Monitoring water quality in inland water bodies is critical for environmental management, yet traditional sampling methods are costly and spatially limited. Remote sensing offers a viable alternative by enabling large-scale assessment of water quality parameters
such as chlorophyll-a (Chl-a) concentrations through empirical models linking spectral indices to in situ measurements. However, the accuracy of these models depends on representative sampling strategies that capture spatial and temporal variability.
This study evaluates a clustering-based approach to optimize sampling site selection in the Solıs Dam, Mexico, using Sentinel-2 imagery. We analyzed a one-year time series of Sentinel-2 data to compute spectral indices related to Chl-a and turbidity. Unsupervised K-means clustering was applied to stratify the reservoir into zones of distinct water quality variability, guiding the placement of 20 sampling sites. Field campaigns during dry and wet seasons (2024) provided Chl-a measurements, which were correlated with spectral indices. The Gi033BDA index showed the strongest correlation (R2 > 0.7, p < 0.01) and was used to develop a linear regression model for Chl-a estimation. Results confirmed that clustering-derived sampling points effectively represented spatial variability, though temporal mismatches (1-day lag) and samples location inaccuracies introduced minor errors. The method demonstrates how pre-stratification using remote sensing can enhance sampling efficiency while maintaining model accuracy. This approach is particularly valuable for large-scale monitoring, reducing reliance on exhaustive field campaigns. Future work should address temporal dynamics and sensor resolution trade-offs for broader applicability.

Lake Cuitzeo, the second-largest freshwater lake in Mexico, is highly eutrophic and exhibits pronounced spatial and seasonal variability. Monitoring chlorophyll-a, a key indicator of phytoplankton biomass and trophic status, is critical for understanding ecological dynamics and supporting local communities reliant on the lake. In this study, hyperspectral EnMAP imagery from the dry (April 2025) and wet (August 2025) seasons was used to assess chlorophyll-a distribution of the East section of the lake. Water pixels were identified using the Modified Normalized Difference Water Index (MNDWI), and chlorophyll-a concentrations were estimated with the Normalized Difference Chlorophyll Index (NDCI), effective in optically complex waters. Lake-wide mean NDCI values showed no significant seasonal difference (t = 1.27, p = 0.204); however, spatial analysis revealed heterogeneity. Elevated wet-season chlorophyll values were concentrated in the central-western basin, while higher dry-season values occurred along shallow eastern shorelines. These findings highlight the importance of spatially explicit monitoring for detecting localized eutrophication and informing effective management strategies to preserve the ecological and socio-economic health of Lake Cuitzeo.

La percepción remota ha cobrado mayor relevancia como herramienta para el monitoreo ambiental, particularmente en la evaluación de la calidad del agua de cuerpos superficiales. Esta metodología ofrece ventajas operativas, temporales y espaciales frente al muestreo tradicional, especialmente en presas que cumplen funciones estratégicas de abastecimiento y regulación. En este estudio se evaluó el desempeño de 17 índices espectrales derivados de Sentinel-2 para la estimación indirecta de parámetros biológicos, ópticos y fisicoquímicos: clorofila-a, transparencia, turbidez y sólidos suspendidos totales (SST) en la presa Cointzio, Michoacán. Entre noviembre de 2023 y marzo de 2025 se recolectaron 92 muestras in situ y se procesaron 162 imágenes multiespectrales mediante el paquete waterquality en R. Se emplearon tres ventanas temporales según la coincidencia entre los índices y los datos de campo: coincidencia exacta (0 días), desfase de 1-2 días y una base con ambas. Los análisis de la base combinada mostraron correlaciones significativas (p < 0.0001), destacando la relación negativa entre transparencia y turbidez (r = –0.9165, R² = 0.8401), así como entre transparencia y SST (r = –0.8522, R² = 0.7263). Los índices TurbBow06RedOverGreen (R² = 0.8219) y Be16NDTIblue (R² = 0.6145) representaron eficazmente la transparencia. La clorofila-a mostró correlación con turbidez (r = 0.7098) y transparencia (r = –0.8318), destacando el índice Am092Bsub con R² = 0.5141. Se observó relación entre turbidez y SST (r = 0.8943, R² = 0.7998), siendo los índices Be16NDTIblue y TurbLath91RedOverBlue los más representativos (R² > 0.61); y los mejores estimadores de SST fueron estos mismos índices, con coeficientes de determinación superiores a 0.63. Estos resultados respaldan el uso de índices espectrales como herramientas eficaces para la estimación remota de parámetros clave de calidad del agua aun con desfase temporal de 1-2 días. No obstante, su aplicabilidad no debe considerarse estandarizada, ya que cada cuerpo de agua presenta respuestas particulares ante factores ambientales como el cambio climático, las presiones de demanda hídrica y la dinámica ecosistémica local. En este sentido, los hallazgos contribuyen al desarrollo de marcos metodológicos adaptables que pueden orientar el diseño de protocolos replicables, pero sujetos a calibraciones específicas para cada cuerpo de agua, que coadyuven en la toma de decisiones en políticas hídricas.

The impacts of climate change in the Mediterranean basin are characterized by the alternation between prolonged droughts and episodes of high-intensity torrential rainfall. These hydrometeorological extremes contribute to severe soil erosion, the extent of which is strongly influenced by the land cover within the catchment area, ultimately affecting the quality of the receiving water body. Taking advantage of the increased availability of hyperspectral imagery in recent years, this study aims to estimate both the concentration and composition of suspended matter in inland waters situated in the eastern Iberian Peninsula. For this purpose, we have relied on 55 in situ samples and their respective reflectance spectra, for 36 sites from the L2D product of the PRISMA mission, and for 19 sites from the L2A product of the EnMAP mission. To obtain the algorithms, we first use the ARTMO toolbox to identify the best-performing ones and then determine the same retrieval algorithm for both sensors. Trying to explain the theoretical basis of the algorithms using simple bio-optical theory. For the organic fraction, its strong relationship with chlorophyll-a concentration in our data has influenced the selection of its estimation algorithm, which uses the 714 nm band—an inflection point caused by maximum phytoplankton scattering and water absorption. For the inorganic fraction, the best-performing algorithm was based on the ratio between the ~705 nm band—near the reflectance peak—and the 551 nm band, which corresponds to a region of minimal phytoplankton absorption. Finally, the total suspended matter is the sum of both fractions.