University of Washington | Tacoma
Three-Dimensional Campus Map
A three-dimensional map of the University of Washington - Tacoma (UWT) campus buildings. These include both academic buildings as well as those leased to Tacoma businesses.
Starting Up
The only layers used for this project were a UWT buildings polygon layer, LiDAR data of the campus area, and an orthophoto of downtown Tacoma. The LiDAR data included both a layer of bare earth returns as well as a layer of all returns.
Figure 1: Original UWT buildings layer on top of the orthophoto.
Obtaining Building Heights
Two methods were used to obtain UWT building heights. The first step for each of these involved interpolating both sets of LiDAR data using the inverse distance weighted (IDW) method, resulting in "bare earth" and "all returns" grids of the UWT campus (figures 2,3).
​
For the raster values method, new fields were then added to the buildings layer attribute table for raster values of both bare earth and all returns and point height, as well as a mean height field for the zonal statistics method.
Figure 2: Bare earth returns interpolated surface
Figure 3: All returns interpolated surface
Raster Values Method
In order to calculate building heights using raster values, the buildings polygon layer was converted to points, followed by using "extract value to points" for both all returns and bare earth, interpolating the values at the point locations. These extracted point value layers were then joined one at a time to the original buildings polygon layer. The field calculator was used to subtract the bare earth values from the all returns values, resulting in building heights.
Zonal Statistics Method
Tables were created using the "zonal statistics as table" tool for both the all returns grid and the bare earth grid, and the buildings polygon layer as the feature zone data. The 'buildings' field was used as the zone field. The resulting tables were subsequently joined to the buildings polygon layer. All statistics were removed except the means, which were then used to calculate building heights using the field calculator to subtract the mean bare earth from the mean all returns.
Tables were created using the "zonal statistics as table" tool for both the all returns grid and the bare earth grid, and the buildings polygon layer as the feature zone data. The 'buildings' field was used as the zone field. The resulting tables were subsequently joined to the buildings polygon layer. All statistics were removed except the means, which were then used to calculate building heights using the field calculator to subtract the mean bare earth from the mean all returns.
Figure 4: New fields of raster values, point heights, zonal statistics means, and mean height in buildings layer attribute table.
Three-Dimensional Work
A session was performed for ground-truthing the heights of the buildings. The buildings layer was adjusted in editing sessions to reflect heights from the down-slope side of the buildings as well as to add new campus building polygons that were not represented on the original buildings layer.
ArcScene was used to model the building heights in 3D on top of the downtown Tacoma orthophoto. The orthophoto, buildings polygon layer, and the bare earth grid were added to the data frame. All layers were set to 'floating on a custom surface' (the bare earth grid) with cell resolution size of 5x5. The buildings layer was then extruded using the mean heights from the zonal statistics method and applied by adding it to the feature's minimum height. The buildings were then symbolized as unique values.
Figure 5: Buildings layer with adjusted heights and added building polygons.
Figure 6: 3D view in ArcScene of extruded buildings.