Abstract

Heat-flow data along La Ristra seismic line crossing New Mexico from the Colorado Plateau to the Great Plains suggest that the Jemez Lineament and the Rio Grande rift are high heat-flow geologic provinces, which, from the heat-flow data, appear separated by a triangular wedge of Colorado Plateau. Estimated mid-depths of upper-mantle thermal sources for the Jemez Lineament and the Rio Grande rift generally agree with the mid-depths of the greatest upper-mantle seismic velocity anomalies, supporting the suggestion that these seismic anomalies are caused in part by increased temperature. Mid-depths of shallower thermal sources are near the mid-depth between the top of the crustal seismic velocity anomaly and the top of the upper-mantle seismic anomaly. Heat flow returns to intermediate values in the Colorado Plateau and the eastern rift flank of the Rio Grande rift, suggesting similar crustal and upper-mantle thermal conditions in the two regions and low values in the Great Plains, indicating cooler crustal and upper-mantle temperatures. Heat-flow values are generally consistent with seismic velocities, although differences between the data sets are observed for the Rio Grande rift eastern flank.

INTRODUCTION

In this study I present heat-flow analyses along La Ristra seismic line across New Mexico and discuss supplementary as well as complementary results to the seismic studies. Heat-flow data are used to estimate crustal and upper-mantle temperatures. In well-resolved data regions, upper-mantle temperatures predicted from heat-flow data and seismic studies are thought to be compatible (Goes and van der Lee, 2002). Heat-flow data may show short distance variability due to crustal radiogenic variations and/or groundwater flow (Decker et al., 1988; Reiter and Mansure, 1983, respectively). The complex character of heat-flow transition zones has been discussed by Blackwell (1978). By filtering heat-flow values along profiles that extend from highest to lowest values across a region, smoothed data can be obtained that better relate to crustal and upper-mantle thermal sources. Along five transitions crossing the Southern Rocky Mountains, Reiter (2008) has shown that gravity anomaly sources and seismic velocity anomalies generally have compatible depths with thermal sources, suggesting the influence of temperature on gravity anomaly sources and seismic velocity anomalies. Figure 1 shows heat-flow data along the La Ristra seismic line. There are 58 heat-flow values projected onto the line representing La Ristra. The heat-flow data supplement the horizontal resolution of the seismic studies and indicate the general compatibly in depth between seismic velocity anomalies and thermal source estimates.

DATA ANALYSIS AND UNCERTAINTIES

Similar to gravity data analyses for density step discontinuities, heat-flow data profiles can be analyzed for temperature discontinuities or steps that relate to thermal source depths by considering 1/4–3/4 widths (half-widths) in data components, although considerable ambiguity does exist (Reiter, 2008). Heat-flow anomaly source considerations involve uncertainties associated with the data traverse and the source location, depth, geometry, timing, and continuity. Sources can appear shallower or deeper depending on the data traverse with respect to the source location and geometry (Reiter, 2008). Source depth estimates can also vary depending only on the source geometry, e.g., lower temperatures at shallower depth having a sloping isotherm can produce the same heat-flow profile as deeper higher temperatures having a step isotherm (Clarkson and Reiter, 1987). The geometries of anomalous Rio Grande rift seismic regions are discussed later herein and suggest that a step temperature discontinuity is a reasonable approximation to the thermal sources in this study. Considering thermal diffusion times, an isotherm stepping to a 30 km depth produces an anomaly (as compared to equilibrium values) of ~80% magnitude and half-width at 10 m.y.; the same level of recognition requires ~20 m.y. for an equivalent step-up to 50 km (Clarkson and Reiter, 1987). Short-lived crustal and mantle heat sources would likely be unnoticed, lending support to the idea that source replenishment is typically involved in noticeable heat-flow anomalies (Reiter and Clarkson, 1983). Volcanism in the Rio Grande rift near La Ristra (from ~24 Ma to ~0.1 Ma; Kelley and Kudo, 1978) suggests timing consistent with recognizable, approximately steady-state, heat-flow anomaly sources in the crust and upper mantle.

By using the first term in a sine series to fit heat-flow measurements across the San Juan Basin, the 1/4–3/4 width method provides a half-width estimate that is the same as the mid-depth value of a best-fit steady-state temperature step configuration by Clarkson and Reiter (1987). If the data trend goes from highest to lowest values across a transition, then potential temperature step mid-depths in the crust and upper mantle can be estimated from the heatflow profile by filtering high-frequency components (using a fast Fourier transform [FFT] algorithm) and fitting a fourth-order sine series to the smoothed data. The 1/4–3/4 width of the first sine term is one-sixth of the peak-to-peak distance and approximates the mid-depth of the deepest recognizable temperature step. Successive terms by analogy have half-widths of one half, one third, and one quarter of the first term. Using an FFT algorithm (Table Curve 2D, version 5.01, SYSTAT Software), I removed higher frequencies so that the data were well represented and the fourth-order sine series best fit the smoothed values by obtaining the maximum F statistic. The sine series is given by
formula
(1)
The resulting sine series fits are compared qualitatively because the F statistics are typically so large as to indicate statistical significance at 99% confidence.

PREVIOUS STUDIES

Several seismic studies have analyzed data along La Ristra profile (Fig. 1). Analyses by Gao et al. (2004) and West et al. (2004) considered the entire profile to depths of ~600 km and ~350 km, respectively. Gao et al. (2004) showed both compressional and shear velocity perturbations. At 50–100 km depth, a compressional wave perturbation of ~−1.5% to −2% extends ~320 km, from ~105.4°W to just east of 108°W (Fig. 1). This perturbation comes much closer to the surface in the western portion of the Rio Grande rift near seismic station 30 (Fig. 1), and it becomes deeper (~200 km) in the western and eastern sections of the anomaly. The largest P wave perturbation (~–3%) is at ~50–60 km depth, and it extends across the Rio Grande rift from the eastern boundary with the Rio Grande rift eastern flank west to between the boundaries of the Rio Grande rift and Jemez Lineament (from about seismic stations 27 to 32; Fig. 1). Both to the east and west of negative compressional velocity perturbations, the perturbations quickly become positive between 50 and 150 km depth in the Colorado Plateau and Great Plains (+1.5% to +2.0%).

The geometries of the shear velocity perturbations are somewhat similar to the compression velocity perturbations, although they are not quite as voluminous (Gao et al., 2004). The shear velocity perturbations of ~–3% at 100 km have about the same lateral extent as the compressional wave perturbations (Gao et al., 2004; West et al., 2004). The shear velocity perturbations in the Colorado Plateau and Great Plains indicate cooler crust and upper mantle (Gao et al., 2004), where temperatures in the Great Plains may be somewhat cooler than in the Colorado Plateau (West et al., 2004). The largest shear velocity perturbation (~–5% or ~–7%; Gao et al., 2004; West et al., 2004, respectively) is at ~55 to ~75 km depth, and it extends either from east of the eastern boundary of the Rio Grande rift to within the Rio Grande rift eastern flank (Gao et al., 2004), or from the western boundary of the Rio Grande rift to within the Rio Grande rift eastern flank (West et al., 2004).

Wilson et al. (2005) focused on the Rio Grande rift and relatively nearby areas (~104.7°W to 108.3°W) to depths of ~150 km (Fig. 1). At 100 km depth, the anomalous shear wave velocity of 4.4–4.45 km s−1 extends from the center of the Rio Grande rift to ~170 km to the southeast and ~130 km to the northwest. Shear wave velocities at 100 km depth are ~4.45 to ~4.6 km s−1 in the Colorado Plateau and ~4.45–4.7 km s−1 in the Great Plains. Shear wave velocities of 4.2–4.35 km s−1 in the upper mantle have an almost bicuspid-like geometry centered on the Rio Grande rift, with shear velocities of 4.2–4.25 km s−1 in the crown (from ~75 to 50 km depth at the middle of the anomaly), and shear velocities of 4.25–4.35 km s−1 dipping down into the upper mantle as roots separated along the center of the Rio Grande rift. Shear wave velocities of 4.25–4.30 km s−1 extend from ~43 km depth to ~118 km depth in the root northwest of the center of the Rio Grande rift, and from ~43 km to below 150 km in the root southeast of the Rio Grande rift center.

The horizontal resolution of the surface wave velocity model is 55 km in the crust and 105 km in the upper mantle, whereas the depth of the Moho is ~±1.5 km, representing one standard deviation of crustal thickness (Wilson et al., 2005). Accounting for vertical exaggeration, the anomalous shear wave velocity of 4.25–4.30 km s−1 dips on average ~73°E to the east of the center Rio Grande rift (from ~43 to ~145 km depth), and ~62°W to the west of the center Rio Grande rift (from ~43 to 118 km depth; Wilson et al., 2005). The horizontal boundary of the region having anomalous shear wave velocities of 4.2–4.25 km s−1 is near vertical east of the center Rio Grande rift (from ~53 to 76 km depth), and it dips on average ~60°W to the west of center Rio Grande rift (from ~54 to 87 km depth, although the boundary is rounded). Anomalous shear wave velocities are also shown in the crust from ~14 to 27 km depth; the eastern boundary is essentially vertical, while the western boundary varies (depending on velocity) from vertical to east dipping at ~35° (Wilson et al., 2005). These anomalous seismic velocity region boundaries suggest that a temperature step model is a reasonable first-order approximation of the thermal sources at depth, assuming that temperature increase causes the seismic velocity anomalies.

Morgan et al. (1986) discussed the complex two-stage Cenozoic thermal and tectonic evolution of the Rio Grande rift and showed the heatflow transition from the Great Plains to the Rio Grande rift in southern New Mexico. Heat-flow studies in New Mexico have proceeded since the work by Herrin and Clark (1956) in southeast New Mexico. Data references given in the caption for Figure 1 indicate work over the past five decades. Although individual values can be suspect, consistency of neighboring data provides confidence in local or regional trends. For example, ~30 km southeast of the Socorro magma body, three data points in the Rio Grande rift eastern flank demonstrate a regional intermediate heat flow (~62 mW m−2) that is quite different from the values >90 mW m−2 in the Rio Grande rift (Fig. 1). These data suggest that the Rio Grande rift eastern flank has a different thermal regime than the Rio Grande rift.

DATA PRESENTATION

Figure 2 shows the heat-flow data profile along La Ristra. The data are presented as individual site values and three-point moving averages (Figs. 2A and 2B, respectively). In the northwest, values averaging ~65 mW m−2 represent the western San Juan Basin and Four Corners area of the Colorado Plateau. Going eastward, the values north of the Zuni Mountains show considerable variability, probably in part because of groundwater flow from the nearby Zuni Mountains (Figs. 1, 2A, and 2B). Near the center of the profile, two significant high heatflow anomalies appear, one associated with the Jemez Lineament and the other associated with the Rio Grande rift. These anomalies are separated by lower values near the western boundary of the Rio Grande rift (Fig. 1). To the east of the Rio Grande rift, the heat-flow values decrease across the Rio Grande rift eastern flank and decrease once again in the Great Plains (Fig. 1).

Heat-flow data along La Ristra appear to separate the Rio Grande rift and Jemez Lineament, as suggested by values ~80 mW m−2 near the western boundary of the Rio Grande rift along La Ristra (Figs. 1 and 2). Figure 1 shows a triangular wedge of the Colorado Plateau between the Rio Grande rift and the Jemez Lineament along La Ristra that may be consistent with intermediate heat-flow values. Data to the northwest of the Jemez Lineament return to Colorado Plateau heat-flow values over a shorter distance than data to the southeast of the Rio Grande rift return to Great Plains values. Because the heat flow is not quite symmetrical about the western boundary of the Rio Grande rift, data are considered separately (southeast and northwest of the Rio Grande rift western boundary) to estimate components of thermal anomalies. The southeast profile extends ~400 km from the western boundary of the Rio Grande rift (heat flow of 78 mW m−2) to the Great Plains. The northwest profile extends ~350 km from the 78 mW m−2 heatflow site to the Four Corners region in southeastern Utah. Because data trends from highest to lowest values, one may filter high frequencies from the profile data using a FFT algorithm and fit the resulting smoothed data with a sine series to provide approximations of depths for crustal and upper-mantle thermal sources causing the heat-flow anomalies (discussed earlier).

Site heat-flow data along the southeast part of La Ristra are shown in Figure 3A, along with the smoothed data and the sine series fit to smoothed data. Two new values are tabulated in the southern Albuquerque Basin (Fig. 1; Table 1). The deepest thermal source for data along La Ristra appears to have a midstep depth of between 98 km and 92 km, depending on the data set used (Table 2). A second important temperature midstep occurs at ~49–46 km depth. Although these depths may possibly be related to seismic velocity anomalies, the La Ristra seismic line runs oblique to the north-south axis of the Rio Grande rift, and it is necessary to consider a heat-flow profile that is orthogonal to the rift axis (east-west along latitude) in order to correctly represent the width of the Rio Grande rift heat-flow anomaly and the potential depths of thermal sources. Seismic profiles are projected onto a rift-normal profile (caption Fig. 2; Wilson et al., 2005). The east-west projection of the site distances along La Ristra is sin 45° times the distance along La Ristra. The three-point averaged data projected onto an east-west profile are shown in Figure 3B, along with the smoothed data and the sine series fit. The temperature mid-step discontinuity depths along the east-west profile are 66 km and 33 km (Table 2).

La Ristra is almost perpendicular to the Jemez Lineament, so heat-flow sites need not be projected onto an orthogonal profile (Fig. 1). A much better sine fit to the smoothed data is obtained by using three-point averaged data (Figs. 2B and 3C). Thermal source midstep depths in this case are estimated to be ~57 km, 29 km, and 19 km (Table 2).

DISCUSSION

La Ristra seismic line crosses New Mexico from the northwest to the southeast, intersecting the Jemez Lineament and the southern Albuquerque Basin of the Rio Grande rift. Considering three-point moving averaged data, the midstep thermal source depths estimated for the Rio Grande rift data (southeast profile), projected normal to the Rio Grande rift, and the Jemez Lineament data (northwest profile) are similar (66 km and 57 km, respectively; Table 2). This is generally consistent with near symmetry of the thermal regime across the area and the notion of pure shear suggested by Wilson et al. (2005). The distance across the extent of the shear wave velocity anomaly of 4.25–4.30 km s−1 at ~75 km depth under the Jemez Lineament and Rio Grande rift is ~120 km (Wilson et al., 2005), which is about the same as the distance across the Jemez Lineament and the Rio Grande rift along the 80 mW m−2 heat-flow value (Fig. 2A, ~110 km; Figs. 3B and 3C). The thermal source mid-depth estimates in the upper mantle compare to the mid-depths of the greatest shear wave velocity anomalies (4.20–4.25 km s−1; Wilson et al., 2005): ~66 km versus 64 km (49–78 km) under the Rio Grande rift and ~57 km versus 61 km (47–74 km) under the Jemez Lineament (Table 2, note depths are measured at ~10–12 km from the center of the Rio Grande rift and vary more with distance to the west). The seismic study Moho depths are ~±1.5 km; horizontal resolution is not as good (discussed previously).

A second thermal source has midstep depths of ~33 km and ~29 km (Rio Grande rift and Jemez Lineament, respectively; Table 2), which is near the Moho depth in the area. The Moho appears to gradually shallow, from ~44 km to ~36 km, coming from both the east and west, starting ~115 km from the center of the Rio Grande rift (Wilson et al., 2005). This Moho geometry may be quite different from that of the region of anomalous seismic velocity in the crust, which appears most notably from ~14 to 27 km depth (Wilson, et al., 2005). The crustal anomalous seismic region has a near-vertical step ~105 km east of center Rio Grande rift along La Ristra, and a near-vertical to a 36°E-dipping boundary (depending on the crustal velocity contour) west of center Rio Grande rift. The anomalous crustal seismic velocities suggest heat is being supplied to the crust, but the nature of the gradual Moho depth change compared to sharp boundaries of the crustal seismic anomaly may suggest that more than just advection by the Moho (likely associated with upper-mantle thermal sources) is delivering heat to the crust. The temperature steps estimated at ~33 and 29 km depth are midway between the top of the anomalous upper-mantle seismic velocity region and the top of the anomalous crustal seismic velocity region (~44 to ~14 km depth). The Moho would appear to be near the middle of a heat-transfer process between these depths. If the Moho temperature were approximately maintained as it moved closer to the surface, one might expect the crustal anomalous seismic velocity region east of center Rio Grande rift to have more gently sloped boundaries. West of center Rio Grande rift, the crustal anomalous seismic region may have sloping boundaries that mirror the sloping of the Moho and upper-mantle seismic anomalous region, depending on the chosen crustal velocity contour.

The shallower midstep source at 19 km under the Jemez Lineament is in the middle of the shear wave velocity anomaly in the crust, which dips from ~14 km depth down to ~28 km depth (Wilson et al., 2005). This step does not appear as significant in the southeast profile, where the crustal shear wave velocity anomaly is also strongly present (Table 2). The 19 km midstep along the northwest profile suggests crustal heat sources (heat injection or deep groundwater flow) and may be associated with the higher heat flows noticed northwest of the Zuni Mountains and the much narrower anomaly across the Jemez Lineament (Figs. 1 and 3A).

The heat-flow decrease near the western boundary of the Rio Grande rift indicates a likely separation of thermal sources between the Rio Grande rift and the Jemez Lineament. The horizontal resolution of seismic data (~55 km for crustal data and ~105 for mantle data) may have been unable to denote this separation, which is only ~40 km along La Ristra; though mid-crustal shear wave velocity contours becoming shallower near the western Rio Grande rift are consistent with intermediate heat flow (West et al., 2004). The heat-flow separation of the Rio Grande rift and Jemez Lineament would be compatible with a triangular wedge of the Colorado Plateau coming between the regions along La Ristra (Fig. 1), although values of ~80 mW m−2 suggest a warmer region than in the Four Corners area. This segment of the Colorado Plateau may be warmed by a shallower upper mantle than elsewhere in the Colorado Plateau, while the structural integrity of the crustal segment impedes mantle heat sources from entering the crust as they do along the Rio Grande rift and Jemez Lineament. If upper-mantle temperatures are brought closer to the surface, geothermal gradients and heat flow are increased. For example, moving in situ temperature from 45 km to 35 km will increase intermediate heat-flow values ~15 mW m−2 (Reiter and Barroll, 1990). For a Colorado Plateau heat flow normally ~65 mW m−2, this would increase the heat flow to the observed ~80 mW m−2 in the triangular Colorado Plateau wedge without involving crustal heat sources. However, much larger heat flow likely involves crustal heat sources. The compressional wave perturbation coming nearer to the surface at seismic station 30 (Gao et al., 2004) is consistent with the high heat-flow values of 107 and 110 mW m−2 and crustal heat sources in the southwestern Albuquerque Basin of the Rio Grande rift (Fig. 1).

The subsurface thermal regime of both the Rio Grande rift and Jemez Lineament likely vary along and normal to structural axes (Fig. 1). North of the seismic line, high heat flows are noted at the intersection of the Jemez Lineament and the Rio Grande rift and also near the middle of the Jemez Lineament (Fig. 1), but heat flow appears lower in the western part of the Jemez Lineament, generally consistent with a deepening Moho and reduction of the shear wave velocity anomaly (Gao et al., 2004; West et al., 2004; Wilson et al., 2005).

Shear wave velocities of ~4.40–4.45 km s−1 and velocity perturbations in the upper mantle appear to indicate a velocity anomaly (smaller in magnitude than the anomaly associated with the Rio Grande rift and Jemez Lineament) that extends southeast from the Rio Grande rift, a considerable distance into the Rio Grande rift eastern flank and possibly into the Great Plains (Fig. 1; Gao et al., 2004; West et al., 2004; Wilson et al., 2005). The heat-flow data suggest the thermal regime of the crust and upper mantle changes at the Rio Grande rift–eastern rift flank boundary. The Rio Grande rift widens south of Socorro, and La Ristra crosses the Rio Grande rift eastern flank rather close to the Rio Grande rift (Fig. 1). Perhaps the nearby Rio Grande rift is influencing the seismic data recorded along La Ristra in the eastern rift flank. The heat-flow values in the eastern rift flank are about the same as the values in the Colorado Plateau near the Four Corners area, suggesting a similar thermal regime in the crust and upper mantle.

I thank Tom Kaus and Leo Gabaldon for helping prepare Figure 1. R.M. Russo and several reviewers made many suggestions to improve the manuscript.