We report 35 new and 9 revised terrestrial heat flow measurements overlying the Andean subduction zone in Bolivia and Peru. The measurement sites are distributed in the Andean Cordillera, the Altiplano, the sub‐Andean ranges, and the sedimentary platform and basins to the east of the Andes. They fall in the distance range 75–900 km from the Peru‐Chile trench. Sites in the Peruvian Cordillera have a mean heat flow of 41 mW m−2, whereas those in the Bolivian Cordillera and Altiplano average 84 mW m−2. The sub‐Andean ranges and the adjacent sedimentary platform have a mean heat flow of 50 mW m−2. The higher heat flow of the Bolivian Cordillera and Altiplano lies to the east of Quaternary volcanoes along the Bolivia‐Chile border and in southernmost Peru and thus can be recognized as a “back arc” heat flow high. Neither Quaternary volcanism nor high heat flow are present in central and northern Peru. This contrasting along‐strike pattern correlates with the variable angle of subduction of the Nazca plate beneath the region. Beneath Bolivia the subduction is at about 30°–35°, whereas beneath Peru it is subhorizontal and may be providing a cold underplate to the overlying Peruvian lithosphere. Extensive Miocene volcanism in Peru suggests, however, that heat flow in Peru 10 Ma ago was likely similar to that in Bolivia in the present day and implies a change in subduction and a rapid diminution of heat flow in Peru over the past 10 Ma. Imbrication of cold oceanic lithosphere beneath Peru, analogous to that beneath western British Columbia, may provide a mechanism for rapid reduction of the heat flow.

The natural remanent magnetization of the upper Keweenawan Nonesuch Shale and Freda Sandstone has been analyzed with thermal, alternating field, and chemical demagnetization techniques. The results of this study are in good agreement with previously published works by DuBois and by Vincenz and Yaskawa, but place a tighter constraint on the North American apparent polar wander path. Fifty-eight samples, representing nearly 900 m of section, have been collected from the flanks of the Porcupine Mountain uplift. From principally thermal demagnetization analyses, a mean direction of primary magnetization has been calculated for the Nonesuch Shale, with declination 279.8°, inclination +9.8°, yielding a virtual geomagnetic pole position at 176.5° E, 10.3° N, and for the Freda Sandstone, with declination 271.3° inclination + 0.7°, yielding a virtual geomagnetic pole at 179.5° E, 1.2° N. A group of intermediate (secondary) components of magnetization is removed between temperatures of 350 °C and 550 °C, yielding well clustered directions. Its mean direction with declination 280.6°, inclination −9.5°, resulted in a virtual geomagnetic pole at 169.2° E, 3.7° N. This secondary magnetization is assumed to be of chemical origin and is most likely associated with the late Precambrian copper mineralization of the Nonesuch Shale. By thorough sampling of the stratigraphic column it is possible to infer the general direction of motion of a plate as the sediments were deposited. The motion of the North American plate as observed in the upper Keweenawan magnetizations is in agreement with the previously published polar wander paths for the late Precambrian.

Seventy-live samples from the Freda sandstone (1050 Ma old) have been drilled from the same slab in order to investigate the best method of sample preparation and to optimize procedures for chemical demagnetization of red sandstones, Expanding on the chemical demagnetization procedures used by Park, several variables have been tested. These include: multicoring or slotting of samples to permit access of the acid, temperature and normality of the acid, zero-magnetic field during leaching, and vacuum conditions before leaching steps. Through proper sample preparation and leaching procedures, it is possible to separate the solubility spectra and associate the components of magnetization with particular magnetic carriers.Visual inspection of cores sectioned during progressive chemical demagnetization has allowed the comparison of the solubility spectra to the blocking temperature spectra of thermal demagnetizations. This comparison has linked the coarser-grained specularite to the highest blocking temperatures, and the fine-grained hematilic cement to intermediate blocking temperatures. There exists a component of magnetization which appears to be insoluble lo the leaching process. It has been shown to be carried by detrital titanomagnetite. Through the combined use of chemical, thermal, and alternating field demagnetization a magnetic history involving deposition. lithilication. and later overprinting has been unraveled from this slab of Freda.

We present a method of estimating subsurface temperatures and true regional heat flow in the presence of perturbing topography, variable surface temperature, and subsurface thermal conductivity contrasts. The method involves solution of the steady‐state three‐dimensional heat conduction equation by finite‐difference numerical techniques. The topography is represented by an irregular upper boundary and the variable surface temperature as a boundary condition along the irregular upper surface. Internal structural configurations and conductivity contrasts are easily accommodated. The principal variable input into the system is the deep basal (unperturbed) heat flow. The best value of heat flow is obtained by minimizing, in a least‐squares sense, the differences between observed and calculated temperatures. Temperature observations commonly are distributed irregularly in the near‐surface (perturbed) environment, in multiple vertical or inclined boreholes, tunnels, and/or mine galleries. The method is particularly suited to simultaneous analysis of an ensemble of distributed observations, in contrast to methods that focus on the perturbation to the temperature gradient in the vicinity of a single borehole. We used the method to reduce data obtained at fifteen newly established heat flow sites in the Bolivian and Peruvian Andes. We illustrate with three examples—a two‐dimensional model from the Bolivar Mine, Bolivia; (2) a three‐dimensional model using variable conductivity from the Cerro Verde Mine, Peru; and (3) a three‐dimensional model at the Colquiri Mine, Bolivia where temperature measurements were few and the distance between the individual boreholes was fairly large.

No abstract