Cybershake NZ 18.6 run details:
Srf generation:
1- Use Leonard (2010) relation for active shallow crustal (double-check it)
2- Use stoch sub-fault size of 2 km
3- Use srf sub-fault size of 0.1 km for ALL FAULTS
4- Set the number of rupture realizations for a given source based on Magnitude
5- Create hypocenter in a randomized fashion in down-dip, along-strike directions, and use a single slip realization per hypocenter
6- Remove (i.e., combine) the fault segments in Stirling ERF that are smaller than 1 km in length.
VM generation:
1- PGV threshold of 2 cm/s (previously 5cm/s)
Vs30 model:
1- Use updated Vs30 model (v18.4)
2- Compare the old and new Vs30 models
Runs:
1- Make sure dt is set to 0.02s
3- LF calculation is at 400m grid with min Vs=500m/s
4- Use ‘200 meter’ non-uniform grid for storing the outputs
Hazard calculations:
1- Generate csv file with faults and magnitudes for hazard calculation.
2- run hazard calculation on Kupe
3- Empirical hazard calculations independent of openSHA
Item | 17p8 | 18p6 | Comment |
Magnitude scalingrelashionship | Leonard (stable continental region relation) | Leonard (Shallow crustal) | The difference is 0.2 Mw. Using stable continental region relation was a legacy from the previous codes. |
Stoch file | 2km | 2km |
|
| If fault_length < 50 Sub_fault = 0.1 elseif 50<fault_length < 100 Sub_fault = 0.2 else Sub_fault = 0.5 end | 0.1 km for everything |
|
Number of simulations | Place hypocenter every 20 km, with minimum of 1 hypo for fault_length < 20 | Using the ‘model’ we have decided based on magnitude |
|
Hypocenter perturbation | Stratified – every 20 km along the strike | Randomized – Monte Carlo Normal along strike; Weibull along dip |
|
Slip realizations | Three per each hypocenter | One per each realization |
|
|
|
|
|
Vs30 comparison for 400 meter non-uniform grid: old (left) versus new (right) | |
Comparison for the 400 (left) and 200(right) new Vs30 outputs | |
Hypocenter realizations | Mw scaling relationship |
Estimated core-hours for a single simulation of considered sources |
|
Estimated core-hours for all simulations
rup_name hypo slip nx ny nz dt sim_duration core_hours core_hours_total
AhuririR 1 1 681 882 113 0.02 141.76 14.44 14.44
AlpineF2K 1 1 732 2213 138 0.02 359.1 120.45 120.45
HopeConwayOS 1 1 743 1226 120 0.02 192.92 31.64 31.64
Wairau 2 1 742 1352 123 0.02 211.54 39.17 39.17
Sample simulation domain:
| AlpineF2K | HopeConwayOS | Wairau | AhuririR |
PGV = 5cm /s threshold |
PGV = 2 cm /s threshold |