Recibido: 31/10/2024
Aceptado: 20/02/2025
Disponible en línea: 31/03/2025
Publicado: 31/07/2025

REVISTA DE TELEDETECCIÓN
Asociación Española de Teledetección
(2025) 66, 22733
ISSN 1133-0953
EISSN 1988-8740
https://doi.org/10.4995/raet.2025.22733

Baseflow measurement in mountain rivers using LSPIV: A case study of the Tarqui and Yanuncay rivers in the Ecuadorian Andes

Santiago A. Ochoa-García^*1, Leandro Massó^2,3, Antoine Patalano^2,3, Carlos M. Matovelle-Bustos⁴, Paola V. Delgado-Garzón⁵

¹ Universidad Católica de Cuenca (UCACUE), Grupo de Investigación Ambiente, Ciencia y Energía, Carrera de Ingeniería Civil, Av. de las Américas y Humboldt, Cuenca, Ecuador.

² Universidad Nacional de Córdoba (UNC), Facultad de Ciencias Exactas, Físicas y Naturales (FCEFyN), Av. Vélez Sarsfield 1611, Ciudad Universitaria, Córdoba, Argentina.

³ Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Instituto de Estudios Avanzados en Ingeniería y Tecnología (IDIT), Av. Vélez Sarsfield 1611, Ciudad Universitaria, Córdoba, Argentina.

⁴ HYDROLAB, Centro de Investigación, Innovación y Transferencia de Tecnología, Universidad Católica de Cuenca, Cuenca, Ecuador.

⁵ Universidad Católica de Cuenca (UCACUE), Carrera de Ingeniería Civil, Av. de las Américas y Humboldt, Cuenca, Ecuador.

Abstract: This study is motivated by the difficulty of applying experimental techniques to characterize base flows in mountain rivers. Intrusive instruments are not optimal for measuring low flow rates, as they require a minimum depth to be submerged and to measure flow velocity. The LSPIV methodology was applied using an Autel Evo II RTK Series 3 UAV. The results were validated through measurements taken with a Redback current meter, showing that the flow rates and velocity fields obtained with the presented techniques are of the same order of approximation. The flow velocity fields resulting from the application of LSPIV enabled the identification of typical flow characteristics in mountain rivers with gravel and boulder beds: zones of acceleration and turbulent mixing, stagnation areas due to obstacles within the flow, flow recirculation, and shear regions caused by interaction with existing morphological structures. Thus, the LSPIV technique is presented as a valuable tool for characterizing extreme flows in mountain rivers using non-intrusive methods.

Key words: LSPIV, surface flow, flow measurement, base flows.

Medición de los caudales de estiaje en ríos de montaña usando LSPIV: Un caso de estudio de los ríos Tarqui y Yanuncay en los Andes ecuatorianos

Resumen: Este trabajo se motiva en la dificultad de aplicar técnicas experimentales para la caracterización de caudales base en ríos de montaña, debido a que los instrumentos intrusivos no son óptimos para medir láminas de caudal de estiaje, ya que es necesaria una altura mínima para que los instrumentos queden sumergidos y midan la velocidad del flujo correspondiente. Se aplicó la metodología LSPIV con el uso del dron Autel Evo II RTK Serie 3. Los resultados fueron validados a partir de medidas realizadas con un correntímetro Redback, mostrando que los caudales y campos de velocidad obtenidos con las técnicas presentadas se encuentran en el mismo orden de aproximación. Los campos de velocidad de flujo resultantes de la aplicación de LSPIV permitieron distinguir características típicas de los flujos en ríos de montaña con lechos de gravas y cantos rodados: secciones de aceleración y mezcla turbulenta, zonas de estancamiento debido a la existencia de obstáculos sobre el flujo, recirculación del flujo y regiones de corte debido a la interacción del flujo con las estructuras morfológicas existentes. En este sentido, la técnica LSPIV se presenta como una valiosa herramienta para caracterizar flujos extremos en ríos de montaña mediante técnicas no intrusivas.

Palabras clave: LSPIV, flujo superficial, medición de caudales, flujos de estiaje.

1. Introducción

Flow measurement in canals and rivers is a key area of hydrometry, the science and practice of water measurement. Methods of flow control vary depending on stream type (Alongi, 2022). Stream types can be classified based on eight primary variables: width, depth, velocity, flow, slope, roughness of bed and bank materials, sediment load, and sediment size (Singh et al., 2003). Additionally, experimental quantification of hydrological variables in rivers during peak and base flow events is essential for the effective management of water resources and the validation of hydrological models (Díaz-Lozada, 2019).

In general, the experimental conditions present during these atypical events make it impossible to use traditional experimental techniques, such as mechanical velocimeters, Acoustic Doppler Current Profilers (ADCP), and Acoustic Doppler Velocimeters (ADV) (Patalano et al., 2017). Techniques using intrusive instruments are unsuitable for measurements in shallow water channels, large-scale flow phenomena (e.g., flood flows), or inaccessible locations (e.g., densely vegetated systems) (Perks et al., 2020).

Floods and base flows are events with a low probability of occurrence and are, therefore, often poorly characterized. Accurate estimation of low water flows is crucial for the effective management of water resources in cases of hydrological scarcity. Intrusive instruments are not optimal for measuring low water flow depths, as a minimum height is required for the instruments to be submerged and to measure flow velocity (Lobo, 2019).

In recent years, non-intrusive measurement techniques for characterizing flows in large-scale river environments have gained considerable popularity (Aberle et al., 2017; Jolley et al., 2021). Studies have shown that remote sensing techniques can quickly, reliably, and comprehensively quantify surface flow velocities. Both the scientific community and the commercial sector have shown growing interest in the ‘image velocimetry’ approach, particularly when combined with optical remote sensing technologies. Particle image velocimetry has advanced rapidly over the past decade. Adrian (1991) demonstrated that a two-dimensional image velocimetry technique could provide accurate, high-quality measurements of instantaneous fields in a range of gas and liquid flows at laboratory scale, covering velocities from millimeters per second to several hundred meters per second. Fujita et al. (1998) further developed Adrian’s technique by applying it in real river environments, coining the term Large Scale Particle Image Velocimetry (LSPIV).

Imaging velocimetry techniques use optical cameras mounted on ground-based or aerial platforms, such as Unmanned Aerial Vehicles (UAVs), to capture images in inaccessible environments at relatively low cost. By tracking the motion of visible particles, imaging velocimetry generates a two-dimensional velocity field (Koutalakis et al., 2019). Particle Image Velocimetry (PIV), for example, is a commonly applied image velocimetry algorithm in laboratory settings within experimental fluid mechanics (Aberle et al., 2017). However, advancements over the past three decades have enabled the application of conventional PIV in large-scale river environments, often referred to as Large-Scale PIV (LSPIV) (Sharif, 2022).

Surface flow velocity measurements derived from LSPIV can help reveal flow characteristics in water systems. Studies (e.g., Fujita et al., 1998; Meselhe et al., 2004; Muste et al., 2008; Le Coz et al., 2010; Strelnikova et al., 2020; Bandini et al., 2021; Sharif, 2022) indicate that LSPIV provides valuable insights into hydrodynamic, morphological, and ecological processes crucial for river hydraulic engineering applications. While recent advancements highlight LSPIV’s significant potential, it is essential to recognize that LSPIV technology and its results are still far from perfect (Detert, 2021; Massó et al., 2024). Several technical and practical challenges remain unresolved, and experience with UAV-based LSPIV is currently limited (Jolley et al., 2021). Therefore, LSPIV is still considered a novel approach for non-intrusive remote sensing of river flows and remains an active field of research (Sharif, 2022).

Additionally, the high temporal and spatial variability of flow and turbulence conditions in steep mountain streams makes parameter characterization challenging. High slopes, large particles, and highly turbulent flow complicate the accurate quantification of bedload transport, especially in secondary streams, which further hinders precise flow velocity estimates (Carrillo-Serrano, 2021). The presence of large-diameter elements such as gravels and boulders, typical of mountain streams in the Ecuadorian Andes, results in variations in flow hydraulics, turbulence intensity, hydraulic jumps, zones of flow acceleration and deceleration, and significant spatial variability in boundary shear stress (Papanicolaou et al., 2001; Wilcox and Wohl, 2007).

Therefore, the objective of this research is to apply and validate the LSPIV technique using UAVs to characterize low water flows in the Tarqui and Yanuncay rivers in the city of Cuenca, which exhibit typical mountain characteristics of the Ecuadorian Andes.

2. Materials and Methods

2.1. Study Area

The study area is located in the city of Cuenca, in southern Ecuador, which is the third most politically and economically important city in the country, after Quito and Guayaquil. Cuenca is known for its four rivers (Tomebamba, Yanuncay, Tarqui, and Machángara), which have played a crucial role in the continuous development of the southern region of Ecuador. These rivers are part of the tributaries of the Paute River basin, where significant hydroelectric projects generate nearly 30% of Ecuador’s hydroelectric energy (CELEC EP, 2024). Due to their mountainous terrain, the tributaries of the upper Paute River basin are classified as mountain rivers, characterized by steep slopes and large hillsides that contribute to flash floods, posing risks of channel overflow and flooding in residential areas of Cuenca. The experimental characterization was conducted near the confluence of the Yanuncay and Tarqui rivers, located in the southern part of Cuenca (Figure 1).

The flow of four rivers through the city of Cuenca creates three important confluences within the urban area (Figure 1). In this work, we present the results obtained in the vicinity of the first confluence, where the Tarqui River meets the Yanuncay River. Both rivers exhibit characteristics typical of torrential or mountain rivers, with material sizes ranging from silts (<0.063 mm) to large boulders (>500 mm). In these types of rivers, the primary source of coarse material in the basin (such as cobbles and gravels) is the conglomeratic lithology along the edge of the depression, resulting in cobbles that are already rolled by the time they reach the rivers in the study area. The bed’s morphological characteristics create secondary currents, areas of stagnation, and islands formed by coarse material, complicating the characterization of velocities using traditional methodologies.

2.2. Experimental Methods

To apply the measurement technique, control sections were established both upstream and downstream of the confluence of the Tarqui and Yanuncay rivers. This work presents four measurement campaigns conducted during periods of low water levels, specifically between February 1 and April 17, 2024. Five control sections (Figure 2) with varying levels of turbulence were selected for the measurements.

Regarding the control sections presented in Figure 2, section 1 (S1) corresponds to the flow of the Yanuncay river downstream of its confluence with the Tarqui river; this section presents flow characteristics with high turbulence, irregularities in the bottom with mean slope in the order of 2%. Sections S2 and S3 represent the Yanuncay river upstream of the confluence, and their characteristics differ in flow depth, section irregularity and bottom slope. In S2, very low flow depths influence the formation of surface irregularities due to the interaction of turbulent flow with the irregular solid boundaries of the riverbed; in this section the mean bottom slope is approximately 4%. In contrast, S3 presents calm waters (ideal for flow control sections) with greater depths than S2, a regular section and mean bottom slope on the order of 0.5%. The characteristics of the Tarqui River sections, S4 and S5, are similar to those of S2 and S3 with average bottom slopes on the order of 3% and 0.3%, respectively. The control sections are illustrated in Figure 3.

Images were captured using an Autel Evo II RTK Series 3 UAV, with a resolution of 3840×2160 px² and 30 fps, with the camera positioned directly above the water surface. Plastic cones were placed on the banks to delineate the analysis cross-section. The distance between the cones, measured with a tape measure, was used to define the pixel scale at ground level for each video. The water surface was seeded with wood chips to serve as surface flow tracers. The recorded videos averaged 6 minutes in length, which were then reduced to 30-second segments that provided the best natural lighting and shadow conditions. The artificial seeding of tracers in the section was visible in the images around 30 seconds into the recordings, allowing for optimal space-time heterogeneity of the tracer particles over the characterized control sections.

To validate the LSPIV results, velocity measurements were conducted by wading across the cross-section using a HyQuest Redback current meter. Measurements were taken at 20%, 60%, and 80% of the depth from the water surface across the full width of the cross-section, following ISO 748 standards (ISO 748, 2007).

2.3. Image Processing

Image extraction and processing were performed using RIVeR v2.6 (Patalano et al., 2017), which includes a set of tools for characterizing free surface flows through the application of PIV (Particle Image Velocimetry) and PTV (Particle Tracking Velocimetry) techniques. For this application, image extraction was conducted in grayscale, with the spatial resolution reduced to 2560×1440 px² to minimize computational costs. A time interval of 100.1 milliseconds between images was selected, extracting every third available frame due to the small pixel displacement of the tracers between consecutive images.

To perform image velocimetry calculations, PIVlab v2.59 (Thielicke and Stamhuis, 2014) integrated into RIVeR v2.6 was used. Pre-processing of the images was conducted to enhance the visualization of surface tracers. For each case, regions of interest (ROI) were defined and centered on the analysis section, with masks applied to static elements when necessary (e.g., riverbanks). The PIV algorithm employed was a multi-window (two-pass) fast Fourier transform (FFT) with warping. The interrogation window (IA) sizes for the first and second passes were 128×128 px² and 64×64 px², respectively, with a 50% overlap. Quasi-instantaneous surface velocity fields of the flow within the ROI were obtained for each reach, including the selected control sections.

If the primary objective of the analysis is to determine the mean velocity field (e.g., for flow discharge estimation), the tracers —whether natural (such as foam, tree branches, and coherent structures on the water surface) or artificial (such as wood chips or other ecologically harmless and biodegradable materials)— should be homogeneously and densely distributed across the water surface. In this case, the images should be analyzed in an Eulerian framework using PIV. Conversely, if the tracer density is sparse and individual trajectories are needed —such as for characterizing the flow velocity field near hydraulic structures— then PTV should be employed within a Lagrangian framework of analysis (Patalano et al., 2017).

PIV analysis was conducted using PIVlab, which incorporates several filters to improve image processing, including Contrast Limited Adaptive Histogram Equalization (CLAHE). This filter enhances intensity histograms across the entire data range (from 0 to 255 in 8-bit images) and operates on small areas of the image (Pizer et al., 1987). It is applied to increase the likelihood of detecting valid vectors in experimental images. The pixel shift can be calculated using either the robust direct correlation algorithm (DCC) (Stamhuis, 2006) or the discrete Fourier transform (DFT), which computes the correlation matrix in the frequency domain utilizing the FFT (Raffel et al., 2007).

3. Results

Applying the methodology presented in Section 2, four measurement campaigns were conducted to evaluate the LSPIV technique for characterizing mountain flows in the Ecuadorian Andes. Five cross-sections (S1, S2, S3, S4, and S5) were selected, each exhibiting different levels of turbulence and specific morphological characteristics in the riverbed that complicated the application of velocity measurements using a HyQuest Redback current meter. This study will consider information from the four measurement campaigns (C1, C2, C3, and C4), as detailed in Table 1. The characteristics of the control sections are illustrated in Figure 4.

Table 1 shows that the flow and velocity variables monitored using the two methods are of the same order of magnitude. According to the Froude number obtained for all cases, the analyzed flows are subcritical (Fr < 1) with low flow velocities. Sections S1, S2, and S4 exhibit similar characteristics, with Froude numbers ranging from 0.2 to 0.42, indicating velocities that generate turbulence in acceleration zones. In contrast, Sections S3 and S5 have very slow flow characteristics, with Fr < 0.2, which facilitates the identification of seeded particles during processing due to the lower variability in the three-dimensional turbulent velocities of the flow, reducing the likelihood of particle sinking.

It is also noted that in the first campaign (C1), control points on the cross-section were taken every 2 meters (Figure 4). However, inconsistencies in the continuity of the estimated flow rates were observed; thus, in campaigns 2, 3, and 4 (C2, C3, and C4), observations were taken every meter across the control cross-sections to provide greater detail in the velocity field, resulting in more precise flow rate estimates. The Absolute Percentage Error (APE) of the flow variables characterized according to the measurement techniques applied is presented in Table 2.

As shown in Table 2, most of the observed flow variables have an Absolute Percentage Error (APE) on the order of 10%. However, it is noted that the measurements taken during the last two campaigns exhibit greater stability in their variables. This improvement is attributed to the ease of applying the measurement techniques in sections with less turbulence and greater depths. This aspect will be discussed in more detail when contrasting the velocities observed along the control cross-sections (Figure 5).

The velocities measured with the analyzed techniques exhibit values consistent with those presented in Figure 5. When applying the LSPIV technique, the correction factor “α” is used to represent the ratio between the surface velocity and the depth-averaged velocity (Biggs et al., 2021). The correction factor α can fluctuate between 0.6 and 1.2 (Le Coz et al., 2011; Fujita, 2018). In the observations studied in this paper, there was a good correspondence between the velocities measured with LSPIV and the mean velocities obtained with the Redback windlass, with a correction factor equal to 1. This can be attributed to the low flow levels, the irregularity of the control sections, and the presence of recirculation regions typical of the characterized mountain rivers. Consequently, it is assumed that the logarithmic variation of the vertical velocity profile can be neglected. To evaluate the correlation of velocities obtained with the two methodologies in greater detail, Figure 6 presents scatterplots of the characterized velocity variables.

Figure 6 shows a strong correlation between the velocities observed with LSPIV and those measured with the Redback, with the slopes of the linear fits tending toward the 45° line, indicating a good fit. This observation can also be numerically justified by evaluating the correlation coefficients between the velocity series measured with LSPIV and Redback across the control sections. The Coefficient of Determination (R²), Kling-Gupta Efficiency (KGE), and Root Mean Square Error (RMSE) are presented in Table 3.

According to the results presented in Table 3, the errors when contrasting the observed velocities with LSPIV and Redback are acceptable in all sections, except for S1C1. In terms of correlation coefficients, good results are observed in 6 of the 8 analyzed sections (S4C1, S1C2, S2C2, S3C3, S4C4, and S5C4).

One variable that could increase uncertainty in surface velocity field measurements is wind speed during image acquisition. Wind can introduce several unwanted effects in flow characterization, including: wind shear stress (e.g., wind velocity profiles and turbulence), fetch, surface roughness (which interacts with wind-generated surface waves), surface tracer types (i.e., surface particles such as wood shavings or features like boils and eddies), and even the turbulent mixing characteristics of the channel itself (i.e., vertical mixing of wind-disturbed surface water) (Biggs et al., 2021). As previously mentioned, the presence of large-diameter elements, such as gravels and boulders, leads to variations in flow hydraulics and turbulence intensity that complicate mean flow measurements with intrusive instruments (Wilcox and Wohl, 2007). Therefore, the surface velocities characterized with LSPIV can be considered representative of the observed base flows. Figure 7 presents the surface flow fields corresponding to the analyzed control sections.

In the observations made, the macrorough flow stands out with marked heterogeneity in the spatial distribution of velocity, including the presence of ‘lateral’ flow towards the central zone of section S2 (Figure 7(b) and (e)). This case shows the largest discrepancy in velocity profiles between the two techniques (greater Froude difference, higher Flow APE, and larger Mean Flow Velocity APE). The marked bidirectionality of the flow and shallow depths are factors that may increase uncertainty in the windlass measurement, as it is an intrusive instrument that measures in 1D.

It is also observed that section S2 presents a geometry with minimum depth values in the central zone (Figure 4(b) and (e)), generating two main channels near the margins. U-shaped velocity profiles are then expected based on the section’s geometry (Figure 7(b) and (e)). A similar pattern is observed in section S4, where maximum depth zones are located near the left margin (Figure 4(c) and (f)), resulting in a profile with maximum velocities in that area (Figure 7(c) and (f)). Both techniques adequately represent the geometry in the velocity distribution for these cases.

Figure 7 demonstrates that the flow velocity fields measured with LSPIV reveal distinct characteristics of flows in mountain rivers, such as acceleration and turbulent mixing pathways, stagnation zones caused by obstacles in the flow, flow recirculation, and shear regions due to the interaction of the flow with existing morphological structures. This level of detail is unattainable with conventional techniques, which suffer from both low spatial resolution and intrusiveness. Therefore, the LSPIV technique is presented as a crucial tool for characterizing extreme flows in mountain rivers using non-intrusive methods.

4. Conclusions

In the context of the challenges associated with applying experimental techniques to characterize extreme flows in mountain rivers, the Large-Scale Particle Image Velocimetry (LSPIV) methodology was implemented using an Autel Evo II RTK Series 3 unmanned aerial vehicle (UAV). The results were validated against measurements taken with a Redback current meter, demonstrating that the flow rates and velocity fields obtained with both techniques were comparable. The ratio of the surface velocity observed with LSPIV to the average velocity measured with the Redback current meter was 1. This finding is attributed to the low flow levels, the irregularity of the control sections, and the presence of recirculation regions typical of the characterized mountain rivers.

We analyzed conditions that could increase the uncertainty in the characterization of flows using LSPIV, particularly concerning the observed surface velocity field relative to actual velocities. Specifically, the effect of wind speed during image acquisition can negatively impact the evaluation of flows with free surfaces. Additionally, the presence of large-diameter elements, such as gravels and boulders, introduces unique characteristics in the flow hydraulics of mountain rivers, complicating the application of techniques for assessing extreme flows.

The flow velocity fields obtained from the application of LSPIV highlighted characteristics typical of flows in mountain rivers with gravel and boulder beds. These include sections of acceleration and turbulent mixing, stagnation zones caused by obstacles in the flow, flow recirculation, and shear regions resulting from the interaction of the flow with existing morphological structures. In this context, the LSPIV technique is presented as a valuable tool for characterizing extreme flows in mountain rivers using non-intrusive methods.

5. Acknowledgements

We appreciate the funding from UCACUE within the PIC5P23-28 Research Project, thanks to this collaboration it was possible to have the equipment to measure the tributaries of Cuenca city.

References

Aberle, J., Rennie, C.D., Admiraal, D.M., Muste, M. 2017. Methods, Instrumentation, Data Processing and Management. Volume II: Instrumentation and Measurement Techniques. CRC Press/Balkema. https://doi.org/10.1201/9781315158921

Adrian, R.J. 1991. Particle-Imaging Techniques for Experimental Fluid Mechanics. Annu. Rev. Fluid Mech., 23, 261-304. https://doi.org/10.1146/annurev.fl.23.010191.001401

Alongi, F. 2022. River flow monitoring: LS-PIV technique, an image-based method to assess discharge. Doctorate Degree in Civil, Environmental and Materials Engineering. Dipartimento di Ingegneria. Universitá degli Studi di Palermo.

Bandini, F., Lüthi, B., Peña-Haro, S., Borst, C., Liu, J., Karagkiolidou, S., Hu, X., Lemaire, G.G., Bjerg, P.L., Bauer-Gottwein, P. 2021. A Drone-Borne Method to Jointly Estimate Discharge and Manning’s Roughness of Natural Streams. Water Resources Research, 57(2). https://doi.org/10.1029/2020WR028266

Biggs, H., Smart, G., Holwerda, N., Doyle, M., McDonald, M., Ede, M. 2021. River discharge from surface velocity measurements - A field guide for selecting alpha. Envirolink Advice Report. Christchurch, New Zealand.

Carrillo-Serrano, V. 2021. Sediment transport in steep rivers with coarse material. PhD thesis, Universidad de Cuenca, Escuela Politécnica Nacional and Universidad Técnica Particular de Loja, Ecuador.

CELEC EP. CELEC SUR. Información Técnica. Retrieved Sep 2, 2024, from https://www.celec.gob.ec/celecsur/

Detert, M. 2021. How to Avoid and Correct Biased Riverine Surface Image Velocimetry. Water Resources Research, 57(2). https://doi.org/10.1029/2020WR027833

Díaz-Lozada, J.M. 2019. Avances en la Cuantificación hidrológica y caracterización hidráulica del flujo en el sistema fluvial del Río Carcarañá utilizando ADCP. Tesis Doctoral. Doctorado en Ciencias de la Ingeniería. Facultad de Ciencias Exactas, Físicas y Naturales. Universidad Nacional de Córdoba.

Fujita, I. 2018. Principles of surface velocity gaugings. The 4th IAHR-WMO-IAHS Training Course on Stream Gauging.

Fujita, I., Muste, M., Kruger, A. 1998. Large-scale particle image velocimetry for flow analysis in hydraulic engineering applications. Journal of Hydraulic Research, 36(3), 397-414. https://doi.org/10.1080/00221689809498626

ISO 748. 2007. Hydrometry — Measurement of liquid flow in open channels using currentmeters or floats. International Organization for Standardization, 2007-10-15, Fourth edition.

Jolley, M.J., Russell, A.J., Quinn, P.F., Perks, M.T. 2021. Considerations When Applying Large-Scale PIV and PTV for Determining River Flow Velocity. Frontiers in Water, 3, 709269. https://doi.org/10.3389/frwa.2021.709269

Koutalakis, P., Tzoraki, O., Zaimes, G. 2019. UAVs for Hydrologic Scopes: Application of a Low-Cost UAV to Estimate Surface Water Velocity by Using Three Different Image-Based Methods. Drones, 3(1), 14. https://doi.org/10.3390/drones3010014

Le Coz, J., Duby, P., Dramais, G., Camenen, B., Laronne, J., Zamler, D., Zolezzi, G. 2011. Use of emerging non-intrusive techniques for flood discharge measurements. 5th International Conference on Flood Management. Tokyo, Japan.

Le Coz, J., Hauet, A., Pierrefeu, G., Dramais, G., Camenen, B. 2010. Performance of image-based velocimetry (LSPIV) applied to flash-flood discharge measurements in Mediterranean rivers. Journal of Hydrology, 394(1-2), 42-52. https://doi.org/10.1016/j.jhydrol.2010.05.049

Lobo, A.P. 2019. Implementación de la técnica experimental de velocimetría por seguimiento de partículas (PTV) para cuantificar el recurso hídrico superficial en cursos fluviales de la provincia de Catamarca. Tesis de Maestría. Facultad de Ingeniería y Ciencias Hídricas. Universidad Nacional del Litoral.

Massó, L., Patalano, A., García, C.M., Ochoa-García, S.A., Rodríguez, A. 2024. Enhancing LSPIV accuracy in low-speed flows and heterogeneous seeding conditions using image gradient. Flow Measurement and Instrumentation, 100(2024), 1-16. https://doi.org/10.1016/j.flowmeasinst.2024.102706

Meselhe, E.A., Peeva, T., Muste, M. 2004. Large Scale Particle Image Velocimetry for Low Velocity and Shallow Water Flows. Journal of Hydraulic Engineering, 130(9), 937-940. https://doi.org/10.1061/(ASCE)0733-9429(2004)130:9(937)

Muste, M., Fujita, I., Hauet, A., 2008. Large-scale particle image velocimetry for measurements in riverine environments. Water Resources Research, 44(4), 0-19. https://doi.org/10.1029/2008WR006950

Papanicolaou, A.N., Diplas, P., Dancey, C. L., Balakrishnan, M. 2001. Surface Roughness Effects in Near-Bed Turbulence: Implications to Sediment Entrainment. Journal of Engineering Mechanics, 127(3), 211–218. https://doi.org/10.1061/(ASCE)0733-9399(2001)127:3(211)

Patalano, A., García, C.M., Rodríguez, A. 2017. Rectification of Image Velocity Results (RIVeR): A simple and user-friendly toolbox for large scale water surface Particle Image Velocimetry (PIV) and Particle Tracking Velocimetry (PTV). Computers and Geosciences, 109, 323-330. https://doi.org/10.1016/j.cageo.2017.07.009

Perks, M.T., Fortunato, S., Sasso, D., Hauet, A., Le Coz, J., Pearce, S., Peña-Haro, S., Tauro, F., Grimaldi, S., Hortobágyi, B., Jodeau, M., Maddock, I., Pénard, L., Manfreda, S. 2020. Towards harmonization of image velocimetry techniques for river surface velocity observations. Earth System Science Data, 12, 1545-1559. https://doi.org/10.5194/essd-12-1545-2020

Pizer, S.M., Amburn, E.P., Austin, J.D., Cromartie, R., Geselowitz, A., Greer, T., Ter Haar Romeny, B., Zimmerman, J.B., Zuiderveld, K. 1987. Adaptive histogram equalization and its variations. Computer Vision, Graphics, and Image Processing, 39, 355-368. https://doi.org/10.1016/S0734-189X(87)80186-X

Raffel, M., Willert-Christian, E., Wereley, S.T., Kompenhans, J. 2007. Particle Image Velocimetry, Experimental Fluid Mechanics. Springer Berlin Heidelberg, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-72308-0

Sharif, O. 2022. Measuring surface water flow velocities by a drone and large-scale particle image velocimetry (LSPIV). Master Thesis. University of Twente.

Singh, P., Ramasatri, K.S., Kumar, N., Bhatnagar, N.K. 2003. Suspended sediment transport from the Dokriani Glacier in the Garhwal Himalayas. Hydrology Research, 34(3), 221-244. https://doi.org/10.2166/nh.2003.0005

Stamhuis, E.J. 2006. Basics and principles of Particle Image Velocimetry (PIV) for mapping biogenic and biologically relevant flows. Aquatic Ecology, 40, 463-479. https://doi.org/10.1007/s10452-005-6567-z

Strelnikova, D., Paulus, G., Käfer, S., Anders, K.H., Mayr, P., Mader, H., Scherling, U., Schneeberger, R. 2020. Drone-Based Optical Measurements of Heterogeneous Surface Velocity Fields around Fish Passages at Hydropower Dams. Remote Sensing, 12(3), 384. https://doi.org/10.3390/rs12030384

Thielicke, W., Stamhuis, E.J., 2014. PIVlab – Towards User-friendly, Affordable and Accurate Digital Particle Image Velocimetry in MATLAB. Journal of Open Research Software, 2(1). https://doi.org/10.5334/jors.bl

Wilcox, A.C., Wohl, E.E., 2007. Field measurements of three-dimensional hydraulics in a step-pool channel. Geomorphology, 83(3–4), 215–231. https://doi.org/10.1016/j.geomorph.2006.02.017

To cite this article: Ochoa-García, S.A., Massó, L., Patalano, A., Matovelle-Bustos, C.M., Delgado-Garzón, P.V. 2025. Baseflow measurement in mountain rivers using LSPIV: A case study of the Tarqui and Yanuncay rivers in the Ecuadorian Andes. Revista de Teledetección, 66, e22733. https://doi.org/10.4995/raet.2025.22733

* Corresponding author: santiago.ochoa@ucacue.edu.ec