OneWave Timemap (Figure 2). The main challenge is to reduce it to the minimum time possible.

OneWave timemap
Figure 2: OneWave timemap

The OneWave Timemap Data Processing step is defined as the time necessary to process the altimetry data from reception at the ground station until the Sea Level Anomaly (SLA), an altimetry Level 2 product, is delivered. This process is currently performed by the Sentinel-3 Payload Data Ground Segment (PDGS), which is responsible for exploitation of instrument data and dissemination of Level 1B and Level 2 products to the user community. A key point for the improvement of data latency performance is, therefore, a correct understanding of the data processing of altimetry products from instrument raw data (Level 0) to Level 2 Sea Level Anomaly. Reducing the data processing time will require a complete analysis of the influence of the different corrections (position, atmospheric and tidal) and a prioritisation of the correction process.

Working closely on-site with the altimetry experts (before COVID) meant they could provide crucial support for understanding of the data processing chain and, as a result, the essential information for the processing from Level 0 to Level 2 concept was defined.

Threshold for parameters for timeliness

Parameters Minimum value Maximum value
Sea Surface Height -130 m 100 m
Sea Level Anomaly -2 m for NRT/STC, -7 m for NTC 2 m for NRT/STC, 7 m for NTC
Standard deviation on the range 0 0.12 + 0.02*SWH m if instrument mode is SAR, 0.2 m otherwise
Nb measurements of range 10 DV
Dry troposphere correction -2.5 m -1.9 m
Dynamical Atmospheric correction -2 m 2 m
Wet troposphere correction -0.5 m -0.001 m
Sea State Bias -0.5 m 0.00 m
Standard deviation of backscatter coefficient 0 0.7 dB if instrument mode is SAR, 1 dB otherwise
Oceanic tide -5 m 5 m
Earth tide -1 m 1 m
Pole tide -15 m 15 m
Altimeter wind speed 0 m/s 30 m/s
Backscattering coefficient 5 dB 28 dB
Significent wave height 0 m 15 m
Filtered ionosphere correction -0.4 m 0.04 m

Each morning PML provided a guide of the tasks to be undertaken and the team met to brief on the objectives of the day and create a plan. At the end of the day a Q&A session was organised to consolidate the knowledge acquired and confirm information and data sources.

Meeting with experts during visit to PML (pre-COVID)
Figure 3:  Meeting with experts during visit to PML (photo taken prior to social distancing requirements)

The visit concluded with a presentation of the OneWave concept to a wider group of scientists at PML. They provided further feedback and there was a discussion of different strategies to improve altimetry data filtering to better estimate Sea Level Anomaly associated to a tsunami event. The support provided by PML was very useful for the validation of the concept of tsunami verification using altimetry data and gaining further understanding of the Sentinel-3 altimetry data processing chain to minimise the processes involved, reducing the data latency of the OneWave system.