Overlying the Socorro Magma Body is a ~5000 km2 region of accentuated seismicity in the central Rio Grande rift of New Mexico [Sanford et al., 1995] which has been designated the Socorro Seismic Anomaly (SSA) Figure 2. The SSA covers only about 2 percent of the total area of the state but accounts for ~45% percent of the state's seismicity above magnitude 2.5. The lateral extent of the SSA is defined by the existence of an aseismic halo surrounding the area of intense seismic activity.
In this paper we present a new map of the geographic extent of the Socorro magma body. Our map is constructed using all three commonly identified reflected phases. We used over 5400 reflections, 24 times the number of reflection points used by Rinehart et al., [1979]. We have found that the magma body extends farther south, southeast, and northwest than had been previously mapped, and its northern limit has been well constrained for the first time. The western and eastern boundaries are less well known, primarily due to the recording geometry dictated by the station configuration of the Socorro seismic network. We have also found that the upper surface of the magma body displays no measurable dip, or regional variations.
Mapping Procedure and Data
The data we used were the reflected phases identified on seismograms Figure 3 recorded with the Socorro seismic network Figure 4 between 1975 and 1995. Utilizing the {\bital Hartse et al.} [1992] velocity model Figure 5 we traced rays between earthquake hypocenters and the stations which recorded reflections to find reflection-point positions. Plotting the reflection-point positions in map view provides an estimate of the magma body's lateral extent.
Figure 6 shows the epicenters of the 1181 events in our data set. The majority of events detected within the SSA occur in swarms, which appear as clusters of epicenters on Figure 6. Between five and 16 stations recorded each earthquake. Hypocenters were determined using the computer program SEISMOS [Hartse, 1991] which incorporates arrival times of reflections off the midcrustal magma body as well as direct P and S. All of the events in the data set have latitude, longitude and depth errors less than or equal to 1.0 km at one standard deviation; the average errors were 0.31 km, 0.37 km and 0.46 km, respectively. Magnitudes (Md) range between -0.8 and 2.6.
Our final reflected phase data set consisted of 697 PzP, 2169 SzP, and 2589 SzS arrivals identified on the seismograms of 1181 earthquakes. An average event had between 4 and 5 observed reflections. The maximum number of reflections from a single event is 15.
Lateral Extent of the Socorro Magma Body
We estimate the lateral extent of the SMB by estimating the reflection-point positions of the observed SzP, and SzS arrivals Figure 7. PzP reflections are used only to fill in the region defined by the SzP and SzS reflections and to obtain additional constraints on hypocenter estimates.
The segments of the COCORP lines marked with bold lines on Figure 7 denote locations along the profiles where the magma body is clearly imaged. Other portions of the lines do not clearly show reflections, even where we observe many microearthquake reflections, such as near Lines 3 and 4. Most of the processing and reprocessing of the COCORP data emphasized upper crustal structure, sometimes degrading deeper images that may include magma body reflections.
The outline surrounding the reflection-point positions shown in Figure 7 is our new map of the Socorro magma body. This map indicates that the minimum lateral extent of the magma body is >3400 km squared. To better determine which borders of the magma are well constrained we plotted hypothetical reflection-point positions Figure 8 . We assumed that every station which recorded a direct arrival from each of the 1181 events could also have recorded SzP and SzS arrivals. If similiar distributions of hypothetical reflection-point positions are found along both sides of the boundary, then we assume those portions are constrained by our data. Thus, our microearthquake reflection data constrain the south, southeast, north, and northwest sides of the magma body.
The northern boundary of the magma body is primarily limited by an absence of observed reflections at the Albuquerque Seismic Lab stations ALQ and ANMO, ~110 km northeast of Socorro. Two swarms are responsible for the clusters of possible reflection points north of the magma body outline (Figure 8): The San Acacia swarm events [Balch et al., 1994, 1995] show no reflected phases of any kind at the Albuquerque stations; and the Bernardo swarm events [Sanford et al., 1993] have no midcrustal reflections, but a matched pair of P and S Moho reflections are observed. The existence of these shear wave reflections is further evidence that midcrustal magma is not present northeast of our boundary.
Figure 9 compares the magma body outline estimated by Rinehart et al. [1979] to the outline we have found. The new outline clearly indicates the minimum lateral extent of the magma body is much greater than had been previously estimated. Some portions of the magma body were not imaged by Rinehart et al. because they had a more limited data set. Stations deployed in the 1970's were more tightly clusterd than the current network, and they used 220 SzS reflections while we used all three phases and 5455 reflections. Further, the data of Rinehart et al. was gathered in a three year period, While our data covers 20 years and has a more complete sampling of the overall pattern of seismicity.
Only the northernmost part of the magma body map presented by Rinehart et al. [1979] extends beyond the new boundaries. In this region, Rinehart et al. did not observe any reflections and this part of their outline was originally interpolated and displayed with a dashed line.
A strong spatial correlation exists between our map of the midcrustal magma body and the extent of the Socorro Seismic Anomaly Figure 10, suggesting that the seismicity is causally related to the magma body. The ~80 km north-south extent of the magma body falls just inside the ~100 km north-south range of the SSA, which covers about 5000 km squared and has the same elliptical shape as the outline of the 3400 km squared magma body. The close spatial relation between the SSA and the midcrustal magma body along their northern and southern boundaries suggests that the observed seismicity can be used to place limits on the eastern and western boundaries. Using the distribution of earthquakes within the SSA over the last 34 years, the maximum east-west extent of the SMB is ~60 km, our observed value from the map of reflection points is ~50 km.
Evidence for Flatness
Because Hartse et al. [1992] assumed a flat reflector when inverting for a velocity model, arrival time residuals which do not fit assumed timing errors can indicate dip or significant relief. Jackson [1972] quantifies how well residuals match timing error by:
If R is much greater than 1.0, then assumed picking errors are too small, or the assumed model is too simple. If R is much less than 1.0, then the assumed picking errors are too large, or the assumed model is too complex. For the 5455 reflected phases we found R=1.06, indicating that the reflected phase residuals are very similiar to the assumed timing errors. If dip or any relief were significant, then residuals should exceed assumed timing error, and hence, R for reflections should be large. This is especially true when considering that the reflection-point positions Figure 7 are from such a wide geographic distribution and are associated with both large and small source-receiver distances.
An argument could be made that R ~1.0 for the 1181 event data set because event focal depths are being adjusted within the inversion to compensate for reflector dip or relief. However, each event has up to 15 phases which sample the reflecting surface at different geographic positions. Assuming dip or any significant relief, it is unlikely that focal depth adjustments could compensate for several different reflector depths at once and still fit a velocity model which fits the observations so well. Furthermore, Hartse et al. [1992] demonstrated that when the SzP and SzS arrival times from a single station can be included in the earthquake location problem, the two arrivals define a unique reflector depth and focal depth. Because we are holding the velocity model (and reflector depth) fixed, any dip or unevenness on the reflector should force a poor fit between the SzP - SzS observations and the SzP - SzS arrival times predicted by the fixed velocity model. Clearly, R ~1.0 for the reflected phases indicates that this is not happening, even though we 1328 SzP - SzS pairs in our data set.
The other way we tested for dip was to examine the distribution of the residuals associated with the reflection-point positions. Figure 11 shows the positions of reflection points that have negative reflection time residuals and Figure 12 shows positive reflection time residuals. If the magma body displayed a regional dip, then the positive and negative residuals should display roughly separated geographic distributions. For instance, if dip were down to the north, then positive residuals should be more prevalent to the north and negative residuals more prevalent to the south (positive residuals imply later observed arrival times). No such patterns are apparent, which implies random timing errors, rather than a complicated model, are responsible for the residual patterns.
Typical timing errors for the reflections are between 0.25 and 0.35 s. For the short-offset (10 to 15 km) reflected phases, reflector depths, which actually differ by +- 0.50 km from the assumed 18.75 km flat model, will produce SzS traveltimes which will differ by 0.25 to 0.30 seconds from the times of the assumed model. Thus, if the true depth of a reflection point varies by more than +- 0.50 km from the actual model, then segregated patterns of positive and negative residuals should be apparent. Assuming a maximum depth variation of about 1 km (18.75 +- 0.50 km) over an 80 Km distance, this allows for a maximum north-south dip on the upper surface of the magma body of less than 1 degree. A width of 50 km limits east-west dip to a maximum of about 1 degree.
Conclusions
We have re-estimated the lateral extent of the Socorro magma body. Our mapping was primarily based on calculating the reflection-point positions of 5455 PzP, SzP, and SzS reflected phases identified on the seismograms of 1181 microearthquakes. Additional constraints on the northeast limit of the magma body were obtained from COCORP profiles recorded in the Socorro area in 1975 and 1976. Our map indicates that the magma body extends farther to the south, southeast, and northwest than had been previously mapped, and we have found that the minimum size of the magma body is around 3400 km squared. A maximum possible extent of the midcrustal magma body may be the size of the Socorro Seismic Anomaly, which occupies a somewhat larger area of 5000 km squared and includes our magma body outline.
Our study indicates the upper surface of the magma body is quite flat. The primary reasons for this conclusion are the exceptionally good fit between observed and theoretical reflection times and the even distribution of reflection-point positions associated with positive and negative reflected-phase residuals across the entire mapped surface of the magma body. The timing errors associated with our data allow for maximum relief on the magma's upper surface of about +- 0.50 km, which translates to a maximum north-south dip of < 1 degree and a maximum east-west dip of about 1 degree.
References
Ake, J. P., and A. R. Sanford (1988). New evidence for the existence and internal structure of a thin layer of magma at mid-crustal depths near Socorro, New Mexico. Bull. Seismol. Soc. Am., 78, 1335-1359.
Balch, R. S., (1992). New constraints on the Socorro Magma Body based on improved hypocenter estimates, Geophysics Open-file Report 63, New Mexico Tech, Socorro, 120 pp.
Balch, R. S., A. R. Sanford, L. H. Jaksha, H. E. Hartse, and L. S. House (1994). Non-Double-Couple Focal Mechanisms Observed in the Central Rio Grande Rift of New Mexico, abstract, EOS, 75 no. 44, p. 446.
Balch, R. S., A. R. Sanford, L. H. Jaksha, H. E. Hartse and L. S. House (1995). Exotic Seismic Sources Observed in the Central Rio Grande Rift, abstract, Seis. Res. Letters, 66 no. 2., p 25.
Brocher, T. M. (1981). Geometry and physical properties of the Socorro, New Mexico, magma bodies, J. Geophys. Res., 86, 9420-9432.
Brown, L. D., C. E. Chapin, A. R. Sanford, S. Kaufman, and J. Oliver (1980). Deep structure of the Rio Grande rift from seismic reflection profiling, J. Geophys. Res., 85, 4773-4800.
Brown, L. D., P. A. Krumhansl, C. E. Chapin, A.R. Sanford, F. A. Cook, S. Kaufman, J. Oliver, and F. S. Schilt (1979). COCORP seismic reflection studies of the Rio Grande rift, in Rio Grande Rift: Tectonics and Magmatism, edited by R. E. Riecker, AGU, Washington, D.C., 169-184.
Cape, C. D., S. McGeary, and G. A. Thompson (1983). Cenozoic normal faulting and the shallow structure of the Rio Grande rift near Socorro, New Mexico, Geol. Soc. Am. Bull., 94}, 3-14.
de Voogd, B., L. Serpa, and L. Brown (1988). Crustal extension and magmatic processes: COCORP profiles from Death Valley and the Rio Grande rift, Geol. Soc. Am. Bull., 100, 1550-1567.
Hartse, H. E. (1991). Simultaneous hypocenter and velocity model estimation using direct and reflected phases from microearthquakes recorded within the central Rio Grande rift, New Mexico, Ph.D. Dissertation, New Mexico Institute of Mining and Technology, Socorro, New Mexico, 251 pp.
Hartse, H. E., A. R. Sanford, and J. S. Knapp (1992). Incorporating Socorro Magma Body reflections into the earthquake location process, Bull. Seismol. Soc. Am., 82, 2511-2532.
Jackson, D. D. (1972). Interpretation of inaccurate, insufficient, and inconsistent data, Geophys. Jour. Roy. Astron. Soc., 28, 97-109.
Olsen, K. H., G. R. Keller, and J.N. Stewart (1980). Crustal structure along the Rio Grande rift from seismic refraction profiles, in Rio Grande Rift: Tectonics and Magmatism, edited by R. E. Riecker, AGU, Washington, D.C., 127-143.
Olsen, K. H., D. J. Cash, and J. N. Stewart (1982). >Mapping the northern and eastern extent of the Socorro Midcrustal Magma body by wide-angle seismic reflections, New Mexico Geological Society guidebook 33, 179-185.
Rinehart, E. J., and A. R. Sanford (1981). Upper crustal structure of the Rio Grande rift near Socorro, New Mexico, from inversion of microearthquake S-wave reflections, Bull. Seismol. Soc. Am., 71, 437-450.
Rinehart, E. J., A. R. Sanford, and R. M. Ward (1979). Geographic extent and shape of an extensive magma body at mid-crustal depths in the Rio Grande rift near Socorro, New Mexico, in Rio Grande Rift: Tectonics and Magmatism, edited by R. E. Riecker, AGU, Washington, D.C., 237-251.
Sanford, A. R., and C. R. Holmes (1961). Note on the July 1960 earthquakes in central New Mexico, Bull. Seismol. Soc. Am., \bold 51, 311-314.
Sanford, A. R., and L. T. Long (1965). Microearthquake crustal reflections, Socorro, New Mexico, Bull. Seismol. Soc. Am., %5, 579-586.
Sanford, A. R., O. Alptekin, and T. R. Topozada (1973). Use of reflection phases on microearthquake seismograms to map an unusual discontinuity beneath the Rio Grande rift, Bull. Seismol. Soc. Am., 63, 2021-2034.
Sanford, A. R., R. S. Balch, J. A. Lakings, H. E. Hartse and L. S. House (1993). A Link Between Listric Faulting and Recent Seismicity in the Central Rio Grande Rift of New Mexico, abstract EOS, 74 no. 43, p. 417.
A. R. Sanford, A. R., R. S. Balch, K. Lin (1995). A seismic anomaly in the Rio Grande rift near Socorro, New mexico, abstract, Seis. Res. Letters, 66 no. 2, p. 44.
Sanford, A. R., R. P. Mott, P. J. Shuleski, E. J. Rinehart, F. J. Caravella, R. M. Ward, and T. C. Wallace (1977). Geophysical evidence for a magma body in the vicinity of Socorro, New Mexico, in The Earth's Crust: Its Nature and Physical Properties, Geophys. Monogr. Ser., 20, edited by J. G. Heacock, AGU, Washington, D.C., 385-404.