An Approach to Remove the Systematic Bias from the Storm Surge forecasts in the Venice Lagoon
Abstract
In this work a novel approach is proposed for removing the systematic bias from the storm surge forecast computed by a two-dimensional shallow-water model. The model covers both the Adriatic and Mediterranean seas and provides the forecast at the entrance of the Venice Lagoon. The wind drag coefficient at the water-air interface is treated as a calibration parameter, with a different value for each range of wind velocities and wind directions. This sums up to a total of 16-64 parameters to be calibrated, depending on the chosen resolution. The best set of parameters is determined by means of an optimization procedure, which minimizes the RMS error between measured and modeled water level in Venice for the period 2011-2015. It is shown that a bias is present, for which the peaks of wind velocities provided by the weather forecast are largely underestimated, and that the calibration procedure removes this bias. When the calibrated model is used to reproduce events not included in the calibration dataset, the forecast error is strongly reduced, thus confirming the quality of our procedure. The proposed approach it is not site-specific and could be applied to different situations, such as storm surges caused by intense hurricanes.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2017
- Bibcode:
- 2017AGUFMNH43B..04C
- Keywords:
-
- 1821 Floods;
- HYDROLOGY;
- 1847 Modeling;
- HYDROLOGY;
- 4333 Disaster risk analysis and assessment;
- NATURAL HAZARDS;
- 4341 Early warning systems;
- NATURAL HAZARDS