GEOPHYSICAL RESEARCH LETTERS, VOL. 20, NO. 15, PAGES i635-1638, AUGUST 6, 1993 IMPLICATIONS OF NEW GRAVITY DATA FOR BAIKAL RIFT ZONE STRUCTURE C. Ruppel Woods Hole Oceanographic Institution, Woods Hole, Massachusetts M.G. Kogan Institute of Physics of theEarth, Moscow, Russia M.K. McNutt Massachusetts Institute of Technology, Cambridge, Massachusetts Abstract. Newly available, 2D Bouguergravity anomalydata from the Baikal Rift zone, Siberia, indicate that this discrete, intracontinental rift system is regionally compensated by an elastic plate-50 km thick. However, spectral andspatial domain analyses and isostaticanomaly calculations show that simple elastic plate theorydoes not offer an adequate explanation for compensation in the rift zone, probablybecause of significant lateral variations in platestrength andthe presence of subsurface loads. Our results and other geophysical observations support the interpretation that the Baikal Rift zone is colder than either the East African or Rio Grande rift. Introduction This studyfocuses on standard forwardmodeling and spectral domain analyses of Bouguer gravityanomaly datafrom the Baikal Rift Zone (BRZ), which were first made available to us in early 1992. In thepast, workers have used unpublished BRZ gravity data to estimate lithospheric thickness [Zorin et al., 1989] and to constrain possible asthenospheric upwelling [Zorin, 1981] and passive thermal upwelling [Diament andKogan,1990]processes at depth. In this study, we undertake a conventional analysis of the newlyavailable data to estimate lithospheric flexural rigidityand to determine thedegree of localvs.regional compensation. Tectonic Setting The BRZ lies along a Paleozoic suture [Zamarayev and Ruzhich, 1978; Zonenshain et al., 1990] separating the Siberian platformon the west from the Paleozoic Sayanfolded zone on the east. Lake Baikal dominates the rift's axial zone, but comprises only threeof the rift's -15 major basins. The rift basinsform a sinistral (S- shaped) pattern stretching from 400 km westof the lake's southern tip, through Lake Baikal, to 600 km east of the lake's northern tip. The principal tectonic elements of the BRZ are shown in Figure la and includethe Tunka-Sayan strike-slip region southwest of Lake Baikal, the rift's axial en echelon basins, the seismically active Barguzin region, andthe Udokan left-lateral shear zone. Portions of the BRZ have been undergoing active extensionat least since Pliocene time [Nikolayev et al., 1985] and probably since asearlyas Oligocene to Miocene time [Logatchev and Florensov, 1978]. Data Topography andgravitydatafor this study weredigitized from contour maps at 7.5' by 5' intervals (-10 km x 10 km) by the Topographic Service of the ArmedForces of Russia (TSAFR). The map shown in Figure la representsa compilation of TSAFR topography data and bathymetric data from Hutchinson et al. Copyright 1993 by theAmerican Geophysical Union. Paper Number 93GL01873 0094-8534/93/93GL-01873503.00 [1992]. Elevations in the BRZ rangefrom a low of 1045 m below sealevel (water depth 1500 m) in Lake Baikal'sCentralBasinto a high of over 3000 m in the Tunka Alps, southwest of the lake. Elevations average over 1000 m in the Tunka region and folded zone, but only 400-700 m in the Siberian platform. The Bouguer gravity anomaly map shownin Figure lb was compiled from measurements takenwith a SovietGAK-3 Worden- type gravimeter (meter precision: 0.02-0.03 mgal; anomaly accuracy: -2-4 mgal). Individualsurveys weretied to regional base stations, where gravity values had been calibrated using both absolute and Lacoste-Romberg type gravimeters. The original researchers reduced the data usingthe 1967 International Gravity formula andcrustal density of 2300kg/m 3 and report onlythe complete Bouguer anomaly, including terrain corrections within 200 km of gravity stations. Bouguer slab corrections, but not terrain corrections, were subsequently recalculated usinga crustal density of 2670kg/m 3. The 10%difference in densities between theold andnew datareductions means thatterrain corrections may be too low byseveral milligals inrugged areas ofthe riftzone. The spacingof the original surveypoints and the quality of elevation measurements used for Bouguer corrections areimpossible to assess without as-yet unavailable original (ungridded) data. Overinterpolation of sparse data likely accounts for some unexpected featuresin the gravity map, including prominentcomers in the contoured data (label d in Figure 1 b). Errors of 10 m in elevation changethe Bouguer correction by >1 mgal. We conservatively estimate datauncertainties to be aslargeas 10 mgal in rugged areas. Bouguer anomalies range from a low of -275 mgal overthe thick sedimentary fill of Lake Baikal's South Basinto a high of -35 mgal over rift flanking topographic highs. Important featuresin the Bouguer field (Figure lb) are: (a) sharp horizontal gradients across the Udokan shear zone andreactivated Main Sayan fault, (b) long- wavelength lows over the Barguzin and Tunka areas, and (c) lineations parallel to dominant NE-trending topographic fabricof the Sayan foldedzone. Sediment Corrections Recentmultichannel seismic surveys [Hutchinson et al., 1992] haverevealed over7 km of sedimentary fill in the South andCentral basins of Lake Baikal and over 4 km in the North basin. These thick sedimentary sequences constitute an important low-density subsurface load in the rift zone and are responsible for large- amplitude, short-wavelength, Bouguer anomaly lows over Lake Baikal. Attempts to match these observed Bouguer anomalies lows assuming a wide range of density-depth relations for the sedimentary fill--including constant density models andexponential compactionschemes [Cowie and Karner, 1991]--have proved largely unsuccessful, particularly over the South Basin wherethe observed anomaly is profoundly negative [Ruppel,1992]. For the analyses below,the sediment gravity correction has been calculated assuming compaction parameters thatpredict a very negative signal (high surface porosity •0 =0.6 andlarge depth to whichporosity is reduced to 1/e of itsinitial value kp=4 kin). These values are close to •0 measured by S. Colman (pers.comm.) in Lake Baikal basin 1635