Volcanic influence on sediment dynamics and flow distribution at the Mezcalapa River bifurcation, southeast Mexico
Influencia volcánica en la dinámica sedimentaria y la distribución del flujo en la bifurcación del río Mezcalapa, sureste de México
Alejandro Mendoza1, Fabian Rivera-Trejo2,*, Gastón Priego-Hernández3, Moisés Berezowsky1
1 Instituto de Ingeniería, Universidad Nacional Autónoma de México. Circuito Escolar S/N, Ciudad Universitaria, Coyoacán, 04510 CDMX, México.
2 División Académica de Ingeniería y Arquitectura, Universidad Juárez Autónoma de Tabasco. Carretera Cunduacán-Jalpa km. 1, Col. La Esmeralda, 86690, Cunduacán, Tabasco, México.
3 División Académica de Ciencias Básicas, Universidad Juárez Autónoma de Tabasco. Carretera Cunduacán-Jalpa km. 1, Col. La Esmeralda, 86690, Cunduacán, Tabasco, México.
* Corresponding author: (F. Rivera-Trejo) This email address is being protected from spambots. You need JavaScript enabled to view it.
How to cite this article:
Mendoza, A., Rivera-Trejo, F., Priego-Henández, G., & Berezowsky, M. (2026). Volcanic influence on sediment dynamics and flow distribution at the Mezcalapa River bifurcation, southeast Mexico. Boletín de la Sociedad Geológica Mexicana, 78(1), A150126. https://doi.org/10.18268/BSGM2026v78n1A150126
Manuscript received: July 4, 2025. Corrected manuscript received: November 7,2025. Manuscript accepted: November 25, 2025.
ABSTRACT
The 1982 eruption of El Chichón volcano, one of the most destructive volcanic events of the 20th century in Latin America, ejected large volumes of pyroclastic material into the Mezcalapa River catchment, located in southeastern Mexico. Approximately 75 km downstream from the location of the volcano, the Mezcalapa River bifurcates into the Samaria and Carrizal rivers, which supply water to highly populated and agriculturally important areas in Tabasco state. This bifurcation is a key fluvial characteristic and variations in its f low distribution have historically caused significant social and economic impacts due to f looding and infrastructure stress. Around one year after the eruption, the bifurcation’s f low distribution started a new trend, where the dominant branch began gradually to capture less f low, reversing the trend observed in the preceding decades. This study investigates how and whether a volcanic eruption influences sediment dynamics and hydromorphological processes that could explain this shift. First, historical records and field measurements were analyzed to assess post-eruption changes in sediment load. Second, a morpho-hydraulic numerical model was applied to evaluate the effects of variations in sediment supply on the bed morphology and f low distribution of the bifurcation. Results show that suspended sediment loads increased by a factor of 3 to 10 after the eruption. Even 32 years later, the Platanar River still exhibited sediment concentrations twice as high as pre-eruption values. Modeling revealed that increased sediment supply may modulate differential erosion and sedimentation processes in the branches of the bifurcation, and it can modify the f low distribution. These findings underscore the long-term impact of volcanic eruptions on sediment transport and river morphology. They highlight the need to integrate volcanic sediment dynamics into fluvial hazard assessments and the importance of monitoring and modelling strategies to anticipate downstream consequences in sediment-sensitive elements of fluvial systems such as the Mezcalapa bifurcation.
Keywords: river bifurcation, flow distribution, volcanic impacts, sediment dynamics, geophysical hazards.
RESUMEN
La erupción del volcán El Chichón en 1982, uno de los eventos volcánicos más destructivos del siglo XX en América Latina, expulsó grandes volúmenes de material piroclástico en la cuenca del río Mezcalapa, en el sureste de México. Aproximadamente 75 km aguas abajo de la ubicación del volcán, el río Mezcalapa se bifurca en los ríos Samaria y Carrizal, los cuales abastecen de agua a zonas densamente pobladas y con actividad agrícola relevante en el estado de Tabasco. Esta bifurcación constituye un elemento fluvial estratégico, cuya variabilidad en la distribución del flujo ha ocasionado históricamente impactos sociales y económicos importantes, incluyendo inundaciones y daños en la infraestructura hidráulica. Alrededor de un año después de la erupción, la distribución del flujo en la bifurcación inició una nueva tendencia. El difluente dominante comenzó gradualmente a captar menos caudal, revirtiendo la tendencia observada en las décadas anteriores. Este estudio examina si la erupción volcánica influyó en la dinámica de los procesos hidromorfológicos que podrían explicar dicho cambio. En primer lugar, se analizaron registros históricos y mediciones de campo para evaluar las variaciones en el transporte de sedimentos tras la erupción. En segundo lugar, se aplicó un modelo numérico morfohidráulico para simular los efectos del aporte sedimentario sobre la evolución del lecho y la distribución del flujo en la bifurcación. Los resultados mostraron que las cargas de sedimento en suspensión aumentaron entre 3 y 10 veces tras la erupción. Incluso 32 años después, el río Platanar, afluente directamente influenciado por la erupción, presenta concentraciones de sedimento dos veces mayores que las condiciones previas. La modelación evidenció que el incremento en la disponibilidad de sedimentos puede modular procesos diferenciales de agradación y erosión en los difluentes de la bifurcación, y regular la distribución del flujo. Estos hallazgos destacan los impactos a largo plazo de las erupciones volcánicas sobre la morfología fluvial y el transporte de sedimentos. Destacan la necesidad de integrar la dinámica de los sedimentos volcánicos en las evaluaciones de los riesgos fluviales y la importancia de las estrategias de seguimiento y modelado para anticipar las consecuencias aguas abajo en los elementos de los sistemas fluviales sensibles a los sedimentos, como la bifurcación de Mezcalapa.
Palabras clave: impactos volcánicos, bifurcación fluvial, distribución del flujo, dinámica sedimentaria, riesgos geofísicos.
1. Introduction
Explosive volcanic eruptions are major geological events that can impact regional sediment dynamics by rapidly modifying the landscape and significantly increasing sediment supply to river systems. In volcanically active regions such as southeast Mexico, these eruptive events can disrupt the sedimentological equilibrium, alter fluvial morphology, and generate long-term hydrological and geomorphological changes. Understanding how volcanic processes trigger fluvial responses is crucial for managing volcanic hazards related to the displacement of sediment generated by eruptions, particularly in Latin America, where numerous active volcanoes are located near densely populated river valleys.
Alluvial rivers tend to achieve a state of dynamic equilibrium to transport the supplied flow discharge and sediment load by adjusting their channel slope, cross-section, planform morphology, and channel roughness (Huang and Nanson, 2007; Nanson and Huang, 2017; Su et al., 2021). The dynamic equilibrium can be disturbed by anthropogenic causes (e.g., Zhou et al., 2021), natural factors (e.g., Pierson and Major, 2014), or a combination of both. For instance, hydraulic structures that interact with rivers may disrupt the balance between sediment supply and sediment load capacity, altering the river morphology (Reisenbüchler et al., 2019). An example is the dams; they disrupt the sediment supply downstream of reservoirs (e.g., Juwono et al., 2019; Smolar-Zvanut and Mikos, 2014). Another example is climate change, which can alter rainfall patterns and consequently modify erosion rates and sediment loads delivered to rivers (Tian et al., 2020).
In the case of volcanic eruptions, they are one of the natural causes of increased sediment supply to rivers (Hayes et al., 2002; Sclafani et al., 2018). An explosive volcanic eruption disrupts the equilibrium of fluvial systems by supplying large amounts of excess sediment into the fluvial system (Pierson and Major, 2014). The sediment load attained maximum levels in the years immediately following the eruption (Gob et al., 2016; Tanarro et al., 2010). Furthermore, there could be an increase in the magnitude and frequency of floods, and the excess sediment input could change the morphology of streams (Park and Schmincke, 2020). Some direct effects of volcanic eruptions are (1) damage to the vegetation in the catchment ( Jennerjahn et al., 2013), limiting the interception and evapotranspiration capacity; (2) reduction of infiltration capacity by up to two orders of magnitude compared to the pre-eruption conditions, caused by deposition of pyroclastic material in the catchment (Pierson and Major, 2014); (3) alteration of the hydraulic regime of the streams, caused by the reduction of the roughness triggered by the deposited sand that buries gravel beds (Dorava and Meyer, 1994); (4) change of the bed elevation of the channels near the eruption emission point/area, increasing the slope and bed shear stress. An example was the eruption of Mount Hood in the United States. It caused a sediment wave that aggraded the Sandy River, increasing the riverbed elevation in a canyon by 23 m in the lower part of the watershed (Pierson et al., 2011); and (5) intense rainfall events on volcanic slopes can mobilize stored pyroclastic deposits, triggering lahars or low-concentration sediment-laden flows with considerable geomorphic and social impact, as reported for the Pico de Orizaba volcano (Rodríguez et al., 2006). The temporal scales of the consequences can also be of different magnitudes, with long-lasting effects, such as the case of Mount St. Helens, where, after the eruption in 1980, it continued to supply sediment load exceeding almost four decades later, causing aggradation in the Turtle-Cowlitz fluvial system (Sclafani et al., 2018). Furthermore, there were mid-term effects such as a transient change in the hydrological response and an increase in peak flows, which lasted only about five years after the eruption (Major and Mark, 2006).
River bifurcations are a transitory characteristic of the morphology of fluvial systems (Kleinhans et al., 2012), they are searching for the dynamic equilibrium; however, they may be unstable concerning the flow and sediment distribution along the diffluent channels when the bed is composed of sand or gravel (Bolla-Pittaluga et al., 2015). Multiple factors can disrupt the stability of the flow distribution in bifurcations, including differences in energy gradient of the diffluentchannels, width-to-depthratios, upstream bends, sediment sorting (Kleinhans et al., 2008), or upstream and downstream boundary conditions (Salter et al., 2019). However, when multiple factors interplay and balance each other to maintain an incipient dynamic equilibrium of flow distribution, river bifurcations can last from a few years to centuries before a complete avulsion occurs (Kleinhans et al., 2011). There are documented cases where flow distribution changes occurred relatively quickly, generating problems for communities located downstream of the bifurcation (Rivera-Trejo et al., 2010). The mechanics of these changes are a complex interaction between flow and sediment load distributions driven by the bed morphology of the main and diffluent channels (Hackney et al., 2018; He et al., 2019). The sediment supply modified by anthropogenic or natural causes may trigger a change in the flow distribution of river bifurcations. An example of anthropogenic alterations that have affected a river bifurcation is the system of dams in the Mezcalapa River, southeast Mexico, as characterized by Mendoza et al. (2022). They argued that sediment retention by upstream reservoirs significantly altered the flow distribution at the Mezcalapa River bifurcation by inducing channel incision and reinforcing feedback mechanisms that enhance flow capture in the diffluent with a higher energy gradient. Additionally, flow homogenization (suppression of peak discharges by the dams) reduced the sediment transport capacity, inducing aggradation in the lower-energy branch of the bifurcation and further limiting its hydraulic conveyance capacity. Flow records show that the flow distribution has been persistently changing since the late 1960s, after the commissioning of a system of four dams upstream of the bifurcation, the first of which was completed in 1967. Between 1969 and 1983, one of the diffluent channels captured gradually less flow, and unexpectedly, after 1983 this same diffluent changed its trend and gradually captured more flow (Mendoza et al., 2019). Mendoza et al. (2022) hypothesized that the dam system upstream of the bifurcation triggered the flow distribution trend observed between 1969 and 1983; however, the causes of the increased capacity of the Carrizal River observed after 1983 were unclear. This is especially relevant because the alteration of the flow distribution has caused flooding, resulting in economic and social consequences for the communities downstream of the bifurcation (Rivera-Trejo et al., 2010). As mentioned earlier, multiple factors control the flow distribution in river bifurcations, making it difficult to identify all the natural and anthropogenic events that may have impacted the fluvial system. An event in the catchment of the Mezcalapa River that could have influenced the dynamics of the fluvial system was El Chichón volcano eruption between March 28 and April 4, 1982 (Scolamacchia and Macías, 2005). Inbar et al. (2001) classified El Chichón eruption as one of the largest during the twentieth century and the worst volcanic disaster in Mexico (Tilling, 2009); they also reported that substantial erosion processes occurred in the basin in the first year after the eruption, decreasing with time. Other studies (e.g., Iroumé et al., 2020; Ulloa et al., 2015; Jennerjahn et al., 2013; Inbar et al., 1994) found similar patterns to those reported by Inbar et al. (2001). They indicated that immediately after a volcanic eruption, erosion rates in the watershed were several orders of magnitude larger compared to pre-eruption conditions; gradually, the erosion rate decreased exponentially with time.
Our study examines whether—and by what mechanisms—the 1982 El Chichón eruption influenced sediment dynamics and hydromorphological processes that could explain the observed shift in flow distribution at the Mezcalapa-Samaria-Carrizal bifurcation. To address these interrogations, two questions were formulated to explore the possible relationships between the 1982 eruption of the El Chichón volcano and the morphological changes of the Mezcalapa River bifurcation. First, what was the magnitude of change in suspended sediment load within the Mezcalapa River system after the El Chichón eruption? Second, what morphological processes caused by sediment supply variations at the bifurcation can be related to the observed flow distribution change? To answer these questions, we analyzed historical records and field measurements to assess post-eruption changes in suspended-sediment load. Later, we applied a morpho-hydraulic numerical model to evaluate how variations in sediment supply affect bed morphology and the partitioning of flow at the bifurcation.
2. Study area
2.1. EL CHICHÓN VOLCANO
El Chichón volcano is located in the northern part of the state of Chiapas, Mexico. It lies approximately 75 kilometers southwest of Villahermosa, Tabasco, and 70 kilometers northeast of Tuxtla Gutiérrez, capital city of the state of Chiapas. It is part of the Chiapanecan volcanic arc (Figure 1), which is associated with the abduction of the Cocos Plate under the North American Plate (Alatorre-Ibargüengoita et al., 2021). The 1982 eruption of El Chichón has been classified as one of the most destructive volcanic events in Mexico during the 20th century due to its unexpected explosivity and widespread impact (Macías, 2005). The fluvial system connected directly to the surroundings of the volcano consists of: the Magdalena River, which drains towards the Grijalva River upstream of Peñitas Dam; the Platanar River, with headwaters at the flanks of El Chichón cone draining towards the Mezcalapa River upstream of the bifurcation; and the Comuapa River, which also drains to the Mezcalapa River upstream of the bifurcation (Figure 2a). In the case of the Platanar River, it converges into the Mezcalapa River 31 km downstream of the Peñitas Dam; the Comuapa River, located to the north of El Chichón, converges into the Mezcalapa River 23 km upstream of the bifurcation (Figure 2a). The discharges of the catchments of both rivers are small compared to the annual average flow at the upper Grijalva River (Table 1).
Before 1982, the volcano flanks were covered by vegetation, and according to a 1970 census, the surrounding area consisted of 80% pasture, 1.5% crops, and 17% forest (Inbar et al., 2001). The most recent data indicate that grasslands and agriculture are the prominent land uses in the Platanar and Comuapa catchments (Figure 2b). The predominant soil type in the Platanar catchment is Cambisol, while Acrisol is the most common in the Comuapa catchment. In the lower part of both basins, the common type is Gleysol (Figure 2c).
The El Chichón eruption, which occurred between March 28 and April 4, 1982, consisted of three explosive events; the global event had a Volcanic Eruption Index (VEI) of 5 (Alatorre-Ibargüengoita et al., 2021). The eruption buried nine villages under pyroclastic fallouts and density currents (Scolamacchia and Schouwenaars, 2009), and the estimated volume of ejected material was 0.5 km3 (Sigurdsson et al., 1984). The volcanic ash covered an area of 30 000 km2 (Medina-Martinez, 1986). The consecutive eruptive phases produced different layers of deposit units, conformed by pyroclastic falls and flows (Sigurdsson et al., 1984; Scolamacchia and Schouwenaars, 2009). The topography played a crucial role in the distribution of pyroclastic density currents during the eruption, and the fallout deposits were influenced by the predominant wind direction during the event (Pitari and Visconti, 1984; Scolamacchia and Macías, 2005). Ash aggregates were evaluated by Scolamacchia and Macías (2005) in different pyroclastic units at distances between 0.3 and 9.3 km from the crater; the aggregates had diameters between 0.2 and 11 mm. After the eruption, El Chichón cone experimented intense erosion due to its hilly topography, steep slopes, and large volumes of unconsolidated eruption material (Thouret, 1999). The major erosion occurred mainly during the first year after the eruption; the estimated sediment yield drained by the rivers near the cone (Magdalena, Platanar, and Moba) from the surrounding area was 105 ton km–2, or the equivalent of about 0.01 km3 (Inbar et al., 2001).
2.2. THE MEZCALAPA RIVER SYSTEM
The Grijalva River has its headwaters in Guatemala and drains towards the Gulf of Mexico, crossing the states of Chiapas and Tabasco. Along the river are four reservoirs (Figure 3a); the last is Peñitas Dam. Downstream of the Peñitas Dam, the Grijalva River is renamed as Mezcalapa River, and 75 km downstream, it bifurcates into the Samaria and Carrizal rivers (Figure 3b). The Samaria River follows a path of 97 km to the Gulf of Mexico, while the Carrizal River has a course of 167 km to the Gulf of Mexico. The slope of the rivers decreases progressively downstream (Figure 3c, note the logarithmic scale); this is important for the Mezcalapa River and the diffluent rivers, since a consequence is that their sediment load capacity also decreases progressively. An implication of the different river lengths is that the Samaria River has a hydraulic gradient advantage. In addition, a bend in the final reach of the Mezcalapa River (Figure 3b) guides the flow towards the Carrizal River, which helps counterbalance its lower hydraulic gradient and equilibrate the flow distribution.
Before commissioning the Grijalva River dams, the average suspended sediment yield was 57.05 x 106 tons year-1 (Mendoza et al., 2019). According to the analysis performed by Mendoza et al. (2022), the combined effect of flow peak suppression and the shortage of sediment supply caused by the reservoirs triggered a process at the bifurcation, where the Samaria River gradually captured more flow. This pattern reversed by 1983, when the Carrizal River gradually began to capture more flow over the next two decades, surpassing 50% by 2002. This condition forced local authorities to limit the flow entering the Carrizal River, initially using spur dikes and later implementing a control structure with a system of gates.
3. Methods
An evaluation of historical records from gauging stations was conducted to assess the impact of El Chichón eruption on the sediment load in the fluvial system of the Mezcalapa River. Given the significant geophysical characteristics of the eruption, such as its explosive nature and the substantial volume of pyroclastic material ejected (~0.5 km³), understanding these parameters provided a critical context to interpret the magnitude and persistence of sedimentological impacts observed downstream. The analysis was complemented with data from a sediment load measurement field campaign conducted in 2014. Since the sediment load measurements at the gauging stations were interrupted in the late 1980s, satellite images obtained from Google Earth were utilized to characterize the displacement of sediment deposits along the rivers downstream of the volcano. Incorporating this geological context into the sediment dynamics analysis enhanced the interpretation of morphological responses in the fluvial system. To address the question of how the variation in sediment load can drive morphological bed and hydraulic processes, the alteration of sediment load at the Mezcalapa River could alter the flow distribution. A morpho-hydraulic model of a representation of the river bifurcation was developed.
3.1. HISTORICAL RECORDS
The gauging stations with suspended sediment load records situated upstream and downstream of the Mezcalapa River bifurcation are depicted in Figure 2a. Table 2 details the location of the stations and the period during which information on suspended sediment load records was available. Station 30094 measured the suspended sediment load drained towards the Mezcalapa River and the bifurcation coming from the headwaters of the Platanar River, where the volcano’s cone is located. Since there are no stations near the bifurcation, the ones located downstream (30005 and 30062, see Figure 2a) were utilized to characterize the suspended sediment load before and after the eruption. Station 30015 is also of interest since it is located on the Mezcalapa River upstream of the bifurcation; however, the occurrence of the eruption coincided with the construction process of the Peñitas Dam, a condition that affected the suspended sediment load measurements after the eruption.
3.2. FIELD MEASUREMENTS
In November 2014, a sediment load measurement campaign was carried out in the main rivers of the Mezcalapa catchment downstream of the Peñitas Dam (Rivera-Trejo and Rosas-Figueroa, 2014). The field campaign considered the Platanar and Comuapa rivers. Figure 4a shows a large sand bar on the right bank of the Platanar River exposed during low flow conditions; on the other hand, Figure 4b shows the Comuapa River, and the coloration of the flow indicates it transports a large amount of wash load. During the campaign, suspended and bed loads were measured. Sediment samples from the bed were taken using a hand dredge to characterize sediment size (Figure 4c). The suspended sediments were taken using a DH-78 (Figure 4b) or DH-59 (Figure 4d) suspended sediment sampler. The bedload transport was estimated using a Helley-Smith (H-S) bed sampler with an extendable handle (Figure 4e) and the H-S 8020 sampler with an 8 kg weight (Figure 4f). Rivera et al. (2008) described the methodology followed to process the measurements.
3.3. SATELLITE IMAGERY ANALYSIS
A qualitative estimation of the displacement of the sediment deposits in the rivers that drained from the cone of the volcano was performed with satellite imagery. The images utilized were obtained from Google Earth® in the period from 1984 to 2003; the image resolution is variable, with a medium resolution of 15m. The sediment displacements were calculated by analyzing the changes in the position of the colored features. This analysis helped to identify the displacement of the sediment deposits along the Platanar and Magdalena rivers. However, the Magdalena River flows towards the Peñitas Reservoir, and the sediment is trapped by the dam, which cannot be the cause of the high sediment rate in the Mezcalapa River. Therefore, we employed a morpho-hydraulic model to analyze the sedimentation process.
3.4. MORPHO-HYDRAULIC MODEL
A morpho-hydraulic model was utilized to identify potential interactions between the sediment load increment caused by the eruption and the flow distribution changes observed at the Mezcalapa River bifurcation. The driver of the flow distribution change is the bed evolution processes within the bifurcation. Two scenarios were modeled, related to pre- and post-eruption conditions. For the pre-eruption conditions, zero sediment load was considered at the inlet of the main channel (most of the sediment load is retained in the upstream dams); this is an approximation since downstream of the Peñitas dam, the Mezcalapa River has sediment inputs from the Platanar and Comuapa rivers. For the post-eruption conditions, the equilibrium sediment load at the entrance of the main channel was considered (sediment load provided by the material of the eruption); this is also an approximation since the sediment load from the Platanar River is lower than the equilibrium capacity of the Mezcalapa River.
Regarding flow conditions, the annual average hydrograph of the Mezcalapa River was considered. In both scenarios, the simulation was conducted over a five-year period, and the flow distribution at the bifurcation was characterized. The conceptual geometry of the bifurcation utilized for the morpho-hydraulic model is shown in Figure 5. It considers the main and diffluent channels, with the key characteristics of the Mezcalapa River bifurcation: the larger energy gradient of the left branch (Samaria River) and an upstream bend just before the bifurcation directing the flow towards the right branch (Carrizal River), conditions described early in the study area section. The hydraulic boundary conditions considered at the inlet were the average annual hydrograph developed after the commissioning of the first dam on the system (in 1967) and a constant sea level at the outlet of the diffluents for both scenarios modeled. Regarding sediment load conditions, pre-eruption conditions were considered with a zero input sediment load, and post-eruption conditions were modeled with an equilibrium sediment load (Figure 5). The model is an adaptation of the morpho-hydraulic model presented by Mendoza et al. (2022), as it considers the final bed configuration of the channels as the initial bed for this study.
4. Results
4.1. HISTORICAL RECORDS
Figures 6a to 6d show the monthly suspended sediment load records from the gauging stations used in the analysis. Although the records were discontinuous for extended periods, the stations located downstream of the volcano revealed an increase after the eruption, except for the one located immediately downstream of the Peñitas Dam (Figure 6a). As mentioned earlier, the increase in sediment load could be attributable to the large volume of pyroclastic material expelled (~0.5 km³) and the extensive spatial distribution of volcanic ash deposits covering approximately 30 000 km². The station located in the Mezcalapa River (30015) had only seven months of recorded data after the eruption, and no increase in sediment load was noted after the eruption (Figure 6b). This was driven by the construction of the Peñitas dam which limited the free flow. The stations on the Platanar (30094), Samaria (30005), and Carrizal (30062) rivers showed an increase in suspended load observed after the eruption (Figures 6b-6d). The monthly data of flow volume and suspended sediment volumetric concentration from the Platanar River (30094) before the eruption are shown in Figure 6e; there, the average annual volumetric concentration is 99.6 x 10–6 m3 m–3. Immediately after the eruption, the concentration raised to 475.6 x 10–6 m3 m–3 (the average of the discontinuous 12 months recorded after the eruption, see data points in Figure 6b expressed in tonnes); the elevated sediment concentrations in the post-eruption period can be attributed to the response of the fluvial system to the increase in sediment supply caused by the pyroclastic material provided by the eruption. Regarding the flow distribution at the bifurcation between 1972 and 2004, as shown in Figure 6f, a tipping point was identified in 1983, marking the beginning of an increasing trend in the Carrizal River’s ability to capture more flow. Post-eruption sediment records span only a few years because suspended-sediment monitoring is no longer performed systematically in Mexico since four decades ago.
4.2. FIELD MEASUREMENTS
The suspended load, bedload, and flow discharge were measured in the Platanar and Comuapa rivers during the 2014 field campaign (Table 3). Over the measurement dates, the Platanar River exhibited significant flow discharge variation; whereas the Comuapa River exhibited a more uniform flow discharge. Despite these hydrological differences, the volumetric concentration of the suspended load remained consistent within each river throughout the two days of measurements, with 330 and 260 x 10–6 m3 m–3 in the Platanar River and 530 and 470 x 10-6 m3 m–3 in Comuapa River (Table 3). The field measurements show that the Platanar River, at the time of the 2014 field campaign, had a larger concentration than before the eruption (compared to the annual average concentration of 99.6 x 10–6 m3 m–3 or the average concentration for November of around 135 x10-6 m3 m–3). Figure 6e shows that three decades after the eruption, the sediment concentration is larger, which may be caused by the long-lasting impacts of El Chichón eruption or other factors, such as land use changes in the region.
4.3. SATELLITE IMAGERY ANALYSIS
The displacement of sediment deposits along the rivers draining the vicinity of El Chichón after the eruption was assessed using satellite images (Figures 7a to 7d). The displacement of the deposits along the Magdalena River was identified here with the position of its front (white dots) by 1988 (Figure 7b), 1993 (Figure 7c), and 2003 (Figure 7d); it is important to note that in these figures, it is visible that the sediment deposits filled progressively part of the reservoir of the Peñitas Dam. Additionally, in the Platanar River, the region of deposited sediment along the stream is noticeable, as indicated in 1988, 1993, and 2003 by a white dotted rectangle (Figures 7b to 7d). After 2003 (Figure 7d), the deposited material was either washed towards downstream reaches or covered by vegetation, and it was no longer visible in the images.
4.4. MORPHO-HYDRAULIC MODEL
The results of the scenarios for pre- and post-eruption conditions (without and with increased sediments, respectively) were assessed in terms of i) the bed evolution, expressed as the non-dimensional change in bed elevation ∆z/H0 (negative = erosion; positive = deposition; H0 is the initial mean f low depth at each section), and (ii) the f low partition between the diffluent channels, with three control sections at the bifurcation (Figure 8a): Section 1 (Mezcalapa River), Section 2 (Samaria River), and Section 3 (Carrizal River). The model runs five years of each scenario after a brief warm-up. Where t = 0 is therefore not a calendar date. Pre- and post-eruption bed evolution are plotted on the same normalized time axis for comparability across scenarios, without implying historical synchrony. Figure 8b shows the temporal bed evolution of the three rivers; note that after the second year, the bed of the Mezcalapa (1) and Samaria (2) rivers tends towards a dynamic equilibrium with the post-eruption condition, with ∆z/H0 close to zero. In contrast, under the pre-eruption condition, these rivers continuously experienced erosion (e.g., ≈ -0.1 to -0.25 in the Mezcalapa River and ≈ -0.1 to -0.2 in the Samaria River). In contrast at Carrizal River both scenarios exhibit aggradation ( ≈ + 0.2 to +0.4). Finally, the temporal evolution of the f low distribution is shown in Figure 8c. Note that, under the post-eruption conditions, the distribution became quasi-steady after the second year on the order of ~6–7% of the total discharge. Conversely, in the pre-eruption conditions, the Carrizal River tended to capture a gradually decreasing fraction of the f low. This trend is consistent with the persistent incision observed at sections 1–2 under pre-eruption conditions, which increases the conveyance of the main branch and reduces the relative share diverted to the Carrizal.
5. Discussion
5.1. SEDIMENT LOAD ALTERATION
The satellite imagery analysis revealed the displacement of sediment deposits on the two rivers that descend from the cone (Platanar and Magdalena rivers) towards the Mezcalapa River following the eruption. The Magdalena River transported part of the deposits generated by the eruption, and later, they were trapped by the Peñitas Dam (Figures 7b-7d). On the other hand, the sediment deposits on the Platanar River are more clearly visible after the eruption and up to the 1990s (Figures 7b-7d), with the displaced material moving towards the confluence with the Mezcalapa River. There are no records of bedload at the gauging stations before or in the years following the eruption; however, the displacement of sediment deposits after the eruption along the Platanar River allows us to infer an increase in bedload. The sediment deposits on the riverbed increased temporally the hydraulic gradient, allowing the transference of sediments towards downstream reaches, potentially up to the bifurcation.
Regarding suspended sediment load, it increased after the eruption in three of the four gauging stations located downstream of El Chichón volcano. The station that did not experience an increase in suspended load (Figure 6a) is located downstream of the Peñitas reservoir; this is attributed to the retention of sediments in the Peñitas Dam, which was under construction at the time of the eruption. In contrast, the suspended sediment load in the gauging stations in the Platanar, Samaria, and Carrizal rivers shows higher values approximately two years after the El Chichón eruption (Figure 6b to 6d).
Due to the interruption of suspended sediment measurements at the stations after 1986, the field measurements carried out in 2014 were used here to complement the available gauging data following the eruption. The annual average volumetric concentration, computed using data from the Platanar gauging station (30094) prior to the eruption, was 99.6 × 10–6 m3 m–3 (Figure 6e). Immediately after the eruption, the concentration increased to 475.6 × 10–6 m3 m–3; 20 years later, the average concentration measured in November 2014 in the Platanar River was 295 × 10–6 m3 m–3; it is still 2.1 times larger compared to pre-eruption conditions (average of 136 × 10–6 m3 m–3 for November; Figure 6e).
A statistical analysis was performed here to assess the significance of the changes observed in the records of suspended sediment load. Figure 9a shows the annual averages of suspended load before and after the eruption. In the case of the Platanar River, the suspended load increased by a factor of five. On the other hand, the Samaria River experienced a 2.6-fold increase, while the Carrizal River developed an 11-fold increase. Note that the data after the eruption is limited, and all measurements were interrupted by the end of 1986. A test of hypothesis and significance was performed to validate the increment of suspended load after the eruption. A test of two populations with different standard deviations was utilized (Weiss, 2016), considering the null hypothesis H₀, that the monthly average suspended load before and after the eruption belongs to the same population; with the alternative hypothesis H1, that the monthly average suspended load after the eruption is larger. The T-critical values were calculated for a one-tailed test with a level of significance of p = 0.1; Figure 9b summarizes the statistical comparison of pre- and post- eruption of suspended-sediment loads. For the Samaria (30005) and Carrizal (30062) stations, the null hypothesis of no change is rejected, indicating a statistically significant post-eruption increase. In contrast, for the Platanar (30094) and Mezcalapa (30015) stations, the null hypothesis cannot be rejected, which we attribute to the limited post-eruption records, only 12 discontinuous months at Platanar and 7 months at Mezcalapa, reducing statistical power. Notably, despite this lack of significance, the post-eruption mean load at Platanar is approximately five times the pre-eruption mean.
Besides the eruption, other factors may have contributed to the increased sediment input into the fluvial system. Even though human settlements have existed for several centuries, the Grijalva River basin began to degrade due to deforestation in the mid-19th century. This process has been accelerated since 1946 with the use of mechanical tools (Muñoz-Salinas et al., 2016). Ochoa-Gaona and Gonzalez-Espinosa (2000) estimated that between 1974 and 1990, the annual deforestation rate within the Grijalva basin ranged from 1.1% to 3.4%. From 1990 to 1996, it reached values of 5% annually in the highlands of Chiapas, located upstream of the Mezcalapa catchment. Land use change impacts erosion rates and sediment yield in the watershed. However, the analysis presented here indicates an abrupt increase in sediment load in the Mezcalapa River downstream of the Platanar River following the eruption of El Chichón (Figure 9a), which cannot be directly attributed to deforestation and the resulting erosion in the watershed given the annual average rate of deforestation of 2.25% observed during that period.
5.2. EFFECT ON THE HYDROMORPHOLOGY OF THE BIFURCATION
In order to identify the potential hydromorphological processes driven by the change of sediment load produced by the excess of pyroclastic material drained by the fluvial system of the Mezcalapa River and the possible relation with the change of the trend of the flow distribution at the bifurcation observed by 1983, a morphohydraulic model was used (Figure 6f). According to the results of the modeling, the effect of the increment of sediment load promoted the stabilization of bifurcation that otherwise tends to the gradual erosion of the Samaria River by: (1) the difference of hydraulic gradient between the two diffluents (note the larger gradient in the Samaria River) and (2) the retention of sediments upstream of the bifurcation in the dams. The model of the bifurcation used here is simplified, and multiple factors controlling the flow distribution are not considered. However, according to the model’s results, the increment of sediment load produced by the eruption limits the incision process developed in the main (Mezcalapa) and left (Samaria) channels. The results of the modeling showed that the flow distribution for post-eruption conditions attained stability; this is different from what happened at the beginning of 1983 in the bifurcation, where the flow of the Samaria River diminished gradually (Figure 6f). However it highlights the sensitivity of the flow distribution in the bifurcation to variations of sediment load supplied from upstream reaches.
6. Conclusions
The flow distribution into the branches of the Mezcalapa River bifurcation has changed over the decades; by 1983, a new trend emerged where the dominant branch gradually began to capture less flow. On the other hand, the El Chichón volcano erupted in 1982, depositing a significant volume of pyroclastic material upstream of the Mezcalapa River catchment. Data from gauging stations with suspended sediment load measurements showed an increase in sediment load in the fluvial system ranging from 2.6 to 11 fold higher than pre-eruption conditions in the years following the eruption. On the other hand, the data obtained from fieldwork conducted in 2014 (32 years after the eruption) showed that the sediment load concentration in the Platanar River, which descends directly from the cone of El Chichón, was 2.1 times larger compared to the concentration before the eruption.
The morpho-hydraulic model utilized here to assess the effect of the increment of sediment load in the fluvial system showed that the variations of sediment load were able to modulate different processes of erosion and sedimentation in the main and diffluent rivers of the bifurcation. As a result, the differential evolution of the bed among the channels can modulate the flow distribution. Before the eruption, the bifurcation had a shortage of sediment supply due to the commissioning of the upstream dams, the first of which was commissioned in 1967. The shortage of sediment supply induced erosion in the Samaria River, which gradually increased its hydraulic capacity. The simulations with the morpho-hydraulic model showed that an increase in sediment supply can lead to a process where erosion and sedimentation balance each other to attain a dynamic equilibrium in the bed of the Samaria River and, consequently, to stabilize the flow distribution. The results highlight the sensitivity of the flow distribution in river bifurcations due to variations in the sediment load supply caused by natural or anthropogenic factors.
From a geological perspective, this study underscores the critical relationship between volcanic processes and fluvial sediment transport dynamics. Understanding the patterns and temporal evolution of sediment transport following explosive volcanic eruptions, such as that of the El Chichón, is fundamental for assessing volcanic hazards and managing associated risks. Enhanced knowledge of post-eruption sediment dynamics can significantly contribute to more accurate predictions of sedimentary processes, guiding the design of effective mitigation and management strategies in volcanically active regions throughout Latin America.
Supplementary data
Some or all data, models, or code that support the findings of this study are available from the corresponding author upon reasonable request.
Contributions of authors
Conceptualization: AMR; (2) Data analysis or acquisition: FRT, GPH; (3) Methodological/technical development: AMR, FRT; (4) Original manuscript writing: AMR; (5) Revised and edited manuscript writing: FRT, MB; (6) Graphic design: GPH; (7) Fieldwork: FRT, GPH; (8) Interpretation: AMR, MB.
Financing
This research was supported by the UNAM-PAPIIT Grant IA102623.
Conflict of interest
The authors declare no conflicts of interest.
Handling editor
Dmitri Rouwet.
References
Alatorre-Ibargüengoita, M., Hernández-Urbina, K., & Ramos-Hernández, S. (2021). Long-and short-term volcanic hazard assessment of El Chichon Volcano (Mexico) through Bayesian inference. Natural Hazards, 106, 1011–1035. https://doi.org/10.1007/s11069-021-04506-1
Bolla-Pittaluga, M., Coco, G., & Kleinhans, M. (2015). A unified framework for stability of channel bifurcations in gravel and sand fluvial systems. Geophysical Research Letters, 42(18), 7521–7536. https://doi.org/10.1002/2015gl065175
Dorava, J., & Meyer, D. (1994). Hydrologic hazards in the lower Drift River basin associated with the 1989-1990 eruptions on Redoubt Volcano, Alaska. Journal of Volcanology and Geothermal Research, 62, 387–407. https://doi.org/10.1016/0377-0273(94)90044-2
Gob, F., Gautier, E., Virmoux, C., Grancher, D., Tamisier, V., Primanda, K. W., Wibowo, S. B., Sarrazin, C., de Belizal, E., Ville, A., & Lavigne, F. (2016). River responses to the 2010 major eruption of the Merapi volcano, central Java, Indonesia. Geomorphology, 273, 244–257. https://doi.org/10.1016/j.geomorph.2016.08.025
Hackney, C., Darby, S., Parsons, D., Leyland, J., Aalto, R., Nicholas, A., & Best, J. (2018). The influence of flow discharge variations on the morphodynamics of a diffluence-confluence unit on a large river. Earth Surface Processes and Landforms, 43, 349–362. https://doi.org/10.1002/esp.4204
Hayes, S., Montgomery, D., & Newhall, C. (2002). Fluvial sediment transport and deposition following the 1991 eruption of Mount Pinatubo. Geomorphology, 45, 211–224. https://doi.org/10.1016/s0169-555x(01)00155-6
He, G., Fang, H., Wang, J., & Zhang, T. (2019). From fluvial dynamics to eco-fluvial dynamics. International Journal of Sediment Research, 34, 531–536. https://doi.org/10.1016/j.ijsrc.2019.05.002
Huang, H., & Nanson, G. (2007). Why some alluvial rivers develop an anabranching pattern. Water Resources Research, 43(7). https://doi.org/10.1029/2006wr005223
Inbar, M., Hubp, J., & Villers Ruiz, L. (1994). The geomorphological evolution of the Paricutin cone and lava flows, Mexico, 1943-1990. Geomorphology, 9(1), 57–76. https://doi.org/10.1016/0169-555x(94)90031-0
Inbar, M., Reyes Enriquez, A., & Graniel Graniel, J. (2001). Morphological changes and erosion processes following the 1982 eruption of El Chichon volcano, Chiapas, Mexico. Géomorphologie: relief, processus, environment, 7(3), 175–184. https://doi.org/10.3406/morfo.2001.1100
Iroumé, A., Paredes, A., Garbarino, M., & Morresi, D. (2020). Post-eruption morphological evolution and vegetation dynamics of the Blanco River, southern Chile. Journal of South American Earth Sciences, 104, 1–14. https://doi.org/10.1016/j.jsames.2020.102809
Jennerjahn, T., Jänen, I., Propp, C., Adi, S., & Nugroho, S. (2013). Environmental impact of mud volcano inputs on the antropogenically altered Poron River and Madura Strait coastal waters, Java, Indonesia. Estuarine, Coastal and Shelf Science, 130, 152–160. https://doi.org/10.1016/j.ecss.2013.04.007
Juwono, P., Hidayat, F., Ruritan, R., Rianto, A., & Taufiqurrachman, M. (2019). The study of optimization of sediment flushing efficiency from cascade reservoirs as mitigation to the secondary impact of volcanic hazard. In J. Tournier, T. Bennet, & J. Bibeau, Sustainable and safe dams around the world (pp. 152–160). https://doi.org/10.1201/9780429319778-104
Kleinhans, M., Cohen, K., Hoekstra, J., & IJmker, J. (2011). Evolution of a bifurcation in a meandering river with adjustable channel widths. Earth Surface Processes and Landforms, 36(15), 2011–2027. https://doi.org/10.1002/esp.2222
Kleinhans, M., Ferguson, R., Lane, S., & Hardy, R. (2012). Splitting rivers and their seams: bifurcations and avulsion. Earth Surface Processes and Landoforms, 38(1), 47–61. https://doi.org/10.1002/esp.3268
Kleinhans, M., Jagers, H., Mossselman, E., & Slof, C. (2008). Bifurcation dynamics and avulsion duration in meandering rivers by one-dimensional and three-dimensional models. Water Resources Research, 1-31. https://doi.org/10.1029/2007wr005912
Macías, J. L. (2005). Geología e historia eruptiva de algunos de los grandes volcanes activos de México. Boletín de la Sociedad Geológica Mexicana, 57(3), 379–424. https://doi.org/10.18268/bsgm2005v57n3a6
Major, J., & Mark, L. (2006). Peak flow responses to landscape disturbances caused by the cataclysmic 1980 eruption of Mount St. Helens, Washington. GSA Bulletin, 118(7-8), 938–958. https://doi.org/10.1130/b25914.1
Medina-Martinez, F. (1986). Análisis de las columnas eruptivas del volcán Chichón, Marzo-Abril 1982. Geofísica Internacional, 25, 233–249. https://doi.org/10.22201/igeof.00167169p.1986.25.1.806
Mendoza, A., Berezowsky, M., Caballero, C., & Arganis-Juarez, M. (2022). Alteration of the flow distribution at a river bifurcation caused by a system of upstream dams: Case of the Grijalva River basin, Mexico. Earth Surface Processes and Landforms, 47(2), 509– 521. https://doi.org/10.1002/esp.5265
Mendoza, A., Soto-Cortes, G., Priego-Hernandez, G., & Rivera-Trejo, F. (2019). Historical description of the morphology and hydraulic behavior of a bifurcation in the lowlands of the Grijalva River Basin, Mexico. Catena, 176, 343–351. https://doi.org/10.1016/j.catena.2019.01.033
Muñoz-Salinas, E., Castillo, M., Sanderson, D., Kinnard, T., & Cruz-Zaragoza, E. (2016). Using three different approaches of OSL for the study of young fluvial sediments at the coast plain of the Usumacinta-Grijalva River basin, southern Mexico. Earth Surface Processes and Landforms, 41(6), 823–834. https://doi.org/10.1002/esp.3880
Nanson, G., & Huang, H. (2017). Self-adjustment in rivers: Evidence for least action as the primary control of alluvial-channel form and process. Earth Surface Processes and Landforms, 42(4), 575–594. https://doi.org/10.1002/esp.3999
Ochoa-Gaona, S., & Gonzalez-Espinosa, M. (2000). Land use change and deforestation in the highlands of Chiapas, Mexico. Applied Geography, 20(1), 17–42. https://doi.org/10.1016/s0143-6228(99)00017-x
Park, C., & Schmincke, H. (2020). Multistage damming of the Rhine River by tephra fallout during the 12,900 BP Plinian Laacher See Eruption (Germany). Syn-eruptive Rhine damming I. Journal of Volcanology and Geothermal Research, 389, 1–26. https://doi.org/10.1016/j.jvolgeores.2019.106688
Pierson, T., & Major, J. (2014). Hydrogeomorphic effects of explosive volcanic eruptions on drainage basins. Annual Review of Earth and Planetary Science, 42, 469–507. https://doi.org/10.1146/annurev-earth-060313-054913
Pierson, T., Prongle, P., & Cameron, K. (2011). Magnitude and timing of downstream channel aggradation and degradation in response to a dome-building eruption at Mount Hood, Oregon. GSA Bulletin, 123, 3–20. https://doi.org/10.1130/b30127.1
Pitari, G., & Visconti, G. (1984). Global transport of volcanic aerosol from El Chichon eruption studied with a three-dimensional circulation model. Geofísica Internacional, 23(3), 415–431. https://doi.org/10.22201/igeof.00167169p.1984.23.3.1161
Reisenbüchler, M., Bui, M., Skublics, D., & Rutschmann, P. (2019). An integrated approach for investigating the correlation between floods and river morphology: a case study of the Saalach River, Germany. Science of the Total Environment, 647, 814–826. https://doi.org/10.1016/j.scitotenv.2018.08.018
Rivera, F., Vázquez, J., Mendoza, J., Estrada, M., & Soto, G. (2008). Sediment distribution alteration in a plain river bifurcation due to induced morphological changes. Ingeniería Hidráulica en México, 23, 49-63.
Rivera-Trejo, F., & Rosas-Figueroa, C. (2014). Study for the characterization and quantification of sediments (in Spanish). Tabasco: PROHTAB-UNAM.
Rivera-Trejo, F., Soto-Cortés, G., & Méndez-Antonio, B. (2010). The 2007 flood in Tabasco, Mexico: An integral analysis of a devastating phenomenon. International Journal of River Basin Management, 8, 255– 267. https://doi.org/10.1080/15715124.2010.508746
Rodríguez, S. R., Mora-González, I., & Murrieta-Hernádez, J. L. (2006). Flujos de baja concentración asociados con lluvias de intensidad extraordinaria en el flanco sur del volcán Pico de Orizaba (Citlaltépetl), México. Boletín de la Sociedad Geológica Mexicana, 58(2), 223–236. https://doi.org/10.18268/bsgm2006v58n2a5
Salter, G., Voller, V., & Paola, C. (2019). How does the downstream boundary affect avulsion dynamics in a laboratory bifurcation? Earth Surface Dynamics, 7, 911–927. https://doi.org/10.5194/esurf-7-911-2019
Sclafani, P., Nygaard, C., & Throne, C. (2018). Applying geomorphological principles and engineering science to develop a phased Sediment Management Plan for Mount St Helens, Washington. Earth Surface Processes and Landforms, 43(5), 1088–1104. https://doi.org/10.1002/esp.4277
Scolamacchia, T., & Macías, J. (2005). Distribution and stratigraphy of deposits produced by diluted pyroclastics density currents of the 1982 eruption of El Chichón volcano, Chiapas, Mexico. Revista Mexicana de Ciencias Geológicas, 22(2), 159–180.
Scolamacchia, T., & Schouwenaars, R. (2009). High-speed impacts by ash particles in the 1982 eruption of El Chichón, Mexico. Journal of Geophysical Research, 114, 1–14. https://doi.org/10.1029/2008jb005848
Sigurdsson, H., Carey, S., & Espindola, J. (1984). The 1982 eruptions of El Chichón Volcano, Mexico: Stratigraphy of pyroclastic deposits. Journal of Volcanology and Geothermal Research, 23(1-2), 11–37. https://doi.org/10.1016/0377-0273(84)90055-6
Smolar-Zvanut, N., & Mikos, M. (2014). The impact of flow regulation by hydropower dams on the peripython community in the Soca River, Slovenia. Hydrological Sciences Journal, 59(5), 1–14. https://doi.org/10.1080/02626667.2013.834339
Su, T., Huang, H., Carling, P., Yu, G., & Nanson, G. (2021). Channel-form adjustment of an alluvial river under hydrodynamic and eco-geomorphologic controls: insights from applyingequilibrium theorygoverningalluvial channel flow. Water Resources Research, 57(11). https://doi.org/10.1029/2020WR029174
Tanarro, L., Andrés, N., Zamorano, J., Palacion, D., & Renschler, C.S. (2010). Geomorphological evolution of a fluvial channel after primary lahar deposition: Huiloac Gorge, Popocatépetl volcano (Mexico). Geomorphology, 122, 178–190. https://doi.org/10.1016/j.geomorph.2010.06.013
Thouret, J. (1999). Volcanic geomorphology - an overview. Earth-Science Reviews, 47, 95–131. https://doi.org/10.1016/s0012-8252(99)00014-8
Tian, P., Lu, H., Feng, W., Guan, Y., & Xue, Y. (2020). Large decrease in streamflow and sediment load in Qinghai-Tibetan Plateau driven by future climate change: A case study in Lhasa River Basin. Catena, 187, 1–12. https://doi.org/10.1016/j.catena.2019.104340
Tilling, R. I. (2009). El Chichón’s “surprise” eruption in 1982: Lessons for reducing volcano risk. Geofísica Internacional, 48(1), 3–19. https://doi.org/10.22201/igeof.00167169p.2009.48.1.96
Ulloa, H., Iroumé, A., Picco, L., Korup, O., Lemzi, M. L. M. & Ravazzolo, D. (2015). Massive biomass flushing despite modest channel response in the Rayas River Following the 2008 eruption of Chaitén volcano, Chile. Geomorphology, 250, 397–406. https://doi.org/10.1016/j.geomorph.2015.09.019
Weiss, N. A. (2016). Introductory statistics (10th ed). Pearson.
Zhou, M., Deng, J., Lin, Y., Zhang, L., He, S., & Yang, W. (2021). Evaluating combined effects of socio-economic development and ecological conservation policies on sediment retention service in the Qiantang River basin, China. Journal of Cleaner Production, 286, 124961. https://doi.org/10.1016/j.jclepro.2020.124961
Peer Reviewing under the responsibility of Universidad Nacional Autónoma de México.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/)