Field Monitoring in Sample Sites: Hydrological Response of Slopes with Reference to Widespread Landslide Events

Available online at www.sciencedirect.com ScienceDirect

Procedia Earth and Planetary Science 9 ( 2014 ) 44 – 53

The Third Italian Workshop on Landslides Field monitoring in sample sites: hydrological response of slopes with reference to widespread landslide events

Giovanni Gullàa*

aCNR-IRPI – U.O.S. Cosenza, Via Cavour n. 4, Rende 87036, Italy

Abstract

The field monitoring is an essential tool to define geotechnical model able to assess landslide stability conditions, with particular reference to the hydrological response of the slopes in presence of widespread landslide events — the occurrence of several landslides trough wide areas. The results relative to two sample sites, representative of study areas characterised by homogeneous geological elements and where field monitoring has been carried out for adequate time intervals, are illustrated in the paper. In particular, we consider an area with outcroppings formed by a sort of “melange structure” made up of blocks and fragments of phyllites, clays, shales, ect., in a prevalently clayey matrix (Lungro sample site), and an area where are present rocks and soils deriving from weathering of crystalline rocks (“Serra di Buda”, Acri, sample site). In the Lungro sample site a piezometer monitoring network and a rain gauge give indications about the hydrological response of the slopes in an urban area; in the “Serra di Buda” sample site the piezometer levels measured for a long time period, and the rainfall, permit to identify some relationships between the cumulative rainfall and the piezometric levels. Values of cumulative rainfall of about 700 mm on 120 days represent a necessary, but not sufficient, condition for critical stability conditions in the considered sample sites, in relation to possible scenarios of widespread landslide events.

© 2014 Elsevier B.V. This is an open access article under the CC BY-NC-ND license © 2014 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and peer-review under responsibility of Dipartimento di Ingegneria Civile, Design, Edilizia e Ambiente, Seconda SelectionUniversità and di Napoli.peer-review under responsibility of Dipartimento di Ingegneria Civile, Design, Edilizia e Ambiente, Seconda Universit di Napoli. Keywords: monitoring, integrated network, landslide, hydrological respons, piezometric level.

* Corresponding author: +39-0984-841458 E-mail address: [email protected]

1878-5220 © 2014 Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and peer-review under responsibility of Dipartimento di Ingegneria Civile, Design, Edilizia e Ambiente, Seconda Universit di Napoli. doi: 10.1016/j.proeps.2014.06.008 Giovanni Gullà / Procedia Earth and Planetary Science 9 ( 2014 ) 44 – 53 45

1. Introduction

Landslides are activated especially during rainy season1,2,3,4,5. Either prolonged rainy periods, characterized by low daily intensities, or short and intense storms can trigger landslides, often resulting in damage and casualties6,7,8,9. Calabria is one of the Italian regions mostly affected by mass movements10,11,12,13,14 because of its complex geological evolution (Fig. 1a), its geographical location and distribution of mountain ranges, causing great space variability of climatic conditions and frequent extreme rainfall events17 (Fig. 1b). The geological complexity of Calabria is the result of the tectonic history of the region during its formation18,19. Almost the entire region is made up of crystalline–metamorphic nappe units, defined as Calabrian Arc18, and this condition represents a relevant predisposing factor to landslides20,21. However, landsliding phenomena triggered by precipitation involve almost all outcropped lithologies10,11,12,13,14. In the described context field monitoring assumes particular relevance. In effect, field monitoring is an essential tool to define geotechnical model able to assess landslide stability conditions, with particular reference to the hydrological response of the slopes, and then to assess susceptibility, hazard and risk, to design and realize reduction and/or mitigation risk measures, to study the influence of predisposing and triggering factors on the slope stability conditions.

(a) (b)

Fig. 1. (a) Lithostructural map of the Calabria region and Integrated Monitoring Networks (from [14], [15] and [16], modified); (b) Average yearly rainfall (1921-2000 rainfall) (from [5], modified).

Therefore, starting from 1986, Integrated Monitoring Networks have been realized in some sample sites in Calabria region (Fig. 1a). In particular, the monitored landslides involve, prevalently, soils coming from weathering of rocks. The monitored landslides are shallow (2-3 m depth), medium deep (about 30 m depth) and deep-seated (deeper than about 30 m). The sample site of San Pietro in Guarano represents an example of the useful results that can be achieved from the Integrated Monitoring Network20,22,23,24,25,26,27,28. A critical point in field monitoring is the availability of measures for a significant time period, considering that in most cases field monitoring is not available for landslide phenomena that take place each year. Consequently, it is useful to consider the monitoring results with reference to the occurrence of several landslides triggered by rainfall through wide areas and throughout periods ranging from a few days to a few months5. 46 Giovanni Gullà / Procedia Earth and Planetary Science 9 ( 2014 ) 44 – 53

This paper deals with the results, in terms of hydrological response of the slopes, obtained from the field monitoring carried out for adequate time intervals on two sample sites. In particular, taking into account the rainfall events that have generated widespread landslide events in Calabria region, we propose the analysis of the piezometric monitoring data for sample sites representative of study areas characterised by homogeneous geological elements. Preliminary indications can be found about the processes that determine slope instability conditions with reference to possible scenarios of widespread landslide events.

2. Method and data

In the field monitoring, the availability of measures for a significant time period represents a critical point, and for this reason the method proposed in this paper assumes to consider the measurements with reference to the occurrence of Widespread Landslide Events (WLE)5. Starting from the Landslide database used for Calabria region (historical data on the damage caused by landslides triggered by rainfall referred to 3451 Landslide Records (LRs) occurred in 409 municipalities from 1921–2009), we define Landslide Event (LE) as the occurrence of one or more LRs during one or more consecutive days preceded and followed by at least one day without a LR, and select the LEs that hit at least 20 municipalities (Fig. 2). Then, using the rainfall data relative to the Calabria region a rainfall database of Calabria [17-5] was created, and a rainfall events database was extracted (Rainfall Event (RE) is a group of rainy days preceded and followed by at least one day with no rain). Comparing selected Landslide Events (LEs) and relative Rainfall Events (REs) we can individuate Widespread Landslide Events (Fig. 2). In Figure 3 the interactions between Widespread Landslide Events and monitoring periods are shown. Referring to the geological contexts, to the thickness of the landslides, and to the involved lithologies, we have selected two sample sites to analyse some results obtained from field monitoring: Lungro9, Serra di Buda (Acri)29. The comparison of the piezometric levels and the cumulative rainfall, with reference to WLE, allows to verify the presence of typical values of cumulative rainfall that can be assumed as indicatives of possible instability conditions, able to generate Widespread Landslide Events.

2.1. Lungro sample site

The Lungro sample site9 is located on the northwest sector of the Calabria region, characterized by an extreme geological and structural complexity for the reason that the Units of the Calabrian Arc are overthrusted on the Meso- Cenozoic sedimentary sequences of the Appennine domains30 (Fig. 4a). The lithologies outcropping in this area are represented by phyllites and slate (Diamante-Terranova Unit), and form a sort of “melange structure” made up of blocks and fragments prevalently of phyllites, clays and shales in a clayey matrix. In the area, the slopes are mainly affected by slides, slide-flows and landslide zones (Fig. 4b), prevalently medium-deep landslides. The grain size envelope of the soils produced by degradation processes of phyllites in Lungro territory is delimitated by curves from sandy clay with silt to sandy silty gravel (Fig. 4c). The hydraulic properties at full saturation, obtained from laboratory and the in situ tests, show for these soils a permeability ranging from 3.62*10-7 m/s to 5.00*10-5 m/s (Fig. 4d).

2.2. Serra di Buda (Acri) sample site

The “Serra di Buda” (Acri) sample site29,31,32 is located on north-western border of the Sila Massif characterized by Quaternary extensional N-S faults superimposed at WNW-ESE strike-slip faults33. The N-S striking faults are arranged into a westward down-stepping system (Fig. 5a). Crystalline rocks are mainly represented by medium-to- high-grade metamorphic rocks of the Sila Unit, generally intensely weathered and fractured. In the study area, deep- seated gravitational deformations and deep-seated landslides have been recognized, mainly in the transition zone between the above cited fault-systems33,34,35. In particular, in the Acri area have been identified and mapped three types of landslide: debris flow (generally shallow landslides), debris slide (generally medium-deep landslides) and rock slides (generally deep-seated landslides)34. The Serra di Buda deep-seated landslide (Fig. 5b), located near the Giovanni Gullà / Procedia Earth and Planetary Science 9 ( 2014 ) 44 – 53 47 town of Acri (Cosenza), is part of a wide Sackung-type Deep-seated Gravitational Slope Deformation32,34,35. The grain size envelope of the classes VI and V gneissic rocks of Serra di Buda is delimitated by curves from clayey silt with sand to sandy gravel (Fig. 5c). The laboratory and the in situ tests show that the permeability of these soils and rocks ranges from 5.00x10-7 m/s to 1.25x10-3 m/s (Fig. 5d).

1085LEs 200,000REs

20municipalitieshit

25WLEs

Fig. 2. Flow-chart showing the methodology proposed for individuating WLEs (from [5], modified).

48 Giovanni Gullà / Procedia Earth and Planetary Science 9 ( 2014 ) 44 – 53

Fig. 3. Interactions between Widespread Landslide Events and monitoring periods.

(a (b)

(c (d)

Fig. 4. (a) Lithostructural map of the Lungro area (from [30], modified); (b) Lithostructural and landslide map of the Lungro sample site and piezometer-inclinometer monitoring networks (from [9], modified) - Legend: 1) landslide debris; 2) colluvial soil; 3) slope debris; 4) alluvional deposit; 5) phyllites; 6) cataclastic dolomite; 7) meta-limestone; 8) fault; 9) active landslide; 10) dormant landslide; 11) active landslide scarp; 12) dormant landslide scarp; 13) inclinometer; 14) piezometer; (c) Grain size envelope of the colluvial soils and degraded phyllites; (d) Permeability values of the colluvial soils and degraded phyllites. Giovanni Gullà / Procedia Earth and Planetary Science 9 ( 2014 ) 44 – 53 49

(a) (b)

(d) (c)

Fig. 5. (a) Lithostructural map of the Acri area (from [13], modified); (b) ”Serra di Buda” landslide map and integrated monitoring network; (c) Grain size envelope of the classes V and VI from gneiss; (d) Permeability values of the classes V and VI from gneiss.

2.3. Serra di Buda (Acri) sample site

The interactions between WLEs and monitoring period relative to piezometers in Serra di Buda sample site are shown in Figure 3. In particular, specific LRs have detected in the WLEs 22,24,25, and in the WLS that have been generated from January to March 2010 (WLE 25*). In correspondence of the central part of the Serra di Buda slope (Fig. 5b), and with reference to WLEs, it is possible to take some observations for the hydrological response of slope. In the S3 piezometer vertical, about five years after the starting of the monitoring, we can observe on 14 March 2005 a clear rising in the piezometric levels, (1) in Figure 7, about 14-19 m below g.s., in presence of cumulative rainfall on 120 days of about 800 mm. Five years later, on 19 March 2010, a greater rising of the piezometric levels of about 11-14 m below g.s., (2) in Figure 7, can be observed in presence of values of cumulative rainfall of about 800 mm. Either these cases are characterized by the presence of snow (Fig. 7). In absence of snow, only on 21 May 2009 we register a sudden rising of the piezometric levels, (3) in Figure 7, of about 11-19 m with a cumulative rainfall of about 900 mm (Fig. 7). For the sample site of “Serra di Buda” more than 11 years of monitoring indicate that, in presence of snow, a cumulative rainfall on 120 days with values of about 800 mm determines significant variations in piezometric levels. Significant, but sudden rising in piezometric levels could be recorded also in absence of snow with values of cumulative rainfall greater than 800 mm. In Figure 8 two examples of the effects of the hydrologic response of the “Serra di Buda” slope are shown. In particular, from December 2004 to June 2005, in concomitance to the minimum piezometric levels (about 14-19 m 50 Giovanni Gullà / Procedia Earth and Planetary Science 9 ( 2014 ) 44 – 53

below g.s.), the measures carried out in the GPS benchmark P01 (Fig. 5b) have recorded a maximum rate of displacement of about 25 mm/day (Fig. 8a) and from June 2005 to March 2009, with piezometric levels greater than 15 m below g.s., the measures carried out in the same benchmark P01 (Fig. 5b) have recorded a constant rate of displacement of about 0.4 mm/day (Fig. 8b).

Fig. 6. Rainfall, piezometric levels, and inclinometric displacements in the Historical centre of the Lungro sample site.

3. Conclusions

Field monitoring is an essential tool to assess landslide stability conditions, but presents many critical points, in particular relative to long time periods measurements. The results obtained in the sample sites considered in this paper, through piezometric measures carried out for an adequate time period, provide some interesting indications relatively to the hydrologic response of two slopes in homogeneous geological contexts, and for medium deep and deep-seated landslides. For the translational slides with a maximum depth of about 30 m, located in the Historical centre of the Lungro sample site9, the cumulative rainfall on 120 days with values greater than about 700 mm determines a significant variation in the piezometric levels (about 2-4 m) that produces an increase of the displacement rate in the mass movements, involving colluvial soils and degraded phyllites. Giovanni Gullà / Procedia Earth and Planetary Science 9 ( 2014 ) 44 – 53 51

Fig. 7. Rainfall and piezometric levels (S03 piezometer vertical) in “Serra di Buda” sample site.

(a) (b)

Fig. 8. Typical hydrological responses of the “Serra di Buda” slope and measured displacements (GPS benchmark P01): (a) December 2004- June 2005; (b) June 2005-March 2009. 52 Giovanni Gullà / Procedia Earth and Planetary Science 9 ( 2014 ) 44 – 53

In the “Serra di Buda” sample site, for a deep-seated landslide involving intensely weathered and fractured gneissic rocks34,35, in presence of snow and with values of 120-days cumulative rainfall of about 800 mm we find variations in piezometric levels (about 3-6 m) that produce significant changes in the landslide kinematic (rate of displacement from 0.4 mm/day to 25 mm/day). The useful observations obtained from the field monitoring in the sample sites of Lungro and “Serra di Buda”, essential to model the “reality”, allow to infer indications about the monitoring time period needed: about 6 years for medium deep landslides involving colluvial soils and degraded phyllites, and about 11 years for deep-seated landslides in weathered and fractured gneissic rocks. Field monitoring is very expensive, time consuming, and affected by high risk of failing. The deep study of Widespread Landslide Events5 permit to identify sample sites, representative of landslides that produce high social and economic impact, with reference to homogeneus geological contexts, to defined landslide typologies and depths. In these identified sample sites, preliminary studies (geological, hydrological, geotechnical) allow to design and realize Integrated Monitoring Networks useful for multi-purposes. Moreover the sample sites with Integrated Monitoring Network permit to generate an “Observatory of Networks” with the aim to study the instability processes, to define efficient measures of risk prevention and reduction, and to validate and develop technologies.

Acknowledgements

The author would like to thank Luigi Aceto, Luigi Borrelli, Sarah C. Maiorano, and Gino Cofone for their valuable collaboration and for the usefull discussions. Thanks also to Roberto Coscarelli for the fruitful suggestions, and to Claudio Reali, Salvatore Guardia, Enzo Valente and Duilio D'Onofrio for the measurements carried out on the monitoring networks. This paper is part of the researches carried out under the “Commessa TA.P05.012 – Tipizzazione di eventi naturali e antropici ad elevato impatto sociale ed economico” of the National Research Council (CNR)-Department “Scienze del sistema terra e tecnologie per l’ambiente”.

References

1. Brand EW. Relationship between rainfall and landslides in Hong Kong. In: Proceedings of the 4th International Symposium on Landslides, Toronto 1984;1. p. 37-384. 2. Crozier MJ. Landslides: causes, consequences and environment. Croom Helm, London. 252 pp; 1986. 3. Glade T. Establishing the frequency and magnitude of landslide-triggering rainstorm events in New Zealand. Environ Geol 1998;35(2-3):160- 174. 4. Gullà G, Antronico L, Iaquinta P, Terranova O. Susceptibility and triggering scenarios at a regional scale for shallow landslides. Geomorphology 2008;99:39-58. 5. Gullà G, Caloiero T, Coscarelli R, Petrucci O. A proposal for a methodological approach to the characterization of Widespread Landslide Events: an application to Southern Italy. Nat Haz Earth System Sci 2012;12:165-173. 6. Guzzetti F. Landslide fatalities and evaluation of landslide risk in Italy. Eng Geol 2000;58:89-107. 7. Remondo J, Soto J, Gonzàlez-Dìez A, Dìaz de Teràn J R, Cendero A. Human impact on geomorphic processes and hazards in mountain areas in northern Spain. Geomorphology 2005;66(1–4):69-84. 8. Petrucci O, Pasqua AA, Gullà G. Landslide damage assessment using the Support Analysis Framework (SAF): the landsliding event in Calabria (Italy). Adv Geosci 2010;26:13-17. 9. Antronico L, Borrelli L, Peduto D, Fornaro G, Gullà G, Paglia L, Zeni G. Conventional and innovative techniques for the monitoring of displacements in landslide affected area. In: C. Margottini et al., editors. Landslide Science and Practice 2013;2:125-131. 10. APAT. Rapporto sulle frane in Italia: il progetto IFFI-metodologia, resultati e rapporti regionali. Rapporto 78/2007. Agenzia per la protezione dell'ambiente e per i servizi tecnici 2007, Rome, Italy. (in Italian) 11. Antronico L, Sorriso-Valvo M, Tansi C, Gullà G. Carta litologico-strutturale e dei movimenti in massa della Stretta di Catanzaro (Scala 1:50000). Consiglio Nazionale delle Ricerche – Gruppo Nazionale per la Difesa dalle Catastrofi Idrogeologiche, LINEA 2 – Previsione e Prevenzione di Eventi Franosi a Grande Rischio, Unità Operativa 2.56 (Pubblicazione n. 2119) 2001. (in Italian) 12. Sorriso-Valvo M, Antronico L, Gaudio R, Gullà G, Iovine G, Merenda L, Minervino I, Nicoletti PG, Petrucci O, Tettanova O. Carta dei dissesti causati in Calabria meridionale dall'evento meteorologico dell'8-10 settembre 2000. Geodata n. 45 2004, CNR-GNDCI, Pubblicazione n. 2859, Rubbettino, Soveria Mannelli (CZ). (in Italian) 13. Borrelli L, Critelli S, Gullà G, Muto F. Rilievo del grado di alterazione di rocce cristalline: Presentazione della “Carta del grado di alterazione e dei movimenti in massa della porzione centro-occidentale del bacino del Fiume Mucone (Calabria, Italia)”. Geologi Calabria 2011;11:3-46 (attached “Carta del grado di alterazione e dei movimenti in massa della porzione centro-occidentale del bacino del Fiume Mucone (Calabria, Italia)”, Scala 1:10.000). (in Italian) Giovanni Gullà / Procedia Earth and Planetary Science 9 ( 2014 ) 44 – 53 53

14. Gullà G, Antronico L, Borrelli L, Caloiero T, Coscarelli R, Iovine G, Nicoletti PG, Pasqua AA, Petrucci O, Terranova O. Indicazioni conoscitive e metodologiche connesse all’evento di dissesto idrogeologico dell’autunno-inverno 2008-2009 in Calabria. Geologi Calabria (Rivista dell’Ordine dei Geologi della Calabria) 2009,10(1):4-21. (in Italian) 15. Sorriso-Valvo M, Tansi C. Grandi frane e deformazioni gravitative profonde di versante della Calabria. Note illustrative della carta al 250.000. Geogr Fis Quat 1996;19:395.408. (in Italian) 16. Borrelli L, Cofone G, Gullà G. Procedura speditiva per la redazione di una carta del grado di alterazione a scala regionale. Rend Online Soc Geol It 2012;21:528-530. (in Italian) 17. Brunetti M, Caloiero T, Coscarelli R, Gullà G, Nanni T, Simiolo C. Precipitation variability and change in the Calabria region (Italy) from a high resolution daily dataset. Int J Climatology 2012;32:57-73. 18. Amodio-Morelli L, Bonardi G, Colonna V, Dietrich D, Giunta G, Ippolito F, Liguori V, Lorenzoni S, Paglionico A, Perrone V, Piccarreta G, Russo M, Scandone P, Zanettin-Lorenzoni E, Zuppetta A. L'Arco Calabro–Peloritano nell'orogene appennino-magrebide. Memorie della Società Geoligica Italiana 1976;17:1-60. (in Italian) 19. Tortorici L. Analisi delle deformazioni fragili dei sedimenti postorogeni della Calabria settentrionale. Bollettino Società Geologica Italiana 1982;100:291-308. (in Italian) 20. Cascini L, Critelli S, Di Nocera S, Gullà G. A methodological approach to landslide hazard assessment: A case history. In: Proc. of the 6th International Symposium on Landslides; Christchurch, 10-14 February 1992, Vol. 2: 899-904. 21. Borrelli L, Greco R, Gullà G. Weathering grade of rock masses as a predisposing factor to slope instabilities: Reconnaissance and control procedures. Geomorphology 2007;87:158-175. 22. Cascini, L. 1986. Movimenti discontinui di una coltre di detrito della Sila Grande. In;: Atti 16th Convegno Nazionale di Geotecnica; Bologna, Italy, November 1986; Vol. 1. p. 259-262. (in Italian) 23. Cascini L, Gullà G. Caratterizzazione fisico-meccanica dei terreni prodotti dall'alterazione di rocce gneissiche. Rivista Italiana di Geotecnica 1993;XXVII( 2):125-147. (in Italian) 24. Cascini L, Gullà G. Sulle acque sotterranee di una coltre di detrito in frana. In: Atti del Convegno cartografia e monitoraggio dei movimenti franosi (Sessione Monitoraggio); C.N.R.-G.N.D.C.I., Bologna, 10-11 Novembre 1988, 85-95. (in Italian) 25. Cascini L, Versace P. Relationship between rainfall and landslides in a gneissc cover. In: Bonnard C, editor. Proceedings of the 5th International Symposium on Landslides 1988; Lausanne, Switzerland, 10-15 July 1988. A.A. Balkema, Rotterdam, The Netherlands; Vol. 1. p. 565-570. 26. Cascini E, Cascini L, Gullà G. A back-analysis based on piezometers response. In: Proc. of the 6th International Symposium on Landslides 1992, Christchurch, 10-14 February 1992, Vol. 2. p. 1123-1128. 27. Gullà G, Sorbino G. Soil suction measurements in a landslide involving weathered gneiss. In: Proc. of the 7th International Symposium on Landslides 1996; 17-21 June 1996, Trondheim, Norway. p. 749-754. 28. Cascini L, Gullà G, Sorbino G. Groundwater modelling of a weathered gneissic cover. Canad Geotech J 2006;43:1153-1166. 29. Gullà G, Aceto L, Antronico L, Cilento M, Niceforo D, Perna E, Terranova O. Failure and post failure conditions of a landslide involving weathered and degraded rocks. In: Proceedings of the 9th International Symposium on Landslides 2004; Rio de Janeiro, June 28 - July 7, 2004, 1241-1246. 30. Iannace A, Bonari G, D’Errico M, Mazzoli S, Perrone V, Vitale S. Structural and tectonic evolution of the Apennine Units of northern Calabria. Tectonics 2005;CR Geosci 337:1541-1550. 31. Gullà G, Niceforo D, D’Onofrio D, Reali C. Monitoraggio conoscitivo degli spostamenti con tecniche GPS e gestione di una condizione di criticità per la frana di Serra di Buda (Acri-CS). Atti della 9^ Conferenza Nazionale ASITA 2005, Catania, 15-18 novembre 2005. (in Italian) 32. Sorriso-Valvo M, Gabriele S, Gullà G, Antronico L, Tansi C, Greco R, Fantucci R. Studio geologico-geomorfologico-geotecnico e monitoraggio della frana di Serra di Buda (Acri). CNR-IRPI_Sezione di Cosenza, Regione Calabria-Assessorato LL.PP. e Acque, Autorità di Bacino Regionale 2005; Rubbettino Industrie Grafiche ed Editoriali, pp. 60. (in Italian) 33. Borrelli L, Perri F, Critelli S, Gullà G. Minero-petrographical features of weathering profiles in Calabria, southern Italy. Catena 2012;92:196- 207. 34. Borrelli L. Categorie di frane ed elementi caratteristici nei profili di alterazione. PhD Thesis. University of Calabria. 2008; Dottorato di ricerca in Scienze della Terra (XX Ciclo): 395 pp. 35. Borrelli L, Gullà G. Modello geologico-tecnico della frana di Serra di Buda (Acri - Cs). Rend Online Soc Geol It 2012;21:525-527. (in Italian)