Basic data such as political boundaries, roads or river networks are important to delineate study areas, to identify features of interest or to create meaningful maps. Datasets of national and sub-national boundaries as well as many other types of features are available online globally. They are perfect for crating broad-scale overview maps (at continental, national or also provincial levels). For studies at more detailed levels, however, those data available from national or provincial agencies are sometimes more useful due to their potentially higher accuracy.
This website provides free access to administrative boundaries at the national, regional and local level. The data are available as shapefile or geodatabse and are downloadable for the entire world or separately for each country.
The following list contains the globally available DEMs. It has to be considered that all of them are rather suitable for regional-scale studies or overviews than for detailed studies at the local scale.
For detailed studies, DEMs of higher resolutions are necessary. Such can be generated by stereo matching of high-resolution satellite or aerial imagery, by the digitization of contour lines of topographic maps, or by employing laser scanning techniques. For some areas, laser scans or other high-resolution DEMs are available from national or regional agencies.
When working with digital terrain data, one has to distinguish between Digital Elevation Models (DEMs), which are sometimes also called Digital Terrain Models (DTMs), and Digital Surface Models (DSMs). Whilst DEMs represent the land surface without buildings or forest, DSM include these features. This distinction is particularly important when working with high-resolution datasets, e.g. Laser Scans. Whilst DSMs are very useful e.g for urban planning purposes, DEMs are required for most geological applications.
The GTOPO30 DEM is freely available worldwide at a resolution of 30 arc seconds, equivalent to approx. 1 km at the equator. It is suitable for maps or investigations at a very broad scale (national to continental).
The Shuttle Radar Topographic Mission (SRTM), carried out in 2000, resulted in a DEM for the entire world between 60 degree North and South. it is freely available at a resolution of 1 arc second (30 m at the equator) for the United States and at 3 arc seconds (90 m at the equator) for the rest of the world. A disadvantage are data gaps in mountain areas due to topographic shading.
Free download of SRTM with gaps filled (SRTM V4) via CGIAR-CSIFree download of original data tiles via Earth Explorer
The ASTER Global Digital Elevation Model, version 2 (ASTER GDEM V2) was derived by stereo-matching of a huge number of ASTER scenes. It is available for free and represents a good - or even better - alternative to the SRTM DEM. Hovewer, the finer cell size (approx. 30 x 30 m) is to some extent offset by a larger degree of noise.
A comprehensive introduction to the principles of remote sensing would be out of scope of the present course. However, some basics shall be summarized. Remote sensing is a methodology to gain information on the Earth's surface (including land, sea and atmosphere) from a distant location which is usually (but not necessarily) the air or space, usually using electromagnetic radiation. The sensor recording the information is usually fixed to an aircraft or a satellite, but it can also be attached to a helicopter or fixed on a mountain top. In principle, each ordinary photo is a remotely sensed dataset, in this case the camera (or, strictly spoken, its CCD) would be the sensor. A fundamental distinction has to be made between active and passive sensors:
Active sensors emit radiation by themselves. Part of this radiation is reflected by the Earth surface (or the atmosphere). Measuring the intensity or travel time of the reflected radiation, information about the properties of the reflecting surface or the distance of the reflecting surface to the sensor can be derived. The SRTM DEM (see above) was derived this way. The most common applications of active sensors are radar remote sensing (SAR) and laser scanning (LiDAR). The advantage of active sensors is their independence from natural radiation, meaning that the sensors can operate independently from daytime and cloud cover. The analysis of data from active sensors, however, is less straightforward than analysis of data from passive sensors.
Passive sensors just record the incoming radiation, which is usually the reflected solar radiation. Therefore, passive sensors cannot operate during the night or with cloud cover (except if the clouds are the object of interest). Most passive sensors record radiation in the visible and infrared spectrum. They usually have one channel including a broad range of the spectrum (e.g the entire visible part) which is referred to as panchromatic. In addition, they have a number of channels with a more narrow spectral range, e.g. blue, green or red light or near infrared. These channels are called multispectral. View the channels of the Landsat sensors. All sensors listed below are passive ones.
One of the major characteristics of remotely sensed data is their resolution, which determines its applicability for different purposes. Four types of resolution are distinguished:
Spatial resolution: size of one image pixel, determining the level of detail recognizable in the image. Pixel sizes of different sensors reach from <1 m to >1000 m.
Temporal resolution: revisit time or frequency at which a certain place of the Earth's surface is recorded by the sensor. The coarser the spatial resolution, the finer is usually the temporal resolution which may range from few hours to several weeks.
Spectral resolution: ability to distinguish between different wavelengths of the spectrum. Hyperspectral sensors can distinguish more than 200 spectral ranges whilst most earth observation satellites have between 4 and 14 channels which is sufficient for many purposes.
Radiometric resolution: the ability to detect differences in the recorded signal, i.e. the number of possible values an image pixel can assume. In often-used 8-bit images, 256 values can be distinguished (0 - 255).
The following list contains an (incomplete) selection of the most-used earth-observation satellites and sensors in the visible and infrared spectrum. Other data requiring more advanced processing techniques, e.g. radar imagery, are not included.
All the data listed below has a global or at least continental coverage. For detailed studies, however, it may also be useful to search for alternatives to satellite imagery, e.g aerial imagery which is usually available from national or regional agencies.
The Landsat program, run by the NASA and the NOAA, consists of several satellites from Landsat 1 to Landsat 8. The first satellite was launched in the early 1970s. All satellites provide medium-resolution imgery from the Earth's land surface in the visible and infrared spectra. The spatial resolution varies between 15 and 120 m, depending on the sensor and the spectral channel. The size of one scene is approx. 170 x 183 km. In principle, data for each spot of the earth is available for each 16 days, but the quality is often limited by cloud cover. The latest sensor ETM+ (Enhanced Thematic Mapper+) carried by Landsat 7 provides only reduced quality imagery since 2003 due to hardware failure.
EO-1 can be considered kind of continuation of the Landsat program. The satellite was launched by the NASA in 2000 and carries two sensors. The Advanced Land Imager (ALI) records nine channels in the visible and infrared spectrum, at a spatial resolution of 10 m panchromatic and 30 m multispectral. Hyperion is a hyperspectral sensor with more than 200 spectral channels. EO-1, with smaller scene size, does not yet provide the comprehensive worldwide cover of the Landsat program, but this may change with increasing time of operation.
ASTER (Advanced Spaceborne Thermal Emission and reflection Radiometer) is a sensor of the Terra satellite, which was launched by the NASA in 1999. It records data for 14 channels in the visible and infrared spectrum with a spatial resolution of 15 to 90 m, depending on the channel. Scene size is approx. 60 x 60 km. Coverage is quite comprehensive. The sensor was built in Japan.
MODIS (Moderate-Resolution Imaging Spectroradiometer) is also carried by the Terra satellite. It records data for 36 spectral channels, with a spatial resolution of 250 to 1000 m. Due to its relatively low spatial resolution, the temporal resolution is high (1-2 days). MODIS is used for the detection of changes in the lower atmosphere (cloud cover), the land surface (biological processes, snow cover) and the sea surface (sea ice) on a broad spatial scale and at various time scales.
SPOT (Satellite Pour l'Observation de la Terre) is a series of French earth observation satellites run by Airbus Defence & Space and operating in the visible and near infrared spectrum. There are five satellites, SPOT-1 to SPOT-7, with a spatial resolution of 1.5 m to 20 m, depending on the sensor and the spectral channel.
Ikonos, launched in 1994, was the first satellite sensor providing publicly available imagery at a resolution of 1 m (panchromatic) and 4 m (multispectral). It is run by the United States company DigitalGlobe. The GeoEye-1 satellite, with 0.4 m and 1.6 m resolution, respectively, was launche din 2008 and is operated by the same company.
Instructions how to purchase can be found or requested on the website of DigitalGlobe
QuickBird was launched in 2001. Like Ikonos and GeoEye-1, it is a high-resolution earth observation sensor with 0.6 m panchomatic and 2.4 m multispectral spatial resolution. It is run by the United States company DigitalGlobe. WorldView-1 was launched in 2007 and is run by the same company. It provides panchromatic data at 0.5 m spatial resolution. WorldView-2, with 0.5 m panchromatic and 1.8 m multispectral resolution, was launched in 2009. WorldView-3 was launched in 2014 (0.31 m panchromatic and 1.24 m multispectral resolution).
Instructions how to purchase can be found or requested on the website of DigitalGlobe
The Corona program was a series of strategic reconnaisence satellites run by the CIA and the US Air Force. It recorded imagery of the USSR and the People's Republic of China between 1959 and 1972. The imagery was declassified and made publicly available in 1996 for an affordable price. Due to the effective resolution of the scanned imagery of up to approx. 5 m, it provides a valuable reference for studies of environmental changes, for example in the mountains of central Asia.