RECENT ADVANCES IN SEISMIC MICROZONATION STUDIES IN COLOMBIA: THE MANIZALES CITY CASE

13 th World Conference on Earthquake Engineering Vancouver, B.C., Canada August 1-6, 24 Paper No. 284 RECENT ADVANCES IN SEISMIC MICROZONATION STUDIES IN COLOMBIA: THE MANIZALES CITY CASE Luis YAMÍN 1, Omar D. CARDONA 2, Mauricio GALLEGO 1, Camilo PHILLIPS 1, Santiago ARÁMBULA 1 SUMMARY The article presents the recent advances on the seismic microzonation studies that have been carried out for some Colombian cities, making emphasis in the microzonation study for the city of Manizales, which is located on volcanic ash soils. An analysis of the Colombian territory's seismological variables was carried out and the attenuation laws for different types of seismological sources were defined based on regional recent strong motion records. The non-linear dynamic properties of the city's characteristic soils were studied using in situ and laboratory dynamic tests. A non-linear transfer function evaluation model for the volcanic ash deposits was developed in order to obtain Fourier Amplitude Spectra (FAS) at ground surface level. Using the FAS at surface level, uniform hazard spectra were constructed at different locations according to the classic hazard theory. Simultaneously, sensibility analyses for the different variables involved were carried out as well as Montecarlo simulations of non -stationary processes. Bidimensional analyses were also performed in order to evaluate the amplification effects generated by abrupt topographic characteristics of the ground on which the city has been built. Artificial earthquakes were generated for the time response analysis of the soil deposits and representative structures using the seismological theory of the Omega Squared model following a Green empirical function procedure. Recommendations for the installation of a seismic instrumentation network were made in order to improve the proposed models. The spectra establish the basis for the city's seismic design code and city's earthquake loss scenarios for emergency response planning. Finally, taking into account disaster risk management and insurance public policies, a seismic risk evaluation was performed on a representative group of government's building infrastructure in order to establish the optimum seismic risk transfer and retention strategies for the city. INTRODUCTION The three major cities in the Colombian Coffee Growing Region are, in their order, Pereira (38 inhabitants), Manizales (37 inhabitants) and Armenia (27 inhabitants). Considering the high seismic activity characteristic of the zone, these three cities are undertaking, from some years ago, seismic microzonation studies aimed to estimate the expected seismic response of the several distinctive zones for purposes of structural design and rehabilitation, to estimate damage and loss future scenarios and to 1 Universidad de Los Andes, CITEC, Bogotá, Colombia, lyamin@uniandes.du.co 2 Universidad de Los Andes, CEDERI, Bogotá, Colombia, ocardona@uniandes.edu.co

develop prevention and emergency management plans. The following general methodology has been used to carry on the above mentioned microzonation studies: Conduction of basic studies, including the following: regional and local geology, neotechtonics, historical and instrumental seismicity, accelerographic local network set up, geophysical characterization (seismic refraction and microtremors), geotechnical characterization by compiling previous studies, geomorphology, topography, and cartography. Assessment of local seismic hazard and definition of design earthquakes, based on existing studies at national level [1, 2, 3, 4, 5] and the existing instrumental and historical seismic information. Detailed geotechnical investigation by means of borings in the several different characteristic zones of the city, field tests to measure wave velocities and laboratory and in situ tests for soil dynamic behavior characterization. Assessment of dynamic response of the different characteristic deposits by using unidimensional analytical models. Evaluation of dynamic response related to geometric effects such as fills, slopes and hill zones, by using two-dimensional analytical models. Definition of earthquake-resistant uniform hazard design spectra. Instrumentation (using accelerographic networks) in order to calibrate the analytical results. Seismic risk evaluation Seismic risk management strategies SEISMIC HAZARD ANALYSIS An analysis of the Colombian territory's seismological variables was carried out. Figure 1 presents the analytical model developed to evaluate the seismic hazard for Colombia. The most important seismic sources are shown. Each one of them is characterized using the seismic information available from the national accelerographic network. Bahia Solano CHOCÓ Itsmina Buenaventura Quibdó San Pablo Dagua Condoto Novita Sipi VALLE DEL CAUCA Cali Restrepo Yumbo Caicedo Copacabana Caracoli Bello San Vicente Itagui Peñol Guatape Urrao Ebejico San Carlos Granada Pto Nare Betulia Medellín Amaga Cocorna La Union SanLuis Salgar venecia La Ceja Pto Triunfo Fredonia Montebello Bolivar San Francisco Santa Barbara ANTIOQUIA Abejorral Jericó Sonson Argelia jardin Andes Aguadas La Dorada LLoro Nariño Caramanta Samana Pacora Bagadó Supia Riosucio Pensilvania Marmato Salamina Mistrato CALDAS Victoria Quinchia Filadelfia Marquetalia Tadó Pto Rico Marulanda Arnzazu Manzanares RISARALDA Mariquita Anserma Neira Fresno Apia Viterbo Manizales Santuario Palestina La Celia Villamaria San José del Palmar El Aguila Balboa Dosquebradas Ansermanuevo Pereira Santa Rosa de Cabal El Cairo Argelia Cartago Alcala Toro Versalles Salento Filandia La Union El Dovio Vijes Murindo Roldanillo Garrapatas Trujillo San Pedro Guacari Ginebra Bugalagrande Buga El Cerrito Palmira Tulua Zarzal Riofrio Cauca Cauca Benniof Intermedia Sevilla Armenia Calarca QUINDIO Caicedonia Rioblanco Pijao Genova Cordoba Chaparral Romeral Roncesvalles Cajamarca San Antonio TOLIMA Ataco Puerto Salgar HUILA BOYACA Caparrapi SANTANDER Tausa CUNDINAMARCA Honda Guaduas Nocaima Pacho Suesca Falan Supata Herveo Libano Armero Sasaima Nemocon Sesquile San Francisco Chaguani Alban Tabio Guatavita Tocancipa Lerida Bituima Sopó Guayabal Manta Tenjo Chia Murillo Villeta San Juan Santa Rosa de Cabal Facatativa Guasca Zipacon Cota Gacheta Ambalema Madrid Quipile Funza La Calera Santa Isabel Venadillo Anolaima Bojaca Puli Mosquera Tena Beltran Anzoategui San Antonio Choachi La Mesa Santa Isabel Jerusalen Soacha El Colegio Fomeque Coello Ubaque Apulo GranadaSibate Anapoima Chipaque Ibague Guataqui Silvania Caqueza Meta Nariño Une Fosca El Calvario Agua de Dios Fusagasuga Girardot Tibacuy Pasca Quetame Melgar Rovira Arbelaez Gutierrez Flandes Venecia Guayabetal Pandi Espinal San Bernardo Icononzo Acacias Villavicenc San Luis Venecia Guamo Carmen de Apicala Cubarral Guamal Villarica Ortega Saldaña Cabrera Ibague Coyaima Alpujarra Palestina Magdalena Prado Dolores Natagaima Purificación Colombia Pto. Boyaca Utica Yacopi La Uribe La Palma Nimaima Topaipi Mesetas Otanche Pauna Tunungua Pauna Maripi Paime Bogotá D.C. Bolivar La Belleza Muzo Supatasa Villagoméz Lejanias Frontal META El Castillo Landazuri Albania El Peñon Puente Nacional Saboya Caldas Simijaca Susa Fuquene Ubate Salinas Cucunaba Choconta Macheta Figure 1: Analytical model for seismic hazard analysis

Itsmina Buenaventura San Pablo Dagua Condoto Novita Sipi Restrepo Yumbo Caicedo Copacabana Caracoli Bello San Vicente Landazuri Itagui Peñol Guatape Urrao Ebejico San Carlos Granada Pto Nare El Peñon Betulia Amaga Cocorna SanLuis Bolivar La Union Salgar venecia La Ceja Pto. Boyaca Pto Triunfo Fredonia Montebello La Belleza Bolivar San Francisco Santa Barbara Abejorral Albania Puente Nacional Jericó Sonson Otanche Argelia Pauna Saboya jardin Tunungua Andes La Dorada Yacopi Aguadas Nariño Pauna Caldas LLoro Caramanta Samana Maripi Pacora Bagadó Simijaca Supia Riosucio Pensilvania Muzo Caparrapi Susa Marmato Salamina Paime Fuquene Mistrato Victoria Ubate La Palma Filadelfia Marquetalia Cucunaba Quinchia Supatasa Tadó Pto Rico Topaipi Marulanda Villagoméz Choconta Arnzazu Manzanares Utica Nimaima Mariquita Tausa Anserma Neira Fresno Macheta Apia Honda Viterbo Guaduas Nocaima Pacho Suesca Falan Supata Herveo Santuario Palestina Libano Armero Sasaima Nemocon Sesquile Villamaria San Francisco La Celia Chaguani Alban Guatavita Tabio San José del Palmar Tocancipa Lerida Bituima Sopó Balboa Guayabal Manta El Aguila Villeta Tenjo Chia Dosquebradas Murillo San Juan Santa Rosa de Cabal Facatativa Guasca Zipacon Cota Ansermanuevo Santa Rosa de Cabal Gacheta Ambalema Madrid El Cairo Quipile Funza Venadillo Anolaima Bojaca La Calera Argelia Santa Isabel Cartago Puli Alcala Mosquera Toro Tena Versalles Salento Beltran Anzoategui San Antonio Choachi Filandia La Mesa Soacha Santa Isabel Jerusalen El Colegio Fomeque La Union Coello Ubaque El Dovio Calarca Apulo Granada Sibate Anapoima Chipaque Cajamarca Guataqui Silvania Caqueza Cordoba Nariño Zarzal Une Fosca El Calvario Pijao Agua de Dios Fusagasuga Roldanillo Caicedonia Girardot Tibacuy Pasca Quetame Melgar Genova Rovira Sevilla Arbelaez Gutierrez Trujillo Bugalagrande Flandes Venecia Guayabetal Pandi Espinal San Bernardo Tulua Roncesvalles Icononzo Acacias Villavicencio San Luis Venecia Guamo Carmen de Apicala Riofrio San Pedro Guamal Cubarral San Antonio Villarica Ortega Saldaña Cabrera Vijes Guacari Ginebra El Cerrito Palmira Buga Rioblanco Chaparral Ataco Coyaima Alpujarra Prado Dolores Natagaima Purificación Colombia La Uribe Mesetas Lejanias El Castillo Attenuation laws for different types of seismological sources were defined based on regional recent strong motion records. Using a probabilistic integration scheme, exceedence rates for firm soil deposits are estimated for the city. Figure 2 illustrates the different acceleration exceedence rates for firm soil for each one of the seismic sources of the zone. In addition, the maximum acceleration exceedance rates considering the integrated effects of all relevant seismic sources is presented as well as the relative participation factor for each one of them in the global hazard curve for the city. 1 Tasas de Excedencia en Roca para Manizales Participación de las Fuentes para un Tr = 475 años Participación 7% 6% 5% 4% 3% 2% 1 1 Aceleración (gal) 1 Total Subducción Benioff I Benioff I Benioff III Benioff Prof Cauca Frontal Ibague Palestina Romeral 1% % Romeral Benioff II Subducción Benioff Prof Bennioff I Fuente Figure 2: Exceedance rates for relevant seismic sources and relative participation factors Following the general random vibration theory, different seismic intensity maps such us acceleration, velocity, displacement or any other related parameter, are plotted for different return periods. Figure 3 presents the maximum probable acceleration in firm soil for a 475 year return period, corresponding to the design return period established by local structural design codes. Simultaneously, uniform hazard spectra are presented for the city, for different periods of analysis return. Murindo Benioff III Palestina Cauca Latitud 6 5 5 4 4 3 CHOCÓ Quibdó VALLE DEL CAUCA Cali Aceleración Máxima del suelo; Tr=475 años, (gal) ANTIOQUIA RISARALDA Pereira Armenia QUINDIO Medellín Manizales CALDAS TOLIMA Ibague HUILA Puerto Salgar BOYACA SANTANDER CUNDINAMARCA Bogotá D.C. META -77-76 -76-75 -75-74 -74 Longitud Meta Pseudoaceleración (gal) 7 6 5 4 3 2 1 Espectros de Amenaza Uniforme; Manizales 1 1 2 2 3 Periodo (Segundos) Tr=5 años Tr=1 años Tr=2 años Tr=475 años Tr=1 años Figure 3: Maximum probable acceleration for firm soil in the area of Manizales and uniform hazard spectra for different return periods

Based on the hazard information available, artificial earthquakes were generated for the time response analysis of the soil deposits and representative structures using the seismological theory of the Omega Squared model following a Green empirical function procedure. Some representative records are presented in Figure 4. Figure 5 presents the response spectra of the artificial earthquakes generated as compared to the design spectrum defined by the local code. Aceleración (g).3.2.1 -.1 -.2 -.3 1 2 3 4 5 Tiempo (seg) Aceleración (g).2.15.1 5-5 -.1 -.15 -.2 1 2 3 4 5 Tiempo (seg).2.15.1.2.15.1 Aceleración (g) 5-5 Aceleración (g) 5-5 -.1 -.1 -.15 -.15 -.2 -.2 2 4 6 8 1 2 4 6 8 1 Tiempo (seg) Tiempo (seg) Figure 4: Representative artificial records Sa (g) 1.75 Espectros de Aceleración Sismo Romeral Deconvolución Sismo Romeral Sintético Sismo Benioff Sintético Sismo Benioff Calima NSR - 98 S=1 NSR - 98 S=2.25 1 1 2 2 3 Figure 5: Acceleration response spectra for artificial earthquakes DYNAMIC SOIL RESPONSE From the stand point of dynamic response analysis, it is necessary to characterize surface deposits which due to its dynamic behavior features (wave velocity, stiffness or density) show a contrasting condition with deeper deposits. The City of Manizales is located at the western flank of the Colombian Central

mountain range, wherein there are found more than 23 of the identified active volcanoes of the country, with high activity in the past. For such reason, surface deposits of the entire zone are characterized by the presence of volcanic ashes or pyroclastic fall deposits of variable thickness, ranging from 3 m to 35 m. These materials were deposited over Tertiary and Cretaceous units or over more recent deposits made up by debris flows, alluvial or colluvial deposits. Also plenty of hydraulic, mechanical, or sanitary fills are found at surface level with variety of geometry and high heterogeneous distribution. The fills were used to form flat zones in initially wavy topography with the purpose of developing housing construction activities. In order to study the dynamic properties of these soils, a comprehensive boring plan was carried out as appears illustrated in Figure 6. A total of 24 boreholes were carried out at different depths which were supplemented with more than 2 previous boreholes. Figure 6: Distribution of boreholes Geotechnical information as well as microtemors measurements carried out in different locations of the city allowed to develop a thickness map of superficial volcanic ash soils deposits. Based on the information obtained, recommendations for the installation of a seismic instrumentation network were made in order to improve the proposed models. The location of the accelerographs should match the representative explored locations of the city selected from the points indicated in Figure 6. The installation of at least fifteen (15) strong motion accelerographs at surface level and at least two (2) at base rock level was recommended in order to determine the characteristics of typical earthquake records and the empiric transfer functions for each one of the points of analysis within the study area. The non-linear dynamic properties of these characteristic soils deposits were studied using in situ and laboratory cyclic tests on unaltered samples. Cyclic triaxial tests (standard ASTM D5311-92), resonant column tests (standard ASTM D415-92), and wave velocity tests on unaltered samples confined in triaxial chamber ("Bender Element", standard ASTM D2845-95) were made. In addition, seismic cone in situ test, cyclic in situ presionmeter test and down hole shear wave velocity tests were performed. Findings from the tests conducted show in general a high variability of ash dynamic behavior. Typical results from in situ cyclic presiometer test as well as correlation between different tests is presented in Figure 7.

25 2 15 1 5 Antigua alcaldia 14.8 m El cable 5m el cable 8m el cable 12.8 m 2 4 6 8 1 12 14 16 u (mm) 1.9.8.7.6.4.3.2.1 1 1 1 1 1.1 1 Bender Element Triaxial cíclico Columna Resonante Presiometro Figure 7: Typical results from in situ cyclic presiometer test and correlation between different tests γ UNIFORM HAZARD SPECTRA A non-linear transfer function evaluation model for the volcanic ash deposits was developed in order to obtain Fourier Amplitude Spectra (FAS) at ground surface level. Using the FAS at surface level, uniform hazard spectra were constructed at different locations according to the classic hazard theory. Simultaneously, sensibility analyses for the different variables involved were carried out as well as Montecarlo simulations of non-stationary processes. The spectra establish the basis for the city's seismic design code. Figure 8 illustrates uniform hazard spectra at different locations of the city as compared with deterministic analysis using classical non-linear equivalent one-dimensional response analysis with four different accelerographic artificial records. Espectros de Aceleración en Superficie Sondeo 1 Romeral Deconvolución Espectros de Aceleración en Superficie Sondeo 5 Romeral Deconvolución Romeral Sintético Romeral Sintético Benioff Sintético Benioff Calima Benioff Sintético Benioff Calima Espectro de Peligro Uniforme Tret = 475 años Espectro de Peligro Uniforme Tret = 475 años 1 1 2 2 3 1 1 2 2 3 Espectros de Aceleración en Superficie Sondeo 13 Espectros de Aceleración en Superficie Sondeo 4 Romeral Deconvolución Romeral Deconvolución Romeral Sintético Romeral Sintético Benioff Sintético Benioff Calima Benioff Sintético Benioff Calima Espectro de Peligro Uniforme Tret = 475 años Espectro de Peligro Uniforme Tret = 475 años 1 1 2 2 3 1 1 2 2 3 Figure 8: Uniform hazard and deterministic spectra for different locations

TOPOGRAPHIC EFFECTS The effects of the highly abrupt superficial topography of the city in the surface dynamic response, is studied. A two-dimensional model is resolved by using QUAD-4M computer program [6, 7]. Figure 9 shows the scheme of the model developed and results regarding maximum superficial acceleration for different locations in the modeled slope shown. Figure 1 presents acceleration response spectra and amplification factors for different locations. Amplification factors are evidenced for the period range under analysis in relation with the equivalent one-dimensional response. The analysis is made for several different points along the slope. 4 Aceleración Máxima del Terreno Modelo 5 - h=3 m y θ=3º 3 2 1 4 8 12 16 2 Longitud (m) Romeral Deconvolución Benioff Sintético Benniof.ace Benioff Calima Figure 9: Two- dimensional model and maximum acceleration at different locations Factor de Amplificación 2 1.75 1 1.25 1.75 Amplificación Modelo 3m 45 grados Romeral por Deconvolución.25 1 1 2 2 3 Punto 1 Punto 2 Punto 3 Punto 4 Punto 5 Punto 6 Punto 7 Punto 8 Punto 9 Punto 1 Punto 11 Punto 12 Punto 13 Punto 14 Punto 15 Punto 16 Punto 17 Punto 18 Punto 19 Punto 2 Punto 21 Punto 22 Punto 23 Aceleración (g) 1.2 1.8.6.4.2 Espectros de Aceleración Modelo 3m 45 grados Romeral por Deconvolución 1 1 2 2 3 Shake 91 Punto 1 Punto 2 Punto 3 Punto 4 Punto 5 Punto 6 Punto 7 Punto 8 Punto 9 Punto 1 Punto 11 Punto 12 Punto 13 Punto 14 Punto 15 Punto 16 Punto 17 Punto 18 Punto 19 Punto 2 Punto 21 Punto 22 Punto 23 Figure 1: Acceleration response spectra and amplification factors at different locations

SEISMIC MICROZONATION GIS Considering the high variability of soils deposits, both from the geometric and physical, mechanical and dynamic properties point of view, the results of seismic microzonation are presented using a geographical information system. The system allows the management of all the city s geographical information including general cartography, rivers, roads, places of interest, location of essential and special buildings, geology, geotechnical and in general all the related information available. The system allows visualizing the different seismic intensity parameters (acceleration, velocity or displacement) for any return period and for any structural period of analysis. Simultaneously, for each specific location the system calculates, by means of interpolation, the acceleration spectrum for different return periods and for the damping coefficient selected. One of the main advantages of the system is that it specifies the design spectra for each location, eliminating the use of traditional seismic microzones which generally bear a considerable overdesign due to the need of specifying enveloping spectra for the specified uniform microzones. Figure 11 presents the main window of the Geographical Information System to calculate seismic hazard for the city and the typical results given by the system for a particular location. Figure 11: System main window and typical results SEISMIC RISK EVALUATION A seismic risk evaluation was performed on a representative group of government s constructions in order to establish the basis for more detailed analyses related to general seismic risk management for the city. To carry out seismic risk assessment and to evaluate expected losses caused by earthquakes, a modified version of the RN-COL v 2.1 system developed by Evaluación de Riesgos Naturales, ERN Colombia, was used. This system includes the most comprehensive information available on the country s seismic conditions. The system was integrated to the geographical information system for hazard evaluation in the city, including subsoil amplification effects. Additionally, the system includes the information that characterizes building typologies more often used in the city. The geographical location of the 4 main city buildings analyzed, corresponding to the municipality s public buildings is presented in Figure 12.

Figure 12: Geographical location of analyzed public risks Number of Buildings 15 14 13 12 11 1 9 8 7 6 5 4 3 2 1-5 5-1 1-2 2-5 5-1 1-15 15-2 2-5 Insured Value (Thousands $USD) Figure 13: presents a frequency curve of insurable values of the different buildings analyzed. The results of the analysis are presented in Figures 14 to 16. Figure 14 shows a histogram of expected losses for the buildings analyzed. These results indicate that expected values of the individual loss of the buildings present a great variability with values ranging from 8% to 96% with respect to insurable values. Figure 15 presents a concentration curve of the group of buildings. This curve indicates that about 5% of the most critical buildings concentrate nearly 85% of risk for the city administration. On the other hand, Figure 16 depicts probable maximum loss variation, PML, for the group of buildings as a function of the return period of analysis. It is concluded that the city s public buildings PML varies between 31% and 41% for a range of return periods between 5 and 15 years. These results set the basis for a risk transfer and retention strategy for the city administration.

Expected Loss (%) 1 9 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 1 11 12 13 14 15 16 17 18 19 2 21 22 23 24 25 26 27 28 29 3 31 32 33 34 35 36 37 38 39 4 Building Number Figure 14: Expected losses for the buildings analyzed 111% 9% 8% 7% 6% 5% 4% 3% 2% 1% % Cumulative prime Concentration Index IC =.91 % 2% 4% 6% 8% 1% Number of buildings Figure 15: Concentration curve of the group of buildings PML (%) 6 5 4 3 2 1 37.8 31.2 2.1 8.9 4.9 43.1 49.7 1 2 3 4 5 Return period (years) Figure 16: Probable Maximum Loss variation

CONCLUSIONS Based on the previous findings, the following conclusions can be drawn: Uniform hazard spectra resulting from an analysis using the random vibrations theory together with a non-linear transfer function evaluation model developed in order to obtain Fourier Amplitude Spectra (FAS) at ground surface level, indicates very similar results as compared with acceleration spectra resulting from the classical non-linear equivalent one-dimensional response analysis with a group of accelerographic artificial records. The characterization of soils dynamic properties for volcanic ash deposits requires different laboratory and field tests in order to adequately characterize their behavior for the entire range of deformations. The results given by different laboratory tests such as wave velocity, cyclic triaxial and resonant column present an acceptable agreement with equivalent results obtained from field tests such as down-hole, cyclic presiometer and seismic cone test. Abrupt topography can generate geometric amplification effects that can be in the order of 1.2 to 1.3 in the range of structural periods between.2 and seg (rigid buildings) depending on the relative location of the building within the topographical profile. The results of hazard analyses in microzonation studies are interpreted in a more convenient way by using a geographical information system (GIS) which can evaluate the design spectra at each exact location by means of continuous interpolation, thus eliminating traditional seismic microzones which generally bear a considerable over design. The results of seismic risk evaluation for public buildings of the city of Manizales indicate that expected values of the individual loss of the buildings present a great variability with values ranging from 8% to 96% with respect to insurable values. City s public buildings PML varies between 31% and 41% for a range of return periods between 5 and 15 years. These results set the basis for a risk management strategy for the city. REFERENCES 1. AIS, Estudio General de Amenaza Sísmica de Colombia. Santa Fe de Bogotá, Colombia. October 1996, 254pp. 2. CARDER, Exploración geotécnica, Investigación de laboratorio y Microzonificación sísmica de las áreas urbanas y suburbanas de las ciudades de Pereira, Dosquebradas y Santa Rosa de Cabal. Pereira, Colombia. 1999, pp3-37. 3. OMPAD, Estudios de amenaza sísmica y evaluación sismogeotécnica preliminar para la zonificación sísmica de Manizales. Colombia, March 1996, pp39-65. 4. OPES, Proyecto Microzonificación Sísmica de Santa Fe de Bogotá. Santa Fe de Bogotá, Colombia. November 1996, pp 57-63. 5. Universidad de los Andes, Microzonificación Sísmica de la Ciudad de Manizales, Alcaldía de Manizales, Colombia, 22. 6. Hudson, M., Idriss, I. and Beikae, M, QUAD-4M: A computer program evaluate the seismic response of soil structures using finite element procedures and incorporating a compliant base. Davis: University of California. May 1994, 68pp. 7. Idriss, I., and Sun, J., SHAKE-91: A computer program for conducting equivalent linear seismic response analyses of horizontally layered soil deposits. Davis: University of California. November 1992, 35pp.