mechanism, and a pencil records sea level changes by means of a 

 float-pulley system. A second instrument, a Fisher-Porter digital tide 

 gage, was installed adjacent to the analog gage by the NOS in 

 November 1973. This is an electrically operated system which 

 punches digital data on foil tape. Both gages use the same 21.6 cm 

 diameter float and have operated simultaneously since November 

 1973. The stilling well, which serves as a low pass filter for oscilla- 

 tions with periods greater than a minute , consists of a 30 . 5 cm diame- 

 tersteel pipe with a 2.5 cm diameter orifice at the bottom. Both gages 

 are checked for accuracy of time and height and are annotated about 

 five times per week. 



Data Processing and Reduction. — Continuous tide traces 

 obtained from the analog gage during the period 20 July 1963 

 through 31 December 1974 were manually digitized for use in this 

 study by Ocean Data Systems, Inc., Monterey, Calif. Datums were 

 reviewed and data were reduced to hourly sea level heights using 

 standard NOS procedures (Coast and Geodetic Survey 1965). Data 

 from the digital gage for the period 1 January 1974 through 31 Sep- 

 tember 1976 were processed for hourly heights by the NOS and pro- 

 vided for use in this study. Data from both gages were recorded in feet 

 and in this study converted to centimeters. The hourly heights from 

 both analog and digital gages are accurate to about 0. 1 ft (3.0 cm) and 

 times of observation (Pacific Standard Time) are accurate to within 6 

 min. A small percentage of the hourly sea level data was missing, 

 either rejected as erroneous or lost due to equipment malfunctions. 

 As a result, some monthly means contain less than a full month of 

 data. Missing data of duration of a day or longer are listed in Appen- 

 dix A. 



All hourly heights were measured relative to the station datum 

 established by the NOS in November 1973. Mean sea level for the 

 period 1963 through 1978 lies at 184.4 cm and the National Geodetic 

 Vertical Datum lies 182.88 cm above the station datum. 



Merging of Analog and Digital Tide Data. — To obtain the long- 

 est possible continuous tide record, it was necessary to merge the 

 older analog data with the more recent digital data. Before the data 

 sets were combined, the response of the two gauges was analyzed by 

 comparing the hourly heights from both tide records for the calendar 

 year 1974. The correlation coefficient between the analog and digital 

 data sets exceeds 0.99, as anticipated. 



The differences (digital-analog) between the two sets of hourly sea 

 levels for the calendar year 1974 had a mean value of -0.06 cm. The 

 frequency distribution of the differences (Fig. 2) resembles a normal 

 distribution, with a standard deviation of 3.7 cm. Nearly all of the 

 differences can be attributed to the fact that the digital data were 

 recorded as instantaneous values, which can include short-term sea 

 level fluctuations such as long period waves and seiches, whereas in 

 the analog data, these short-term fluctuations were filtered out by 

 manually smoothing the tidal curve before digitizing. 



It was concluded that differences between the two data sets were 

 negligible, and that the analog and digital data could be combined 

 without significant error. Thus, analog data from the period 20 July 

 1963 through 31 December 1974 were combined with digital data 

 from the period 1 January 1975 through 31 August 1976 to form a 

 13-yrtime series containing 107,954 hourly observations. 



Long Period Sea Level Changes.— Tide gages monitor the 

 height of the sea level relative to land. Thus, changes in mean sea 

 level over periods of years or decades can result from the addition 

 or removal of water from the oceans due to global climatic varia- 

 tions, from subsidence or emergence of the land upon which the 



-30 -15 15 30 



Digital Minus Analog Tide Height (cm) 



Figure 2. — Comparison of hourly tide measurements in Monterey Bay, Calif., from 

 digital and analog gages for calendar year 1974. Total number of observations Mas 

 107,954. 



gage is located, or from long-period astronomic tides. For exam- 

 ple, some long-period trends in sea level records, such as the rise in 

 sea level in Panama described by Roden (1963) or the drop in sea 

 level in the Juneau, Alaska, area described by Hicks (1973), clearly 

 result from local or regional land subsidence or uplift. 



To determine trends in the Monterey sea level record during the 

 period 1963 through 1978. a least-squares linear fit was made to the 

 time-series on monthly mean values. The fit showed a relative rise in 

 sea level of about 0.01 cm/yr. The variability in sea level due to 

 oceanographic and meteorological processes greatly exceeds this 

 trend and thus the effects of long term trends were neglected in this 

 study. 



Of the long-period astronomic tides, the nodal tidal constituent, 

 which results from the changing declination of the moon over a pe- 

 riod of 18.61 yr, has the greatest amplitude. The theoretical ampli- 

 tude of this constituent varies with latitude, with maximum effects at 

 the Equator and the poles and minimum effects near lat. 35°N and 

 35 °S (Lisitzin 1974). A second significant long period constituent, 

 the annual solar tide, has an amplitude approximately one-fifth of the 

 nodal tide component. The effects of this tidal constituent vary with 

 latitude in a manner similar to that of the nodal tide. Monterey, 

 located near lat. 36°N, is in a region where the ranges of both of these 

 long period tides are about 1 cm, so these effects were neglected in 

 this study. 



Ocean and Atmospheric Data 



The atmospheric pressure and wind data used in this study were 

 derived from 6-h synoptic surface pressure fields prepared by Fleet 

 Numerical Oceanography Center (FNOC). The pressure fields, 

 interpolated onto a grid with a mesh length of 3° latitude and longi- 

 tude, were used to compute geostrophic winds, from which wind 

 stress, Ekman transport, and Sverdrup transport estimates were cal- 

 culated at a deep water site approximately 14 km west of Monterey 

 (Fig. 1). A description of the methods and computations used in these 

 calculations is given by Bakun (1975). Briefly, the geostrophic wind 

 was computed for the point lat. 36.6°N. long. 122. 1°W and an esti- 

 mate of the wind near the sea surface was made by rotating the geos- 

 trophic wind vector 15° to the left and reducing its magnitude by 



