This is an educational outreach product intended for a general audience. Feedback is welcome.
Winter surf, North Shore, Oahu. Photo: Steve Businger,
Dept. of Atmospheric Sciences, University of Hawaii
Various entities, such as lifeguards and
Surf News Network, report daily visual observations of the surf.
Historically, surfers have often shared their observations casually in conversation. Inherently, the
observations are of coarse resolution.
Beginning in 1968 for the north
shore and 1972 for the south shore of Oahu, Mr. Larry Goddard took on the task of writing down the daily observations. In addition to his own surf monitoring,
he also was given reports from experienced watermen such as Peter Cole, Randy Rarick, Bernie Baker
and Albert Benson. The reporting sites were most often the areas along the coast of highest breakers,
such as Sunset Beach on the north shore and Ala Moana on the south side. Only one value was recorded per
day for each shore, representing the time of day with the highest reported breakers. Each value represented the
common though less frequent larger sets
of breakers, nominally referred to as the H1/10, or the average of the 10% highest waves.
In 1987, Mr. Goddard moved to the mainland and his series ended. Fortunately in 1987,
Mr. Patrick Caldwell independently began a database of daily surf reports using a similar scheme (H1/10, highest
reporting spots, highest time of day). In the late 1990s, Mr. Goddard returned to Hawaii and shared his historic
set with Mr. Caldwell. The combined database of visual surf observations was the target for a research paper,
The Validity of North Shore, Oahu, Surf Observations ,
which showed the historic readings in the colloquial Hawaii scale have been consistent in time. The uncertainty
averages near 15% with increasing observational error with increasing surf size.
The next task was to
translate the observations from Hawaii scale to trough-to-crest heights
(peak face), which was published with other research in the
Journal of Coastal Research .
The observations recorded in Hawaii scale were converted to peak face using a factor of 2 over the
full range (0-40); for example, 10 Hawaii scale equals 20 peak face (feet). For days
when waves are so large that Sunset Beach is closed out (H1/10 > 15 Hawaii scale, or 30' peak face),
the heights represent zones of high refraction on outer reefs.
The visual surf observation database is approaching 50 years in length and has been the basis for various
surf research projects. This set of daily visual surf
observations are referred to as the Goddard-Caldwell Database.
Nearshore wave buoys in the Hawaiian Islands have been maintained and made
available by the
Pacific Islands Ocean Observing System (PacIOOS) and the
Coastal Data Information Program
(CDIP) in partnership with the Dept. of Oceanography, University of Hawaii. The sampling rate is 1 Hz, and the acquisition time is 20 minutes.
Significant wave height (Hs),
peak wave
period (Tp) and peak wave direction (Dp) are reported on 30-minute cycles.
Weather makes the wind; wind makes the waves. The north to west exposures of Oahu receive swell generated primarily in the
mid latitudes of the North Pacific, and on rare occasions from tropical sources.
Ocean surface winds are strongest near intense surface low pressure areas, which reach a maximum of intensity in the
northern hemisphere winter. Subsequently, the
North Shore annual cycle of surf follows suit. The large range between the minimum, average,
and maximum on any given day draws attention
to the need for up-to-date, accurate surf forecasts for safe planning.
In the general discussion of the
NOAA Oahu Collaborative Surf Forecast, the surf for a given forecast period is compared to averages for context.
North Shore surf statistics are defined for select time periods. The mean is the most common statistic to denote average conditions. The median and mode are other ways to define the most common heights that occur over a given period.
These parameters tend to be lower than the mean for North Shore surf, which leans higher due to the influence in the calculation
of the mean from select extreme days being well beyond 2 standard deviations above the mean. The difference
between the median or mode and the mean gets larger for peak season.
Most commonly in the Collaborative Forecast, the heights for a given day are referred to the average of the active season
(September to May), which if averaged over the mean, median and mode is nominally 12' peak face for the H1/10 breakers in zones of high refraction.
The standard deviation (St. Dev.) denotes how much variation from the mean can occur
during the given time period. As seen in the annual cycle between the maximum and minimum for any calendar
day, the variability in range of heights is greatest during the peak month (January) and season (winter). This helps fine tune the interpretation of the active season average (September to May), which is 12' +/- 8' peak face. In other words, during the active season, the heights fall mostly within 4-20' peak face for the H1/10 breakers.
The Waimea Buoy
provides an additional perspective on defining the averages. Monthly means
of significant wave height (Hs) and peak wave period (Tp) show an annual cycle centered on winter months. This
calculation was performed on all available data using the "Averages: monthly" option in the
CDIP Summaries: Min Max Mean. The average Hs and Tp over the active season (September to May) are 6' at 12 second intervals.
From the Hs and Tp, estimates of surf heights, referred to as the surf compute, can be made based on
an empirical method. This is the same formula
used daily in the Collaborative Surf Forecast, which predicts a deep water Hs and Tp, from which breaker heights
are computed.
The mean H1/10 breaker heights computed from the monthly mean Hs and Tp compare well with the
independently calculated monthly means from the visual surf observations (Goddard-Caldwell Database).
The buoy surf compute is slightly lower than the visual observations for select months. This is due
to the influence of shorter period wind waves (trades and konas) that are included in the monthly mean Tp
calculation, giving a low bias to Tp compared to if the wind waves were filtered out. For the active season average, the buoy average (6'@12s) results in a surf compute
11' peak face, which also has a slight low bias relative to the estimate from the visual observations, for reasons described above.
As an additional test, the monthly means of the visual observations were performed over differing time periods. It shows very
little deviation, giving confidence to the statistics.
Wave direction from the Goddard-Caldwell data shows how the
dominant direction varies through the active season (September to May). More northerly components are most
common during fall and spring. During winter, the west to northwest components happen more often. This is related
to the jet stream track, which shifts equatorward in winter. In turn,
the lower-latitude, surface low pressure systems aim more
westerly swell component at Hawaii.
Southerly exposures of Oahu typically receive swell that is generated primarily in
the mid latitudes of the South Pacific.
On rare occasions, swell reaches Oahu from north Pacific tropical cyclones
south of Hawaii or from more distant tropical cyclones in the southern hemisphere.
The mid latitudes have the strongest, broadest surface low pressure
systems within fall through spring of the respective hemisphere, which centers the
South Shore annual cycle of surf around the
austral winter, or summer in Hawaii. Surf is much lower relative to the North Shore, because the
South Shore sources are remote. The southerly swell loses size as it traverses
the vast distance beyond 3000 nm to Hawaii.
South Shore surf statistics
show June as the peak month and place spring closely behind summer
for the peak season. The southerly surf maximum
in late spring to early summer is likely related to less Antarctic ice sheet
during this time (austral fall to early winter), which allows
more open ocean for wave generation.
The median and mode share insight into most common conditions. Most days during
the active season (March to October) were reported in H1/10
as either 2 or 3 Hawaii scale,
or 4' or 6' after converting to peak face. The coarse resolution hinders
accuracy. The mean is nearer to the median and mode for the South Shore relative
to the North Shore, since the percentage of days of large events is less
on the South Shore.
In the Collaborative Forecast discussion, heights on southerly
exposures are often referred to the
average of the active season (March to October). Averaging over the
mean, median and mode gives 5.5' peak face H1/10 breakers.
Keep in mind this refers to high refractive zones, such
as Ala Moana and Diamond Head. Elsewhere, the average for
less refractive zones, such as the Waikiki area,
would be lower (~ 4').
The standard deviation for the South Shore varies much less than the
North Shore, since the
range of breakers is much lower on the South Shore. The active
season (March to October) can better be described as most commonly
5.5' +/- 2.3', in other words, most of the time H1/10 surf heights
fall within 3'-8' peak face.
Since the available PacIOOS/CDIP nearshore buoys off the southern
shores of the Hawaiian islands can receive swell from north Pacific sources,
and the magnitude of the southerly swell is typically low, a similar
estimate of monthly mean Hs and Tp as done for the Waimea Buoy off the
North Shore was not calculated for southerly surf.
Boreal winter months (December to March) provide infrequent surf for
south shores from north Pacific sources. Extra-tropical cyclones of
lower latitude in the western Pacific can generate westerly-component,
long-period swell (Hs~2-4', Tp~14-20s, wave direction less than 280 degrees).
H1/10 surf heights at exposed locations typically range from ~5'
to as high as 12' for peak face value. In addition,
Kona weather patterns occur within fall through spring
when sharp upper atmospheric troughs or cut off lows and
their associated surface low pressure features
unfold to the immediate west of Hawaii. The Kona winds (from within SE to W)
generate windswell (Hs~6-12', Tp~6-10s) with rough, choppy H1/10 breaker
heights within 4-10'. There is great interannual variability in
both of these winter sources.
Surf observations in the Goddard-Caldwell Database for Windward Oahu began in 1987. North Beach was used into
the 1990s, then a combination of Makapuu or a "windward index" based on Caldwell observations in the Kailua-Kaneohe
area, which would include windswell as well as north-component longer period swell. Ideally, only Makapuu
would be the best choice for an exclusive trade windswell breaker estimate. However, it is not digitally available. Thus,
available wind and buoy data are studied.
Mean winds provided by the
NOAA/NESDIS National Centers for Environmental Information from land-based observations in Hawaii show variability from place
to place due to local topography. The mean wind at Kaneohe Bay MCAS is biased
low due to shadowing by Mokapu crater. A more optimal location would be
wind measurements offshore. Buoy 51001 is too far northwest and buoy 51002 and 51004
are too far south and southeast, respectively. The best located buoy,
NOAA Buoy 51026
was positioned north of Molokai from 1993 to 1996.
However, the short time-series limits statistical confidence.
The
buoy 51026 monthly wind speed statistics reflect a complicated
annual cycle with three maximas (Nov-Dec, Apr, Jun-Jul). The
annual mean wind speed is ~14 knots (16 mph). The agreement with
the Kahului, Maui wind averages increases confidence in this statistic.
Such a moderate breeze would make seas of 5 feet with 6 second peak wave
periods, and in turn, breakers from windswell of H1/10 peak face ~3 feet. This estimate for a mean breaker height from trade windswell seems low based
on Caldwell's 30 years of wind- and kite-surfing in the Kailua area surf and cross-comparisons with nearshore buoy readings.
A better estimate can be made from the offshore buoy data. Buoy 51026 is not used due to exposure to north-component swell.
The
Mokapu buoy off Kailua, Oahu has less north swell exposure and
is better suited for analysis of the trade windswell.
From the 30-minute samples over a 16-year period (2000-2016), these
data were filtered to remove the long period north Pacific swell
and the windswell beyond the range of
the trades. Only readings within peak wave direction 30 to 100 degrees and peak wave period less than 13 second
were analyzed. No effort was made to remove any tropical cyclone swell of wave periods less than 13 seconds,
since such sources rarely exceed levels generated
by strong to near gale trades. The
Mokapu buoy trade windswell sample statistics give a mean annual
significant wave height (Hs) = 6.4 feet at peak wave period (Tp) of 8.3 seconds. This is larger than one
would expect given the mean local wind. The waves are higher and longer since the fetch area to the east to northeast of
Hawaii is vast, meaning windswell is propogating into the area, not just being generated locally.
The empirical method, or surf compute, was
created matching long-period swell at the Waimea Buoy to the North Shore visual surf observations. The
method is unrealistic for wave periods less than ~10 seconds. Based on Caldwell's unofficial comparison under trade windswell
conditions between buoy readings
and breaker size for high refraction reefs in the Kailua-Kaneohe area, Hs=6.4' at Tp=8.4 seconds should give an H1/10 breaker
of approximately 5 feet, which is the assumed trade windswell average surf height.
Surf Research NOAA/NESDIS/NCEI Hawaii Liaison/Pacific Region Science Officer
Aucan Directional Wave Climatology Hawaii
Wave Climate SOEST Coastal Hazards Analysis Report,
Coastal Geology Group, University of Hawaii
Hawaii Wave Climatology from Hindcast Department of Ocean and Resources Engineering, University of Hawaii
NOAA/NESDIS National Centers for Environmental Information (NCEI) Historic Weather, Climate and Oceanographic Data
Central North Pacific Climatology Overview NOAA/NESDIS/NCEI
Wind-wave Climate of the Pacific Ocean CSIRO, Australia
Important Notices:
Data Sources
North Shore, Oahu
South Shore, Oahu
Trade Windswell
Associated Links
Higher Level Agencies: