Let's break down the steps to **develop the seismic hazard curve using PSHA** (Probabilistic Seismic Hazard Analysis) for the site at (0,0) using the information provided. ## **Step 1: Fault Discretization** Each fault (source) should be divided into at least 20 increments (segments). ### **Fault Source 1:** - Endpoints: (-40, 70) to (-10, -30) - Divide into 20 segments, obtain segment coordinates \((x_i, y_i)\) using linear interpolation. ### **Fault Source 2:** - Endpoints: (-12, 80) to (60, 55) - Divide into 20 segments, obtain segment coordinates \((x_j, y_j)\). --- ## **Step 2: Distance Calculation** For each segment, calculate the distance to the site (0,0): \[ R = \sqrt{(x_{\text{segment}} - 0)^2 + (y_{\text{segment}} - 0)^2} \] --- ## **Step 3: Magnitude Discretization** - Magnitude range: 3.5 to 8.0 (discretize, e.g., in steps of 0.1 or 0.2). --- ## **Step 4: Ground Motion Prediction (Cornell et al. 1979)** The **attenuation relationship** (Cornell et al., 1979) for Peak Ground Acceleration (PGA): \[ \log(\text{PGA}) = a + bM - \log(R) - cR \] Where: - \( M \) = magnitude - \( R \) = distance (km) - \( a, b, c \) = empirical coefficients (typically: \( a = -1.024 \), \( b = 0.249 \), \( c = 0.00255 \) for PGA in g, but **confirm with your reference**) - Standard deviation of \(\log(\text{PGA}) = 0.2\) --- ## **Step 5: Probability Calculation** For each magnitude and segment: - Calculate mean \(\log(\text{PGA})\). - For a given PGA threshold \( x \), probability that shaking exceeds \( x \) is: \[ P(\log(\text{PGA}) > \log(x)) = 1 - \Phi \left( \frac{ \log(x) - \mu }{ \sigma } \right) \] where \( \mu \) is the mean from the attenuation equation, \( \sigma = 0.2 \), and \( \Phi \) is the standard normal CDF. --- ## **Step 6: Rate of Exceedance** For each fault segment and magnitude increment, sum the contributions: \[ \lambda_{\text{exceed}} = \sum_{\text{segments}} \sum_{\text{magnitudes}} P(\text{exceed}) \times \text{annual rate} \] Assume Gutenberg-Richter recurrence for magnitude distribution (if not provided, use standard values). --- ## **Step 7: Hazard Curve** - For a range of PGA values, plot **annual probability of exceedance** (\(\lambda_{\text{exceed}}\)) vs. **PGA**. --- ### **Summary Table** | Step | Task | Formula/Action | |------|-----------------------------------------------------------------------|--------------------------------------------------------------------| | 1 | Discretize each fault (20 segments) | Linear interpolation of endpoints | | 2 | Calculate distance from site to each segment | \( R = \sqrt{(x - 0)^2 + (y - 0)^2} \) | | 3 | Discretize magnitude range (e.g., 3.5 to 8.0 in 0.1 increments) | - | | 4 | Compute mean PGA from attenuation relation | \( \log(\text{PGA}) = a + bM - \log(R) - cR \) | | 5 | Compute exceedance probability for each PGA threshold | \( P(\text{exceed}) = 1 - \Phi \left( \frac{ \log(x) - \mu }{ \sigma } \right) \) | | 6 | Sum over segments and magnitudes (with occurrence rate) | \( \lambda_{\text{exceed}} \) | | 7 | Plot hazard curve | \( \lambda_{\text{exceed}} \) vs. PGA | --- ## **What you need to provide:** - The Gutenberg-Richter parameters (b-value, annual rate of events, etc.) for each fault, **unless you assume standard values**. - The specific coefficients for the Cornell et al. (1979) attenuation relationship if different from above. --- ### **Example Calculation (for one segment and one magnitude):** Suppose segment at (-10, -30): \[ R = \sqrt{(-10)^2 + (-30)^2} = \sqrt{100 + 900} = \sqrt{1000} \approx 31.6 \text{ km} \] For \( M = 7 \): \[ \log(\text{PGA}) = -1.024 + 0.249 \times 7 - \log(31.6) - 0.00255 \times 31.6 \] \[ = -1.024 + 1.743 - 1.5 - 0.08 = -0.861 \] If you want \( \text{PGA} > 0.1g \): \[ \log(0.1) = -1 \] \[ z = \frac{-1 - (-0.861)}{0.2} = \frac{-0.139}{0.2} = -0.695 \] \[ P(\text{exceed}) = 1 - \Phi(-0.695) \approx 1 - 0.244 = 0.756 \] --- If you provide the recurrence parameters, I can help you set up the full calculation or Python/Excel code for hazard curve generation!