TY  - JOUR
A1  - Shebalin, Peter N.
A1  - Narteau, Clement
A1  - Zechar, Jeremy Douglas
A1  - Holschneider, Matthias
T1  - Combining earthquake forecasts using differential probability gains
JF  - Earth, planets and space
N2  - We describe an iterative method to combine seismicity forecasts. With this method, we produce the next generation of a starting forecast by incorporating predictive skill from one or more input forecasts. For a single iteration, we use the differential probability gain of an input forecast relative to the starting forecast. At each point in space and time, the rate in the next-generation forecast is the product of the starting rate and the local differential probability gain. The main advantage of this method is that it can produce high forecast rates using all types of numerical forecast models, even those that are not rate-based. Naturally, a limitation of this method is that the input forecast must have some information not already contained in the starting forecast. We illustrate this method using the Every Earthquake a Precursor According to Scale (EEPAS) and Early Aftershocks Statistics (EAST) models, which are currently being evaluated at the US testing center of the Collaboratory for the Study of Earthquake Predictability. During a testing period from July 2009 to December 2011 (with 19 target earthquakes), the combined model we produce has better predictive performance - in terms of Molchan diagrams and likelihood - than the starting model (EEPAS) and the input model (EAST). Many of the target earthquakes occur in regions where the combined model has high forecast rates. Most importantly, the rates in these regions are substantially higher than if we had simply averaged the models.
KW  - Probabilistic forecasting
KW  - Earthquake interaction
KW  - Forecasting and prediction
KW  - Statistical seismology
Y1  - 2014
U6  - https://doi.org/10.1186/1880-5981-66-37
SN  - 1880-5981
VL  - 66
PB  - Springer
CY  - Heidelberg
ER  - 
TY  - JOUR
A1  - Shebalin, Peter
A1  - Narteau, Clement
A1  - Holschneider, Matthias
T1  - From alarm-based to rate-based earthquake forecast models
JF  - Bulletin of the Seismological Society of America
N2  - We propose a conversion method from alarm-based to rate-based earthquake forecast models. A differential probability gain g(alarm)(ref) is the absolute value of the local slope of the Molchan trajectory that evaluates the performance of the alarm-based model with respect to the chosen reference model. We consider that this differential probability gain is constant over time. Its value at each point of the testing region depends only on the alarm function value. The rate-based model is the product of the event rate of the reference model at this point multiplied by the corresponding differential probability gain. Thus, we increase or decrease the initial rates of the reference model according to the additional amount of information contained in the alarm-based model. Here, we apply this method to the Early Aftershock STatistics (EAST) model, an alarm-based model in which early aftershocks are used to identify space-time regions with a higher level of stress and, consequently, a higher seismogenic potential. The resulting rate-based model shows similar performance to the original alarm-based model for all ranges of earthquake magnitude in both retrospective and prospective tests. This conversion method offers the opportunity to perform all the standard evaluation tests of the earthquake testing centers on alarm-based models. In addition, we infer that it can also be used to consecutively combine independent forecast models and, with small modifications, seismic hazard maps with short- and medium-term forecasts.
Y1  - 2012
U6  - https://doi.org/10.1785/0120110126
SN  - 0037-1106
VL  - 102
IS  - 1
SP  - 64
EP  - 72
PB  - Seismological Society of America
CY  - Albany
ER  -