Mechanisms and timing of exhumation of collision-related metamorphic rocks, southern Brooks Range, Alaska: Insights from 40 Ar/ 39 Ar thermochronology

Transcription

1 TECTONICS, VOL. 21, NO. 3, /2000TC001270, 2002 Mechanisms and timing of exhumation of collision-related metamorphic rocks, southern Brooks Range, Alaska: Insights from 40 Ar/ 39 Ar thermochronology James J. Vogl, 1 Andrew T. Calvert, 2 and Phillip B. Gans Department of Geological Sciences, University of California, Santa Barbara, California, USA Received 6 November 2000; revised 9 October 2001; accepted 13 December 2001; published 17 May [1] New 40 Ar/ 39 Ar thermochronologic data document the spatial distribution of cooling ages and rates across a metamorphic culmination in the south central Brooks Range, providing important constraints on the timing and processes responsible for exhumation of deep-seated metamorphic rocks. The data indicate widespread episodic exhumation-related cooling, with short-lived events in the mid-albian to early Cenomanian and Paleocene to Eocene. Hornblende ages indicate that cooling from the metamorphic peak began at Ma across the epidoteamphibolite facies core of the culmination. Mica ages, however, increase southward from 90 Ma to 100 Ma, indicating that average cooling rates increased southward. Thermal ing suggests that Albian cooling is the result of a short-lived exhumation event that ended by the early Cenomanian, and that exhumation rates were between 1 and 8 mm/yr, with higher rates toward the south. The southward increase in cooling/exhumation rates is attributed to a southward increase in the component of tectonic exhumation by normal faulting on the south flank of the range. Rapid erosion, which is recorded by thick molasse deposits in flanking basins that are coeval with rapid cooling, may have been aided by surface uplift in the footwall of these faults. The northward younging of mica ages may not be explained by a simple core complex involving metamorphic rocks being progressively dragged out from beneath an extensional fault. The overall spatial distribution of metamorphic zones and cooling ages/rates is best explained by a two-step process: (1) a brief episode of extensional faulting producing a regional northward tilt with superimposed doming during the Albian- Cenomanian and (2) relative uplift of the core during renewed Tertiary contraction, which enhanced the domal geometry. Diffusion-domain ing of K-feldspar data indicates a period of little or no cooling that spanned much of the Late Cretaceous followed by rapid increase in cooling rates in the Paleocene to Eocene. Increased Tertiary cooling rates are attributed to renewed contraction leading to surface uplift and erosion. INDEX TERMS: 1035 Geochemistry: Geochronology; 8102 Tectonophysics: Continental contractional orogenic belts; 8109 Tectonophysics: Continental tectonics extensional (0905); KEYWORDS: Brooks Range, exhumation, 40 Ar/ 39 Ar thermochronology, collisional tectonics, Alaska, diffusion-domain ing 1 Now at Department of Geology, University of Florida, Gainesville, Florida, USA. 2 Now at U.S. Geological Survey, Menlo Park, California, USA. Copyright 2002 by the American Geophysical Union /02/2000TC001270$ Introduction [2] During collisional orogenesis, supracrustal rocks are commonly buried to depths of several tens of kilometers, resulting in metamorphism and ductile deformation. It is generally accepted that erosion and normal faulting are the dominant processes responsible for exhumation of these rocks, and that these processes may operate during and/or after the contractional phase of orogenesis [Ring et al., 1999]. The contributions and interactions of erosion, extension, and contraction during the exhumation process, however, have not been widely documented. Methods for directly delineating exhumation histories through the reconstruction of depth (pressure)-time paths are currently not available. In contrast, cooling histories may be thoroughly documented through application of argon and fission track thermochronology. Although the relationship between exhumation and cooling is complicated by the advection of heat, much can be learned about exhumation processes through documentation of the spatial distribution of cooling histories and by investigation of the relationship between specific structures and the variation in cooling histories. [3] In this paper we have used 40 Ar/ 39 Ar thermochronology to investigate the processes and timing of exhumation of metamorphic rocks formed in the Brooks Range collisional orogen of northern Alaska. Previously published K-Ar and 40 Ar/ 39 Ar ages from the metamorphosed southern Brooks Range, which generally range from 130 Ma to 80 Ma [Brosge and Reiser, 1964; Turner et al., 1979; Christiansen and Snee, 1994; Patrick et al., 1994; Hannula and McWilliams, 1995; Blythe et al., 1998; Gottschalk and Snee, 1998; Toro, 1998], reveal little about the rates, episodicity, and spatial patterns of cooling, and therefore have shed little light on the timing and processes (i.e., erosion, late-orogenic extension, postorogenic extension) that led to cooling and exhumation of the metamorphic core. Furthermore, there are no quantitative cooling histories from K-feldspar to constrain the history between mica and apatite fission track ages. [4] To address these issues, we conducted a systematic 40 Ar/ 39 Ar thermochronology traverse across a metamorphic culmination in the southern Brooks Range and integrated the thermochronology with microstructural studies and quantitative thermobarometry. The culmination comprises an epidote-amphibolite facies core flanked by greenschist and lower-grade rocks and therefore provides exposure of a range of crustal levels. We document the spatial variation in cooling ages as a function of structural depth, metamorphic grade and N-S (across strike) position. We also present the first quantitative cooling histories using K-feldspar diffusion-domain ing for the Brooks Range, which delineate the cooling history between 300 C and 150 C, 2-1

2 2-2 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY Figure 1. (a) Terrane and metamorphic map of northern Alaska. CH, Cosmos Hills; DA Doonerak antiform. Modified from Moore et al. [1994]. Tertiary folds from Mull [1985]. (b) Simplified cross section through western Brooks Range. Modified from Moore et al. [1997]. and bridge the gap between 40 Ar/ 39 Ar mica ages and apatite fission track ages. Our new data allow us to delineate the cooling rates, the spatial variation in cooling rates, and the episodicity of cooling. This information is integrated with structural, metamorphic, and sedimentological data to provide insight into how and when the metamorphic rocks were unroofed. 2. Tectonic Setting of the Brooks Range [5] The Brooks Range orogen is the result of collision between the Koyukuk island arc and a Devonian-to-Jurassic south facing passive margin [e.g., Moore et al., 1994]. The northern half of the Brooks Range comprises a number of stacked and internally imbricated allochthons that define a north-vergent fold-thrust belt (Figure 1). The structurally highest (Angayucham) terrane consists of a package of dominantly ultramafic and gabbroic rocks that structurally overlie imbricated basaltic and sedimentary rocks, which are believed to represent fragments of an ocean basin that once separated the Alaskan continental margin from the Koyukuk arc to the south [e.g., Patton and Box, 1989; Moore et al., 1994]. The Angayucham terrane is exposed as klippen in the western Brooks Range as well as in fault slivers along the south flank of the range (Figure 1). Below the obducted Angayucham terrane, the allochthons consist dominantly of sedimentary packages that represent various facies of the continental margin. These allochthons (De Long Mountains subterrane, Endicott Mountains allochthon, and North Slope subterrane of Figure 1) are stacked such that the more distal facies are exposed in the structurally higher allochthons [Mayfield et al., 1988; Moore et al., 1994]. Stratigraphic reconstructions suggest a minimum of 250 to 600 km of N-S shortening across the Brooks Range [Mull et al., 1987; Oldow et al., 1987; Mayfield et al., 1988; Moore et al., 1994]. [6] Deeper levels of the orogen are exposed in the southern half of the range (Figure 1), where rocks are more metamorphosed and penetratively strained. The metamorphic rocks contain dominantly greenschist-grade mineral assemblages, which locally overprint blueschist-facies minerals in a belt along the southern part of the range (Figure 1) [Nelson and Grybeck, 1981; Dusel-Bacon et al., 1989]. This greenschist-overprinted blueschist belt is separated from the Angayucham terrane by south side down extensional faults and sheared greenschist-grade rocks of the Slate Creek terrane (Figure 1) [e.g., Gottschalk and Oldow, 1988; Moore et al., 1994]. The foreland basin to the north, and the Koyukuk basin to the south of the Brooks Range are filled with severalkilometer-thick sections of clastic rocks that record the erosional denudation of the orogen [e.g., Mull, 1985; Mayfield et al., 1988; Molenaar et al., 1988; Nilsen, 1989; Patton and Box, 1989]. [7] On the basis of isochron analysis (age spectra are highly disturbed) of biotite and hornblende 40 Ar/ 39 Ar data from the metamorphic sole beneath the Angayucham terrane ophiolites, Wirth et al. [1993] suggested that thrusting may have initiated in the Middle Jurassic. Most of the juxtaposition and internal short-

3 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY 2-3 ening of the allochthons occurred between the Jurassic and mid- Cretaceous during the evolution of the fold-thrust belt. Latest Cretaceous to Paleocene contraction is well documented by tight thrust-cored anticlines and open synclines in the foreland basin north of the Brooks Range front (Figure 1) [e.g., Mull, 1985]. It is less clear whether or not contraction continued between these two periods. [8] A Middle to Late Jurassic age for the blueschist-facies metamorphism is often inferred on the basis of Rb-Sr ages [Armstrong et al., 1986], K-Ar dates [Turner et al., 1979], and a single 40 Ar/ 39 Ar plateau age [Christiansen and Snee, 1994], but the data are not compelling. Other studies have suggested that a minimum age for blueschist-facies metamorphism is given by oldest step ages in hump-shaped 40 Ar/ 39 Ar white mica spectra [Gottschalk and Snee, 1998]; however, Hannula and McWilliams [1995] presented similarly shaped 40 Ar/ 39 Ar spectra from the Seward Peninsula and noted that unreasonably old step ages are likely due to excess argon. Till and Snee [1995] suggested that a 120 Ma white mica 40 Ar/ 39 Ar plateau age effectively dates a younger blueschist metamorphic event that affected basement rocks in an antiformal stack in the western Brooks Range. 3. Structure and Metamorphism of the Study Area [9] Most of the area along the study transect is composed of quartzose and calcareous schist/phyllite, marbles, and lesser amounts of quartzite and metabasite (Figure 2). The area also exposes a Devonian granitic orthogneiss (Arrigetch Peaks orthogneiss) that has locally intruded adjacent units and has been metamorphosed along with the host rocks (Figure 2). Throughout most of the section, lithologic layering dips gently northward (Figure 3). Toward the southern end of the cross section, layering is steep, giving way to moderate southward dips at the southern edge of the Brooks Range (Figure 3) [Dinklage, 1998]. [10] The metamorphic-grade distribution along the transect defines a metamorphic culmination, with an epidote-amphibolite facies core centered near the southern margin of the Devonian Arrigetch Peak granitic orthogneiss (Figure 3a). The flanks of the culmination consist of greenschist and lower grade rocks that overprint high-p/low-t assemblages on the south flank (Figure 3a) [Patrick et al., 1994; Dinklage, 1998]. For our discussion we divide the transect into five metamorphic zones (Figures 3a and 3b): northern greenschist (NGS) zone, northern epidote-amphibolite (NEA) zone, central epidote-amphibolite (CEA) zone, southern epidote-amphibolite (SEA) zone, and blueschist-greenschist (BGS) zone. The zones are distinguished by differences in mineral assemblages, peak pressure-temperature (P-T) estimates, and metamorphic histories (J. J. Vogl, Thermal/baric structure and P-T evolution of the Brooks Range metamorphic core, Alaska, submitted to Journal of Metamorphic Geology, 2000, hereinafter referred to as Vogl, submitted manuscript, 2000). Diagnostic mineral assemblages for metabasites and metapelites are shown in Figure 3a, and a peak metamorphic temperature profile across the culmination is provided in Figure 3b. Metamorphic field gradients are generally smooth across the five metamorphic zones, except for the juxtaposition of greenschist-grade (NGS zone) and lowermost epidote-amphibolite grade rocks (NEA zone) across the NW trending high-angle Takahula Fault. On the basis of fission track ages on nearby similarly oriented faults [Blythe et al., 1997], the Takahula Fault is believed to be Tertiary in age. [11] Vogl [2002] has described Cretaceous deformation in the study area in terms of three deformation events. D 1 involved north directed deformation that produced a pervasive foliation that is generally parallel to the lithologic layering along the transect. This event, which includes formation of most of the north directed foldthrust belt, probably spanned most of the Cretaceous and may have ended in the Aptian Vogl [2002]. The post-d 1 structural history of the Brooks Range has been more highly debated. Recent structural mapping by Vogl [2002] has shed light on this history and is summarized as follows. Porphyroblasts overgrow the D 1 foliation in the deepest levels of the NGS zone (kyanite, chloritoid, or albite) and in the NEA zone (garnet, biotite, or chloritoid), indicating post- D 1 heating. Greenschist-grade rocks on the north flank of a metamorphic culmination (NEA zone) display mesoscopic southvergent F 2 folds with shallowly to moderately dipping S 2 axialplanar crenulation cleavages, as well as small-displacement north side down normal faults (Figure 3a). These structures indicate a sequence of D 2 south directed backfolding and D 3 extension following the protracted phase of north directed D 1. At deeper levels, epidote-amphibolite facies rocks (NEA zone) display a pervasive gently north dipping S 2 crenulation cleavage that is axial-planar to kilometer and smaller scale tight to isoclinal recumbent F 2 folds. Post-D 1 porphyroblast-matrix relationships at the deeper levels are similar to those associated with D 2 backfolds at higher levels, suggesting that folds at the deeper levels also initiated during synmetamorphic D 2 backfolding. Much of the strain recorded by S 2 crenulation cleavages, the tightening of folds, foliation boudinage, and north side down shear bands, however, is interpreted to have formed during postmetamorphic D 3 extension, which produced normal faults at higher levels. [12] Dinklage [1998] has proposed a similar structural sequence for the south flank of the culmination. He proposed that the north dipping layering south of the Arrigetch Peaks orthogneiss was overturned during regional-scale backfolding and rotated to shallow dips during subsequent extension. He described south side down extensional shear bands, as well as shallowly south dipping S 2 crenulation cleavages, which are locally axial-planar to northvergent folds. Vogl [2002] suggested that the vergence reversal between the north and south flanks of the culmination delineate the regional-scale south closing backfold (Figure 3a). [13] The S 2 crenulation cleavages dip oppositely on opposite sides of the culmination and therefore may define a structural dome centered in the southern part of the Arrigetch Peaks orthogneiss. The highest grade rocks exposed along the transect appear to coincide with the region of S 2 -dip reversal, which is consistent with a structural dome interpretation. Following the observations on the north flank of the culmination, it was proposed by Vogl [2002] that the S 2 fabrics in epidote-amphibolite facies rocks initiated during D 2 backfolding, but have been domed during D 3 extension and Tertiary deformation. The fabrics were probably intensified and accommodated shearing away from the core during the D 3 extension [Dinklage, 1998; Vogl, 2002]. [14] Regardless of the kinematic interpretation of the S 2 fabrics, it is clear that N-S extension affected the area. East and west of our transect, south side down extensional faults and shear zones omit several kilometers of structural section along the south flank of the

4 2-4 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY Figure 2. Geologic map with 40 Ar/ 39 Ar locations, based on Nelson and Grybeck [1980], Vogl [2000] (north of Arrigetch Peaks orthogneiss), and Dinklage [1998] (south of Arrigetch Peaks orthogneiss).

5 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY 2-5 Figure 3. (a) Cross section showing structure and metamorphic zones. See Figure 2 for location. SCT, Slate Creek terrane; AT, Angayucham terrane; Kc, Cretaceous clastic sedimentary rocks of the Koyukuk basin. Mineral assemblages in metabasites (top) and metapelites (bottom) that define the metamorphic zones are shown above the section. Mineral abbreviations: Ab, albite; Bt, biotite; Cld, chloritoid; Grt, garnet; Hbl, hornblende; Olig, oligoclase. Also shown is the approximate limit of early blueschist facies metamorphism. (b) Peak temperature profile, with pressures where available. Data from Vogl [2000]. (c) 40 Ar/ 39 Ar ages from this study. Also shown are data from Patrick et al. [1994], P; A. Till and L. Snee, (unpublished data), TS. Dashed line shows observed trend of mica cooling ages. Vertical shaded box in BGS zone shows range of published 40 Ar/ 39 Ar and K-Ar ages for rocks contiguous with the BGS zone rocks that underwent similar metamorphic history. range [Gottschalk and Oldow, 1988; Christiansen and Snee, 1994; Little et al., 1994; Carlson, 1985; Law et al., 1994]. Similar faults exist in the study area and are shown in our sections; however, the faults have not been mapped in detail in this area. These faults are likely synchronous with the smaller displacement north side down normal faults (D 3 ) on the north flank of the culmination. As noted above, D 3 extensional strain at deeper levels also involved oppositely dipping extensional shear bands on opposite sides of the culmination, and may have also involved the tightening of folds and associated subhorizontal S 2 crenulation cleavage development. 4. The 40 Ar/ 39 Ar Analytical Methods [15] The 40 Ar/ 39 Ar ages were obtained by conventional stepheating experiments [e.g., McDougall and Harrison, 1988] using a Staudacher-type resistance furnace in the argon geochronology laboratory at University of California, Santa Barbara. Samples were crushed, and appropriate size fractions were concentrated using heavy liquids, paper shaking, and magnetic separation techniques. Samples were cleaned in an ultrasonic bath and handpicked under reflected light. Separates were packaged in copper foil and loaded into a quartz vial with packaged flux monitors. Vials were irradiated in the Dummy Fuel Element of the TRIGA reactor at Oregon State University during six irradiations over 5 years. Neutron flux was monitored with Fish Canyon sanidine (assigned age of 27.8 Ma) or Charcoal Ovens Tuff (assigned age of Ma). Ca interference was monitored with fluorite. [16] Samples and monitors were heated in the resistance furnace, and gas was purified continuously during extraction by two SAES ST-172 porous getters. Argon isotopes were analyzed on a MAP 216 mass spectrometer fitted with a Baur-Signer source and a Johnston

6 2-6 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY Table 1. Summary of Argon Thermochronology Data a Sample Mineral Spectrum Plateau Age (Ma) %/Steps Isochron Age (Ma) Points MSWD 40 Ar/ 36 Ar Total Fusion Age (Ma) Northern Greenschist Zone AVL95-23 WM staircase none ± ± AVL95-26 WM saddle ± / ± ± AVL94-14 WM plateau ± / ± ± AVL94-87A WM stepping plateau none none 104 AVL97-2 WM stepping plateau none none AVL94-98 Bt plateau ± / ± ± Northern Epidote-Amphibolite Zone (Northern Part) AVL95-96 Hbl plateau ± /13 none AVL94-100b Hbl saddle none none AVL94-100b Bt plateau with loss ± /14 none AVL95-43 WM plateau 87.6 ± / ± ± AVL94-62 WM plateau 89.9 ± / ± b 297 ± AVL94-60 WM plateau, with loss 92.0 ± /4 none 285 ± APK Bt plateau 91.0 ± / ± ± AVL94-78 Bt plateau with loss 89.5 ± / ± ± AVL95-43 Bt plateau with loss 84.8 ± /75/ ± ± AVL95-52 WM plateau, saddle 92.8 ± / ± b 296 ± AVL95-52 Bt plateau 94.6 ± / ± ± AVL95-51 Bt plateau ± / ± ± Northern Epidote-Amphibolite Zone (Southern Part) AVL95-53 WM plateau 93.1 ± / ± ± AVL95-56 WM plateau 92.3 ± / ± ± AVL95-56 Bt plateau 92.6 ± / ± ± Central Epidote-Amphibolite Zone APK90-38 Hbl saddle none ± ± ATi-230 A Bt plateau ± / ± b 272 ± a Abbreviations are as follows: MSWD, mean square of weighted deviate; WM, white mica; Bt, biotite; Hbl, hornblende. A complete listing of argon data is given in Table A1, which is available as electronic supporting material. b Bad isochron fit. MM1 multiplier with a sensitivity of moles/volt. Our Staudacher-type resistance furnace uses a coiled tungsten filament, a tantalum crucible with no liner and tungsten-rhenium (C type) thermocouple. Analyses are blank-corrected and our typical system blanks on m/e 40 vary from moles at low temperatures (<1000 C) and climb to moles at 1300 C for 15 min heating steps. All parts of the step-heating experiments are automated with pneumatically actuated, all-metal valves, a Macintosh computer, and custom software. [17] K-feldspar s were analyzed over 50 to 65 step experiments, with individual step isolation times ranging from 15 min to 11 hours. Initial steps commonly yielded old apparent ages, but replicate steps at low temperatures allow estimates of lowtemperature ages. The diffusion and age data from the step-heating experiments were ed (diffusion-domain ing) to reconstruct the moderate- to low-temperature cooling history. Modeling is based on the theory of Lovera and Richter [1989] and Lovera et al. [1991], that complicated K-feldspar age spectra and nonlinear Arrhenius plots are caused by the presence of multiple diffusion domains of varying size, resulting in a range of closure temperatures, typically between 150 and 300 C. A log r/r o versus % 39 Ar released plot [Richter et al., 1991] can be produced from the diffusion data, where r is the size of a given diffusion domain, and r o is the size of a reference domain determined by fitting a straight line through the linear array of low-temperature steps on the Arrhenius plot. A number of domains with a given size and volume fraction can be estimated and log r/r o plots can be calculated and compared to plots produced from the diffusion and age data. Once the domain-size distribution is determined, the age spectra can be synthesized by iterative calculation of different cooling histories and matched to laboratory-obtained spectra. 5. The 40 Ar/ 39 Ar Data [18] The age data across the metamorphic culmination are presented in the following sections. We also integrate our new data with previously collected data from other authors. A summary of all of our age data is given in Table 1, and a complete listing of argon data is given in Table A1, which is available as electronic supporting material. 1 The spatial distribution of all ages along the cross section is shown in Figure 3. Herein, we use the term plateau in a general descriptive sense rather than assigning a more quantitative, but arbitrary, definition. Two-sigma errors are reported on plateau and inverse isotope correlation diagram (isochron) ages. 1 Supporting Table A1 is available via Web browser or via Anonymous FTP from ftp://kosmos.agu.org, directory append (Username = anonymous, Password = guest ); subdirectories in the ftp site are arranged by paper number. Information on searching and submitting electronic supplements is found at esupp_about.html.

7 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY 2-7 that hornblende from this white mica-rich is complexly zoned in several elements, perhaps leading to the inconsistencies in the step ages. [23] A second hornblende (APK90-121) from the CEA zone analyzed by A. Till and L. Snee (unpublished data) yielded a spectrum with three contiguous steps (each 25 30% of the total gas) with apparent ages of ± 0.2, ± 0.2, and ± 0.3 Ma (A. Till, written communication). The yielded an isochron age of ± 0.6 Ma (A. Till, written communication). Given the slight saddle shape and the few number of steps, we suggest Ma is a maximum age for cooling through hornblende T C SEA zone. [24] We analyzed one hornblende (APK90-38) from a metabasite within the southern part of the SEA zone. The yielded a subdued asymmetric saddle-shaped spectrum with step ages ranging from ± 0.6 Ma to ± 0.3 Ma (Figure 4c). A six-point isochron yields an age of ± 1.5 Ma and a 40 Ar/ 36 Ar ratio of K/Ca ratios for individual steps range from to 0.096, similar to the range ( ) measured on the microprobe. Microprobe analyses of the hornblendes reveal zoning and overgrowths, with Si cations per formula unit ranging from 7.2 in the cores to 6.3 in the rims. K 2 O/CaO ratios generally increase toward the rim. This is located near the Hbl-in isograd and 1 km north (upgrade) of rocks yielding Grt-Bt temperatures of 480 Figure 4. (a i) The 40 Ar/ 39 Ar age spectra from all metamorphic zones. Uncertainties shown on individual steps represent 1s errors without error in J parameter. For s with plateau ages, steps used in age determination are shaded. Bt, biotite; Hbl, hornblende, Wm, white mica Hornblende [20] We have analyzed two new hornblende s from the NEA zone and one from the SEA zone; two other hornblende s from the CEA zone were analyzed by Patrick et al. [1994] and A. Till and L. Snee (unpublished data). One hornblende (AVL95-96; NEA zone) comes from a millimeter-scale mafic layer 1 2 cm inside the Arrigetch Peaks orthogneiss, and yields a plateau with an age of ± 0.1 Ma (Figure 4a). Microprobe analyses indicate 2.5 wt % K 2 O, consistent with K/Ca ratios of measured for all steps defining the plateau, indicating a very pure. [21] The other NEA zone hornblende (AVL94-100b), from a strongly foliated and lineated metavolcanic, produced a v, or saddle-shaped, spectrum with a minimum age of 111 Ma (Figure 4b). Consistent K/Ca ratios of.04 were measured for nearly all steps and agree well with microprobe analyses, indicating a pure separate. The shape of the hornblende spectrum is probably the result of excess argon. Excess argon is also indicated by a biotite separate from the same that yielded an anomalously old weighted-mean plateau age of ± 0.2 Ma (Figure 4b). [22] Patrick et al. [1994] presented one hornblende analysis (APK90-119) from the CEA zone that yields an isochron age of 110 Ma. The spectrum is crankshaft -shaped with apparent step ages between 105 and 110 Ma. Our microprobe analyses indicate Figure 4. (continued)

8 2-8 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY Figure 4. (continued) 510 C [Vogl, 2000]. Given these temperatures (and the uncertainties in estimated T C for hornblende), the older, higher Si cores may have grown slightly below T C and may not have been sufficiently degassed during final heating that produced rim growth. Therefore the range in individual step ages could be reflecting hornblende growth ages or partially degassed cores. Alternatively, the asymmetric saddle shape is the result of excess argon, which may be reflected in the high 40 Ar/ 36 Ar ratio for the isochron. Whichever interpretation is correct, the youngest step-age of ± 0.3 Ma probably represents a maximum age for cooling of this below 500 C. On the basis of the youngest step-age and isochron age, cooling of the SEA zone through hornblende T C probably occurred at Ma. [25] Overall, the hornblende data indicate that the entire epidote-amphibolite facies core of the metamorphic culmination cooled through hornblende T C at Ma White Mica and Biotite [26] Analyses were performed on several s from the NEA zone between the northern margin of the Arrigetch Peaks orthogneiss and the Takahula Fault. Two white micas (AVL94-60 and AVL94-62) and two biotites (AVL94-78 and APK94-105), from four different s produced weighted mean plateau ages ranging from 89.5 to 92.0 Ma (Figures 4d and 4e). One other (AVL95-43) gives slightly younger ages of 87.6 ± 0.8 Ma (white mica) and 84.8 ± 0.4 Ma (biotite; Figures 4d and 4e). All of the biotites and one of the white micas show slightly younger ages at low gas fractions, suggestive of argon loss. [27] We have analyzed two biotites and a white mica from a single outcrop at the western contact of the orthogneiss body a few kilometers to the south (Figure 2). Biotite from the orthogneiss (AVL95-51) yields a plateau with a weighted mean age of 90.0 ± 0.3 Ma (Figure 4f). White mica and biotite from a micaceous quartzite (AVL95-52) several meters away from the contact yield slightly older plateau ages of 92.8 ± 0.2 Ma and 94.6 ± 0.2 Ma, respectively (Figure 4f ). Approximately 3 5 km farther south, along the northern contact area of the narrow southwest margin of the Arrigetch Peaks orthogneiss (Figure 2), three s (AVL95-53, 56) yield plateaus with weighted-mean ages between 92.3 and 93.1 Ma (Figure 4g). [28] Patrick et al. [1994] presented a single white mica plateau age of 96 Ma from the CEA zone near the south margin of the Arrigetch Peaks orthogneiss (Figure 4g). Several other unpublished biotite and white mica ages (A. Till, written communication) yield similar ages. [29] From the SEA zone, Patrick et al. [1994] presented a hump-shaped white mica spectrum with step ages ranging from 122 Ma to 102 Ma ( APK90-95; Figure 4h). Our microprobe analyses indicate that the white micas are phengite and muscovite, with Si content ranging from 6.2 to 6.8 cations p.f.u. X-ray maps and quantitative analyses show that the dominant foliation is defined by high-si phengitic micas. Lower-Si white micas occur as rims and as grains that cut the main foliation. The hump-shaped spectrum from this suggests that either this area was locally affected by excess argon or that early formed D 1 white micas formed during high-p/low-t metamorphism were incompletely degassed despite reaching post-d 1 temperatures of 500 C and remaining above 500 until Ma. [30] We analyzed one biotite from the SEA zone ( 90ATI-230a). The biotite spectrum yields a plateau age of 100 Ma, similar to the flanks of the hump-shaped white mica spectrum (Figure 4h). We interpret cooling of the SEA through mica T C to have occurred around 100 Ma. [31] No consistent relationship between biotite and white mica ages from the same, or from closely spaced s, was observed. Overall, mica ages from the epidote-amphibolite facies rocks of the core of the culmination display a well-defined northward younging trend from 100 Ma in the SEA zone to < 92 Ma in the NEA zone. [32] Samples from the NGS zone north of the Takahula Fault yield scattered ages. A biotite (AVL94-98) yields a plateau age of ± 0.2 Ma and an isochron age of ± 0.4 Ma (Figure 4i). A white mica (AVL94-14) yields a plateau age of ± 1.3 Ma (Figure 4i). Two other white micas (AVL94-87A and AVL97-2) from the same area yield stepping plateaus with individual step ages ranging from 97 Ma to 103 Ma (Figure 4i). Two other s yield less interpretable spectra (see Tables 1 and A1). The two s yielding the younger ages contain only a single generation of micas (D 1 ), suggesting that the young ages represent cooling through mica closure rather than new mica growth. The youngest step ages of 97 Ma are probably a minimum age but

9 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY APK Arrigetch Peaks orthogneiss APK Apparent age (Ma) o AVL95-51 Arrigetch Peaks orthogneiss AVL95-51 Apparent age (Ma) log r/r 40 log r/ro AVL95-56 Arrigetch Peaks orthogneiss AVL95-56 Apparent age (Ma) Cumulative 39 Ar released log r/ro Cumulative 39 Ar released Figure 5. Measured K-feldspar 40 Ar/ 39 Ar age spectra and log r/r o plots derived from step-heating experiments. closely approximate the final cooling age. Thus the difference in cooling ages (97 Ma and Ma) across the Takahula Fault is consistent with the NE side down displacement of 2 3 kilometers suggested by the difference in metamorphic grade. Given the metamorphic grade in the NGS zone, no hornblende data are available to estimate cooling rates. [33] We have no analyses from south of the SEA zone, and no meaningful analyses have been reported. Structurally, petrologi-

10 2-10 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY APK APK Apparent age (Ma) Apparent age (Ma) Apparent age (Ma) APK PCBR Cumulative 39 Ar released Figure 5. (continued) log r/ro log r/ro log r/ro APK PCBR Cumulative 39 Ar released

11 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY 2-11 cally and lithologically similar rocks to the east and west [Hitzman et al., 1986; Christiansen and Snee, 1994; Gottschalk, 1998; Gottschalk et al., 1998] yield K-Ar and 40 Ar/ 39 Ar mica ages that range from 130 Ma to 100 Ma [Brosge and Reiser, 1964; Turner et al., 1979; Christiansen and Snee, 1994; Blythe et al., 1998; Gottschalk and Snee, 1998] K-Feldspar [34] To obtain information about the moderate- to low-temperature ( C) part of the cooling history, we performed detailed step-heating experiments on five K-feldspar s from the Arrigetch Peak orthogneiss (APO) and one from a small orthogneiss body in the Cosmos Hills (Figure 1). Both orthogneiss bodies belong to a Devonian suite of granite/granodiorite plutons exposed in the southern Brooks Range. The orthogneiss was variably strained during Cretaceous deformation, leading to locally developed recrystallized porphyroclastic textures. Synkinematic lower amphibolite-facies metamorphism produced final assemblages that include K-feldspar + phengite + muscovite ± albite ± oligoclase ± garnet ± biotite ± epidote ± titanite ± fluorite. [35] All five K-feldspar s from the APO produce similarly shaped age spectra with subplateaus at temperatures < C, spanning 40 60% of 39 Ar released (Figure 5). The subplateaus range in age from 70 Ma to 50 Ma, and show a general northward decrease in age (Figure 5). At higher temperatures, apparent ages increase smoothly to maximum ages between 110 Ma and 130 Ma (Figure 5). Locally, older spikes appear on the sloping higher temperature part of the spectra, but they are relatively minor except for s AVL95-51 and APK (Figure 5). Such spikes are likely the result of excess argon and were removed before ing. The Cosmos Hills (PCBR-65) produced an age spectrum similar to the APO s but with a slightly younger plateau and much younger maximum ages at the high-temperature steps (Figure 5). The amount of 39 Ar extracted above the melting temperature ranged between 2% and 28%. [36] Thermal histories derived from the diffusion-domain ing are shown in Figure 6. Fifty cooling curves were produced for each, specifying monotonic cooling. Model age spectra calculated from the thermal history and diffusion domain size distribution produced good fits with the measured age spectra (Figure 5). The plateau and high-temperature limit of the plateau are reproduced particularly well, while a slight discordance between the and measured age spectra appears in the higher temperature steps of some s (Figure 5). The discordance is most noticeable in s AVL95-51 and APK90-108, which show the most evidence for excess argon, in the form of spikes and high-temperature ages that are older than the Ma hornblende ages (Figure 5) Temperature ( C) T-t curves from K-feldspar ing 1. APK AVL AVL APK APK Age (Ma) Figure 6. Temperature-time curves for the various metamorphic zones. Diverging curves toward older ages represent reasonable end-member thermal histories for unconstrained early part of the history. K-feldspar curves with thin solid/ dashed lines represent continuation of cooling history from higher temperature parts of CEA and NEA zones, respectively. Other K-feldspar ing curves are from s located between these two zones.

12 2-12 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY [37] Sample APK was also ed allowing reheating. All thermal histories derived from this ing display the same Latest Cretaceous to early Eocene rapid cooling event (Figure 6; similar to the monatonic s) but have widely varying earlier thermal histories. All the possible early thermal histories also produce good fits to the measured age spectra, but most yield geologically unrealistic cooling and heating cycles. Although minor reheating is possible, there is no field or petrographic evidence for large amounts of reheating. Thus we infer that the cooling histories are well represented by the monotonic cooling s. [38] The major conclusions drawn from the ing are as follows. 1. The Arrigetch Peaks region underwent relatively rapid cooling (up to 15 C/m.y.) during the early Tertiary, resulting in a minimum of C of cooling. Cooling rates appear to have slowed following the early Tertiary cooling event, since most of the APO has apatite fission track ages between 21 Ma and 27 Ma, with mean track lengths between 12.3 and 13.6 m [Blythe et al., 1997]. 2. Early Tertiary rapid cooling was preceded by a period of little or no cooling that spanned most of the Late Cretaceous. Although most of the s were constrained as monatonic, the temperatures at the onset of early Tertiary cooling ( C) allow little cooling following passage through mica closure. 3. Temperatures prior to early Tertiary cooling were distinctly higher in the northern s than in the southern s, although temperature estimates for AVL95-51 may be somewhat high (perhaps because of excess argon) given the Ma mica cooling ages. Thus, during the Late Cretaceous the northern part of the Arrigetch Peaks orthogneiss resided at depths of 1 2 km below the southern s. This result is consistent with the northward younging of mica cooling ages described in section 5.2, which suggests that the southern margin of the APO was at a higher structural level than the northern margin by 96 Ma. [39] Modeling results for the Cosmos Hills has produced unsatisfactory results, and further ing is underway. Nonetheless, a thermal history similar to the Arrigetch Peaks s is suggested by the K-feldspar age spectrum, as well as an 40 Ar/ 39 Ar white mica cooling age of Ma from the same [Christiansen and Snee, 1994]. 6. Discussion [40] The thermal histories for the different metamorphic zones along the transect have been reconstructed in Figure 6 by combining the K-feldspar, mica, and hornblende 40 Ar/ 39 Ar data with peak metamorphic temperature estimates. The data delineate two distinct cooling events: (1) an Albian-Cenomanian event defined by hornblende and mica data and (2) an early Tertiary event delineated by the K-feldspar ing. These two events will be discussed separately, in sections 6.1 and 6.2. [41] For simplicity in graphic representation and discussion we use T C of 500 C for hornblende and 325 C for micas. A single T C for micas was chosen because there is no obvious consistent systematic difference between white mica and biotite ages from the same areas. Different choices for closure temperatures do not affect the general conclusions drawn from the isotopic ages Mid-Cretaceous Cooling and Exhumation Cooling History. [42] The hornblende data from the epidote-amphibolite facies core of the metamorphic culmination suggest that the SEA, CEA, and NEA zones all cooled through hornblende closure temperature at Ma. Since estimates of peak metamorphic temperatures from the lowermost epidoteamphibolite facies SEA and NEA zones (Figure 3) [Vogl, 2000] are very similar to estimated hornblende T C, (500 C ± 50 ) [McDougall and Harrison, 1988] the Ma ages closely approximate the onset of cooling from peak temperatures in the core of the culmination. [43] Although initial cooling from the metamorphic peak began at approximately the same time throughout the entire epidoteamphibolite facies metamorphic core, biotite and white mica ages display a well-defined northward younging trend from 100 Ma in the SEA zone to 90 Ma in the NEA zone (Figure 3). The southern part of this trend is defined by only one of our new biotite dates ( APK90-38) and one published white mica age ([Patrick et al., 1994] APK90-119); however, several other unpublished ages (A. Till, written communication) strongly reinforce this trend. The age trend indicates that time-averaged cooling rates between hornblende and biotite closure increased systematically southward across the metamorphic culmination. Using the closure temperatures defined above we calculate average cooling rates of 35 C/m.y., 22 C/m.y., and 13 C/m.y. for the SEA, CEA, and NEA zones, respectively. [44] The K-feldspar data and a zircon fission track age of 76 Ma for the SEA zone [Blythe et al., 1997] indicate that the epidoteamphibolite facies core resided at temperatures near C during most of the Late Cretaceous (Figure 6). Thus cooling had slowed dramatically around the time of, or shortly after, passage through mica closure temperatures. Such prolonged residence at these temperatures is also consistent with the decrease in step ages in the low-temperature steps of the biotites in the NEA zone (Figure 4e). [45] The lack of Cretaceous plutons in the Brooks Range indicates that the regional cooling event beginning circa Ma is likely the result of exhumation. We have performed simple one-dimensional ing to provide some first-order constraints on the rate and duration of exhumation within the NEA and SEA zones. Results suggest that exhumation rates of 1 to 8 mm/yr are necessary to reproduce the hornblende and mica ages in the NEA and SEA zones, with the lower end of the range corresponding to the NEA zone. These relatively rapid exhumation rates could not have been sustained for more than a few million years because more prolonged moderate/rapid exhumation would have produced cooling below 250 C, which would conflict with K-feldspar and zircon data. Thus exhumation rates must have slowed considerably by the earliest Cenomanian Exhumation mechanisms and history. [46] Our data indicate that during the Albian-Cenomanian exhumation event the entire metamorphic culmination was being exhumed (Figure 7). Exhumation rates increased southward, however, suggesting that exhumation occurred in a hinged fashion, with a hinge located north of the study area (Figure 7). The hinged exhumation likely produced a north side down tilt to the region (Figure 7). We interpret the hinged exhumation as the result of large-scale south

13 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY 2-13 Figure 7. Diagram depicting exhumation mechanisms for metamorphic rocks along the transect by using a simple reconstruction at 96 Ma (immediately following the mid-albian to early Cenomanian exhumation event) of presently exposed geology. Dashed lines show northward tilt produced from more rapid exhumation in south due to normal faulting. Upper dashed line is an approximate minimum estimate for the amount of post metamorphic-peak exhumation. Arrows represent total amount of cooling prior to 96 Ma, which qualitatively shows the differential exhumation of the different areas. Differences in arrow lengths suggest probable slight doming centered on CEA during this event. S 2 foliation orientations and associated late-stage shear sense are shown across the culmination. Penetrative thinning refers to gently dipping crenulation cleavages, extensional shear bands, and boudinage that record shortening in the vertical direction. Grid shows thermal gradient of 30 C/km. Movement on the Takahula Fault and 10 of Tertiary northward tilting north of the Takahula Fault are removed. side down normal faulting along the south flank of the range. Extensional faulting may have directly exhumed the regions now composing the southern flank of the culmination (Figure 7). Farther north, this faulting may have contributed to exhumation indirectly by creating surface uplift from isostatic rebound of the footwall, leading to high rates of erosional denudation recorded by molasse deposits in the flanking basins (Figure 7). Without knowing the original fault dip it is difficult to quantify the relative contributions of erosion and normal faulting to exhumation. Modeling of the cooling age data suggest that this extensional event was short-lived, lasting only a few million years. [47] Although the distribution of exhumation/cooling rates can be related to extensional faulting in a simple way, the northward younging of mica ages away from the normal faults is more difficult to relate to extensional faulting. In core complexes, cooling ages in the footwalls of major extensional faults typically become younger toward the fault, with such age trends commonly attributed to the progressive dragging out of the metamorphic rocks beneath an extensional detachment [e.g., Harms and Price, 1992; John and Foster, 1993; John and Howard, 1995; Lee, 1995; Foster, 2000]. The mica-age trend recorded along our transect is opposite to the trend commonly observed, suggesting that a core complex analogy [Little et al., 1994] is not appropriate and that other processes were operative. In addition to the southward increase in cooling/exhumation rates, two important features must be accounted for by any exhumation. (1) The highest-grade metamorphic rocks occur well to the north of the large-displacement extensional faults on the south flank of the range (Figure 3). (2) The well-documented northward younging mica-age trend cuts across the S 2 fabrics and the metamorphic zonation on the north flank of the culmination (Figure 3). Although the first observation can be explained by postmetamorphic doming, the second observation suggests a more complicated evolution because, in detail, the cooling ages do not young inward from the lower grade flanks toward the highest-grade zone (Figure 3). [48] The metamorphic and cooling-age spatial distribution is best explained by two episodes of tilting/doming that are superimposed on the regional northward tilt inferred to have formed during extensional faulting. The first is a slight doming centered on the highest grade metamorphic rocks near the southern margin of the Arrigetch Peaks orthogneiss (CEA zone). The combination of peak temperature estimates and mica age data suggests that the largest amount of exhumation prior to 96 Ma likely occurred in the CEA zone (Figure 7). Thus this initial doming probably occurred during the Albian event. Doming was accommodated by normal faults and extensional shear bands that dip away from culmination core (Figure 7; see section 3 for description). Furthermore, the last stages of movement along S 2 crenulation cleavages may have also had a component of shear with movement directions away from the core [Dinklage, 1998; Vogl, 2002].

14 2-14 VOGL ET AL.: EXHUMATION FROM 40 AR/ 39 AR THERMOCHRONOLOGY [49] The second tilting/doming episode was a north side up rotation between the NEA zone and CEA/SEA zones, which is required to account for the 90 Ma and younger mica ages in the NEA zone (Figure 7). If cooling rates were slow by Ma, as indicated above, this tilt could be very small. This north side up rotation may have occurred during a period of renewed doming in the Tertiary and/or SW side up movement on the Takahula Fault. These Tertiary events are discussed in detail in section 6.2. [50] While the cooling patterns may relate well to an extensional event, the geodynamic setting (lithospheric-scale extension or syncontractional extension driven by deeper level thrusts) of the extensional exhumation is not very well constrained. Given the available data, we favor a postcontractional setting for two reasons: (1) The mica and K-feldspar data presented here suggest very little cooling during most of the Late Cretaceous, and cooling ages between 85 and 60 Ma are extremely rare overall. The lack of cooling may be an indication of tectonic quiescence beginning at the start of the Late Cretaceous. (2) The youngest identifiable Cretaceous contractional structures in the foreland region are syndepositional folds of early- to mid-albian deposits of the Fortress Mountain Formation interpreted by Cole et al. [1997] to have formed above active thrust faults. Tertiary folds and thrust faults have also been documented, however, structures of intermediate (Late Cretaceous) age have not been identified. Structures of this age may exist, but may be difficult to identify. [51] The synchronicity between relatively rapid mid-albian to Cenomanian exhumation and high depositional rates of molasse in flanking basins indicates that erosion played an important role in exhumation. In the foreland basin to the north, the Fortress Mountain Formation, Torok Formation, and Nanushuk Group together compose an Albian to early Cenomanian coarsening and shallowing upward flysch-to-molasse transition sequence up to 6 km thick that prograded northward and eastward, filling the foreland basin [e.g., Mull, 1985]. Barremian and Aptian sedimentary rocks are not exposed in the foreland basin. A possible sequence (<1.5 km thick) of this age has been identified in seismic sections by Cole et al. [1997], whose subsidence ing suggested that the foreland basin during this time was highly underfilled. This contrasts with Albian-Cenomanian sedimentation, which filled the basin. These relationships suggest that the onset of rapid cooling approximately coincides with the onset of high depositional rates in the foreland basin. As noted above, this increase in erosion rates may have been related to footwall uplift below the large-scale normal faults. [52] In the Koyukuk basin south of the Brooks Range, an Albian to Santonian flysch-to-molasse sequence up to 4 km thick is observed [Nilsen, 1989]. The basal deep-water flysch of this sequence overlies Valanginian shallow-water volcanogenic strata, indicating Hauterivian to Aptian subsidence in the Koyukuk basin. The subsidence significantly predates the onset of cooling at Ma, suggesting that basin subsidence and cooling/exhumation are not both related to the same extensional event as suggested by Miller and Hudson [1991]. The coarser clastic sedimentation, however, was coeval with cooling, further suggesting a strong erosional component of exhumation Regional Extent of Mid-Albian to Cenomanian Cooling Event. [53] Several lines of evidence suggest that the mid-albian to Cenomanian exhumation, recorded in the study area, affected much of the southern Brooks Range: 1. K-Ar and 40 Ar/ 39 Ar mica ages of Ma have been reported from metamorphic rocks in many areas to the east and west of the study area [Brosge and Reiser, 1964; Turner et al., 1979; Christiansen and Snee, 1994; Toro, 1998]. Although these ages may be found in close proximity to locally more numerous older ages (typically Ma), they are widespread suggesting that Albian-Cenomanian cooling/exhumation affected a large area. 2. On the basis of 40 Ar/ 39 Ar mica data, Christiansen and Snee [1994] suggested that extensional faulting and shearing along the southern flank of the range in the Cosmos Hills (Figure 1) occurred at, and after, 103 Ma. Their white mica 40 Ar/ 39 Ar age spectra are variably disturbed with most yielding individual step ages of 103. Except for the deepest with a preferred age of 103 Ma, which was interpreted as a cooling age, all of the white micas were interpreted by Christiansen and Snee [1994] as having grown synkinematically during extensional shearing below mica T C. No hornblende data were available to determine cooling rates. 3. On the basis of apatite fission track data, O Sullivan et al. [1991] and Blythe et al. [1998] inferred a period of anomalously high cooling rates of 10 C/m.y. between 100 Ma and 95 Ma for relatively unmetamorphosed rocks of the southern fold-thrust belt along the Dalton Highway. [54] Because our conclusions are drawn from cooling data, we have not assessed synmetamorphic exhumation. Vogl (submitted manuscript, 2000) has semiquantitatively reconstructed P-T paths across the culmination. On the basis of inclusion thermobarometry, garnet zoning, and the lack of high-pressure minerals he suggested that the NEA zone did not undergo more than a few kilometers of exhumation prior to the metamorphic peak and initial cooling at 103 Ma. In contrast, significant amounts of synmetamorphic exhumation were required for blueschist-facies rocks on the south flank to escape epidote-amphibolite grade overprinting during the Albian. Thus, significant synmetamorphic exhumation was limited in regional extent, whereas postmetamorphic exhumation that produced cooling affected the entire Brooks Range Tertiary Cooling and Exhumation [55] Our K-feldspar ing indicates a period of relatively rapid cooling beginning in the earliest Tertiary (Figure 6). Onedimensional thermal ing suggests that denudation rates of 0.5 mm/yr are required to produce the cooling rates estimated from the diffusion-domain ing. The early Tertiary in the Brooks Range was characterized by contractional deformation, recognized by east-west trending open synclines and tight, locally thrust-cored anticlines that involve Upper Cretaceous sedimentary rocks of the foreland basin [e.g., Mull, 1985]. During this time the Brooks Range is believed to have been carried upon a detachment presently at a depth of 12 km beneath the range front and >25 km in the southern Brooks Range [Fuis et al., 1997]. Surface uplift and erosion associated with early Tertiary contractional deformation has been suggested to account for Paleocene cooling that is well documented by apatite fission track data [e.g., O Sullivan et al., 1991, 1993, 1998; Blythe et al., 1998]. This cooling affected much of the Brooks Range, including the foreland basin, range front, fold-thrust belt, and metamorphic rocks of the southern Brooks Range. Thus the Latest Cretaceous to early Tertiary cooling of the epidote-amphibolite facies culmination documented here through K-feldspar diffusion-domain ing is also likely related to this north directed thrusting and erosional denudation. Specific struc-