Scientific Improvements
| Functionality | To be included in this version | Might be included | Not this version | What we actually did? |
|---|---|---|---|---|
| VM v2.0.2+ | Aim: All basins that are included in the latest NZVM (as published in Thomson et al 2019) for which validation workflow has shown produce enhanced results over neglect of basins (targets incl: Cant, Cheviot, Waiau, Hanmer, Kaikoura, Nelson, Blehneim, Marlborough, Wellington) | Basins that are currently undergoing consideration for inclusion (e.g. 'enhanced' versions of current basins; new basins such as Hamilton, Hauraki, Napier etc) | Cant, Cheviot, Waiau, Hanmer, Kaikoura, Nelson, Blehneim, Marlborough, Wellington, basins included | |
Inclusion of Subduction-Interface ruptures:
| Aim: Need to run sufficient scenario simulations with the functionality in order to have confidence that we are running correctly. (pseudo-validation via 50 realizations of a specific fault geometry; then examine statistics of the outputs in terms of macro-parameters). | Not in the part 1 runs of this version (which will run in May and comprise only shallow crustal), but should be appended to complete this version | No Subduction included | |
| SRFgen improvements for shallow crustal earthquakes | Validation pending. Will need to be pseudo-validation from the perspective that we don't have validation for finite fault SRF ATM | Still using same SRF generation | ||
| 200m simulation realisations | Not in the initial version but will be done in a supplementary version (based on disagg outputs) | Did not end up computing | ||
| Sim Duration | Reducing the simulation duration by 66%-75% (75% is Karim's recommendation). Also need to consider the depth of the simulation (currently X and Y dimensions are strongly optimized, whereas depth and sim duration are poorly optimized). | Sim duration calculation updated. Average 70% of 18p6 | ||
| Additional Realisation Parameters (e.g. Magnitude) | Not to consider in this version (will wait for validation with uncertainties to get a better constraint on this. | |||
| New Non-uniform Grid | Update in Christchurch / Wellington Region Finer Grid Real station update | No changes | ||
| vs30 improvements | Update to Kevin's 18p12 model | Recalculated the vs30 for all sites based on Kevin's 18p12 model | ||
| BB sim methodology improvements | From Robin's Validation already identified changes to HF path duration and LF site amp (i.e. no LF site amp). Apply these advancements. | gmsim_v18.5.3 - see the wiki page for more information | ||
| Depth of VM increase | Equation is taking dbottom rather than hdepth now. | Equation is taking dbottom rather than hdepth now. VMs 33% deeper |
Core Hour Estimations
LF
Original estimate: 133082
Last run: 132544
New estimate with sim-duration change: 104173
New estimate with sim-duration and z change: 112709
Subduction To-Do
Check SRFgen 5.4.2
Magnitude Scaling
Run quick animation for all faults
Core Hour Usage:
LF: