IV. The Study Data

The data cover portions of three separate 3 arc second data sets. 30m datasets were selected for each 3 arc second coverage based on several criteria. First of all, I wanted to make comparisons within USGS production levels as well as across them. That is, since the available 30m data is produced using a variety of methods, any comparative work has to be done with the production method in mind. Secondly, I tried to pick DEMs with varying ranges of elevation. High relief datasets are those with a wide margin between minimum and maximum elevation, while low relief datasets are characterized by a relatively short range of values. Finally, some attention was paid to the spatial distribution of the 30m DEMs. I wished to avoid clumps of neighboring datasets; in this I was not always successful due to data availability factors as well as conflicts in meeting my other criteria. Table 4.1 presents all datasets employed in this work, and Figure 4.1 shows the spatial distribution of these DEMs.
Quadrangle Name LevelRelief
Los Angeles-w
Santa Barbara 1high (0-1209 m.)
Santa Barbara 2high (0-1217 m.)
Salisbury Potrero 1high (782-1713 m.)
Cuyama Peak1 high (796-1979 m.)
Oxnard1 low (0-35 m.)
Camarillo2 low (0-414 m.)
Bakersfield-w
Fellows1 high (245-1114 m.)
Elkhorn Hills 2high (453-1112 m.)
N. of Oildale 2low (146-372 m.)
Buena Vista Dry Lake 1low (71-273 m.)
Death Valley-e
Devils Speedway 2high (-83-1525 m.)
Furnace2 high (-82-1277 m.)
High Peak2 high (707-1341 m.)
Ashton2 low (761-1101 m.)

Table 4.1. Study area data sets with production and landscape characteristics.

Figure 3.1. Map of the study region. Los Angeles-w is the quadrangle in the lower left, Bakersfield-w is just above it, and Death Valley-e is in the upper right. 30m quads are displayed above the elevation data.

Los Angeles-w is the one degree quadrangle bounded by the 119th and 120th meridians on east and west and by the 35th and 34th parallels on north and south. It is characterized by high relief of the coastal ranges, punctuated by low-lying coastal plains. Elevation ranges from 0 to 2682 meters. The 30 meter datasets in this quadrangle have some interesting relationships which made them notable for study. The Santa Barbara DEM available from the Map and Imagery Library at UCSB (as late as November 1996) is categorized as level 1 by the USGS. Level 1 is the specification which denotes those DEMs not produced from digital line graph (DLG) hypsography and hydrography files. A previous section indicated the range of production techniques which qualify as level 1. This one was created in 1976, but I was unable to determine the exact method. I suspect it was using GPM-II, since a vast majority of level 1 DEMs were produced using that technology. A more recently produced Santa Barbara DEM is available from the archives at Cal State Northridge (http://geogdata.csun.edu). This DEM is level 2 and was released in 1994. A USGS level 2 DEM is generated from DLGs. Having DEMs produced using different methods for the same location offered some additional research directions. Both the Salisbury Potrero and the Cuyama Peak DEMs are level 1 and produced in 1977 from the same source as the Santa Barbara level 1 DEM. They are located in mountainous terrain north of Santa Barbara. The Oxnard DEM shares its heritage with the other level 1 DEMs in this area. Its extremely flat terrain made for a helpful contrast with the other datasets. Camarillo, which borders the Oxnard quadrangle, is a level 2 DEM generated in 1994 from DLGs originally produced in 1967.

Bakersfield-w is bounded by the 119th and 120th meridians on east and west and by the 36th and 35th parallels on north and south. Much of this region is very flat - the southern San Joaquin Valley dominates the area. Unfortunately, a large part of the valley is not covered by the 30m DEM data set, so most of the available 30m data is concentrated in the southwest along the Temblor Range. Fellows and Elkhorn Hills are the higher relief datasets I chose from this region. Fellows was probably generated with the PhotoMapper, while Elkhorn Hills was produced in 1995 from contour lines. North of Oildale is a level 2 DEM produced in 1993. Buena Vista Dry Lake Bed is level 1.

Death Valley-e is bounded by the 116th and 117th meridians on east and west and by the 37th and 36th parallels on north and south. The terrain includes sections of Death Valley itself along with rugged, high relief terrain to the east. I chose it partly to see how the USGS dealt with negative elevations, and also to check on the ability of my code to handle them. Three of the four 30m DEMs in this area were produced in 1994 or 1995 from DLGs using identical technology. The fourth, Ashton (I wonder why I picked that one?) uses an alternative method, but is also level 2.