Troubleshooting High PGA Bias in Classical PSHA: Validating Smoothed Seismicity & ZMAP Declustering

[Aslam]

Dear OpenQuake Community,

I am currently working on a Classical PSHA involving multiple sources: fault sources, background sources, subduction interface, and Benioff zone sources.

I developed the background seismicity for a depth of 0–50 km using the following workflow:

1. Declustering: I performed declustering in ZMAP by dividing the catalog into two depth clusters (0–25 km and 25–50 km) and then combining them.

2. Grid & FMD: I generated a 0.1-degree grid (approx. 11 km area) and calculated the Frequency-Magnitude Distribution (FMD) for each grid cell. I derived the a-values using the available HMTK smoothed seismicity tool (hmtk_smoothed.py) and the b-values using the regional Weichert method (Smoothed_grid.py).

3. Source Generation: I created center points for each cell to represent the 0.1-degree grids using the HMTK source generator (Source_model_generator.py). I then inputted simple parameters based on typical fault geometry probabilities.

4. The Issue: The resulting Peak Ground Acceleration (PGA) shows significant bias. Despite using the exact same Ground Motion Prediction Equations (GMPEs) and overall parameters, the PGA values seem off.

My questions for the community are:

• Question 1: Why is there a significant discrepancy between the smoothed rate and the observed rate? The two source codes I used only differ in how they handle the a-value, yet the outputs vary greatly.

• Question 2: Could I have missed or incorrectly configured a parameter in the OpenQuake code (source_model_Shallow)? Alternatively, is it possible that my ZMAP declustering process inadvertently removed major, critical earthquakes?

I would greatly appreciate any insights, discussion, or guidance. I have been going in circles trying to troubleshoot this and validate my OpenQuake results to ensure they align closely with EZ-FRISK or USGS PSHA analyses.

I have attached my codes, the output files, and the reference outputs I am using for validation.

Thank you so much in advance for your time and expertise!

Best regards, Aslam