SANSA Earth Observation Services

SANSA Earth Observation is associated with the collection of information on the Earth's surface or from its atmosphere from space-based systems such as instruments aboard satellites.

Data archive and Catalogue

services1New: SPOT 6 & SPOT 7 Imagery now available at SANSA Earth Observation (View)

SANSA has a satellite imagery archive dating back to 1972 from the first Landsat satellite. The archive has grown over the years to over one million scenes including scenes from the following satellites: Landsat 2-5, 7; SPOT 1-5; SumbandilaSat; CBERS-2B; SAC-C; ERS-1 and 2 radar imagery. The majority of the imagery available within the archive includes images that fall within the footprint region of the ground receiving station of Hartebeesthoek. This footprint region covers most of the SADC region.  The chart below shows the breakdown of the SANSA image archive as of January 2013. This archive forms the part of the Earth Observation Data Centre (EODC) of the Department of Science and Technology (DST).
 
services2The majority of the SANSA imagery archive is available for purchase or download via the SANSA Earth Observation Online Catalogue. The search and ordering facility is available to the public, governmenet departments and research institutions to find available imagery over a given area of interest for a specific date or date range. The results of the search may be ordered online and according to the licencing agreements for each sensor, may be purchased or downloaded without a payment fee.

services3
The Data Catalogue can be accessed via the following link: SANSA Earth Observation Catalogue.

Following the link will provide a new catalogue user with a brief guideline on performing an image data search on the SANSA Earth Observation catalogue.

 

Data reseller

ZA2 L2A 100220 Cape Town stadiumNew: SPOT 6 & SPOT 7 Imagery now available at SANSA Earth Observation (View)

As a data reseller, SANSA Earth Observation is able to offer a variety of different sensors such as WorldView, GeoEye and TerrSAR-X. This allows a customer to engage with a knowledgeable representative at the directorate, who can help determine what the optimal imagery would be for their usage. Acting as the data reseller, SANSA's Earth Observation directorate can then place the order with both foreign and local suppliers and deliver the data to the customer through many different channels. The directorate may also assist in the processing of data, provided resources are available to do so. The data reseller is an important link between SANSA Earth Observation and their stakeholders, as the unit is the primary point of contact in terms of data provision, including the roll-out of the annual SPOT-5 mosaic, and the delivery of the Fundisa disk through to the respective research institutes.

Photo: A SumbandilaSat image of the Cape Town stadium

 

Applications development

applicationsA key component in remote sensing is image processing. Raw satellite images need to be presented in a usable and meaningful format before various remote sensing tasks can be performed on the images. Image processing algorithms are usually executed on raw imagery to produce meaningful products. Typical tasks namely image enhancements, atmospheric corrections; pan-sharpening and orthorectification are usually performed on raw imagery. Recently, the amounts of data received from various sensors have increased drastically, hence the need for automated processing chains for the above tasks in order to get the data out timeously. The Research and Applications Development (RAD) team within SANSA's Earth Observation directorate is responsible for developing and improving these automated processing chains in order to speed up and ensure accuracy for the delivered satellite image products.

With the advancement of high performance computing, research is on-going with regards to improving these processing chains. In addition, the RAD team follows up on the delivery of these products by educating potential users on the best processing algorithms for carrying out their specific remote sensing tasks.  While SANSA Earth Observation's RAD team focuses much of their energy on improving and fine tuning these processing chains, they also look into the development of additional value-added products that can be derived from the images archived. The goal of applications development is to make applications and methods available for end users, so that they may get the most out of the satellite imagery. Below is an example of an application developed within the RAD team, which demonstrates how 3D object modelling can be performed using stereopair satellite images.

 

Image production

SANSA Earth Observation provides image processing for the following satellite imagery:

Landsat programme: Landsat 2, 3, 4, 5, MSS (LS 2-5 MSS); Landsat 5 TM (LS 5 TM); Landsat 7 ETM+ (LS 7 ETM+)

SPOT programme: SPOT 1, 2, 3 HRV; SPOT 4 HRVIR; SPOT 5 HRG

A variety of image processing levels are offered on the above listed sensors. The table below details the different processing levels and additional image processing options.

productionThe first processing option is to receive the images at raw level. With raw level images only radiometric and geometric correction at the sensor level are applied. The next processing option is to have the satellite image geolocated. The image is geolocated using either the satellite position or GCP points. Orthorectification is the next processing level. The satellite image is orthorectified using reference imagery and a 20m DEM. Orthorectified images are suitable for visual interpretation, spectral signature analysis, classification and derivation of indices (NDVI, EVI).
Together with to the above mention image processing levels, additional image processing options are available. Normalisation of a satellite image involves converting the DN values of the image to reflectance values at top of atmosphere. Some of the sensors contain a panchromatic band and as a result can have pan-sharpening applied to them. Pansharpening involves the fusion of the multispectral bands with the panchromatic band to produce an image at the resolution of the panchromatic band. Mosaicking involves the stitching together of many images into a larger single image. Clipping allows the user to select a subset of a satellite image. Reprojection allows for the specification of the projected coordinate system (UTM or geographic coordinate). Format change lets the user select which file format to have the images delivered in (Geo TIFF, JPEG2000 or PCI pix). The 'Subset S5 Mosaic' option is currently only available for SPOT-5 imagery. The user has the option of ordering a subset of 2.5m true colour SPOT-5 imagery created from the annual SPOT mosaics.

In addition to these broad processing options the image production service also provide clients with specific value addition services such as fire scar mapping and the annual SPOT-5 national mosaic.

 

Fire scar mapping

Clients have been calling on the fire scar mapping expert within SANSA Earth Observation's Image Production team to outline the history of specific fire scars. The areas that are investigated are specified according to a client's individual request. The fire scar mapping expert will then investigate the chronological sequence of a fire event, through the acquisition of a minimum of two images, one prior the event and the second post event. Images selected will be constrained by the revisit times of the sensor and as a result a mix of sensors (Landsat, SPOT, NOAA, MODIS and Meteosat) may be considered.

The images will be geographically corrected by orthorectification according to predetermined accuracy parameters. The objective of aligning the two images to the same reference will create the basis for true comparisons. Following this alignment of the images, SANSA's expert will then perform an image interpretation of the fire. A few examples of the different types of fire products produced will be illustrated below.
The first example shows the area before the fire event (Figure A), followed by an image acquired after the event; which then illustrates the extent of the fire after the burning (Figure B). Here is also a final classification on the area burnt (Figure C). The classified map includes the areas of trees destroyed as well as the grass patches.

fire
The second example shows how a fire can be tracked with the aid of Meteosat. Meteosat takes an image of the earth every 15 minutes. The resolution of the imagery is however very coarse (5km resolution). The pie chart fire progression is therefore displayed on a higher resolution Landsat image (30m resolution) for ease of visualisation. To interpret this image, each pie piece designates burning at a particular image time interval (in this case every 15 minutes). This fire starts with a red pie and progressively moves through the greens, and on to the purple. In the case below, it can be seen that the fire started on the left side (the red/orange pie slice) of the image and progressively moved to the right. It can be seen that a number of sections (pixels) were burning at the same time (all have the same colour pie piece, thus indicating that they were burning at the same time). This information in conjunction with meterological data for the area and time; highlight that this was an extensive fire that burnt rapidly due to high winds.

fire2

 

SPOT 5 South African mosaic

Mosaic 2011
An important national product provided by the SANSA Earth Observation directorate is the annual production of a country wide true colour composite image. The SPOT 5 mosaic has been produced annually since 2006. The creation of the SPOT 5 annual mosaic takes just over a year to produce, from tasking the collection of satellite imagery, selecting the required imagery, extracting them off archive, to a final mosaic product. The final mosaic is then packaged and distributed to approximately 30 government departments as well as to between 15 and 20 universities and research institutes. It is possible for subsets of this image to be ordered through the Customer Services division. On the right is an example of the mosaicked image produced in 2011.

 

Dont' miss the latest SANSA updates! FacebookTwitterYoutube