Conference PaperPDF Available

Spatial Algal Bloom Characterization by Landsat 8-Oli and Field Data Analysis

  • Instituto de Altos Estudios Espaciales Mario Gulich
Andrea Guachalla Alarc´on1,2, Alba Germ´an3,4, Alejandro Aleksink´o1, Fernanda Garc´ıa
Ferreyra1, Carlos Marcelo Scavuzzo1, and Anabella Ferral1,4
1Instituto Mario Gulich, Comisi´on Nacional de Actividades Espaciales (CONAE), C´ordoba, Argentina
2Instituto de Investigaciones Farmac´euticas y Bioquim´ıcas, Universidad Mayor de San Andr´es, La Paz, Bolivia
3Universidad Blas Pascal, C´ordoba, Argentina
4Departamento de Hidrolog´ıa, SRH de la provincia de C´ordoba, Argentina
Water pollution is an important problem around the
world as it is closely related to human and environ-
mental health. Field campaigns are expensive, time
consuming and may provide little information. Remote
sensing provides synoptic spatio-temporal views and
can lead to a better understanding of lake ecology. In
this work an extreme algal bloom event which occurred
in a reservoir is characterized by LANDSAT8-OLI sen-
sor and in situ sampling. Chlorophyll-a concentration
and algae abundance data are measured on samples
collected simultaneously with satellite pass and used
to build semiempirical models. Two linear functions to
calculate chlorophyll-a from satellite data are presented
and compared. A linear model from band 2 (blue)
and band 5 (NIR) presents the best performance with
a determination coefficient equal to 0,89. In situ
and satellite chlorophyll-a lead comparable trophic
class assessment, hypertrophic. Both Models fail to
predict chlorophyll-a concentration near river intrusion
(North), where low values of reflectance are recorded.
Keywords: chlorophyll-a, phytoplankton, eutrophi-
cation, LANDSAT8-OLI, linear regression
Water pollution is one of the most serious problems
around the world and it is closely related to human
and environmental health. This kind of pollution can
proceed from: mismanagement of organic wastes, oil
spill, mining drainage, and others. Particularly, the
increase of organic wastes in water leads to the prolifer-
ation of microorganisms that can cause eutrophication.
Eutrohpication is a process in which primary produc-
tion in water is enhanced and photosynthetic microor-
ganisms proliferate due to the presence of nitrogen and
phosphorus in unbalanced proportions[1]. These mi-
croorganisms inhabit water surface and interfere with
solar rays reaching lower strata of the water column,
causing: an increase in biomass and a narrowed micro-
bial biodiversity, an unbalanced ecological niche, and a
higher mortality rate in aquatic animals due to anoxic
episodes. Chlorophyll-a concentration can be used as
an index to monitor algal abundance by remote sens-
ing technique since it presents active optical properties
in visible and near infrared regions of the electromag-
netic spectrum [2]. In this work, we analyze an algal
bloom in San Roque Reservoir with satellite imagery
from Landsat 8-OLI and field data. This reservoir is
the main source of drinking water for Cordoba city, Ar-
gentina, which has a population of 1,8 millon people.
On the other hand, coastal residents use water for do-
mestic purposes, tourism and fishing. The rivers and
streams that flow into the reservoir contribute a large
amount of nutrients which has been causing, in the last
decades, a process of severe eutrophication, sometimes
reaching the hypereutrophication level[3, 4] and releas-
ing cyanotoxins [5]. This research is a continuation of
previous studies carried out with LANDSAT 5-TM and
TERRA-MODIS sensors to monitor water quality and
harmful algae blooms on this reservoir [6].
2.1 Study area
Figure 1 shows San Roque reservoir which is located in
Villa Carlos Paz in Cordoba, Argentina at 3122’4”S
6428’10” and 643 masl (meters above sea level). It
shows sampling sites for In Situ measurements, chemi-
cal analysis and satellite imagery processing.
Figure 1: San Roque Reservoir and sampling sites
2.2 Field data
Eight sampling sites were considered in this work as
shown in Figure 1. In Situ parameters including Sec-
chi depth and coordinates were measured in each site,
while two liters of water were collected aseptically. Sam-
ples for algae counting were collected in separate flasks,
and were preserved with lugol until processing. Sam-
ples then were transported to the laboratory where
chlorophyll-a, and algal species abundance were mea-
sured using standardized methods[7].
2.3 Satellite data
Satellite imagery from the scene of algal blooming in
San Roque Reservoir in February 22nd 2017 was re-
covered from Landsat 8-OLI, available in ”Earth Ex-
plorer“ ( A level
2 LANDSAT 8-OLI product, atmospheric corrected sur-
face reflectance and 30 meter spatial resolution, was
used. The scene corresponds to path 229, row 82 ac-
quired on February 22nd of 2017.
2.4 Chlorophyll-a models
Correlation between chlorophyll-a measured in samples
and reflectance from Landsat 8-OLI imagery was per-
formed with Rstudio. Reflectance recorded at seven
bands by OLI sensor was correlated with in Situ data
by bivariate analysis, obtaining Pearson correlation co-
efficient. For this, B5-B2 and B5/B4 mathematical
functions were used to generate semiempirical models.
Modeled chlorophyll-a was calculated using simple and
multiple line repression to build mathematical semiem-
pirical models. Garganta point was excluded from re-
gressions because it is located in a very narrow lake
2.5 Trophic state maps
Statistical data was used to generate chlorophyll-a maps
and Trophic State Index (TSI) maps with two semiem-
pirical models: B5-B2 and B5/B4, using ENVI 4.8.
Carlson index was used to generate the TSI maps from
chlorophyll-a concentration data, see table in Figure 2.
Figure 2: Carlson scale for Trophic State Index[8].
TSI=30.6 + 9.81*ln(chlorophyll-a)
3.1 Field data analysis
Secchi disk depth and chlorophyll-a concentration are
directly associated with the trophic state of a water
body, and are used to evaluate eutrophication in differ-
ent settings. Table 1 shows that all the sampling sites
displayed high concentration of chlorophyll-a as well as
a shorter Secchi depth than expected for oligotrophic
Table 1: Field data from San Roque Dam, 22/02/2017
Site Chlorophyll-a (ug/L) Secchi depth
Centro 127,1 0,55
Garganta 197,1 0,5
Zona A 53,8 0,3
Zona B 288,5 0,3
SAT 1 27,6 0,3
SAT 2 132,2 0,5
SAT 3 94,7 0,6
SAT 4 56,7 0,5
From the results a case of eutrophic or hypereu-
trophic state in San Roque reservoir can be inferred[8].
Biological analysis showed different species of algae,
cyanobacteria and diatomeas (data not shown). The
most abundant species were: Microcystis sp. in Zona B
and SAT2, and Ceratium sp. in Centro and Garganta.
Other species found in the samples were: Oscillatoria
sp.,Cyclotella sp.,Aulacoseria sp., and Clostertopsis
sp. All the species described in the samples are com-
monly found inhabiting aquatic environments, and are
not related to diseases directly, therefore they do not
represent a threat for human health. However, Micro-
cystis sp., a genre that belongs to Cyanobacteria, is
able to produce neurotoxins and hepatotoxins that are
highly harmful for humans and animals and was found
in this reservoir [5].
3.2 Satellite data analysis
Satellite data was obtained from Earth Explorer for
Landsat 8-OLI imagery with a 30 meter spatial reso-
lution. A pre-processed image RGB (4,3,2) is shown in
Figure 3, where a green tone can be appreciated, due
to the presence of photosynthetic microorganisms.
Figure 3: Landsat 8-OLI subset (432), 22/02/2017
Surface reflectance (SR) data from the satellite im-
agery was obtained in every sampling point. Figure
4 presents SR as a function of the wavelength center
of OLI sensor bands for all monitoring sites. It can
be observed that ZB and Garganta show the greatest
values at green band (565 nm) in agreements with the
greatest values obtained over those sites for field mea-
surements of Chlorophyll-a concentration, 288,5 and
197,1 ug/L respectively. Although algae composition
for ZB and Garganta sites are markedly different, spec-
tral signatures obtained from OLI sensor are similar for
both sites, indicating that distinguishing between dif-
ferent algae abundances by means of this sensor is not
a straightforward issue.
Figure 4: Reflectance obtained from Landsat 8-OLI,
3.3 Chlorophyll-a semiempirical models
Two semiempirical mathematical models were used to
evaluate eutrophication in San Roque reservoir based
on the correlation obtained between reflectance and
chlorophyll-a concentration. It was found that B4 (Red)
and B5 (NIR) were highly and significantly correlated
with chlorophyll-a concentration since Pearson coeffi-
cients greater than 0,9 were calculated. Previous stud-
ies used mathematical functions including B5, B4 and
B2 reflectances to assess the trophic state on inland
waters [9, 2]. Equations (1) and (2) show the most
adequate models. (1) obtained by Multiple regression,
negative with B2 and positive with B5 and (2) which
depends positively and linearly with the ratio B5/B4.
Model 1 presents the better determination coefficient,
0.89 (p<0,05), while Model 2 presents a determination
coefficient equal 0.87 (p<0,05).
[clorof ila a] = 5749.5B59951.1B2 + 128.7 (1)
[clorof ila a] = 101.03 B5/B448.04 (2)
3.4 Trophic state maps
Figures 5 and 6 present chlorophyll-a maps obtained by
Model 1 and 2. In situ and satellite chlorophyll-a and
TSI lead comparable values. Both models fail to predict
chlorophyll-a concentration near Cosquin river intru-
sion (North), where low reflectance data are recorded.
In that place, negative values of this variable were cal-
culated that affected the building of TSI maps. Trophic
State Index maps showed that the lakes center tend to
a higher state of eutrophication than the northern and
southern region where river intrusion with fresh water
may dilute the bloom. Artificial aeration system seems
to be not functioning that day. Both models had some
pixels with no data, due to the negative values afore-
mentioned that could not be taken into account by TSI
formula. In this sense, Model 1 seems to respond better
for low chlorophyll-a concentration. Further studies are
being carrying out to improve and extend these algo-
rithms to perform temporal series analysis and incor-
porate low value of chlorophyll-a.
Combined use of satellite and in situ chlorophyll-a data
improves phytoplankton bloom delimitation, an impor-
tant aspect to manage water supply reservoirs. In ad-
dition, remote sensing enhance understanding of spa-
tial patterns associated with natural and anthropogenic
interventions as rivers intrusion or artificial aeration
systems results. In situ and satellite chlorophyll-a
(a) Chlorophyll-a map
(b) Trohpic State Index map
Figure 5: Semiempirical model using B5 and B2 bands
lead comparable trophic class assessment, mostly hy-
pertrophic. More studies need to be done in order to
extend these models to perform temporal studies and
to improve low chlorophyll-a concentration values esti-
ACKNOWLEDGMENT: A.Guachalla Alarcon
thanks CELFI for the fellowship granted.
[1] D. J. Conley, R. W. Paerl, H. W.and Howarth, D. F. Boesch,
S. P. Seitzinger, K. E. Havens, C.E. Lancelot, G. E. Likens,
et al. Controlling eutrophication: nitrogen and phosphorus.
Science, 323(5917):1014–1015, 2009.
[2] Katja D¨ornh¨ofer, Philip Klinger, Thomas Heege, and
Natascha Oppelt. Multi-sensor satellite and in situ monitor-
ing of phytoplankton development in a eutrophic-mesotrophic
lake. Science of The Total Environment, 612:1200–1214,
[3] A. Ferral, V. Solis, A. Frery, A. Orueta, I. Bernasconi, J. Bres-
ciano, and C. M Scavuzzo. Spatio-temporal changes in water
quality in an eutrophic lake with artificial aeration. Journal
of Water and Land Development, 35(X-XII):27–40, 2017.
(a) chlorophyll-a map
(b) Trophic State Index map
Figure 6: Semiempirical model using B5 and B4 bands
[4] A. Germn, C. Tauro, M. C. Scavuzzo, and A. Ferral. De-
tection of algal blooms in a eutrophic reservoir based on
chlorophyll-a time series data from modis. In 2017 IEEE
International Geoscience and Remote Sensing Symposium
(IGARSS), pages 4008–4011, July 2017.
[5] M. Ruiz, L. Galanti, A. L. Ruibal, M.I. Rodriguez, and Ma. V.
Wunderlin, D.and Am´e. First report of microcystins and
anatoxin-a co-occurrence in san roque reservoir (c´ordoba, ar-
gentina). Water, Air, & Soil Pollution, 224(6):1593, 2013.
[6] A. Germ´an, C. Tauro, Ver´onica Andreo, I. Bernasconi, and
A. Ferral. An´alisis de una serie temporal de clorofila-a a par-
tir de im´agenes modis de un embalse eutr´ofico. In Biennial
Congress of Argentina (ARGENCON), 2016 IEEE, pages 1–
6. IEEE, 2016.
[7] Water Environmental Federation, American Public Health
Association, et al. Standard methods for the examination
of water and wastewater. APHA: Washington, DC, USA,
[8] Robert E Carlson. A trophic state index for lakes. Limnology
and oceanography, 22(2):361–369, 1977.
[9] A. Ferral, E. Luccini, V. Solis, A. C. Frery, A. Aleksinko,
I Bernasconi, and C. M. Scavuzzo. In-situ and satellite mon-
itoring of water quality of an eutrophic lake with an artifi-
cial air diffusion system. IEEE Latin America Transactions,
16(2):627–633, 2018.
... Multispectral data have been widely used in freshwater algal bloom studies, particularly due to its easiness and cost-effectiveness to mount on airborne and spaceborne platforms compared to hyperspectral sensors. The most widely used multispectral data series for algal bloom remote sensing studies is the Landsat (TM, ETM + , OLI) series (e.g., Chang et al. 2004;Veettil and Bianchini 2014;Veettil and Quang 2018;Alarcon et al. 2018;Li et al. 2018;Xu et al. 2021). In addition to cyanobacterial blooms, multispectral reflectance has been used for the optical detection of other algal blooms (e.g., Bangyi et al. 2013) because the spectral reflectance properties of the phytoplankton are the key variable in detecting the bloom. ...
... In addition, the authors (Alba et al. 2020) used different combinations of Sentinel-2 bands to perform bi-variate (logarithms, indexes, and ratios) analysis and Pearson correlation for chlorophyll-a estimation and a ratio model was created and validated using multi-date data. A slightly similar method was used by Alarcon et al. (2018) using Landsat-8 OLI data; chlorophyll-a concentration and algae abundance data were measured using simultaneous data from the field and remote sensing that were used to build semi-empirical models and a linear model from blue (0.450 to 0.515 µm) and NIR (0.845 to 0.885 µm) presented the best results with a determination coefficient of 0.89. ...
Full-text available
A large number of freshwater lakes around the world show recurring harmful algal blooms, particularly cyanobacterial blooms, that affect public health and ecosystem integrity. Prediction, early detection, and monitoring of algal blooms are inevitable for the mitigation and management of their negative impacts on the environment and human beings. Remote sensing provides an effective tool for detecting and spatiotemporal monitoring of these events. Various remote sensing platforms, such as ground-based, spaceborne, airborne, and UAV-based, have been used for mounting sensors for data acquisition and real-time monitoring of algal blooms in a cost-effective manner. This paper presents an updated review of various remote sensing platforms, data types, and algorithms for detecting and monitoring algal blooms in freshwater lakes. Recent studies on remote sensing using sophisticated sensors mounted on UAV platforms have revolutionized the detection and monitoring of water quality. Image processing algorithms based on Artificial Intelligence (AI) have been improved recently and predicting algal blooms based on such methods will have a key role in mitigating the negative impacts of eutrophication in the future.
... Landsat series, developed by NASA, and Sentinel-2 satellite constellation, by ESA within the framework of the Copernicus program, have proved its usefulness for estimating chlorophyll-a [9], [22], [23], Secchi disk depth [24], suspended solids [25] and water extent [26], [27]. Particularly, Sentinel-2 satellites have some of their bands in VIS-NIR region with 10 m of spatial resolution, 5 days of temporal resolution, free access to their image repository and high precision results [28] making them one of the most chosen data sets. ...
Conference Paper
Water has historically been considered a renewable resource, however in the last century a sustained degradation of its quality has been observed, both in continental and oceanic systems due to anthropic impact. In this context, the management of water resources by satellite technology represents a central point in public policies since they allow anticipating and adapting to these disturbances. In this work, Sentinel 2-MSI multispectral and PRISMA hyperspectral sensors are used to carry out an anal- ysis of different optical water types in North-East of Argentina. Convolutional procedure was used to compare sensors responses for atmospheric corrected products and RMSE (root mean square error), BIAS and MAE (mean absolute error) metrics were used to assess their performance. Differences below 1 percent were obtained in all cases, indicating an excellent match-up between both sensors. From hyperspectral PRISMA data it was possible to detect quantitative shift towards reddish wavelengths as turbidity of Parana River increases along a transect as well as an increase of the peak value in magnitude. This work opens new opportunities to monitor water quality changes related to optical constituents with deeper details in space and time in highly urbanized and perturbed regions.
... Two spectral indices, which are sensitive to chlorophyll-a concentration, were used to assess eutrophication of water bodies: Ch1_1 = 14.9 * (b5 -b4) / (b2 +b3)) + 5.94, where b2 is the blue band (460-520 nm), b4 is the red band (650-680 nm), b5 is the red edge band (700-710 nm) (Alarcon et al., 2018). ...
The aquatic macrophytes are widely used to indicate the trophic status of water bodies. For Ukrainian water bodies, which represent a wide ecological diversity, the indicator scales of the trophicity regime have not been developed. The purpose of the research is to develop an approach to adapt the Macrophyte Biological Index for Rivers (MBIR) for the conditions of the middle course of the Dnipro River. The difficulty in solving the problem is that the effect of eutrophication occurs against the variability of a complex of environmental factors and processes of natural and anthropogenic origin. The extraction of the eutrophication gradient is a prerequisite for assessing the response of plants to this factor and for constructing adequate indicator scales. The influence of natural factors on aquatic macrophyte communities is assumed to be spatially structured due to the long-term coevolution of the environment and biota. The pure plant response to the effects of eutrophication was assessed after extracting the spatial component of plant community variation. The pure response of aquatic macrophytes to the eutrophication factor provided an opportunity to construct an adequate scale for indicating the trophic status of water bodies, which correlates strongly with the scores of the MBIR. The studies were carried out within the Dnipro-Orilskiy Nature Reserve in the waters of the Dnipro reservoir, which was created on the Dnipro River. The aquatic macrophyte surveys were conducted at 76 locations. Studies were conducted in 50–100 m long plots along the shoreline. The remote sensing data from the Sentinel-2 satellite was used to estimate the indexes of the eutrophication, turbidity, and transparency of the water. The distance-based Moran's eigenvector maps were produced from the geographic coordinates of the sampling points to be used as the spatial variables. Three versions of the Canonical Correspondence Analysis (CCA) were performed: 1) the CCA with the environmental factors and spatial variables as constrained predictors; 2) the CCA with the spatial variables as constrained predictors; 3) the CCA with the environmental factors as constrained predictors and the spatial variables as conditional predictors. The species scores on the corresponding ordinate axes were considered to estimate the species optimum. The variations of the species scores on the corresponding ordinate axes were considered to estimate the species tolerance. Species scores were converted into the CSd indicator values, and tolerance was converted into the Ed coefficients of ecological amplitude. The indicator values and environmental amplitude coefficients obtained for the ordination accounting for spatial effects had no statistically significant correlation coefficients with other indicator scales. The CSd indicator values that were obtained for community ordination accounting for pure environmental effects were statistically significantly correlated with the MBIR and other of the indicator scales. The coefficient of environmental amplitude Ed was negatively correlated with the E. The MBIR revealed that the level of eutrophication in 102 locations (67.1%) was estimated as Bad, in 37 locations (24.3%) it was estimated as Poor, and in 13 locations (8.6%) it was estimated as Moderate. The extraction of the spatial component of macrophyte community variation resulted in the evaluation of indicator scales that correlated statistically significantly with the original MBIR scale.
... Para que una imagen de clorofila-a (perepción remota) pueda ser calibrada con datos in-situ, la captura de ambos tipos de datos debería tener una diferencia temporal mínima porque la concentración de clorofila-a varía estacionalmenteAranguren-Riaño et al. (2018); Abella and Martínez (2012);Barrera et al. (2019);Cordero et al. (2005), de forma que la variación espacial de la concentración de clorofila-a no varíe significativamente y sea equiparable. En el Lago Panguipulli (ubicado en el centro sur de Chile), Huovinen et al. (2019) estudia las concentraciones de clorofila-a y otros componentes en el cuerpo de agua en diferentes imágenes satelitales usando muestras in-situ tomadas en tiempos distintos para determinar la variación espacial de los componentes medidos permitiendo una distancia temporal de entre 0 y 6 días, excepto un conjunto de datos con diferencia temporal de 9 días; para generar mapas de clorofila-a del Embalse de San RoqueGuachalla et al. (2018) usó datos de percepción remota e in-situ con diferencia temporal de captura de 0 días;Yu et al. (2021) indicó que en un lago (Lago Lugu, China) sin desembocaduras importantes de ríos y con presencia de actividades humanas como la agricultura, descarga de aguas residuales domésticas y desarrollo turístico (como el LT), las mediciones de la Inversión de profundidad de Secchi (relacionadas con la calidad del agua y la concentración de clorofila-a) varían temporalmente con un patrón que se repite anualmente;Cheng et al. (2019) encuentra también variaciones temporales en las concentraciones de clorofila-a en el Lago Taihu. Debe buscarse entonces emparejarse de conjuntos de datos cuya diferencia de captura sea mínima. ...
Experiment Findings
Full-text available
We found that remote sensing, machine learning, and Bayesian statistical methods working simultaneously on a conventional computer allow detection of chlorophyll-a concentrations in Lago de Tota, Colombia.
... Eutrophication is a process in which the primary production in water is increased and photosynthetic microorganisms reproduce due to the availability of nitrogen and phosphorus in unbalanced proportions (Conley et al., 2009). These microorganisms inhabit the water surface and prevent sunlight from penetrating into the lower layers of the water column, causing an increase in biomass and a narrowing of microbial biodiversity, an unbalanced ecological niche, and increased mortality of aquatic animals due to anoxic episodes (Alarcon et al., 2018). One of the main reasons for water quality deterioration is the increased nutrient loading, which may directly or indirectly cause a number of environmental problems, including a decrease in the concentration of dissolved oxygen or hypoxia (Nixon, 1995;Cloern, 2001;Boesch, 2002;Miltner, 2018). ...
Full-text available
The aim of this article was to investigate the influence of structural features of the floodplain water network on the spatial and temporal dynamics of chlorophyll-a concentration as an indicator of eutrophication. The research was conducted in the waters of the “Dnipro-Orilskiy” Nature Reserve. The geographic information base with polygonal objects which represented water bodies of the reserve was created on the basis of detailed geographical maps and the high resolution space images. The water bodies were characterized using such parameters as the distance of the water body centroid from the nearest shore of the Dnipro River, the area of the water body, the order of the water body and the connectivity of the water body. Chlorophyll-а concentration was estimated based on the surface algal bloom index. The information was obtained about 148 water bodies, 141 of which are water bodies in the floodplain of the Dnipro River. The area of floodplain water bodies within the reserve was 3.28 million m2. The area of floodplain water bodies ranged from 300–232,500 m2. Trophic State Index allows us to estimate the trophic level of Dnipro River waters as mesotrophic, water bodies of first and second order as eutrophic, and water bodies of third and fourth order as hypereutrophic. The dynamics of chlorophyll-a content in water followed the seasonal course of temperatures. The concentration was lowest in the cold period of the year and reached its maximum in the second half of summer. The autumn decrease occurred at the end of September. The seasonal course of air temperature was superimposed on the peculiarities of the temperature regime of a particular water body, which depended on its depth and flow rate. The time, water body area, distance from the Dnipro River channel, connectivity and order of water bodies were the statistically significant predictors of chlorophyll concentration in water and were able to explain 85% of the variation of this indicator. The increase in chlorophyll-a concentration with increasing order of a water body is due to a decrease in the intensity of water exchange and a decrease in the depth of water bodies of higher order. An increase in the order of a water body is accompanied by a branching network of water bodies, the ability of water bodies to clear sediments decreases. Sediment accumulation leads to a decrease in their depth. Warming of shallow ponds and accumulation of organic matter in them are factors of intensive growth of blue-green algae. The evacuation of surplus organic matter, which results from mass vegetation development with excessive nutrient inputs, is a key driver of the eutrophic regime of water bodies. The increasing importance of regulatory processes develops in agreement with an increase in chlorophyll-a concentration in a water body. The importance of the considered factors reaches the highest level in summer time, when simultaneous maximum warming of water bodies and minimum water level in them take place. Accordingly, the differences between deep and relatively cool water bodies and shallow water bodies that warm up quickly, which significantly stimulates the growth of organic mass, reach the greatest contrast. The spatial patterns of variation in chlorophyll-a concentration have a complex multiscale structure, indicating the multiple nature of the acting factors. The spatial variability was represented as a composition of broad-scale and medium-scale spatial processes. The broad-scale process is most dependent on connectivity, whereas for the medium-scale process the leading one is the effect of water body order.
... In this framework, remote sensing becomes an essential and effective monitoring tool of algal bloom. In the passed decades, Landsat and MODIS program had been successfully used to monitor [Chl-a] in San Roque reservoir [4,5,6]. However, their lower spatial resolution or sparse acquisition disabled the joint extraction of spatial patterns and temporal trends. ...
... In this framework, remote sensing becomes an essential and effective tool to monitor algae bloom. In the passed decades, Landsat and MODIS program had been successfully used to monitor chlorophyll-a in San Roque reservoir (Germán et al., 2017b, Alarcón et al., 2019. However, their lower spatial resolution (MODIS) and sparse acquisition (Landsat), disabled the joint extraction of spatial patterns and temporal trends. ...
Full-text available
Eutrophication is characterized by excessive plant and algal growth due to the increased of organic matter, carbon dioxide and nutrients in water body. Although eutrophication naturally occurs over centuries as lakes age, human activities have accelerated it processes and caused dramatic changes to the aquatic ecosystems including elevated algae blooms and risk for hypoxia as well as degradation in the quality of drinking water and fisheries. Monitoring eutrophic processes is therefore highly important to human health and to the aquatic environment. However, the spatial and seasonal distribution of the phenomena and its dynamic are difficult to be resolved using conventional methods as water sampling or sparse acquisition of remote sensing data. This research work proposes a methodology that takes advantage of the high temporal resolution of Sentinel-2 (S2) for monitoring eutrophic reservoir. Specifically, it uses large temporal series of S2 images and advanced temporal unmixing model to estimate the abundance of [Chl-a] and algae species in San Roque reservoir, Argentina, in the period August 2016 to August 2019. The spatial patterns and the temporal tendencies of these aquatic indicators, that have a direct link to Eutrophication, were analysed and evaluated using in situ data in order to assess their contribution to the local water management.
... In this framework, remote sensing becomes an essential and effective monitoring tool of algal bloom. In the passed decades, Landsat and MODIS program had been successfully used to monitor [Chl-a] in San Roque reservoir [4,5,6]. However, their lower spatial resolution or sparse acquisition disabled the joint extraction of spatial patterns and temporal trends. ...
Eutrophic reservoirs are characterized by excessive presence of plant and algal growth due to favourable environmental conditions, temperature, light and nutrients. Human activities accelerate this phenomenon and provoke dramatic changes to the aquatic ecosystems. The monitoring of water quality of these ecosystems and the study of the effects they have on the environment demand a large amount of spatial and temporal information, which is almost exclusively provided by Earth Observations (EO). This study uses a large temporal series of Sentinel-2 (S2; 2016 till 2019) images to characterize the temporal and spatial distribution of chlorophyll-a [Chl-a] in San Roque Reservoir, Cordoba Province, Argentina. A robust method that combines empirical modelling of [Chl-a] and data mining analysis is employed. Model results showed significant fit (R² = 0.77) between [Chl-a] measured in the reservoir and the ratio between the NIR and red bands of S2. An analysis of spatio-temporal patterns demonstrated that [Chl-a] distribution in San Roque is complex and influenced by seasonal changes, aeolian forces, hydrodynamic flows, bathymetry, water levels, and pollution sources. The study also found a correlation between algae bloom events and areas with extreme levels of [Chl-a] (>850 mg/m3) in the water body. Additionally, advanced data mining tools such as slope analysis and spatial anomalies indexes, identified regions in the reservoir where water quality had improved or deteriorated. The results show the added value of using large Sentinel-2 data series to assess the concentration of Chlorophyll-a in eutrophic reservoirs over a variety of spatial and temporal scales.
Full-text available
In this work we present novel results concerning water quality changes in an eutrophic water body connected with an artificial aeration system installed in it. Sixty one in-situ and laboratory measurements of biogeochemical variables were recorded monthly between October 2008 and June 2011 to evaluate temporal and spatial changes in San Roque reservoir (Argentina). t-Student mean difference tests, carried out over the whole period, showed with 95% confidence that a monitoring point located at the centre of the water body is representative of the chemical behaviour of the reservoir. Thermal stratification was observed in all sampling sites in the summer, but the frequency of these episodes was markedly lower in bubbling zones. Mean chlorophyll-a concentrations were 58.9 μg·dm⁻³ and 117.0 μg·dm⁻³ in the absence and in the presence of thermocline respectively. According to the t-Student test, this difference was significant, with p < 0.001. Phosphate release from sediments was corroborated under hypoxia conditions. ANOVA one way analysis did not show significant spatial differences for any variable. Mean normalize spatial index (MENSI) was developed to compare data from different regions affected by high temporal variability. It proved to be useful to quantify spatial differences. Structure analysis of temporal series was used to scrutinize both chemical and spatial association successfully. Three chemically different zones were determined in the reservoir. This study demonstrated that spatial comparisons by means of marginal statistics may not be an adequate method when high temporal variation is present. In such a case, temporal structure analysis has to be considered.
Full-text available
The aim of this study was to evaluate the presence of microcystin-LR, microcystin-RR, microcystin-YR, and the neurotoxin anatoxin-a in water samples collected monthly during 1 year in San Roque reservoir (Córdoba, Argentina) to identify the environmental factors that could promote the presence of these cyanotoxins. The HPLC-UV and MS/MS analysis showed the presence of microcystin in most of the sampling times, even when Cyanobacteria were subdominant. Microcystin concentrations varied from not detectable levels to 119.0 μg L−1. Thus, they frequently surpassed the guidelines suggested by WHO for drinking water (1 μg L−1) and recreational exposure (20 μg L−1). To the extent of our knowledge, this is the first report of anatoxin-a in freshwaters in South America. Anatoxin-a concentrations varied from not detectable levels to 6.6 ng L−1, a thousand times below the provisional guideline adopted by New Zealand for drinking water. Microcystin showed significant correlation with Microcystis and Pseudoanabaena while anatoxin-a correlated with Oscillatoria and Anabaena counts. Linear discriminant analysis showed that higher pH levels and more variable chlorophyll-a concentrations were measured in San Roque reservoir when cyanotoxins were present. Lower inorganic nitrogen concentrations were observed in autumn, when the prevalence of Anabaena became significant in Cyanobacteria composition and highest anatoxin-a levels were measured. The observed dynamic of phytoplankton going together with the cyanotoxins occurrence could be explained by the hypothesis of cyanotoxins acting as allelopathic compounds. The microcystin levels measured plus the presence of anatoxin-a show the need of stronger management efforts to preserve human and wildlife health.
Full-text available
A numerical trophic state index for lakes has been developed that incorporates most lakes in a scale of 0 to 100. Each major division ( 10, 20, 30, etc. ) represents a doubling in algal biomass. The index number can bc calculated from any of several parameters, including Secchi disk transparency, chlorophyll, and total phosphorus. My purpose here is to present a new ap- proach to the trophic classification of lakes. This new approach was developed because of frustration in communicating to the pub- lic both the current nature or status of lakes and their future condition after restoration when the traditional trophic classification system is used. The system presented hcrc, termed a trophic state index (TSI), in- volves new methods both of defining trophic status and of determining that status in lakes. All trophic classification is based on the division of the trophic continuum, howcvcr this is defined, into a series of classes termed trophic states. Traditional systems divide the continuum into three classes: oligotrophic, mesotrophic, and cutrophic. There is often no clear delineation of these divisions. Determinations of trophic state are made from examination of several di- verse criteria, such as shape of the oxygen curve, species composition of the bottom fauna or of the phytoplankton, conccntra- tions of nutrients, and various measures of biomass or production. Although each changes from oligotrophy to eutrophy, the changes do not occur at sharply defined places, nor do they all occur at the same place or at the same rate. Some lakes may be considered oligotrophic by one criterion and eutrophic by another; this problem is
A temporal-spatial analysis is made of the effect of an underwater artificial air diffusion system for the period 2008-2011, on the water quality of Embalse San Roque, Córdoba, through field data and LANDSAT5-TM satellite imagery. The temperature estimated using TM6 band, resulting with a determination coefficient of 0.94 and a mean square error of 0.4ºC. A multiple regression model was used to calculate log(chlorophyll-a) by means of TM1 and TM4 bands, obtaining a determination coefficient of 0.64 which was validated with a control group with and agreement equal 83 %. The processed images show the localized effect of the underwater artificial air diffusion system, improving the water quality in their neighbours.
Phytoplankton indicated by its photosynthetic pigment chlorophyll-a is an important pointer on lake ecology and a regularly monitored parameter within the European Water Framework Directive. Along with eutrophication and global warming cyanobacteria gain increasing importance concerning human health aspects. Optical remote sensing may support both the monitoring of horizontal distribution of phytoplankton and cyanobacteria at the lake surface and the reduction of spatial uncertainties associated with limited water sample analyses. Temporal and spatial resolution of using only one satellite sensor, however, may constrain its information value. To discuss the advantages of a multi-sensor approach the sensor-independent, physically based model MIP (Modular Inversion and Processing System) was applied at Lake Kummerow, Germany, and lake surface chlorophyll-a was derived from 33 images of five different sensors (MODIS-Terra, MODIS-Aqua, Landsat 8, Landsat 7 and Sentinel-2A). Remotely sensed lake average chlorophyll-a concentration showed a reasonable development and varied between 2.3 ± 0.4 and 35.8 ± 2.0 mg·m− 3 from July to October 2015. Match-ups between in situ and satellite chlorophyll-a revealed varying performances of Landsat 8 (RMSE: 3.6 and 19.7 mg·m− 3), Landsat 7 (RMSE: 6.2 mg·m− 3), Sentinel-2A (RMSE: 5.1 mg·m− 3) and MODIS (RMSE: 12.8 mg·m− 3), whereas an in situ data uncertainty of 48% needs to be respected. The temporal development of an index on harmful algal blooms corresponded well with the cyanobacteria biomass development during summer months. Satellite chlorophyll-a maps allowed to follow spatial patterns of chlorophyll-a distribution during a phytoplankton bloom event. Wind conditions mainly explained spatial patterns. Integrating satellite chlorophyll-a into trophic state assessment resulted in different trophic classes. Our study endorsed a combined use of satellite and in situ chlorophyll-a data to alleviate weaknesses of both approaches and to better characterise and understand phytoplankton development in lakes.
Improvements in the water quality of many freshwater and most coastal marine ecosystems requires reductions in both nitrogen and phosphorus inputs.
Analisis de una serie temporal de clorofila-a a partir de imágenes modis de un embalse eutrófico
  • A Germán
  • C Tauro
  • Verónica
  • I Andreo
  • A Bernasconi
Análisis de una serie temporal de clorofila-a a partir de imágenes modis de un embalse eutrófico
  • A Germán
  • C Tauro
  • I Verónica Andreo
  • A Bernasconi
  • Ferral
A. Germán, C. Tauro, Verónica Andreo, I. Bernasconi, and A. Ferral. Análisis de una serie temporal de clorofila-a a partir de imágenes modis de un embalse eutrófico. In Biennial Congress of Argentina (ARGENCON), 2016 IEEE, pages 1-6. IEEE, 2016.