A Procedure for Operational Use of Wave Hindcasts to Identify Landfall of Heavy Swell

Abstract

he wave pattern on the Brazilian coastline is composed of both wind waves and swell. The wave systems (WSs), extracted from the spectra near the coast produced by numerical wave models, reveal the occasional presence of intense swells, with small significant wave height (HS) and large average period (Ta). This kind of event has nearly no effect over deep water, but its landfall can be accompanied by inundation, mainly when coupled with favorable tides and storm surge. Since these events are not clearly evident in the bulk parameters, this study proposes a methodology (i) to identify intense swells simulated by a coarse grid resolution wave modeling system (CWS) and (ii) to evaluate their importance. In this methodology, monitoring sites are defined along a 100-m isobath contouring the Brazilian coast, where the CWS hindcasts the spectra for a 31-yr period, from 1979 to 2009, obtained by the WAVEWATCH wave model. The spectra are partitioned into WSs, which are used to build cumulative distribution tables (CDTs) for each site. The variable used in the CDT is the flux of energy per unit length perpendicular to the wave propagation (PW), which contains in its definition both HS and Ta. The direction of propagation of a WS is used to compute the components of PW parallel and perpendicular to the coast. From the CDT of the perpendicular component of PW, the percentile of an incoming WS can be found and its intensity ranked. To illustrate the feasibility of this proposal, the method is used to find the 50 most powerful distantly generated swells for two sites: one on the northern and another on the southern Brazilian coast. In addition, the method is applied in two case studies, both accompanied by coastal flooding and erosion: one represents a very powerful WS arriving at the northern coast and the other a less energetic event occurring on the southeastern coast. The analysis of bulk parameters fails to identify the second case as potentially destructive, while the proposed methodology clearly gives some indication.