Q: What is the CHRIS?
A: CHRIS is an acronym for the Compact High Resolution Imaging Spectrometer. You can find more about the instrument here.
Return to top

Q: What is the PROBA?
A: PROBA is an acronym, standing for the PRoject for On-Board Autonomy, a project of the European Space Agency (ESA). ESA maintains a web-page for the PROBA platform here
Return to top

Q: What is so special about the CHRIS-on-PROBA mission?
A: Apart from demonstrating a significant new concept in platform design, the mission provides, for the first time, remotely-sensed Multi-View-Angle data at high spatial resolution, and in superspectral/hyperspectral wavelength resolutions. The richness of this data sampling is unique.
Return to top

Q: Where can I get some data?
A: If you are not one of the approved experimenters of the mission, you will probably not be able to get hold of any data for a while. If you are one of these experimenters, you will already have been given the necessary details.
Return to top

Q: Where can I obtain details of the data formats etc.?
A: Details are provided in the document "CHRIS_Data_Format" located on the CHRIS FTP site (accessible only to CHRIS PIs).
Return to top

Q: At what level of processing are the data supplied?
A: The data are provided in what is generally known as level 1a form. That is, they are corrected to top-of-atmosphere radiance values.
Return to top

Q: What are the units of the image data?
A: The data are given in radiance unitsof microwatta per square metre per nanometre per steradian.
Return to top

Q: Why does my data contain negative numbers?
A: Early versions of the processing software show what appear to be perforated lines, and on examination the low radaince pixels in these lines can contain large negative numbers. These arise from a hardware fault in the PPU which results in the loss of data from alternate columns in some image lines. This can occur in the raw spectral data, and also in the smear band. During the image processing, the smear band values are subtracted from the raw band data. If the raw spectral data is zero as a result of the PPU fault, the corrected value will be a negative number. The location of the lost data is random. Occasionally the drop-out happens on the smear band itself, and in these cases the perforation appears as a row of bright pixels.
Return to top

Q: At what angles are the images acquired?
A: This varies according to how close the site is the to the sub-satellite track. A more detailed explanation is given in the document "CHRIS_Data_Format", available from the FTP site.
Return to top

Q: How do I correct the data for the atmosphere?
A: That's up to you. The data were provided on the understanding that the experimenters were competent to do any further processing, atmospheric or geometric, that was required for the application. If this is not the case, please tell us so we can remove your site from the list of planned acquisitions. A number of software packages exist which can be used to carry out atmospheric correction: 6S and Modtran are widely used and ATCOR has its fans.
Return to top

Q: How do I co-register the different views?
A: That's up to you. The data were provided on the understanding that the experimenters were competent to do any further processing, atmospheric or geometric, that was required for the application. If this is not the case, please tell us so we can remove your site from the list of planned acquisitions. The simplest approach to geometric correction for flat scenes (core sites are generally flat) would appear to be by ground-control-pointing, although this is tricky for coastal sites or indeed, any site that lacks a decent spatial spread of recognisable features. As the images are so small, the earth's curvature is negligible and for flat areas the images should be simple affine transforms of each other (in other words, three ground control points should be enough for an adequate correction, provided those points are not nearly collinear). However, there is a chance that the spatial resolution in the along track direction may vary slightly from "top" to "bottom" of an image, so more points would be advisable. A more serious concern is that variations in relief across the image lead to a breakdown of the affine property, and the registration is not then readily treated by simple rubber-sheeting corrections of the kind that are to be found in most commercial image processing packages. Geometric correction of scenes of high relief is likely to be a severe technical challenge.
Return to top

Q: Can I have the lat-long coordinates of the image corners, please?
A: Not yet, and possibly not at all. The reason appears to be that the pointing vector of the satellite is not recorded, or at any rate not easily accessible from those engineering data which are telemetered to the ground station.
Return to top

Page last changed 3 May 2003