Earthquake Hazard GIS Modelling in the Greater Vancouver Regional District

Error & Problems

I didn't encounter any major problems in the processes of this modelling. However there are some problems that impact the result of my analysis:

• The degree of uncertainty of the description values in the orginal soil.rdc and bedrock.rdc. Since there were no metadata provided from the orginal source, I have to search for information that are closed to the descriptions on the internet in order to make my reclassification. For example, the descriptions in LITHOLOGY are classified based on values found QUATERNARY GEOLOGY OF SOUTHEASTERN VANCOUVER KLANlOl AND GULF ISLANDS by Ministry of Energy and Mines. Note that the orginal LITHOLOGY data were from the GVRD department that were available in the SIS Lab as IDRISI soil.mdb and soil.rdc, so the reclassification might be inaccurate and add errors to the final result. Also, for the Soil Texture description DESCRIPTIN, most values are cut off, and I've to reclassify based on the available values, this will also add errors to my final results.

• The degree of assumptions I made for the entire model. This includes:
  1. My weight values for the MCE analysis. No information can be found on which geotechnical hazards (soil amplification, soil liquefaction, slope instability, and earthquake intensity) have greater effect on earthquake. Also, no information on which slope factors (bedrock instability, aspect, degree) have greater effect on slope instability. Therefore, the Weight Values I assigned for the MCE analysis of slope stability and relative earthquake hazard mapping were done by guessing. This totally add errors on my final result maps.
  2. My assumption on earthquake intensity based on closeness to fault lines. The activity of the GVRD fault lines are not known and there is no way that one can determined where an earthquake will locate before the earthquake occurs.
  3. The assumption of landslides occuring only in non-developed areas and in areas with > 5m. This affects my slope instability map which in turn affects my final relative earthquake hazard map. Since the non-developed areas are derived from landuse.rdc which have no values on the Northerneast corner, using this non-developed areas as constraint in my slope instability MCE analysis resulted in visual flatness in the entire Northern East corner. This area should be removed so that it doesn't mislead someone visually. However, due to time constraints of this project, this is not feasible.

• The generalized data (bedrock.shp was very generlized) that I used and the generalization that I made in defining each Geotechnical hazard factor. For example, for amplification, I didn't encounter the softness of soil texture because this information was not available in my dataset.

• Missing soil data in the Northern parts of GVRD (including some Northern parts of North Vancouver, West Vancouver, and Maple Ridge) and in Vancouver Area. As a result, you can see strange square patterns in the northern part and very low values in Vancouver in my final relative earthquake hazard map. This also add errors into my map.

In summary, my entire approach for mapping the relative earthquake hazard needs further research for verification.