ASTER-Derived 30-Meter-Resolution Digital Elevation Models of Afghanistan

Metadata:

Identification_Information:

Originator: U.S. Geological Survey

Publication_Date: November 2007

Title: ASTER DEMs of Afghanistan

Geospatial_Data_Presentation_Form: remote-sensing image

Series_Name: USGS Data Series 318

Publication_Information:

Publication_Place: Reston

Publisher: U.S. Geological Survey

Online_Linkage:

Abstract:

The Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) is an imaging instrument aboard the Terra satellite, launched on December 19, 1999, as part of the National Aeronautics and Space Administration's (NASA) Earth Observing System (EOS). The ASTER sensor consists of three subsystems: the visible and near infrared (VNIR), the shortwave infrared (SWIR), and the thermal infrared (TIR), each with a different spatial resolution (VNIR, 15 meters; SWIR, 30 meters, TIR 90 meters). The VNIR system has the capability to generate along-track stereo images that can be used to create digital elevation models (DEMs) at 30-meter resolution.

 

Currently, the only available DEM dataset for Afghanistan is the 90-meter-resolution Shuttle Radar Topography Mission (SRTM) data. This dataset is appropriate for macroscale DEM analysis and mapping. However, ASTER provides a low cost opportunity to generate higher resolution data. For this publication, study areas were identified around populated areas and areas where higher resolution elevation data was desired to assist in natural resource assessments. The higher resolution fidelity of these DEMs can also be used for other terrain analysis including landform classification and geologic structure analysis.

 

For this publication, ASTER scenes were and processed and mosaicked to generate 36 DEMs which were created and extracted using PCI Geomatics’ OrthoEngine 3D Stereo software. The ASTER images were geographically registered to Landsat data with at least 15 accurate and well distributed ground control points with a root mean square error (RMSE) of less that one pixel (15 meters). An elevation value was then assigned to each ground control point by extracting the elevation from the 90-meter SRTM data.

 

The 36 derived DEMs demonstrate that the software correlated on nearly flat surfaces and smooth slopes accurately. Larger errors occur in cloudy and snow-covered areas, lakes, areas with steep slopes, and southeastern-facing slopes. In these areas, holes, large pits, and spikes were generated by the software during the correlation process and the automatic interpolation method. To eliminate these problems, overlapping DEMs were generated and filtered using a progressive morphologic filter.

 

The quadrangles used to delineate the DEMs in the publication were derived from the Afghan Geodesy and Cartography Head Office's (AGCHO) 1:100,000-scale maps series quadrangles. Each DEM was clipped and assigned a name according to the associated AGCHO quadrangle name.

 

The geospatial data included in this publication are intended to be used with any GIS software packages including, but not limited to, ESRI's ArcGIS and ERDAS IMAGINE.

 

Purpose:

To create high-resolution DEMs, generated from the ASTER sensor on the Terra satellite, to be used for natural resource assessments of Afghanistan.

Calendar_Date: unknown

Currentness_Reference: publication date

Progress: Complete

Maintenance_and_Update_Frequency: As needed

West_Bounding_Coordinate: 66.489108

East_Bounding_Coordinate: 67.008797

North_Bounding_Coordinate: 37.009443

South_Bounding_Coordinate: 36.657142

Theme_Keyword: ASTER

Theme_Keyword: digital elevation model

Theme_Keyword: digital terrain model

Theme_Keyword: DEM

Theme_Keyword: DTM

Place_Keyword: Afghanistan

Access_Constraints: None

Use_Constraints: None

Contact_Person: Peter G. Chirico

Contact_Organization: U.S. Geological Survey

Contact_Position: Geographer

Address_Type: mailing and physical address

Address: National Center, MS 926A

Address: 12201 Sunrise Valley Drive

City: Reston

State_or_Province: Virginia

Postal_Code: 20192

Country: USA

Contact_Electronic_Mail_Address: pchirico@usgs.gov

Native_Data_Set_Environment: Version 6.0 (Build 6000) ; ESRI ArcCatalog 9.2.2.1350

Data_Quality_Information:

Positional_Accuracy:

Horizontal_Positional_Accuracy:

Horizontal_Positional_Accuracy_Report:

All DEMs generated were referenced to orthorectified, 15-meter-resolution Landsat imagery with a root mean square error (RMSE) value of ±1 pixel (15 m).

Quantitative_Horizontal_Positional_Accuracy_Assessment:

Horizontal_Positional_Accuracy_Explanation:

Previous studies indicate that the vertical accuracy of DEMs generated from ASTER have a root-mean-square error (RMSE) in elevation between ±7 m and ±15 m (Hirano and others, 2003). However, in more mountainous terrain, research suggests that the RMSE values increase to ±15 m to ±20 m in hilly terrain and about ±30 m in mountainous terrain (Eckert and others, 2005).

 

Eckert, S., Kellenberger T., and Itten K., 2005, Accuracy assessment of automatically derived digital elevation models from aster data in mountainous terrain: International Journal of Remote Sensing, v.26, p.1943-1957.

 

Hirano, A., Welch R., and Lang H., 2003, Mapping from ASTER stereo image data:  DEM validation and accuracy assessment:  ISPRS Journal of Photogrammetry and Remote Sensing, v.57, p.356-370.

 

Process_Description:

To generate DEMs from ASTER, a single-level 1A or 1B scene is required. The ASTER nadir (3N) and backward-looking (3B) satellite images were imported into PCI Geomatics’ OrthoEngine 3D Stereo software module (version 9.1.7) for the sensor geometry modeling, stereo-matching, and elevation extraction stages of DEM development.

Ground control points (GCPs) were collected directly from 15-m, pan sharpened, orthorectified Landsat imagery, and each GCP was then assigned an elevation value by extracting the elevation from an acquired 30-m-resolution Shuttle Radar Topography Mission (SRTM) DEM. Two hundred automatic and manually collected tie points (TPs) were also generated to tie the 3N and 3B images together. After GCP and TP collection, the images were reprojected into an epipolar projection, which reduces the amount of registration error between images and improves the pixel matching process. The software algorithm then calculates an elevation based on the sensor geometry, GCPs and measured parallax, or difference of each pixel set. Following the epipolar pair generation, the DEM was extracted at 30-m-resolution.

Like most DEMs produced using the stereo-autocorrelation process, ASTER DEMs contain artifacts, or pits and spikes, in the extracted DEM values. These errors are introduced due to pixel mismatching, missing data, and cloud masking. A progressive morphological filter was developed to remove erroneous elevation values and improve the quality of the DEM. The filter was applied using an ArcMacro Language (AML) script in ArcGIS (version 9.0). The filter iteratively compares individual raw elevation values to a set of focal neighborhood statistics and a user-defined threshold value. Elevation differences between the raw value and the neighborhood statistics are compared to the threshold value. Raw values that exceed the threshold are replaced with a focal minimum, focal maximum, or focal median value based on the characteristics of the elevation value in question. For a 30-m-resolution DEM, a vertical threshold value of 30-m was specified for the filtering process. The filter progresses through four stages whereby elevation values are compared to increasingly smaller neighborhoods and a progressively reduced threshold value. The result is that only elevation values that exceed the defined parameters are replaced; all other values remain unchanged and the overall output quality is improved without degrading the high-resolution fidelity of the DEM.

Source_Used_Citation_Abbreviation:

Process_Step:

Process_Description: Metadata imported.

Source_Used_Citation_Abbreviation:

Spatial_Data_Organization_Information:

Direct_Spatial_Reference_Method: Raster

Raster_Object_Information:

Raster_Object_Type: Pixel

Row_Count: 1268

Column_Count: 1516

Vertical_Count: 1

Spatial_Reference_Information:

Horizontal_Coordinate_System_Definition:

Planar:

Grid_Coordinate_System:

Grid_Coordinate_System_Name: Universal Transverse Mercator

Universal_Transverse_Mercator:

UTM_Zone_Number: 42

Transverse_Mercator:

Scale_Factor_at_Central_Meridian: 0.999600

Longitude_of_Central_Meridian: 69.000000

Latitude_of_Projection_Origin: 0.000000

False_Easting: 500000.000000

False_Northing: 0.000000

Planar_Coordinate_Information:

Planar_Coordinate_Encoding_Method: row and column

Coordinate_Representation:

Abscissa_Resolution: 30.000000

Ordinate_Resolution: 30.000000

Planar_Distance_Units: meters

Geodetic_Model:

Horizontal_Datum_Name: D_WGS_1984

Ellipsoid_Name: WGS_1984

Semi-major_Axis: 6378137.000000

Denominator_of_Flattening_Ratio: 298.257224

Distribution_Information:

Resource_Description: Downloadable Data

Standard_Order_Process:

Digital_Form:

Digital_Transfer_Information:

Metadata_Reference_Information:

Metadata_Date: 20070820

Metadata_Contact:

Contact_Information:

Contact_Organization_Primary:

Contact_Organization: U.S. Geological Survey

Contact_Person: Peter G. Chirico

Contact_Position: Geographer

Contact_Address:

Address_Type: mailing and physical address

Address: National Center, MS 926A

Address: 12201 Sunrise Valley Drive

City: Reston

State_or_Province: Virginia

Postal_Code: 20192

Country: USA

Contact_Voice_Telephone:

Contact_Electronic_Mail_Address: pchirico@usgs.gov

Metadata_Standard_Name: FGDC Content Standards for Digital Geospatial Metadata

Metadata_Standard_Version: FGDC-STD-001-1998

Metadata_Extensions:

Online_Linkage: <http://www.esri.com/metadata/esriprof80.html>

Profile_Name: ESRI Metadata Profile

Generated by mp version 2.8.6 on Mon Aug 20 09:21:13 2007