International Journal of Geosciences, 2012, 3, 899-907
http://dx.doi.org/10.4236/ijg.2012.325092 Published Online October 2012 (http://www.SciRP.org/journal/ijg)
Accessing the Potential of Satellite and Telemetric Data to
Evaluate the Influence of the Heat Flux Exchange in the
Water Column Mixing and Stratification
Enner Alcântara
Department of Cartography, Sao Paulo State University (UNESP), Presidente Prudente, Brazil
Received August 3, 2012; revised September 1, 2012; accepted October 2, 2012
ABSTRACT
The objective of this work is to evaluate the feasibility of moderate resolution satellite data estimating the surface heat
balance in a tropical hydroelectric reservoir. Each component of the heat flux balance was computed using the MODIS
(Moderate Resolution Imaging Spectroradiometer) water surface temperature (WST) level 2, 1 km nominal resolution
data (MOD11L2, version 5) from 2003 to 2008. The consequence of the heat flux exchange in the water column ther-
mal structure is also investigated. The passage of cold front over a region decreases the atmospheric pressure and air
temperature, enhancing the relative h umidity. The sensible flux presents a small variability b ut an increase occurs due to
a convective turbulence caused by front passage. The latent flux decrease but insufficiently to cause a co nd en sation, just
the evaporation decreases. The upwelling events are the responsible to maintain the loss o f heat after the cold front pas-
sage.
Keywords: Thermal Infrared; Heat and Cooling; Mixing; Stratification
1. Introduction
Aquatic systems continually respond to climatic con-
ditions (hydro-metrological processes) that vary over
broad scales of space and time. The primary control of
the seasonal cycle of water temperature at a given
location is the seasonal cycle of incoming shortwave
radiation. The response of each water body to meteo-
rological conditions is revealed firstly by the thermal
structure of the water column [1]. The precise knowl-
edge of reservoir heat flux dynamics is of paramount
relevance for hydrobiological and water quality studies
as physical control of the biotic structure in reservoirs is
even more important than in natu ral lakes [2].
Thermal infrared remote sensing applied to freshwater
ecosystems has aimed to map surface temperatures [3-5],
bulk temperatures [6], circulation patterns [7] and to
characterize upwelling events [8]. However, the appli-
cation of thermal infrared images to estimate the net heat
flux in tropical hydroelectric reservoirs is scarce.
Besides the power of thermal remote sensing images
to study the water surface temperature, the satellite data
is only from the upper most layer of the water surface
(skin temperature). To see below the surface some au-
thors had used telemetric data collected by an anchored
buoy [9,10].
Based on this, the objective of this paper is to estimate
the influence of the heat flux exchange between the wa-
ter surface and the atmosphere in the water column
stratification.
2. Methodology
2.1. Study Area
Our study site is the Itumbiara hydroelectric reservoir
(18˚25'S, 49˚06'W), located in a region stretched between
Minas Gerais and Goiás States (Central Brazil) that was
originally covered by tropical grassland savanna, cover-
ing an area of approximately 814 km2 and a volume of
17.03 billion m3 (Figure 1).
The climate in the region is characterized by an aver-
age precipitation ranging from 2.0 mm in the dry season
(May-September) to 315 mm in the rainy season (Oc-
tober-April). In the rainy season the wind intensity ran-
ges from 1.6 to 2.0 m·s–1 and reaches up to 3.0 m·s–1 in
the dry season, with the preferential wind direction from
the southeast. The air temperature in the rainy season
ranges from 25˚C to 26.5˚C dropping to 21˚C in June as
the dry season starts. The relative humidity has a pattern
similar to tha t of th e air temp erature, bu t with a s mall shift
in the minimum value towards September (47%). More-
ver, during the rainy season the humidity can reach 80%. o
C
opyright © 2012 SciRes. IJG

E. ALCÂNTARA 900
Figure 1. Localization of Itumbiara hydroelectric reservoir in Brazil’s central area. Shows the location of the meteorological
station and the SIMA buoy.
These hydro-meteorological patterns and the opera-
tional routine for energy generation drive the water level
fluctuations in the reservoir (Figure 2). The water level
rising period starts in December and extends until May
(with a mean period water change of = 0.031 m·day–1);
from May to June the water level is high (with a mean
period water change of 0.006 m·day–1). Due to the use of
water for power generation and evaporation rates, the
water level recedes until November (with a mean period
water change of 0.032 m·day–1). From November, the
water reaches the low level condition until December
(with a mean period water chang e of 0.023 m·day–1).
2.2. Hydrometeorological Data
The daily mean air temperature (˚C), relative humidity
(%), wind intensity (m·s–1) and precipitation (mm) from
2003 to 2008 were used for the study. These data were
obtained from a meteorological station near the dam. The
daily mean of each variable was converted into monthly
means to convert for consistency with the time scale of the
satellite data.
2.3. Buoy Data
The water temperatu re at four depths (5, 12, 20 and 40 m)
and shortwave radiation was collected by a moored buoy
called SIMA (Integrated System for Environmental
Monitoring [11] every hour from 28th March 2009 to
28th February 2010.
SIMA consis ts of an anchored buo y and its electron ics
which is instrumented with a suite of meteorological and
water quality sensors (see Figure 1 for position of
SIMA). The SIMA data are collected in preprogrammed
time intervals (1 hour) and are transmitted via satellite
link in quasi-real time to any user in a range of 2500 km
from the acquisition point.
2.4. Satellite Data
MODIS (Moderate Resolution Imaging Spectroradiome-
ter) water surface temperature (WST) level 2, 1 km
nominal resolution data (MOD11L2, version 5) were
obtained from the National Aeronautics and Space Ad-
ministration Land Processes Distributed Active Archive
Center [12,13]. All available clear-sky MODIS Terra
imagery between 2003 and 2008 were selected by visual
inspection, resulting in a total of 786 daytime images and
473 nighttime images (Figure 3).
A shoreline mask to isolate land from water was built
using the TM-Landsat-5 image in order to isolate any
anomalously cold or warm pixels remaining near the
shoreline of the reservoir. The WST-MODIS data were
extensively validated for inland waters and were consid-
ered accurate [3-5].

E. ALCÂNTARA 901
Figure 2. Climate patterns of Itumbiara reservoir: average (2003-2008) monthly mean of (a) precipitation (mm·month–1) and
wind intensity (m ·s–1), (b) air temperature (˚C) and humidity (%).
Figure 3. Information about the MODIS thermal infrared images: (a) Hour of passage; (b) Number of satellite imagery
available for daytime; (c) Nighttime.
2.5. Estimating the Heat Balance
A study of the energy exchange between the lake and
atmosphere is essential for understanding the aquatic
system behavior and its response to possible changes of
environmental and climatic conditions. In aquatic fresh-
water systems, such as lakes and reservoirs, the heat
fluxes are controlled by water surface and atmosphere
interactions and are largely affected by the stability of the
atmospheric boundary layer (ABL) [14,15]. The wind

E. ALCÂNTARA 902
intensity, temperature and humidity gradient between
water surface and air, short and longwave incoming ra-
diation and air pressure are the main environmental
variables related wit h the he at exchange s. The heat budget
in water surface, (W·m–2), can be summarized as
[16]: tot

w se le



totswswlw r
1    (1)
where, sw is the shortwave incoming radiation that
reach the water surface (W·m–2), sw

is the water al-
bedo for shortwave radiation, lw is the longwave ra-
diation that reach the water surface (W·m–2), rw is the
longwave radiation emitted by the water surface (W·m–2),
se is the sensible heat flux (W·m–2) and is the
latent heat flux (W·m–2).


le 

24
017
aa
CT
The shortwave incoming radiation can be measured
directly by radiometer or spectroradiometer and is con-
sider a penetrative term since this sort of radiation can
penetrate in the water column following the Beer-Lam-
bert Law [16]. All the other terms of heat budget are
considered non penetrative and generally are estimated
indirectly (see Equations (2)-(5)) [15,16]:


lw lw
11

   (2)
where, lw

is the water al bed o f or lo ngwa ve radi ation, C
is the cloud cover (fraction), a

is the air emissivity
(dimensionless),

is the Stefan-Boltzmann Constant
(5.699 × 10–8 W·m–2·K–4) and is the air temperature
(˚C). a
T
4
rwww
T

  (3)
where, w

is the water emissivity (dimensionless) and
is the water surface temperature (˚C).
w
T


se 10aaHaw
UCC TT   

(4)

1
le 100.622
aVEaw
ULCeep 

 (5)
where a

is th e air density (kg·m–3), 10 is wind speed
at 10 meters above the water surface (m·s–1), a is spe-
cific heat of air (1003 J·kg–1·K–1), V is latent heat of
vaporization (J·kg–1), a is vapor pressu re (mbar), w is
saturated vapor pressure at (mbar), is the atmos-
pheric pressure (mbar) and
UC
L
e e
w
T p
H
C and
E
C are the bulk
coefficients for sensible and latent heat transfer respec-
tively (dimensionless).
Due to the complexity of these fluxes and the limita-
tions of the atmospheric data available for the area of
study, some constraints were imposed. The air tempera-
ture (Ta) and wind intensity

V were considered the
same for the whole reservoir because only one meteoro-
logical station was available. Other constraint is that this
heat flux balance is due two periods of a day, that is,
daytime (10:30 h) and nigh ttime (23:30 h).
3. Results and Discussion
The estimated annual cycle of heat fluxes is shown in
Figure 4. It gives the spatially-averaged monthly mean
of reservoir using meteorological and satellite time series
data of the six years (from 2003 to 2008).
The incoming shortwave radiation is the term with
largest contribution to the heat balance. To obtain the
relative error of our estimated shortwave radiation we
used the measured (Novalynx sensor, accuracy ±5%)
shortwave by SIMA buoy covering the period from 1st
March 2009 to 28th February 2010 (Figure 4(a)).
Figure 4. Energy budget components for (a) Day t ime (10:30 h) and (b) Nighttime (23:30 h). SW is the shortwave radiation.

E. ALCÂNTARA 903
The RMSE in shortwave radiation showed here affects
mainly the surface heat flux balance (tot , because the
longwave radiation and the sensible heat flux are mainly
driven by the difference between air and water tempera-
ture, and the latent flux by wind intensity. Based on this
the expected error in tot is on average 10% higher
than was observed (Figu re 4 (a)).


The measured shortwave radiation by a radiometer
mounted in a buoy in the sea presents also an error from
6.4 to 11.3 Wm–2 (if the mast is in the vertical or tilted
15˚, respectively).
3.1. Latent Heat Flux
The latent heat flux was positive, correspond ing to a heat
gain in all months during daytime and nighttime. The
nighttime computation was set at 23:30 h, and at this
time, the water surface did not lose all the heat gained
during daytime. From January to June (summer to au-
tumn), the latent flux was near zero for both daytime and
nighttime. From August to October (winter to spring),
the latent flux during daytime was more pron ounced than
that of nighttime because of the high temperature ampli-
tude.
This amplitude occurs because of the formation of fog
just above the surface, which warms the water by pre-
cipitating into it, as observed in Figure 4(a) (mainly at
the beginning of the rainy season in September). The
positive latent flux generally occurs when the atmos-
phere above the water is stable, with little turbulent mix-
ing in the atmospheric boundary layer [17]. From No-
vember to December, the latent flux decreases again, and
the cycle recommences.
3.2. Sensible Heat Flux
A negative sensible heat flux occurs when the surface
loses heat by convective and advective processes and is
positive when the surface gains heat. For daytime, the
sensible heat flux was only negative in Janu ary (summer),
indicating that the water surface was colder than in De-
cember and February. It was positive for the others
months (see Figure 4(a)). A typical case of a heat gain in
the sensible flux occurred in October (spring), when this
variable reached its highest value (15.01 W·m–2) and
when the water surface temperature reached its highest
value. For nighttime, the sensible flux was negative for
all months of the year with a peak in September (–20.37
W·m–2) which can be related to advection caused by by
relatively high wind intensity (3.1 m·s–1). The lowest flux
occurred in February (–0.14 W·m–2), when the wind in-
tensity was two times lower than in September (1.7
m·s –1). This pattern of sensible heat distribution over
time was also observed by [18] in a reservoir in Spain.
3.3. Net Longwave Radiation
The net longwave radiation as computed by equation 3
expresses the net balance between outgoing long-wave
radiation from the lake and incoming long-wave radia-
tion from the atmosphere. Positive values indicate a loss
of energy by the reservoir. The more contrast there is
between the water and air temperature, the larger the flux
is. Net longwave radiation consists of a loss of energy
throughout th e year at daytime and nighttime. Losses are
greater during daytime because the difference between
the lake and atmosphere temperatures decreases. Their
seasonal patterns differ slightly.
Daytime longwave radiation increases from March to
August with a maximum value of 110 W·m–2 and de-
creases until January with a minimum value of about 50
W·m–2. Nighttime longwave radiation increases from
March until June with a maximum value of about 74
W·m–2 and decreases until October with a value of about
48 W·m–2. It then increases again slightly in November
and December and diminishes in January, reaching simi-
lar values as those in October.
The contrast between daytime and nighttime values is
at its greatest in September (53 W·m–2) and at its lowest
in January (3 W· m–2). During daytime, long-w ave values
are higher from May (95.94 W·m–2) to September
(106.36 W·m–2) than from January (51.28 W·m–2) to
April (78.56 W·m–2) and October (96.12 W·m–2) to De-
cember (84.14 W·m–2). The period of high flux of long-
wave radiation occurs during autumn-winter in the dry
season, that is, when the cloud cover is very low. The
lowest values occur in the rainy season with high cloud
cover (spri n g-summ e r).
3.4. Net Flux
The net heat flux during nighttime always corresponded
to a loss of energy from the lake (negative values) be-
cause the short wave radiation stopped. The daytime flux
is always positive, corresponding to a source of energy.
This is because loss terms (back longwave radiation,
sensible heat flux and latent heat flux) do not counter-
balance the source terms (shortwave and atmospheric
longwave radiation).
The daytime net radiation and heat flux followed a
nearly similar seasonal pattern as the shortwave radiation.
It had a maximum around 300 W·m–2 during January. It
steadily decreased until June, with a value of about 60
W·m–2. It remained with low values (less than 100
W·m–2) until August and then increased steadily until
December. The nighttime net heat and radiation flux also
followed a seasonal pattern, with absolute maximum
values occurring from June until August and lowest ab-
solute values during summer. The daytime heat flux was
always that of nighttime, except in June and July. The

E. ALCÂNTARA 904
balance between daytime and nighttime followed a sea-
sonal pattern with high er positive values during summer,
steadily decreasing during autumn until reaching nega-
tive values in winter. It then increases steadily again
(Figure 4). This seasonal pattern conforms to the water
surface temperature variation along the year.
When we proceed with a balance between the daytime
heat budget (Figure 4(a)) and nighttime heat budget
(Figure 4(b)), we have the effective (diurnal) heat budg-
et (Figure 5). The effective heat budget shows that dur-
ing the rainy season the reservoir gains heat, and it loses
heat during dry season. This pattern is due to the incom-
ing shortwave radiation (see Figure 4(a)), which, during
the day, adds heat to the water column and at night loses
heat due to convective cooling.
For January, February and April the northwest section
of the reservoir gains more heat than the southeast sec-
tion. This is because the dominant wind direction is from
southeast to northwest as the wind drives the warm
masses into the littoral zo ne by adv ection . Fo r March, the
southwest heats more than the northeast. During May,
the greatest area of the reservoir loses heat, and only a
small area in the main body of the reservoir gains heat.
From June to July, heat loss dominates the whole reser-
voir, and the northwest losses are lower than the south-
east losses. From August to December, the reservoir
heats from the littoral zone to the center of the reservoir;
but in October, it presents the greatest heat gradient be-
tween the littoral zone and the center of the reservoir.
3.5. The Water Column Response
The primary effect of the surface heat balance pattern is
perceptible in the thermal structure of the water column.
The Figure 6 shows that during the months when the
heat balance are positive (heat gain) the water columns
stratify; and when the heat balance are negative the water
column exhibits mixing. The main effect of this differen-
tial heat and cooling of the water in the reservoir is on
the water quality. Iron and manganese, due to circulation
and mixing, can be released after a period of reducing
conditions in the bottom, promoted by the stratification,
resulting in increased costs of water treatment for drink-
ing purpo ses [19].
Figure 5. Spatially surface heat flux balance (W·m–2) over
the Itumbiara reservoir.
Figure 6. Thermal structure of the Itumbiara reservoir.

E. ALCÂNTARA 905
Also during the mixing period the events of upwelling
events can occur, bringing nutrient-rich waters to the
surface. If prolonged, the upwelling events could induce
the growth of algae in tropical reservoirs in Brazil.
3.6. The Effect of Cold Front Passage during Dry
Season
From Figure 6 is clear that during the low shortwave
radiation, low air temperature the water can mix. How-
ever, the cold front passage over the reservoir can in-
crease the possibility to o ccur the overturn events.
In a research conducted to check the side effects of
cold fronts passage in th e Itumbiara reservoir, the authors
[10] had showed that during this meteorological event
the water can losses from 81.57 to 569 cal·cm–2 in first
six days after the passage. This heat content modification
due to the cold front will reflect in the water column
temperature and stability (Figure 7(a)). Before the pas-
sage of the front the water column presented a little te m-
perature difference between the epilimnion and metalim-
nion; with the passage of the cold front the water tem-
perature of the top-most layer decrease and the difference
of temperature in the water column decreases also.
To indicate the degree of stability and mixing in the
reservoir, due to the passage of cold front, the LN was
used [15]. The LN characterizes the dynamic stability of a
lake and is a ration of moments about the center of lake
volume of the wind force at the surface of the lake and
the gravity restoring force to the stratification.
1.5
*
1
1
T
t
g
z
gS Hz
uA
N
L
H










, (6)
where g is grav ity,

is the density of water, T is the
height to the cen ter of the metalimnion, z
g
z is the height
to the center of volume of the lake,
A
is lake area,
H
is the depth of the lake, * is the friction velocity in
water and t (gcm–1cm–2) is an estimate of the stability
of the reservoir calculated as [20]:
u
S

0
zm
tg
Sz

 
dzAzzz  (7)
where
g
z


can be obtained as:
0
0
d
d
zm
zm
zA zz
zg
A
zz


 (8)
The analysis of Lake Number (LN) are sh ow in Figure
7(b). When LN > 1 there is no deep upwelling and when
LN < 1 the cold deep, often nutrient rich, water from the
hypolimnion will reach the surface layer during the wind
episode [21].
Figure 7. Thermal structure (a) and the lake number—LN
(b) for the Itumbiara reservoir.
For LN as high as 60, little turbulent mixing is ex-
pected in the hypolimnion [22]. In this case all LN > 1
occurred dur ing the dayti me when the in cident sh ortwave
radiation is present, but after the passage of the cold front
the values of LN increase during the heating phase. Often
LN < 1 occurred during the nighttimes, the unique excep-
tion is the day during the cold passage with LN less than
1. After the passage of the front the water from hypo- li-
minion progressively cooler and the mixed layer goes up
to the top layer. The fact of the LN increases after the
front passage during the daytime could be explained by
the fact that during the cold front passage the water
losses energy to the atmosphere and when the cold front
dissipate the incident shortwave radiation heats the sur-
face creating the condition enhancing the stability of the
water column.
4. Conclusions
The use of satellite data to estimate the heat balance is
useful and facilitates more accurate analysis of the phys-
ical processes in the whole surface water. With these
results it is important to consider the predictive capacity
of the combined use of satellite, meteorological and in
situ temperature data to study the interactions between
the surface water and the surrounding atmosphere.

E. ALCÂNTARA 906
The passage of cold front over a region decreases the
atmospheric pressure and air temperature, enhancing the
relative humidity. With the formation of cloud cover the
longwave radiation increase and transfer heat by turbu-
lent convection to the water surface. The sensible flux
presents a small variability but an increase occurs due to
a convective turbulence caused by front passage; in other
hand the latent flux decrease but insufficiently to cause a
condensation, just the evaporation decreases. The up-
welling events are the responsible to maintain the loss of
heat after the cold front passage.
There is a high dependency of water column stability
to the climate and increase the knowledge about atmos-
pheric events in these systems can help in the water re-
sources management.
5. Acknowledgements
The authors would like to thank the FAPESP Project
2007/08103-2, INCT for Climate Change project (grant
573797/2008-0 CNPq). Enner Alcântara thanks CAPES
grant 0258059.
REFERENCES
[1] W. Ambrosetti, L. Barbanti and N. Sala, “Residence Time
and Physical Processes in Lakes,” Journal of Limnology,
Vol. 63, No. 1, 2002, pp. 1-15.
[2] D. Uhlmann, “Reservoirs as Ecosystems,” International
Review of Hydrobiology, Vol. 83, 1998, pp. 13-20.
[3] D. Oesch, J.-M. Jaquet, R. Klaus and P. Schenker, “Mul-
ti-Scale Thermal Pattern Monitoring of a Large Lake
(Lake Geneva) Using a Multi-Sensor Approach,” Inter-
national Journal of Remote Sensing, Vol. 29, 2008, pp.
5785-5808. doi:10.1080/01431160802132786
[4] A. Reinart and M. Reinhold, “Mapping Surface Tem-
perature in Large Lakes with MODIS Data,” Remote
Sensing of Environment, Vol. 112, No. 2, 2008, pp.
603-611. doi:10.1016/j.rse.2007.05.015
[5] E. T. Crosman and J. D. Horel, “MODIS-Derived Surface
Temperature of the Great Salt Lake,” Remote Sensing of
Environment, Vol. 113, No. 1, 2009, pp. 73-81.
doi:10.1016/j.rse.2008.08.013
[6] S. Thiemann and H. Schiller, “Determination of the Bulk
Temperature from NOAA/AVHRR Satellite Data in a
Midlatitude Lake,” International Journal of Applied
Earth Observation and Geoinformation, Vol. 4, No. 4,
2003, pp. 339-349. doi:10.1016/S0303-2434(03)00021-7
[7] S. G. Schladow, S. O. Palmarsson, T. E. Steissberg, S. J.
Hook and F. J. Prata, “An Extraordinary Upwelling Event
in a Deep Thermally Stratified Lake,” Geophysical Re-
search Letters, Vol. 31, 2004, pp. 1-4.
[8] T. E. Steissberg, S. J. Hook and S. G. Schladow, “Char-
acterizing Partial Upwellings and Surface Circulation at
Lake Tahoe, California-Nevada, USA with Thermal In-
frared Images,” Remote Sensing of Environment, Vol. 99,
No. 1-2, 2005, pp. 2-15.
[9] E. Alcântara, E. Novo, J. Stech, J. Lorenzzetti, C. Barbosa,
A. Assireu and A. Souza, “A Contribution to Under-
standing the Turbidity Behavior in an Amazon Flood-
plain,” Hydrology and Earth System Science, Vol. 14,
2010, pp. 351-364. doi:10.5194/hess-14-351-2010
[10] E. Alcântara, M.-P. Bonnet, A. T. Assireu, J. L. Stech, E.
M. L. M. Novo and J. Lorenzzetti, “On the Water Ther-
mal Response to the Passage of Cold Fronts: Initial Re-
sults for Itumbiara Reservoir (Brazil),” Hydrology and
Earth System Sciences Discussions, Vol. 7, No. 6, 2010,
pp. 9437-9465. doi:10.5194/hessd-7-9437-2010
[11] J. L. Stech, I. B. T. Lima, E. M. L. M. Novo, C. M. Silva,
A. T. Assireu, J. A. Lorenzzetti, J. C. Carvalho, C. C. F.
Barbosa and R. R. Rosa, “Telemetric Monitoring System
for Meteorological and Limnological Data Acquisition,”
International Association of Theoretical and Applied
Limnology, Vol. 29, No. 4, 2006, pp. 747-1750.
[12] Z. Wan, “New Refi nements and Validation of the MODIS
Land-Surface Temperature/Emissivity Products,” Remote
Sensing of Environment, Vol. 112, No. 1, 2008, pp. 59-74.
doi:10.1016/j.rse.2006.06.026
[13] C. C. Justice, et al., “The Moderate Resolution Imaging
Spectroradiometer (MODIS): Land Remote Sensing for
Global Change Research,” IEEE Transactions on Geo-
science and Remote Sensing, Vol. 36, No. 4, 1998, pp.
1228-1247. doi:10.1109/36.701075
[14] P. Verburg and J. P. Antenucci, “Persistent Unstable At-
mospheric Boundary Layer Enhances Sensible and Latent
Heat Loss in a Tropical Great Lake: Lake Tanganyika,”
Journal of Geophysical Research, Vol. 115, 2010, pp. 1-
13.
[15] J. Imberger and J. J. C. Patterson, “Physical Limnology,”
Advances in Applied Mechanics, Vol. 27, 1990, pp. 303-
475. doi:10.1016/S0065-2156(08)70199-6
[16] M.-P. Bonnet, M. Poulin and J. Devaux, “Numerical
Modeling of Thermal Stratification in a Lake Reservoir:
Methodology and Case Study,” Aquatic Sciences, Vol. 62,
No. 2, 2000, pp. 105-124. doi:10.1007/s000270050001
[17] B. Lofgren and Y. Zhu, “Surface Energy Fluxes on the
Great Lakes Based on Satellite-Observed Surface Tem-
peratures 1992 to 1995,” Journal of Great Lakes Re-
search, Vol. 26, No. 3, 2000, pp. 305-314.
doi:10.1016/S0380-1330(00)70694-0
[18] T. Serra, J. Vidal, X. Casamitjana, M. Soler and J. Co-
lomer, “The Role of Surface Vertical Mixing in Phyto-
plankton Distribution in a Stratified Reservoir,” Limnol-
ogy and Oceanography, Vol. 52, No. 2, 2007, pp. 620-
634. doi:10.4319/lo.2007.52.2.0620
[19] J. G. Tundisi, T. Matsumura-Tundisi, J. D. Arantes Junior,
J. E. M. Tundisi, N. F. Manzini and R. Ducrot, “The Re-
sponse of Carlos Botelho (Lobo, Broa) Reservoir to the
Passage of Cold Fronts as Reflected by Physical, Chemi-
cal and Biological Variables,” Brazilian Journal of Biol-
ogy, Vol. 64, No. 1, 2004, pp. 177-186.
doi:10.1590/S1519-69842004000100020
[20] G. E. Hutchinson, “A Treatise on Limnology, Geography,
Physics and Chemistry,” John Wiley, New York, 1957,
1015 p.
[21] J. Antenucci and J. Imberger, “The Seasonal Evolution of

E. ALCÂNTARA 907
Wind/Internal Wave Resonance in Lake Kinneret,” Lim-
nology and Oceanography, Vol. 48, No. 5, 2003, pp.
2055-2061. doi:10.4319/lo.2003.48.5.2055
[22] M. Hondzo and H. G. Stefan, “Long-Term Water Quality
Predictors,” Water Research, Vol. 30, No. 12, 1996, pp.
2835-2852. doi:10.1016/0043-1354(95)00286-3