The Mechanics of Plotting MMI Intensity Maps in ShakeMap Format

This website presents Modified Mercalli Intensity (MMI) maps for the great San Francisco earthquake of April 18, 1906. These new maps combine two disjoint but complementary developments. First, we have comprehensively re-evaluated the intensities inferred from the damage and shaking reports compiled by Lawson (1908) for California, Oregon, and Nevada, and added intensities inferred from local newspaper accounts, survivor narratives, and damage to rural cemeteries. Second, we use the recent ShakeMap methodology (Wald et al., 1999) to map these intensities. This methodology gauges the intensity estimates using site amplifications derived from the surficial geology. The irregularly spaced estimates are then interpolated rather than contoured, and the mapped geology used to reapply site amplifications everywhere. This methodology yields detailed and spatially variable maps of estimated intensity.

The new Modified Mercalli Intensity maps (Map index page) are built from the same fundamental data as Stover and Coffman’s (1993) contoured MMI intensity map and the contoured Rossi-Forel intensity maps published in Lawson (1908). For many sites in the Bay Area, as well as in Santa Cruz, Monterey, Sonoma, Mendocino, and Humboldt counties, the descriptions of damage and shaking are relatively complete. For more distant sites in the Sacramento and San Joaquin Valleys, the Sierra Nevada foothills and mountains, or in Oregon and Nevada, however, the descriptions can be extremely brief, often no more than a list of localities where the earthquake was either “slightly felt” or not felt. In generating the new MMI intensity maps, all of the localities cited by Lawson (1908) were identified, located, and geocoded, including many small towns, post offices, and “stations” that have since disappeared, been renamed, or been inundated by reservoirs.

Care has been taken in assigning intensities to these shaking and damage descriptions, particularly for sites that suffered the most severe shaking. The large moment release and long duration of shaking in the 1906 earthquake modified the phenomenology of many shaking effects: long-period effects such as stopped clocks and seiches occur at lower intensities than in the Modified Mercalli Scale revised by Richter (1958). We have adjusted the intensities associated with these effects to maintain the general grouping of effects reported in Lawson (1908). These groupings are encapsulated and discussed in the 1906 Modified Mercalli Scale. We diverge from some previous researchers in stressing damage to chimneys, structures, cemetery monuments, and trees more strongly than ground failure effects such as liquefaction and landslides.

The sheer number of sites (some 624 locations outside the City of San Francisco) described by Lawson (1908) yields a remarkably dense set of intensity estimates in many areas. We have added to this set of intensities by reviewing other historical sources, thereby obtaining intensity estimates at many more sites than previous researchers. But the clearest difference between the old and new maps is derived from the different mapping techniques. Both Lawson (1908) and Stover and Coffman (1993) contour the intensity data: in drawing contours at relatively small scales, they smooth the variation in shaking between soil and rock. In contrast, the interpolation in ShakeMap allows this variation to be as detailed as the digitization of the map.

ShakeMap fills out the unequally spaced intensity estimates by interposing false sites in areas without data. The attenuation relations used to generate the intensity estimates at these sites are mixtures of the PGA and PGV relations from Boore et al. (1997) where the overall amplitudes (the source terms) are adjusted to fit the actual PGA and PGV data. To fit the 1906 intensities at distances beyond the source-receiver distances for which Boore et al. (1997) derive their regression curves, we have added an anelastic attenuation term, exp[-0.0035 rJB] where rJB is the Joyner-Boore distance to the fault rupture, in kilometers. This modification of the Boore et al. (1997) attenuation relation is shown in a separate Graph, and the fitting procedure is described in detail in the accompanying caption.

Plotting the intensity estimates against the Boore et al. (1997) attenuation curves yields two unexpected results. First, the fitted attenuation curves for PGA and PGV both overestimate the 1906 intensities near the fault (that is, for rJB < 10 km). Previous authors such as Borcherdt et al. (1975) obtain systematically high intensities (MMI ≥ 10) near the fault. This difference in MMI intensities is derived from different interpretations of landslide reports in the Richter (1958) and Stover and Coffman (1993) MMI scales: we follow Stover and Coffman (1993) and only assign MMI 9 to reports of massive landslides, such as the landslide that buried the Loma Prieta sawmill and killed nine mill workers.

Second, the intensities at distances of 40 < rJB < 90 km from the fault appear to systematically exceed the intensity predicted from the attenuation curves. This difference is derived from the breakdown of the Boore et al. (1997) attenuation relations for great strike-slip earthquakes: at these distances, the 1906 fault rupture should be modeled as a line source rather than a point source, and the amplitude of the radiated waves should falloff more slowly with distance. We note, too, that if the rupture velocity of the earthquake to the north exceeds the S-wave velocity in places on the fault, as Song et al. (2004) claim, we expect the attenuation with distance from the fault to differ from the attenuation obtained for smaller, slower, earthquakes.

There is substantial variability along-strike, as might be expected for sites arrayed along more than 400 km of fault, with a potential variety of crustal structures. Our map agrees with Stover and Coffman’s (1993) map for most of these variations: for example, the MMI 7 area in the southwestern San Joaquin Valley associated with the towns of Newman, Volta, and Los Banos, and the MMI 3 area in the Sierra Nevada foothills, running from Angels Camp north to Oroville. But there are also disagreements: the new map obtains MMI 8 for the Sacramento delta which Stover and Coffman (1993) map as MMI 7, and MMI 7 for the area west of Clear Lake which Stover and Coffman (1993) contour as MMI 8. The MMI 8 area that Stover and Coffman (1993) infer at the northern end of Anderson Valley (~20 km southeast of Ukiah) is derived from a mislocation of the town of Navarro.

Overall, there are relatively few interpretative disagreements between the new interpolated map and the previous contoured map. The greater difference is derived from the increased number of intensity sites in the new map and ShakeMap’s ability to resolve densely sampled data. Developed more than 90 years after the 1906 earthquake, ShakeMap provides a pliant and detailed canvas for the 1906 intensity survey. The power of these new intensity maps is a tribute to the diligence and foresight of Andrew Lawson and his co-authors John Branner, Roderic Crandall, Arthur Eakle, Harold Fairbanks, Karl Gilbert, Ruliff Holway, Esper Larsen, George Louderback, François Matthes, Gerald Waring, Harry Wood, and George Zoffman.


Many colleagues and friends made significant contributions to this work. Mary Lou Zoback’s request for a ShakeMap of the 1906 earthquake was the initial impetus for our study, and her interest and support sustained us during the very long course of the work. Bill Bakun gave us a digital copy of the Stover and Coffman (1993) intensity file and reviewed the report. Tousan Toppozada also reviewed the report and made many helpful recommendations. Linda Seekins determined site categories for the intensity sites. Emily Ellis wrote many of the intensity summaries in the "Site List" section of the report. Luke Blair digitized the Stover and Coffman (1993) isoseismal map. Aron Meltzner compiled a digital copy of Toppozada and Parke’s (1982) intensity file, and identified a renamed town in Mendocino County. Seena Hoose’s Professional Paper contributed many ground failure sites not in Lawson (1908). Ellin Beltz’s online post of the Ferndale Enterprise articles on the 1906 earthquake convinced us of the importance of local newspaper reports. The Blue Lake Advocate, edited by Gus Perigot, compiled remarkable reports on the earthquake effects in Humboldt County. Lew Rosenberg suggested that we read the Watsonville Pajaronian. Glennda Chiu of the San Jose Mercury News mustered her readers to help us locate sites. David Durham’s authoritative Gazetteer verified almost all of our relocations. Carl Wentworth challenged us to consider cemeteries that the Lawson co-authors had not visited. Phoebe Boatwright helped us find and canvass cemeteries in Marin, Sonoma, Mendocino, Santa Cruz, and San Benito Counties; Sumner Hearth, Lauren Hall, Karen Kuhlman, and Peggy Flick also helped us canvass cemeteries. Anita Crabtree’s online photographs of the Upper Lake Cemetery obviated a trip there. Amie Hill’s history of Occidental and John Schubert’s article gave us two critical intensity sites in western Sonoma County. Craig DePolo pointed us to a set of Lawson (1908) reports for Nevada that we had missed. Bob Nason’s (1982) Open-File Report added five sites in Contra Costa County. Sylvia Bartley’s collection of earthquake narratives significantly improved our coverage in Mendocino County. We also found useful narratives in the Bancroft Library’s The 1906 San Francisco Earthquake and Fire collection. Heather Lackey relocated some crucial Bancroft photographs. A visit to Point Reyes with Cindy Carpien and John McChesney of National Public Radio helped us locate some of Gilbert’s sites. In addition to noting the lack of the Lawson intensity sites in southeastern Santa Cruz County, Nan Shostak contributed critical intensity reports for Santa Clara County. We are thankful for their help; this work would not have been completed without it.


Bartley, Sylvia E., (2006). The 1906 Earthquake on California’s North Coast, Fort Bragg-Mendocino Coast Historical Society and Mendocino Museum, Fort Bragg, CA.

Blue Lake Advocate, The Earthquake (April 21, 1906), Blue Lake, CA.

Boatwright, J. and D.K. Keefer, (2005). The distribution of violent shaking in the 1906 San Francisco earthquake (abs.), Seism. Res. Lett., 74, in press.

Boore, D.M., W.B. Joyner, and T.M. Fumal, (1997). Estimation of response spectra and peak accelerations from western North American earthquakes: a summary of recent work, Seism. Res. Lett., 66, 128-153.

Borcherdt, R.D., J.F. Gibbs, and K.R. Lajoie, (1975). Maps Showing Maximum Earthquake Intensity Predicted for Large Earthquakes on the San Andreas and Hayward Faults, U.S. Geological Survey Misc. Field Studies Map MF-709, Scale 1:125,000.

David L. Durham, (1998). California’s Geographic Names, Word Dancer Press, Clovis, CA.

Ferndale Enterprise, Terrific shock of earthquake (April 20, 1906) and And still the earth trembles (April 24, 1906), Ferndale.

Fortuna Beacon, Reports of the Earthquake (April 20, 1906), Fortuna, CA.

Healdsburg Enterprise-Tribune, Healdsburg not so bad as many other places, (April 21, 1906), Healdsburg.

Hill, A. and H. Lapham, (1997). An Historical and Anecdotal Walking Tour of the Small (but Fascinating) Village of Occidental, Occidental.

Joyner, W.B., and D.M. Boore, Peak horizontal acceleration and velocity from strong-motion records including the records from the 1979 Imperial Valley, California, earthquake, Bull. Seism. Soc. Amer., 71, 2011-2038, 1981.

Keefer, D.K. (1984). Landslides caused by earthquakes, Bull. Geol. Soc. Am., 95, 406-421.

Lawson, A.C. (ed.), (1908). The California Earthquake of April 18, 1906, 1908, Reprinted 1969 by the Carnegie Institution of Washington, D.C.

Mallet, R. (1862). Great Neapolitan Earthquake of 1857. The First Principles of Observational Seismology. London, Chapman and Hall.

Nason, R., (1982). Damage in Alameda and Contra Costa counties, California, in the earthquake of 18 April 1906. U.S. Geological Survey Open-File Report 82-63.

Nason, R., (1984). Food store disturbance as a seismic intensity indicator, Bull. Seism. Soc. Am., 74, 988-991.

Richter, C., (1958). Elementary Seismology, W.H. Freeman and Co., San Francisco.

Santa Cruz Sentinel, Earthquake Notes (April 20, 1906), Santa Cruz, CA.

Schubert, John, (2006). West County Quake Memoirs, Sonoma Historian, no. 1, Simone Wilson, ed., Sonoma County Historical Society, Santa Rosa, CA.

Song, S., G.C. Beroza, and P. Segall, (2004). Another look at the source of the 1906 San Francisco earthquake, Eos Trans. AGU, S13-E08.

Stover, C.W., and J.L. Coffman, (1993). Seismicity of the United States, 1568-1989 (Revised), U.S. Geological Survey Professional Paper 1527.

Titus, Louis Ehrmann, The Mountain Moves, Bancroft Library, MSS 73/122 c:88.

Toppozada, T. R., and D. L. Parke, (1982). Areas damaged by California earthquakes, 1900-1949, Calif. Div. Mines Geol. Open-File Report. 82-17 SAC, 65 pp.

Ukiah Dispatch-Democrat, Mendocino County’s Total Loss $1,000,000 and From the Valleys (April 27, 1906), Ukiah, CA.

Wald, D.A., V. Quitoriano, T.H. Heaton, and H. Kanamori, Relationships between peak ground acceleration, peak ground velocity, and modified Mercalli intensity in California, Earthquake Spectra, 15, 557-564, 1999.

Watsonville Evening Pajaronian, This Section Visited by a Fearful Earthquake (April 18, 1906) and Doings of Our Late and Very Unwelcome Visitor (April 20, 1906), Watsonville, CA.

Willits News, Effects of the Earthquake (April 26, 1906), Willits, CA.

Youd, T.L., and Hoose, S.N., (1978). Historical ground failures in Northern California triggered by earthquakes: U.S. Geological Survey Professional Paper 993.


Modification of the Boore et al. (1997) Attenuation Relation

Map showing Nevada and Eastern Sierra Sites