Introduction

Systematic, long-term time-series studies of selected aquatic and terrestrial habitats have yielded significant contributions to earth and ocean sciences through the characterization of climate trends. Important examples include the recognition of acid rain (Hubbard Brook long-term ecological study, Vermont; Likens et al., 1977), the documentation of increasing carbon dioxide (CO₂) in the earth's atmosphere (Mauna Loa Observatory, Hawaii; Keeling et al., 1976) and the description of large scale ocean-atmosphere climate interactions in the equatorial Pacific Ocean (Southern Oscillation index; Troup, 1965).

Long time-series observations of climate-relevant variables in the ocean are extremely important, yet they are rare. Repeated oceanographic measurements are required to gain an understanding of natural processes or phenomena that exhibit slow or irregular change, as well as rapid event-driven variations that are impossible to document reliably from a single field expedition. Time-series studies are also ideally suited for the documentation of complex natural phenomena that are under the combined influence of physical, chemical and biological controls. Examination of data derived from the few long-term oceanic time-series that do exist provides ample incentive and scientific justification to establish additional study sites (Wiebe et al., 1987).

The role of the oceans in climate variability is primarily in the sequestration and transportation of heat and carbon (Barnett, 1978). Both can be introduced into the ocean in one place, only to return to the atmosphere, at a subsequent time, possibly at a distant location. While both heat and carbon can be exchanged with the atmosphere only carbon is lost to the seafloor through sedimentation. The oceans are known to play a central role in regulating the global concentration Of CO₂ in the atmosphere (Sarmiento and Toggweiler, 1984; Dymond and Lyle, 1985). It is generally believed that the world ocean has removed a significant portion of anthropogenic CO₂ added to the atmosphere, although the precise partitioning between the ocean and terrestrial spheres is not well constrained (Tans et al., 1990; Quay et al., 1992; Keeling and Shertz, 1992).

The cycling of carbon within the ocean is controlled by a set of reversible, reduction-oxidation reactions involving dissolved inorganic carbon (DIC) and dissolved organic carbon (DOC) with marine biota serving as the critical catalysts. Detailed information on the rates and mechanisms of removal of DIC from the surface ocean by biological processes, the export of biogenic carbon (both as organic and carbonate particles) to the ocean's interior, and the sites of remineralization and burial are all of considerable importance in the carbon cycle. The continuous downward flux of biogenic materials, termed the "biological pump" (Volk and Hoffert, 1985; Longhurst and Harrison, 1989), is a central component of all contemporary studies of biogeochemical cycling in the ocean and, therefore, of all studies of global environmental change.

During the embryonic phase of ocean exploration more than a century ago (Thomson, 1877), it was realized that a comprehensive understanding of the oceanic habitat and its biota would require a multidisciplinary experimental approach and extensive field observations. Progress toward this goal has been limited by natural habitat variability, both in space and time, and by logistical constraints of ship-based sampling. Consequently, our current view of many complex oceanographic processes is likely biased (e.g., Dickey, 1991; Wiggert et al., 1994). The synoptic and repeat perspective that is now available from research satellites is expected to improve our understanding of oceanic variability, despite certain limitations.

In 1988, two deep ocean time-series hydrostations were established with support from the U.S. National Science Foundation (NSF): one in the western North Atlantic Ocean near the historical Panulirus Station (Bermuda Atlantic Time-series Study [BATS]; Michaels and Knap, 1996) and the other in the subtropical North Pacific Ocean near Hawaii (Hawaii Ocean Time-series [HOT]; Karl and Lukas, 1996). These programs were established and are currently operated by scientists at Bermuda Biological Station for Research and the University of Hawaii, respectively.

The primary research objectives of these ocean measurement programs are to establish and maintain deep-water hydrostations for observing and interpreting physical and biogeochemical variability. The initial design called for repeat measurements of a suite of core parameters at approximately monthly intervals, compilation of the data and rapid distribution to the scientific community.

A deep-ocean weather station network was established in the post-World War II period as a ship-based observation program designed to improve global weather prediction capabilities. One of the sites, Station November, was located in the eastern sector of the North Pacific Ocean gyre at 30N, 140W and was occupied during 121 cruises between July 1966 and May 1974. The intercruise frequency ranged from a few days to a few weeks with a typical cruise duration of 2-3 weeks, including transits. Water samples were collected from approximately 12-14 depths in the range of 0-1500m using bottles equipped with deep-sea reversing thermometers. Salinity and, on occasion, dissolved oxygen concentrations were measured from the discrete water samples.

During the 1970s, most of the U.S. weather ship stations were phased out of operation and were eventually replaced with more cost-effective, unattended ocean buoys. These buoys measure standard meteorological parameters as well as basic wave characteristics (e.g., significant wave direction, height, period and spectrum) but few, if any, hydrographic variables.

Physical and biogeochemical time-series investigations of the North Pacific subtropical region are sparse and consist of a series of unrelated research programs including CliMAX, GOLLUM, NORPAX, VERTEX, ADIOS and most recently HOT. CliMAX I occupied a series of stations near 28N, 155W during August-September 1968 and CliMAX II reoccupied the site during September of the following year. Since that time, scientists from the Scripps Institution of 0ceanography have revisited the "CliMAX region" (26.5 to 31N, 150.5 to 158W) on 18 cruises between 1971 and 1985 (Hayward, 1987). It is important to emphasize that the temporal coverage in this time-series is biased with respect to season because approximately 70% of the cruises occurred in summer (June-Sept) and 35% were in August alone. These observations are also aliased by the annual cycle because no cruises were conducted in 1970, 1975, 1978-79, 1981 or 1984. Nevertheless, observations made during this extensive series of cruises, especially the measurements of plankton distributions, nutrient concentrations and rates of primary production, provided an unprecedented view of the oligotrophic North Pacific ecosystem structure and dynamics.

From January 1969 to June 1970, a deep ocean hydrostation (Station GOLLUM) was established by scientists at the University of Hawaii at a location 47 km north of Oahu (22 10'N, 158 00'W; Gordon, 1970). The water depth was 4760m and the location was selected to be beyond the biogeochemical influences of the Hawaiian Ridge (Doty and Oguri, 1956). On approximately monthly intervals, 13 two-day research cruises were conducted to observe and interpret variations in particulate organic matter distributions in the water column and other parameters (Gordon, 1970).

A major advance in our understanding of biogeochemical processes in the sea was made during the NSF International Decade for Ocean Exploration (IDOE)-sponsored Geochemical Ocean Sections Study (GEOSECS) Pacific Ocean expedition (August 1973 - June 1974). Although repeated ocean observations were not made during GEOSECS, the high-precision data, including numerous radioactive and stable isotopic tracers, that were collected from selected stations in the central North Pacific Ocean can be used as the basis for assessing "change," especially for the concentration and ¹³C isotopic composition of the total dissolved carbon dioxide pool (Quay et al., 1992). In particular, GEOSECS stations #202 (33 6'N, 139 34'W), #204 (31 22'N, 150 2'W), #212 (30N, 159 50'W) and #235 (16'45'N, 161'19'W) are the most relevant to our current biogeochemical investigations at Station ALOHA.

In the early 1970's the North Pacific experiment (NORPAX) was initiated as an additional component of the NSF-IDOE. Research was focused on large scale interactions between the ocean and the atmosphere (e.g., El Niño), and the application of this knowledge to long-range climate forecasting. The Anomaly Dynamics Study was one component of NORPAX aimed at understanding interannual variability of the mid-latitude, North Pacific upper ocean thermal structure. Long-term ocean observation programs were fundamental to the success of NORPAX and, accordingly, the Trans-Pac XBT program and the Pacific Sea Level Network were established. Furthermore, the extensive 15 cruise Hawaii-to-Tahiti Shuttle time-series experiment (January 1979 - June 1980) was conducted to obtain direct measurements of the temporal variations in thermal structure of the equatorial Pacific region. These cruises also supported extensive ancillary research programs on chemical and biological oceanography, and provided a rich dataset including measurements of DIC and primary productivity (Wyrtki et al., 1981).

With the abandonment of the central North Pacific Ocean weather ship stations and time-series programs such as Station GOLLUM, there remained very few sites where comprehensive serial measurements of the internal variability of the ocean were continuing. The Intergovernmental Oceanographic Commission (IOC) and World Climate Research Program (WCRP) Committee on Climate Change in the Ocean (CCCO) recognized this deficiency, and in 1981 endorsed the initiation of new ocean observation programs. Reactivation of Station GOLLUM was an explicit recommendation (JSC/CCCO, 1981).

In 1986, a biogeochemical time-series station was established in the northeast Pacific Ocean (33'N, 139'W) as one component of the NSF-sponsored Vertical Transport and Exchange (VERTEX) research program. A major objective of the VERTEX time-series project was to investigate seasonality in carbon export from the euphotic zone in relation to contemporaneous primary production. During an 18-month period (October 1986 - May 1988), the station was occupied for seven 1 -week periods on approximately 3-month intervals. In addition to standard hydrographic surveys, samples were also collected for the measurement of dissolved inorganic and organic nutrients, particulate matter elemental analysis, primary production, nitrogen assimilation rates, microbial biomass and particle flux (Knauer et al., 1990; Harrison et al., 1992). Significant variability was observed in rates of primary production and particle flux and no clear relationship was found between new production and primary production. Despite the comprehensive scope and intensity of this research project, the sampling frequency was clearly inadequate to resolve much of the natural variability in this oligotrophic oceanic ecosystem.

In response to the growing awareness of the ocean's role in climate and global environmental change, and the need for additional and more comprehensive oceanic time-series measurements, the Board on Ocean Science and Policy (BOSP) of the National Research Council (NRC) sponsored a workshop on "Global Observations and Understanding of the General Circulation of the Oceans" in August 1983. The proceedings of this workshop (National Research Council, 1984a) served as a prospectus for the development of the U.S. component of the World Ocean Circulation Experiment (U.S.-WOCE). U.S.-WOCE has the following objectives: (1) to understand the general circulation of the global ocean, to model with confidence its present state and predict its evolution in relation to long-term changes in the atmosphere and (2) to provide the scientific background for designing an observation system for long-term measurement of the large-scale circulation of the ocean.

In a parallel effort, a separate research program termed Global Ocean Flux Study (GOFS) focused on the ocean's carbon cycle and associated air-sea fluxes of carbon dioxide. In September 1984, NRC-BOSP sponsored a workshop on "Global Ocean Flux Study" which served as an eventual blueprint for the GOFS program (National Research Council, 1984b). In 1986, the International Council of Scientific Unions (ICSU) established the International Geosphere-Biosphere Programme: A Study of Global Change (IGBP), and the following year JGOFS (Joint GOFS) was designed as a Core Project of IGBP. U.S.-JGOFS research efforts focus on the oceanic carbon cycle, its sensitivity to change and the regulation of the atmosphere-ocean CO₂ balance (Brewer et al., 1986).

The broad objectives of U.S.-JGOFS are: (1) to determine and understand on a global scale the time-varying fluxes of carbon and associated biogenic elements in the ocean and (2) to evaluate the related exchanges of these elements with the atmosphere, the sea floor and the continental boundaries (SCOR, 1990, JGOFS Rept. #5). To achieve these goals, four separate program elements were defined: (1) process studies to capture key regular events, (2) long-term time-series observations at strategic sites, (3) a global survey of relevant oceanic properties (e.g., CO₂) and (4) a vigorous data interpretation and modeling effort to disseminate knowledge and to generate testable hypotheses.

In 1987, two separate proposals were submitted to the U.S.-WOCE and U.S.-JGOFS program committees to establish a multi-disciplinary, deep water hydrostation in Hawaiian waters. In July 1988, these proposals were funded by the National Science Foundation and Station ALOHA was officially on the map.

The primary objective of HOT is to obtain a long time-series of physical and biochemical observations in the North Pacific subtropical gyre that will address the goals of the U.S. Global Change Research Program. The objectives specific to the Physical Oceanography (P.O.) program are to:

• Document and understand seasonal and interannual variability of water masses.
• Relate water mass variations to gyre fluctuations.
• Determine the need and methods for monitoring currents at Station ALOHA.
• Develop a climatology of short-term physical variability.

In addition to these general primary objectives, the physical oceanographic component of HOT provides CTD/rosette sampling support for the biogeochemistry & ecology (BEACH) time-series sampling program, and supports development of new instrumentation for hydrographic observations.

The objectives of HOT specific to the BEACH program are to:

• Document and understand seasonal and interannual variability in the rates of primary production, new production and particle export from the surface ocean.
• Determine the mechanisms and rates of nutrient input and recycling, especially for nitrogen (N) and phosphorus (P) in the upper 200m of the water column.
• Measure the time-varying concentrations of dissolved inorganic carbon (DIC) in the upper water column and estimate the annual air-to-sea CO₂ flux.

In addition to these primary objectives, the HOT Program provides logistical support for numerous complementary research programs.

There are both scientific and logistical considerations involved with the establishment of any long-term, time-series measurement program. Foremost among these is site selection, choice of variables to be measured and general sampling design, including sampling frequency. Equally important design considerations are those dealing with the choice of analytical methods for a given candidate variable, especially an assessment of the desired accuracy and precision, and availability of suitable reference materials, the hierarchy of sampling replication and, for data collected at a fixed geographical location, mesoscale horizontal variability.

The HOT program was initially conceived as being a deep-ocean, ship- and mooring- based observation experiment that would have an approximately 20-year lifetime. Consequently, we selected a core suite of environmental variables that might be expected to display detectable change on time scales of several days to one decade. Except for the availability of existing satellite and ocean buoy sea surface data, the initial phase of the HOT program (Oct 1988 - Feb 1991) was entirely supported by research vessels. In February 1991, an array of five inverted echo sounders (IES) was deployed in an approximately 150 km² network around Station ALOHA (Chiswell, 1996) and in June 1992, a sequencing sediment trap mooring was deployed a few km north of Station ALOHA (Karl, 1994). In 1993, the IES network was replaced with two strategically-positioned instruments: one at Station ALOHA and the other at Station Kaena. Except for brief service intervals, both the Station ALOHA IES transducer and ALOHA sediment trap mooring have been collecting data since their respective initial deployments.

We evaluated several major criteria prior to selection of the site for the HOT oligotrophic ocean benchmark hydrostation. First, the station must be located in deep water (>4000 m), upwind (north-northeast) of the main Hawaiian islands and of sufficient distance from land to be free from coastal ocean dynamics and biogeochemical influences. On the other hand, the station should be close enough to the port of Honolulu to make relatively short duration (<5 d) monthly cruises logistically and financially feasible. A desirable, but less stringent criterion would locate the station at, or near, previously studied regions of the central North Pacific Ocean, in particular Station GOLLUM.

After consideration of these criteria, we established our primary sampling site at 22' 45'N, 158' 00'W at a location approximately 100 km north of the island of Oahu, and generally restrict our monthly sampling activities to a circle with a 6 nmi radius around this nominal site. Station ALOHA is in deep water (4750 m) and is more than one Rossby radius (50 km) away from steep topography associated with the Hawaiian Ridge. We also established a coastal station W-SW of the island of Oahu, approximately 10 km off Kahe Point (21' 20.6'N, 158' 16.4'W) in 1500 m of water. Station Kahe serves as a coastal analogue to our deep-water site and the data collected there provide a near-shore time-series for comparison to our primary open ocean site. Station Kahe is also used to test our equipment each month before departing for Station ALOHA, and to train new personnel at the beginning of each cruise. From January 1997 to October 2000, a physical-biogeochemical mooring was deployed to obtain continuous measurements of various atmospheric and oceanographic parameters. The mooring was located at 22' 28' N, 158' 8' W and was designated as Station HALE-ALOHA.

In August 2004, HALE-ALOHA was redeployed at a site 6 nautical miles west of Station ALOHA (22' 46' N, 158' 5.5' W) as part of the Multi-diciplinary Ocean Sensors for Environmental Analyses and Networks (MOSEAN) project. MOSEAN is directed toward new technologies that will lead to increased observations that are essential for solving a variety of interdisciplinary oceanographic problems. These include: biogeochemical cycling, climate change effects, ocean pollution, harmful algal blooms (HABs), ocean ecology and underwater visibility. This site, also called Station 51, is a collaboration with the University of California Santa Barbara and WET Labs.

Also in August 2004, a surface mooring outfitted for meteorological and oceanographic measurements was deployed 6 nautical miles east of Station ALOHA (22' 46' N, 157' 54' W). Ever since, CTD casts have been taken during various HOT cruises near the mooring for calibration of the moored instruments. This site, named WHOTS is a collaboration with the Woods Hole Oceanographic Institution. It has also been called Station 50. It is intended to provide long-term, high-quality air-sea fluxes as a coordinated part of the HOT program and contribute to the goals of observing heat, fresh water and chemical fluxes. The approach is to maintain a surface mooring by successive mooring turnarounds (Plueddemann, et al., 2006).These observations will be used to investigate air sea interaction processes related to climate variability.

HOT program cruises are conducted at approximately monthly intervals; the exact timing is dictated by the availability of research vessels. To date, our field observations have not been severely aliased by month, season or year, except perhaps for a slight under representation of data collected during April and slight over representation in October (Karl and Lukas, 1996). From HOT-1 (October 1988) to HOT-65 (August 1995), with the exception of HOT-42 and HOT-43 (November and December 1992), each cruise was 5 days in duration. Beginning with HOT-66 (September 1995) the standard HOT cruise was reduced to 4 days in order to accommodate additional mooring-based field programs within a fixed per annum allocation of ship days.

From HOT-1 (October 1988) to HOT-32 (December 1991), underway expendable bathythermograph (XBT; Sippican T-7 probes) surveys were conducted at 13 km spacing on the outbound transect from Station Kahe to Station ALOHA. These surveys were later discontinued because the space-time correlation of the energetic, internal semi-diumal tides made it difficult to interpret these data. From February 1995 until December 1997 we added an instrumented, 1.5 m Endeco towfish package (Sea- Bird CTD, optical plankton counter, fluorometer) to our sampling program (Tupas et al. 1997). Upper water column currents are measured both underway and on station using a hull-mounted Acoustic Doppler Current Profiler (ADCP), when available (Firing, 1996).

Underway near-surface measurement of a variety of physical, chemical and biological properties were made possible by sampling seawater through a pumped intake system positioned in the hull of the R/V Moana Wave. In May 1995, a thermosalinograph was installed in line to the ship's seawater intake system. In July 1996, the existing system was replaced with a noncontaminating PVC/stainless steel system. A flow-through fluorometer was installed in 1996. The R/V Ka'imikai-o-Kanaloa is outfitted with a similar seawater intake system to which the existing instruments were installed when R/V Moana Wave was retired. The R/V Kilo Moana also has a similar system.

The majority of our sampling effort, approximately 60-72 h per standard HOT cruise, is spent at Station ALOHA. High vertical resolution environmental data are collected with a Sea-Bird CTD having external temperature (T), conductivity (C), dissolved oxygen (DO) and fluorescence (F) sensors and an internal pressure (P) sensor. A Sea-Bird 24-place carousel and an aluminum rosette that is capable of supporting 24 12-L PVC bottles are used to obtain water samples from desired depths. The CTD and rosette are deployed on a 3-conductor cable allowing for real-time display of data and for tripping the bottles at specific depths of interest. The CTD system takes 24 samples s^-1 and the raw data are stored both on the computer and, for redundancy, on VHS-format video tapes.

In February 2006, before cruise 178, we replaced our 24 aging 12-L PVC rosette bottles with new 12-L bottles fabricated at the University of Hawaii Engineering Support Facility, using plans and specifications from John Bullister (PMEL).

Up until HOT-96 (August 1998), we routinely conducted a dedicated hydrocast to collect "clean" water samples for biological rate measurements, using General Oceanics Go-Flo bottles, Kevlar line, a metal-free sheave, Teflon messengers and a stainless steel bottom weight. During HOT-97 through HOT-118, due to the frequency of mis-trips & the inability to know the exact depth from which samples were collected, replicate samples were taken from the CTD rosette and the Go-Flo bottles. Comparisons with the Go-Flo collected samples showed there was no statistical difference in rates of ¹⁴C-primary production derived from samples collected using the Go-Flo bottles or the CTD rosette. As a result, beginning with HOT-119 (October 2000), we have collected samples for biological rate measurements only from the rosette.

A free-drifting sediment trap array, identical in design to the VERTEX particle interceptor trap (PIT) array (Knauer et al., 1979), is deployed at Station ALOHA for an approximately 60 hour period to collect sinking particles for chemical and microbiological analyses.

Sampling at Station ALOHA typically begins with sediment trap deployment followed by a deep (>4700 m) CTD cast and a "burst series" of at least 13 consecutive 1000m casts, on 3-hr intervals, to span the local inertial period (~31 hr) and three semidiurnal tidal cycles. The repeated CTD casts enable us to calculate an average density profile from which variability on tidal and near-inertial time scales has been removed. These average density profiles are useful for the comparison of dynamic height and for the comparison of the depth distribution of chemical parameters from different casts and at monthly intervals. For example, by fitting the distribution of inorganic nutrients to this average density structure, the depth of the nutricline can be defined each month, independent from the short time scale changes in the density structure of the upper water column (Dore and Karl, 1996). This sampling strategy is designed to assess variability on time scales of a few hours to a few years. Very high frequency variability (<6 hr) and variability on time scales of between 3-60 days are not adequately sampled with our ship-based operations.

Water samples for a variety of chemical and biological measurements are routinely collected from the surface to within 10 m of the seafloor. To the extent possible, we collect samples for complementary biogeochemical measurements from the same or from contiguous casts to minimize aliasing caused by time-dependent changes in the density field. This approach is especially important for samples collected in the upper 350 m of the water column. Furthermore, we attempt to sample from common depths and specific density horizons each month to facilitate comparisons between cruises. Water samples for salinity determinations are collected from every water bottle to identify sampling errors. Approximately 20% of the water samples are collected and analyzed in duplicate or triplicate to assess and track our precision in sample analyses.

The suite of core measurements provides a database to validate and improve existing biogeochemical models. Our list of core measurements has evolved since the inception of the HOT program in 1988, and now includes both continuous and discrete physical, biological and chemical ship-based measurements, optical, in situ biological rate experiments, and observations and sample collections from bottom-moored instruments and buoys. Continuity in the measurement parameters and their quality, rather than continuity in the methods employed, is of greatest interest. Detailed analytical methods are expected to change over time through technical improvements. In addition to the core data, specialized measurements and process-oriented experiments have also been conducted at Station ALOHA.