ATCOR for IMAGINE - Overview
ATCOR for IMAGINE has reached end-of-life with the upcomping ERDAS IMAGINE version 2018 in March 2018. The product will not be further supported. Please use the modern, radically improved ATCOR Workflow for IMAGINE.
ATCOR stands for Atmospheric and Topographic CORrection. ATCOR for IMAGINE is a professional tool for any user of satellite data, no matter whether the aim is to produce brilliant images without haze or clouds or whether you want to retrieve physical properties of the surface. ATCOR for IMAGINE is your choice to get better results.
Earth-observing satellite sensors map the Earth’s surface properties. However, haze from water vapour and aerosol particles influence the recorded signal. In addition, in rugged terrain, varying illumination conditions (well illuminated and shadowed slopes) mask the “true” spectral behaviour of surfaces. Therefore, atmospheric and topographic correction is essential to retrieve physical properties of the surface such as surface reflectance, emissivity and temperature.
ATCOR was originally developed by Dr. Rudolf Richter at DLR (German Aerospace Center). GEOSYSTEMS GmbH integrated ATCOR into ERDAS IMAGINE.
ATCOR-2 is the ATCOR-version of choice if the terrain covered by the imagery to be calibrated is almost flat or if a proper Digital Elevation Model (DEM) is not available. The underlying algorithm only requires an average ground elevation of the area. ATCOR-2 can also be used for fast checks and preliminary testing. For non-flat areas ATCOR-3 is recommended.
ATCOR-3 is the ATCOR-version of choice for the correction of satellite imagery acquired over rugged terrain. With ATCOR-3 a combined atmospheric-topographic correction of satellite image data is performed. Therefore, using ATCOR-3 a DEM is required. In addition to illumination effects caused by topography, ATCOR-3 also considers the elevation dependence of atmospheric effects.
Often optical satellite imagery is affected by haze and clouds. In cloudy areas there is no information about the ground surface, whereas in areas affected by haze the image still contains valuable spectral information. ATCOR contains a haze removal module for retrieving this information. It automatically detects hazy areas and produces a clear, haze-free image.
Haze removal is an optional pre-processing step that enhances your imagery visually. It is applied on at-sensor radiance data prior to the physically-based image correction that converts at-sensor radiance to surface reflectance. If you just need a visually appealing image without being interested in physical quantities, haze removal without any further processing will satisfy your needs.
The ATCOR haze removal module
- can be applied to multispectral data as provided by most common sensors with bands in the visible and near-infrared spectral region
- eliminates haze, but not clouds. The de-hazing results depend on the percentage of haze in the image and on the 'transparency' of the hazy areas especially in the red and the near-infrared band.
- is currently not suitable for haze over water.
Atmospheric correction converts raw pixel values to quantities with a physical meaning, i.e. surface reflectance in case of solar bands and surface temperature in case of thermal bands. From these quantities ATCOR derives a variety of products, comprising vegetation indices, albedo, and quantities relevant for surface energy balance investigations. Key inputs for applications related to climatology and all kinds of biophysical modelling. ATCOR provides the following products:
- Soil adjusted vegetation index (SAVI)
- Leaf area index (LAI)
- Fraction of photosynthetically active radiation (FPAR)
- Surface albedo
- Absorbed solar radiation flux
Additional products in case of at least one thermal band:
- Thermal air-surface flux difference
- Ground heat flux
- Latent heat
- Sensible heat flux
- Net radiation
ATCOR for IMAGINE Spatial Modeler
Later this year we will release a NEW ATCOR for IMAGINE Spatial Moder. This new ATCOR
will enhance and optimize your personal Spatial Modeler Workflow.