Manuscript Click here to access/download;Manuscript;Final_Manuscript.docx A Strong-Motion Database of Costa Rica: 20 Years of Digital Records Aarón Moya-Fernándeza, Luis A. Pinzónb, Victor Schmidt-Díaza, Diego Hidalgo-Leivaa* and Luis G. Pujadesb a Earthquake Engineering Laboratory, Universidad de Costa Rica, San José, Costa Rica b Department of Civil and Environmental Engineering, Universitat Politècnica de Catalunya, Barcelona, Spain *Contact information of correspondence author: Diego Hidalgo-Leiva, Ph.D. Address: Laboratorio de Ingeniería Sísmica, Ciudad de la Investigación, Universidad de Costa Rica, San Pedro, San José, Costa Rica Mobile phone: +506 2511-6675 Email: diego.hidalgo@ucr.ac.cr 1 1 Abstract 2 In this paper, we present a strong-motion database from earthquakes recorded by the Earthquake Engineering 3 Laboratory at the University of Costa Rica. The database consists of 2471 three-component accelerograms from 155 4 digitally recorded events. It covers the last 20 years of measurements, including records from the Nicoya earthquake 5 of Mw 7.6 on 2012 September 05. The engineering and seismological communities can use this data either to conduct 6 new research or to improve seismic or hazard studies in the region. A catalog is also available with metadata of each 7 record containing several intensity measures from the ground-motion time-histories. 8 Introduction 9 The convergence of the Cocos and Caribbean plates along the Pacific coast of Costa Rica is the major 10 source of seismic activity for the country (Alvarado et al., 2017; Arroyo et al., 2017). As a result, many 11 earthquakes occur along the subduction zone as well as active volcanism in the continental part. The outer 12 slope side of the place generates normal faulting while reverse faulting takes place at depths between 15 13 and 50 km (Quintero and Güendel, 2000; DeShon et al., 2003; Norabuena et al., 2004). At depths between 14 50 and 280 km, intraplate or intra-slab earthquakes (deep subduction) occur and in general normal type 15 mechanisms predominate (Guendel and Protti, 1998). 16 The Benioff zone gets shallower in the southern part of Costa Rica, where the Cocos mountain range 17 subducts. The Panama Fracture Zone is a dextral fault system that separates the Cocos plate from the Nazca 18 plate (Schmidt-Díaz, 2014). At the southern end of the Burica Peninsula lies the triple junction where the 19 Cocos, Caribbean and Panama block meet. There is also a high number of seismic events that take place 20 along the Northern Panama Deformed Belt (NPDB) and Central Costa Rica Deformed Belt (CRDB). These 21 are a series of cortical deformation zones with a high density of active faults (Goes et al., 1993; Guangwei 22 Fan et al., 1993; Montero, 2001). This complex tectonic framework has resulted in numerous destructive 23 earthquakes (i.e., 1991 Limon Mw 7.7; 2012 Nicoya Mw 7.6), and, consequently, a concern to develop and 24 improve the seismic hazard and risk studies of the country. 2 25 The Earthquake Engineering Laboratory at the University of Costa Rica (LIS-UCR for its acronym in 26 Spanish) started operations in 1983. That year, the United States Agency for International Development 27 (USAID), donated several SMA-1 Kinemetrics strong-motion accelerographs to Costa Rica. They were 28 located along the Pacific coast and the highly populated Central Valley. That was known as the Faculty of 29 Engineering's Accelerographic Network. It was an analog network, which meant that after a strong 30 earthquake took place, the collection and processing of the information took several days to weeks to get 31 ready for analysis. 32 In 1989 the name was changed to LIS-UCR. New digital instruments were acquired, and the geographic 33 coverage of the stations increased. At the time of writing this document, the LIS-UCR has more than 160 34 digital, 24-bit strong-motion units located in free-field conditions, boreholes, and inside buildings. The LIS- 35 UCR is in charge of recording, processing and storing all acceleration records for academic and research 36 purposes. The accelerograms used in this document were recorded only by sensors in free-field conditions. 37 The time span for the database provided in this paper ranges from 1998 to the present. The objective of this 38 article is to give an overview and provide easy access to this database, and therefore, expanding its use on 39 research. 40 Strong-Motion Network 41 The strong-motion network of the LIS-UCR began operating in 1983 with the installation of SMA-1 42 Kinemetrics analog sensors. In June 1991, digital processing started with the installation of several SSA-2 43 Kinemetrics type sensors. In 2010 many analog instruments were replaced by Ref Tek technology, and in 44 2012 Güralp and Nanometrics sensors were also added to the network. Nowadays, there are a total of 130 45 free-field stations [most of them with FBA (force-balanced-accelerograph) sensors, but MEMS (Micro- 46 Electro-Mechanical Systems) as well] as shown in figure 1. 47 There are four soil types according to the Costa Rican Seismic Code (CRSC) (CFIA, 2016). This 48 classification is similar to that proposed in the ASCE 7-16 (ASCE, 2017) with some differences. Soil types 3 49 A and B are called S1 (rock). Soil types C, D, and E of ASCE 7-16 are equivalent to S2 (stiff soil), S3 (soft 50 soil) and S4 (very soft soil). There is no F type of soil in the CRSC classification. 51 Due to the complexity of data acquisition and the cost of the geotechnical studies, we used the classification 52 method proposed by Zhao et al. (2006). The results can be found in Schmidt-Díaz (2011). The method is 53 based on the horizontal-to-vertical (H/V) 5% damped response spectral ratio. From that, the fundamental 54 period can also be obtained. We then used a classification index for each station. When available, geological 55 and geotechnical information was also used as a reference. 56 A total of 42.0% stations are classified as soft soil (S3), 33.1% as stiff soil (S2), 17.2% are classified as 57 very soft soil (S4), and 7.7 % as rock sites (S1). In order to get a better site characterization, we are also 58 conducting MASW measurements to define the Vs30 parameter. Currently, 35 stations have Vs30 and we 59 are conducting measurements in 30 more stations (data available upon request via email). Figure 2 shows 60 the site classification described above. There is also a table with a summary of the site conditions available 61 at the LIS-UCR website (see Data and Resources). 62 Strong-Motion Database 63 The strong-motion database we present here has a total of 2471 three-component accelerograms. They 64 correspond to 155 earthquakes recorded from 1998 to the present. The database is being updated 65 automatically with new events as they trigger the Accelerographic Monitoring System (SMA in Spanish, 66 Moya-Fernández, 2018). Figure 3 shows the distribution of ground motion recordings per year. The number 67 has increased in recent years because at present there are more stations. 68 The SMA threshold requires that 30 stations surpass a value of 10 defined as follows: Every 15 seconds the 69 SMA computes the PGA at every station for the last 60 seconds and stores its value. Only the NS component 70 is used. It compares the PGA from the current minute (PGAC) and the previous one (PGAP). If the ratio 71 (PGAC/PGAP) is larger than 10 in 30 sites, the SMA processing begins. PGA is computed for the three 72 components of every station once the SMA gets activated using the whole waveform. Once the SMA closes 4 73 the event, we select the records that have at least one horizontal PGA greater than 2 gals in order to include 74 them in the final database. 75 The earthquake’s location (coordinates in WGS84 system), depth, and magnitude are calculated 76 automatically by the SMA (Moya-Fernández, 2018). The magnitude used to characterize the database is 77 the moment magnitude (Mw). Table 1 shows a statistical summary of the number of records per different 78 ranges of magnitude, depth, and epicentral distance. Figure 4a shows the relation magnitude vs hypocentral 79 distance. The hypocentral distance for the events in the database ranges from 5 to 400 km. There are 1509 80 records from earthquakes with Mw ≥ 5 (61.1 %) (see Figure 4b). Figure 5 shows the location of the events 81 in the database. Only one of the recorded earthquakes has an Mw > 7, the 7.6 Mw Nicoya earthquake of 2012. 82 A total of 71 stations recorded that event which shook the whole country. The largest peak ground 83 acceleration was 1.6 g at GNSR station, which was the closest to the epicenter (Schmidt-Díaz et al., 2014). 84 The LIS-UCR stores strong-motion data in an ASCII format called “lis-format” (Moya-Fernández, 2006). 85 This is a special type of format developed for researchers and students to have access to time-series data. 86 The files contain a header of 34 lines with relevant station and earthquake information of each record, after 87 which there are the three independent columns corresponding to the north-south (N00E), vertical (UPDO), 88 and east-west (N90E) component. Metadata from the header includes the earthquake source (subduction or 89 local), site to event distance [epicentral (Repi), hypocentral (Rhypo), Joyner-Boore (Rjb) and the closest 90 distance to rupture (Rrup)], site condition, soil classification, among others. The Rjb and the Rrup were 91 computed following the methodology proposed by Thompson and Worden (2018). Earthquake source 92 information is given in the database as local (LOCAL) or subduction (SUBDU) type events. This 93 classification is a general one, and it is based on the epicenter location and depth. Earthquakes located place 94 along the Pacific coast are usually classified as subduction type events. Earthquakes further inland in the 95 rest of the country at shallow depths (less than 30 km) are classified as local ones. Deeper earthquakes 96 (more than 30 km) along the subducted Cocos plate are classified as intraslab (INSLB) events. We use 97 “UNDEF” for those earthquakes happening in complex tectonic settings or where a simple classification 5 98 cannot be made. The slab model for Central America from USGS was used to help define which events 99 happened along the subducted slab in Costa Rica (Hayes et al., 2012). This metadata is available in a catalog 100 on the LIS-UCR website (see Data and Resources). 101 102 Data Processing 103 Each station transmits real-time data to the LIS-UCR servers in miniSEED format. When an earthquake is 104 strong enough to trigger 30 stations, the SMA extracts a pre-defined time-window and converts waveform 105 data to SAC format (Goldstein et al., 2003) in cm/s2. A baseline correction is applied by removing the mean 106 value. After tapering on both ends, a second-order Butterworth bandpass filter is used. The SMA then 107 processes the source parameters, calculate peak values, and gathers station information and soil type to save 108 data into lis-format. Notice that the entire process is automatic, for that reason, the data is later inspected 109 by eye in order to identify events with a low signal-to-noise ratio or with processing issues. Records that 110 are not suitable are removed from the database. 111 Data processing is proposed to satisfy the requirements of an Engineering Strong Motion (ESM) database. 112 Frequency bandpass is set to include and overcomes the frequency range for civil structures. Over the years, 113 corner frequencies have change according with technology, equipment brands and internal requirements on 114 LIS. For example, before 1998, the LIS's network was made of Kinemetrics type instrumentation only. The 115 default filtering from the K2 and ETNA strong motion records from Vol2 format was 0.12 to 47 Hz. When 116 Reftek was introduced in 2010, the range was set at 0.1 to 40 Hz. After 2017, when the SMA took care of 117 the automatic signal processing, it was decided that the range 0.05 to 25 Hz best fitted the needs for most 118 engineering purposes in Costa Rica. In this way, new technologies such as Guralp and Nanometrics that 119 were later introduced could be used with common values. It is recommendable for the reader to take care 120 when this parameter is sensitive and read the corner frequencies for each record. 6 121 Intensity Measures 122 In addition to the database and the catalog, we computed a series of intensity measures (IMs) based on 123 ground motion time-histories (Table 2) and peak responses (Table 3). The IMs for each record are also 124 available in the LIS-UCR website (see Data and Resources). The IMs based on time histories are available 125 in a single table where each column represents a single IM. In the case of the IMs based on peak responses, 126 they were calculated with absolute spectral acceleration (SA) and a 5% damping. Despite the most 127 commonly used IM in Ground Motion Prediction Equations (GMPE) is the pseudo-spectral acceleration 128 (PSA), we estimate the SA with the Nigam and Jennings (1969) exact solution of the differential equation 129 governing the response. For small damping, these two IMs are equivalent (Chopra, 2007). They are 130 presented in single tables as a function of several oscillator periods. 131 We used the acceleration time-histories to calculate the IMs in Table 2. They have been widely used in the 132 development of ground-motion prediction equations and seismic hazard studies (Boore et al., 1997; 133 Watson-Lamprey and Boore, 2007; Mezcua et al., 2008; Schmidt-Díaz, 2014; Douglas, 2017), as well as 134 in the evaluation of expected damage (Park et al., 1987; Kostinakis et al., 2015; Muin and Mosalam, 2017). 135 Figure 6 shows the relation between PGA (PGAN00E, PGAN90E and PGAZ from Table 2) and hypocentral 136 distance in the database. There are 7413 individual time-histories corresponding to the 2471 records. Of 137 them, 39.5% have a PGA larger than 10 cm/s2. Comparing the mean values of several PGA definitions, we 138 got differences of 1.45% between PGALarger(3) and PGALarger(2), and 12.5%, 15.6% and 14.0% between 139 PGALarger(2) and PGAN00E, PGAN90E, and PGAGM respectively. Figure 7 shows the relation for the rest of the 140 IMs with hypocentral distance. 141 The IMs from peak responses in Table 3 are commonly used in the development of GMPE and hazard maps 142 (Douglas, 2017). The SAGM (where GM means geometric mean) has gained popularity in the development 143 of GMPEs in recent years (Douglas, 2003; Campbell and Bozorgnia, 2008; Bindi et al., 2011) because the 144 dispersion in the averaging procedure in GMPE is significantly reduced (Baker and Cornell, 2006; Watson- 145 Lamprey and Boore, 2007; Stewart et al., 2011). However, this IM has a dependence on the recording 7 146 sensor orientation (this means that if the recording sensor is oriented along the polarization direction, the 147 GM of the response spectra of the as-recorded ground motion tends to zero, Boore et al. 2006). The 148 SAGMRotDpp and the SAGMRotIpp developed by Boore et al. (2006) (where GM means geometric mean, Rot: 149 rotation, D and I: period-dependent and independent rotations, and pp is the percentile) were proposed in 150 order to eliminate the sensor orientation dependency of the SAGM. The IM SAGMRotIpp, for the 50th percentile 151 (SAGMRtI50), was used as an intensity parameter in the Next Generation Attenuation (NGA) project (Chiou 152 et al., 2008; Power et al., 2008). Later on, Boore (2010) proposed the usage of the orientation-independent 153 SARotDpp and SARotIpp IMs without computing the geometric mean. Finally, the SARotD50 IM was used to 154 develop the NGA-West2 (Boore et al., 2013; Bozorgnia et al., 2014) and NGA-East (PEER, 2015) GMPEs 155 models. 156 Figure 8 shows a comparison between the rotated spectra and the SARotD100 IM (following Boore (2010)) 157 for the 2012 Nicoya earthquake at GNSR station. It is clear from the figure that the SARotD100 is the envelope 158 of the rotated spectra. Because this IM represents the maximum value of the vector composition, it could 159 be used for the design (or risk assessment) of structures of special importance such as historical-cultural 160 heritage buildings or other high-risk constructions (Pinzón, Pujades, Hidalgo-Leiva, et al., 2018). Figure 9 161 shows the rest of IMs: SARotD100, SALarger, SAGMRotD50, SAGMRotI50, SARotD50, and SAGM calculated in the in 162 the range of 0.10s to 0.25s. SAGMRotD50, SAGMRotI50 and SARotD50 correspond to the median values of the 163 rotated spectra and have similar values compared to SAGM. SARotD100 and SALarger represent the maximum 164 spectral values. SARotD100 is 9% larger than SALarger on average for the entire database. A statistical summary 165 for all the IMs can be found on the LIS-UCR website (see Data and Resources). 166 Conclusions 167 The database presented in this paper contains 2471 three-component digitally recorded strong-motion 168 records from the last 20 years in Costa Rica. They correspond to 155 earthquakes with maximum 169 hypocentral distances of 400 km. Data will continue to be added as new earthquakes get recorded by the 170 LIS-UCR network. In addition, a catalog with earthquake and station metadata is also available. Several 8 171 time-history IMs and peak responses were also calculated for each component. The IMs will be useful for 172 developing new seismic hazard studies for the region or for updating the current GMPEs established for 173 Costa Rica (Schmidt-Díaz, 2014). The database, the catalog with the metadata, and the estimated IMs are 174 available at the LIS-UCR website (see Data and Resources). 175 Data and Resources 176 The link to the LIS-UCR website is http://www.lis.ucr.ac.cr/. A table with the site conditions of each station 177 is available in the following link: http://www.crsmd.lis.ucr.ac.cr/?id=Estaciones. To request the database 178 of accelerograms please access the following link: http://www.crsmd.lis.ucr.ac.cr/?id=BD, and fill out the 179 form or send an e-mail to lis.inii@ucr.ac.cr. 180 The catalog is available at http://www.crsmd.lis.ucr.ac.cr/?id=BD. The IMs and statistical summary can be 181 found in the following link: http://crsmd.lis.ucr.ac.cr/crsmdb.zip. 182 Acknowledgments 183 This research was partially funded by the National Emergency and Risk Prevention Law N° 8933 from 184 Costa Rica, the UCREA funds from the University of Costa Rica through the project referenced as B9780 185 and the Spanish Government’s Ministry of Economy and Competitiveness (MINECO) through the project 186 referenced as CGL2015-65913-P. Luis A. Pinzón is supported by a Ph.D. scholarship grant from the 187 Government of Panama’s Institute for the Training and Development of Human Resources (IFARHU) and 188 the National Secretariat of Science, Technology, and Innovation (SENACYT). 189 190 191 9 192 References 193 Alvarado, G. E. et al., 2017, The new Central American seismic hazard zonation: Mutual consensus based 194 on up today seismotectonic framework, Tectonophysics, 721, no. October, 462–476, doi: 195 10.1016/j.tecto.2017.10.013. 196 Arias, A., 1970, A measure of earthquake intensity, Cambridge, MA, M.I.T. Press, 438–483. 197 Arroyo, M., K. Godínez, and L. Linkimer, 2017, Completitud del catálogo de la red sismológica nacional 198 de Costa Rica durante 1975-2014, Bol. Geol., 39, no. 3, 87–98, doi: 10.18273/revbol.v39n3- 199 2017006. 200 ASCE, 2017, Minimum Design Loads and Associated Criteria for Buildings and Other Structures 201 (ASCE/SEI 7-16) Reston, VA, doi: 10.1061/9780784414248. 202 Baker, J. W., and C. A. Cornell, 2006, Which Spectral Acceleration Are You Using?, Earthq. Spectra, 22, 203 no. 2, 293–312, doi: 10.1193/1.2191540. 204 Beyer, K., and J. J. Bommer, 2006, Relationships between median values and between aleatory 205 variabilities for different definitions of the horizontal component of motion, Bull. Seismol. Soc. 206 Am., 96, no. 4 A, 1512–1522, doi: 10.1785/0120050210. 207 Bindi, D., F. Pacor, L. Luzi, R. Puglia, M. Massa, G. Ameri, and R. Paolucci, 2011, Ground motion 208 prediction equations derived from the Italian strong motion database, Bull. Earthq. Eng., 9, no. 6, 209 1899–1920, doi: 10.1007/s10518-011-9313-z. 210 Bolt, B. A., 1973, Duration of strong ground motion, in 5th World Conference on Earthquake 211 Engineering, 1304–1313. 212 Bommer, J. J., A.-S. Elnashai, and A. G. Weir, 2000, Compatible acceleration and displacement spectra 213 for seismic design codes, in 12th World Conference on Earthquake Engineering, paper no. 207. 10 214 Boore, D. M., 2010, Orientation-Independent, Nongeometric-Mean Measures of Seismic Intensity from 215 Two Horizontal Components of Motion, Bull. Seismol. Soc. Am., 100, no. 4, 1830–1835, doi: 216 10.1785/0120090400. 217 Boore, D. M., W. B. Joyner, and T. E. Fumal, 1997, Equations for estimating horizontal response spectra 218 and peak acceleration from western North American earthquakes: A summary of recent work, 219 Seismol. Res. Lett., no. 68, 128–153. 220 Boore, D. M., and T. Kishida, 2016, Relations Between Some Horizontal-Component Ground-Motion 221 Intensity Measures Used In Practice 1, Bull. Seismol. Soc. Am., 107, no. 1, 334–343, doi: 222 10.1785/0120160250. 223 Boore, D. M., J. P. Stewart, E. Seyhan, and G. M. Atkinson, 2013, NGA-West2 Equations for Predicting 224 Response Spectral Accelerations for Shallow Crustal Earthquakes, Pacific Earthquake Engineering 225 Research Center, California. 226 Boore, D. M., J. Watson-Lamprey, and N. A. Abrahamson, 2006, Orientation-independent measures of 227 ground motion, Bull. Seismol. Soc. Am., 96, no. 4 A, 1502–1511, doi: 10.1785/0120050209. 228 Bozorgnia, Y. et al., 2014, NGA-West2 Research Project, Earthq. Spectra, 140227055104009, doi: 229 10.1193/072113EQS209M. 230 Bradley, B. A., and J. W. Baker, 2015, Ground motion directionality in the 2010–2011 Canterbury 231 earthquakes, Earthq. Eng. Struct. Dyn., 44, 371–384, doi: 10.1002/eqe.2474. 232 Campbell, K. W., and Y. Bozorgnia, 2008, NGA ground motion model for the geometric mean horizontal 233 component of PGA, PGV, PGD and 5% damped linear elastic response spectra for periods ranging 234 from 0.01 to 10 s, Earthq. Spectra, 24, no. 1, 139–171, doi: 10.1193/1.2857546. 235 CFIA, 2016, Código Sísmico de Costa Rica 2010 (Revisión 2014). Editorial Tecnológica de Costa Rica, 236 Cartago, Costa Rica. 11 237 Chiou, B., R. Darragh, N. Gregor, and W. Silva, 2008, NGA project strong-motion database, Earthq. 238 Spectra, 24, no. 1, 23–44, doi: 10.1193/1.2894831. 239 Chopra, A. K., 2007. Dynamics of Structures: Theory and Applications to Earthquake Engineering, 240 Prentice Hall, New Jersey. 241 DeShon, H. R., S. Y. Schwartz, S. L. Bilek, L. M. Dorman, V. Gonzalez, J. M. Protti, E. R. Flueh, and T. 242 H. Dixon, 2003, Seismogenic zone structure of the southern Middle America Trench, Costa Rica, J. 243 Geophys. Res. Solid Earth, 108, no. B10, doi: 10.1029/2002jb002294. 244 Dobry, R., I. M. Idriss, and E. Ng, 1978, Duration characteristics of horizontal components of strong- 245 motion earthquake records, Bull. Seismol. Soc. Am., 68, no. 5, 1487–1520. 246 Douglas, J., 2003, Earthquake ground motion estimation using strong-motion records: a review of 247 equations for the estimation of peak ground acceleration and response spectral ordinates, Earth- 248 Science Rev., 61, nos. 1–2, 43–104, doi: 10.1016/S0012-8252(02)00112-5. 249 Douglas, J., 2017, Ground motion prediction equations 1964-2016, Glasgow, UK. 250 Garini, E., and G. Gazetas, 2013, Damage potential of near-fault records: Sliding displacement against 251 conventional “Intensity Measures,” Bull. Earthq. Eng., 11, no. 2, 455–480, doi: 10.1007/s10518- 252 012-9397-0. 253 Goes, S. D. B., A. A. Velasco, S. Y. Schwartz, and T. Lay, 1993, The April 22, 1991, Valle de la Estrella, 254 Costa Rica (Mw=7.7) earthquake and its tectonic implications: a broadband seismic study, J. 255 Geophys. Res., 98, no. B5, 8127–8142, doi: 10.1029/93JB00019. 256 Goldstein, P., D. Dodge, M. Firpo, and L. Minner, 2003, SAC2000: Signal processing and analysis tools 257 for seismologists and engineers, in International Handbook of Earthquake and Engineering 258 Seismology W. H. K. Lee, H. Kanamori, P. C. Jennings, and C. Kisslinger (Editors), Academic 12 259 Press, London, 1613–1614. 260 Guangwei Fan, S. L. Beck, and T. C. Wallace, 1993, The seismic source parameters of the 1991 Costa 261 Rica aftershock sequence: evidence for a transcurrent plate boundary, J. Geophys. Res., 98, no. B9, 262 15759, doi: 10.1029/93JB01557. 263 Guendel, F., and M. Protti, 1998, Sismicidad y sismotectónica de América Central, Física la Tierra, 10, 264 nos. 0214–4557, 19–51. 265 Hayes, G. P., D. J. Wald, and R. L. Johnson, 2012, Slab1.0: A three-dimensional model of global 266 subduction zone geometries, J. Geophys. Res. Solid Earth 117, no. 1, 1–15, doi: 267 10.1029/2011JB008524. 268 Housner, G. W., 1975, Measures of severity of earthquake ground shaking, in Proc. U.S. Natl. Conf. 269 Earthquake Eng., 25–33. 270 Husid, L. R., 1969, Características de terremotos. Análisis general., Rev. del IDIEM, 8, no. 1, 21–42. 271 Kostinakis, K. G., A. M. Athanatopoulou, and K. Morfidis, 2015, Correlation between ground motion 272 intensity measures and seismic damage of 3D R/C buildings, Eng. Struct., 82, 151–167, doi: 273 10.1016/j.engstruct.2014.10.035. 274 Mezcua, J., R. M. García Blanco, and J. Rueda, 2008, On the strong ground motion attenuation in Spain, 275 Bull. Seismol. Soc. Am., 98, no. 3, 1343–1353, doi: 10.1785/0120070169. 276 Montero, W., 2001, Neotectónica De La Región Central De Costa Rica, Rev. Geol. Am. Cent., 24, 29–56. 277 Moya Fernández, A., 2018, Accelerographic Monitoring System of the Earthquake Engineering 278 Laboratory, Rev. Ing., 28, no. 1, 96–114, doi: 10.15517/ri.v28i1.30874. 279 Moya Fernández, A., 2006, Nuevo formato de datos para el Laboratorio de Ingeniería Sísmica del 280 Instituto de Investigaciones en Ingeniería de la Universidad de Costa Rica, Rev. Ing., 16, no. 2, 63– 13 281 74. 282 Muin, S., and K. M. Mosalam, 2017, Cumulative Absolute Velocity as a Local Damage Indicator of 283 Instrumented Structures, Earthq. Spectra, 33, no. 2, 641–664, doi: 10.1193/090416EQS142M. 284 Nigam, N.C. and Jennings, P. C., 1969, Calculation of Response Spectra from Strong-Motion Earthquake 285 Records, Bull. Seismol. Soc. Am. 59, no. 2, 909–922. 286 Norabuena, E. et al., 2004, Geodetic and seismic constraints on some seismogenic zone processes in 287 Costa Rica, J. Geophys. Res. Solid Earth, 109, no. 11, 1–25, doi: 10.1029/2003JB002931. 288 Park, Y. J., A. H. S. Ang, and Y. K. Wen, 1987, Damage-Limiting Aseismic Design of Buildings., 289 Earthq. Spectra, 3, no. 1, 1–26, doi: 10.1193/1.1585416. 290 PEER, 2015, NGA-East : Median Ground-Motion Models for the Central and Eastern North America 291 Region, Peer Rep. 2015/04, no. April 2015. 292 Pinzón, L. A., L. G. Pujades, S. A. Diaz, and R. E. Alva, 2018, Do Directionality Effects Influence 293 Expected Damage? A Case Study of the 2017 Central Mexico Earthquake, Bull. Seismol. Soc. Am., 294 108, no. 5A, 2543–2555, doi: 10.1785/0120180049. 295 Pinzón, L. A., L. G. Pujades, D. A. Hidalgo-Leiva, and S. A. Diaz, 2018, Directionality models from 296 ground motions of Italy, Ing. Sismica, 35, no. 3, 43–63. 297 Pinzón, L. A., Vargas-Alzate, Y. F., Pujades, L. G., and Diaz, S. A., 2020, A drift-correlated ground 298 motion intensity measure: Application to steel frame buildings. Soil Dynamics and Earthquake 299 Engineering, 132, 106096. doi: 10.1016/j.soildyn.2020.106096 300 Power, M., B. S. J. Chiou, N. A. Abrahamson, Y. Bozorgnia, T. Shantz, and C. Roblee, 2008, An 301 overview of the NGA project, Earthq. Spectra, 24, no. 1, 3–21, doi: 10.1193/1.2894833. 302 Quintero, R., and F. Güendel, 2000, Stress field in Costa Rica, Central America, J. Seismol., 4, no. 3, 14 303 297–319, doi: 10.1023/A:1009867405248. 304 Reed, J. W., and R. P. Kassawara, 1990, A criterion for determining exceedance of the Operating Basis 305 Earthquake, Nucl. Eng. Des., 123, nos. 2–3, 387–396. 306 Sarma, S. K., 1971, Energy flux of strong earthquakes, Tectonophysics, 11, no. 3, 159–172. 307 Sarma, S. K., and K. S. Yang, 1987, An evaluation of strong motion records and a new parameter A95, 308 Earthq. Eng. Struct. Dyn., 15, 119–132, doi: 10.1002/eqe.4290150109. 309 Schmidt-Díaz, V., 2014, Ground motion prediction models for Central America using data from 1972 to 310 2010, Rev. Geológica América Cent., 50, 7–37. 311 Schmidt-Díaz, V., 2011, Soil classification based on spectral ratios where Central American 312 accelerographic stations are located. Cases of El Salvador, Nicaragua and Costa Rica, Rev. 313 Geológica América Cent., no. 44, 9–26, doi: 10.15517/rgac.v0i44.3443. 314 Schmidt-Díaz, V., D. A. Hidalgo-Leiva, A. L. Acuña, A. Moya Fernández, E. Cordero, S. C., and E. 315 López, 2014, Aceleraciones del terremoto de Sámara del 05 de setiembre del 2012, Rev. En Torno a 316 la Prevención, 12, 38–47. 317 Stewart, J. P. et al., 2011, Representation of bidirectional ground motions for design spectra in building 318 codes, Earthq. Spectra, 27, no. 3, 927–937, doi: 10.1193/1.3608001. 319 Sucuoǧlu, H., and A. Nurtuǧ, 1995, Earthquake ground motion characteristics and seismic energy 320 dissipation, Earthq. Eng. Struct. Dyn., 24, no. 9, 1195–1213, doi: 10.1002/eqe.4290240903. 321 Thompson, E. M., & Worden, C. B. (2018). Estimating rupture distances without a rupture. Bulletin of 322 the Seismological Society of America, 108(1), 371–379, doi: 10.1785/0120170174. 323 Trifunac, M. D., and A. G. Brady, 1975, A study on the duration of strong earthquake ground motion., 324 Bull. Seismol. Soc. Am., 65, no. 3, 581–626. 15 325 Tso, W. K., T. J. Zhu, and A. C. Heidebrecht, 1992, Engineering implication of ground motion A/V ratio, 326 Soil Dyn. Earthq. Eng., 11, no. 3, 133–144. 327 Watson-Lamprey, J. A., and D. M. Boore, 2007, Beyond SaGMRotI: Conversion to SaArb, SaSN, and 328 SaMaxRot, Bull. Seismol. Soc. Am., 97, no. 5, 1511–1524, doi: 10.1785/0120070007. 329 Zhao, J. X., K. Irikura, J. Zhang, Y. Fukushima, P. G. Somerville, A. Asano, Y. Ohno, T. Oouchi, T. 330 Takahashi, and H. Ogawa, 2006, An empirical site-classification method for strong-motion stations 331 in Japan using H/V response spectral ratio, Bull. Seismol. Soc. Am., 96, no. 3, 914–925, doi: 332 10.1785/0120050124. 333 334 16 Authors mailing addresses Aarón Moya Fernández Email: cesar.moya@ucr.ac.cr Address: Laboratorio de Ingeniería Sísmica, Ciudad de la Investigación Universidad de Costa Rica, San Pedro, San José, Costa Rica Luis A. Pinzón Email: luis.pinzon@upc.edu Address: C. Jordi Girona 1-3 (Campus Nord-UPC), D2 303, 08034, Barcelona, Spain Victor Schmidt-Díaz Email: victor.schmidt@ucr.ac.cr Address: Laboratorio de Ingeniería Sísmica, Ciudad de la Investigación Universidad de Costa Rica, San Pedro, San José, Costa Rica Diego Hidalgo-Leiva Email: diego.hidalgo@ucr.ac.cr Address: Laboratorio de Ingeniería Sísmica, Ciudad de la Investigación Universidad de Costa Rica, San Pedro, San José, Costa Rica Luís G. Pujades Email: lluis.pujades@upc.edu Address: C. Jordi Girona 1-3 (Campus Nord-UPC), D2 303, 08034, Barcelona, Spain 17 Tables Table 1 Magnitude, depth and epicentral distance statistics for the entire database. Number and percentage of three-components records per interval. Depth (km) Magnitude (Mw) < 10 10—25 25—50 50—100 100—150 ≥ 150 3.0 - 4.0 38 (1.5%) 27 (1.1%) - - - - 4.0 - 5.0 76 (3.1%) 348 (14.1%) 365 (14.8%) 108 (4.4%) - - 5.0 - 6.0 57 (2.3%) 472 (19.1%) 229 (9.3%) 144 (5.8) - 28 (1.1%) 6.0 - 7.0 12 (0.5%) 232 (9.4%) 227 (9.2%) 7 (0.2%) 30 (1.2%) - ≥ 7.0 - 71 (2.9%) - - - - Epicentral distance (km) Magnitude (Mw) < 10 10—25 25—50 50—100 100—150 ≥ 150 3.0 - 4.0 17 (0.7%) 31 (1.3%) 5 (0.2%) 8 (0.3%) 4 (0.2%) - 4.0 - 5.0 30 (1.2%) 130 (5.3%) 325 (13.1%) 326 (13.2%) 53 (2.1%) 33 (1.3%) 5.0 - 6.0 18 (0.7%) 58 (2.3%) 170 (6.9%) 316 (12.8%) 182 (7.4%) 186 (7.5%) 6.0 - 7.0 3 (0.1%) 12 (0.5%) 26 (1.0%) 104 (4.2%) 75 (3.0%) 288 (11.7%) ≥ 7.0 - 1 (0.1%) 1 (0.1%) 8 (0.3%) 13 (0.5%) 48 (2.0%) 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 18 Table 2 List of intensity measures based on ground motion time histories Intensity measure Acronym Formulation Units PGAN00E max|𝑎𝑁00𝐸(𝑡)| Peak ground acceleration PGAN90E max|𝑎 2𝑁90𝐸(𝑡)| cm/s PGAZ max⁡|𝑎𝑍(𝑡)| Larger value of the two horizontal components of 𝑚𝑎𝑥|𝑎 (𝑡)| 2 acceleration (Douglas, 2003; Beyer and Bommer, 2006; PGA 𝑁00𝐸Larger(2) max [ ] ⁡ 𝑚𝑎𝑥|𝑎 cm/s 𝑁90𝐸(𝑡)| Pinzón, Pujades, Hidalgo-Leiva, et al., 2018) 𝑚𝑎𝑥|𝑎𝑁00𝐸(𝑡)| 2 Larger value of the three components of acceleration PGA Larger(3) max[𝑚𝑎𝑥|𝑎𝑁90𝐸(𝑡)|] ⁡ cm/s 𝑚𝑎𝑥|𝑎𝑧(𝑡)| Geometric mean of the PGA of the two horizontal components (Beyer and Bommer, 2006; Pinzón, Pujades, PGA GM √𝑃𝐺𝐴𝑁00𝐸 ∗ 𝑃𝐺𝐴𝑁90𝐸 2cm/s Hidalgo-Leiva, et al., 2018) Peak ground velocity PGV max|⁡𝑣(𝑡)| cm/s PGV-to-PGA ratio (Tso et al., 1992; Sucuoǧlu and Nurtuǧ, max⁡|𝑣(𝑡)| PGV/PGA s 1995; Bommer et al., 2000) max⁡|𝑎(𝑡)| 𝜋 𝑡𝑓 Arias intensity (Arias, 1970) I A ∫ 𝑎(𝑡)2 𝑑𝑡 cm/s 2 𝑔 𝑡𝑖 Root-mean-square (RMS) of acceleration (Housner, 1975; 1 𝑡 95%acc √ ∫ 𝑎(𝑡)2RMS 𝑑𝑡 g Dobry et al., 1978) ∆ 𝑡5% Root-mean-square (RMS) of velocity (Garini and Gazetas, 𝑡 1 95%vel 2 RMS √ ∫ 𝑣(𝑡) 𝑑𝑡 cm/s 2013; Kostinakis et al., 2015) ∆ 𝑡5% Specific energy density (Sarma, 1971; Sarma and Yang, 𝑡𝑓 2 SED ∫ 𝑣(𝑡)2 𝑑𝑡 cm /s 1987) 𝑡𝑖 Characteristic intensity (Park et al., 1987) I 𝑎𝑐𝑐 1.5C 𝑅𝑀𝑆 √𝑡𝑓 - 𝑡𝑓 Cumulative absolute velocity (Reed and Kassawara, 1990) CAV ∫ |𝑎(𝑡)| 𝑑𝑡 cm/s 𝑡𝑖 Significant duration (Husid, 1969; Bolt, 1973; Housner, ∆ 5-95% of Arias intensity s 1975; Trifunac and Brady, 1975), Duration-PGV intensity (Pinzón et al., 2020) I -PGV 𝑃𝐺𝑉𝛼 ∆𝛽 - 350 • a(t) and v(t) represents the acceleration and velocity time histories of an earthquake. 351 • ti is the beginning of the record, tf is the total duration of the record. 352 • 5% and 95% of the Arias intensity marks the beginning (t5%) and end (t95%) of the strong phase. 19 Table 3 List of intensity measures based on peak responses Intensity measure Definition SA and SA N00E N90E Response spectra of the as-recorded horizontal orthogonal components The larger of the two horizontal components (Douglas, 2003; Beyer and Bommer, 2006; SA Larger Bradley and Baker, 2015; Boore and Kishida, 2016; Pinzón, Pujades, Hidalgo-Leiva, et al., 2018) Geometric mean of the response spectra of the two as-recorded horizontal components SA GM (Beyer and Bommer, 2006; Bradley and Baker, 2015; Boore and Kishida, 2016; Pinzón, Pujades, Hidalgo-Leiva, et al., 2018) Percentile (pp) value of the geometric mean of the response spectra of the two as-recorded SA GMRotDpp horizontal components rotated onto all non-redundant azimuths (Boore et al., 2006; Boore and Kishida, 2016) Percentile (pp) value of the geometric mean of the response spectra of the two as-recorded SA GMRotIpp horizontal components rotated onto all non-redundant period- independent azimuths (Boore et al., 2006; Boore and Kishida, 2016) Percentile (pp) values of the response spectra of the two as-recorded horizontal SA RotDpp components rotated onto all non-redundant azimuths (Boore, 2010; Pinzón, Pujades, Diaz, et al., 2018) 20 List of figures Figure 1. Station distribution for the LIS-UCR strong-motion network. White lines correspond to administrative divisions by provinces and gray lines to major roads. Figure 2. Station distribution and soil classification. Figure 3. Number of ground motions recorded per year. Figure 4. (a) Magnitude as a function of the hypocentral distance for the 2471 records and (b) magnitude distribution. Figure 5. Epicenter location for the earthquakes recorded between 1998 and 2019. Figure 6. Peak ground acceleration as a function of the hypocentral distance for the three as-recorded components. Figure 7. Several intensity measures as a function of the hypocentral distance: (a) PGV, (b) PGV/PGA, (c) Arias intensity SAT1, (d) accRMS, (e) velRMS, (f) Specific Energy Density, (g) Characteristic intensity, (h) Cumulative Absolute Velocity and (i) Significand duration. Figure 8. Comparison of the 5% damped response spectra estimated with RotD100, GM, horizontal acceleration components (N00E and N90E) and the rotated components (º rot) from the 7.6 Mw Nicoya earthquake recorded at station GNSR, which occurred on 5 September 2012. Figure 9. Comparison of the 5% damped response spectra estimated with RotD100, Larger, RotD50, GMRotI50, GMRotD50 and the GM using the 7.6 Mw Nicoya earthquake recorded at station GNSR, which occurred on 5 September 2012. 353 21 Figure 1 Nicaragua 11˚ Station 10˚ 9˚ 10 km 75 km 8˚ −86˚ −85˚ −84˚ −83˚ P a n a m a Figure 2 Nicaragua Soil type 11˚ S1 (Rock) S2 (Stiff soil) S3 (Soft soil) S4 (Very soft soil) 10˚ 9˚ 10 km 75 km 8˚ −86˚ −85˚ −84˚ −83˚ P a n a m a Figure 3 F(iag)ure4 (b) Figure 5 12˚ 100 km Nicaragua 10˚ Panama 8˚ 0 100 Earthquakes from 1998−2019 M < 3.9 M 4−4.9 M 5−5.9 M 6−6.9 M > 7 −86˚ −84˚ −82˚ Figure 6 F(aig)ure 7 (b) (c) (d) (e) (f) (g) (h) (i) Figure 8 Figure 9