Analysis of the Impact Area of the 2022 El Tejado Ravine Mudflow (Quito, Ecuador) from the Sedimentological and the Published Multimedia Documents Approach

Abstract

Quito (Ecuador) has a history of mudflow events from ravines that pose significant risks to its urban areas. Located close to the Pichincha Volcanic Complex, on 31 January 2022, the northwest and central parts of the city were hit by a mudflow triggered by unusual rainfall in the upper part of the drainage, with 28 fatalities and several properties affected. This research focuses on the affected area from collector overflow to the end, considering sedimentological characteristics and behavior through various urban elements. This study integrates the analysis of videos, images, and sediment deposits to understand the dynamics and impacts of the mudflow using a multidisciplinary approach. The methodology includes verifying multimedia materials using free software alongside the Large-Scale Particle Image Velocimetry (LSPIV) to estimate the kinematic parameters of the mudflow. The affected area, reaching a maximum distance of 3.2 km from the overflow point, was divided into four zones for a detailed analysis, each characterized by its impact level and sediment distribution. Results indicate significant variations in mudflow behavior across different urban areas, influenced by topographical and anthropogenic factors. Multimedia analysis provided insights into the mudflow’s velocity and evolution as it entered urban areas. The study also highlights the role of urban planning and infrastructure in modifying the mudflow’s distribution, particularly in the Northern and Southern Axes of its path, compared with a similar 1975 event, seven times larger than this. It also contributes to understanding urban mudflow events in Quito, offering valuable insights for disaster risk management in similar contexts.

1. Introduction

Mudflow, in a broad sense, refers to the movement of a mixture of sediment particles and other materials in a watery medium. This is a multi-phase component, where the water acts as the controlling force for the movement [1]. The solid phase of the mixture can include sediment grains of various sizes and compositions, as well as vegetation, all of which flow down from higher areas through narrow or channel-confined shapes, such as ravines or valleys. The driving force behind this flow is gravity. The most common trigger for mudflow is heavy rainfall over a high hydrogeological basin [1,2,3].

The energy developed in that kind of event depends on the potential energy (changes in elevation from head to toe of the flow) but always has a destructive power, sometimes very high, and can produce the loss of human lives and damage and modify topography. Of course, when that kind of phenomenon impacts an urbanized area, the losses can increase and cause a disaster [3].

The urban area of the Metropolitan District of Quito (MDQ), the capital of Ecuador, is located in a basin raised at 2800 m.a.s.l., to the west of the so-called Interandean Valley. It was developed in a geological context marked by the presence of the slopes associated with the Pichincha Volcanic Complex (PVC) to the west and a series of hills that are the expression of a system of reverse faults called the Quito Fault System (to the east) [4].

In the western urban limit of the MDQ, about 85 streams are present, fed by the runoff water from the PVC [5]. Throughout the history and development of the city’s urban area, they have been replaced or modified by fills or sewers [6]. The city and the citizens’ relationship with the streams have been complex because they are used as sites for garbage disposal and the accumulation of debris. The lack of public policies or compliance with them has allowed a high degree of exposure in residential areas to phenomena linked to the evolution and functioning of streams, such as mass movements, subsidence, mudflows, and floods, among others. Since the 20th century, the impact of mudflows on sites related to human-intervened drainages has been better documented and reported in Quito’s written press [5]. One of the most remembered is the mudflow or aluvión (used in Ecuador as a local term) of La Gasca (Quito central residential sector related to the whole study area) on 25 February 1975, with an approximate volume of 52,000 m³. It left a thickness of 20 to 50 cm, which mobilized debris, metric-sized blocks, and tree-cutting fragments, resulting in two deaths, the partial destruction of several buildings and vehicles, and damage to roads in the sector [7].

According to Vidal et al. [8], since 1996, the Metropolitan Public Company of Drinking Water and Sanitation of Quito (Spanish acronym: EPMAPS) has been responsible for reducing the risk of floods, landslides, and mudslides through plans and projects that integrate the management of the slopes of the PVC. These actions were carried out with the help of loans from the Inter-American Development Bank (IDB) and allowed for the construction of engineering works to protect river beds. In a report issued by the EPMAPS in 2022, for the event of January 31 of that year, it is noted that between 1997 and 2000, the “Pichincha Hillsides Protection Program” was executed, which built 48 control structures (dams, reservoirs, transfer tunnels, catchment structures) in 33 streams on the Pichincha hillsides [9].

Since 2019, the city began to face again the direct impact of mudflows generated in the intervened drains on the slopes of Pichincha, with the case with the most significant impact due to the number of fatalities (more than 20) being that which occurred on the afternoon from 31 January 2022.

This study aims specifically at the mudflow impact area from an overflooded collector (in the Santa Clara de San Millán, also named as La Comuna sector) through the end of sediment and flood reaching 6 de Diciembre Avenue in downtown Quito city, analyzing the mudflow behavior from the sedimentology point of view and across the media publications (from newspapers, affected people, and the Net).

2. Geographical Location and Previous Knowledge

The urban area of the Metropolitan District of Quito (MDQ) is the capital of the Ecuador Republic and the second most populated city (1.91 million people, INEC, 2017, https://www.ecuadorencifras.gob.ec, accessed on 12 March 2024). It is located in a longitudinal basin of 37 km in length and 6 km wide (on average) and at 2800 m above sea level (m.a.s.l.) in the so-called Interandean Valley (Figure 1A,B).

The geological context was marked by the presence of the PVC (the Guagua and Rucu Pichincha volcanoes) just at the west and close to the center area of the city. The volcanic cones have steeper slopes facing the east and are limited by the east by a series of hills expressing a system of reverse faults.

In the western urban limit of the MDQ, about 85 streams and ravines (locally called quebradas) are fed by runoff water from that Pichincha volcanic complex. Throughout the history and development of the city, they have been replaced or modified by fills or sewers [10,11,12].

Figure 1. Situation map of the study area. Ecuador map and Quito location (A). Metropolitan District of Quito (MDQ) area with urban Quito area (B). Urban studied area and El Tejado Ravine micro-basin relationship (C). Modified from Google Maps [13].

Figure 1. Situation map of the study area. Ecuador map and Quito location (A). Metropolitan District of Quito (MDQ) area with urban Quito area (B). Urban studied area and El Tejado Ravine micro-basin relationship (C). Modified from Google Maps [13].

Figure 1C represents the urban study area, which spreads from the collector in the La Comuna sector (UTM coordinates: 776,779 mE, 9,978,915 mS) through the end of the affected area on the 6 de Diciembre Avenue (UTM coordinates: 779,768 mE, 9,977,639 mS) where the flood and the mudflow reached and has a minimum presence (Distal Zone). That picture also shows the hydrological micro-basin of El Tejado ravine, which is 3080 m long and 560 m wide in the upper area and 90 m, in the nearest area, to the collector (lower area before the water is conducted under the urban area of the city). The area of the micro-basin is 725,000 m² with a 33% inclination on average and two vertical jumps (cascades).

The surficial geology comprises volcanic sediments, such as ashes, pumice, and pyroclastic flows, deposited over andesitic lavas from ancient PVC eruptions. Some natural sections of the surficial ground (sediments over lava basement) can have a thickness of up to 20 m [6,7,8]. The geologic materials are covered by organic soil with low-to-medium-high bushed vegetation in the upper area and allochthonous tree vegetation (eucalyptus) in the lower area (see Figure 2).

On Monday, 31 January 2022, between 6:00 p.m. [14] and 6:30 p.m. [15] local time, a mudflow was reported that overflowed the catchment structure (sewage collector, a system designed to collect and transport the wastewater and also the runoff rainwater from the ravine) at the actual end of the El Tejado ravine. This event had a significant impact on the western urban boundary of the MDQ, particularly its infrastructure (Figure 1 and Figure 3).

The videos from the surveillance cameras of the ECU-911 (the National Integrated Security System) from two intersections, La Gasca Avenue and Francisco Viteri Street and La Gasca Avenue and America Avenue, indicate that the arrival time of the flow front at the América Avenue (an NE-SW main communication axis of the city) was at 6:38 p.m.

According to a published video on the social network TikTok (www.tiktok.com, accessed on 12 April 2023), it is suggested that an hour before the overflow occurred, the clogging of the collector of the El Tejado stream was reported [17].

This event caused damage to people, buildings, public goods, essential services, and road infrastructure. According to the National Risk and Emergency Management Service (Spanish acronym SNGRE), the event caused 28 people to die, one missing person, 52 injured, 170 people affected, 53 families affected, seven buildings destroyed, 41 buildings semi-destroyed, 60 public properties destroyed (poles electricity and garbage containers), 18 public assets affected (Community Police Unit with patrol cars and motorcycles), private vehicles destroyed and affected (vehicles and motorcycles), and the closure of several roads due to the emergency [18].

In the report prepared by the Technical Commission of the Central University of Ecuador for an analysis of the event, it is noted that from Friday, 28 January to Monday, 31 January 2022 heavy rainfall was reported that caused soil saturation and increased runoff. According to data from EPMAPS and Rondal [9,17], 75.6 mm of rainfall in less than one hour was detected, and 145.5 mm of accumulated precipitation in the previous eight days was recorded in the Cruz Loma pluviometry station (upper area of the El Tejado ravine basin), which is a highly abnormal value in Quito rains.

Water infiltration caused landslides on the channel’s slopes at the middle and upper areas of the El Tejado ravine [19]. According to the Special Meteorological Bulletin of 31 January (obtained online the same day from https://www.inamhi.gob.ec/informacion-en-linea/; INAMHI government institute), it was noted that the rainfall in January 2022 was 32% greater than that of January 2021 and was 31% above the average values for a traditional January [14].

The study area is located in Quito’s north-central and western sectors, in the Belisario Quevedo urban parish (Figure 4). It includes the sectors called La Comuna and La Gasca and focuses on the streets and avenues affected by the direct impact of the flood of 31 January 2022, that is, from the pluvial sewage collector construction on Fulgencio Araujo Street to 6 de Diciembre Avenue (in the NW-SE direction) and from Núñez de Bonilla Street to Indoamérica Square (in the N-S direction).

The urban neighborhoods affected by the 2022 event are La Comuna, Pambachupa, La Gasca, La Colón, Santa Clara, and La Mariscal. It had a severe affectation on sectors of La Comuna and La Gasca in the upper part, causing the loss of lives and personal and real property, and high affectation in La Gasca and Pambachupa, causing the accumulation of mud and debris on the main avenues and intersections. Less impact was caused in Santa Clara and La Mariscal sectors (América, Cristóbal Colón, Amazonas and 10 de Agosto Avenues, and Alonso de Mercadillo, General Ulpiano Páez, Luis Cordero, 9 de Octubre, and Diego de Almagro Streets), causing the accumulation of sludge and clogging of the sewage system [14].

3. Methodology

The methodology starts with the observation and cartography of the impact and affected sectors defining the flooded area and the urban elements involved (curbs, sidewalks, speed-breakers, street furniture, and street inclination). Also, information about the sedimentological characteristics of the deposited material (thickness and geological classification and composition) was taken. Later, graphical information (pictures and videos) published on the Net from the area’s neighborhood and affected people was collected, and some testimonials were provided in the sectors. The Interchangeable Image File Format (EXIF) metadata was verified to prevent fake news and old or fake graphical material, which allows for the filtering and discharging of out-of-time or non-useful news.

The soil laboratory was the last investigation phase, which used a granulometric analysis and binocular observations.

3.1. Video, Image, and Visual Media Approach

The methodology used in this study was based on multimedia material (images and videos) related to the mudflow. The information obtained, ordered, and categorized (coding) is a compilation of images and videos from social networks such as Facebook (www.facebook.com), YouTube (www.youtube.com), TikTok (www.tiktok.com), Instagram (www.instagram.com), and Twitter/X (www.twitter.com), digital press (television channels, and digital newspapers and the ECU-911) [20].

The treatment of the information was divided into a detailed analysis of the videos and images to determine their spatial location (using fixed reference objects, see an example in Figure 5A,B) and temporal location (from four videos with the time shown; see

Table 1. Temporal location of videos and their position. Modified from Cañar [20].

Used Code	Address	Local Time	Length
80-CJB	José Berrutieta Street (corner)	18:09:10 p.m.	13 s
122-N24CNB	Núñez de Bonilla Avenue crossing García Dec	18:29:10 p.m.	4 min
15-GA	La Gasca Avenue crossing Francisco Viteri Street	18:39:37 p.m.	52 s
39-GA	La Gasca Avenue crossing América Avenue	18:39:50 p.m.	49 s

and Figure 5C).

In the second stage, the kinematic parameters of the mudflow were defined through a detailed field survey of the areas selected from the videos, and these parameters were calculated using the Fudaa-LSPIV software (v. 1.9.2) [21].

Two hundred and sixty-one multimedia files were handled: 151 videos were in MP4 format, and 110 images were in JPG format throughout the affected area.

The multimedia file sources were searched using free access software tools: InVID-Project (https://www.invid-project.eu/, accessed on 15 September 2023) and Google Lens (https://lens.google/intl/es-419/, accessed on 15 September 2023), which allows for a determination of the different websites where the multimedia records of the event were uploaded and shared. That procedure aims to eliminate (debug) the fake videos from the database (those that are not actual or of the event), validate them, and gain accuracy.

In calculating the kinematic parameters (surface velocities, average velocity, and caudal flow), methodology was used based on the study between specific points of the digital video frames of the debris flows [22,23]. The developed technique allows for a precise evaluation of the dynamic characteristics of debris flows, thus facilitating the understanding of their behavior and potential impact.

It determines the spatial (two-dimensional) coordinates of selected features and objects from pixels identified in video images, as shown in Figure 5A,B. The average velocity of the flow surface between two video frames is then calculated as the ratio of the distance traveled to the elapsed time, according to the adapted velocity for video frames (Equation (1)):

$$v=\frac{d·V_{fot}}{N_{fot}}$$

(1)

where v represents the surface speed (m/s), d denotes the distance traveled (m), V_fot stands for the recording speed of frames (s²), and N_fot is the number of frames in the video [22].

Furthermore, the LSPIV (Large Scale Particle Image Velocimetry) technique was applied for good-quality videos of recordings in along-channel flow events (see Figure 6) [24,25,26]. Due to the development of software and programming, the technique applies to videos recorded from fixed digital cameras. Recently, it has been adapted for the implementation to videos filmed by the public and shared by different social networks, obtaining good results [24].

In Arattano and Grattoni [22], the technique is applied to videos recorded with instrumented cameras, with minimal disturbances of the video frames. However, the videos of the mudflow of the El Tejado ravine present sudden movements and disturbances in the frames. Most of the videos are amateur, that is, recorded from cell phones by residents of the affected sectors or surveillance cameras, and, therefore, we have low-resolution videos that present slight movements of the camera to different sides. That generated a slight distortion of the frames and blurring of the recorded scene, which made its analysis difficult and presented a certain degree of uncertainty in the results. Instead, it can be considered that it has obtained a good result.

3.2. Sedimentological Approach

Between the dates 1 February and 22 April 2022, four field trips were carried out in the areas affected by the 31 January 2022 mudflow in the following sectors: La Comuna, La Gasca, Pambachupa Park, and 10 de América, 10 de Agosto, Cristobal Colón, and 6 de Diciembre Avenues, and also on the streets that cross these Avenues.

Data on deposit and sediment thickness, flood footprint, extension, and flow characteristics were recorded. Data on the slopes of the streets were also recorded to understand how this factor, added to the presence of different urban structures, such as sidewalks, speed breakers, or flower beds, influenced the flow behavior.

For the textural and compositional analysis, three samples were used (see Figure 4 for localization): M1 taken on Núñez de Bonilla Street (UTM: 777,773.4 mE, 9,978,480.4 mS), M2 collected on Alonso de Mercadillo Street (UTM: 778,525.3 mE, 9,977,973.4 mS), and M3 in the Atacames Passage (UTM: 777,121.0 mE, 9,978,310.5 mS). It can be considered that M1 and M2 samples were obtained from sediment stuck on walls, and the M3 sample was recollected from fresh sediment.

Once the samples were homogenized and quartered, they were sieved to obtain a granulometric distribution of the components [27]. The histograms and statistical parameters were created, which include the mean, median, mode, standard deviation, skewness coefficient, and kurtosis.

Using a binocular magnifying glass, the material retained on the sieves was again quartered for visual recognition. The components were grouped into four main classes and subclasses (

Table 2. Classes and subclasses for mudflow sedimentary components classification. Rondal [17].

Class	Subclass	Reference Image
Organic material	Branches, leaves, bark and tree fragments
Anthropic material	Concrete, brick, paper, plastic, coal
Lithic fragments	Light gray lithic fragments
	Dark gray lithic fragments
	Reddish lithic fragments
	Pumice
Mineral fragments	Amphibole
	Plagioclase
	Volcanic glass
	Magnetite
	Quartz
Undefined	Aggregates with plastic

). Additionally, the characteristics of the clasts were recorded, such as the roundness, sphericity, shape, and composition.

4. Results

The observation and the cartography of the sediments spread in the affected area determined that the mudflow traveled a distance of 3.2 km from the La Comuna sector (sewage collector area) to 6 de Diciembre Avenue (in a straight-line considered) [19].

It changed its direction to every crossing street due to the urban design of streets and the urban elements (shape of the street sections and the constructed parts), maintaining an NW-SE main flow meaning (see Figure 5 and Figure 7).

Regarding the multimedia material, only 65 videos were found to document the mudflow’s development, identified and georeferenced from the total number of video recordings available. However, only four videos had a time record with enough accuracy to establish the start time and progress (see Table 1). Even so, these times have a high degree of uncertainty given that there needs to be more information on the accuracy of the equipment that recorded the event (Figure 8) [20].

By considering the behavior of the flow, it was divided into four zones for this analysis (Figure 7):

• The Direct Impact Zone (collector area and surroundings);
• Northen Axis (along the Núñez de Bonilla Street);
• The Southern Axis (along the La Gasca Avenue);
• The Distal Area (between América and 6 de Diciembre avenues);

The North and South axes, with their unique sections, exhibit different behaviors influenced by conditions, such as flow dynamics and deposition materials. However, they are pretty different and result from anthropic elements’ influence (urban waste containers and other materials, street design and structure, and sections), direction, and general topographical inclination. It can be noticed that the Pambachupa Park area plays a special role in the distribution of the mudflow along the Northern Axis and made some changes in the distribution of the flood, which eroded the North Axis during its development. In contrast, the Southern Axis presents a more straight-line shape and fewer constrained conditions.

4.1. Direct Impact Zone

The Direct Impact area corresponds to the La Comuna sector, where the flow occupies the streets Fulgencio Araujo, Padre Semanate, and José Berrutieta before entering the Belisario Quevedo volleyball sports field (Figure 9). At the intersection of José Berrutieta Street with Antonio de Herrera Street, the flow was split into two branches.

That flow presented a turbulent behavior, and those zones were considered an erosional area due to the street slopes, between 5° and 12°. The observed deposit had a maximum thickness of 1 cm, and the flood footprint reached 2 m on building walls.

The analysis of the multimedia images and footage determined that only two videos were helpful for the analysis of the flow behavior [20]. The entrance of the mudflow to the urban area through José Berrutieta Street could be identified, and it was characterized as a continuous flow with a high sediment load and high energy (Figure 10A). The estimated surface velocity was 4 to 5 m/s based on mapping specific frame points.

It must be emphasized that the flow showed a high carrying capacity, evidenced by the damage to the buildings in the sector and the height of the flood footprint [17]. A change in flow behavior was also determined by the increase in speed and energy caused by its passage through a tunnel under Mariscal Sucre Avenue (approximately 6.0 m in diameter; Figure 10B). That was evidenced in the dragged and transported vehicles, tree trunks (up to 3 m long), mesh fences, and 1.5 m-in-size rubble.

According to the Municipality of Quito [14], in the evaluation of affected properties before the flow entering the tunnel, it was established that six buildings were still habitable homes; two non-habitable properties on the first floor and one property had damage to the structure and other structural pathologies (Figure 10B). The three homes most affected were those that received the direct impact of the flow upon entering the street. Meanwhile, to the east of the tunnel (after the passage of the flow), the properties were classified as three habitable homes and four non-habitable homes on the first floor; that is, even though the home did not present structural damage, the first floor was affected by the entry of flow and debris.

4.2. Northen Axis

This axis follows a straight path according to the input direction, and upon reaching the volleyball sports field, the flow destroyed its structures and carried away about 50 people [28]. Before that area, the mudflow had been divided into the branches mentioned: the one that continued along N24C Street (the north side) and the other along José Berrutieta Street until it reached La Gasca Avenue (south side). The Northern Axis (Figure 11) spanned from the volleyball sports field, along N24C and Núñez de Bonilla main streets (NW-SE direction), until it reached América Avenue, where it ends. It also flowed through the transversal streets: Francisco Lizarazu, Domingo Espinar, and Fernández de Recalde (mentioned those from west to east) [19].

In this section, the flow presented a different behavior than the previous flow and the Southern Axis because, to the east of the volleyball sports field, there was a 35° inclination slope and 8 m height from the first step to the next level down. Under the volley field, there was a stepped waste dump with material of an unconsolidated urban origin (presence of construction remains, pipes, plastics, and others) of 2100 m² (~6000 m³ in volume) approximately (Figure 12). That slope caused a cascade effect where the direct impact of the flow on the deposit has an erosive process that allowed the Northern Axis flow to be enriched with that material (sediments, debris and garbage) [19].

According to Rondal [17], on N24C Street, a sedimentary deposit with an average thickness of 7 cm and 40 to 20 cm was found on the ground floors of the buildings located east of the waste dump that were directly impacted. Reference [20] points out that on this street, the flow had a high-energy turbulent and erosive flow behavior that destroyed the structures (enclosing walls) of several buildings and showed a high load capacity by dragging cars, tree trunks up to 2.5 m, and debris from 10 cm to 200 cm in diameter. An analysis of several frames estimated a surface speed between 10 and 11 m/s.

The damage in the east area of the waste dump is summarized as homes with total collapse, a high degree of structural damage, and non-habitable homes on the street elevation ground [14]. Therefore, the waste dump and the change in the main slope (10° inclination) influenced the variation in flow behavior (higher speed and energy) that determined a high degree of impact on the homes with a greater exposure (by the closeness) to this diverse material accumulation area.

To the east of N24C street, the flow entered the Pambachupa Park, a linear area of 440 m long and 30 m wide, approximately (Figure 11), characterized by green fields on stepped terraces with a main inclination of 4° towards the southeast. The presence of two buildings (a communal house and a community police unit) inside the park generated artificial barriers where the flow collided. It allowed a part of the flow to enter through transversal streets from the park (Domingo Espinar and Fernández de Recalde streets). It was determined that the entrance through Francisco Lizarazu Street (at the upper head of the park) was the product of an inclination towards the south (Figure 13). According to Cañar [20], the surface speed, calculated from the surveillance camera frames on Domingo Espinar Street (where it has a 1° to 2° inclination), was less than 2 m/s. In addition, the mudflow flowed with a dense behavior able to transport trees larger than 1 m and some cars.

The flow that advanced through the transversal streets ended at La Gasca Avenue, feeding the flow that advanced along the Southern Axis.

The arrival time of the flow to the area located at approximately 1.05 km from the clogged collector was 6:29 p.m., obtained from an analysis of the video of a surveillance camera located in Núñez de Bonilla Street, in the east of Pambachupa Park (UTM: 7770811.9E, 9978725.6S). That local time is highly uncertain due to the lack of information about its calibration.

Furthermore, it was established that the flow behavior was waved. It was characterized by a main front and smaller secondary waves, which can still carry medium-sized tree trunks, cars, and garbage containers. Using the Fudaa-LSPIV 1.9.2 software [21], an average velocity of 3.28 m/s and a flow rate of 4.89 m³/s were calculated.

The behavior of waves was evidenced again through videos up to 1.3 km from the collector, in an area with a 2.5° inclination, where the deposited material showed a thickness of 5 cm associated with organic material (trunks, branches up to 30 cm and different kind of chips), garbage containers, and unclassified garbage.

4.3. Southern Axis

The Southern Axis was defined as spreading along José Berrutieta Street through La Gasca and América Avenues, entering the transversal streets of Atacames, Francisco Javier Lizarazu, and Romualdo Navarro. That flow fed a new split branch that crossed from west to east along Diego Zorrilla Street for approximately 500 m. This branch constitutes the southern boundary of the affected area and is located approximately 400 m wide to the south of the Northern Axis, being the most widespread (Figure 14).

In the area where the principal axes were divided, it could be observed that the flow moved south along José Berrutieta Street, forming a kind of barricade of tree trunks and vehicles that prevented its entry into a residential complex, the Colinas de La Gasca urbanization (Figure 15), and continued its journey with heading southeast, along the same street, until it reached La Gasca Avenue.

A video recorded by a private security camera located on José Berrutieta Street (UTM: 776,953.9 mE, 9,978,729.3 mS) indicates that the start of the event south of the volleyball sports field and prior to the formation of the barricade was at 6:09 p.m. (also with a high degree of uncertainty because there is no information about the camera’s time calibration). In addition, it can be seen that the flow advance has a high energy, evidenced by the towing capacity of several vehicles (see Figure 9 and Figure 15A,B), which was part of the barricade at the entrance to the habitational complex, creating a dam-wall that limited the access of the flow through Antonio Herrera Street.

It can be observed that on La Gasca Avenue, the flow still had a high load capacity. The videos and pictures show the transporting of cars, garbage containers, several sizes of debris, and tree trunks of up to 2.5 m. The average thickness of the deposit along that avenue was 5 cm. In this area, the wave’s behavior was changed because the largest fraction of material was deposited (up to 3.0 m). At the same time, the aqueous flow and finest parts overflowed the accumulated material, generating a new wave-flux phenomenon [17].

Additionally, it is noted that the footprint of the maximum flood left on the walls of La Gasca Avenue was approximately 35 cm. Cañar [20] points out that the analysis of the videos allowed the flow to be characterized by arrhythmic (non-periodic) waves in which the front, the body, and the tail of laminar layers could be identified. The surface velocity calculated with the support of the video frames was 5 to 6 m/s in an area 430 m away from the flow entrance through La Gasca Avenue (where it has a slope of 5°). Recording footage obtained from the ECU-911 security cameras, approximately 730 m from the top entrance of the flow to La Gasca Avenue, indicates a record of the event passing at 6:39 p.m. local time (the front of the flow is not recorded). The time from that kind of records can be considered with less uncertainty than others (official cameras against private ones). In this video record, intermittent waves are observed in all areas (3° inclination), and the analysis of the frames allowed us to estimate a speed of 3 to 4 m/s.

The flow reached the intersection with América Avenue, approximately 1.2 km from the entrance to La Gasca Avenue, at 6:38 p.m. (local time). During 30 min, the flow developed with an average of four waves per minute [17]. At the intersection with América Avenue, it was observed that the deposit had a thickness of 5.6 cm and was made up of organic matter (trunks up to 1 m long and branches) and garbage with the lithic sedimentary deposit made up of sand.

4.4. Distal Area

The end of Southern Axis flow was América Avenue, and from there, it continues to advance along Alonso de Mercadillo Street just now in the Distal Area. That also affected Cristóbal Colón Avenue and the transversal streets (all in NW-SE direction) until the main avenues, such as 10 de Agosto and 6 de Diciembre Avenues, which have NE-SW directions (Figure 16). During the information gathering on 2 February 2022, it was determined that the deposit in this area was made up of primary and remobilized material (lithic sediments) as a product of the artificial washing of La Gasca Avenue (for removing the mud).

The average thickness was 5 cm, except for 10 de Agosto Avenue and its intersection with Alonso de Mercadillo Street, where a large amount of material was accumulated. It covered a total area of 2937 m² with an average thickness of 54.9 cm [17]. The largest accumulation area was located near Santa Clara de Millán Park, approximately 2.0 km from the clogged collector (Direct Impact zone), and in a plain area with no slope (0°). The deposit comprises sandy silt with organic matter (branches, chips, and 50.0 cm pieces of tree trunks) and garbage.

That change in thickness sedimentation behavior is because, in this area, there was a change in slope from 5° that gradually passes to 0° and because of the influence of vial structures (street steps and sidewalks approximately 30 cm high from street pavement elevation). Those conditions acted as barriers to the advancement of the mud flow towards the east and became deposit areas for most sediment. In addition, this caused the flow to be driven to 10 de Agosto Avenue and advanced through Luis Cordero, Alonso de Mercadillo, and Ignacio de Veintimilla streets. The most extended branch is the one located to the north, over Luis Cordero Street, which, in its advance, entered through Cristóbal Colón Avenue until the intersection with 6 de Diciembre Avenue. That was 3.2 km from the collector.

According to the analysis of the multimedia material, the flow advances through the Distal Zone as a flood of water with a low sediment load, speed, and energy [20].

The furthest deposit was identified approximately 3.2 km from the collector on Cristóbal Colón Avenue, which was made up of fine sand and had a lesser presence of organic matter (tree bark, leaves, small branches, and vegetation) and garbage. The deposit was less than 1 cm thick (Figure 17).

After defining the impacted surface, it was considered that the total area of the deposit in the affected zones was 100,000.9 m². That value was used to calculate the total volume of the mudflow deposit considering the variation in thickness all over the different branches according to the inclination of the streets and the flow behavior (erosive or depositional phase). In summary, the thickness on the Northern Axis was 1 to 40 cm, the Southern Axis was between 5 and 10 cm, and the Distal Zone was between 1 to 5 cm, with a special accumulation zone of 54.9 cm around the Santa Clara de Millán Park sector. Therefore, the deposit volume was estimated at 7163 m³, but this was not considered the sediment in suspension, driven by the flux running into the sewers and drainage areas below the street levels [17].

Also, it was considered that the flood footprint on walls and buildings presented average values of 90 cm in the direct impact zone and an average of 57 cm in the Northern Axis. On the Southern Axis, it had an average value of 19 cm, and in the Distal Zone, it reached an average value of 9 cm. From those values, it was calculated that the flow volume of the whole aqueous mass and sediments was about 46,219 m³ and reached and covered a total area of 122,790.7 m².

4.5. Sedimentological Patterns

The result of the mechanical sieving of the three collected samples is shown in Table 3. The differences in sample M3 are due to the previous treatment that the sample had for the granulometry analysis via hydrometry.

Table 3. Results of the sieving of the collected samples (see a graphic representation in Figure 18 and their location in Figure 4).

Sieve #	M1 Sample	M2 Sample	M3 Sample
	Retained Weight (gr)	Retained Weight (gr)	Retained Weight (gr)
4	5.40	18.76	-
10	21.61	25.80	9.03
20	41.82	34.25	-
40	77.18	69.86	26.8
60	60.10	60.35	-
100	62.88	80.22	-
200	45.62	51.34	53.02
End tail	82.50	82.54	-
Total weight	397.11	423.12	88.85

# Number of sieve elements of American Standard. The last file highlights the total values of every sample.

Table 3. Results of the sieving of the collected samples (see a graphic representation in Figure 18 and their location in Figure 4).

Sieve #	M1 Sample	M2 Sample	M3 Sample
	Retained Weight (gr)	Retained Weight (gr)	Retained Weight (gr)
4	5.40	18.76	-
10	21.61	25.80	9.03
20	41.82	34.25	-
40	77.18	69.86	26.8
60	60.10	60.35	-
100	62.88	80.22	-
200	45.62	51.34	53.02
End tail	82.50	82.54	-
Total weight	397.11	423.12	88.85

# Number of sieve elements of American Standard. The last file highlights the total values of every sample.

Figure 18. Graphic representation of the granulometric analysis (see Table 3).

Figure 18. Graphic representation of the granulometric analysis (see Table 3).

Sample M1 (collected on Núñez de Bonilla Street on the Northern Axis, approximately 1 km from the clogged collector) comprised 8.6% fine gravel (composed of the undefined aggregates) and 91.4% sand. Its central tendency parameters were a mode of 1.66, median of 1.45, and mean of 1.45, showing a poor draw (1.52), bias towards coarse components (−0.19), and mesokurtic components (1.05). The histogram presents a unimodal distribution with the highest frequency (26.01%) in the medium sand particles (1 and 2 Φ).

Sample M2 (collected on Alonso de Mercadillo Street in the Distal area, approximately 1.9 km from the collector) comprised 13.1% fine gravel (composed of the undefined aggregates) and 86.9% sand. The central tendency parameters were a mode of 2.37, mean of 1.57, and median of 1.22, with a poor draw (1.67) and bias towards thicker and mesokurtic components (1.07). The histogram shows a unimodal distribution (29.81) in fine sand particles (between 2 and 3 Φ).

Sample M3 (collected in the Atacames Passage on the Southern Axis, less than 500 m from the collector) comprised 54% sand and 46% silt, corresponding to silty sand. This sample was the only one taken wet, and the drying and sieving processes before carrying out the granulometric analysis destroyed the gravel-sized aggregates found in samples M1 and M2 [17]. The measures of central tendency were a mode of 2.53, mean of 1.42, and median of 1.71; poor draw (1.52); bias towards the coarsest particles (−0.16); and very platykurtic components (0.65). The histogram shows a unimodal distribution most frequently in sand size (1.3 and 3.9 Φ). The sample has 34.64% coarse sand (0.1 Φ) and 65.36% fine sand (2.5 Φ).

In summary, the material from the waste dump with unconsolidated anthropic deposits located east of the field contaminated the material of samples M1 and M2. Sample M3 was obtained south of the Atacames Passage in the Ritter Park area without the influence of material from the waste dump, so it did not present evidence of contamination, such as anthropic material or aggregates of silty material [17].

The samples generally contained 15.5% organic components, 2.5% anthropic components, 55.0% lithic fragments, and 27.0% mineral fragments. The lithological component comprised pumice clasts, rock fragments of an andesitic and dacitic composition, and mineral fragments of quartz, plagioclase, amphibole, and magnetite. The fine gravel-sized aggregates comprised silty material, including plastic remains (Figure 19A). The cohesive nature of these materials favored the formation of clots or flocs due to Van Der Waals forces [31].

5. Discussion

The analyzed mudflow rises from the El Tejado ravine micro-basin after a short period of heavy rainfall, where a high downsized inclination and soft overburden soil are presented. These factors increased the risk of floods and landslides. They could apport significant amounts of sediments that hit challenging-manner urban areas, such as the La Comuna and downtown sectors in Quito [32].

The flow behavior analysis has been presented as a variable event, starting from the collector and then along N24C, Núñez de Bonilla streets, and the upper and middle part of La Gasca Avenue as a turbulent regime all around the western area. In these areas, it reached a speed of up to 11 m/s, and the flow was also characterized by presenting successive waving behavior phases. That energy decreased towards the eastern zone and was degraded by the intersections between the transversal streets and by the changes in slope (reaching the distal zone where the slope is 0°). The speeds observed in the final zone were less than 4 m/s [32,33,34].

The analyzed mudflow on 31 January 2022 had characteristics similar to the 1975 mudflow event that spread along the same axis, La Gasca Avenue (Figure 20A,B). These include the extreme meteorological conditions due to heavy rains in a short period, the erosion of drainage at high-elevation areas of the hydrographic basin, and the occurrence of small landslides on the edges of the stream. The information about that previous event is limited to a geological note by Feininger [7] and without scientific data or investigations. It also had short information from newspapers and their pictures, where the big boulders that reached the mid part of La Gasca Avenue can be seen (Figure 20C).

Compared to the 1975 event, a strong anthropic influence altered the flow of 2022 due to the presence of limiting structural elements, such as the tunnel and the presence of new constructions and living areas (volleyball court, park, and waste dump) and the constructions on the edge of the Broken. Those elements did not exist in the case of the 1975 flow because the Mariscal Sucre Avenue over the tunnel and the same tunnel had still not been started to be built, and the urban pressure was less significant [7].

With these values and taking into consideration the classification of Cruden and Varnes [35], the mud flow is a small, high-speed event with a water content of 45.67% and a density of 1547 gr/cm³, which classifies it as a flow with enough water to behave like a liquid and a hyper-concentrated flow according to Hungr et al. [36] (obtained from M3 sample).

The sediment load has been high in the area of higher elevations, as well as when the flow crosses the waste dump area. In the distal area, this flow behaves as a water compound with fine sediments because the coarser components are deposited in other points, such as the Santa Clara de Millán park area [37].

The stream and drainage area in the 1975 event was in a wild state with native vegetation (shrub and herbaceous) and without eucalyptus plantations and woody trees [7]. In the flow of 2022, the drainage area and the basin of El Tejado were occupied, for the most part, by non-native forest exploitation plantations on its slopes and even near the riverbed (eucalyptus). It also had a significant load of garbage and anthropogenic remains (especially construction) deposited uncontrolled by the city’s residents. This garbage-accumulation process is believed to have been enhanced during the COVID-19 pandemic period (almost two years of restrictions were suffered in Ecuador) as there was no regular garbage collection [37].

The city’s design in the area that was crossed by the flow (direction and inclination of the streets, above all) determined the behavior of the flow and, therefore, the areas of impact and most intense affectation, as indicated in Karkani et al. [37]. The layout of the streets and the waterproof materials that cover them (asphalting and concrete) condition the direction and behavior of the flow [11,12,32,33,37]. For example, the slopes of the main streets and their intersections (avenues) have been fundamental in the flood’s path. Also, some of the buildings and destroyed elements were involved in the sediment. That can be prevented using reinforcement in the structures, as indicated by Lou et al. [38].

The sedimentological characteristics of the flow (pumice, andesites, and dacites) could be assimilated to a lahar type of secondary origin due to both the existing relationship with the drainage area (slopes of the Pichincha volcano) and its rheological behavior [35,36]. However, it must be highlighted that it is highly anthropogenic-influenced drainage, especially from the interventions of the EPMAPS [16] action projects in the stream (flood control) and the urban occupation of the stream’s banks in the collection environment and entrance to the collector. That has resulted in the appearance of a complex mass flow composed of rock blocks, fine lithic material, plant remains, and debris (of anthropic origin). It is unknown if the sewer collector is the correct size, or it can be due to the current climatic conditions and the drainage volume from the hydrographic basin, as commented by Karkani et al. [37] and Sierra [39].

It has been evident that the presence of plastics and microplastics (detected sizes of 8 to 25 mm) that were accumulated as garbage in the ravine before the event has generated agglomerates (named here as aggregates) of gravel-size to soft small boulders (>10 mm) that cannot be classified in known geological terms. For this reason, they have been classified as undifferentiated (see Table 2).

The analysis of the basin requires further and deeper investigations due to the interventions carried out in the area’s drainage since 1970 for the construction of roads and streets (for example, the Mariscal Sucre Avenue), increasing the occupation (population) of the area as residential, and the construction of the cable-car towards PVC.

Social networks have become fundamental tools in this research due to the variety of points of view shown in images and ideas and because of the access to them (free of charge). With their contribution, it has been possible to complete and complement the data from the official recordings (those belonging to the municipality), which have been scarce (only four control and reliability cameras were available) [28,29,30].

6. Conclusions and Recommendations

The obtained data determine that the 2022 flow constitutes a mass-flow movement of a small magnitude, and compared to the previous 1975 flow described by Feininger [7], it is seven times smaller, according to data published in the media at the time (we consider that it can be an overestimation). However, the structural damage and loss of life were more significant on this occasion.

Anthropic conditions have conditioned both the trajectory and behavior of this 2022 flow. On the one hand, the collapse at the entrance of the collector (by tree trunks and stone blocks up to 3 m in diameter and 7 tons in weight [14]) prevented the accumulating water and sediments from being conducted correctly and, on the other hand, the new constructions and the use of the soil in the environment of the sewage collector and the surroundings. The less-intervened environment in the 1975 mudflow allowed stone blocks to reach the middle of La Gasca Avenue [7]. It is worth highlighting the importance they have had in the behavior of the flow of the volleyball court and the waste dump, as well as their staggered layout (terracing). In these cases, erosion was generated, enriching the incoming flow with sediments.

The characteristics of the 2022 flow determine the need to define new parameters and conditions of these events in urban or urbanized environments because these areas do not adapt to the parameters of natural environments and their known behaviors. For example, in urban environments, there is a need to include a new kind of grain conforming to a plastic nucleus that has never been seen before in Ecuador and without other scientific references.

In addition, land use and territorial planning must consider the occurrence of these extraordinary events to avoid creating spaces with high human exposure (volleyball court) and the placement of elements that influence the phenomenon’s behavior (slag heap and garbage). Streets and buildings constrain the mudflow shape and do not have a typical fan shape of this phenomenon.

The consequences of the pandemic (deposition of garbage in the stream) and the lack of sanitation of the collectors and natural drainage areas concerning this type of special event must be evaluated in the medium term. Also, managing the drainage area is a valuable need [40]. A future insight is to work on defining the micro-basin of El Tejado ravine characteristics and hydrogeological conditions.

The possible expiration of works that have exceeded the return periods of their construction and whether they respond to the demands related to climate change and the city’s urban development must also be analyzed [41,42].

Moreover, the different recordings and images consulted are low-to-medium quality and available through the network. However, education of the population could be considered so that these graphic documents are recorded with the best possible quality for later use in this type of assessment and improvements in the municipal video surveillance system.