IMPORTING USGS 3-ARC SECOND DTED INTO AN ARC LATTICE

This module will describe the relatively simple task of importing 3-arc second Digital Terrain Elevation Data (DTED) into an ARC Lattice.

The U.S. Geological Survey has developed a program to make available through the information super highway, digital data of various forms. One such data set is the entire collection of 3-arc second DTED data covering the 50 states. This data is available through the EROS data center. Click here if you want to go to the site containing this data.

These files are named according to the standard USGS 1:250,000 scale quadrangles. To find the file that you want, just find the quadrangle name on any 1:250k quadrangle index. The DTED data covers a 1 degree of longitude by 1 degree of latitude area and therefore one file covers 1/2 of a 1:250k quad. Each half quad is appended with a "-e" or a "-w" to identify east or west half.

This is an example for the Salt Lake City 1:250k quadrangle:

salt_lake_city-e.gz

- East half of the Salt Lake City Quad.

salt_lake_city-w.gz

- West half of the Salt Lake City Quad.

The ".gz" extension means that this file is compressed.

We downloaded the salt_lake_city-e.gz and salt_lake_city-w.gz files using the Mosaic interface. We could also have downloaded it by using standard FTP.

To uncompress these files we used the "gunzip" utility:

gunzip salt_lake_city-e.gz

This made the file salt_lake_city-e.

According to the EROS Data Center people, these files do NOT contain record delimeters. Once we download and decompressed the files we added delimeters using the following UNIX command:

dd if=salt_lake_city-e of=salt_lake_city-edd ibs=4096 cbs=1024 conv=unblock

dd is the command

if is the input file

of is the output file

ibs is the input block size in bytes

cbs is the output block size in bytes

conv=unblock Converts fixed length ASCII records to new-line terminated records.

The standard record size for USGS DTED data is 1024. However, these files are not delimeted at every 1024 bytes. If you don't add the delimeters at every 1024 bytes, the DEMLATTICE command in ARC will fail.

In ARC we used the DEMLATTICE command to convert the ASCII DTED data called salt_lake_city-edd to a lattice. BE CAREFUL, ARC will not recognize grid files with long names. To get around this we had to come up with a shorter name for the output grid. The following is an example of how to use DEMLATTICE:

Arc: demlattice

Usage: DEMLATTICE [in_dem] [out_lattice] {USGS | TAME} {z_factor}

Arc:

Arc: DEMLATTICE salt_lake_city-edd SLC-E

In this example we changed the output file name to SLC-E to make it shorter.

If the conversion worked, we should have a grid file in ARC format.

ARC imports these files as floating point values. That is, each elevation value is stored as a real number and not as an integer. Since the data are originally stored as integers, having them as real numbers is not necessary and takes up valuable space.

To convert all values to integers and therefore reduce the storage requirements and processing speed, we used the GRID comand called INT. This is how to use INT.

Grid: outputgrid = INT(inputgrid)

so

Grid: SLC-E-int = INT(SLC-E)

Once out of Grid, we killed the original SLC-E grid and copied the new SLC-E-int file to SLC-E. SLC-E-int was killed.

This is what SLC-E looks like when displayed in GRID: