Thermal variation intra and interannual subsuperficial near the Tropic of Cancer (Mazatlán, Mexico).
David Serrano and Arnoldo Valle Levinson
ABSTRACT…
1.INTRODUCCTION
Sea temperature is one of the most relevant oceanographic parameters associated with climate and weather systems. Sea surface temperature is widely used in studies for atmospheric and ocean problems, including oceanic environmental management andfisheries.Subsurface temperaturehas a complex and variable vertical structure that is related to air-sea fluxes of heat, atmospheric forcing, ocean turbulence,advective processes, internal waves andvariations of sea surface level among others. Changes in mixed layerdepth, thermocline and in the depth of a specific isothermare usually analyzed to describe the spatiotemporal variations of subsurface temperature seawater. Variations in the sea surface height also are related to fluctuations of the thermocline in the vertical level.Despiteof their relevancein climate studies, oceanographic observations in situ are not sufficient to fully represent the kinematics and thermal states of the oceans.
Seasonal, interseasonal and longer timescale variations in sea temperature (surface and subsurface) have been described and analyzed in different regions of the world.In the South China Sea, Zhou and Gao (2002)concluded that the subsurface temperature intraseasonal variability is determined by the vertical displacement of the thermocline, and produces subsurface temperature variations of large amplitude on the same time scales. In addition, Ishii et al. (2003) founded that dynamic height anomalies estimated with analyzed temperature and climatological salinity was highly correlated in the tropical Pacific with sea level height observations.Moreover, in the western Pacific Ocean for the 1993–1998 period, along 165ºE using conductivity-temperature-depth (CTD)measurements, Maes and Behringer (2000) showed that the altimetry observations such as from TOPEX/Poseidon was representative of the thermal variability as well as the haline variability, and that altimetry signal may be projected downward into the upper oceanic layers. In the North Atlantic Hughes et al. (2009)used time series of sea surface temperature (SST) and sea subsurface temperature,for demonstrate that variations in the surface temperature are reflected in the subsurface sea condition, they concluded that surface temperature in situ may or may not have a close relationship with subsurface temperature; therefore, it is important to consider the local hydrographyusing SST data as an approximation of subsurface conditions.
For the Mexican Pacific coast, several studies describe relevant information on the subject; for example, in an ecological studyCarballo and Nava (2007) described the sea surface temperature throughout 2001-2003 in the Bay of Mazatlán,they determined that there is a wide temperature range along the seasonal cycle, reaching 12°C (January-August 2001).In Bahía Concepción,a bay on the west coast of the Gulf of California López-Cortés et al. (2003)found that the subsurface temperature in the bay for the year 1998 was modified by El Niño,recording higher temperatures with respecting values for years 1997 and 1999, along witha 2 to 3 month delay in the stratification period.In a subsequent study, conducted in 2005 in the same bayCheng et al. (2010), studied the water temperature variation along a seasonal cycle, they found that the temperature in the water column remains almost homogeneous in the winter whereasin the summer is stratified; they also determined that the advection of cold waters from the Gulf of California has an important role in the thermal structure in the water column.
Moreover, to analyze the spatiotemporal variation of the subsurface thermal structure of the ocean, several authors have used time series analysis; empiricalorthogonal function; the calculation of the heat fluxes at the ocean-atmosphere interface,thus as surface sea level variations in relation to fluctuations in time and space of thethermocline. For example, in the Canary and IberianBasins of the North Atlantic,Müller and Siedler (1992)used an array of 22 subsurface mooringsand determined that the vertical structure can be well approximated by the barotropic and first-order baroclinic dynamical modes as well as with an empirical orthogonal function.Along the Atlantic Ocean, about 26ºN,Szuts et al. (2012)found that subsurface fluctuations in the center of the basin are large, are well described by a first baroclinic mode, have long periods and this signals are accurately described by sea surface height.In the South China Sea Liu et al. (2001)determined that at three oceanographicstations, the amplitudes of the time series of the 22ºC isotherm depth were two orders of magnitude greater than the sea surface height. They also concluded that the first baroclinic mode, represented by the ratio between sea surface height and thermocline depth anomaly, is dynamically important in the central South China Sea.In the Tropical Pacific OceanSmith and Chelliah (1995) using harmonic analysis showed that in general, subsurface temperature variations are much larger than surface variations, and that most of subsurface variations are associated with changes in thedepth of the thermocline.Cheng et al.(2010)calculated the net heat flux at ocean-atmosphere interface and used the heat content in the water column for determining the horizontal advection coming from the Gulf of California.
The present study, wasconducted using four years of water column observations in the vicinity of Mazatlán, Sinaloa, (Fig. 1), located in the Pacific Mexican, at the mouth of the Gulf of California. The main objectives for this study were to analyze the changes oftemperature, salinity and density in time and space and determine the role of sea surface height fluctuations in the thermocline;as well as the role of different water masses andthe consequences of advective processes in the heat content of the water column. Section 2 briefly describes the area of study and data collection.Section 3, describes the changes in time and space fortemperature and salinity in the water columnfrom the surface to 127 m depth,and also identifies the main harmonicsthat make up the time series and the dominant modes of temperature variability are presented.The relationship between the sea surface height and the isotherm of 18°C are addressedand the heat content in the water column and the net heat flux are also shown in this section. Finally the discussion and the conclusions are presented in the section 4 and 5 respectively.
2.STUDY AREA AND DATA COLLECTION
Mazatlán Bay is localized in northwestern Mexico in the coastal of the state of Sinaloa, located at 23°12’N and 106º25’W in the southeastern Gulf of California (Fig. 1),this region is influenced by the water masses of the Gulf of California,Subtropical Subsurface water, Tropical Surface water and California Current water (Castro et al. 2000, Hendrickx and Serrano, 2010). The tide in the bay is mixed, mainly semidiurnal, with form number F = 0.575. The maximum air temperature can reach 40ºC (August-September) and the minimum temperature 15°C(January-February) -using data from the weather station at the MazatlanAirport. The influence of tropical meteor occurs from May to November; at the entrance of the Gulf of Californiahave been recorded from December to March winds from the northwest, reaching magnitudes of 12 ms-1(Douglas et al. 1993).
Hydrographic data was obtained from 49 oceanographic surveys carried out from August 2005 to August 2009. CTD-SEABIRD-9 casts were realized in one station located at 23º05’N and 106º36’W to ~20 km of coastline (Fig. 1). The samplings were carried out on board the boat Miztli, of ICMyL. The sampling interval was about a month; the station depth was 127 m; sampling frequency in the vertical was 2 Hz;the temperature and salinity were interpolated each meter depth in the vertical and all readings weredone between 9 and 11 a.m.The data analysis in this paper was realized each meter depth (127 time series);constructed from 49surveys of the water column carried out with the CTD. The time series were constructed using piecewise cubic Hermite interpolation,with Δt= 5 days, the interpolation generated matrices of 127 rows by 294 columns of temperature, salinity and.In order to relate the surface changes of sea level with the subsurface temperature structure, a time series of region (23ºN 106º30’W) from September 2005 to September 2009of sea surface height (SSH) was used with readings available from TOPEX/Poseidon satellite ( order to examine the horizontal advection,the heat content of the water column wascalculated. To identify the external forcing heat,the net heat flux was calculated.To approximate the heat flow of the area ofstudy, meteorological data from MazatlanAirport (23º10’N and 106º16’W)formAugust 2005 to August 2009were used.( in addition, the time series of sea surface temperaturewas elaborated using data from NOAA( heat content (HC, in J m-2) of the water column was calculated in accordance with Cheng et al. (2010)
Where is the specific heat of sea water, is water density, is water temperature, is a reference temperature, which is arbitrarily set to zero ºC, is the depth of the water column (127 m); was calculated in agreement with Millero et al. (1973).
The net flux heat of the surface (W m-2) was calculated as the sum of:
Where is the net radiationheat flux, or the difference between the incomingheat gains and outgoing radiationlosses through the air-sea interface at the earth’s surface, is the net radiation flux was calculatedwith as the difference between the net downward flux of solar radiation and the net upward flux of long-wave radiation from the ocean (Gill 1982), is the sensible heat flux due to air-sea temperature differences, and is the latent heat flux due to water vapor transport.
The sensible and latent heat fluxes were estimated using the standard bulk formulae from the tropical ocean global atmosphere/coupled ocean atmosphere response experiment (Fairall et al. 1996)
aAnd
Whereis the air density, is the specific heat of the air, is the sensible heat transfer coefficient (Stanton number), is the wind speed, is the air temperature, is the latent heat of evaporation of sea water, is latent heat transfer coefficient, is the specific humidity at air temperature and is the specific humidity at sea surface.Finally, with the purpose of identifyingthe water masses that constitutein the water column, a TS-diagram was constructedusing data form the 49surveys.and Power spectra were calculated to determine the most important harmonics that formed the 127 mtime series of temperature,the power spectra were calculated foratdepths of 1, 30, 60, 90 and120 m depth[A1].
3.RESULTS
3.1.Temperature, salinity and
The time series of temperature for a four year period (2005-2009)from the surface to 127 m depth are shown in Figure 2a.In the time series, three full and two partial seasonal cycles full and two partial seasonal cycles are notoriousevident.The maximum surface temperature of 31.47°C was recorded in August 2009and the minimum temperature of 12.25°C in December 2007 at127 m depth[A2].;Aton the surface, the minimum temperature was 18.32 ºC recorded on the same dayalso in December 2007. The average surface temperature throughout the observation period was 26°C. Subsurface(50-80 m depth) intrusions of cold water without apparent external thermal forcing (atmospheric heat fluxes) was were recorded between September 2006 and February 2007,andalso between August 2008 and February 2009.The isotherm of 18°Ccomes can be regarded as the lower limit of the thermocline (-0.08ºC m-1).
The time series for salinity in the water column is shows a faint seasonal cyclenin (Figure 2b). The seasonal cycle is not clear[A3]; however, the higher Highest salinities were recorded in May and June between 40 m and the surface, with a;the maximum of was35.35 g kg-1. The lowestr salinitiesyoccurred between July and November (the wet or tropical season).;Tthe minimum value was 31.08 gr kg-1recorded at the surface in October 2008. The salinity intervalthrough time decreases with depth[A4], with 4.27, 0.87, 0.77, 0.41 and 0.24 g kg-1 for 1, 30, 60, 90 and 120 m, respectively. The fact that the highest temperatures coincided with the lowest salinities suggested that these relative warm and fresh waters were associated with river plumes advected from shore. This is the first time that freshwater influence is reported in this part of the entrance to the Gulf of California.
The time seriesphase diagram (depth vs time) of throughout four years and from the surface to 127 m depth is shown in Figure 2c. The distribution ofDistribution was similar to the temperature distribution[A5]. Four seasonal cycles are clear. Values less than 20 kg m-3 were recorded between September and October 2008. The greatest Significant fluctuations were recorded between September 2006 and February 2007 and between August 2008 and February 2009 between from 40 and to 60 m depth (corresponding to fluctuations in temperature also registered).The vertical lines [A6]on 2a, 2b and 2c indicated the depth to it was recordedassociated with the Current of California[A7]water.,Tthis occurred only between August and January, with depths between 35 m, December 2007 and 105 m, in August 2009[A8].
The TS diagram shown in Fig. 3 [A9]indicates the water masses recorded in the water column during the study.This diagram was conformed mainly by the following water masses: Tropical Surface and Subtropical Subsurface, with scarce recordsfew observations of California Current and Gulf of California.The greater stratification occurred between July and October and a relatively weak stratification occurred between the months of December to and March.
The relative warm and cold periods spells of the observation period for the water column were determined for the water column by with thetemperature difference with the average temperature for each meter depthanomaly from the average temperature profile(Fig.4a[A10]).The four cold cool periods on at the surface lastedbetween 195 and 150 days,in winter 2005-2006 and 2008-2009, respectively.,with aThe mean duration for the four cool periods of was 176 days. The three warm periods lasted between 205 and 179 days, for summer of 2006 and 2007, respectively.,with anThe average ofwarm period was 189 days.Moreover, the dates where when surface warm periodsstarted began were: 05-18-06, 05-22-07, 05-31-08 and 05-01-09, also, the dates where when cold cool periods started were: 11-04-05, 12-09-06, 11-17-07and 12-20-08.
It should be noted that tThe duration of the warm and cold periods wasnot homogeneous the same in the water column.,Ffor example,;for the first warm period at the 80 mdepth tooklasted for 270 days,with two cold cool periods lasting approximately one month each (Fig. 4a). The second warm period at the 80 mdepth lasted for 70 days.The third period was more homogeneous upsimilar from the surface to ~100 m.;Nnevertheless, in thise third period, a cooling period was observed between 60 and 80 m with a duration of ~1 month.At the bottom (127 m) this third warm period lasted for 80 days.Additionally, the main largest positive differences from the averageanomalies(for the z depth) (Fig. 4a) were recorded between 30 and 60 m depth, with values of 8ºC. The largest negative differences anomalies were recorded observed between 60 m and the surface, reaching values of -6ºC.
The negative temperature gradient in the vertical is shown in Figure 4b.The isoline of 0.08ºC m-1was chosen as the lower limit to determine the zone of the biggest changes in the vertical temperature (base of thermocline). The thermal stratification was recorded observed at different depths throughout four years.,Stratification even reacheding the bottom in three warm periods:,June-September 2006, July-August 2008 and June-September 2009.;the exception wasDuring the second warm periodwhere the isoline of 0.08ºC m-1did not reach the bottom as it was located at 100 m depth,during forSeptember-October 2007.Values greater than 0.4°C m-1 were recorded between 10 and 50 m depth, located in the switchat the transitionbetween from warm to- cold to -warm periods. Values lower than0.08ºCm-1 were recorded found in cold periods at depths greater than 80 m. The differences in temperature over time for the any depth z is are showned in Figure 4c.Differences Values greater than0.25°C day-1occurred from the surface to 80 m depth; these differences were located in during the warm-cold-warm transitions. These were period and the apparently associated with theadvection of warm or cold water.
3.2.Power spectral density and harmonic analysis (HA)
The temperature power spectra um of temperature for the depths of 1, 30, 60, 90 and 120 mdepth is are shown in Figure 5[A11]. A peak located was found between 0.5 and 1 cycles per year (CPY)was evident in the five spectra (ºC2/CPY), although the amplitude of this the peak decreaseds with depth. A peak near the frequency 2 CPY wasevident in the five spectra, increasing but its relative amplitude as the depth increasesincreased with depth, reaching almost the same amplitude as the signal of 1 CPY to at 120 m depth. Peaks with frequencies between 2.5 and 4 CPY werenotorious appreciable for depths of 30, 90 and 120 m.
The variationfor of the water column temperature for MazatlánBay is represented as the sum of the harmonics quadrennial, biennial, annual, biannual, four-monthly and three-monthly [A12]for each meter of the water column with the following equation.
(1)
whereis the average temperature at the depth; are the amplitude, frequency and phase of the harmonics mentioned above. The amplitude is defined asand.Where and were calculated using the discrete Fourier transform (Jenkins and Watts, 1969;Bendat and Piersol, 1986).The amplitudes and phases of the six harmonics from surface to 127 m depth are shown in Figure 6.