Urs data conversion

URS Overview

The URS data encapsulating all hazard points, faults, subfaults and events within faults is provided in a set of files and that's how it was used in TsuDAT1. TsuDAT2 puts all that information into a set of database tables, so there needs to be a method to extract URS data from a set of files into a form that can be loaded into database tables.

The GitHub repository contains programs that were used to generate database load files for the initial implementation of TsuDAT2. These programs can be used as a starting point for the proposed Indonesian version of TsuDAT2.

Data sizes

For the Australian dataset the following numbers of entities were found. The [x,y] indicates the numeric index range of the data:

• 22 return periods
• 52 zones (53 if Sumba_normal is included)
• 3851 hazard points, thinned to 883 and keeping original indices [0,3850]
• 1275 subfaults (maximum of 77 in one fault) [0,1274]
• 76170 events [1, 76170]

Note that hazard point and subfault indices start at zero, and those for events start at 1. This is the way the data is, it's not a bug. This difference is, however, a rich source of bugs when using the URS data!

The programs

There is not much to say here beyond the README.

One thing should be noted. The programs used to load data into the first version of TsuDAT2 were changed because the original programs used data derived from the dataset when TsuDAT1 was written. The code that created that derived data can no longer be found. I realised that this derived data would not exist for the proposed Indonesian dataset, so I rewrote two programs here to use the raw URS data, not the derived data. These changes have not been tested. When these programs are used to load the Indonesian data (or reload the Australian data) the data must be checked for correctness. Of course, you would do that anyway, wouldn't you:)