Spatial Analysis - UCLA Satellite Campus

Updated Rainfall in Los Angeles County Interpolation

Lab 3

Lab 2

Lab 1

Lab #7 - Interpolation

Interpolation is a powerful procedure that can be used in ArcGIS to predict the value of cells at locations that do not have sample points. It is based on the idea of spatial autocorrelation which is premised upon Tobler's first law of geography: "Everything is related to everything else, but near things are more related than distant things..." Of the different interpolation methods, I chose to use inverse distance weighted (IDW) and kriging to create surface area maps of precipitation in Los Angeles county. The IDW method of interpolation estimates cell values by averaging the values of sample data points in the area of each processing cell. The closer the point is to the center of the cell being estimated, the more influence, or weight, it has in the averaging process. Kriging is similar to IDW in that it weights the surrounding measured values to derive a prediction for an unmeasured location. I used ordinary kriging in my maps, which assumes that the variation in z-values is free of any structural component (drift).

My first map using IDW shows that the most rainfall is in the central region of the map, while there is less precipitation in the northern and southern parts of Los Angeles County. Overall, this season's precipitation is not too different from normal precipitation patterns. I think that IDW is an appropriate interpolation technique for this data set because IDW requires a set of points that is dense enough to capture the extent of local surface variation. Since there are over sixty automatic rain gage points, I feel that there was enough to produce a quality surface map for analysis of LA County's precipitation.

The second map I created using kriging showed overall less variance in rainfall from this season compared to the season normal. It also emphasizes in both the normal season and this current season-to-date that the most rainfall is in the central part of the county, while the least is in the north (northeast especially) and in the south. Compared to the IDW map, the kriging maps does not show as much of the local variation in precipitation, causing me to favor the IDW procedure for interpolation since I find it yields more detailed results. Ultimately, while it is predicted to be an El Nino year (which means higher levels of precipitation), these maps show that for the 2009-2010 season there does not seem to be significantly higher levels of rainfall--in Los Angeles County at least.

Source: Los Angeles County Department of Public Works and ESRI

Lab 6 - Fire Hazard Mapping

In 2009, the Los Angeles County Station Fire ignited in the Angeles National Forest on August 26 and burned until October 16. The total area burned covered 160,577 acres and 209 structures were destroyed, including 89 homes.

To do my own analysis of the Station Fire and surrounding area, I first went to the USGS National Map Seamless Server and downloaded a DEM for the region. Next I went to the U.S. Forest Service website to download vegetation cover data. Lastly, from the Los Angeles County Enterprise GIS website I downloaded a shapefile of the Station Fire's perimeter. Next I loaded all of this data into ArcMap to begin my spatial analysis.

I first used the Surface Analysis tool to create a raster showing hillshade from the DEM. I then made a map of the Station Fire area's elevation and hillshade, also adding in a layer of major roads and highways to help the viewer orient themselves. I had wanted to create a layer showing the slope in percent as I did in the Modeling the Wildland/Urban Interface exercise, however, when I tried the values I got just did not seem right (they were very high), which is why I decided not to include this layer in my final map.

For the second part of my map, I took the vegetation cover shapefile of California and used the editor to clip the shapefile to only show my focus area. Then I converted this feature to a raster, classified by land cover type. To make my third map, I used the table from the Modeling the Wildland/Urban Interface exercise showing the NFPA classes for fuel types to reclassify my land cover raster, creating a new fuel grid. Finally, I used the raster calculator to add the fuel risk to the elevation, making my fourth map of the risk for fire in the area--as the value increases, so does the risk.