Hydrography and geomorphology of Antananarivo High City (Madagascar)

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tjom20

Journal of Maps

ISSN: (Print) 1744-5647 (Online) Journal homepage: https://www.tandfonline.com/loi/tjom20

Hydrography and geomorphology of Antananarivo

High City (Madagascar)

William Frodella, Daniele Spizzichino, Andrea Ciampalini, Claudio Margottini

& Nicola Casagli

To cite this article: William Frodella, Daniele Spizzichino, Andrea Ciampalini, Claudio Margottini & Nicola Casagli (2020): Hydrography and geomorphology of Antananarivo High City (Madagascar), Journal of Maps, DOI: 10.1080/17445647.2020.1721343

To link to this article: https://doi.org/10.1080/17445647.2020.1721343

© 2020 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group on behalf of Journal of Maps View supplementary material

Published online: 13 Feb 2020.

Submit your article to this journal

Article views: 294

View related articles

Hydrography and geomorphology of Antananarivo High City (Madagascar)

a

Earth Sciences Department, University of Firenze, Florence, Italy;bISPRA, Geological Survey of Italy, Roma, Italy;cEarth Sciences Department, University of Pisa, Pisa, Italy

ABSTRACT

The city of Antananarivo is located in the central highlands of Madagascar, and is the largest urban center of the island. Considering the frequent heavy cyclonic rains frequently affecting the area, its geomorphological context is particularly prone to geo-hydrological hazards, such as landslides andflash floods, as recently proved during the disastrous events of the winters of 2015 and 2018. Field data and high-resolution remote sensing data interpretation (DEMs and satellite orthophotos) were combined in order to produce detailed hydrographic and geomorphological maps. The aim was to understand the processes acting in the Analamanga hill area, with special regards to the effect of human activity in modeling the natural landforms and exacerbating the geo-hydrological hazards. The obtained maps will provide management-planning tools to be used as a first step towards a risk reduction strategy in the Antananarivo historical urban center.

ARTICLE HISTORY Received 14 August 2019 Revised 20 January 2020 Accepted 22 January 2020 KEYWORDS Geomorphological mapping; GIS; remote sensing; cultural heritage; landslides

1. Introduction

Antananarivo is the capital city of Madagascar and is located in the central highlands region (18.55′ South; 47.32′ East), about 160 km from the east coast and 330 km from the west coast, at approximately 1200 m a.s.l. above sea level (Figure 1). Antananarivo was called Analamanga (the ‘blue forest’), until 1610, when the merina King Andrianjaka built hisfirst woo-den royal palace complex (the so-called Rova) and a fortified village on the flat hilltop of the highest relief of the area (Analamanga hill); its privileged position assured an ideal defense and observation post on the lowermost Ikopa river plane (Ciampalini, Frodella, Margottini, & Casagli, 2019) (Figure 2). Thefirst core of the city (the High City) further developed from the hilltop on the hillslopes (Middle town), by the late nineteenth century reaching the river valley during the colonial period (Low town). During the latter dec-ades, the city has continued to evolve to nearly 1.7 million inhabitants today, covering an urban area of approximately 86.4 km2 (namely ‘The Great Antana-narivo’,Figure 1(b)).

Currently, the High City represents an important cultural heritage site of Madagascar, encompassing the new stone brick-built Rova and Chapel, the high dignitaries’ buildings, such as the baroque-style archi-tecture Andafiavaratra palace, the Cathedrals of Ando-halo and Ambohipotsy, built by thefirst missionaries in the second half of nineteenth century (Figure 3). For these reasons, the High City is currently included

in the UNESCO World Heritage tentative list. Detailed thematic maps are of basic importance for the study of landslide processes (Calista et al., 2016; Frodella, Morelli, Fidolini, Pazzi, & Fanti, 2014;Smith & Ellison, 1999) and for analysing the landform evolution during the Quaternary (De Muro, Ibba, Simeone, Buosi, & Brambilla, 2017; Karymbalis, Papanastassiou, Gaki-Papanastassiou, Tsanakas, & Maroukian, 2013; Pucci et al., 2015), with special regard to the interactions with human activity (Paliaga, Luino, Turconi, & Faccini, 2018; Roccati, Faccini, Luino, Ciampalini, & Turconi, 2019; Visser, 2014). In this framework, the integration with remote sensing technologies can over-come the limitations of afield approach, by allowing a complete coverage of the analyzed phenomena over wide and inaccessible areas reducing costs and ensur-ing the safety of thefield operators (Bardi et al., 2017; Ciampalini et al., 2019; Del Soldato et al., 2018; Frodella, Gigli, Morelli, Lombardi, & Casagli, 2017a; Gigli et al., 2014). The protection of Cultural Heritage from instability phenomena requires a specific inter-disciplinary approach, which should be planned con-sidering the geological and geomorphological charac-teristics of the site, as well as the typology of the related hazard (Margottini, Fidolini, Iadanza, Trigila, & Ubelmann, 2015a, 2015b; Nolesini, Frodella, Bianchini, & Casagli, 2016; Pastonchi et al., 2018). The rapid urban development of Antananarivo, together with the lack of a proper urban planning, due to political and governance issues, have caused several environmental problems, such as intense

© 2020 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group on behalf of Journal of Maps

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

CONTACT William Frodella william.frodella@unifi.it Earth Sciences Department, University of Firenze, Via La Pira 4, 50121 Florence, Italy This article has been republished with minor changes. These changes do not impact the academic content of the article.

Figure 1.Study area location: regional setting (a), the city urban setting (b).

Figure 2.Lithography and Historical pictures showing the urban development of Antananarivo (courtesy of Musee de la Photo, Antananarivo): (a) The High city in the end of the 16thcentury; (b) vegetation cover and rock outcrops along the Analamanga hill slopes in the end of the nineteenth century. Pictures of the modern city configuration: south-eastern slope (c) and southwestern slope (d).

deforestation and quarrying, unauthorized slope cut-ting and terracing for the construction of illegal hovels, moreover, there is a general lack of proper drainage and sewer systems. This context makes the Anala-manga hill area particularly prone to soil erosion, land-slide andflash flooding phenomena (Ciampalini et al., 2019), especially during the frequent heavy cyclonic rain events affecting the central highlands (Table 1). In this context, in March 2015 Antananarivo was hit by the tropical storm Chedza, resulting in severe flood-ing in the river plain area of the Low city, and in wide-spread shallow landslides along the hillslopes of the Middle Town. This event caused diffuse structural damage affecting several quarters, also in the Upper town’s historical buildings, including 20 casualties and an estimate of 36,000 evacuees. Rock falls also occurred in the winter of 2018, affecting the buildings of the Middle town’s depressed neighborhoods, unfor-tunately causing several casualties. Following these events, a detailedfield mapping campaign was carried out (October 2017 and May–June 2019), in order to

understand the geological, geomorphological and hydrographic features of the Analamanga hill area. The surveys also enabled the identification of possible new UNESCO Core and Buffer zones. The outcomes of this activity are presented in this work by showing detailed geological and hydrographical maps, which constitute an update of the work ofCiampalini et al. (2019). Thefinal purpose was to improve the geological and morphological knowledge of an area characterized by limited scientific data, in order to understand the interaction between the natural landforms, the urban evolution and the acting geomorphological processes, as thefirst step towards risk management and a conser-vation strategy of the Antananarivo historical center.

1.1. Geomorphological– geological setting

The Antananarivo plateau presents a widespread sur-face of low relief, lying at about 1300 m.a.s.l., within the central sector of Madagascar (Figure 1(b)). Its urban area develops on several low rocky hill domes

Figure 3.Antananarivo historical center quarter areas, including the possible new UNESCO Core and Buffer zones. The High City cultural heritage: (a) The Rova Palace; (b) Andafiavaratra Palace; (c) The Royal Chapel; (d) Ambohipotsy Cathedral; (e) Andohalo Cathedral.

and ridges dotting the alluvial plain of the Ikopa river (Figure 1(b)). The landscape of the central highlands consists of high and moderate-sized hills, where the forest, which formerly covered a much greater area of Madagascar, has almost completely disappeared. From a geological point of view, the area of Antananar-ivo preserves a neo-Archean geological history: the Antananarivo Craton forms the basement of much of Central Madagascar and is dominated by a thick sequence of migmatites, orthogneisses, and granites

that crop out in a broad north–south band across the island (Collins, 2006;Kroener et al., 2000). Granitiza-tion occurred at various epochs (2500 million years, 1100 million years, 550 million years), producing gran-ite intrusions in the form of batholgran-ites or dikes or, by means of granitization of some beds of the pre-Cam-brian series, forming stratoid granites of variable thick-nesses (from a few tens of centimeters up to 500 m or more; Nédélec, Ralison, Bouchez, & Grégoire, 2000). The Madagascar central basement is mantled by

Table 1.List of tropical storms and cyclones affecting the area of Antananarivo since 1972.

Start End Name Class (MFR) Casualties Peak intensity wind (Km/h) Maximum rainfall 10/02/1972 21/02/1972 Eugenie Tropical Cyclone n.a. 120 350 mm/day 01/03/1972 11/03/1972 Hermione Severe Tropical Storm n.a. 100 757.5 mm/day 03/01/1973 12/01/1973 Charlotte Moderate Tropical Storm n.a. 75 813 mm/day 10/01/1976 29/01/1976 Terry-dayanae Intense Tropical Cyclone 50 175 250 mm/day 01/03/1981 11/03/1981 Johanne Tropical Cyclone n.a. 140 316 mm/day 30/01/1982 06/02/1982 Electre Moderate Tropical Storm 13 80 151.1 mm/day 16/03/1982 25/03/1982 Justine Tropical Cyclone n.a. 150 404.5 mm/day 16/01/1984 02/02/1984 Dayomoina Severe Tropical Storm 242 100 950 mm/day 29/01/1984 04/02/1984 Galy Moderate Tropical Storm n.a. 65 44.5 mm/day 03/04/1984 16/04/1984 Kamisy Intense Tropical Cyclone 69 185 232 mm/day 07/03/1986 23/03/1986 Honorinina Tropical Cyclone 99 205 455 mm/day 22/01/1988 01/02/1988 Dayoaza Tropical Cyclone n.a. 135 113.9 mm/day 11/04/1990 14/04/1990 Hanta Moderate Tropical Storm n.a. 65 75 mm/day

16/02/1991 19/02/1991 Cynthia Tropical Cyclone 36 125 420 mm/day

02/03/1993 07/03/1993 Ionia Moderate Tropical Storm 8 65 356 mm/day 16/03/1994 01/04/1994 Nadayia Intense Tropical Cyclone 252 175 126 mm/day 03/01/1996 20/01/1996 Bonita Intense Tropical Cyclone 42 250 170 mm/day 19/02/1996 29/02/1996 Edaywige Tropical Cyclone n.a. 150 369 mm/day

05/02/1997 16/02/1997 Josie Tropical Cyclone 36 140 80 mm/h

01/02/2000 29/02/2000 Leon-Eline Intense Tropical Cyclone 722 215 400 mm/day 27/02/2000 10/03/2000 Gloria Severe Tropical Storm 66 120 215 mm/day 25/03/2000 09/04/2000 Hudayah Very Intense Tropical Cyclone 114 230 80 mm/h 30/12/2001 03/01/2002 Cyprien Severe Tropical Storm 2 100 92 mm/h 14/02/2002 23/02/2002 Guillaume Intense Tropical Cyclone n.a. 205 500 mm/day 05/03/2002 17/03/2002 Hary Very Intense Tropical Cyclone 4 260 372 mm/day

02/05/2002 11/05/2002 Kesiny Tropical Cyclone 33 130 480 mm/day

02/05/2003 13/05/2003 Manou Tropical Cyclone 89 150 360 mm/day

04/12/2003 20/12/2003 Cela Tropical Cyclone n.a. 120 317 mm/day

26/01/2004 05/02/2004 Elita Tropical Cyclone 33 175 200 mm/day

01/03/2004 15/03/2005 Gafilo Very Intense Tropical Cyclone 363 260 250 mm/day 16/01/2005 23/01/2005 Ernest Very Intense Tropical Cyclone 78 165 240 mm/day 26/01/2005 03/02/2005 Felapi Moderate Tropical Storm n.a. 90 160 mm/day

20/01/2006 05/02/2006 Boloetse Tropical Cyclone 6 185 175 mm/day

15/12/2006 28/12/2006 Bondayo Intense Tropical Cyclone 11 250 180 mm/day

25/12/2006 04/01/2007 Clovis Severe Tropical Storm 6 115 40 mm/h

11/02/2007 23/02/2007 Favio Intense Tropical Cyclone 14 220 215 mm/day 19/02/2007 02/03/2007 Gamedaye Intense Tropical Cyclone 4 195 800 mm/day 11/12/2007 23/12/2007 Celina Moderate Tropical Storm 1 130 40 mm/h 29/12/2007 03/01/2008 Elnus Moderate Tropical Storm n.a. 65 150 mm/day

22/01/2008 01/02/2008 Fame Tropical Cyclone 13 130 365 mm/day

07/02/2008 22/02/2008 Ivan Intense Tropical Cyclone 93 230 50 mm/h 02/03/2008 16/03/2008 Jokwe Intense Tropical Cyclone 16 195 300 mm 17/01/2009 22/01/2009 Fanele Intense Tropical Cyclone 10 185 60 mm/h 01/02/2009 09/02/2009 Gael Intense Tropical Cyclone 2 185 50 mm/h 03/04/2009 10/04/2009 Jadaye Severe Tropical Storm 15 110 50 mm/h 09/02/2011 17/02/2011 Bingiza Intense Tropical Cyclone 34 185 50 mm/h 07/02/2012 22/02/2012 Giovanna Intense Tropical Cyclone 35 249 350 mm

25/02/2012 12/03/2013 Irina Severe Tropical Storm 77 167 65 mm/h

26/01/2013 03/02/2013 Felleng Intense Tropical Cyclone 9 232 40 mm/h

18/02/2013 25/02/2013 Haruna Tropical Cyclone 39 185 91 mm/h

26/03/2014 05/04/2014 Hellen Very Intense Tropical Cyclone 17 259 44 mm/h 14/01/2015 19/01/2015 Chedza Severe Tropical Storm 80 93 279.5 mm/day 05/02/2015 08/02/2015 Fundayi Severe Tropical Storm 5 101 120 mm/day 07/03/2015 10/03/2015 Haliba Moderate Tropical Storm 26 101 95 mm/day 02/03/2017 09/03/2017 Enawo Intense Tropical Cyclone 96 231 215 mm/day

27/12/2017 09/01/2018 Ava Tropical Cyclone 73 176 700 mm

01/03/2018 06/06/2018 Dayumazile Intense Tropical Cyclone n.a. 213 160 mm/h 20/04/2018 25/04/2018 Fakir Severe Tropical Storm 2 120 418 mm/day 04/03/2019 16/03/2019 Idayai Intense Tropical Cyclone 1303 205 600 mm/day 21/04/2019 29/04/2019 Kenneth Intense Tropical Cyclone 52 231 730 mm/day Source:https://trmm.gsfc.nasa.gov/;https://emergency.copernicus.eu/mapping/list-of-components/EMSR348.

lateritic clays produced by chemical weathering of the crystalline bedrock; this laterisation has probably been going on since the Pliocene (Cox, Zentner, Rako-tondrazafy, & Rasoazanamparany, 2010). Covered with a sparse vegetation of grasses, the thick, loose lateritic soils overlying the bedrock are exposed to the acceler-ated weathering and erosion, which produce spectacu-lar effects on the relief: the ‘Lavakas’ (the Malagasy word for‘hole’ with steep sides usually on the side of a hill), a typical gully erosional feature in the Central Highlands (Ramifehiarivo et al., 2017).

2. Methodology

Thefield surveys were performed on a 5.2 square kilo-meter by integrating traditional geological and geo-morphological field surveying data with differential Real Time Kynematic GPS. The maps were prepared using a base topographic contour map 1:5000 (1:5000 scale), a post-2015 cyclone event aerial LiDAR data (1 m resolution) and a Pleiades-1A satellite orthophoto (0.5 m resolution, acquired on May 2015). The available data were homogeneized in an ESRI ArcGIS 10 environment (https://www.esri.com/en-us/home), allowing for contour, photo and digital relief visual interpretation, in order to detect morphological fea-tures related to instability phenomena (large gullies, scarps, counterscarps, hummocky topography, slope scars and bare sectors) (e.g.Brardinoni, Slaymaker, & Hassan, 2003; Brunsden, 1993; Guzzetti et al., 2012). The complexity of the urbanization of the study area did not allow for an accurate map of all the hydro-graphic flow-paths; therefore, the drainage network was automatically mapped on the LiDAR data using the ArcMap 10.4.1 Hydrology Tools package (Flow Direction, Accumulation and Strem Order functions) (e.g. Frodella, Morelli, & Pazzi, 2017b; Margottini et al., 2015a;Strahler, 1957). The geological map legend was arranged based on the granite weathering products classification ofLee and De Freitas (1989) and Dewan-del, Lachassagne, Wyns, Maréchal, and Krishnamurthy (2006).

3. Results

The enclosed map shows the main features of the area and incorporates the hydrographic features (Main map on the left) and the geomorphology of the area (Main mapon the right) that will be described in the following paragraphs. The adopted scale was chosen to focus on the UNESCO Buffer zone (an area that must guarantee an additional level of protection to the assets recog-nized as world heritage, UNESCO, 1977) bounding the Analamanga hill area. The boundary of the chosen UNESCO Core zone, encompassing the High City and its cultural heritage is also shown. Other smaller maps show large and regional scale geographic information,

as well as altitude at a local scale (maps a, b, c in the upper right and central section). Urban features such as the High City quarters and roads are also reported (map d in the lowermost section).

3.1. Study area geomorphological setting

The Analamanga hill is a N–S oriented rocky ridge, 3.3 km in length, rising up to 200 m above the Ikopa river alluvial plain (Figure 4). From a morphologic point of view, the hilltop is represented by a flat area (1436 m.a.s.l. at its highest point), rising gradually from north and alternating southward with three sad-dles; two lower saddles link the Analamanga hill foot-slopes on the south-western and central-eastern sides to two adjacent hilltops that dot the Ikopa river plain area. The Analamanga hill southern sector is character-ized by 40° slopes, where diffuse and intense quarrying activity has carved several large niches (Figures 4 and 5). The western hillside shows a high slope angle in its uppermost sector, due to the widespread outcrop-ping of the granite bedrock along a sub-vertical rock wall, while the slope toe is represented by low energy surfaces gently linking with the plain area, severely affected by lavaka-like gully erosional features. On the contrary, the eastern hillside has a general minor slope angle (mean values around 30°) (Ciampalini et al., 2019). This slope being more exposed to the cyclones coming from the Indian Ocean, it is affected by sheet-overlandflow erosion, as testified by slope scars expos-ing eluvial–colluvial deposits overlyexpos-ing the granite bed-rock. Residual lateritic soils (up to a few tens of meters of maximum thickness) widely crop out at the hill foot-slopes, bounding the lower sector of the whole Anala-manga hill, particularly in the western sector.

3.2. Hydrography

Within the study area, the complexity of the hydro-graphic network is due to the strong influence of urbanization (roads, bridges, buildings, anthropic channels), which modifies the natural flow-paths of the hydrographic network. Channeling and narrowing often affect the creeks, especially in their terminal stretch, where at times they experience thefirst culvert-ing and then swampculvert-ing (Figure 8). Swamps were exten-sively reclaimed during the colonial period, creating artificial ponds and rice-paddies. The latter are drained by means of channels, the larger of which are located in the northern Great Antananarivo urban area, which converge into the Ikopa river main course (Figure 1). According to afirst field surveying approach, the gen-eral hydrographic pattern of the Analamanga hill shows radial-like features, with small creeks, deeply cutting into both the hill slopes. The performed GIS analysis allowed to map this‘anthropic’ hydrographic network and to identify 10 small watersheds,

characterized by streams up to the fifth order; their lowermost sector is delimited by the boundary of the UNESCO Buffer Zone (Table 2). The western slope is characterized by a general lower stream order of the creeks, which often form deep and narrow small valleys trending WNW-ESE, due to the linear erosion of the thick lateritic soil cover (=large gullies, or lavakas; Basins 6–7). On the contrary, the eastern slope is characterized by a more developed hydrographic net-work (sub-dendritic pattern-like), with higher stream order and wider basins trending NE-SW (Basins 1– 4). In these catchments, sheet erosion phenomena are more frequent, as testified by frequent bare soil exposures. Generally, the creeks are present also in the dry season, thanks to the presence of springs and watercress (cressonieres), testifying the presence of

perched water tables in the upper slope area. The catch-mentsflow direction is influenced by a combination of anthropic-natural factors: (i) the two morphological lower saddles (Figures 4 and5) force the network of Basins 1 and 7–10 to flow North, while the network of Basins 2–5 flows South into the large rice fields area; (ii) In the northern sector the hill morphology and the streets, the Antananarivo Stadium and ponds deeply influence the flow path direction of Basins 8-10, while on the southern sector the frequent quarry slope cuts heavily affect Basin 5.

3.3. Study area geological features

From a geological point the view the Analamanga hill is composed of a migmatitic granitoid batholite, covered

discontinuously by the various products of the pro-gressive stages of weathering of the crystalline bedrock. These latter are formed by an irregular sequence reported here, from bottom to top, following the classifications ofLee and De Freitas (1989) and Dewan-del et al. (2006) (Figure 6(a)):

(i) granite fresh rock (=F), in which various frac-ture network develop, widely outcropping in the steep rock walls at the top of the western slopes, and in the south-eastern slope toe (in correspondence with the abandoned quar-ries) (Figure 6(b));

(ii) laminated and slightly moderately weathered granite (saprock = SW-MW), cropping out on top of the rock walls (Figure 6(b));

(iii) from highly to completely weathered granite (saprolite = HW-CW), discontinuously cropping out especially in the hilltop and on the eastern hillside (Figure 6(a));

(iv) completely decomposed residual soils (laterite = RS), formed by mainly loamy sands, completely encircling both the of hillslope toes (Figure 6

(c)). The higher thickness of these deposits (up to a few tens of meters) is located in correspon-dence with the low energy surfaces at the bottom of the western slope;

(v) loose sandy eluvial/colluvial cover (=SC), formed mainly by slightly pebbly sands, irregularly out-cropping especially in the hilltop and the eastern hillside (Figure 6).

Slope deposits mantle the western and eastern slopes in fan-like shapes, and are constituted by cobbles, blocks and scattered boulders, deposited by slope grav-itational processes, in a coarse sandy matrix formed by the weathering of the bedrock (Figure 6(e)). Further-more, the intense anthropic activity in the last centuries has created detrital deposits composed of heteroge-neousresidual building materials, fromfine to coarse-grained (anthropic deposits = AD;Figure 6(d)). These may cover the above-mentioned sequence in corre-spondence with the inhabited areas. Alluvial deposits (=AL), formed by organic-rich clays and silts, mantle the river plain areas in correspondence with the rice-paddies.

Table 2.Features of the identified watershed.

Watershed n° Watershed name Area (m2) Area (ha) Perimeter (m) Stream order (n°)

1 Ambohitsiroa 1,228,350 122.8 5708 5 2 Manjakamiadana East 477,143 47.7 3295 5 3 Ambohimitsimbina 439,913 43.9 3753 4 4 Ambohipotsy East 614,499 61.4 3965 5 5 Andohamandry 315,608 31.5 2319 4 6 Ambohipotsy West 282,737 28.2 2387 4 7 Manjakamiadana West 168,188 16.8 1982 3 8 Mahamasina 483,142 48.3 3050 5 9 Lac Anosy 653,370 65.3 3215 4 10 Ambatovinaky 547,122 54.7 4076 4

Figure 6.Geological features of the study area: road cut showing an example of different granite weathering products (a); fresh granite (on the bottom) and limited granite (on the top) (b); laterite terrace slope cut for housing (c); eluvial–colluvial cover mixed with rubbish on vegetated slope (d); detail of heterogeneous slope deposits (e).

3.4. Slope Instability processes

In this geological and geomorphological context four main groups of instability phenomena were identified, following the classification of Cruden and Varnes (1996), according to which the above-mentioned materials are grouped as follows: (i) rock (F-SW-MW); (ii) debris (HW-CW-AD); (iii) earth (RS):

(1) Debris rotational/translational slides occurring in HW-CW-RS-AD (Figure 7(a)). These phenomena generally occur on the hilltop and the south-east-ern hillside, and are characterized by small volumes and reduced mobility (usually they accumulate at the slope toe and do not reach the hydrographic network). Nevertheless, in the case of cyclonic rain, these can evolve in unchanneled debrisflows moving on the slope surface.

(2) Earth-debris rotational/translational slides invol-ving RS and AD. These phenomena occur along the slopes of the main creek channels deeply cut-ting the slope toe at the foot of the western hillside (gullies; Figure 7(b)), and may evolve into chan-neled debrisflows following intense rainfall events during the cyclones. Their related level of risk is due to the interaction with man-made structures, such as buildings located directly inside the

channels, which often restrict their section, and road paving often causing their culverting in the downstream sector.

(3) Rock fall and rock slides (planar and wedge), from metric to decametric dimension, involving the granite fresh rock (F) (Figure 7(c)). These instabil-ity processes affect the steep rock walls, located both in the western hillside and at the south-east-ern slope toe, and are characterized by high risk, as testified by the severe building damage and casual-tiesin the March 2019 event.

(4) Complex rotational/translational slide phenomena involving various materials such as SW-MW-HW-CW and AD (Figure 7(d)). One of these phenom-ena, from small to medium size (1 m3< V < 15 m3), occurred in January 2019 in correspon-dence with the top of the western rock wall. It evolved down the rock wall as a debris fall-ava-lanche: a few houses located at the foot of the rock wall were destroyed causing several casualties. Therefore, this type of phenomena is characterized by a high risk.

4. Discussion and conclusions

The morphological setting of the Analamanga hill granted a strategic commanding position above the

Figure 7.Examples of the detected instability phenomena: (a) Debris rotational/translational slides; (b); Earth-debris rotational/ translational slides; (c) Rock fall and rock slides; (d) Complex rotational/translational slide.

Ikopa river valley and the surrounding highlands, favoring the development of the first fortified village around the royal palace, the‘High City’, on its flat hill-top in the sixteenth century. The further development of Antananarivo as the Capital city of Madagascar at the end of the eighteenth century caused the beginning of the urban growth along the hillslopes, creating the ‘Middle City’. During the colonial period, this process was enhanced, reaching the river plain areas and creat-ing the ‘Low City’, while in more recent times the urban area has sprawled covering approximately 86.4 km2(‘Great Antananarivo’). This rapid expansion was not controlled by proper urban planning and did not take into account the geological and geomorpholo-gical processes acting on the slopes and the river plain. The steep slope sectors, the widespread weathering and fracturing of the granitoid bedrock, together with the

heavy cyclonic rainfall frequently hitting the highland area, represent natural predisposing/triggering factors for geo-hydrological hazards. In this context the anthropic activity is exacerbating soil erosion, slope instabilities and flooding due to overbuilding, diffuse burning of the vegetation cover, quarry excavation and slope cutting, garbage dumping, hovels lacking proper sewer and drainage systems, narrowing, culvert-ing and swampculvert-ing of the creeks. The identification of landslide processes on thefield was not an easy task, due to the rapid modification of the landforms and slope cover caused by the intense hovel urbanization. Nevertheless, the combination of field surveys and the interpretation of remote sensing data has provided the detection of landslide processes and of the hydro-graphic network with a satisfactory resolution from the local to a large scale. Regarding landslide hazards,

Figure 8.Natural and Urbanized features of the study area hydrographic network: Lavakas in the western footslope (Basin 7; a); narrowing by concrete buildings (Basins 6; b); Culverting causing road pavement erosion, and garage dumpingfilling culvert (Basin 8; c, d); Ricefields in Basin 1 (e); general view of an artificial pond (Lac Anosy, Basin 9; f).

the most exposed areas are the depressed southern sec-tor of the Middle City area, particularly the western hillslope (Basins 6-7-8), where hovels have been built at the foot of rock slopes, as testified by the winter 2019 rock fall events. Debris rotational/translational slides involving the saprolithe and the overlying elu-vial/colluvial cover were mapped on the south-eastern hillside, while Earth-debris rotational/translational slides involving residual soil were identified along the slopes of the main creek channels deeply cutting the slope toe at the foot of the western hillside. The Anala-manga hill hydrography is characterized by a radial-like network formed by 10 small watersheds, ranging in the surface from 122 to 16 hectares, and characterized by streams up to the fifth order. Intense linear erosion of the soft lateritic soil cover creates large gullies along all the western slope toe, which are rapidly expanding and damaging the road pavement and buildings. Sheet erosion phenomena are more frequent on the eastern slope. The catchmentsflow direction is influenced by a combination of the natural landforms and the urban structures: regarding flood hazard, the most exposed areas are again Basins 6-7-8, where often hovel and con-crete buildings are located within the large creek gullies. The produced maps represent thefirst step for land-use planning and a general master plan of mitigation measures for the High City and surroundings; in the near future, detailed susceptibilty mapping, slope stab-ility analysis and accurate landslide inventory are necessary for a complete hazard assessment.

Software

ESRI ArcMap 10.4.1 was used for the data integration, the automatic hydrographic network analysis and DEM derived products (Elevation map, Hillshade, Slope), for digitizing landforms recognized by means of field surveys and DEM-ortophoto interpretation and for cartographic work. Map layout andfinal edit-ing were performed usedit-ing Adobe Illustrator CS5.

Acknowledgements

This work was carried out in the framework of the activi-ties of the UNESCO chair on prevention and sustainable management of geo-hydrological hazards of the University of Florence, Italy (www.unesco-geohazards.unifi.it), car-ried out in cooperation with the Région Île-de-France. The authors would like to thank: Paris Region Expertise-Madagascar (PRX) and in particular Tamara Teisseidre-Philip, Alexandrine Wadel, Coline Mollaret for providing all of the available topographic data, and especially kind-ness and support during our work; Francois Cristofoli and Filippo De Dominicis (RC Heritage/RCh consultants) for the logistic support and suggestions; Helihanta Rajao-narison of Antananarivo University for her precious knowledge and guidance, Niandr Bania Ramboason and Andria Rou’hrouh for their valuable assistance and enthusiasm.

Disclosure statement

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

ORCID

Nicola Casagli http://orcid.org/0000-0002-8684-7848

References

Bardi, F., Raspini, F., Frodella, W., Lombardi, L., Nocentini, M., Gigli, G.,… Casagli, N. (2017). Monitoring the rapid-moving reactivation of earth flows by means of GB-InSAR: The April 2013 Capriglio landslide (northern Appennines, Italy). Remote Sensing, 9(2), 165.

Brunsden, D. (1993). Mass movement; the research frontier

and beyond: A geomorphological approach.

Geomorphology, 7, 85–128.

Brardinoni, F., Slaymaker, O., & Hassan, M. A. (2003). Landslide inventory in a rugged forested watershed: A comparison between air-photo and field survey data. Geomorphology, 54(3–4), 179–196.

Calista, M., Miccadei, E., Pasculli, A., Piacentini, T., Sciarra, M., & Sciarra, N. (2016). Geomorphological features of the Montebello sul Sangro large landslide (Abruzzo, Central Italy). Journal of Maps, 12(5), 882–891.

Ciampalini, A., Frodella, W., Margottini, C., & Casagli, N. (2019). Rapid assessment of geo-hydrological hazards in Antananarivo (Madagascar) historical centre for damage prevention. Geomatics, Natural Hazards and Risk, 10(1), 1102–1124.

Cox, R., Zentner, D. B., Rakotondrazafy, A. F. M., &

Rasoazanamparany, C. F. (2010). Shakedown in

Madagascar: Occurrence of lavakas (erosional gullies) associated with seismic activity. Geology, 38(2), 179–182. Cruden, D. M., & Varnes, D. J. (1996). Landslides: Investigation and mitigation. Chapter 3-Landslide types and processes. Transportation research board special report (247).

Collins, A. S. (2006). Madagascar and the amalgamation of Central Gondwana. Gondwana Research, 9(1), 3–16. Del Soldato, M., Riquelme, A., Bianchini, S., Tomás, R., Di

Martire, D., De Vita, P.,… Calcaterra, D. (2018). Multisource data integration to investigate one century of evolution for the Agnone landslide (Molise, Southern Italy). Landslides, 15(11), 2113–2128.

De Muro, S., Ibba, A., Simeone, S., Buosi, C., & Brambilla, W. (2017). An integrated sea-land approach for mapping geo-morphological and sedimentological features in an urban microtidal wave-dominated beach: A case study from S Sardinia, western Mediterranean. Journal of Maps, 13(2), 822–835.

Dewandel, B., Lachassagne, P., Wyns, R., Maréchal, J. C., & Krishnamurthy, N. S. (2006). A generalized 3-D geological and hydrogeological conceptual model of granite aquifers controlled by single or multiphase weathering. Journal of Hydrology, 330(1–2), 260–284.

Frodella, W., Morelli, S., Fidolini, F., Pazzi, V., & Fanti, R.

(2014). Geomorphology of the Rotolon landslide

(Veneto Region, Italy). Journal of Maps, 10(3), 394– 401.

Frodella, W., Gigli, G., Morelli, S., Lombardi, L., & Casagli,

N. (2017a). Landslide mapping and characterization

through infrared thermography (IRT): Suggestions for a methodological approach from some case studies. Remote Sensing, 9(12), 77–84.

Frodella, W., Morelli, S., & Pazzi, V. (2017b). Infrared ther-mographic surveys for landslide mapping and characteriz-ation: The Rotolon DSGSD (Northern Italy) case study. Italian Journal of Engineering Geology and Environment, 1, 77–84. Special Issue.

Gigli, G., Intrieri, E., Lombardi, L., Nocentini, M., Frodella, W., Balducci, M.,… Casagli, N. (2014). Event scenario analysis for the design of rockslide countermeasures. Journal of Mountain Science, 11(6), 1521–1530.

Guzzetti, F., Mondini, A. C., Cardinali, M., Fiorucci, F., Santangelo, M., & Chang, K. T. (2012). Landslide inven-tory maps: New tools for an old problem. Earth-Science Reviews, 112, 42–66.

Lee, S. G., & De Freitas, M. H. (1989). A revision of the description and classification of weathered granite and its application to granites in Korea. Quarterly Journal of Engineering Geology and Hydrogeology, 22(1), 31–48. Karymbalis, E., Papanastassiou, D., Gaki-Papanastassiou, K.,

Tsanakas, K., & Maroukian, H. (2013). Geomorphological study of Cephalonia Island, Ionian Sea, Western Greece. Journal of Maps, 9(1), 121–134.

Kroener, A., Hegner, E., Collins, A. S., Windley, B. F., Brewer, T. S., Razakamanana, T., & Pidgeon, R. T. (2000). Age and magmatic history of the Antananarivo Block, central Madagascar, as derived from zircon geo-chronology and Nd isotopic systematics. American Journal of Science, 300(4), 251–288.

Margottini, C., Fidolini, F., Iadanza, C., Trigila, A., & Ubelmann, Y. (2015a). The conservation of the

Shar-e-Zohak archeological site (Central Afghanistan):

Geomorphological processes and ecosystem-based miti-gation. Geomorphology, 239, 73–90.

Margottini, C., Antidze, N., Corominas, J., Crosta, G. B., Frattini, P., Gigli, G.,… Vacheishvili, N. (2015b). Landslide hazard, monitoring and conservation strategy for the safeguard of Vardzia Byzantine monastery com-plex, Georgia. Landslides, 12(1), 193–204.

Nédélec, A., Ralison, B., Bouchez, J. L., & Grégoire, V. (2000). Structure and metamorphism of the granitic basement around Antananarivo: A key to the Pan-African history of central Madagascar and its Gondwana connections. Tectonics, 19(5), 997–1020.

Nolesini, T., Frodella, W., Bianchini, S., & Casagli, N. (2016). Detecting slope and urban potential unstable areas by means of multi-platform remote sensing techniques: The Volterra (Italy) case study. Remote Sensing, 8(9), 746. Paliaga, G., Luino, F., Turconi, L., & Faccini, F. (2019).

Inventory of geo-hydrological phenomena in Genova municipality (NW Italy). Journal of Maps, 15(2), 28–37. Pastonchi, L., Barra, A., Monserrat, O., Luzi, G., Solari, L., &

Tofani, V. (2018). Satellite data to improve the knowledge of geohazards in world heritage sites. Remote Sensing, 10 (7), 992.

Pucci, S., Villani, F., Civico, R., Pantosti, D., Del Carlo, P., Smedile, A.,… Gueli, A. (2015). Quaternary geology of the middle Aterno valley, 2009 L’Aquila earthquake area (Abruzzi Apennines, Italy). Journal of Maps, 11(5), 689– 697.

Ramifehiarivo, N., Brossard, M., Grinand, C.,

Andriamananjara, A., Razafimbelo, T., Rasolohery, A., & Albrecht, A. (2017). Mapping soil organic carbon on a national scale: Towards an improved and updated map of Madagascar. Geoderma Regional, 9, 29–38.

Roccati, A., Faccini, F., Luino, F., Ciampalini, A., & Turconi, L. (2019). Heavy rainfall triggering shallow landslides: A susceptibility assessment by a GIS-approach in a Ligurian Apennine Catchment (Italy). Water, 11(3), 605. Smith, A., & Ellison, R. A. (1999). Applied geological maps

for planning and development: A review of examples from England and Wales, 1983 to 1996. Quarterly Journal of Engineering Geology and Hydrogeology, 32 (Supplement), S1–S44.

Strahler, A. N. (1957). Quantitative analysis of watershed geomorphology. Transactions, American Geophysical Union, 38, 913–920.

UNESCO, 1977. Operational guidelines for the

implemen-tation of the world heritage convention.

Intergovernamental committee for the protection of the world cultural heritage and natural heritage. First Session, Unesco, Paris, 27 June-1July 1977.

Visser, F. (2014). Rapid mapping of urban development from historic Ordnance Survey maps: An application for pluvialflood risk in Worcester. Journal of Maps, 10(2), 276–288.