gns-science / opensha-ucerf3

UCERF3 code and data

opensha-ucerf3

UCERF3 code and data

opensha-ucerf3's Issues

pull and validate with upstream changes in opensha/opensha-ucerf3

We want latest changes from upstream, and these are expected to break a few tests in nshm-nz-opensha.

• create a tag here for downstream builds to pin to (?? check this is right)
• refactor any local tests if needed
• merge to master

Produce some Coulomb-based NZ crustal ruptures for comparison

As SRM team

we want to see some coulomb based ruptures built from CFM 3.0

so that we can compare the coulomb plausibility approach with the current, UCERF3 approach.

Done when:

• we have a NZ rupture set built using defaults as per ClusterRuptureBuilder.main()
• we have identified the main Coulomb parameters to tune/test on.

Prepare RupSetDiagnosticsPageGen for reuse NSHM reporting and promotion

This file is huge and unwieldy, so care is needed and we should promote these changes ASAP.

In nshm-ns-opensha there are 2 reports awaiting this change, see GNS-Science/nshm-nz-opensha#109

Done when

• figure out the minimum changes required - (NO FORMATTING!)
• Add integration tests here, showing our use pattern in https://github.com/GNS-Science/nshm-nz-opensha/
• PR is merged upstream

Upstream: raise 1 PR + 1 Issue

• my @chrisbc changes on `src/org/opensha/sha/earthquake/faultSysSolution/ruptures/strategies/ClusterConnectionStrategy.java' need a PR back to upstream otherwise this will hit us again.
• still the ERF test is failing. This needs an issue on upstream.

Originally posted by @chrisbc in #8 (comment)

removing state from MagScalingRelationship (in org.opensha.commons.calc.magScalingRelations)

picked up in PR #24

Involving two repos

opensha-commons
in org.opensha.commons.calc.magScalingRelations.MagScalingRelationship there's a public method .setRake()
and in org.opensha.commons.calc.magScalingRelations.magScalingRelImpl.ScalingRelationships we have the new class TMG2017CruMagAreaRel

opensha-ucerf3
in scratch.UCERF3.enumTreeBranches.ScalingRelationships we have this....

TMG_SUB_2017("Thingbaijam et al.(2017) Subduction", "TMG_SUB_2017") {
		// units of the input dimensions are in m or m^2 
		public double getAveSlip(double area, double length, double origWidth) {
			tmg_sub_magArea.setRake(90.0d); // interface

which seems to break thread-safety, since TMG2017SubMagAreaRel should not have state.

A couple of suggestions...

make the setRake method protected or even private.
create a constructor that passes in rake.

but we need to agree this with Kevin, since it's an API change.

scratch.UCERF3.griddedSeismicity.FaultPolyMgr should support any region

NSHM.NZ need to calculate fault section polygons for the NZ region instead of California.

PlausibilityResult logicalOr() method

PlausibilityResult should provide a logicalOr() method to complement the existing logicalAnd()

This is needed to build an investigative plausablity filter for GNS-Science/nshm-nz-opensha#28.