Seasonal Variability of the Global Spectral Wind Wave Climate

Abstract

Our understanding of the seasonal variability in the global wind wave field is revisited here using a novel analysis that resolves the directional wave spectra. Empirical orthogonal function analysis was applied to modeled wave spectral data from a WAVEWATCH III hindcast across a sparse global grid to identify the main patterns of the climatological wave spectral variability at each grid point. Prior methods have focused on the variability of two modes of the wave field—locally generated sea and the primary swell component. Our results also consider additional wave modes at each location, enabling us to track the passage of the less dominant swell modes. Consistent with existing climatological knowledge, the main modes of wave spectral variability at high latitudes are related to eastward propagating waves that disperse equatorward as swell following great circle paths. However, despite being the less energetic mode, the Northern Hemisphere generated swell is found to propagate into the Southern Hemisphere further than the more energetic Southern Hemisphere swell which propagates northward. In the equatorial zone, a complex multimodal wave climate is found, with the spectra variability modulated by remotely generated swell and higher‐frequency waves associated with the prevailing winds. The evolution of these patterns throughout the year is clearly depicted. Overall, our approach captures a more complete picture of the seasonal variability of the global wave field, by accounting for all the wave modes observed in the spectra at each location together with their temporal variability.

Influence of the Pacific‐South American Modes on the Global Spectral Wind‐Wave Climate

“…The approach outlined in Echevarria et al (2019), and briefly described in section 2.2, was implemented in the present study to investigate the patterns of interannual wave spectral variability using daily averaged directional wave spectra from 1979 to 2019, taken from the CAWCR global wave hindcast data set. The time evolution of these patterns of spectral variability at each grid point was then compared to various climate mode indices to better understand the potential drivers of interannual variations in the global spectral wave climate.…”

“…Considering the mean swell direction in the presence of opposing swell waves will not faithfully represent any of the wave components: For example, having two swell systems propagating to the southeast and northeast could yield an average direction to the east. This case may arise in the equatorial Pacific which receives swell waves from high‐latitudes of both hemispheres (Echevarria et al, 2019; Semedo et al, 2011; Young, 1999). Likewise, the estimation of other wave parameters will be affected.…”

“…On the other hand, a directional wave spectrum describes how the waves' energy is distributed across frequencies and directions, hence providing the most complete description of the wave climate in a certain region, since it characterizes all the wave modes that may be present. Several authors (e.g., Echevarria et al, 2019; Hegermiller et al, 2017; Mortlock & Goodwin, 2015; Portilla‐Yandún et al, 2016; Portilla‐Yandún, 2018) outlined potential problems that could arise when describing the wave climate with integrated parameters. For example, defining the peak direction in cases where there are two independent swell modes will only consider the most energetic one and disregard the other.…”

“…EOF (e.g., Preisendorfer & Mobley, 1988) is commonly used in climate sciences to find dominant spatial patterns of variability of a certain climate variable, how this pattern changes through time, and the percentage of the total variance explained by these “modes.” Most commonly, the EOF analysis is performed on spatially gridded data of a scalar quantity that fluctuates in time to find the geographic patterns that depict the main modes of variability of that quantity (e.g., Deser et al, 2010; Gulev & Grigorieva, 2006; for wave‐related studies, see Hemer et al, 2009 or Semedo et al, 2011). Here, following Echevarria et al (2019) and Shimura and Mori (2019), we employ a slightly different approach where the main modes of variability of the wave spectral density in the frequency/direction domain were obtained for each one of the spectral grid points of the model. These patterns represent the main modes of wave variability in the spectral domain, and the methodology allows us to analyze their temporal variability independently.…”

Ocean Swell Comparisons Between Sentinel‐1 and WAVEWATCH III Around Australia

“…The mean spectra of WW3 and Sentinel-1 ( Figure 11) have similar shapes and directionalities throughout the study region. Many of the uniform wave systems expected in this region (Echevarria et al, 2019), such as the Southern and Indian Ocean swell systems, are also well-captured by both Sentinel-1 observations and the model. The mean spectra in the SO, as observed by Sentinel-1A, mainly consist of swell waves with a peak frequency of ∼0.07 Hz (∼14 s period) propagating eastward and can be qualitatively observed to follow great circle paths.…”