Abstract

The increasing magnitude and frequency of maximum streamflow flooding in Central Chile are driving continuous adjustments in the design and maintenance of public and private infrastructure, posing significant challenges in the context of climate change. To assess the historical and future impacts of extreme flooding on Electricity Transmission Infrastructure (ETI), this study implemented coupled hydrological and hydraulic simulations of streamflow in the Tinguiririca River Basin (~34.6°S), Central Chile. Hydrological and topographic data were collected for both the basin and the simulation domain, which includes a critical transmission tower located near the riverbank. Large-scale topography was obtained using ALOS-PALSAR satellite data (~12.5 m resolution), while high-resolution elevation data (~0.5 m) were acquired for the infrastructure site using a LiDAR-equipped Unmanned Aerial Vehicle (UAV). Historical climate data and future projections from the NEX-GDDP dataset (2010–2100) were used to simulate daily and sub-daily streamflow across multiple climate change scenarios. The results reveal that extreme flood events in this section of the river are projected to become more frequent and severe than any recorded in the historical period. The study highlights the strong influence of digital elevation model resolution on flood simulation accuracy and evaluates the resilience of ETI under unprecedented extreme hydrological conditions. Beyond providing a detailed flood risk assessment, the study establishes a practical framework for electric utility companies in Chile, especially in the transmission sector. This framework supports informed decision-making for the design, protection, and adaptation of infrastructure at river crossings under future climate conditions. The findings underscore the importance of regular topographic updates, discharge curve maintenance, and high-resolution simulations as part of a long-term strategy for climate-resilient energy systems.

1 Introduction

The increasing frequency and intensity of extreme flood events pose a growing challenge for the resilience of public and private infrastructure in a world increasingly affected by climate-related hazards (IPCC 2022). In this context, coupled hydrological–hydraulic simulations (Thakur et al. 2017; Abdessamed and Abderrazak 2019; Peker et al. 2024; Verma et al. 2025) have become essential tools to understand and evaluate the impacts of flooding on critical systems, including Electricity Transmission Infrastructure (ETI) (Tari et al. 2021; Violante et al. 2022; Wang et al. 2024; Plata-Rocha et al. 2025; Paredes et al. 2025).

This study focuses on the Tinguiririca River Basin in Central Chile (~34.6°S), where climate-based ensemble simulations of sub-daily and daily streamflow were conducted to assess historical and future extreme flood events. The analysis targets a key transmission tower operated by CELEO, located downstream of the confluence of the  Tinguiririca (in Bajo Briones) and Claro (in the valley) rivers (hereafter referred to as the Tinguiririca and Claro rivers). The hydrological simulations were developed using the HEC-HMS model (Feldman 2000), while hydraulic simulations were carried out with the one-dimensional HEC-RAS model (Brunner 2016). The coupling of both models enables a robust representation of streamflow dynamics and water volumes during extreme flood events (Thakur et al. 2017; Abdessamed and Abderrazak 2019; Maidment and Djokic 2000; Knebl et al. 2005; Peker et al. 2024; Verma et al. 2025; Abegaz et al. 2026).

A critical factor in accurately simulating flood extent is the spatial resolution of the digital elevation model (DEM) used. The DEM’s resolution significantly influences the ability to represent terrain variations and, consequently, the capacity to assess flood impacts on infrastructure. Coarse-resolution models may misrepresent flood extents, leading to under- or overestimation of risks, while fine-resolution models improve precision but require substantially more computational resources (Dabberdt et al. 2000). Therefore, the selection of a DEM must balance the trade-offs between accuracy, computational efficiency, and project constraints.

This research evaluates the implications of using coarse versus fine topographic datasets in flood modeling. Two DEMs were analyzed:

1. A widely available, medium-resolution DEM from ALOS-PALSAR (~12.5 m) (ASF 2023), and
2. A high-resolution DEM (~0.5 m) derived from LiDAR data collected with an Unmanned Aerial Vehicle (UAV).

These DEMs were applied within hydrological–hydraulic models driven by historical and projected climate scenarios from the NEX-GDDP dataset (NASA Earth Exchange Global Daily Downscaled Climate Projections) (Thrasher et al. 2012; Taylor et al. 2012).

This comparison provides new insights into how DEM resolution affects flood simulation outputs under both historical and future climate conditions. Notably, some of the projected extreme flood events have magnitudes and frequencies that exceed those previously recorded in the basin. This underlines the importance of integrating high-resolution data and forward-looking models into ETI planning, particularly in steep Andean basins where short distances between mountains and the coast enhance flood hazard potential. The findings support evidence-based decision-making for the investment, design, and maintenance of electricity transmission infrastructure in Chile, and present a transferable workflow that can be used as a planning standard for other river crossings with critical infrastructure.

The following sections describe the materials and methods used to set up the hydrological and hydraulic models, the calibration and validation process, the generation of climate-driven ensemble simulations, and the results of the comparison between historical and projected extreme flood scenarios in the Tinguiririca River.

2 Materials and Methods

2.1 Study area

The Tinguiririca River Basin is in Central Chile (~34°17’ south latitude), in the O'Higgins Region, specifically in the San Fernando municipality (Figure 1a). The existing ETI corresponds to the tower 126/1 (Figure 1b), owned by Alto-Jahuel Energy Transmission, and operated by CELEO-Chile. Since Chile’s National Electric System is interconnected, the energy transmitted through the Alto Jahuel substation contributes to the electricity supply of multiple regions and millions of people across the country.

Figure 1  Study area including (a) location of Tinguiririca Basin, (b) existing electrical transmission infrastructure corresponding to the tower 126/1, (c) Digital Elevation Model (DEM) of ALOS-PALSAR for the Tinguiririca basin and subbasins, including meteorological stations, and (d) land use in the Tinguiririca basin.

The transmission tower is secured by a riverbank protection infrastructure (flood defense) constructed in 2012. Both the tower and the riverbank protection infrastructure are situated in the outlet of a basin with a drainage area of 1837.8 km² (Figure 1a), where a 78.3% comprises the upper Tinguiririca River Basin, 20% includes the Claro River Basin, and the remaining 1.7% corresponds to the area surrounding the riverbank protection infrastructure (Figure 1c).

The whole contributing basin to the existing electrical transmission infrastructure has an area of 1837.8 km² with an average altitude of 2,359 meters above sea level (m.a.s.l.). The maximum elevation is 4,900 m.a.s.l., and the minimum elevation is 499 m.a.s.l., with an average slope of 40%. This clearly reveals the high hydraulic gradient of the Tinguiririca basin. The Horton's shape factor (Horton 1932) is 0.52, corresponding to a wide basin characterized by high streamflow production and high potential for flooding events. Additionally, the characteristics associated with the drainage networks of such basins often result in multiple flood hydrographs over time, displaying multiple rising and falling curves. The main river channel has a length of 71.9 kms with an average slope of 3.7%. The maximum elevation is 3,163 m.a.s.l., and the minimum elevation of 460 m.a.s.l. The average concentration time was estimated from several methods as an average of 6.5 hours (Rowe et al. 1953; Kirpich 1940; Giandotti 1934; Temez 1978). These morphometric factors show the capacity of the basin to generate extreme floods with high flow velocities and low concentration time.

The land cover in the basin includes five predominant classes (Figure 1d), distributed as 50% bare soils, 20% shrublands, 10% grasslands, 5% forest cover, and 5% snow and ice. In contrast, the Claro River Basin has 35% shrublands, 28% native forest, 20% unproductive areas, 14% grassland, while the remaining 3% corresponds to other classes (Boisier et al. 2018). The degree of human-intervention of each subbasin was estimated from the Human-Intervention Index which relates the average annual flow assigned as surface water rights (permanent and continuous consumptive rights), normalized by the basin's average annual flow. In this context, values close to zero have a low level of intervention, while values close to one would indicate a high degree of intervention. The index calculated for both basins ranges between 0.12 for the Tinguiririca subbasin to 0.2 for the Claro subbasin, therefore, both basins are classified as medium-to-low level of intervention.

2.2 Climatology

The climate in the Tinguiririca River Basin is Mediterranean, with precipitation primarily concentrated between the months of May to August (MJJA) (Valdés-Pineda et al. 2015; 2016). According to records from the La Rufina rain gauge station, located 15 kms upstream from the outlet of the contributing basin in the Andean foothills (at an altitude of approximately 743 m.a.s.l.), the average accumulated precipitation between May and August (MJJA) is 602 mm (1984–2019). This accumulation represents on average 81% of the average annual precipitation accumulation observed over this basin (Figure 2).

Figure 2  Monthly average precipitation and accumulated monthly average precipitation for La Rufina rain gauge for the 1984–2019 period.

In addition to the previous analysis, a total of 42 years of available rainfall records (1979–2020) provided by the gridded product CR2MET (Boisier et al. 2018) were used to map the spatial and interannual variability of the annual rainfall accumulation in the Tinguiririca River Basin (Figure 3). Using the gridded product with a spatial resolution of 5 km², it can be observed that the accumulation is higher towards the Andes Mountains and decreases towards the valley that encompasses the drainage point of the basin. The seasonal rainfall cycle represented by the CR2MET product aligns with the observed cycle from the available rain gauge stations (Figure 4), depicting a rainfall pattern with higher accumulations during the winter months (MJJ) and a reduction in accumulation during the summer months (DJF).

Figure 3  Interannual variability of annual rainfall accumulation between 1979 and 2020 at 5 km² spatial resolution extracted from the CR2MET gridded product.
NOTE: The red dots represent the rain gauge stations of the General Water Directorate (DGA) available for the basin.

Figure 4  Scatter plots comparing monthly rainfall accumulation from DGA rain gauges and CR2MET gridded data (at ~5 km² of spatial resolution).
NOTE: The rainfall records of the DGA rain gauges: (a) Tinguiririca River in Bajo Briones, and (b) La Rufina were compared to their corresponding CR2MET grid-cell.

2.3 Hydrological regime

The Tinguiririca river exhibits a nivo-pluvial hydrological regime, resulting in higher monthly average flows during the summer months of December and January due to snow and glacier melt in the upper part of the basin (Cortés et al. 2011; Bravo et al. 2017). According to data for the period 1985–2020, an average streamflow of 90 m³/s was recorded between the months of December and January (Figure 5a). The flow duration curve for the period 1985–2020 revealed that the streamflow with 85% of probability of occurrence (Q₈₅) is 17.3 m³/s for this river (Figure 5b). On the other hand, the Claro River has a pluvial regime, with its highest streamflow records occurring in the winter season (June to October), reaching an average value of 15 m³/s (Figure 5c). As this basin lacks contributions from snow and glaciers, the flows in the summer period are mainly base flow. The flow duration curves at the Claro River station in the Valley report a Q₈₅ of 1.2 m³/s for the period 1985–2020 (Figure 5d). Despite both basins revealing low average streamflow and Q₈₅ values, extreme flooding events can exceed average conditions by several orders of magnitude.

Figure 5  (a) Monthly average streamflow; (b) Flow duration curve for the period 1985–2020 at the Tinguiririca River in Bajo Briones stream gauge; (c) Monthly average streamflow; and (d) Flow duration curve for the period 1985–2020 at the Claro River in the Valley stream gauge.

2.4 Maximum flooding events

The frequency analysis of maximum peak flows in the Tinguiririca Basin was applied using both direct and indirect methods of calculation. The Direct Method is the classical frequency analysis of peak flows using probability distribution functions for extreme values, such as Gumbel (Extreme Values Type I), Normal, Log-Normal, and Pearson Type III, among others. The goodness-of-fit for the direct method was evaluated using the Kolmogorov-Smirnov (K-S) test and the coefficient of correlation (R²). The Indirect Method to estimate maximum peak flows in ungauged basins, included the implementation of the DGA-AC Method calibrated for basins of central Chile (DGA 1995), the Modified Verni and King Method (Verni and King 1977), and the Rational Method (Chow et al. 1994) (see details in Table 1). Given the large size of the Tinguiririca Basin, the Rational Method was used only as a comparative reference alongside the other indirect methods, and it was not used in any subsequent analyses nor in the climate‑change‑driven predictive simulations.

Table 1  Direct and indirect methods utilized for maximum peak flow estimation in the Tinguiririca River Basin.

Method	Number	Function		Eq.#	Variables
Direct	Probability Distribution Function of Flows			(1)	F(x) = selected probability distribution function. P = probability of exceedance. x = observed flow. X = design flow for a return period T (years). T = return period (years).
Indirect	DGA-AC			(2)	Q10  = mean daily flow (m3/s) for a return period of 10 years. AP = pluvial area of the basin (km2).  = maximum daily precipitation (mm) for a return period of 10 years.
				(3)	QT=x = instantaneous flow for a return period T = x years. QT=10 = mean daily flow (m3/s) for a return period T = 10 years. ∝T=x = conversion factor for a return period T = x years, where ∝T=2 = 0.43, ∝T=5 = 0.74, ∝T=10 = 1.00, ∝T=20 = 1.28, ∝T=50 = 1.71, and ∝T=100 = 2.08
	Modified Verni and King Method			(4)	Q = maximum instantaneous flow (m3/s) for the return period T (years). C(T) = empirical coefficient for return period T. P24 = maximum daily precipitation (mm) associated with the return period T. A = pluvial area of the basin (km2).
	Rational Method			(5)	Q = maximum flow (m3/s) for the return period T. C = runoff coefficient for the basin for return period T. I = design rainfall intensity (mm/h) A = area of the basin (km2).

2.5 High-resolution topographic mapping

The spatial resolution used to represent surface water processes requires Digital Elevation Models (DEMs) that can accurately represent variations of terrain or infrastructure. For this study, ALOS-PALSAR satellite product (12.5 m of spatial resolution) (ASF 2023) was extracted for the whole basin from the NASA Alaska Satellite Facility Distributed Active Archive Center (https://search.asf.alaska.edu/#/). Two raster images (AP_264684320 and AP_264684330) were utilized to delineate the basin, perform morphometric analyses, and perform hydraulic simulations in the section of interest (Figures 6a, and 6b). Additionally, we also generated a hyper-resolution DEM (~0.5 m of spatial resolution) for the section of interest using a LiDAR sensor mounted in an UAV (Figures 6c, and 6d).

Figure 6  (a) ALOS-PALSAR raster images AP_264684320 and AP_264684330; (b) basin delineated from the drainage point of interest at tower 126/1 using the ALOS-PALSAR DEM and used to perform hydrologic (hourly or daily streamflow) simulations; (c) high-resolution image (0.5 m); and (d) high-resolution DEM (0.5 m) for the hydraulic (flooding) simulation domain generated with UAV-LiDAR sensor.

2.6 Climate projections for the Tinguiririca Basin

To assess the potential impact of future floods on the existing electrical transmission infrastructure, climate projections of daily precipitation, and minimum/maximum temperatures from NASA/NEX-GDDP (Thrasher et al. 2012) were extracted for the whole basin for the period 1950–2100 (see details of models in supplementary material). The climate models were validated through a comparison of precipitation and temperature records against the Tinguiririca at Bajo Briones meteorological station for the period 1980–2014. The three best climate-assimilated models were used to establish the most pessimistic possible flooding scenario considering extreme precipitation and temperature conditions in the Tinguiririca basin. The validated datasets were used as forcings for the calibrated hydrological model and for developing projected streamflow projections.

2.7 Construction and calibration of hydrological model

The HEC-HMS hydrological model was built and calibrated to simulate streamflow in the Tinguiririca River basin and to improve the understanding of flooding events within the domain of interest. For the model calibration, the DGA network including two meteorologic stations:

1. Teno River after the confluence with Río Claro
2. Tinguiririca River in Bajo Briones

Two stream gauging stations:

1. Claro River in the Valley
2. Tinguiririca River in Bajo Briones

One snow gauge station:

1. Termas del Flaco

The latter is located at approximately 2600 m.a.s.l. and generates Snow Water Equivalent (SWE) and temperature records. Furthermore, as most of the basin is in the Andes mountains and the Andean foothills, the effects of temperature during rainstorm events were also considered as they play a crucial role in the partitioning of precipitation into rain or snow. For example, higher temperatures can lead to:

• Increased accumulation of precipitation as rain, and
• Elevation rise of the snow line, both resulting in augmented runoff from snow melting processes and from an increase in snow-free areas.

For this reason, the model included the following components:

• Temperature Index to represent snowmelt contributions,
• SCS Unit Hydrograph for the precipitation-runoff model,
• Recession method for base flow estimates, and
• Surface method for soil surface representation, constant losses, and flood routing for both rivers.

The parameter calibration process was carried out using observed streamflow records from the wet season of 2006 where extreme successive flooding events occurred between May 2 and August 8 with a maximum instantaneous recorded peak flow for the Tinguiririca river of around 1304 m³/s. The final calibrated model for this season event had an NSE (Nash-Sutcliffe Efficiency) efficiency of 0.81, which suggests a relatively high level of agreement between the model simulations and the observed streamflow data, where about 81% of the variability in the observed data is explained by the model. This model was subsequently used to generate sub-daily and daily streamflow simulations using historical and projected climate data.

2.8 Construction and coupling of 1D hydraulic model

HEC‑RAS was implemented and coupled with HEC-HMS simulations using a one‑dimensional (1D) hydraulic model configuration, which is appropriate for the geomorphological characteristics of the Tinguiririca River at the study site. Although the channel widens from approximately 80–100 m upstream to nearly 400 m in the vicinity of the transmission tower, the system maintains a well‑defined channel during flooding events, with a predominantly longitudinal flow direction and a wide inundation surface that behaves as an expansion zone rather than a distributary network. Under these conditions, a 1D formulation is adequate to represent the propagation of flood waves, water‑surface elevations, and hydraulic energy relevant for assessing the impact on the electrical infrastructure. The model was run under steady and unsteady flow conditions using a kinematic‑wave formulation and a mixed flow regime (subcritical and supercritical). The hydraulic domain consisted of 57 cross‑sections spaced at 20 m intervals, constructed from both the ALOS‑PALSAR (12.5 m) and UAV‑LiDAR (0.5 m) DEMs. Details of the geometry and the assumptions of the constructed HEC‑RAS model are presented in the supplementary material.

The geometry used for the simulation domain in the areas of interest, i.e., the hydraulic network, riverbanks, and main flood zones, was constructed based on the DEMs presented in Section 2.6. The bottom slope of the river for the simulated section was estimated to be 0.010 m/m. The cross-sections constructed along the central axis of the river had a spacing of 20 m, defining a total of 57 cross-sections in the simulated area of interest (see details in supplementary material).

Using data collected from a site visit (see details in supplementary material) and information obtained from the gridded land cover product available for Chile (Zhao et al. 2016) (Figure 7), Manning's roughness coefficients were estimated for the riverbed, riverbanks, and the floodplains using the Cowan method (1956) (see details in supplementary material).

Figure 7  (a) Land cover map from Zhao et al. (2016) cropped for the simulation domain; (b) roughness coefficients estimated for the simulation domain.

2.9 Simulation of extreme flooding events

Through the construction and coupling of hydrological and hydraulic models (HEC-HMS and HEC-RAS), it was possible to simulate, classify, and evaluate historical and future extreme floods for the area of interest. Based on the observed historical hydrographs (Figure 8) and the projected streamflow simulations, three representative scenarios were defined (Table 2):

• Scenario 1, corresponding to the most extreme historical event recorded in the basin (May 27, 2012; peak flow of 1,634 m³/s);corresponding to 1,430 m³/s from Tinguiririca and 204 m³/s from Claro); 
• Scenario 2, a scaled synthetic extreme event derived from the 2012 hydrograph using the GHD (2012) estimate of 3,838 m³/s; and
• Scenario 3, representing the most extreme climate‑driven event projected for 2010–2100 across all NEX‑GDDP climate models used for this basin, whose peak discharge reached 4,260 m³/s. Details of the climate models used in this study are presented in the supplementary material.

Figure 8 Hourly streamflow hydrographs measured at the Tinguiririca in Bajo Briones stream gauge.
NOTE: The flow duration curve (FDC) of Tinguiririca River is included to illustrate the probabilities of occurrence of flood events.

Table 2 Summary of the three flood simulation scenarios developed using coupled hydrological (HEC-HMS) and hydraulic (HEC-RAS) models.

Scenario	Description	HEC-HMS Input	HEC-RAS Simulation	Peak Discharge (m³/s)	Notes
1	Simulation of the most extreme observed historical flood (Tinguiririca River, May 27, 2012)	Observed hourly rainfall and discharge	Dynamic wave routing using 1D Saint-Venant equations	1,634 (Tinguiririca 1,430 m3/s + Claro 204 m3/s)	Based on observed events. Used as baseline.
2	Scaled extreme event using hydrological assumptions	Synthetic hydrograph scaled linearly from 2012 event using GHD (2012) streamflow estimate	Dynamic wave routing using 1D Saint-Venant equations	3,838 m3/s	Discharge estimated via direct/indirect methods. Valid under assumptions of linearity and superposition (Dooge 1973).
3	Projected extreme flood based on climate change simulations (2010–2100)	Daily projections of precipitation, Tmin, and Tmax from NEX-GDDP	Dynamic wave routing using 1D Saint-Venant equations	>4,000 m3/s	Peak event was selected as the maximum simulated streamflow between all NEX-GDDP models for the simulation period 2010–2100.

NOTE: Each scenario represents different levels of flood severity (based on historical records, scaled synthetic events, and future climate projections) and was used to evaluate the potential impact on the study domain.

2.10 Methodological framework

The simulation of historical and future flooding scenarios in the Tinguiririca basin allowed for the definition of flooding discharges, elevations, areas, and velocities, etc., that were used to assess the potential impact on the electrical transmission infrastructure and to inform appropriate management schemes for its current and future protection and maintenance. The methodological framework used in this study to evaluate the proposed scenarios is presented in Figure 9.

Figure 9  Methodological framework used to perform coupled hydrologic-hydraulic simulations for both steady and unsteady flow.
NOTE: Flooding simulations were performed for historical conditions and climate projections using the NEX-GDDP Climate Models.

3 Results and Discussion

3.1 Comparison of digital elevation models

The comparison between Digital Elevation Models (DEMs) derived from ALOS-PALSAR and UAV-based LiDAR revealed substantial differences in terrain representation within the cross-section that includes the electricity transmission infrastructure (cross-section CS-640). As shown in Figures 10a–10d, the average elevation difference between both models was 7.2 m, with a maximum difference of 12.7 m and a minimum difference of 2.5 m. These differences are primarily explained by the contrasting spatial resolutions of the two DEMs, as well as by morphological changes in the riverbed that have occurred since 2012, when the ALOS‑PALSAR data were acquired. In contrast, the UAV‑LiDAR DEM represents the most recent and highest‑resolution topographic survey, reflecting the current hydraulic configuration of the channel.

Figure 10  Digital Elevation Models (DEMs) obtained from (a) ALOS-PALSAR, and (c) UAV-LiDAR. Histograms of terrain elevation mapped from (b) ALOS-PALSAR, and (d) UAV-LiDAR. (e) Cross-section at the electrical transmission infrastructure.
NOTE: The elevation of area where the transmission tower is located is 475.40 m.a.s.l. from ALOS-PALSAR and 467.55 m.a.s.l. according to the UAV-LiDAR.

These discrepancies are primarily attributed to the differences in spatial resolution between the two datasets. Specifically, the DEM derived from ALOS-PALSAR, with a spatial resolution of approximately 12.5 m, produced a wetted perimeter that was 29.5% smaller than that derived from the UAV-LiDAR DEM, which has a much finer resolution of 0.5 m. This variation in perimeter and cross-sectional geometry directly affects the simulation of floodwater volumes and flow behavior. A coarser-resolution DEM like ALOS-PALSAR may represent a narrower channel or smaller cross-section, potentially overestimating flood extents and misrepresenting the actual behavior of extreme flooding events. Conversely, the finer-resolution UAV-LiDAR DEM captures more detailed topographic features, providing a more accurate basis for simulating water levels and assessing flood impacts.

The observed differences in terrain representation (Figure 10e) have significant implications for flood modeling. The UAV-LiDAR, by providing high-resolution data (0.5 m), allows for detailed capture of topographic variations, such as small depressions, minor slope changes, and the precise location of river defense structures or obstacles. This enables more realistic simulations of streamflow dynamics and water volumes during extreme events, facilitating a more accurate assessment of potential impacts on electrical transmission infrastructure.

While the ALOS-PALSAR DEM remains useful for larger-scale flood simulations (offering a broad overview of basin-wide terrain features) it lacks the spatial detail necessary to evaluate localized risks. As such, its use may lead to underestimation or overestimation of critical flood parameters near infrastructure assets. In the context of climate change, where small-scale variations in topography can significantly influence flood behavior and infrastructure vulnerability, the use of fine-resolution DEMs becomes essential for effective planning, risk assessment, and resilience building.

3.2 Estimation of extreme peak flows

The estimation of peak flow rates for the Tinguiririca River using both empirical and direct statistical methods (as described in Section 2.5) revealed notable differences across return periods and methodologies. The DGA-AC and Rational Method produced lower estimates, with 100-year peak flows remaining below 2,000 m³/s. In contrast, the Verni and King method, along with the direct method using the Gumbel probability distribution function (PDF) calibrated to observed streamflow data, yielded peak flow estimates exceeding 2,300 m³/s (Table 3).

Table 3  Maximum streamflow rates for different return periods from direct and indirect methods.

Return Period	Direct Method (Gumbel) P(x>X) (m3/s)	DGA-AC (m3/s)	Verni and King (m3/s)	Rational (m3/s)
2	557.9	233.3	334.5	738.1
5	966.7	365.2	645.1	1039.4
10	1165.2	507.3	916.4	1235.1
20	1423.4	786.3	1230.4	1430.7
50	1857.8	1283.4	1695.5	1676.7
100	2301.1	1780.6	2338.7	1861.3

Among these, the Gumbel PDF (recognized for its robustness in modeling extreme hydrological events in central Chile) provided the best statistical fit to observed peak discharges, consistent with findings from multiple hydrological studies in Central Chile (Mintegui and Robredo 1993; Temez 1978; Pizarro 1986; Ponce 1989; Chow et al. 1994; Monsalve 1999; Pizarro et al. 2012; 2013). This reinforces the validity of using the Gumbel approach in regions with similar hydroclimatic conditions. Moreover, both the Gumbel-based and Verni and King methods represent worst-case scenarios in terms of potential flood impacts on critical electricity transmission infrastructure.

These discharge estimates were derived using empirical formulas based on physical characteristics of the basin, including contributing area, slope, runoff coefficient, and other geomorphological factors. However, their accuracy is inherently dependent on the specific hydrological context and climatic zone of the study area. It is important to note that the peak flow estimates presented here are significantly lower than those reported by GHD (2012), which proposed a peak discharge of 3,838 m³/s for a 100-year return period, approximately 65% higher than the highest value estimated through the empirical methods used in this study.

To further evaluate extreme flood scenarios, three reference discharges were selected for hydraulic simulations (Figure 11):

• The 2012 observed flood of 1,634 m³/s,
• The GHD (2012) proposed flood of 3,838 m³/s, and
• The maximum projected flood of 4,260 m³/s, derived from climate simulations for the period 2010–2100.

This last scenario, based on the MRI-CGCM3 / RCP-4.5 model, represents a future extreme event associated with an average projected temperature increase of 2.4°C. These projections suggest that future floods could be larger than both historical and recent extremes (Figure 11), reinforcing the importance of incorporating climate change scenarios in infrastructure risk assessments and resilience planning.

Figure 11  Daily streamflow simulations for the Tinguiririca River between 2010 and 2100 using an HEC-HMS model.

3.3 Simulations of extreme flooding extent

The simulation of three extreme flood scenarios (corresponding to peak discharges of 1,634 m³/s, 3,838 m³/s, and 4,260 m³/s) revealed a clear increase in floodplain inundation with increasing discharge, particularly in the area northwest of the transmission tower on the right bank of the river. In all cases, overflow into this sector was evident in both topographic models: the coarser ALOS PALSAR DEM and the higher-resolution UAV-LiDAR DEM (Figures 12a–c and 13a–c for ALOS; Figures 12d–f and 13d–f for UAV-LiDAR). 

Furthermore, it is important to note that the assessment of the most extreme flood obtained from climate projections (4,260 m³/s) revealed that for ALOS PALSAR the flooding water level remained 1.75 m below the level where the electrical transmission infrastructure is located. In the case of UAV-LiDAR data, the flooding water level was only 0.55 m lower. The maximum flow velocities of the flooding over the domain of interest revealed values close to 3.9 m/s under subcritical flow conditions for both finer and coarser model configurations (see details in Table 4).

Table 4  Summary of hydraulic variables simulated at cross-section CS-640 for three extreme discharge scenarios.

Cross-section	1,634 m3/s		3,838 m3/s		4,260 m3/s
	ALOS PALSAR	UAV- LiDAR	ALOS PALSAR	UAV-LiDAR	ALOS PALSAR	UAV-LiDAR
Elevation energy line (m)	472.75	466.24	474.17	467.55	474.4	467.76
Water velocity (m)	0.24	0.38	0.66	0.75	0.75	0.81
Water level elevation (m)	472.51	465.86	473.52	466.8	473.65	466.95
Critical water depth (m)	471.33	465.58	472.57	466.48	472.76	466.62
Energy line slope (m/m)	0.0033	0.0112	0.006	0.0112	0.0065	0.011
Total discharge (m³/s)	1634	1634	3838	3838	4260	4260
Upper section width (m)	306.41	428.21	322.7	437.06	324.43	470.63
Total velocity (m/s)	2.17	2.73	3.59	3.82	3.82	3.98
Flow area (m²)	751.29	597.99	1069.85	1002.59	1113.79	1066.24
Hydraulic depth (m)	2.45	1.4	3.32	2.33	3.43	2.48
Wetted perimeter (m)	306.8	435.28	323.24	438.55	324.98	438.95
Maximum channel depth (m)	3.83	3.12	4.84	4.06	4.97	4.21

The total inundation area for the three scenarios evaluated within the simulation domain was 40.5, 48.1, and 48.9 hectares (Figures 12a, 12b, and 12c) from ALOS PALSAR, and 41.3, 49.1, and 50.6 hectares for UAV-LiDAR (Figures 12d, 12e, and 12f). For most results, the simulated floods caused overflows downstream of the electrical transmission tower and on the right bank of the Tinguiririca River; however, the spatial distribution of the flooded areas also revealed differences when comparing between different scenarios and elevation models used.

Figure 12  Plan view of extreme flooding simulations for the three selected scenarios (1,634 m³/s, 3,838 m³/s, and 4,260 m³/s). The simulations were developed using ALOS-PALSAR model (a), (b), and (c); and UAV-LiDAR model (d), (e), and (f).

Notably, simulations using the UAV-LiDAR DEM (~0.5 m resolution) provided more realistic flow paths compared to ALOS-PALSAR (Figure 13), improved representation of water depth variations, and greater accuracy in delineating localized inundation zones. These results underscore the greater capacity of high-resolution topographic data to capture detailed hydraulic responses under extreme flood conditions, offering a stronger foundation for impact assessments on nearby critical infrastructure.

Figure 13  3D view of extreme flooding simulations for the three selected scenarios (1,634 m³/s, 3,838 m³/s, and 4,260 m³/s). The simulations were developed using ALOS-PALSAR DEM (~12.5 m) (a), (b), and (c); and UAV-LiDAR (~0.5 m) (d), (e), and (f).

3.4 Discussion of extreme flooding impacts on ETI

The flood hydrographs and discharge curves generated from historic and projected streamflow simulations informed the assessment of flooding risk in the CS-640 cross-section that includes the ETI. According to our results and considering the historical background of this basin, flows equal to or lower than 1,500 m³/s are in a low-risk category in terms of impact on the existing riverbank protection infrastructure, and the location of the electrical transmission tower. Similarly, for floods with streamflow between 1,500 and 3,000 m³/s, a medium to moderate impact risk was established (see Figures 14a and 14b), and for all flows exceeding 3,000 m³/s, a high impact risk was assigned (see Figures 14c and 14d). Therefore, mitigation strategies for riverbank protection, and measures to reduce erosive processes that may impact the stability of the infrastructure, should consider the levels of impact established on this risk assessment.

Figure 14  Flooding risk levels and comparison of (a) and (c) Hydrographs, and (b) and (d) discharge curves for scenarios 1 and 3 (1,634 m³/s and 4,260 m³/s) using UAV-LiDAR (~0.5 m). The empirical relationships between streamflow discharge (m³/s) and water levels measured at the CS-640 cross section.

A comparison between the 2012 flood (1,634 m³/s classified as low to moderate risk), and the extreme flood simulated under climate change conditions (4,260 m³/s classified as high risk), revealed that the water levels of both events cannot reach the elevation where the ETI is located. In fact, both events varied between 1.6 m below the infrastructure surface for the 2012 flood (Figure 15a), to 55 cm below the infrastructure surface for the extreme projected flooding event (Figure 15b). However, the energy generated by the kinematic flooding waves can be extremely turbulent and can transport materials that can generate forces on the infrastructure that can cause erosive and mechanical impacts due to additional dynamic pressures and loads, vibrations, and cyclic loads that can weaken or damage the foundation of the ETI. For this reason, it will also be essential to periodically update the analyses that allow evaluation of the hydrodynamic behavior of this river and thus appropriately categorizing the risks and designing structures resilient to the combined effects of water elevations and kinematic waves during extreme floods in a climate change context.

Figure 15  Comparison and evaluation of the potential impact of (a) low- to moderate-risk flooding that occurred in 2012 (1,634 m³/s), and (b) high-risk flooding simulated from future projections of climate change models (2010–2099).

4 Conclusions

This research provides a valuable contribution to the understanding and assessment of extreme flood risks in river basins with critical infrastructure by integrating hydrological and hydraulic modeling with high-resolution topographic data and climate projections. One of the main strengths of the study lies in its comprehensive comparison of flood simulations using different digital elevation models, comparing a coarser ALOS-PALSAR DEM (12.5 m) with a high-resolution UAV-LiDAR model (0.5 m). The results demonstrate that the choice of elevation model has a significant impact on the accuracy of flood extent and depth estimations, which is crucial for risk assessment and infrastructure design.

The streamflow simulations included both historical extreme events and future flood scenarios derived from climate model projections, some of which represent unprecedented flow magnitudes and frequencies in the Tinguiririca River basin. This forward-looking approach enabled the evaluation of the true resilience of a transmission tower located on a riverbank protected by a flood defense, under conditions beyond the observed historical record. The study therefore provides not only a snapshot of current vulnerabilities, but also a foundation for planning adaptive measures to cope with increasingly intense and frequent extreme events due to climate change.

Given that many of Chile’s main rivers are characterized by steep slopes and short distances from the Andes to the Pacific, they are especially prone to extreme flood events intensified by climatic shifts. This makes the regular updating of cross-section mapping and simulation domains a key element in infrastructure resilience planning. For example, the discharge curves derived from these streamflow simulations have proven essential for understanding the hydrological behavior of the Tinguiririca River at the area of interest. Maintaining these updated curves will support ongoing flood simulations and the appropriate sizing of protective infrastructure.

This study has also established a practical framework that can be adopted by electric utility companies in Chile, particularly in the transmission sector. It serves as a reference for current and future investments in high-resolution topographic mapping and continuously updated flood simulations at river crossings with critical infrastructure. The findings are valuable not only for decision-makers and policymakers, but also for engineers, researchers, and utility operators. The proposed methodology supports evidence-based planning, improved hydraulic design, and the long-term maintenance of essential infrastructure, within a framework that aligns energy management with climate change adaptation and mitigation objectives.

Acknowledgments

The authors acknowledge the support and funding provided by CELEO Redes Chile in the implementation of this project. The results and conclusions of this study support the environmental monitoring system of CELEO, contributing significantly to enhancing adaptation and mitigation strategies in response to climate change.

References

1. Abegaz, R., J. Xu, and F. Wang. 2026. “Flood impacts on civil infrastructure systems: A comprehensive review of impacts, modeling tools, emerging technologies, and mitigation strategies.” Natural Hazards Review 27 (1): 03125002. https://doi.org/10.1061/NHREFO.NHENG-2477
2. Abdessamed, D., and B. Abderrazak. 2019. “Coupling HEC-RAS and HEC-HMS in Rainfall–Runoff Modeling and Evaluating Floodplain Inundation Maps in Arid Environments: Case Study of Ain Sefra City, Ksour Mountain. SW of Algeria.” Environmental Earth Sciences 78, 586. https://doi.org/10.1007/s12665-019-8604-6
3. Arora, V.K., J.F. Scinocca, G.J. Boer, J.R. Christian, K.L. Denman, G.M. Flato, V.V. Kharin, et al. 2011. “Carbon Emission Limits Required to Satisfy Future Representative Concentration Pathways of Greenhouse Gases.” Geophysical Research Letters 38 (5): L05805.
4. ASF. 2023. “ALOS PALSAR—Digital Elevation Model (12.5 m spatial resolution)”. Available at: https://asf.alaska.edu/data-sets/derived-data-sets/alos-palsar-rtc/alos-palsar-radiometric-terrain-correction
5. Bentsen, M., I. Bethke, J.B. Debernard, T. Iversen, A. Kirkevåg, Ø. Seland, H. Drange, et al. 2013. “The Norwegian Earth System Model, NorESM1-M.” Geoscientific Model Development 6 (3): 687–720.
6. Bi, D., M. Dix, S. Marsland, S. O’Farrell, H. Rashid, P. Uotila, A.C. Hirst, et al. 2013. “The ACCESS Coupled Model: Description, Control Climate and Evaluation.” Australian Meteorological and Oceanographic Journal 63 (1): 41–64. https://scispace.com/pdf/the-access-coupled-model-description-control-climate-and-32xsky42ip.pdf
7. Boisier, J.P., C. Alvarez-Garretón, J. Cepeda, A. Osses, N. Vásquez, and R. Rondanelli. 2018. “CR2MET: A High-Resolution Precipitation and Temperature Dataset for Hydroclimatic Research in Chile.” EGU General Assembly Conference Abstracts, pp. 19739. Available at: https://ui.adsabs.harvard.edu/abs/2018EGUGA..2019739B/abstract
8. Bravo, C., T. Loriaux, A. Rivera, and B.W. Brock. 2017. “Assessing Glacier Melt Contribution to Streamflow at Universidad Glacier, Central Andes of Chile.” Hydrology and Earth System Sciences 21 (7): 3249–3266. https://doi.org/10.5194/hess-21-3249-2017
9. Brunner, G.W. 2016. HEC-RAS River Analysis System: Hydraulic Reference Manual, Version 5.0. Davis, CA: US Army Corps of Engineers, Hydrologic Engineering Center. Available at: https://www.hec.usace.army.mil/software/hec-ras/documentation/hec-ras%205.0%20reference%20manual.pdf
10. Chow, V.T., D.R. Maidment, and L.W. Mays. 1994. Hidrología Aplicada. Santa Fe de Bogotá, Colombia: McGraw-Hill. Available at: https://dn790003.ca.archive.org/0/items/applied-hydrology-ven-te-chow/Applied%20Hydrology%2C%20Ven%20Te%20Chow_text.pdf
11. Cortés, G., X. Vargas, and J. McPhee. 2011. “Climatic Sensitivity of Streamflow Timing in the Extratropical Western Andes Cordillera.” Journal of Hydrology 405 (1–2): 93–109. https://doi.org/10.1016/j.jhydrol.2011.05.013
12. Cowan, W.L. 1956. “Estimating Hydraulic Roughness Coefficients.” Agricultural Engineering 37 (7): 473–475. Available at: https://ia801801.us.archive.org/10/items/sim_agricultural-engineering_1956-07_37_7/sim_agricultural-engineering_1956-07_37_7.pdf
13. Dabberdt, W.F., A. Crook, C. Mueller, J. Hales, S. Zubrick, W. Krajewski, J.C. Doran, C. Mueller, et al. 2000. “Forecast Issues in the Urban Zone: Report of the 10th Prospectus.” Bulletin of the American Meteorological Society 81 (9): 2047–2064. https://doi.org/10.1175/1520-0477(2000)081%3C2047:FIITUZ%3E2.3.CO;2
14. Dirección General de Aguas (DGA). 1995. Manual de Cálculo de Crecidas y Caudales Mínimos en Cuencas sin Información Fluviométrica. Santiago, Chile: Ministerio de Obras Públicas. In Spanish. https://snia.mop.gob.cl/PIA/handle/20.500.13000/1819
15. Donner, L.J., B.L. Wyman, R.S. Hemler, L.W. Horowitz, Y. Ming, M. Zhao, J-C. Golaz, et al. 2011. “The Dynamical Core, Physical Parameterizations, and Basic Simulation Characteristics of the Atmospheric Component AM3 of the GFDL Global Coupled Model CM3.” Journal of Climate 24 (13): 3484–3519. https://doi.org/10.1175/2011JCLI3955.1
16. Dooge, J. 1973. “Linear Theory of Hydrologic Systems.” Agricultural Research Service, US Department of Agriculture. Tech. Bull. No. 1468. Available at: https://www.tucson.ars.ag.gov/unit/publications/PDFfiles/TB_1468.pdf
17. Dufresne, J-L., M.-A. Foujols, S. Denvil, A. Caubel, O. Marti, O. Aumont, Y. Balkanski, et al. 2013. “Climate Change Projections Using the IPSL-CM5 Earth System Model: From CMIP3 to CMIP5.” Climate Dynamics 40, 2123–2165. https://doi.org/10.1007/s00382-012-1636-1
18. Dunne, J.P., J.G. John, A.J. Adcroft, S.M. Griffies, R.W. Hallberg, E. Shevliakova, R.J. Stouffer, William Cooke, et al. 2012. “GFDL’s ESM2 Global Coupled Climate–Carbon Earth System Models.” Journal of Climate 25 (19): 6646–6665. https://doi.org/10.1175/JCLI-D-11-00560.1
19. Feldman, A.D. 2000. Hydrologic Modeling System HEC-HMS: Technical Reference Manual. Davis, CA: US Army Corps of Engineers, Hydrologic Engineering Center. Available at: https://www.hec.usace.army.mil/software/hec-hms/documentation/HEC-HMS_Technical%20Reference%20Manual_%28CPD-74B%29.pdf
20. Gent, P.R., G. Danabasoglu, L.J. Donner, M.M. Holland, E.C. Hunke, S.R. Jayne, D.M. Lawrence, et al. 2011. “The Community Climate System Model Version 4.” Journal of Climate 24 (19): 4973–4991. https://doi.org/10.1175/2011JCLI4083.1
21. GHD. 2012. “Permit Study for Regularization Works of Natural Watercourses. Tinguiririca River Natural Watercourse, Ancoa – Alto Jahuel 2 x 500 kV Line Project: First Circuit.” In Spanish. 80 p. 
22. Giandotti, M. 1934. “Previsione delle Piene e delle Magre dei Corsi d’Acqua.” Istituto Poligrafico dello Stato 8, 107–117. In Spanish.
23. Giorgetta, M.A., J. Jungclaus, C.H. Reick, S. Legutke, J. Bader, M. Böttinger, V. Brovkin, et al. 2013. “Climate and carbon cycle changes from 1850 to 2100 in MPI-ESM simulations for the Coupled Model Intercomparison Project phase 5.” Journal of Advances in Modeling Earth Systems 5 (3): 572–597. https://doi.org/10.1002/jame.20038
24. Horton, R.E. 1932. “Drainage Basin Characteristics.” Transactions of the American Geophysical Union 13 (1): 350–361. https://doi.org/10.1029/TR013i001p00350
25. IPCC. 2022. “Climate Change 2022: Impacts, Adaptation, and Vulnerability.” Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [H-O. Pörtner, D.C. Roberts, M. Tignor, E.S. Poloczanska, K. Mintenbeck, A. Alegría, M. Craig, S. Langsdorf, S. Löschke, V. Möller, A. Okem, B. Rama (eds.)]. Cambridge University Press. Cambridge, UK and New York, NY, USA, 3056 pp. https://www.ipcc.ch/report/ar6/wg2/
26. Ji, D., L. Wang, J. Feng, Q. Wu, H. Cheng, Q. Zhang, J. Yang, et al. 2014. “Description and Basic Evaluation of BNU-ESM Version 1.” Geoscientific Model Development 7, 2039–2064. https://doi.org/10.5194/gmd-7-2039-2014
27. Kirpich, Z.P. 1940. “Time of Concentration of Small Agricultural Watersheds.” Civil Engineering 10 (6): 362.
28. Knebl, M.R., Z-L. Yang, K. Hutchison, and D.R. Maidment. 2005. “Regional Scale Flood Modeling Using NEXRAD Rainfall, GIS, and HEC-HMS/RAS: A Case Study for the San Antonio River Basin Summer 2002 Storm Event.” Journal of Environmental Management 75 (4): 325–336. https://doi.org/10.1016/j.jenvman.2004.11.024
29. Long, M.C., K. Lindsay, S. Peacock, J.K. Moore, and S.C. Doney. 2013. “Twentieth-Century Oceanic Carbon Uptake and Storage in CESM1 (BGC).” Journal of Climate 26 (18): 6775–6800. https://doi.org/10.1175/JCLI-D-12-00184.1
30. Maidment, D.R. and D. Djokic. 2000. Hydrologic and Hydraulic Modeling Support: With Geographic Information Systems. ESRI, Inc. Redlands, CA.
31. Meyer-Peter, E. and R. Müller. 1948. “Formulas for Bed-Load Transport.” In: IAHSR 2nd Meeting, Stockholm, Appendix 2. International Association for Hydraulic Research.
32. Mintegui, J. and J. Robredo. 1993. Métodos para la Estimación de los Efectos Torrenciales en una Cuenca Hidrográfica. Madrid: Ministerio de Agricultura. In Spanish.
33. Monsalve, G. 1999. Hidrología en la Ingeniería. 2nd ed. Santa Fe de Bogotá, Colombia: Editorial Alfa Omega S.A. In Spanish.
34. Paredes, D.S., E.F. Rivera, P. Baldeón-Egas, and R.M. Toasa. 2025. “The Multiscale Assessment of Infrastructure Vulnerability to River Floods in Andean Areas: A Case Study of the Chibunga River in the Parish of San Luis, Ecuador.” Sustainability 17 (17): 7915. https://doi.org/10.3390/su17177915
35. Peker, İ.B., S. Gülbaz, V. Demir, O. Orhan, and N. Beden. 2024. “Integration of HEC-RAS and HEC-HMS with GIS in Flood Modeling and Flood Hazard Mapping.” Sustainability 16 (3): 1226. https://doi.org/10.3390/su16031226
36. Pizarro, R. 1986. Elementos Técnicos de Hidrología. La Serena, Chile: Corporación Nacional Forestal (CONAF). In Spanish.
37. Pizarro, R., P. Garcia-Chevesich, R. Valdes, F. Dominguez, F. Hossain, P. Ffolliott, C. Olivares, et al. 2013. “Inland Water Bodies in Chile Can Locally Increase Rainfall Intensity.” Journal of Hydrology 481, 56–63. https://doi.org/10.1016/j.jhydrol.2012.12.012
38. Pizarro, R., R. Valdes, P. Garcia-Chevesich, C. Vallejos, C. Sanguesa, C. Morales, et al. 2012. “Latitudinal Analysis of Rainfall Intensity and Mean Annual Precipitation in Chile.” Chilean Journal of Agricultural Research 72 (2): 252–261. https://doi.org/10.4067/S0718-58392012000200014
39. Plata-Rocha, W., E. Avila-Aceves, S.A. Monjardin-Armenta, L.Y. Pérez-Aguilar, J.G. Rangel-Peraza, and Y.G. Zambrano-Medina. 2025. “Urban flood modeling using hydrograph and Rain-on-Grid approaches with HEC-HMS and HEC-RAS: A case study of the Culiacán River.” Journal of South American Earth Sciences 16, 105694. https://doi.org/10.1016/j.jsames.2025.105694
40. Ponce, V.M. 1989. Engineering Hydrology: Principles and Practices. Englewood Cliffs, NJ, Prentice-Hall.
41. Rotstayn, L.D., M.A. Collier, M.R. Dix, Y. Feng, H.B. Gordon, S.P. O'Farrell, I.N. Smith, et al. 2010. “Improved Simulation of Australian Climate and ENSO-Related Variability in a Global Climate Model With an Interactive Aerosol Treatment.” International Journal of Climatology 30 (7): 1067–1088. https://doi.org/10.1002/joc.1952
42. Rowe, P.R., R.L. Thomas, G.A. Tilton, and C.P. Woodin. 1953. California Culvert Practice. Sacramento, CA: Division of Highways, Department of Public Works, State of California, USA.
43. Tari, A.N., M.S. Sepasian, and M.T. Kenari. 2021. “Resilience Assessment and Improvement of Distribution Networks against Extreme Weather Events.” International Journal of Electrical Power & Energy Systems 125, 106414. https://doi.org/10.1016/j.ijepes.2020.106414
44. Taylor, K.E., R.J. Stouffer, and G.A. Meehl. 2012. “An Overview of CMIP5 and the Experiment Design.” Bulletin of the American Meteorological Society 93 (4): 485–498. https://doi.org/10.1175/bams-d-11-00094.1
45. Temez, J.R. 1978. Cálculo Hidrometeorológico de Caudales Máximos en Pequeñas Cuencas Naturales. Madrid: Ministerio de Obras Públicas y Urbanismo (MOPU), Dirección General de Carreteras. In Spanish. Available at: https://www.transportes.gob.es/recursos_mfom/0610400.pdf
46. Thakur, B., R. Parajuli, A. Kalra, S. Ahmad, and R. Gupta. 2017. “Coupling HEC-RAS and HEC-HMS in Precipitation Runoff Modelling and Evaluating Flood Plain Inundation Map.” World Environmental and Water Resources Congress, 240–251. https://doi.org/10.1061/9780784480625.022
47. Thrasher, B.L., E.P. Maurer, C. McKellar, and P.B. Duffy. 2012. “Technical Note: Bias Correcting Climate Model Simulated Daily Temperature Extremes with Quantile Mapping.” Hydrology and Earth System Sciences 16 (9): 3309–3314. https://doi.org/10.5194/hess-16-3309-2012
48. Valdés-Pineda, R., R. Pizarro, J.B. Valdés, J.F. Carrasco, P. García-Chevesich, and C. Olivares. 2016. “Spatio-Temporal Trends of Precipitation, Its Aggressiveness and Concentration, along the Pacific Coast of South America (36–49°S).” Hydrological Sciences Journal 61 (11): 2110–2132. https://doi.org/10.1080/02626667.2015.1085989
49. Valdés-Pineda, R., J.B. Valdés, H.F. Diaz, and R. Pizarro-Tapia. 2015. “Analysis of Spatio-Temporal Changes in Annual and Seasonal Precipitation Variability in South America-Chile and Related Ocean–Atmosphere Circulation Patterns.” International Journal of Climatology 36 (8): 2979–3001. https://doi.org/10.1002/joc.4532
50. Verma, A., S. Kumar, and P.K. Parhi. 2025. “Rainfall Runoff and Flood Plain Inundation Modeling of the Kharkai River, India Using HEC-HMS and HEC-RAS.” Journal of Water Management Modeling 33, C546. https://doi.org/10.14796/JWMM.C546
51. Verni, F. and H. King. 1977. “Estimación de Crecidas en Cuencas no Controladas.” Coloquio, Sociedad Chilena de Ingeniería, 357–374. In Spanish.
52. Violante, M., H. Davani, and S.Dv. Manshadi. 2022. “A Decision Support System to Enhance Electricity Grid Resilience against Flooding Disasters.” Water 14 (16): 2483. https://doi.org/10.3390/w14162483
53. Voldoire, A., E. Sanchez-Gomez, D. Salas y Mélia, B. Decharme, C. Cassou, S. Sénési, S. Valcke, et al. 2013. “The CNRM-CM5.1 Global Climate Model: Description and Basic Evaluation.” Climate Dynamics 40, 2091–2121. https://doi.org/10.1007/s00382-011-1259-y
54. Volodin, E.M., N.A. Dianskii, and A.V. Gusev. 2010. “Simulating present-day climate with the INMCM4.0 coupled model of the atmospheric and oceanic general circulations.” Izvestiya, Atmospheric and Oceanic Physics 46 (4): 414–431. https://doi.org/10.1134/S000143381004002X
55. Wang, Z.A., R. Wei, H. He, and L. Wang. 2024. “Distribution Substation Risk Assessment Under Rainfall Flood Based on HEC-RAS.”  The 9th International Conference on Power and Renewable Energy (ICPRE), 478-482. IEEE.
56. Watanabe, M., T. Suzuki, R. O’ishi, Y. Komuro, S. Watanabe, S. Emori, T. Takemura, M. Chikira, et al. 2010. “Improved Climate Simulation by MIROC5: Mean States, Variability, and Climate Sensitivity.” Journal of Climate 23 (23): 6312–6335. https://doi.org/10.1175/2010JCLI3679.1
57. Wu, T., L. Song, W. Li, Z. Wang, H. Zhang, X. Xin, Y. Zhang, et al. 2014. “An Overview of BCC Climate System Model Development and Application for Climate Change Studies.” Journal of Meteorological Research 28, 34–56. https://doi.org/10.1007/s13351-014-3041-7
58. Yukimoto, S., Y. Adachi, M. Hosaka, T. Sakami, H. Yoshimura, M. Hirabara, T.Y. Tanaka, et al. 2012. “A New Global Climate Model of the Meteorological Research Institute: MRI-CGCM3—Model Description and Basic Performance.” Journal of the Meteorological Society of Japan 90A, 23–64. https://doi.org/10.2151/jmsj.2012-A02
59. Zanchettin, D., A. Rubino, D. Matei, O. Bothe, and J.H. Jungclaus. 2013. “Multidecadal-to-centennial SST variability in the MPI-ESM simulation ensemble for the last millennium.” Climate Dynamics 40, 1301–1318. https://doi.org/10.1007/s00382-012-1361-9
60. Zhao, Y., D. Feng, L. Yu, X. Wang, Y. Chen, Y. Bai, H.J. Hernandez, et al. 2016. “Detailed Dynamic Land Cover Mapping of Chile: Accuracy Improvement by Integrating Multi-Temporal Data.” Remote Sensing of Environment 183, 170–185. https://doi.org/10.1016/j.rse.2016.05.016

Supplementary Material

S1 Site visit

During a site visit in September 2022, it was estimated that the instantaneous flow in the section of interest was 33 m³/s (Figure S1). The Tower 126/1 is protected by a rock riprap revetment flood defense with a height of approximately 4 m from the stream channel bottom to the crest elevation. This protection structure was designed by GHD consultants (GHD 2012) and built by Elecnor Chile S.A. in 2018 considering streamflow for a return period of 100 years. Sediment and gravel deposited as debris from previous floods were also observed during the field visit (Figure 8b). The presence of the sediment allowed defining reference values for the maximum flood elevation to conduct HEC-RAS simulations of both steady and unsteady flow conditions.

Figure S1  (a) Estimated instantaneous streamflow during the field visit in September 2022; (b) Main features of the river defense infrastructure protecting Tower 126/1, including sediments and gravels deposited as debris from previous floods.

Regarding the characteristics of the floodplain area where tower 126/1 is located, the presence of grassland vegetation and thorny shrubs, mainly used as live fences or territorial divisions for cattle grazing, was observed (Figure S2).

Figure S2  Control points for the validation of Manning's roughness coefficients. Point 1 corresponds to rocky riverbed material. Point 2 is an animal grazing zone. Point 3 represents shrubland. Points 4 and 5 correspond to grasslands.

Manning’s “n” coefficients were estimated through Equation S1:

(S1)

Where:

n0	=	base roughness for a straight, uniform, prismatic channel with homogeneous roughness,
n1	=	additional roughness due to surface irregularities of the wetted perimeter along the studied reach,
n2	=	equivalent additional roughness due to the variation in shape and dimensions of the sections along the studied reach,
n3	=	equivalent additional roughness due to existing obstructions in the channel,
n4	=	equivalent additional roughness due to the presence of vegetation, and
m	=	correction factor to incorporate the effect of channel sinuosity or the presence of meanders.

The reference values for each of the roughness parameters considered in the Cowan method (1956) (See supplementary material) were estimated in the simulated section of the Tinguiririca River. To apply the Cowan method, it was necessary to estimate the value of n₀ which was estimated using the Meyer-Peter and Muller method (1948) with the Strickler-type Equation S2:

(S2)

Where:

n0	=	base roughness for a straight, uniform, prismatic channel with homogeneous roughness, and
D90	=	value associated with 90% of the grain size distribution curve constructed based on the sampling of the riverbed material.

Considering the length of the section of interest of approximately 1 km, and the observed channel roughness characteristics estimated during the site visit, the final coefficients for Manning's roughness were established for the mainstream channel bed, the left overbank (LOB), and the right overbank (ROB) following the Cowan method (Cowan 1956) (see Table S1).

Table S1 Estimation of Manning’s coefficients for the HEC-RAS simulation domain.

Surface	n0	n1	n2	n3	n4	m	n Final
Main Channel	0.029*	0.005	0.005	0.005	0.005	1	0.049
ROB - LOB	0.024	0.005	0.005	0.01	0.01	1	0.054

NOTE: ROB: Right overbank; LOB: Left overbank *obtained from Meyer-Peter and Muller (1948).

S2 Climate models

Table (S2) presents the climate models used for hydrological simulations.

Table S2  Climate models used in NEX-GDDP CMIP5.

Id	Model	Institution	Resolution (Lat x Lon)	Reference
1	ACCESS1-1	The Commonwealth Scientific and Industrial Research Organization is an Australian Government agency responsible for scientific research (CSIRO). Bureau of Meteorology Australia	1.875 x 1.25	Bi et al. 2013
2	BCC-CSM1-1	The Commonwealth Scientific and Industrial Research Organization is an Australian Government agency responsible for scientific research (CSIRO) Bureau of Meteorology. Australia	2.8 x 2.8	Wu et al. 2014
3	BNU-ESM	College of Global Change and Earth System Science, Beijing Normal University	2.8 x 2.8	Ji et al. 2014
4	CanESM2	Canadian Centre for Climate. Victoria. Canada	2.8 x 2.8	Arora et al. 2011
5	CESM1-BGC	CMCC-Centro Euro-Mediterraneo per i Cambiamenti Climatici. Bologna, Italy	3.75 x 3.75	Long et al. 2013
6	CNRM-CM5	Centre National de Recherches Meteorologiques, Meteo-France. Toulouse, France	1.41 x l.41	Voldoire et al. 2013
7	CSIRO-Mk3-6-0	CSIRO-QCCCE, Australia	1.875 x 1.875	Rotstayn et al. 2010
8	GFDL-CM3	NOAA. Geophysical Fluid Dynamics Laboratory, USA	2.5 x 2	Donner et al. 2011
9	GFDL-ESM2M	NOAA. Geophysical Fluid Dynamics Laboratory, USA	2.5 x 2	Dunne et al. 2012
10	GFDL-ESM2G	NOAA. Geophysical Fluid Dynamics Laboratory, USA	2.5 x 2	Dunne et al. 2012
11	IPSL-CM5A-LR	Institut Pierre Simon Laplace. Paris, France	1.9 x 3.75	Dufresne et al. 2013
12	IPSL-CM5A-MR	Institut Pierre Simon Laplace. Paris, France	1.25 x 2.5	Dufresne et al. 2013
13	INMCM4	INM, Russia	2 x 1.5	Volodin et al. 2010
14	MIROC5	Japan Agency for Marine-Earth Science and Technology, Japan	1.41 x 1.41	Watanabe el al. 2010
15	MIROC-ESM-CHEM	Japan Agency for Marine-Earth Science and Technology, Japan	2.81 x 2.81	Watanabe et al. 2010
16	MIROC-ESM	Japan Agency for Marine-Earth Science and Technology, Japan	2.81 x 2.81	Watanabe et al. 2010
17	MPI-ESM-LR	Max Planck Institute for Meteorology, Germany	1.875 x 1.875	Zanchettin et al. 2013 / Giorgetta et al. 2013
18	MPI-ESM-MR	Max Planck Institute for Meteorology, Germany	1.875 x 1.875	Zanchettin et al. 2013 /  Giorgetta et al. 2013
19	MRI-CGCM3	Meteorological Research Institute, Tsukuba, Japan	1.125 x 1.125	Yukimoto et al. 2012
20	NorESM1-M.	Norwegian Climate Centre, Norway	2.5 x 1 .875	Bentsen et al. 2013
21	CCSM4	National Center for Atmospheric Research, Colorado, USA	1.25 x 0.94	Gent et al. 2011

S3 River geometry

The geometry of the river included the area of the riverbank protection including tower 126/1 is represented by cross-sections CS-700, CS-680, CS-660, CS-640, CS-620, CS-600, respectively (Figure S3).

Figure S3  (a) Hydraulic characteristics of the geometry in the simulated domain, i.e., hydraulic axis, riverbanks, and main flooding zones; (b) Cross-sections with a 20-m spacing; and (c) Cross-sections identified in the area of the electric transmission infrastructure (from upstream to downstream CS-700, CS-680 CS-660 CS-640, CS-620, CS-600).

S4 Assumptions and limitations of the hydraulic model

The HEC-RAS hydraulic model was used to solve the energy balance equation for the simulated 1D kinematic wave under steady and unsteady flow conditions. Under steady flow conditions, it is assumed that the flow transit and water depths are constant across all cross-sections of the simulated domain. In unsteady flow, the transit is based on an input hydrograph, and the water flow and depth at each cross-section changes over time. The boundary conditions of the model include the main channel bed slope, and a mixed flow regime (subcritical, and supercritical). Considering these boundary conditions, HEC-RAS uses an iterative process to solve the following energy balance equation (Equation S3):

(S3)

Where:

Z1 and Z2	=	channel elevations relative to the datum (L),
Y1 and Y2	=	water depths at each cross-section (L),
V1 and V2	=	average velocities at each cross-section (L/T),
a1 and a2	=	average velocity coefficients (L), and
g	=	gravitational acceleration (L/T2).
he	=	energy losses (L)

Regarding the subdivision of the cross-section for discharge calculation, HEC-RAS implements Manning’s formula using Equation S4:

(S4)

Where:

Q	=	discharge at the cross-section (L3/T),
n	=	Manning roughness coefficient (non-dim),
A	=	cross-sectional area (L2),
R	=	hydraulic radius of the cross-section (L), and
Sf	=	slope of the energy grade line (non-dim).

Table S3 summarizes the key assumptions and limitations associated with the 1D HEC‑RAS configuration applied in this study. These limitations primarily relate to the model’s inability to explicitly resolve lateral flow variability, secondary flow structures, and morphological adjustments during extreme flooding. Despite these simplifications, the 1D approach remains appropriate for the objectives of this study, as the analysis focuses on longitudinal flood wave propagation, peak water levels, and hydraulic energy at the cross‑section where the transmission infrastructure is located.

Table S3 Assumptions and limitations of the 1D HEC-RAS model.

Limitation	Description
Lateral flow variability is simplified	The 1D model represents flow propagation along the main channel axis and does not explicitly resolve lateral flow components across the wide inundation zone.
Complex secondary currents and multi‑thread flows are not simulated	The model cannot capture local recirculation zones, turbulent eddies, or minor distributary threads that may develop across the floodplain during extreme events.
Floodplain storage is approximated geometrically	Overbank storage is included through cross‑section geometry but without representing 2D spatial flow distribution.
Sediment transport and morphological changes are not dynamically modeled	The simulation assumes a fixed bed and does not incorporate sediment mobilization or morphological adjustments that may occur during extreme floods.
Roughness coefficients remain constant	Manning’s n values are estimated but assumed temporally invariant, even though vegetation submergence and sediment deposition may modify roughness during flooding.
Results are cross‑section‑averaged	Water depths, velocities, and hydraulic energies represent section‑averaged values, which may underestimate small‑scale variability near infrastructure elements.