Introduction

This project examines spatial interpolation accuracy and 3D visibility across the UCLA campus. Four interpolation methods were evaluated to assess how effectively they estimate elevation in an urban environment. A 3D campus model was created, using the Bruin Bear statue as a reference observation point for line-of-sight and visibility analysis.

Methodology

A UCLA campus buildings shapefile was used as the primary dataset. Building elevation points were divided into training and testing sets, with one building excluded from the training data to evaluate prediction accuracy. Four spatial interpolation methods—Inverse Distance Weighting (IDW), Ordinary Kriging, Regularized Spline, and Tension Spline—were used to generate digital elevation models. Predicted elevation values at test locations were extracted and compared against known values to calculate prediction error and RMSE. A 3D model of the campus was then used to visualize line-of-sight patterns from the Bruin Bear statue.

Results 

Results indicate notable differences in interpolation performance across methods. IDW and Ordinary Kriging produced the most stable and accurate elevation estimates, with comparatively low RMSE values. In contrast, spline-based methods—particularly tension splines—introduced substantial error and exaggerated surface variation. These findings demonstrate how interpolation choice can significantly influence terrain representation and downstream analyses such as visibility modeling. The 3D visualization highlights how elevation uncertainty directly affects line-of-sight outcomes in dense urban environments.

Introduction

This project evaluates cellular coverage across Los Angeles County using GIS-based viewshed and coverage modeling. Due to complex terrain and urban density, cellular signal distribution is uneven across the county. Multiple infrastructure scenarios were tested to determine which modifications most effectively improve county-wide coverage, including adding new towers, increasing tower height, and expanding signal range.

Methodology

Spatial data included the Los Angeles County boundary, a digital elevation model (DEM), and existing cell tower locations. The DEM was clipped to the county boundary and used to generate viewshed rasters representing signal visibility. Towers were modeled as observers, with baseline parameters set for tower height, human receiver height, and signal radius. Three modification scenarios were tested: adding new towers in topographically strategic locations, increasing tower height, and expanding signal range. Coverage rasters were clipped to the county boundary and pixel counts were extracted to quantify total and percentage coverage.

Results

Baseline coverage accounted for approximately 56% of Los Angeles County. Adding three new towers increased coverage to roughly 62%, representing a notable improvement in localized service areas. Expanding signal range resulted in a similar increase in overall coverage but produced greater redundancy in already-served regions. Increasing tower height produced the largest overall improvement, increasing county-wide coverage by approximately 12%, and emerged as the most effective strategy.