The procedure used to calculate the Community Internet Intensity values was calibrated so that the Community Internet Intensity values should, on average, be similar to the Modified Mercalli Intensity values for the same communities. In the United States, intensities have for many years been assigned on the basis of the Modified Mercalli Intensity scale [Wood and Neumann, 1931, Richter, 1958]. [Dewey et al., 1995] document the current procedures for assignment of USGS MM intensities, which have been slightly modified over the years to account for difficulties encountered using the earlier conventions.
At low-to-moderate shaking levels, the Modified Mercalli Intensities are based largely on postal questionnaires, in which respondents summarize the effects of shaking in their communities. In stronger shaking areas, Modified Mercalli Intensities are based on field study in areas of significant damage, on damage maps produced by emergency response agencies, on reports produced by the earthquake engineering community, and on press reports. For a destructive earthquake, the process of collecting and interpreting damage data and preparing a map of Modified Mercalli Intensities can take months. We hope that the CII map will serve as a useful first approximation to the Modified Mercalli Intensity map in the early hours and days following damaging earthquakes during which time the final Modified Mercalli Intensity maps are being prepared. We envision that the responses collected from the CII questionnaires will be considered in final assignment of USGS Modified Mercalli Intensities, but we do not currently envision that USGS MMI will be based on numerical values of CII that results from equation (3). Because there are major differences in the data and procedures used to assign the two types of intensities, the Community Internet Intensities cannot be considered to be identical to the USGS Modified Mercalli Intensities.
We received over 800 questionnaire responses for the Northridge 1994 event between May, 1998, shortly after our system when online, and July, 1999. We find that there is a good correlation between USGS MMI (numbered circles) assignments of [Dewey et al., 1995] and the CII values derived from the recollections of the respondents (ZIP codes) as shown in Fig. 4. While there are occasional discrepancies of more than 1 intensity unit, these typically occur in ZIP codes where we received fewer than 5 responses.
Figure 4: Community Internet Intensity maps for the Northridge 1994 event using data collected between May, 1998 and July, 1999. Superimposed on the ZIP code areas are the USGS Modified Mercalli intensity values [Dewey et al., 1995] shown with number-filled circles.
The Community Internet Intensity (CII) maps were intended to be compatible with TriNet ``ShakeMap'' Rapid Instrumental Intensity Maps [Wald et al., 1999]. Like ``ShakeMap'', the CII maps are centered on the epicenter (star) of the earthquake and have similar overall dimensions as the ``ShakeMaps''. However, the ``ShakeMaps'' are based on point location measurements of the ground motion as recorded by seismometers; the Instrumental Intensity is inferred by empirically relating the recorded peak ground motions to MMI values and then interpolating the ground motions between the recording sites to complete the maps. In contrast, the CII maps provide intensity values only in areas (communities) where data have been provided; no interpolation is done, and hence the intensity in unrepresented areas is left undefined. Nonetheless, there is generally good overall agreement between the CII maps and the ShakeMap instrumental intensities (Figs. 4-9).
An example of a recent community Internet Intensity map is shown in Figure 6 for the M3.7 earthquake on June 3 1998, near Mt. Palomar, California. The corresponding ``ShakeMap'' Instrumental Intensity map is given in Figure 7. Figures 8 and 9 show a second example of a CII map and a ``ShakeMap'', respectively, for the August 20, 1998, M4.4 earthquake which occurred near Wrightwood, California. Note the area of locally higher intensities near Riverside (approximately 117.3 W., 33.9 N.) on the CII map (Fig. 8) is confirmed by the instrumental measurements shown in Figure 9.
Even with the current sparse dataset, the CII dataset has proven to be a useful tool for calibrating the ShakeMap intensity algorithm for low amplitude shaking [Wald et al., 1999]. Since there is very little recorded (and digitized) ground motion data for areas that experienced Modified Mercalli intensities of IV or less, there is little empirical basis for assigning low intensities based on current ground motion recordings. Using the CII responses that we have collected for various historical and recent earthquakes, we can begin to put constraints on the intensities for low amplitude shaking. Figure 10 shows a comparison of CII and ``ShakeMap'' instrumental intensity values for historical and recent earthquakes. Since the instrumental intensities are derived directly from peak ground motion values, this figure shows how recorded peak motions relate to CIIs. We have determined that the approximate boundary between ``felt'' and ``not felt'' shaking occurs at about 0.2 %g, at least for small to moderate-sized events. These relationships will revised as we obtain more CII data near TriNet seismic stations recording the same events.
Figure 5: Instrumental intensity ``ShakeMap'' from strong motion data recorded during the 1994 Northridge earthquake [Wald et al., 1999]. Fill corresponds to the intensity scale in the legend at the bottom of the figure. The epicenter is shown with a filled star; lines depict highways. Small circles show selected city locations as labeled. Also given in the scale bar are corresponding peak ground motion values, one- or two-word damage and perceived shaking descriptors.
Figure 6: Community Internet Intensity Maps for the June 3, 1998 M3.7 earthquake near Mt. Palomar, California. The Web page version of the map is shown in Figure 3. A total of 161 responses were received, even though this small event occurred late at night in a fairly rural area.
Figure 7: Instrumental intensity ``ShakeMap'' for the June 3 1998, M3.7 earthquake for comparison with the CII shown in Figure 6. Fill corresponds to the intensity scale in the legend at the bottom of the figure. The epicenter is shown with a filled star; lines depict highways. Small circles show selected city locations as labeled. Also given in the scale bar are corresponding peak ground motion values, one- or two-word damage and perceived shaking descriptors.
Figure 8: Community Internet Intensity Maps for the August 20, 1998 M4.7 event near Wrightwood, California. A total of 120 responses were received.
Figure: Instrumental intensity ``ShakeMap'' for the August 20, 1998, M4.4 earthquake for comparison with the CIIM shown in Figure 8 [Wald et al., 1999]. Shading corresponds to the intensity scale in the legend at the bottom of the figure. The epicenter is shown with a filled star; lines depict highways. Small circles show selected city locations as labeled. Also given in the scale bar are corresponding peak ground motion values, one- or two-word damage and perceived shaking descriptors.