Data Voids Filling Occasionally, satellite remote sensing techniques are unable to obtain sufficient textural detail to determine elevation information. The consequences are regions with no data, or data voids. Such voids can become an impediment especially for research requiring continuous data. The original SRTM data contained a total of 3,436,585 voids accounting for about 0.2% of the available data. CGIAR-CSI corrected this problem with a void-filling algorithm which established a void-free DEM, SRTM V3. After this input DEM was interpolated to 1-arcsecond, it was used to fill the voids of the US SRTM DEM.

Data voids in a region are filled in by using an alternative DEM that fully covers the region.

Interpolation

DEM interpolation takes available pixel values and produces new pixel values by comparing each pixel to its surrounding pixels. The result is a higher resolution DEM covering the same region as before, but with more elevation values. Interpolation was essential in producing the consistent 1-arcsecond global DEM presented on this website. The different input data sets used had resolutions ranging from 30 arc-seconds to 1 arc-second, which can be inconvenient in research requiring a consistent resolution. The interpolated DEMs can be easily compared and down sampled to a lower resolution.

Value Replacement The different input data sets represent water regions with different numerical values. GTOPO30, for example, uses the value -9999 to represent water bodies while SRTM uses -32768. For the sakes of consistency, all water regions were replaced with the number 0. This standard also facilitates the viewing of the DEMs in certain programs. Data Shifting The input Alaska DEM is referenced to the NAD27 horizontal datum. As a result, the orginal Alaska DEM is shifted compared to SRTM. In order to provide a smooth transition between different DEM data sets, the output Alaska DEM has been shifted to correlate with SRTM.

All DEMs are broken up into 1x1 degree tiles.

Formatting The different input data sets each have their own format and naming conventions. The output global DEM on this website uses the Geotiff format as its standard. The Geotiff format conveniently embeds header information, which simplifies the use of the DEM. The download interface also allows for the conversion to the binary format which includes a separate header file. Data Division and Naming The output global DEM is conveniently divided into 1x1 degrees tiles with a consistent naming convention corresponding to the south west corner of the DEM.

<< Introduction

Analysis and Assesment >>