Efficient testing of earthquake forecasting models

Rhoades, David A.; Schorlemmer, Danijel; Gerstenberger, Matthew C.; Christophersen, Annemarie; Zechar, J. D.; Imoto, Masajiro

doi:10.2478/s11600-011-0013-5

Cited by 111 publications

(140 citation statements)

References 22 publications

Supporting

Mentioning

140

Contrasting

Order By: Relevance

“…Forecasts are tested against observed seismicity using a set of likelihood-based consistency tests Zechar et al 2010a) and their relative performance to each other is tested to determine the best performing model using a set of comparative tests (Rhoades et al 2011). In contrast, previous earthquake forecasting experiments merely compared one forecast model against a null hypothesis, as described by .…”

Section: Csep Testing Centre Frameworkmentioning

confidence: 99%

“…The information gain score is a test metric that indicates the relative performance of two forecasts at an observed earthquake location (Rhoades et al 2011). Applying the Student's paired t-test, we consider the null hypothesis that the two forecasts perform equally well, and the alternate hypothesis that one forecast can be rejected in favour of the other.…”

Section: T-test and W-testmentioning

confidence: 99%

“…This tests the median information gain per earthquake rather than the mean, only requiring that the information gain distribution is symmetric. The W-test becomes less powerful as the number of observed earthquakes decreases and is incapable of rejecting a forecast with fewer than five events at the 0.05 significance level (Rhoades et al 2011). Therefore, the W-test result should be 244 A.…”

Section: T-test and W-testmentioning

confidence: 99%

“…The CSEP testing centre was built upon prospective evaluation of seismicity forecasts against observed seismicity. Within the CSEP framework, described in the 'CSEP testing centre framework' section, the RELM working group established a set of consistency tests to evaluate each forecast in the RELM experiment against observed seismicity during a testing period Zechar et al 2010a), as well as a set of comparative tests to evaluate the performance of one forecast over another (Rhoades et al 2011). The testing suite provides an unbiased approach to learn about seismicity models' strengths and weaknesses, indicating dimensions (e.g.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Prospective and retrospective evaluation of five-year earthquake forecast models for California

Strader¹,

Schneider

Schorlemmer

2017

Geophysical Journal International

Self Cite

View full text Add to dashboard Cite

Section: Csep Testing Centre Frameworkmentioning

confidence: 99%

Section: T-test and W-testmentioning

confidence: 99%

Section: T-test and W-testmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Prospective and retrospective evaluation of five-year earthquake forecast models for California

Strader¹,

Schneider

Schorlemmer

2017

Geophysical Journal International

Self Cite

View full text Add to dashboard Cite

“…To estimate the significance of the differences in probability gains, we could use the T-test or the W-test (Rhoades et al, 2011). However, the empirical distribution of x points to issues in the interpretation of the results.…”

Section: Comparison Of the Predicted Number Of Events Per Daymentioning

confidence: 99%

Adaptive Smoothing of Seismicity in Time, Space, and Magnitude for Time-Dependent Earthquake Forecasts for California

Helmstetter

Werner

2014

Bulletin of the Seismological Society of America

View full text Add to dashboard Cite

We present new methods for short-term earthquake forecasting that employ space, time, and magnitude kernels to smooth seismicity. These methods are purely statistical and rely on very few assumptions about seismicity. In particular, we do not use Omori-Utsu law, and only one of our two new models assumes a Gutenberg-Richter law to model the magnitude distribution; the second model estimates the magnitude distribution nonparametrically with kernels. We employ adaptive kernels of variable bandwidths to estimate seismicity in space, time, and magnitude bins. To project rates over short time scales into the future, we simply assume persistence, that is, a constant rate over short time windows. The resulting forecasts from the two new kernel models are compared with those of the epidemic-type aftershock sequence (ETAS) model generated by . Although our new methods are simpler and require fewer parameters than ETAS, the obtained probability gains are surprisingly close. Nonetheless, ETAS performs significantly better in most comparisons, and the kernel model with a Gutenberg-Richter law attains larger gains than the kernel model that nonparametrically estimates the magnitude distribution. Finally, we show that combining ETAS and kernel model forecasts, by simply averaging the expected rate in each bin, can provide greater predictive skill than ETAS or the kernel models can achieve individually.

show abstract

Comparative evaluation of physics‐based and statistical forecasts in Northern California

2013

View full text Add to dashboard Cite

[1] We perform a retrospective forecast test using Northern California seismicity for the period between 1980 and 2009. We compare 7 realizations of the short-term clustering epidemic-type aftershock sequence (ETAS) model, and 21 models combining Coulomb stress change calculations and Rate/State theory (CRS) to forecast seismicity rates in 10 day time intervals. We employ a common learning phase (1974)(1975)(1976)(1977)(1978)(1979)(1980) for CRS models to ensure consistency, and we evaluate the forecasts with log likelihood statistics to detect any spatial inconsistencies and compare the total numbers of forecasts versus observed events. We find that: (1) ETAS models are better forecasters of the spatial evolution in seismicity in the near-source region, (2) CRS models can compete with ETAS models away from the mainshock rupture, and for short periods after mainshocks, (3) CRS models with optimally oriented receiver fault planes perform better in the first few days after mainshocks, whereas mapped fault planes should be implemented for longer-term forecasting, and (4) CRS models based on shear stress change calculations have comparable performance with Coulomb stress change models, with the benefit of lesser parameters involved in stress calculations. We conclude that physics-based and statistical forecast models are complimentary to each other and that future forecasts should be based on statistical models for near-source regions, and physical models for longer periods and distances. However, the realization of the CRS models involves a number of critical parameters (reference seismicity rates, regional stress field, and loading rates), which should be retrospectively tested to improve the predictive power of physics-based models.

show abstract

Efficient testing of earthquake forecasting models

Cited by 111 publications

References 22 publications

Prospective and retrospective evaluation of five-year earthquake forecast models for California

Prospective and retrospective evaluation of five-year earthquake forecast models for California

Adaptive Smoothing of Seismicity in Time, Space, and Magnitude for Time-Dependent Earthquake Forecasts for California

Comparative evaluation of physics‐based and statistical forecasts in Northern California

Contact Info

Product

Resources

About