Detailed geomorphology of debris avalanches of El Estribo volcanic complex (Central Mexico)

ABSTRACT

The El Estribo Volcanic Complex, located to the south of Pátzcuaro Lake (Central Mexico), forms an elevation based on a shield volcano crowned by a cinder cone. Two debris avalanches, dated at 28,000 and 14,000 ybp, cover an area of 4 km² with a typical hummocky topography. The zone is a state natural protected area with no previous studies of hazard and risk from mass movement processes. Herein, we present a detailed geomorphological map of the debris-avalanche area, scale 1:20,000. The approach applied used two hierarchical levels, geomorphological landscapes and landforms. The fault scarp was mapped using high-resolution digital elevation models obtained whit unmanned aerial vehicles (UAV) survey. We describe and characterized four main geomorphological units, 20 subunits, 66 hummocks, and the general drainage network. This map is a valuable tool to identify and quantify risks from mass movement processes.

KEYWORDS:

• Mass movement
• unmanned aerial vehicles (UAV)
• hummocks
• Pátzcuaro Lake
• Michoacán–Guanajuato volcanic field
• Trans-Mexican Volcanic Belt

1. Introduction

Geomorphological maps are basic inputs for terrain understanding and serve as the starting point of specific applications such as landslide hazard and risk assessment linked to mass movement processes (Alcalá-Reygosa et al., 2016; Devoto et al., 2012; Echeverría Arnedo, 1997; Loibl & Lehmkuhl, 2013; Quesada-Román & Zamorano-Orozco, 2019a; Tripodo et al., 2012). Our work presents a detailed geomorphological map of the El Estribo volcanic complex (Figure 1) – a regionally important tourist destination inhabited by more than 7900 people (INEGI, 2011). This study belongs to an ongoing geomorphological inventory aimed to identify hazards related to gravitational movements and to evaluate risks in the area. El Estribo is located at the southern part of the Pátzcuaro Lake, an area covered by circa 45 volcanoes (Osorio-Ocampo et al., 2018). The lake has been the site of lacustrine sedimentation since Miocene (Isradé-Alcántara et al., 2005) with suitable conditions for animals (Robles-Camacho et al., 2010) and human occupation (Osorio-Ocampo et al., 2018). Lacustrine sedimentation has been disrupted by tectonic and volcanic processes in the past (Garduño-Monroy et al., 2009a; Garduño-Monroy et al., 2011). Past earthquakes have triggered landslides into the Pátzcuaro Lake such as those recorded for the El Metate, Isla de Janitzio (Garduño-Monroy et al., 2011) and El Estribo volcanoes (Pola et al., 2014a). Such landslides surely unleashed tsunamis of unknown dimensions in the lake. Historical earthquakes have been recorded in 1845 and 1858 in this region (Isradé-Alcántara et al., 2005). The Pátzcuaro Lake and its homonymous city are tourist attractions in the State of Michoacán, both visited by more than 100,000 people a year (Bocco et al., 1999). Therefore, there is an urgent need to produce a detailed geomorphological map of the El Estribo Volcanic Complex and its outcrops of debris-avalanche deposits exposed on the southern shore of the lake. This map will be useful for local authorities responsible for mapping landslide and hazard susceptibility, which is crucial for land-use planning in the area.

Figure 1. Up shows perspective view from the eastern scarp of El Estribo, Pátzcuaro Lake and the island of Janitzio (scoria cone). Down shows El Estribo cinder cone as seen from south (photographs of Gemma Gómez-Castillo).

2. Geographic, geological and tectonic contexts

The El Estribo Volcanic Complex is located within the Trans-Mexican Volcanic Belt (TMVB), an active continental volcanic arc that runs across central Mexico (Figure 2(a)). The TMVB extends 1000 km from the states of Nayarit in the Pacific coast to Veracruz in the Gulf of Mexico (Demant, 1979). It is formed by large stratovolcanoes, calderas and monogenetic volcanoes in certain areas named ‘volcanic fields’ (Macías, 2005). The Michoacán–Guanajuato Volcanic Field (MGVF) is located in the central western part of the TMVB (Hasenaka & Carmichael, 1985), comprises nearly 900 cinder cones, 100 small structures (cones, domes, thick lava flows, etc.), and over 300 medium size volcanoes, but lacks of active ones (Hasenaka, 1994; Hasenaka et al., 1994) (Figure 2(b)). Several volcanoes with partial collapse of their flanks occur within the TMVB (Capra et al., 2002), one of which is the Late Pleistocene El Estribo Volcanic Complex, located south of Pátzcuaro Lake (Pola et al., 2014a; Figure 2(c)). El Estribo is made up of a basal shield volcano (dated at 126,000 ± 7000 ybp) that is crowned by a cinder cone separated by a paleosol dated at 28,360 ± 170 ybp. The shield volcano is cut by an E–W normal fault exposing its basal diameter and 200 m of piled-up lava flows that doesn't affect the cinder cone above. This fault scarp is associated to a basal and an upper debris-avalanche deposits of different ages and an exposed hummocky topography placed north of the volcanic structure and along the southern shore of the Pátzcuaro Lake (Pola et al., 2014a, 2015).

Figure 2. (a) Location of the study area within the Trans-Mexican Volcanic Belt (TMVB), (b) the Michoacán-Guanajuato Volcanic Field (MGVF), (c) Location of the study area to the south of Pátzcuaro Lake. TMVB limits were taken from Ferrari et al. (2012) and MGVF from Mazzarini et al. (2010). White thin lines indicate state boundaries.

Study area is located in the municipality of Pátzcuaro in the State of Michoacán, covering an area of 35.4 km² which includes the El Estribo Volcanic Complex, small parts of the city of Pátzcuaro and the homonymous lake. Average annual precipitation in the area is 943 mm (Figure 3(a)), being July the rainiest month (Figure 3(b)) (CONAGUA, 2017).

Figure 3. Study area precipitation over a 43 year period between 1969 and 2016. (a) Annual precipitation. (b) Average monthly precipitation. Data from CONAGUA (2017), weather station 16,087 ‘Pátzcuaro’, located at 19.516389° latitude, −101.609722° longitude.

The central part of the TMVB is disrupted by the Morelia-Acambay Fault System (MAFS) that has a predominantly extensional regime with E–W normal faults and NNW-SSE directions (Ferrari et al., 2012; Garduño-Monroy et al., 2009b; Isradé-Alcántara et al., 2005; Pasquarè et al., 1991; Suter et al., 1992). In the Pátzcuaro area, MAFS is represented by the E–W Jarácuaro and Pátzcuaro faults that form a semi-graben with a N–S extension regime (Osorio-Ocampo et al., 2018; Pola et al., 2014b). Paleoseismological studies in the area suggest that severe earthquakes have hit the area during the Holocene (Isradé-Alcántara et al., 2005).

3. Materials and methods

The geomorphological map at a 1:20,000 scale of the El Estribo Volcanic Complex was made using the following inputs: 2008 panchromatic SPOT 5 satellite images at a 5 m spatial resolution, a 1:20,000 scale topographic vector digital data (key E14A22D) of Instituto Nacional de Estadística y Geografía (INEGI) produced in 2011, aerial photographs at a 0.1 m spatial resolution obtained with drone flights made between 2015 and 2016 and field inspections in which the distribution of the landforms and the general morphodynamic processes were observed. The applied approach used two hierarchical levels: geomorphological landscapes and landforms (Bocco et al., 1999, 2001; de Boer, 1992; Zinck, 2016).

First, a digital terrain model (DTM) was made from 10-m-equidistant contour lines (INEGI, 2011), then the DTM was used as input in ArcMap to produce a hillshade model, an elevation map, and a slope map. Next, using ArcMap for calculations and measurements, the slope map was reclassified into seven new ranges representing the potential mass movement according to Aceves-Quesada et al. (2014), which was used to make a zonation of landforms relative to the process of vertical erosion, and a relief energy map based on 0.04 km² quadrants – which expresses the maximum elevation change – use to describe the potential of the denudation and relief evolution processes, and the presence of geological structures and threats (Lugo-Hubp, 1988; Quesada-Román & Barrantes-Castillo, 2017). These products supported the interpretation of geomorphological landscapes and landforms following the methodology of Francioni et al. (2019), Miccadei et al. (2012), Miccadei et al. (2013) and Ventura et al. (2005). Hummocks were recognized and identified visually using the hillshade model.

A Phantom 2 Vision Plus (P2V+) and a Phantom 3 Standard (P3S) used for mapping were flown at heights between 100–120 meters above take-off point. Ten flights were completed with P3S (2117 photos), and four with P2V+ (933 photos), acquiring totally 3050 images to map the scarp area. Flights were done at an average speed of 10 m/s (36 km/h), between August 2015 and November 2016. DJI Vision, Pix4DCapture (P2V+), and MapsMadeEasy (P3S) apps were used to achieve homogeneous overlap (80% front, 75% side). Manually flights were also conducted with DJI GO app (P3S). Agisoft PhotoScan Professional (2016) was used to process the information, and a high-resolution DTM at a 0.1 m spatial resolution was generated from a classified dense cloud.

All this information was verified during geological and geomorphological field surveys and the piloting of UAVs. Two different working scales were used for the interpretation of units and hummocks delimitation. In the scarp area, a 1: 500 scale was used with a 4 m² minimum mappable unit because of DTM high resolution (10 cm/pixel). Remaining areas were mapped at 1:5000 scale using a minimum mappable unit of 400 m². The geological data (ages, lithology, volcanic stratigraphy, etc.) were taken from Osorio-Ocampo et al. (2018) and Pola et al. (2014a, 2015). Finally, we designed the geomorphological map by correlating the geomorphological structures with the geology and by establishing categories and subcategories that were easy to identify and interpret in relatively small areas.

4. Results

The main geomorphologic features and characteristics of El Estribo Volcanic Complex are presented in the geomorphological map, such as the main morphometric information (elevation, relief energy, slope), potential mass movement processes and landform distribution. This information was incorporated into six main sections, which are described in the following paragraphs:

4.1. Elevation map

The altimetry analysis of the area identified patterns of the spatial distribution and characteristics of altitudinal contrasts (Quesada-Román & Zamorano-Orozco, 2019b). The elevation map of Figure 4(a) displays the changes in the elevation of the relief. In this map, the highest point corresponds to the El Estribo cinder cone at 2451 m above sea level (hereafter all elevations are given in meters above sea level) and the lowest the Pátzcuaro Lake at 2040m. The largest area includes elevations between 2040 and 2080m, which represents 54.5% of the study area, and mostly corresponds to the lacustrine plain (Figure 4(a)), followed by elevations between 2080 and 2280 m (42.6%), morphologically corresponding to gently hillslope, fault scarp, and upper piedmont, and the smaller area corresponds to elevations above 2280 m (2.9%) coinciding with the cinder cone and some high parts of the lava flows.

Figure 4. Thematic maps of the study area: (a) Elevation, represented by stretch values along a color ramp, four standard deviations types, (b) Relief energy, grid of 200 × 200 meters, (c) Slope and (d) Potential mass movement.

4.2. Relief energy map

The relief energy map was obtained by a spatial search of the maximal and minimal elevations every 40,000 m² followed by the subtraction of the second from the first values, thus finding the maximal difference in elevation within a 200 × 200 m (0.04 km²) grid. A raster map was obtained from the continuous values of the relief energy expressed in m/0.04 km². The map shows that the highest values (of nearly 170 m) are located along the fault scarp and the cinder cone, and the lowest, on the lacustrine plain (Figure 4(b)).

4.3. Slope map

Terrain inclination is one of the most important factors of mass movement processes (Aceves Quesada et al., 2014; Alcalá-Reygosa et al., 2016; Guzzetti et al., 2012; Quesada-Román & Zamorano-Orozco, 2019a; Shrestha & Poudel, 2018) because water erosion has a higher impact in abrupt slopes, which may lead to the acceleration of the process in such areas. The slope map was obtained in ArcMap by algebraic and trigonometric operations using the contour lines to obtain the terrain's inclination degree. The maps show that the predominant slope values on the lacustrine plain vary between 0° and 2°. In the very gentle hillsides surrounding the plains, a value between 2.1° and 6° is dominant. In the upper foothills and gentle slopes that value ranges between 6° and 20°, and in some mountainsides and the fault scarp inclinations are above 30°. Finally, in parts of the cinder cone, the slopes range between 20° and 45°, higher values being only present in the walls of formerly exploited sand and gravel pits (Figure 4(c)).

4.4. Potential mass movement map

The potential mass movement map is aimed to zoning the landforms in relation with gravitational processes (Aceves Quesada et al., 2013, 2014; Shroder Jr. et al., 2011; van Beek et al., 2008). This map reclassifies the slope map into the following seven ranks according to Aceves Quesada et al. (2014): Null movement, from 0° to 1.5°; very weak movement, from 1.5° to 3°; weak movement, from 3° to 6°; moderate to high movement, from 6° to 12°; high movement, from 12° to 20°; very high movement from 20° to 45°; and intense movement in inclinations above 45°. Most of the lacustrine plain shows null to very weak potential mass movement values while in the fault scarp the potential mass movements are very high and, in some areas, intense (Figure 4(d)).

4.5. Distribution and characteristics of hummocks

These two debris avalanches and hummocks were previously described (Pola et al. 2014a, 2014b, 2015), however, the authors did not provide detailed estimations of the volume. Thus, the volume (measured in relation to the nearest contour line), convex summit area, and length and orientation of major axes were estimated for the 66 hummocks identified in the debris-avalanche area. In both debris’ avalanches, the volume of hummocks is approximately 184 million cubic meters. Assuming that the original hillside was placed about 300 m to the north of its present location (Pola et al., 2014a) and reconstructing the scarp contour line, also called morphoisohypses (Lugo-Hubp, 1988; Siebe et al., 1992), we estimated that the volume of the collapsed material was 228 million cubic meters. The purple dashed lines in Figure 5 indicate the limits of the basal and upper debris avalanches. Most of the hummocks of both avalanches have an elongated form and only a few are conical. Figure 5 also shows some trends in the distribution patterns, for example, the largest hummocks originated from the older (basal) and more extensive debris-avalanche deposit – mostly with ample and flattened summits product of their greater age and more prolonged erosion – tend to be distributed towards the outside of the deposit and have a predominantly NE-SW orientation. Instead, the medium and small-sized hummocks are distributed in the center of the deposit and their long axes are variably oriented from east to west. In the case of the younger (upper) debris-avalanche deposit, hummocks have, in general, sharper, more elongate, and narrow summits than in the older avalanche. These hummocks tend to be of low height and small volume in the proximity to the fault scarp, their height and volume increase with distance from the source of the debris up to hummock 43. The orientation of their major axis varies widely in an E–W direction. Table 1 summarizes the values of each hummock's morphological characteristics.

Figure 5. Location, distribution, volume, and length of major axis of hummocks in the study area. The numbers next to the hummocks correspond to their identifiers (Id) in Table 1.

Table 1. Morphological characteristics of hummocks.

Id	Area (m^2)	Perimeter (m)	Height (masl)	Volume (m^3)	Superficial area (m^2)	Length (major axis m)	Strike (°)	Id	Area (m^2)	Perimeter (m)	Height (masl)	Volume (m^3)	Superficial area (m^2)	Length (majoraxis m)	Strike (°)
1	26,307	686	2060	561,247	27,413	297	40	34	5047	274	2050	50,466	5047	108	104
2	15,315	517	2080	250,471	15,592	225	58	35	70,224	1030	2070	373,818	64,571	404	68
3	1411	197	2100	14,644	1418	90	70	36	38,820	793	2060	817,708	40,075	334	97
4	604	122	2100	5977	604	54	83	37	58,624	1019	2070	861,639	55,426	416	104
5	160	48	2100	1537	160	19	93	38	6030	292	2060	56,025	6045	111	74
6	270	66	2100	2682	270	23	64	39	36,161	760	2060	239,233	30,385	301	78
7	31,340	806	2090	433,243	31,537	338	84	40	27,483	674	2060	264,783	27,527	243	89
8	5191	337	2090	53,318	5196	150	71	41	2543	187	2050	25,196	2543	72	148
9	9812	430	2100	128,451	9930	182	69	42	22,512	794	2060	133,726	15,557	357	92
10	24,000	669	2100	375,874	22,845	293	85	43	70,784	1121	2060	1,138,125	72,237	460	87
11	5484	321	2080	8839	5414	129	79	44	16,363	623	2040	259,627	16,809	255	76
12	3981	265	2080	40,796	3987	115	62	45	4908	275	2040	56,469	4923	107	91
13	21,650	606	2090	179,567	21,759	258	64	46	9199	357	2040	150,656	9468	133	100
14	26,892	727	2090	306,580	27,818	312	77	47	31,785	788	2050	302,962	32,399	323	68
15	5266	285	2080	42,677	5352	114	77	48	4412	246	2040	83,369	4465	90	59
16	7370	311	2080	112,660	7590	109	62	49	7633	456	2040	120,810	7741	204	68
17	4269	304	2080	42,702	4270	133	96	50	13,218	446	2040	132,179	13,218	181	108
18	27,457	686	2090	259,805	27,559	283	78	51	77,749	1329	2040	1,054,439	79,080	581	83
19	5489	290	2080	57,180	5492	111	80	52	4952	298	2040	48,307	4952	125	70
20	52,799	914	2080	1,115,579	53,699	376	87	53	94,761	1372	2050	1,179,205	95,875	551	55
21	22,280	617	2060	238,132	22,394	259	99	54	17,363	529	2040	277,623	17,618	194	82
22	4491	272	2060	46,100	4561	115	111	55	5620	275	2040	56,202	5620	98	67
23	11,183	428	2050	88,739	11,295	168	103	56	8927	350	2040	89,270	8927	132	94
24	4036	283	2070	40,139	4036	121	68	57	8386	338	2040	124,095	8770	127	48
25	18,963	542	2070	366,684	19,153	222	99	58	14,730	514	2040	216,482	15,485	219	85
26	3180	210	2070	31,935	3181	82	84	59	31,773	682	2040	473,373	32,215	250	102
27	2923	205	2070	25,702	2947	73	105	60	12,113	461	2040	119,144	12,114	188	96
28	26,614	663	2070	385,145	27,099	277	97	61	6880	331	2040	67,692	6882	130	84
29	38,009	974	2070	644,855	40,667	435	82	62	6750	325	2040	6192	6750	131	287
30	59,916	1159	2070	854,743	62,725	481	80	63	3949	230	2040	37,997	3953	82	74
31	38,644	819	2070	380,831	38,739	276	84	64	10,375	396	2040	150,701	10,894	158	294
32	75,788	1064	2060	945,979	71,307	394	90	65	57,795	917	2040	268,379	58,132	335	57
33	69,004	1056	2060	1,101,558	73,009	426	94	66	135,308	1609	2040	1,067,844	136,610	669	44

Notes: Id: Identifier of hummock. Area (m²): The area in square meters of each hummock calculated in two dimensions. Perimeter (m): Length in meters of the hummocks’ contour. Height: The reference contour line value in meters from which the hummocks’ volumes were calculated. Volume (m³): The space in cubic meters occupied by each hummock calculated from the next lower contour line (Height). Superficial area (m²): The surface area in square meters of each hummock calculated in three dimensions. Length (major axis m): The length in meters of the major axis of each hummock. Strike (°): Direction or bearing relative to true north in degrees of the major axis of each hummock.

4.6. Geomorphological map

The primary landforms in the study area are the product of volcanic and tectonic activities. The erosion of hillsides and accumulation are the result of the runoff and mass movement exogenous processes. The map characterizes the morphogenic categories in a hierarchical classification (Bocco et al., 2001; Cisneros-Máximo, 2016; Zinck, 2016), from which the following major units shown in the map (Main Map) were defined: (I) Highlands; (II) Hills; (III) Piedmont, and (IV) Plains. These units were further subdivided into 18 relief subunits.

• I . – Highlands. It includes seven subunits distributed in lava flows and the fault scarp (beige and brown colors). The subunits are 1 – Very gently slopes (Lvg), 2 – Gently slopes (Lgs), 3 – Anthropogenic rubble dumps (Lrd), 4. – Steep slopes (Lss), 5 – Lava front (Llf), 6 – Hillslope shoulders (Lhs), and 7 – Lava flow top (Lft). Subunits 1–3 are located near the fault scarp and are mainly made up of andesitic lavas from the shield volcano (Pola et al., 2014a). Subunits 4–6 are morphologically compound of the three portions of a lava flow in the eastern area.

• II . – Hills. Includes four subunits of the El Estribo cinder cone (peach color). The first subunit (Hcb) represents the plain in the bottom of the crater composed of alluvial deposits (material remobilized from the crater walls), pyroclastic surges, ash and lapilli falls, and impacts of basaltic andesite blocks. The second subunit (Hvg) corresponds to the very gentle slopes surrounding the crater forming nearly flat surfaces originated by phreatomagmatic explosions that partially collapsed the main crater (Pola et al., 2014a, 2015). These areas are currently planted with maize. The third subunit (His) is the inner slope of the volcanic cone characterized by a slope opposite to the external part of the cone and the same geological composition of the first and second subunits. The fourth subunit (Hos) is the external hillslope of the volcanic cone compound of the pyroclastic fall deposits of ash and dark-gray blocks. The latter deposits are overlaid by remobilized volcaniclastic deposits with sedimentary structures.

• III. Piedmont: Made up of four subunits distributed in the area of the upper debris avalanche and some areas close to slope changes (pink color). The first subunit (Phu) is the hummocks product of the two debris avalanches deposits geologically integrated of light-gray andesitic angular blocks with jigsaw puzzle structure supported by a fine to coarse ash matrix (Pola et al., 2014a). The second subunit (Pif) is the inter-hummocky plain that geologically belongs to the upper debris-avalanche deposit made up of a layer of alluvium and little consolidated lava blocks within a fine to coarse sand matrix (Pola et al., 2014a). The third subunit (Plp) is the lower piedmont located in the lateral gently slopes of the fault scarp. Its geology is diverse and depends on the location. The fourth subunit (Pup) is the upper piedmont to the south of the fault scarp with variable rock compositions.

• IV. Plains: It is divided into three subunits distributed near the lake and in the southern area of the cinder cone. The first subunit (Plf) is the flood plain adjacent to the water surface. It is composed of fine lacustrine sediments, has very low inclinations. Although it is susceptible to flooding, it has remained dry since 1977, in fact, the lake's level has tended to become lower over time (O’Hara, 1993). The second subunit (Plh) is the currently not flooded lacustrine plain, but that has been intermittently flooded during the nineteenth century (O’Hara, 1993). It is also predominantly composed of lacustrine sediments and, in the basal debris-avalanche area, is composed of poorly consolidated gray to reddish massive lava blocks. The third subunit (Pll) corresponds to the inter-lavic plain located to the south between the cinder cone and the lava flows. Geologically, it is made up of three types of rocks, from west to east, the basaltic lavas, the little inclined lava flows from the shield volcano, and the volcaniclastic sediments over which the city of Pátzcuaro is settled.

Besides the above-mentioned subunits, categories were established for a gully (G) and the formerly operating sand and gravel pits (Op). The geological column for the ages has been taken from Cohen et al. (2013).

The use of UAVs allowed identifying other smaller previously unrecognized structures (Figure 6(a)). The use of these tools can be of great help if high-resolution DTMs are unavailable (Cook, 2017; Tanteri et al., 2017) but it requires software and hardware to quickly analyze the aerial photographs, as well as mixed methods to classify the photogrammetrically generated dense cloud. Utilization of UAVs can become a multitemporal analytic tool if it is intended to monitor an area (Hackney & Clayton, 2015; Yu et al., 2017). Its use also eased the identification of human activities and land uses, and the comparison of satellite images from previous years (Figure 6(b)).

Figure 6. Mosaics of high-resolution images obtained from drone flights. The areas outside of the red polygon are references of the contrast in resolution of the information previously available for the study area. (a) Hillshade model; and (b) Rectified orthomosaic.

5. Summary and conclusions

Hierarchical geomorphological mapping has been applied to ecological regionalization (Avers et al., 1993; Bailey, 1987; Bailey et al., 1985, 1994; Bocco et al., 1999; Cleland et al., 1997; Forman, 2014; Fu et al., 2004; Klijn et al., 1994), land-use planning (Bathrellos et al., 2012; Bocco et al., 2001; López-Blanco & Villers-Ruiz, 1995), ecological restoration (Kondolf, 1995; Palik et al., 2000), river and hydrological basin management (Frissell et al., 1986; Gurnell et al., 2016; Hawkins et al., 1993; Lehotský, 2004; Newson & Newson, 2000; Parsons et al., 2003; Thomson et al., 2001), and risk analysis (Kayastha et al., 2013; Piacentini et al., 2015; Sinha et al., 2008; Yu et al., 2017), to mention only some of its applications. In our case, the geomorphological map of the El Estribo Volcanic Complex allowed to analyze a volcanic zone with a very complex tectonic and seismic activity that have triggered two landslides during the last 28,000 ybp (Pola et al., 2014a).

Our methodology and resulting map (Mai Map) are useful to analyze with new lights the morphological characteristics of hummocks of both debris-avalanche deposits, to differentiate useful features, and to understand their distribution and link them to their origin. In fact, the orientation of the major axis of hummocks indicates the transversal direction of the moving avalanche during the collapse of the rock mass (Glicken, 1991; Siebert, 1984; Ui, 1983). Our map did not detect a unique pattern in the orientation of the major axis of hummocks for neither of the two debris avalanches, because the superposition of the two deposits made this task difficult. Authors like Dufresne and Davies (2009), Siebe et al. (1992) and Siebert (1996) have explained this phenomenon as a rotation in the horizontal plane taking place as the material is being emplaced. The concentration of bigger hummocks in the far side of the source of the avalanches (Urandén de Morelos Island, Id 66) could be explained by the lack of physical barriers along the path of the avalanche, to the lubricating action of the water body allowing to the avalanche to move a longer distance (De Blasio et al., 2004; Mazzanti & De Blasio, 2010; Mohrig et al., 1998) and, to the ‘bulking’ process that incorporates material at the front of the avalanche as it displaces forward (Scott et al., 2001; Siebe et al., 1992; Siebert, 1996).

Elaborating a detailed geomorphological map eases the understanding of the endogenous and exogenous processes taking place within a specific area; therefore, it is a valuable tool to aid the identification and quantification of the risks from mass movement processes (Bentivenga & Piccarreta, 2016). In addition, the quality of geomorphological maps can be increased by the use of UAVs to obtain orthophotomosaics and high-resolution DTMs allowing supplementing the information and verifying it in the field (Cook, 2017; Francioni et al., 2018). The methodology applied in this work provides a quick and effective way to identify the main morphostructural features, which might facilitate its use for land planning and for managing and mitigating risks, minimizing the time and cost needed for that. Our review of the literature showed that our geomorphological map (Main Map) is one of the few to be published on a volcanic structure at such high resolution (10 cm/pixel) of inputs and analysis, and therefore, we consider it an excellent tool to analyze the geomorphological risk.

Finally, the combination of geomorphology and geology offers a broader perspective of the behavior and functioning of a volcanic area and its associated tectonic processes. Also, the hierarchical approach used for the geomorphological mapping can be applied at different scales having the advantage of being replicable in similar areas, even those having a larger surface.

Software

Esri ArcMap 10.4 software was used for georeferencing, digitizing, visualizing orthorectified aerial photographs or remote sensing images, for spatial and statistical analysis, and for Layout map design.

Acknowledgements

Thanks are due to CONACyT for the granted PhD scholarship and to the ‘Evaluación de la importancia relativa de bosque húmedo de niebla bajo un enfoque de paisaje’ project. We thank Antonio Pola, Susana Osorio, Laura García, Jaime Paneque, Luis Miguel Morales and Paz Coba for their participation in the field work. Manuel Mendoza acknowledges the sabbatical stay scholarship provided by the PASPA-UNAM and to Dr. Rick Giardino from Texas A&M University for commenting a previous version of this article. This peace of research is part of the Interinstitutional Lab of Hazard and Risk Evaluation of UNAM.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Manuel Mendoza acknowledges PASPA-UNAM [grant number 0179386] and Texas A&M University for its partial support. Nicolás Vargas-Ramírez acknowledges the support of UNAM's Postgraduate Program in Geography as well as the fellowship for his postgraduate studies (CVU/grant holder number: 710901/420676) received from the National Council of Science and Technology of Mexico (CONACYT). This work was also supported by PASPA UNAM [grant number 1410DFA].