WESTERN REGION TECHNICAL ATTACHMENT
NO. 03-06
May 14, 2003

High Wind and Wave Events Along the Northern California Coast During Summer

Jeff Lorens, Weather Forecast Office, Eureka CA

Introduction

a. Background. During the summer months, swell from distant storms normally decays to a few feet or less by the time it reaches the west coast of North America, and is often only a slight contributor to the total significant wave height (with a few significant exceptions, as will be noted later). Instead, wave generation along the northern California coast is largely driven by regional and local wind patterns. Beardsley, et. al. (1987) describes the predominantly low level flow along the California coast in summer. Mass, et. al. (1987) describes particular flow regimes in this same region which result in occasional shifts in the low level flow along the coast to a southerly direction. Additionally, Burke and Thompson (1996) describe a phenomena known as the northerly low level jet. Each of these flow situations are largely driven by synoptic and mesoscale features, including coastal topography, but they will not be further discussed here.

Regardless of the particular flow regime, however, the most direct and persistent influence on surface winds in this region comes from sea level pressure patterns. Pressure patterns here are typically characterized by the presence of a thermally-induced trough along or near the California coast, and the persistent eastern north pacific high pressure cell. As a result of these features, moderate to strong pressure gradients often develop over the northern California coastal waters. The resultant winds (also influenced by the other factors mentioned previously), depending on their strength, duration, and fetch length and orientation, can generate large and/or steep waves in a short period of time. These rapidly-developing wind and wave conditions pose a significant hazard to mariners, especially smaller vessels, and are the primary focus of this study.

Because the thermal trough and eastern north Pacific high are such persistent features in this region, long-term averages of winds and waves are potentially very useful in daily forecasting applications. In the following study, a detailed climatology of summer (June through September) wind and wave information is presented, based on approximately 20 years of data from NOAA environmental data buoys located along the northern California coast (see paragraph 2b, below). Based on this data, long-term average values of wind speed and direction, wave height, and dominant period are presented. Historical extremes (from the period of study) are also included for wind speed and wave height data. Additionally, information on duration of these summer wind and wave events will be presented.

Finally, a series of short case studies will be presented. These case studies will highlight wind and wave characteristics typical of summer wind and wave events along the northern California coast. Examples will focus primarily on two types of events: (1) "periodic" events (dominated by moderate to strong diurnal variations in wind speed and wave height), and (2) "rapid-rise" events (characterized by longer duration build-ups of wind speed and wave height).

b. Data. Quality-controlled wind, pressure, and wave data for the months of June through September, 1981-2001 was obtained from NOAA's National Data Buoy Center (NDBC) archive files (on-line and CD-ROM). The four buoy locations selected were: #46027 (8 nautical miles west of Crescent City), #46022 (17 nautical miles west-southwest of Eureka), #46030 (8 nautical miles west of Cape Mendocino), and #46014 (19 nautical miles northwest of Point Arena). See also Figure 1 for buoy locations. Data elements included for analysis in this study include the following variables:

(1) Wind Speed and Direction (averaged over 8-minute periods),
(2) Wind Gust (peak wind speed over 8-minute periods),
(3) Significant Wave Height (average of the highest one-third of wave heights observed in a 20-minute sampling period), and
(4) Dominant Period (wave period with maximum energy)
(5) Atmospheric Pressure (Mean Sea Level )

Figure 1. Buoy locations (buoys used for this study shown in yellow). (Note: Buoy #46030 is no longer located at position shown; this buoy was relocated off the southern Oregon coast and re-designated as #46015.)

There were occasional periods of missing data throughout the entire period evaluated. Most commonly, one or a few elements were missing for relatively short periods (one to several hours). In a few cases, however, all data from particular buoys was missing for extended periods (up to several months), due to several possible reasons, including sensor or data transmission problems. On rare occasions, buoys have broken their moorings in strong storms and gone adrift. In such cases, data transmission is normally suspended until the buoy can be returned to service. Appendix 1 summarizes the data availability for each of the buoys for this study period.

2. Wind and Wave Averages.

a. Wind speed and direction. Table 1 shows the overall maximum sustained wind speeds and significant wave heights, along with the number of observations each was based on. Figure 2 shows wind speed frequencies (based on all available data for all directions) at each buoy, in terms of increasing speed thresholds. Wind speeds for all locations were 10 knots or greater more than 35 percent of the time (up to 62 percent of the time for buoy #46030). At higher wind speed thresholds (>= 20 knots), buoys #46027 and #46014 tended to dominate.

Table 1. Maximum observed wind speeds (sustained) and significant wave heights for buoy locations for entire period of study (see Appendix 1 for specific periods of data availability).

Figure 3 shows the frequency distribution of wind direction. Directions from 031o through 130o (clockwise) were not included in the analysis, primarily because speeds in this range of directions tended to be relatively light. Additionally, winds from these directions have an offshore component, so fetch lengths are typically shorter. Therefore, the associated wind-waves tend to be lower than those associated with other directions (131o clockwise through 030o). Figures 4 and 5 show the overall average and maximum wind speed and gust (by direction) for each location. Figures 6 through 9 show wind speed and direction frequencies for individual buoy locations.

Figure 2. Wind speed frequency (cumulative) for specified thresholds.
Figure 3. Wind direction frequency (131o through 030o).
Figure 4. Average wind speed by direction.
Figure 5. Maximum wind speed and gust by direction (dashed line denotes missing data).
Figure 6. Frequency of wind speed (knots) classified by direction (buoy # 46027).
Figure 7. Frequency of wind speed (knots) classified by direction (buoy # 46022).
Figure 8. Frequency of wind speed (knots) classified by direction (buoy # 46030).
Figure 9. Frequency of wind speed (knots) classified by direction (buoy # 46014).

b. Significant wave heights. Long-term averages of significant wave heights along the Northern California coast during the summer vary, depending on location and wind direction. This is due, in part, to the resultant effects of direction on fetch lengths. Wind directions from the north, south and west (along-shore and on-shore) have longer (potential) fetch lengths than directions from the east (offshore). Figure 10 shows the frequency of significant wave heights at each buoy, classified according to wind direction. Data for wind directions with a significant offshore component are not included because of the shorter fetch lengths involved. As noted above, north to northwest winds, and (to a lesser extent) south to southwest winds are also associated with higher average wind speeds. The combination of higher wind speeds, longer fetch lengths, and (likely) greater persistence all contribute to higher significant wave heights (i.e. locally and regionally wind-generated waves). As noted previously, wind speeds associated with westerly directions average lower (only about 3-4 knots) and are less frequent than those from the north to northwest or south to southwest. Therefore, despite essentially unlimited fetch lengths, waves associated with westerly directions are lower on average. Additionally, although a directional analysis of wind persistence was not conducted here, longer wind speed persistence values are probable, due to higher frequencies of occurrence (paragraph 2d for a non-directional discussion of wind event durations).

Figure 10. Significant wave heights (by wind direction).

c. Dominant wave periods. Figure 11 shows the frequency of occurrence of dominant wave periods. Figure 12 displays the same information, but in a cumulative fashion. As with wave height, wave period is a function of fetch (length and orientation), wind speed, and wind duration.

The "dominant" wave period is that period associated with the maximum wave energy, and is also one of the two period values reported by the buoys (the other is "average period" -- not used here). Other period values, associated with "wind wave" and "swell," are derived via a Fourier transform technique from the buoy spectral wave data (Tucker, 1991), and are not included here.

As noted previously, and as will be discussed further in the case studies section later, wind speeds and durations vary widely during the summer months in this region. Fetch length, however, is typically limited to distances on the order of a few hundred miles and, given the predominance of north to northwest and (secondarily) south to southwest winds, this commonly results in a fetch orientation parallel to the coast (or nearly so).

Wind-generated waves are typically characterized by short periods, typically less than about 6 to 8 seconds. As waves propagate out from the generation (fetch) area, wind waves gradually transition to swell; periods tend to become longer and wave heights subside, due to dispersion of energy. As will be further discussed in the case studies, one of the most common fetch areas in this area in summer is associated with tighter pressure gradients along the far northern California and southern Oregon coasts, as a result of interaction between the thermal trough and the northeast Pacific high. Given that the most common wind directions here are from the north to northwest, along with the highest average speeds, this typically results in shorter fetch lengths at the north end of this area (near buoy #46027), and conversely, longer fetch lengths at the southern end (near buoy #46014). Figure 13 shows the variation of dominant periods categorized by wind direction. This figure shows the tendency for shorter dominant periods (especially at buoy #46014) with the north to northwest winds so characteristic of this region in summer.

Based primarily on the variation in fetch length, shorter wave periods should typically be expected to the north and longer periods to the south, and the long-term averages in dominant period tend to reflect this. Looking at Figure 11, shorter dominant periods (especially from about 7 to 8 seconds) are more frequent at buoy #46027 than at buoys to the south. At longer periods (from about 8 through 11 seconds), there is a slightly higher frequency of occurrence at the buoys to the south of #46027. At the long end of the range, e.g. periods around 16 seconds, the two southern-most buoys in this area (#46030 and #46014) show a slightly higher frequency of occurrence than at the two northern buoys (#46027 and #46022). Using simple wave decay assumptions, the dominant period would not typically increase by more than a few seconds in propagating southward over the distances involved here (maximum distance approximately 200 miles, from buoy #46027 to #46014). It may be that conditions south of Cape Mendocino allow for a broader spectrum of wave energy to be observed, i.e. the influence of long-period wave energy (due to swell from distant storm systems) becomes greater. Because the difference is so slight at these longer periods, though, it is not certain that the difference is significant.

Figure 11. Frequency of dominant wave periods.

Figure 12. Cumulative frequency of dominant wave periods.

Figure 13. Average of dominant wave period categorized by wind direction.

d. Duration of wind events. Fetch length and orientation tend to be the most persistent factors in wave development in this region. So, wind speed and duration become critical factors in determining wave growth and decay along the coast. Wind data at each buoy were evaluated to determine how long wind speeds in this coastal region persist, based on certain conditions. "High wind events," are defined here using sustained wind speed thresholds of 20, 25, and 30 knots, and durations of at least 12 hours (consecutive). Although directions are not explicitly included, general observation of the data showed the vast majority of events to be of relatively constant direction (i.e. wind directions did not typically vary by more than about 30 degrees when winds were at these levels. Table 2 lists the number of events for each buoy, along with maximum sustained wind speed and significant wave height for each event. Figures 14-17 graphically show the results for each buoy. Clearly, buoys #46027 and #46014, at the north and south ends of this area, had the greatest number of "high wind" events, and also the greatest number of long-duration events (including a few exceeding 72 hours) were much more frequent). It is important to note here that this data may not include the longest durations, highest wind speeds, or highest waves, due to occasional long periods of missing data (refer to Appendix X for summary of missing data). Additionally, as will be seen in the some of the information presented in the following case studies, longer dominant periods (up to 17 seconds) at times indicated that swell from distant sources was a significant contributing factor to the high waves observed.

Table 2. Wind duration summary for specified wind speed thresholds (for events >=12 consecutive hours), with maximum wind speed (sustained) and wave height.

Figure 14. Duration of wind events (buoy # 46027) >= 20 knots and >= 12 hours, with associated maximum wind speeds and wave heights (zero-values denote missing data).

Figure 15. Duration of wind events (buoy # 46022) >= 20 knots and >= 12 hours, with associated maximum wind speeds and wave heights.

Figure 16. Duration of wind events (buoy # 46030) >= 20 knots and >= 12 hours, with associated maximum wind speeds and wave heights.
Figure 17. Duration of wind events (buoy # 46014) >= 20 knots and >= 12 hours, with associated maximum wind speeds and wave heights (zero-values denote missing data).

e. Long-term correlation between pressure gradients (between buoys) and wind speeds (at the buoys). Correlation coefficients were calculated using all buoy data available for the summer months (June-September) through 2001, in order to evaluate relative strength and weakness of the long-term gradient-wind speed relationships. Consideration was given only to the absolute difference of pressure gradients and their relation to wind speed. Wind direction and orientation of the pressure gradients were not evaluated. Correlation coefficients between wind speeds and pressure gradients at the same times were calculated. Additionally, correlation coefficients for specified "lag times" (time of observed pressure gradient - time of observed wind speed) up to 24 hours were calculated to assess possible predictive value. Although buoy #46030 wind and pressure data was available for the period of study, it was not used. As noted previously, buoy #46030 is no longer located off Cape Mendocino, and therefore the results of the analysis at this particular location are of only limited use for future predictive purposes. The correlation coefficient results for the other three buoy locations are given in Figures 18-20.

Overall, correlation was strongest (r = 0.68) between wind speeds at buoy #46014 and the buoy #46022-46014 pressure gradients with a lag time = 0 hours (pressure gradient and wind speed at the same time). Thereafter, correlation coefficients steadily decreased, to 0.40 with a lag time = 24 hours. In some cases, however, correlation coefficients actually increased as lag time increased. At buoy #46027, correlation coefficients (using #46027-46014 pressure gradients) increased from 0.25 at lag time = 0 hours to 0.41 at lag time = 14 hours. Also at bouy #46027, while correlation coefficients initially dropped from 0.50 (at lag time = 0 hours) to 0.11 (at lag time = 10 to12 hours), they then rose back to 0.30 at lag time =24 hours, possibly due (at least in part) to repetitive, and at times strongly diurnal wind patterns at this location. Other correlation coefficients (as shown below) also showed some increasing tendencies as lag time increased, indicating possible predictive value. Pressure gradient and wind speed relationships are evaluated in more detail in some of the case studies to follow.

Figure 18. Wind speed (buoy #46027) and pressure gradient (absolute value) correlation coefficients for specified "lag times," i.e. pressure gradient at time=t and wind speed at time= t + delta-t).

Figure 19. Wind speed (buoy #46022) and pressure gradient (absolute value) correlation coefficients for specified "lag times," i.e. pressure gradient at time=t and wind speed at time= t + delta-t).

Figure 20. Wind speed (buoy #46014) and pressure gradient (absolute value) correlation coefficients for specified "lag times," i.e. pressure gradient at time=t and wind speed at time= t + delta-t).

Appendix 1. Data Inventory.
("X" denotes available data; blanks denote periods of missing data)

Case Studies

#1 CASE01EVENTDESCRIPTION ((All buoys) 12-23 Jun01)
#2 CASE02EVENTDESCRIPTION (#46022 13-16 Jul 82)
#3 CASE03EVENTDESCRIPTION (#46022 9-20 Sep 88)
#4 CASE04EVENTDESCRIPTION (#46022 3-6 Jun 91)
#5 CASE06EVENTDESCRIPTION (#46014 5-12 Jul 93)
#6 CASE11EVENTDESCRIPTION (#46014 15-18 Jul 87)
#7 CASE12EVENTDESCRIPTION (#46014 15-17 Sep 87)
#8 CASE14EVENTDESCRIPTION (#46014 24-30 Sep 97)
#9 CASE15EVENTDESCRIPTION (#46014 11-18 Aug 00)`

References.

Beardsley, R.C., C. E. Dorman, C.A. Friehe, L.K., and C.D. Winant, 1987: Local Atmospheric Forcing During the Coastal Ocean Dynamcis Experiment, 1. A Description of the Marine Boundary Layer and Atmospheric Conditions Over a Northern California Upwelling Region. Journal of Geophysical Research, 92, 1467-1488.

Burke, S.D. and W.T. Thompson, 1996: The Summertime Low-Level Jet and Marine Boundary Layer Structure along the California Coast. Monthly Weather Review, 124(4), 668-686.

Gilhousen, D.B., 1987: A field evaluation of NDBC moored buoy winds. Journal of Atmospheric and Oceanic Technology, 4, 94-104.

Mass, C. F. and M. D. Albright, 1987: Coastal Souterlies and Alongshore Surges of the West Coast of North America: Evidence of Mesoscale Topographically Trapped Response to Synoptic Forcing. Monthly Weather Review, 115, 1707-1738.

Neiburger, M., D.S. Johnson, and Chen Wu Chien, 1961: Studies of the structure of the atmosphere over the eastern Pacific Ocean in summer, Pt. 1: The inversion over the eastern north Pacific Ocean. Publications in Meteorology, Univ. of California, 1(1), 1-94.

Tucker, M.J., 1991: Waves in Ocean Engineering: Measurement, Analysis, and Interpretation. Ellis Horwood, LTD., 431 pp.

Steele, K.E. and T.R. Mettlach, 1993: NDBC wave data - current and planned. Ocean Wave Measurement and Analysis - Proceedings of the Second International Symposium. ASCE, 198-207.