Search

ABSTRACT Remote-sensing data are essential to evaluate dynamic processes such as eutrophication and increases in the concentration of suspended sediments in continental aquatic systems. The aim of this study, therefore, was to develop models to estimate chlorophyll-a concentrations from remote-sensing data in continental waters of the Brazilian semi-arid region. The study area corresponds to the Orós reservoir, located in the state of Ceará. The models were developed based on measurements taken at 20 sampling points. Water samples were collected from the reservoir to i) analyse the chlorophyll-a, electrical conductivity, pH and turbidity; ii) take optical measurements in situ of water transparency and spectral radiance. Radiance measurements were carried out using an ASD FieldSpec®3 Hi-Res spectroradiometer. The spectral data were later transformed into reflectance values and used to test the performance of several models found in the literature for estimating chlorophyll-a. The results showed that for the three-band model, the maximum value for the coefficient of determination (R2), of 0.88, was obtained using the λ1 = 660 nm, λ2 = 690 nm and λ3 = 717 nm spectral bands. The model employing two spectral bands presented the best performance (R2 = 0.87) in the λ1 = 660 nm and λ2 = 690 nm bands. An absolute mean error of 5.35 and 5.00 µg L-1 was found for the three- and two-band models respectively. The developed models are reliable, showing that chlorophyll-a concentrations can be quantified from remote-sensing field data with a high degree of accuracy.

Due to increasing algae bloom occurrence and water degradation on a global scale, there is a demand for water quality monitoring systems based on remote sensing imagery. This paper describes the scientific, theoretical, and methodological background for creating a cloud-computing interface on Google Earth Engine (GEE) which allows end-users to access algae bloom related products with high spatial (30 m) and temporal (~5 day) resolution. The proposed methodology uses Sentinel-2 images corrected for atmospheric and sun-glint effects to generate an image collection of the Normalized Difference Chlorophyll-a Index (NDCI) for the entire time-series. NDCI is used to estimate both Chl-a concentration, based on a non-linear fitting model, and Trophic State Index (TSI), based on a tree-decision model classification into five classes. Once the Chl-a and TSI algorithms had been calibrated and validated they were implemented in GEE as an Earth Engine App, entitled Algae Bloom Monitoring Application (AlgaeMAp). AlgaeMAp is the first online platform built within the GEE platform that offers high spatial resolution of water quality parameters. The App benefits from the huge processing capability of GEE that allows any user with internet access to easily extract detailed spatial (30 m) and long temporal Chl-a and TSI information (from August 2015 and with images every 5 days) throughout the most important reservoirs in the State of São Paulo/Brazil. The application will be adapted to extend to other relevant areas in Latin America.

Nowadays, the monitoring of water is essential for the sustainability and better management of water resources. The use of remote sensing data is important, since it allows evaluation of dynamic problems in aquatic systems, such as the eutrophication of bodies of water and suspended sediment. The aim of this study was to estimate chlorophyll-a concentrations in a reservoir of the semi-arid region of Brazil using simulated orbital-sensor data, as an aid in the management of water resources. The study area corresponded to the Orós reservoir, in the State of Ceará, Brazil. Water samples for analysis of the chlorophyll-a and measurements of the spectral radiance of the aquatic system were collected from 20 points. The radiance was measured by spectroradiometer. The data were collected in June and August of 2011. The model using three bands of the MERIS sensor (7, 9 and 10) presented an R2 of 0.84. For the two-band model (7 and 9), the value of R2 was 0.85. The waters of the Orós reservoir were all classified as eutrophic. The main optically active component in modelling the shape of the spectra was chlorophyll-a. The models showed a mean absolute error (MAE) of 3.45 and 3.61 μg L-1 for the three- and two-band models respectively. The models displayed high coefficients of determination, i.e. the simulations show the feasibility of estimating chlorophyll-a concentration from the data of the MERIS orbital sensor. = O monitoramento da água nos dias atuais é essencial para a sustentabilidade e para a melhor gestão dos recursos hídricos. O uso de dados de sensoriamento remoto é relevante, uma vez que permite a avaliação de problemas dinâmicos em sistemas aquáticos, tais como a eutrofização de corpos de água e dos sedimentos em suspensão. Objetivou-se estimar as concentrações de clorofila-a em reservatório da região semiárida do Brasil, usando dados simulados de sensor orbital, como suporte à gestão dos recursos hídricos. A área de estudo corresponde ao açude Orós, no estado do Ceará. As amostras de água para análise da clorofila-a e as medições da radiância espectral do sistema aquático foram coletadas em 20 pontos. As medidas de radiância foram realizadas utilizando um espectrorradiômetro. Os dados foram coletados nos meses de junho e agosto de 2011. O modelo usando três bandas (7, 9 e 10) do sensor MERIS apresentou R2 igual a 0,84. Para o modelo de duas bandas (7 e 9), o valor de R2 foi de 0,85. As águas do reservatório Orós foram todas classificadas como eutróficas. O principal componente opticamente ativo modelando a forma dos espectros foi clorofila-a. Os modelos apresentaram um erro médio absoluto (EMA) de 3,45 e 3,61 μg L-1, respectivamente, para os modelos de três e de duas bandas. Os modelos apresentaram coeficientes de determinação altos, ou seja, as simulações mostram a viabilidade de se estimar as concentrações de clorofila-a a partir dos dados do sensor orbital MERIS.

The diffuse attenuation coefficient of downwelling irradiance (Kd) is an essential parameter for inland waters research by remotely sensing the water transparency. Lately, Kd semi-analytical algorithms substituted the empirical algorithms widely employed. The purpose of this research was to reparametrize a semi-analytical algorithm to estimate Kd and then apply it to a Sentinel-2 MSI time-series (2017–2019) for the Três Marias reservoir, Brazil. The results for the Kd semi-analytical reparametrization achieved good accuracies, reaching mean absolute percentage errors (MAPE) for bands B2, B3 and B4 (492, 560 and 665 nm), lower than 21% when derived from in-situ remote sensing reflectance (Rrs), while for MSI Data, a derived MAPE of 12% and 38% for B2 and B3, respectively. After the application of the algorithm to Sentinel-2 images time-series, seasonal patterns were observed in the results, showing high Kd values at 492 nm during the rainy periods, mainly in the tributary mouths, possibly due to an increase in the surface runoff and inflows and outflow rates in the reservoir watershed.

AIM: This paper describes an experiment carried out to evaluate in vivo fluorescence (IVF) as an alternative method for chlorophyll-a estimation in optically complex aquatic environment (Amazon floodplain lakes) METHODS: The experiment consisted of collecting in situ measurements at 26 sampling stations distributed throughout Curuai floodplain lakes. For each sampling station the following parameters were measured: temperature, turbidity, depth, Secchi depth, chlorophyll-a (Chl-a) concentration, total suspended solids (TSS) and dissolved organic carbon (DOC), concurrently with several transects of IVF. Two methods were tested for quantifying the fluorescence measurement to be used as input for the chlorophyll-a estimates: instantaneous IFV and average IVF. Global and regional models were tested and assessed by analyzing optically active components (Chl-a, DOC and TSS) of the water. RESULTS: Regardless of fluorescence estimating method, the results indicate that it was not possible to fit a global model for estimating Chl-a from IVF for all the lakes in the Curuai floodplain. Regional models provided contrasting results according to the concentration of optically active components. The best results were observed for aquatic systems with a single dominant component homogenously distributed throughout the lake. The results highlight the influence of the ratios Chl-a/TSS, Chl-a/DOC and Phaephytin/Chl-a in the relationship between IVF and chlorophyll concentration. CONCLUSIONS: It was not possible to develop a global model to account for the entire region of Curuai floodplain. The search for regional models provided insights on the main factors affecting the relationship between IVF and Chl-a concentration. Nevertheless this work reinforces the great potential of fluorometry technique, since even with a small number of samples it was possible to set a good model in the main lake of the Curuai floodplain. In spite the fact that this is not an accurate method, it is very useful for assessing the chlorophyll spatial distribution with relatively low cost. These possibilities are very interesting in the execution of field missions in the Amazon region.OBJETIVO: Este trabalho descreve os experimentos realizados para avaliar as medidas de fluorescência in vivo (IVF) em um ambiente aquático opticamente complexo (planície de inundação amazônica), como método alternativo para a quantificação da clorofila-a na água. MÉTODOS: Foram coletados parâmetros de qualidade da água em 26 estações amostrais distribuídas em diversos lagos da planície de inundação de Curuai: temperatura, pH, turbidez, condutividade, profundidade, transparência ao disco de Secchi, concentração de clorofila-a (Chl-a), total de sólidos em suspensão (TSS) e carbono orgânico dissolvido (DOC) simultaneamente a quilômetros de transectos de IVF. Dois métodos de determinação do valor de fluorescência foram testados para o ajuste de modelos de estimativa da concentração de clorofila: fluorescência instantânea e fluorescência média. Também foram testados um modelo global e modelos regionais, os quais foram analisados em termos dos componentes opticamente ativos na água (Chl-a, DOC e TSS). RESULTADOS: Os resultados indicam que independentemente do método de estimativa de fluorescência empregado, não foi possível ajustar um modelo global para a planície de Curuai. Os modelos regionais apresentaram resultados contrastantes, de acordo com a concentração dos componentes opticamente ativos. Melhores resultados foram observados nos sistemas aquáticos com um único componente dominante, distribuído homogeneamente pelo respectivo lago. Os resultados destacam a influência das razões Chl-a/TSS, Chl-a/DOC e Feofitina/Chl-a no ajuste entre a IVF e a Chl-a. CONCLUSÕES: Não foi possível ajustar um modelo global para a planície de Curuai através de medidas de IVF. Entretanto, a busca por modelos regionais forneceu informações sobre os principais fatores que afetam a relação entre a IVF e a Chl-a. Dessa forma, este trabalho reforça o grande potencial da fluorometria, pois mesmo com um baixo número de amostras foi possível estabelecer um bom modelo no principal lago da planície. Ainda que a técnica não possa ser considerada extremamente precisa, ela é útil para avaliar o padrão de variação espacial da clorofila, com baixo custo. Estas possibilidades são muito interessantes na realização de missões de campo na bacia amazônica.

A utilização de modelos bio-ópticos com o intuito de estimar a concentração de cloroï¬la (Chl-a) em águas do caso 2 tem sido amplamente estudados dentro do sensoriamento remoto, uma vez que esse parâmetro é determinante para a compreensão da qualidade de água. A abordagem empírica foi utilizada em dois reservatórios, com parâmetros limnológicos distintos, e o desempenho dos modelos comparados. Para isso, foram coletadas amostras de água e medidas radiométricas nos reservatórios de Nova Avanhandavan (Nav) e Barra Bonita (BB) localizados no rio Tietê, estado de São Paulo. Dos 20 pontos coletados em Nav, 13 foram aplicados na calibração dos modelos e 7 na validação; enquanto que em BB, 12 pontos foram utilizados para calibração e 8 para validação. As amostras de água foram obtidas para dar suporte às análises laboratoriais visando determinar a concentração de Chl-a por meio da extração em acetona 90%. Foram comparados modelos utilizando primeira e segunda derivadas, duas bandas, sendo uma localizada na região do azul e outra no verde e três bandas baseadas na região espectral do vermelho e infravermelho próximo. Além desses, foi aplicado o índice de Cloroï¬la da Diferença Normalizada (NDCI - Normalized Difference Chlorophyll Index) e uma razão de bandas, denominada TWICE, ajustada por meio da ferramenta web ICE (Interactive Correlation Environment). Com base nos resultados foi observado que o modelo que obteve menor erro associado a estimativa da concentração de Chl-a em Nav foi o TWICE, com ajuste linear (RMSE = 2,89 μg/L; ε = 46,48% e δ = 0,45), e para BB o modelo de duas bandas (2BD) com ajuste polinomial (RMSE = 37,49 μg/L; ε = 42,36% e δ = -1,97).

This study assessed sunglint effects in airborne high spatial and high spectral resolution images acquired by the SpecTIR sensor under different view-illumination geometries over the Brazilian Ibitinga reservoir (Case II waters). These effects were corrected using the Goodman et al. (2008) and the Kutser et al. (2009) methods, and a variant that used the continuum removal technique to calculate the oxygen absorption band depth. The performance of each method to removing sunglint effects was evaluated by a quantitative analysis of pre- and post-sunglint correction reflectance values (residual reflectance images). Furthermore, the analysis was supported by inspection of the reflectance differences along transects placed over homogeneous masses of waters or over specific portions of the scenes affected and non-affected by sunglint. Results showed that the algorithm of Goodman et al. (2008) produced better results than the other two methods, as it approached to zero the amplitude of the reflectance values between homogenous water masses free and contaminated by sunglint. The Kutser et al. (2009) method had also good performance, except for the most contaminated sunglint portions of the scenes. When the continuum removal technique was incorporated to the Kutser et al. (2009) method, results varied with the scene and were more sensitive to atmospheric correction artifacts and instrumental signal-to-noise ratio.

This paper presents a methodology for processing and applying high resolution bathymetric data acquired with a Lowrance-480M ecosounder in the Amazon floodplain supported by Landast/TM images for planning and executing the survey. 4600 km of transects were processed semi-automatically and integrated into a georeferenced database. A digital elevation model with 15 m horizontal resolution and 1 cm vertical resolution was generated for the floodplain. The changes in inundated area and volume of water on the floodplain were estimated. Regression models to predict flood area and water stored volume from water level were constructed. The results of this research show that water level and flooded area mapped from images are good enough for estimating water stored volume in the Lago Grande de Curuai.

Este artigo apresenta uma metodologia para processamento e utilização de dados batimétricos de alta resolução, coletados com o ecobatimetro - Lowrance- modelo 480M, em áreas inundáveis da planície do rio Amazonas. Imagens TM/Landsat foram utilizadas como suporte ao planejamento e execução do levantamento batimétrico. Um conjunto de 4600 km lineares de transectos gerados durante o levantamento, foi submetido a uma seqüência semi-automática de processamento, e o resultado integrado a uma base de dados georreferenciada. Um modelo de elevação digital com resolução horizontal de 15 metros e resolução vertical de 1 centímetro foi gerado para a planície. A dinâmica de área inundada e de volume de água armazenado na planície foi avaliada a partir destes dados. Modelos de regressão para estimativa de área inundada e de volume armazenado na planície, a partir no nível de água, foram construídos. Os resultados deste trabalho mostraram que a informação de nível de água e área inundada mapeada a partir de imagens são um bom "proxy" para a estimativa do volume de água armazenado na planície do Lago Grande de Curuai.

Abstract Originally developed for terrestrial science and applications, the US Geological Survey Landsat surface reflectance (SR) archive spanning ~ 40 yr of observations has been increasingly utilized in large‐scale water‐quality studies. These products, however, have not been rigorously validated using in situ measured reflectance. This letter quantifies and demonstrates the quality of the SR products by harnessing a sizeable global dataset (N = 1100). We found that the Landsat 8/9 SR in the green and red bands marginally meet the targeted accuracy requirements (30%), whereas the uncertainties in the blue and coastal‐aerosol bands ranged from 48% to 110%. We further observed > +25% biases in the visible bands of Landsat 5/7 SR, which can introduce an apparent downward trend when applied in time‐series analyses combined with Landsat 8/9. Users must exercise caution when using this archive for trend analyses, and progress in atmospheric correction is required to foster advanced applications of the Landsat archive for aquatic science.

The Amazon Basin is the largest on the planet, and its aquatic ecosystems affect and are affected by the Earth’s processes. Specifically, Amazon aquatic ecosystems have been subjected to severe anthropogenic impacts due to deforestation, mining, dam construction, and widespread agribusiness expansion. Therefore, the monitoring of these impacts has become crucial for conservation plans and environmental legislation enforcement. However, its continental dimensions, the high variability of Amazonian water mass constituents, and cloud cover frequency impose a challenge for developing accurate satellite algorithms for water quality retrieval such as chlorophyll-a concentration (Chl-a), which is a proxy for the trophic state. This study presents the first application of the hybrid semi-analytical algorithm (HSAA) for Chl-a retrieval using a Sentinel-3 OLCI sensor over five Amazonian floodplain lakes. Inherent and apparent optical properties (IOPs and AOPs), as well as limnological data, were collected at 94 sampling stations during four field campaigns along hydrological years spanning from 2015 to 2017 and used to parameterize the hybrid SAA to retrieve Chl-a in highly turbid Amazonian waters. We implemented a re-parametrizing approach, called the generalized stacked constraints model to the Amazonian waters (GSCMLAFW), and used it to decompose the total absorption αt(λ) into the absorption coefficients of detritus, CDOM, and phytoplankton (αphy(λ)). The estimated GSCMLAFWαphy(λ) achieved errors lower than 24% at the visible bands and 70% at NIR. The performance of HSAA-based Chl-a retrieval was validated with in situ measurements of Chl-a concentration, and then it was compared to literature Chl-a algorithms. The results showed a smaller mean absolute percentage error (MAPE) for HSAA Chl-a retrieval (36.93%) than empirical Rrs models (73.39%) using a 3-band algorithm, which confirms the better performance of the semi-analytical approach. Last, the calibrated HSAA model was used to estimate the Chl-a concentration in OLCI images acquired during 2017 and 2019 field campaigns, and the results demonstrated reasonable errors (MAPE = 57%) and indicated the potential of OLCI bands for Chl-a estimation. Therefore, the outcomes of this study support the advance of semi-analytical models in highly turbid waters and highlight the importance of re-parameterization with GSCM and the applicability of HSAA in Sentinel-3 OLCI data.

The aim of this work is to test the state-of-the-art of water constituent retrieval algorithms for phycocyanin (PC) and chlorophyll-a (chl-a) concentrations in Brazilian reservoirs from hyperspectral PRISMA images and concurrent in situ data. One near-coincident Sentinel-3 OLCI dataset has also been considered for PC mapping as its high revisit time is a relevant element for mapping cyanobacterial blooms. The testing was first performed on remote sensing reflectance (Rrs), as derived by applying two atmospheric correction methods (6SV, ACOLITE) to Level 1 data and as provided in the corresponding Level 2 products (PRISMA L2C and OLCI L2-WFR). Since PRISMA images were affected by sun glint, the testing of three de-glint models was also performed. The applicability of Semi-Analytical (SA) and Mixture Density Network (MDN) algorithms in enabling PC and chl-a concentration retrieval was then tested over three PRISMA scenes; in the case of PC concentration estimation, a Random Forest (RF) algorithm was further applied. Regarding OLCI, the SA algorithm was tested for PC estimation; notably, only SA was calibrated with site-specific data from the reservoir. The algorithms were applied to the Rrs spectra provided by PRISMA L2C products—and those derived with ACOLITE, in the case of OLCI—as these data showed better agreement with in situ measurements. The SA model provided low median absolute error (MdAE) for PRISMA-derived (MdAE = 3.06 mg.m−3) and OLCI-derived (MdAE = 3.93 mg.m−3) PC concentrations, while it overestimated PRISMA-derived chl-a (MdAE = 42.11 mg.m−3). The RF model for PC applied to PRISMA performed slightly worse than SA (MdAE = 5.21 mg.m−3). The MDN showed a rather different performance, with higher errors for PC (MdAE = 40.94 mg.m−3) and lower error for chl-a (MdAE = 23.21 mg.m−3). The results overall suggest that the model calibrated with site-specific measurements performed better and indicates that SA could be applied to PRISMA and OLCI for remote sensing of PC in Brazilian reservoirs.

A propagação de plantas aquáticas nos corpos d’água tem contribuído para o comprometimento dos diversos usos da água. Para realizar o controle dessa vegetação é essencial que se faça a identificação e o monitoramento das espécies presente nos mananciais. O sensoriamento remoto é uma tecnologia que pode ser empregada na coleta de dados sobre macrófitas, sendo o NDVI uma técnica bastante utilizada na discriminação da vegetação. Objetivou-se investigar a aplicabilidade do sensoriamento remoto na avaliação de plantas aquáticas em diferentes níveis de aquisição de dados. O estudo foi realizado no açude Muquêm, localizado no município de Cariús, Ceará, que está inserido na bacia do Alto Jaguaribe, Brasil. Com o resultado obtido a partir da coleta das plantas aquáticas foi possível identificar três espécies (Polygonum ferrugineum Wedd, Salvinia oblongifolia Mart e Oxycaryum cubense (Poepp & Kunth) Palla). A partir das respostas espectrais dos alvos foi observado a diferença entre as espécies de vegetação e fases fenológicas. Através da utilização do NDVI nas imagens de Landsat-8 foi possível mapear a área do espelho d’água ocupado por vegetação aquática. Verificar que houve uma redução na área ocupada por vegetação, passando de 37,62% para 30,21%, para as imagens de julho e setembro de 2014, respectivamente. Verificou-se também que tal redução estava relacionada ao processo de senescência e morte destes indivíduos. Com a aplicação do NDVI também é possível a identificação e mapeamento de diferentes estádios da vegetação: senescência e decomposição.

Although mining plays an important role for the economy of the Amazon, little is known about its attributes such as area, type, scale, and current status as well as socio/environmental impacts. Therefore, we first propose a low time-consuming and high detection accuracy method for mapping the current mining areas within 13 regions of the Brazilian Amazon using Sentinel-2 images. Then, integrating the maps in a GIS (Geography Information System) environment, mining attributes for each region were further assessed with the aid of the DNPM (National Department for Mineral Production) database. Detection of the mining area was conducted in five main steps. (a) MSI (MultiSpectral Instrument)/Sentinel-2A (S2A) image selection; (b) definition of land-use classes and training samples; (c) supervised classification; (d) vector editing for quality control; and (e) validation with high-resolution RapidEye images (Kappa = 0.70). Mining areas derived from validated S2A classification totals 1084.7 km2 in the regions analyzed. Small-scale mining comprises up to 64% of total mining area detected comprises mostly gold (617.8 km2), followed by tin mining (73.0 km2). The remaining 36% is comprised by industrial mining such as iron (47.8), copper (55.5) and manganese (8.9 km2) in Carajás, bauxite in Trombetas (78.4) and Rio Capim (48.5 km2). Given recent events of mining impacts, the large extension of mining areas detected raises a concern regarding its socio-environmental impacts for the Amazonian ecosystems and for local communities.

Satellite data provide the only viable means for extensive monitoring of remote and large freshwater systems, such as the Amazon floodplain lakes. However, an accurate atmospheric correction is required to retrieve water constituents based on surface water reflectance ( R W ). In this paper, we assessed three atmospheric correction methods (Second Simulation of a Satellite Signal in the Solar Spectrum (6SV), ACOLITE and Sen2Cor) applied to an image acquired by the MultiSpectral Instrument (MSI) on-board of the European Space Agency’s Sentinel-2A platform using concurrent in-situ measurements over four Amazon floodplain lakes in Brazil. In addition, we evaluated the correction of forest adjacency effects based on the linear spectral unmixing model, and performed a temporal evaluation of atmospheric constituents from Multi-Angle Implementation of Atmospheric Correction (MAIAC) products. The validation of MAIAC aerosol optical depth (AOD) indicated satisfactory retrievals over the Amazon region, with a correlation coefficient (R) of ~0.7 and 0.85 for Terra and Aqua products, respectively. The seasonal distribution of the cloud cover and AOD revealed a contrast between the first and second half of the year in the study area. Furthermore, simulation of top-of-atmosphere (TOA) reflectance showed a critical contribution of atmospheric effects (>50%) to all spectral bands, especially the deep blue (92%–96%) and blue (84%–92%) bands. The atmospheric correction results of the visible bands illustrate the limitation of the methods over dark lakes ( R W < 1%), and better match of the R W shape compared with in-situ measurements over turbid lakes, although the accuracy varied depending on the spectral bands and methods. Particularly above 705 nm, R W was highly affected by Amazon forest adjacency, and the proposed adjacency effect correction minimized the spectral distortions in R W (RMSE < 0.006). Finally, an extensive validation of the methods is required for distinct inland water types and atmospheric conditions.

The objective of this paper is to compare the performance of medium spatial resolution (250 m and 500 m) Terra MODIS images to finer resolution Landsat ETM+ images. MODIS Terra images have high frequency of acquisition (1 day revisit at high latitudes) and that makes them more useful for inland water studies. Assessing the performance of ETM and MODIS to map relevant features for the functioning of aquatic systems is very important for fostering water resource remote sensing. To carry out the comparison, concurrent ETM+ and MODIS images were acquired over the Lago Grande de Curaui Lake. The images were georeferenced and resample to the 100 x 100 m ground resolution using a next neighbor algorithm to make the data comparable. The resampling rationale was making the digital processing easier, without changing the effective resolution properties of the original data. It was assumed that the improved MODIS radiometric resolution as compared to the ETM+.would make both data set comparable at a 100m x 100m resolution. In the next step, a series of tests were carry out in order to define the best approach to map aquatic system features such as lakes, islands, and levees in the study area. The tests indicated that the best approach was the segmentation of the shadow fraction derived from the application of the linear unmixing model to both ETM+ and MODIS images. The segmentation was then followed by the application of a non-supervised region classifier. The final classes were mapped into two categories: lakes (water) and island (land). The polygon distribution generated by the classification procedure was then statistically analyzed to assess the polygon size frequency of features mapped by each data set. The results showed that ETM+ and MODIS were able to recover the water surface area in the studied region with a difference of 11 %. The analyses of polygon distribution, however, showed that there were many polygons which were detected in MODIS image data, but not in ETM+ image data. This result suggests that the spectral and radiometric resolution improvements presented by MODIS image tend to compensate for the losses in spatial resolution. That is to say that under the boundary conditions adopted in this study, MODIS images performed better than degraded ETM+ images of 100m x 100m. This explains the small difference in performance presented by comparing two sensors with such large differences in nominal spatial resolution.