Submarine hydrogeology of the Hawaiian archipelagic apron 1. Heat flow patterns north of Oahu and Maro Reef

We present two profiles of collocated single-channel seismic reflection and heat flow determinations across the Hawaiian flexural moat: one north of O ahu and the other north of Maro Reef. Seismic reflection data are used as a n aid in determining depth to basement and interpreting moat stratigraphy. Moat sediments are locally up to 2 km thick and result mostly from slumps, debris avalanches and turbidites from volcanoes capping the Hawaiian Ridge. Each heat flow profile is similar to 200 km long and consists of paired th ermal gradient and conductivity measurements made at an interval of 1-2 km. The mean heat flow, uncorrected for the cooling effect of sedimentation, a long the Oahu profile is 63.9 mW m(-2) (s.d. 11.9 mW m(-2)) and along the M aro Reef profile is 59.8 mW m(-2) (s.d. 5.0 mW m(-2)). Our preferred sedime ntation correction model gives a corrected heat flow mean along the Oahu pr ofile of 74.2 mW m(-2) (s.d. 14.6 mW m(-2)) and for the Maro Reef profile o f 64.0 mW m(-2) (s.d. 4.9 mW m(-2)). These values are not significantly dif ferent despite the fact that oceanic crust at Maro Reef is 19 Myr older tha n at Oahu. The mean heat flow along the Oahu profile is within one standard deviation of the global mean value for oceanic crust of the same age, whil e the mean heat flow along the Maro Reef profile differs by greater than on e standard deviation for the global mean. However, the mean heat flow along the Maro Reef profile is within the uncertainty of previous heat flow dete rminations made offswell at Maro Reef [Von Herzen et al., 1989]. Variabilit y along each profile is significantly greater than the measurement uncertai nty (measurement uncertainty is 3 and 2 mW m(-2) for the Oahu and Maro Reef profiles, respectively) and is significantly greater along the Oahu profil e than the Maro Reef profile. While this variability could result from flui d flow within the archipelagic apron or underlying crust, simple moat wide circulation models do not appear to explain the dominant variability in hea t flow. Heat flow means and variability of both profiles are inconsistent w ith simple conductive models of lithospheric reheating.