Tải bản đầy đủ - 0 (trang)
1 The Large-Scale Setting, Water Masses and Ventilation

1 The Large-Scale Setting, Water Masses and Ventilation

Tải bản đầy đủ - 0trang


3 Large-Scale Setting, Natural Variability and Human Influences

Fig. 3.1 The structure of the large-scale wind-driven circulation of the ocean. Note that the

circulation during the southwest monsoon is shown in the Indian Ocean. Taken from Thurman and

Trujillo (1999)

thermocline and nutricline. For example, all subtropical gyres are characterized by

an eastward upward-doming of the thermocline and nutricline, which brings

nutrient-rich water close to the euphotic zone near their eastern boundaries

(Fig. 3.2). It is clear that this shoaling of the nutricline plays a key role in the high

productivity of the big four coastal upwelling systems (see Chap. 10). Theories of

the ventilation of the thermocline, which is intrinsically coupled with the distribution of the oxygen minimum zone (e.g., Brandt et al. 2010), were first developed

by Luyten et al. (1983).


Source Depth of Upwelled Water and Water Masses

Upwelled nutrient-rich water typically comes from relatively shallow depths of

100–200 m. The properties (temperature, salinity, oxygen and nutrient levels) of

water masses at that depth are often controlled by the large-scale circulation rather

than by regional processes. Upwelled water in the four major eastern boundary

coastal upwelling systems, for instance, is derived from complex current systems

that involve relatively oxygen-poor, nutrient-rich water of equatorial undercurrents

feeding into poleward undercurrents that follow the shelf break and provide source

waters to the cross-shore upwelling circulation (see Chap. 10).

Shelf-break currents can also operate as a preconditioning process for the

classical wind-driven upwelling mechanism. Examples are the nutrient-rich intermediate water of the Kuroshio Current which interacts with the shelf break of the

East China Sea (see Sect. 8.2.2) and the Flinders Current that, as a result of its

3.1 The Large-Scale Setting, Water Masses and Ventilation


Fig. 3.2 Large-scale structure of the nutricline a at a depth of 100 m, and b along the equator in

the Pacific Ocean. The dashed white line highlights the tilt of the 15 lM/L nitrate contour. Taken

from Pennington et al. (2010)

interaction with shelf-break canyons at the shelf edge, brings nutrient-rich

sub-surface water onto the southern shelves of Australia (see Sect. 8.2.4).

Water-mass analysis is a technique that allows us to identify the origins and

pathways of individual water masses in the ocean (see Tomczak and Godfrey

2003). Individual water masses, which can be distinguished from each other

through e.g., temperature/salinity or other chemical characteristics, are formed via

air-sea exchanges of heat and freshwater in specific regions of the world ocean

where they also receive their initial chemical signature via gas and particle transfers

and biogeochemical processes in the surface mixed layer. In their formation

regions, water masses are confined to the surface mixed layer from where they are

pushed down or “subducted” into the ocean interior. Subduction takes place either

in the form of density-driven gravity flows or is induced by external forcing such as

the wind-induced convergence of Ekman flows in a region with outcropping

isopyncals (see Tomczak and Godfrey 2003). As water masses in the ocean interior

are transported away from their formation regions, they accrue dissolved nutrients

from the remineralization of sinking detritus and their oxygen concentration drops

(see Chap. 2).


3 Large-Scale Setting, Natural Variability and Human Influences

Water masses formed by subduction in the open ocean that are relevant to

upwelling regions are:

• Central Water, i.e. waters within the permanent thermocline and nutricline, and

• Intermediate Water, typically found below the permanent thermocline.

Central waters generally have relatively high nutrient levels, but nutrient concentrations vary significantly in the vertical and, to some extent, also laterally, due

to the ventilation history of individual water masses. Figure 3.3 shows the distribution of different types of Central Water, which can be distinguished by their

water-mass properties. There are also so-called ‘Mode Waters’ which refer to layers

of nearly vertically homogeneous water found over a relatively large geographical

area. After subduction, mode waters usually occur within or near the top of the

permanent pycnocline, and hence are apparent through the contrast in stratification

within the pycnocline waters. Mode waters are defined by the properties of water

masses at their formation region. Mode waters can also contribute to the formation

of Central Water (Hanawa and Talley 2001).

Intermediate Water is typically found somewhat deeper at depths <1000–1500 m,

but, in some instances, western boundary currents such as the Kuroshio Current can

dynamically lift this nutrient-rich water up across the shelf break and onto the continental shelf (Chen 1996). It is clear that the source depth of upwelled water plays an

important role in the biogeochemical cycles in upwelling systems by determining the

eventual concentrations of nutrients in the water that reaches the surface. Water mass

analysis is also key to inferring changes in the large-scale circulation, which is indeed

important for the understanding of changes in upwelling-induced nutrient fluxes.


Water Mass Properties of Upwelling Water

The proportions of different water masses that make up the water properties in a

region can often be inferred from temperature and salinity data. In turn, this

Fig. 3.3 Global distribution of upper water masses (0–500 m) (from Lalli and Parsons 1993)

3.1 The Large-Scale Setting, Water Masses and Ventilation


information gives clues about chemical characteristics and the circulation of water

masses, and the source depth from which the water is upwelled. In contrast to the

open ocean, coastal ocean water-mass properties are markedly influenced by

amplified seasonal cycles of temperature and salinity because of their shallow

depths and continental influences such as river discharge or outflows of hypersaline

water from inverse estuaries. Thus it is no surprise that measurements of temperature and salinity in coastal waters are highly variable and often difficult to interpret.

Nevertheless, this task is easier for upwelling regions, given that upwelling events

leave behind a pronounced temperature signature. In upwelling regions, in situ

temperature (and salinity) measurements can give information on the source water

properties including nutrient levels of upwelled water.

For instance, let us briefly consider the seasonal coastal upwelling system that

develops on the southern shelves of Australia during austral summer months (Kämpf

et al. 2004). Measurements undertaken during an upwelling event in the upwelling

centre of the region indicate that upwelled water has temperatures <12 °C

(Fig. 3.4a). This water can be traced back to an average depth of *310 m

(Fig. 3.4b) within the nutricline of Subantarctic Mode Water that the Flinders

Current carries westward along the adjacent upper continental slope (Kämpf 2010).

Fig. 3.4 Panel a Symbols are in situ measurements of temperature and salinity of bottom shelf

waters during an upwelling event in the upwelling centre off the southern coast of the Eyre

Peninsula, South Australia. Note how upwelling water tends to get warmer and saltier towards the

surface. Coastal upwelling is fed by shelf waters formed in an adjacent dense-water pool, known as

the Kangaroo Island Pool. The T-S properties of this pool (black dots) are derived from a

hydrodynamic model application. Source Kämpf (2010). Panel b shows vertical profiles of

temperature and nitrate measured in the open ocean adjacent to the upwelling centre. FC is the

Flinders Current. Ellipses indicate the source depth and nutrient level of upwelled water. Modified

from Richardson et al. (2009)



3 Large-Scale Setting, Natural Variability and Human Influences

Seasonal Variability

If the surface ocean was at steady state, with an assumed constant nutrient supply

and an unvarying surface mixed layer having the thickness of the euphotic zone, the

production of phytoplankton in the surface ocean would simply follow the seasonal

cycle of insolation. In reality, both density stratification, which affects the surface

mixed layer thickness, and changes in nutrient availability interrupt this cycle. For

example, a deeper surface mixed layer reduces the contact time of phytoplankton

cells with the euphotic zone. On the other hand, nutrients become depleted over

time in a shallow surface mixed layer in the absence of nutrient supply mechanisms

such as wind-induced entrainment, upwelling or internal waves.

In the open North Atlantic Ocean, for instance, deep convective winter mixing

brings nutrients to the upper ocean without significant consumption. The establishment of a shallow surface mixed layer with reduced turbulence during spring

triggers a widespread phytoplankton spring bloom which recedes in the following

months due to nutrient limitation (Fig. 3.5). A zooplankton maximum follows

during early summer. Autumn storms and atmospheric cooling lead to the

entrainment of nutrient-enriched water into the surface mixed layer, again triggering

a transient phytoplankton bloom. This bloom recedes in winter months given

reduced insolation and substantial convective deepening of the surface mixed layer.

In stark contrast to the two-peak bloom feature of the North Atlantic, the North

Pacific and Arctic waters only experience a single phytoplankton bloom in late

summer (Fig. 3.6). In the North Pacific, the lack of a spring diatom bloom may be

due to iron deficiency (e.g., Gregg et al. 2003), while in the Arctic Ocean the

problem is light limitation and persistently strong density stratification (e.g., Popova

et al. 2010). In tropical waters, there is little seasonal variation in insolation or

density stratification in the upper ocean. Hence, there are continuous levels of

primary and secondary production year-round outside seasonal coastal upwelling

systems (Cushing 1975). In shelf seas subject to nutrient input from upwelling or

Fig. 3.5 Schematic of ambient conditions that trigger the development of spring and fall (autumn)

blooms in the North Atlantic

3.2 Seasonal Variability


Fig. 3.6 Seasonal cycles of

phytoplankton (algae) and

zooplankton (herbivores) in

different regions of the ocean.

Modified from Cushing

(1975). a Arctic. b North

Atlantic. c North Pacific.

d Tropical.

fluvial sources, seasonal cycles of phytoplankton concentration and zooplankton

abundance can be dramatically different when compared to the ambient ocean.



Climate Variability and Climate Change

Modes of Climate Variability

Climate variables are properties of the climate system that influence climate from

regional to global scales. Sea surface temperature, for instance, is a key oceanic

climate variable in the surface heat-flux budget. How global climate varies depends

on the way these climate variables change on timescales of 2–10 years, or even

longer. Each zonal wind system of the general atmospheric circulation can be

associated with a specific mode of climatic variability which is determined statistically from long-term records. For instance, two long-term climate oscillations

have been recognized in the northern hemisphere, the North Atlantic Oscillation

(NAO) and the Arctic Oscillation, also called Northern Annular Mode (NAM), that

appear to be governed by changes in the boreal westerly winds on decadal-scale


3 Large-Scale Setting, Natural Variability and Human Influences

time periods. Changes in the trade winds on 2–5 year time-scales give rise to

another identifiable oscillation, the El Niño Southern Oscillation (ENSO) , whereas

variations of the austral Westerlies—the Roaring Forties—can be described by

Southern Annular Mode (SAM), also known as the Antarctic Oscillation.

Interactions between these climatic modes are referred to as teleconnections.

Because wind speed is driven essentially by atmospheric pressure differences,

these climatic modes are mainly described by atmospheric indices (i.e., the sea level

pressure difference between two selected regions). There is, however, another class

of climatic modes, based on seawater temperatures, that reflects the ocean’s internal

dynamics. This includes the Pacific Decadal Oscillation (PDO) , the Atlantic

Multidecadal Oscillation (AMO), the Antarctic Circumpolar Wave (AACW) and

the Indian Ocean Dipole (IOD).

Oceanic climatic modes typically take place over longer time periods than the

atmospheric modes, and their cycles are typically of the order of one or more

decades. This reflects the much greater viscosity of the ocean compared with the

atmosphere and the slower rates of ocean mixing processes. While there is considerable interest in whether variations in these indices have direct effects on

upwelling systems, and if so, which, the intrinsic variability of ocean-atmosphere

interactions, the multiplicity of potential interactions between different modes, and

the scarcity of data from most regions makes it difficult to determine how large such

effects are (e.g., Reason et al. 2006). Here, we only discuss those modes of variability that are likely to impact the main upwelling regions, so will omit those in the

high latitudes.

The North Atlantic Oscillation (Walker and Bliss 1932; van Loon and Rogers

1978; Wallace and Gutzler 1981) is a mode of climatic variability that depends on

modulation of the strength and storm tracks of the boreal Westerlies. Thus, it

strongly influences the weather in North America and northern Europe. It is defined

in terms of the pressure difference between the Azores (high pressure) and Iceland

(low pressure), and positive NAO indices (large pressure differences) result in

stronger winter storms with a more northerly track, leading to warm wet winters in

Europe and cold, dry winters in northern Canada and Greenland.

The negative (low pressure difference) phase of the NAO results in fewer,

weaker storms that result in a cold northern Europe and a wet Mediterranean.

Although the index varies from year to year, it can become fixed in a positive or

negative mode for several years; thus the period from 1955–1970 was consistently

negative, while during 1970–2000 it was generally in the positive phase. The NAO

also influences the formation of North Atlantic Deep Water—the key driver of the

ocean’s deep circulation—in the Greenland and Labrador Seas (e.g., Dickson et al.

1996), and biological production in the northern Atlantic through storm-induced

interference with the spring bloom (e.g., Zhai et al. 2013). Nonetheless, the North

Atlantic Oscillation has only little impact on the average intensity of coastal

upwelling along the coasts of northwest Africa and Portugal (Narayan et al. 2010).

In the 1920s, Gerhard Schott and Erwin Schweigger in the 1920s (Cushman

2004) and later Klaus Wyrtki (1975) deserve recognition for interpreting oceanic

features of the El Niño Southern Oscillation, which has a profound influence on

3.3 Climate Variability and Climate Change


upwelling off Peru. Later came the first descriptions of decadal-scale variability

associated with the Pacific Decadal Oscillation (PDO) (e.g., Beamish 1993; Latif

and Barnett 1994; Francis and Hare 1994).

The El Niño Southern Oscillation (Karoly 1989; Philander 1990; Burgers 1999;

Barnett et al. 1999), of which the importance was first realized 500 years ago off

Peru, is a climatic mode of variability that influences most of the equatorial Pacific

Ocean and can at times extend into the Atlantic and Indian Oceans as well. It was

named El Niño in association with the coming of the Christ child, as it was usually

noticed in late December. The Walker Circulation (Walker 1923) is a vertical

circulation cell in the atmosphere that is set up along the equator in the Pacific

Ocean (Fig. 3.7a). This circulation consists of the trade winds blowing towards the

west near the sea surface and an eastward return flow of air at higher levels. On

average, warm and moist air rises over the Warm Pool in the western Pacific Ocean,

where it creates tropical cumulus clouds and enhanced precipitation. The location of

Fig. 3.7 Illustration of

processes involved in the

creation of an El Niño event.

Panel a shows the average

situation. Panels b and

c display the onset and full

development of the event.

Courtesy of Billy Kessler,

Pacific Marine Environmental

Laboratory/NOAA. http://


kessler/occasionally-askedquestions.html [accessed on 4

April 2016]


3 Large-Scale Setting, Natural Variability and Human Influences

this thermal updraft region strongly depends on sea surface temperature, which

modulates the flux of moisture (evaporation) into the atmosphere.

An El Niño event commences with a westerly wind burst in the western Pacific

Ocean and a reversal of the trade winds in this sector (Fig. 3.7b). This usually

creates a tropical cyclone pair in the lower atmosphere and an oceanic wave—an

equatorial Kelvin wave—that travels eastward along the equator at a speed of 2–

3 m/s (see Tomczak and Godfrey 2003). It takes *20–30 days for the wave to

cross the Pacific Ocean. This wave lowers the thermocline and suppresses equatorial upwelling along its path (Fig. 3.7c). As a consequence of this disturbance, the

Warm Pool and thermal updraft region of the Walker Circulation is shifted eastward

to the central equatorial Pacific, near the dateline, where it induces enhanced


As the wave hits the west coast of central America, it turns into a pair of coastal

Kelvin waves that propagate along the coasts and away from the equator both

northward and southward. Like the equatorial wave, the coastal Kelvin waves

operate to lower the thermocline along their path, which suppresses coastal

upwelling of cold, nutrient-rich water off the coasts of Chile, Peru and California,

even though the winds in the coastal region are still blowing towards the equator.

Eventually the waves move back towards the centre of the Pacific Ocean as a pair of

westward-propagating Rossby waves and the thermocline along the coast rises

again to “average” conditions.

While ENSO events are triggered by the westerly wind bursts, the process

vacillates irregularly between two phases, called El Niño and La Niña, which are

the two ends of a continuum. El Niño is the warm phase, while La Niña is the cold

phase. The La Niña phase implies an intensification of the regular trade winds (and

enhanced equatorial upwelling) and precipitation in the western equatorial Pacific.

The consequence of an El Niño event is a dramatic increase in sea surface temperatures in the eastern equatorial and tropical Pacific. Figure 3.8 shows an

example of the temperature changes during the 1997/98 El Niño event, which is one

of the strongest ENSO events in recorded history. Typically, ENSO events last from

about 6–18 months but can be very important in terms of their effects on rainfall,

agriculture, fisheries, and changes in biological populations, which can expand or

contract by up to 1,000 km. Recent strong El Niño events occurred in 1957–58,

1982–83 and 1997–98, with weaker events in 1987–88, 1991–92, and 2014–15,

although their occurrence appears to have become less regular since the 1982–83

event. A good introduction to ENSO and its effects on global oceanography is given

by Philander (1990), while specific accounts of the 1982–83 and 1998–99 events

are in Barber and Chavez (1983) and McPhaden (1999).

Similar El Niño type variability also exists in the Atlantic Ocean (see Tomczak

and Godfrey 2003; Shannon et al. 1986). Although westerly wind bursts in Brazil

are thought to control these Atlantic events, they do not seem to be linked directly

to those in the Pacific, and have only been recorded intermittently, in 1934, 1950,

1963, 1984, and 1995 (Shannon et al. 1986; Gammelsrød et al. 1998), although

additional warm events have also been recorded of lesser intensity, for example in

2001 and 2010 (Rouault et al. 2007; Bode et al. 2014; see Chap. 7).

3.3 Climate Variability and Climate Change


Fig. 3.8 Sea-surface temperature (SST) distributions in the equatorial Pacific Ocean for selected

months and years. Panels a and b show the transition between the La-Niña phase and the El Niño

phase of the ENSO in terms of actual SSTs. The impact of the late 1997 El Niño event is best seen

from SST anomalies (Panel c), which displays the SST difference between panels a and b. Panels

d and e show SST anomalies in late 1997 and 1998. Images courtesy of NASA

The intensity of the ENSO can be monitored by means of the atmosphere pressure

difference at sea level between Darwin (Australia) and Tahiti, with the pressure at

Darwin being low when the index is high, and vice versa. This gives the Southern

Oscillation Index (SOI) (Fig. 3.9). There are also oceanic indices, such the Oceanic

Niño Index (ONI), that record ENSO activity based on sea-surface-temperature

anomalies in defined areas of the central equatorial Pacific Ocean. Some El Niño

events have stronger impacts than others. For instance, the global flu pandemic of

1918/1919, which coincided with intense cold in the northeastern United States and

a crippling drought in India, is now blamed at least partly on such an event (Giese

et al. 2010), while the infamous 1997/1998 El Niño (McPhaden 1999) had

far-reaching impacts with droughts in the western pacific islands and Indonesia as

well as in Mexico and Central America and resulted in coral bleaching events in

most tropical regions of the world ocean (Wilkinson and Hodgson 1999). The other

recent strong El Niño, in 1982/1983, which was known at the time as the “El Niño of

the century” led to devastating flooding across Bolivia, Cuba, Ecuador, northern

Peru and in the U.S. states bordering the Gulf of Mexico, and spawned hurricanes in

Hawaii and Tahiti, as well as droughts and fires throughout southern Africa, central

America, Australia, India, Indonesia, and the Philippines (see, e.g., Gill and

Rasmussen 1983; Glantz 1996; Rasmussen and Wallace 1983).


3 Large-Scale Setting, Natural Variability and Human Influences

Fig. 3.9 Time series of the Southern Oscillation Index (SOI) . El Niño events correspond to

negative values. Note that the unit of the y-axis is based on the standard deviation of the data,

based on the normal distribution. Image source: http://www.cgd.ucar.edu/cas/catalog/climind/soi.

html [accessed 4 April 2016]

The Pacific Decadal Oscillation (PDO) is the key mode of oceanic (as opposed

to atmospheric) variability of the Pacific Ocean (e.g., Trenberth and Hurrell 1994),

and is calculated as the first principal component of an EOF analysis of North

Pacific sea surface temperature anomalies, based on a time series from 1900–1993.

The recent study of Shakun and Shaman (2009) indicates that the PDO is an

aftereffect of ENSO events; that is, the dominance of more frequent, stronger and

longer La Niñas leads to the negative phase of the PDO, whereas more frequent,

stronger and longer El Niños trigger the positive PDO mode. Accordingly, it not

surprising that the phases of the PDO have a similar appearance in their SST

anomaly distributions (Fig. 3.10).

Fig. 3.10 Anomalies of sea surface temperatures (colors) and surface winds (arrows) for positive

and negative phases of the Pacific Decadal and El Niño Southern Oscillation. Taken from http://

jisao.washington.edu/pdo/graphics.html [accessed on 12 April 2016]

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

1 The Large-Scale Setting, Water Masses and Ventilation

Tải bản đầy đủ ngay(0 tr)