defining a source logic tree
SourceLogicTree configuration format¶
For use by SourceLogicTree.from_user_config()
The config for a SourceLogicTree uses the json file format. It approximates the dict representation of the SourceLogicTree dataclass with some simplifications to make it easer to create one manually.
Most of the entries are optional, but without including at least some of them, you will not have a useful SourceLogicTree. Comments (//) have been included in the example here, but are not compatible with an actual json file.
The branch set short_name and branch branch_id are used to define correlations. Therefore, if correlations are specified each branch must have a unique branch_set.short_name:branch.branch_id combination. Otherwise, branch set short_name and branch branch_id are optional, as is branch_set long_name.
{
"title": "A sample SRM logic tree definition",
"version": "10.1",
"branch_sets": [
{
"short_name": "PUY",
"long_name": "Puysegur",
"branches": [
{
// A branch branch_id is only necessary if correlations are specified. If so, the branch_id of
// every branch (even between branch_sets must be unique).
"branch_id": "PUY1",
// the weights of the branches in a branch_set must sum to 1.0
"weight": 0.2,
// Rupture rate scaling (if specified) can be used to scale the rates of all
// sources in the branch up or down in a hazard calculation.
"rupture_rate_scaling": 1.1,
// The values are optional, but can be used to build the structure of the logic
// tree. Each entry in the values list is a node and the specific branch on
// that node defined by the value of the parameter in question.
"values": [
{
"name": "dm",
"long_name": "deformation model",
"value": "0.7"
},
{
"name": "bN",
"long_name": "bN pair",
"value": [
0.902,
4.6
]
}
],
// A branch can have multiple sets of earthquake sources.
// Each source (rupture definitions and rates) is identified by a nrml_id.
// This can be any unique identifier string. Branches have two types,
// InversionSource and DistributedSource. Sources are by default Distributed
// unless any one or more parameter from "inversion_id", "rupture_set_id",
// "inversion_solution_type" is included in which case it will be an
// InversionSource. The Source type is not used here, but it can be useful in
// understanding how the source was generated.
"sources": [
{
"nrml_id": "ABC"
},
{
"nrml_id": "XYZ"
},
{
"nrml_id": "SW52ZXJzaW9uU29sdXRpb25Ocm1sOjEyOTE2NjE=",
"rupture_rate_scaling": null,
"inversion_id": "U2NhbGVkSW52ZXJzaW9uU29sdXRpb246MTI5MTM4Mg==",
"rupture_set_id": "RmlsZToxMjkwOTg0",
"inversion_solution_type": ""
},
{
"nrml_id": "RmlsZToxMzA3NTQ=",
"rupture_rate_scaling": null
}
]
},
{
"branch_id": "PUY2",
"weight": 0.8,
"sources": [
{
"nrml_id": "DEF"
}
]
}
]
},
{
"short_name": "HIK",
"long_name": "Hikurangi",
"branches": [
{
"branch_id": "HIK1",
"weight": 0.3,
"sources": [
{
"nrml_id": "GHI"
}
]
},
{
"branch_id": "HIK2",
"weight": 0.2,
"sources": [
{
"nrml_id": "JKL"
}
]
},
{
"branch_id": "HIK3",
"weight": 0.25,
"sources": [
{
"nrml_id": "MNO"
}
]
},
{
"branch_id": "HIK4",
"weight": 0.25,
"sources": [
{
"nrml_id": "PQR"
}
]
}
]
}
],
// Correlations are used to make branches always be paired with others when forming all
// combinations of the branch_set branches. Branches are identified by branch branch_id and branch
// set short_name. The first entry identifies the "primary" branch. The primary branch will only
// be paired with the other branches in it's list (e.g. "HIK1" will only be combined with "PUY1"
// not "PUY2"). However, it is possible for the non-primary branches to be combined with more
// than one branch from the primary list, as shown below. The correlations are not limited to
// two branches, you can have as many entries as there are branch sets.
"correlations": [
["HIK:HIK1", "PUY:PUY1"],
["HIK:HIK2", "PUY:PUY2"],
["HIK:HIK3", "PUY:PUY1"],
["HIK:HIK4", "PUY:PUY2"]
]
}