Heat Flow and Hydrothermal Circulation in the Cascade Range，North-Central Oregon

Quaternary volcanoes of the Cascade Range form 1200-km-long volcanic arc that extends from southern British Columbia to northern California. The arc is related to subduction of the Juan de Fuca Plate beneath North America. Detailed geologic mapping，measurements of advective heat discharge， and numerous conductive heat flow measurements are available for a 135-km-long section of the Cascade Range in north-central Oregon. This data set allow s us to estimate fluxes of heat and mass ( thermal fluid and magma) and to document the role of groundw ater movement in redistributing heat in the upper crust. The results provide some insight into the thermal structure of the arc，and have implications for its geothermal resource potential.

Figure 4. 4. 1 Map showing the location of hot springs，the Quaternary arc，prominent volcanoes ( △) ，the 1500 － m － elevation contour， and the amount of heat transported advectively by the hot spring systems. The total for the southerly group of hot springs ( 26 MW ) is 1. 5 times the value obtained from the individual spring groups ( Table 4. 4. 1) ，because of diffuse input of thermal water into the surface drainage; hot springs: Au，Austin; Ba，Bagby; Br， Breitenbush; Bi，Bigelow ; Be，Belknap; Fo，Foley; Ri， unnamed spring on Rider Creek; Ka，Kahneeta.

In the study area ( Figure 4. 4. 1) ， the High Cascades physiographic subprovince is a broad constructional ridge of upper Pliocene and Quaternary ( ＜ 3. 3 Ma ) volcanic rocks surmounted by several Quaternary stratovolcanoes. The High Cascades are flanked by Oligocene to lower Pliocene volcanic rocks of the Western Cascades to the west and Deschutes basin to the east. Western Cascades rocks also underlie the High Cascades; they are generally less permeable than the younger rocks. In this report，we use Western Cascades and High Cascades as location terms. We also distinguish ① the Quaternary ( ＜ 2 Ma) arc，or area of Quaternary vents ( Figure 4. 4. 1) ， because older magmatic heat sources will generally have cooled to near ambient temperatures，and ② the region where uppermost Miocene ( ＜6 Ma) ，Pliocene，and Quaternary rocks are exposed， because their areal extent is roughly coincident with an extensive area of near-zero near-surface conductive heat flow.

Most of the thermal springs in the study area discharge in deep valleys in the Western Cascades up to 15 km w est of the Quaternary arc. One set of thermal springs discharges from Oligocene or Miocene rocks 18 km east of the Quaternary arc ( Figure 4. 4. 1) . No thermal springs occur in the Quaternary rocks. With tw o exceptions，the thermal w aters are Na-Cl or Na·Ca-Cl w aters. The ratios of Br to Cl are similar to those in seaw ater; these ratios and the high concentrations of Na and Cl ( Table 4. 4. 1 ) suggest that the thermal w aters may have circulated through rocks deposited in a marine environment. Thermal Na · Ca-Cl w aters are typical of rift zones around the w orld，but in North America occur primarily in the Salton trough and in the Columbia embayment of the Pacific Northw est.

Table 4. 4. 1 Geochemical and discharge data for hot springs in the study area

* Discharge based on chloride-flux measurements，except for Bagby Hot Spring，where discharge was measured directly.

** Discharge temperature. ** * Chemical geothermometer temperatures based on anhydrite saturation，except for Kahneeta and Bagby，which are based on the silica and cation geothermometers. § Combined discharge of Bigelow and Belknap Hot Springs. Dashes indicate lack of data. Hot spring locations are shown in Figure 4. 4. 1.

The high Cl content of the thermal w aters makes Cl a useful natural tracer，because surface w aters in the Cascade Range of Oregon contain only about 0. 5 mg / L of Cl. The discharge of groups of thermal springs can be calculated by measuring the increased Cl load of streams passing through hot spring areas. In repetitions done on different occasions，these measurements ( Table 4. 4. 1) have a reproducibility of ± 10% to ± 15% or better，depending on the flow rate of the stream relative to the flow rate of the hot springs and on the Cl concentration in the thermal w ater.

Figure 4. 4. 2 Relation betw een deuterium content and elevation for w aters on or w est of the Cascade crest. Deuterium content ( δD) is expressed as D / H ratios in ‰ relative to SMOW ( Standard Mean Ocean Water) . Filled squares are from Na-Cl and Na·Ca-Cl thermal w aters in the Western Cascades. Open squares are nonthermal samples from low -salinity springs and w ells in zero-or first-order ( unchanneled or headw ater) basins on or w est of the Cascade crest，and represent local meteoric w ater. Line is linear least square fit to these data. Because there is little oxygen-18 shift in the thermal waters，δ¹⁸O values show a similar pattern.

The product of the measured hot spring discharge ( Q ) ，an appropriate density ( ρ) and heat capacity ( c) ，and the difference betw een a chemical geothermometer temperature ( Table 4. 4. 1) and a reference temperature ［Qρc( T_g－ 5℃) ］ gives a measure of the heat transported advectively by the hot spring systems ( Figure 4. 4. 1 ) . The total measured advective heat transport by thermal w ater in the study area is 159 MW ( 159 × 10⁶W ) . For comparison，the Quaternary magma extrusion rate of 3 to 6 km³per kilometer of arc length per million years represents an average heat release of 60 MW to 120 MW in the study area.

The isotopic composition of the thermal w aters in the Western Cascades indicates that they w ere recharged at relatively high elevations in the Quaternary arc，if the isotopic composition of precipitation has not changed significantly since the thermal w aters w ere recharged. This w ould be the case if the thermal w aters w ere recharged during the Holocene， as seems likely. The thermal w aters are much more depleted isotopically than local meteoric w aters in the Western Cascades. Their isotopic composition best matches that of meteoric w aters at elevations of 1300 m to 1990 m near the Cascade crest ( Figure 4. 4. 2 ) . The hot springs are at elevations of 500 m to 700 m. Thus an elevation difference of about 600 m to 1400 m drives the thermal circulation systems. The available data suggest that thermal w aters recharged near the Cascade crest circulate to depth and flow laterally for distances of 10 km to 20 km ( Figure 4. 4. 1) before discharging at relatively low elevations in the Western Cascades. Gravitationally driven thermal fluid circulation transports significant amounts of heat from the Quaternary arc into rocks older than 6 Ma and must profoundly affect the pattern of near-surface conductive heat flow. Gravitationally driven flow of low er temperature groundw ater must also transfer heat from the younger rocks to older rocks at low er elevations，but this effect is difficult to measure directly.

Conductive heat flow data show that the Quaternary arc and adjacent 2-to 6-million-years- old volcanic rocks constitute a large area of near-zero near-surface conductive heat flow that results from dow nw ard and lateral flow of cold groundw ater ( Figure 4. 4. 3) . In contrast，near- surface conductive heat flow is anomalously high in rocks older than 6 Ma exposed at low er elevations in the Western Cascades ( Figure 4. 4. 4) . A similar pattern of low -to-zero conductive heat flow in permeable volcanic highlands and relatively high heat flow in older，less permeable rocks at low er elevations has been observed in the Cascade Range of northern California.

Figure 4. 4. 3 Typical temperature-depth profiles from the Quaternary arc， show ing little or no temperature increase to depths of 150 m or more

Figure 4. 4. 4 Temperature-depth profiles from drill holes collared in rocks older than 6 Ma in the Breitenbush Hot Springs area. The deepest hole w as completed to 2457 m，but w as only logged to 1715 m. The bottom-hole ( 2457 m ) temperature w as higher than 141℃ . The gradient measured over the 1465 m to 1715 m interval ( 31℃ / km) projects to a bottom-hole temperature of 152℃.

On the basis of temperature profiles from the Mt. Hood area，New berry Volcano，and this part of the Cascades，the thickness of the nearly isothermal zone in the younger rocks generally ranges from 150 m to 1000 m. In the study area only tw o drill holes collared in Quaternary rocks w ere deep enough that conductive heat flow beneath the nearly isothermal zone could be measured; the values measured w ere 95 mW / m²and 109 mW / m².

The temperature profiles in the Breitenbush area ( Figure 4. 4. 4 ) suggest that the high conductive heat flow measured in rocks older than 6 Ma may be a relatively shallow phenomenon. Seventeen shallow holes ( ＜ 500 m deep ) had high gradients that generally corresponded to heat flow s higher than 110 mW / m². How ever，a similar gradient in the upper part of the deepest hole ( SUNEDCO 58 － 28) changed abruptly below a zone of thermal fluid circulation at about 800 m depth; that such a change w as observed in the deepest hole suggests that the gradients in the shallow holes are also controlled by groundw ater flow.

We have used a heat budget approach ( Table 4. 4. 2) to compare the magnitude of the heat deficit in the rocks younger than 6 Ma with that of the anomalous heat discharge in the adjacent older rocks，and to estimate the magmatic heat input required to account for the total heat flow anomaly. This analysis ( Table 4. 4. 2) is specific to the section of the arc between 44°00"N and 45°15" N. The regional heat flow map used in our analysis is shown as Figure 4. 4. 5. The conductive components of the budget are defined relative to assumed background heat flow values and are obtained by measurement of areas on Figure 4. 4. 5 with a planimeter. In general，heat flow in a given area is taken as the average of adjacent contours ( for example，70 mW / m² between the 60 mW / m²and 80 mW / m²contours) . We assigned values of 140 mW / m²within the 120 mW / m²contours and 60 mW / m²outside the 80 mW / m²contours east of the Quaternary arc.

Table 4. 4. 2 Components of the heat budget ( in MW)

* Based on discharge temperatures. The difference between the geothermometer and discharge temperatures is due to conductive heat loss and，particularly in the Western Cascades，represents a significant fraction of the conductive anomaly.

Important assumptions in the heat budget analysis are as follow s: ① The background conductive heat flow beneath the isothermal zone in the Quaternary arc is 100 mW / m². This value is typical of areas of Quaternary volcanism and is consistent w ith the tw o measurements in the study area. ② The background heat flow in Tertiary terrane is 60 mW / m². Values over 60 mW / m²are the result of tectonic，magmatic，radiogenic，or hydrologic sources. ③ The heat output of the hot springs represents the anomalous，advective heat discharge from rocks older than 6 Ma. This is a minimum value because it does not include low er temperature advective discharge，w hich is difficult to measure.

The values for hot spring heat output used in the budget are based on discharge temperatures ( T_d) rather than the geothermometer temperatures ( T_g) used previously to calculate advective heat transport. The difference between T_gand T_d( Table 4. 4. 1) results from conductive cooling and presumably appears as part of the conductive anomaly. In the Western Cascades，the thermal power represented by the difference between T_gand T_d( 67 MW) is equal to about half of the conductive anomaly ( compare the values in Figure 4. 4. 1 and Table 4. 4. 2) .

The area of near-zero near-surface conductive heat flow in this part of the Cascade Range is generally coincident with the areal extent of permeable volcanic rocks younger than 6 Ma. On the basis of our assumptions regarding background heat flow ，about 460 MW of heat are swept out of these younger rocks between 44°00"N and 45°15"N by groundwater circulation. This amount is roughly balanced by about 350 MW of anomalous heat discharge in the rocks older than 6 Ma ( Table 4. 4. 2) . Apparently，sufficient heat is removed advectively from the rocks younger than 6 Ma to account for the anomalous heat discharge on the flanks of the Cascade Range. The difference betw een the heat deficit in the younger rocks and the anomaly in the older rocks ( about 110 MW) may occur as lower temperature advective discharge， which was not determined directly. The difference between the heat deficit in the younger rocks and the heat ( T_g) transported advectively by the hot spring systems ( 460 － 160 = 300 MW) is an estimate of the heat removed from the younger rocks by low er temperature groundw ater flow or by yet- unidentified thermal fluids.

Figure 4. 4. 5 Map show ing lines of equal heat flow. Area of near-zero near-surface conductive heat flow rocks younger than 6 Ma is diagonally hatched. This area is estimated conservatively; w e have included in it all areas w here rocks younger than 2 Ma are exposed but only included areas w ith rocks betw een 2 and 6 million years old w here temperature profiles indicate that conditions are near isothermal. Areas of silicic volcanism in the Quaternary arc are show n in region. Names of hot springs are on Figure 4. 4. 1. The heat flow contours are based on 253 temperature profiles，101 of w hich have also been previously published and interpreted.

Because the anomalous heat discharge in the older ( ＞ 6 Ma) rocks can be explained in terms of advection from the younger rocks， w e ne