Lab #7: Spatial Interpolation

My spatial interpolation maps of rainfall throughout Los Angeles county are quite revealing and fascinating. The current rainfall pattern indicates a trend of even and continuous distribution of

rainfall throughout LA county with a few exceptions. The maps showing differences between the current seasonal total and normal seasonal rainfall illustrate the discrepancies between the two effectively. These difference maps show a large discrepancy in rainfall totals within the Angeles National Forest region. This hot spot shows a significant shortage of rainfall where most other areas are only slightly below or above normal seasonal totals. These numbers can be understood within the scope of our current seasonal rainfall sample.

By understanding the current seasonal rainfall totals, we are able to determine important contexts behind the map. As the Department of Public Works Real-Time Precipitation map indicates, "Accumulated Precipitation in Inches Beginning on 10-01-2009 and Ending on 03-04-2010", the current seasonal rainfall totals are created based on 5 months of the most significant rainfall for the Los Angeles county area. This sample date window illustrates why the difference maps show such a minor difference between the current and normal rainfall patterns. Though the exception within the center of the two difference maps is quite interesting and deserves more investigation.

I chose my two spatial interpolation techniques through considering the ideas expressed in Prof. Shin's lecture as well as within the Interpolating Surfaces reading from the website. The reading indicated the IDW technique as being useful in interpolating sample points based on their distance values. I found that this technique would be valuable in mapping rainfall data considering the nature of rainfall intensity being tied to proximity in terms of regionally similar trends. I also chose the Regularized Spline method based on the ability to generate "exact interpolation" by passing directly through the rainfall value points. I believed that passing through the rainfall accumulation values exactly would minimize the significant loss within the fine margin of variation between current and normal rainfall totals. I also felt the Splines ability to create accurate surfaces for continuous temperature phenomena would lend itself well, and ultimately more accurately than the IDW method, to the also fairly continuous phenomena of rainfall.