Conveners: Ramon Arrowsmith and Kate Scharer
Dates: September 8-9, 2012
Location: Hilton Palm Springs Resort, Palm Springs, CA
SCEC Award and Report: 12105

Overview: The 2012 Southern San Andreas Fault Evaluation (SoSAFE) Fieldshop, a combined field activity and workshop, focused on field interpretation and measurement of small (< 50 m) offsets of geomorphic features along faults. Our interest in this issue stems in large part from the proliferation of undated, small geomorphic offsets now used in efforts such as the Uniform California Earthquake Rupture Forecast (UCERF3) to calculate paleoearthquake magnitude and paleoearthquake rupture extent and estimate future timing or slip along the faults given standard recurrence models. The Fieldshop explored the reproducibility of field measurements of geomorphic offsets by multiple investigators and generated a rich discussion on the uncertainties and qualification of these data.

We designed the Fieldshop to mesh with an existing collaboration (between Arrowsmith, Rockwell, and Salisbury) on the repeatability of remote measurements of offset using high-resolution (~50 cm per pixel) topographic digital elevation models (DEMs) typically produced from airborne LiDAR measurements available along high slip rate faults in the western US. For the field activity, the participants travelled to the Mojave section of the San Andreas Fault and worked on two areas southeast of Palmdale, CA. The 27 field participants mapped on two DEMs, one derived from the B4 LiDAR dataset of a 500 m section of the fault near Pearblossom and the other derived from a TLS scan of a 100 m section near Littlerock CA provided by a UC Davis team. The field activity handout and base maps are included as Appendix 2 of this report. Participants were randomly assigned to groups of three people, given base maps and tape measures, and distributed along the fault with instructions to discuss and map the geomorphology of the fault zone and measure geomorphic offsets (see Field Activity Handout). We did not instruct groups how (or what) they should measure, expecting that this would generate discussion and new ideas in the groups. Participants had a mix of experience but the majority was familiar with the goals and typical procedures for measuring offsets. These included graduate students (8) measuring offsets for their own research, early career scientists (6), and mid-career scientists (8). The remaining participants (5) were less experienced in this type of mapping (seismologists, teachers, or undergraduate students). As we were particularly interested in the qualitative and methodological aspects of the field effort, Arrowsmith, Scharer and five graduate students experienced in these types of measurements (Bohon, Haddad, Marliyani, Salisbury, Sato) stationed along the fault took notes on the discussion topics and approaches voiced by the groups as they worked on specific offsets.

The following is a brief summary of the key challenges recorded by the observers on Saturday and further discussed in the workshop on Sunday, organized by theme into Mapping, Measuring, and Quality Ranking. The report ends with consensus highlights and recommendations from the Fieldshop, which were available for review by all participants. Complete results of the Fieldshop will be integrated in a paper by Salisbury et al. (in prep) and a short summary of these issues will be submitted to Seismological Research Letters (Scharer et al., in prep).

Mapping: The field location was picked for its accessibility and concentration of displaced features with a range of offset magnitudes, but it also presented the real challenge of deciphering the contribution of multiple fault traces to offset measurements. Within groups not all participants agreed on the presence of small subsidiary fault strands, or how the strands contributed to cumulative offset (how can you determine if each strand was active during each earthquake?). The most common subject after questions about the location of the fault(s) was the interpretation of the offset itself. Specifically, groups voiced uncertainty about the genesis of individual offsets (how can you determine if a jog was produced by an earthquake rather than geomorphic processes?). This concern was heightened in areas where the fault was multi- stranded, leading to discussion about the significance of each potential (but small) jog through a wide fault zone. Participants also had lengthy discussions about bends in the channels near the fault and how these bends should be handled when reporting offsets (how can you determine what part of the bend predated the last offset?).

Measuring: All groups debated which feature(s) was appropriate to measure; some chose the thalweg (channel centerline), others chose channel walls or margins, and a few people provided measurements for multiple geomorphic components at a single offset. The question of which element was best to measure often depended on the site situation; such as how to project the upstream and downstream features to the fault (to the degree that it had been defined), how to handle sinuous channels, and how much degradation of features affected the measurements. Groups typically provided either (a) a thalweg measurement with channel margin half-width providing the uncertainty or (b) a range (minimum and maximum) with additional error uncertainty based on the shape of the landform. In discussion, participants debated the value of measuring offsets of multiple components of a single feature (to provide more quantitative estimate of uncertainties) as opposed to restricting the measurement to the best- preserved feature (with the goal of obtaining the most accurate estimate, rather than swamping the results with poorly resolved features).

Larger offsets (e.g., >10 m) posed additional problems, causing groups to debate the significance of small inset geomorphic surfaces (were they remnant channel bottoms?), short, very straight reaches of the channel along the fault compared to reaches where the stream bent toward the fault (did the straight section represent the most recent offset?), and degree of erosion (how wide was the channel when it was first offset?). A common conclusion was that the reconstruction itself (epistemic, a model determined by the geomorphic mapping) effectively determined the uncertainty in the offset (aleatory, a quantified measurement).

The group discussed the meaning of the measurement uncertainties, and emphasized that they do not provide a true standard deviation (by definition, this would require many measurements of the same feature) but rather bounds on the amount a feature could have been offset (physical or plausible limits on the offset itself). Many agreed that given uncertainties in feature reconstruction, most offsets should be recorded as a minimum and maximum allowable offset. The range permitted for an offset feature implies that there is an appropriate probability distribution function (pdf) for the offset; outside that range the geologist finds no correlation of features. We discussed the forms of pdfs used in existing studies (uniform/boxcar, triangle, and truncated Gaussians) and generally agreed that unless the best estimate was clear, a uniform pdf was appropriate. Integration of measurements along strike was outside the scope of our workshop, although several participants recommended papers in preparation and in press treat this issue (Arrowsmith and Zielke, in prep; Gold et al., in review; Madugo, et al., in prep; Oskin et al., in prep; Rockwell and Klinger, in review).

Quality ranking: Participants were asked to provide two in-field assessments of offset quality. The first was simply a subjective rank from 1-5 of the overall quality of the measurement. The second was a bivariate that established (low, medium, or high) the definition of the fault zone and the obliquity of the offset feature with respect to the fault. Initial tabulation of the field results show that some sites had fairly uniform quality ranking across groups while other sites received a wide quality ranking (from low to high quality, depending on the group). Groups had differing opinions on whether or not to measure low quality or complex features (this may have been heightened by limited time for the field activity). In discussion, a wide range of factors was recognized to affect quality, including vegetation, competence of the substrate, local geomorphic terrain, number of fault strands, and distinctiveness or geomorphic activity of the offset feature. Participants disagreed whether or not such factors should be individually ranked or lumped into a single quality ranking. A particular challenge is to separate those aspects of the quality of the offset and its reconstruction that are accounted for in the measurement uncertainty (ideally the aleatory aspects) from those that control the epistemic interpretation of the offset. While all agreed that careful field notes and documentation of what was measured were fundamental to such work, the use of the quality ranking when such offset measurements are incorporated into larger projects (e.g. UCERF) remained a concern.

Recommendations and Key Findings:
The Fieldshop highlighted several aspects of measurement geomorphic offsets that should be addressed in future studies:

1. Field mapping was integral to understanding the offset features and getting an accurate measurement. In-field sketch maps and annotated field photos provide important documentation and quantitative details that cannot be obtained from remote (computer-based) studies, even if remote studies provide a better means to measure the offset (e.g. larger or more offsets).
2. Offset measurements feed directly into interpretation of fault behavior and hazards, thus the careful consideration of the possible origin of each feature is needed to understand earthquakes. Studies must determine whether or not each offset is the product of geomorphic rather than tectonic activity. This challenge is increased in multi-stranded sections where offsets may be smaller and less defined.
3. Methods for describing and measuring errors associated with bends in offset channels are needed to improve fault displacement measurements. Very few of the modern channels create highly oblique lines across the fault (even after 153 years of post-earthquake offset storms at the field activity location, for example), so methods for recognizing pre-existing bends should be developed, or methods for formally incorporating uncertainties associated with these features should be standard procedure.
4. Measurement errors should not be conflated with the quality rating. While the ideal feature would have small errors and high quality, it is possible to have features with small measurement errors but poor quality because the geologist has doubts about the geomorphic reconstruction. Aspects of the offset that can be measured (width of channel, angle between piercing line and fault zone, width of fault zone, distance between feature and fault, amount a bend contributes to a perceived offset, etc.) should be reported as quantities, whereas interpretive aspects of the reconstruction (confidence that the offset is the product of faulting, pairing of features across the fault, etc.) should be qualified.
5. Most offsets should be recorded as simply a range that includes the measurement uncertainties. The range should capture the extent of the reconstruction (i.e., there is no evidence that the feature can be reconstructed outside of that range). In cases where the terrain permits, it may be appropriate to provide a best estimate, uncertainties, and associated probability distribution function (pdf, e.g. triangle, truncated Gaussian) to convey more or less likely reconstructions within the range. If this is done, authors should provide clear guidance on the reasons for choosing a non-uniform pdf and emphasize that the uncertainties/pdf does not provide a statistical estimate of the variance.
6. Studies should provide guidance on the subsequent use of these data in other projects (e.g., UCERF3). Recommendations including thresholds of data quality that should be used, certainty of the origin of the offset, shape of offset pdf, and relative weighting compared to other offsets in the study would be valuable.

Participants: Austin Elliott (UC Davis), Beth Haddon (Western Washington), Brian Olson (CGS), Colin Amos (Western Washington), David Haddad (ASU), David Lynch (USGS), Frank Sousa (Caltech), Gayatri Marliyani (ASU), Glenn Biasi (UNR), Guangfu Shao (UCSB), James Dolan (USC), James Hollingsworth (USC), Janis Hernandez (CGS), Jason De Cristofara (USGS), Jerry Treiman (CGS), Justin R. Brown (Caltech), Kate Scharer (USGS), Ken Hudnut (USGS), Kim Blisniuk (UC Berkeley), Koji Okumura (Hiroshima Univ), Lisa Grant Ludwig (UCI), Margaret Gooding (LSA Assoc), Mike Oskin (UC Davis), Mitchell Prante (Utah State), Nate Onderdonk (CSULB), Orlando Teran (CICESE), Peter Gold (UT Austin), Ramon Arrowsmith (ASU), Richard Heermance (CSUN), Sally McGill (CSUSB), Scott Kenyon (CSULB), Sean Bemis (U Kentucky), Thomas Rockwell (SDSU), Ting Lin (Stanford), Tracy Compton (UC Davis), Tsurue Sato (ASU), Wendy Bohon (ASU), and Whitney Behr (UT Austin)

Sunday, September 9, 2012

Return to 2012 SCEC Annual Meeting