Child pages
  • VISAT FAQ
Skip to end of metadata
Go to start of metadata

How can I create a synergy product using MERIS and AATSR?

Assuming that you have two spatially overlapping products map-project the first one using VISAT's map projection or orthorectification tool. Then map-project the second product but this time using the ''co-locate with product'' option in the map projection dialog. You now have two products which use the same geo-reference system. Select one of the map-projected products and go to the band arithmetic tool. You will notice that you can now use bands of both products in mathematic expressions for the creation of new bands. Note that the creation of synergy products with VISAT is not limited to MERIS or AATSR but to any overlapping scenes using the same geo-reference system.

Why do I have no geo-referencing information in my exported products?

To get georeferencing informations into your exported product you have to use map-projection first. If you do not want to map-project your product you have also the option to export the geo-coding informations into an ENVI GCP file. Than your able to import the file into your GIS-Tool, if it suppports this format.

Is there a reason why all the INVALID_* (i.e. INVALID_BOA_VEG) flags do not show up in the list of the Arithmetic Expression Editor?

The "invalid_boa_veg" is the name of a bitmask-definition, its not a flag. If you double-click on the corresponding list entry in the bitmask window, you'll see the expression behind it, which is "l2_flags.LAND AND l2_flags.PCD_17". You can use this expression instead.

How is the PIN file (*.pnf) format defined?

The flat text format for Pin export and import is defined as follows: A head line wich contains the keywords 'Name', 'Lon', 'Lat' and 'Desc' in arbitrary order. Next to the headline must follow the pin values in the same order as defined by the headline. All the keywords and the pin values must be tabulator separated.

Two examples:

 
A well formatted PIN file:
        Name    Lat    Lon    Desc
        pin_1	45.200287	-8.047519
        pin_2	44.92977	-7.4097266
        pin_3	44.72365	-7.9369082
A exportet PIN file by BEAM:
        Productname:	MER_RR__1PNPDK20020421_104345_000002702005_00180_00735_1618.N1
        Date:	Tue Nov 25 09:16:19 CET 2003

        Pin    X    Y    Lon	Lat    Name    Desc    radiance_7    radiance_8
        0	141.5	205.5	-8.047519	45.200287	pin_1		13.9566765	12.663093
        1	194.5	222.5	-7.4097266	44.92977	pin_2		13.317661	12.1334305
        2	159.5	248.5	-7.9369082	44.72365	pin_3		13.644431	12.398261

Which algorithm is used for creating RGB images?

The algorithm used is defined by RGB-Profiles. Each profile consists of 3 arithmetic expressions, one for each RGB channel. In the preferences panel in VISAT your can create such profiles. Each time you create a RGB image you are asked which profile to use. At this time you can also define new profiles.

Why is the import of MODIS products disabled on Mac OS X systems?

The reason for this is the unavailability of HDF4 shared libraries for Mac OS X which are needed for the MODIS Product Reader. When they appear, we will include them in future BEAM releases.

Is it possible to define a spatial subset by geographical coordinates, not by pixel coordinates?

Unfortunately there is no simple way to do this. A practicable solution is, creating a PIN file(*.pnf) with four geographical coordinates definig your region of interest.
Example:

        Lat    Lon
        45.200287	-8.047519
        44.92977	-7.4097266
        44.72365	-7.9369082

Afterwards open your product with VISAT and use the pin manager to import your currently created file. Now go to File -> New menu. Select your product in the Pop-up Menu "Derive from Product". Select the Radio Button "Use Subset". Press the "Define Subset" Button. In the "Define Subset" Dialog enter in "Scene start X" and "Scene Start Y" the Pixel coordinates of your upper left corner (taken from the Pin Manager Window). Enter in "Scene end X" and "Scene end Y" the pixel coordinates of your lower right corner (taken from the Pin Manager Window). Scene Step X and Y should be 1. Press the OK-Button. Now you have created a subset which includes only your area of interest. Save it with "File -> Save as".

How can I extract tie point grids from a product for use with ENVI?

Just save your product as e.g. "prod.dim" in the BEAM/DIMAP format to disk. Open ENVI, go to the directory where you saved your product and open the folder "prod.data" and open the disired image. For example "slant_range_time.img", "slant_range_time.hdr" is the associated ENVI header.

Is it possible to do map-projection, subsetting & subsampling, product geo-boundary extraction or GeoTIFF creation in a batch mode?

Unfurtonately not, but these requests have a high priority on our feature list for the next releases of BEAM.

TIFF images are exported with 8 bit per pixel. Is it possible to export them with 16 bit per pixel?

Unfortunately not, but you can export your product as GeoTIFF file instead. In this case all pixel values are written as 32 bit float values.

How are pixels referenced in VISAT?

BEAM always references pixels by their centers. The upper left pixel of an image has the coordinate (0.5, 0.5).

Where is the jexp.jar (The Java Expression Parser & Evaluator) gone?

The jexp.jar is not required anymore by BEAM. Related classes are now part of beam.jar (and herewith also part of the public source code).

Is it possible to process a MERIS Level 1 image into a Level 2 using BEAM? If not, how can we have this "product" instead of the Level 1 images we already have?

BEAM cannot process MERIS level 1 data to official MERIS level 2 data products. However, there are some special purpose level 2 processors available in BEAM, e.g. FLH/MCI (standard), FAPAR (as plug-in) and Case II Water constituents (as plug-in). For ordering official MERIS level 2 products, please use the same way as you have used for your MERIS level 1.

Is it possible to use other DEMs in Visat instead of the GETASSE30 DEM?

Yes, it is possible by creating a Java plug-in. We account this while the development process. But until now we did not have found other DEMs which

  • have higher resolution (< 30")
  • are global
  • have no value gaps
  • are free to use

I want to create BEAM-DIMAP data product header for my ENVI image files so that I can read them into VISAT. What are the mandatory elements in the BEAM-DIMAP header file?

Have a look at the attached sample.dim XML file.

How can I improve the aspect quality of my pseudo-RGB images?

Here is how you can improve pale or dark looking RGB images in VISAT:

  1. After creating the RGB image from your data product open the constrast stretch / color palette window
  2. Remove 5-10% of the low-value pixels in the red and green channels by adjusting their contrast slider, keep most of the high value pixels in these channels.
  3. Remove 0-5% of the high value pixels in the blue channel by adjusting its contrast slider, keep most of the low-value pixels in the blue channel.
  4. In order to boost blue a bit and to keep cloud textures, adapt gamma as follows: gamma(R)=0.9, gamma(G)=0.9, gamma(B)=0.7

This procedure works best for RGB images derived from MERIS level-1b data products.

Are there plans to support the AATSR products ATS_AR_2P or ATS_MET_2P?

These AATSR products do not fit well in the BEAM data model, because they are composed of degree cells instead of rectangular pixel rasters for each band. Since we have no algorithm nor any requirements how to transform these cells into our pixel rasters, we do not plan to support these products in furture realeses. If you could provide us with some requirements how you want these data products to be represented in VISAT please let us know.

How can I export the spectrum of pixels?

There is more than one possibility to export a spectrum to a text file:

  1. Zoom into the image, move the cursor to the pixel of interest, click right mouse button and select "Copy pixel data to clipboard". Then paste data into MS Excel or similar.
  2. Set a pin using the pin tool, open the pin manager window, click the "filter" button and select the bands of interest in order to display pin pixel values and export data then.
  3. Define a region of interest using the ROI definition tool, click Main menu/File/Export/Export ROI pixels

How can I calculate with bands from products with different geo-location?

  1. Open all the products you want to calculate.
  2. Project one of the opened products with VISAT´s map projection tool Tools --> Map Projection...
  3. Now co-locate all the other products which you want to use in the calculation with the product which you have projected before. Use the co-location switch in the map projection dialog for co-location.
  4. Now select one of the projected products you want to calculate in the product tree view and open the band arithmetic dialog. All the co-located products will be listed in the expression editor dialog.

How can I use a ROI, which is defined in one product, also on other products?

You can use the "apply to other products" button located in the ROI definition window.

The automatic version checking of VISAT fails with an I/O Error. What can I do?

This can happen if you are behind a firewall where an http proxy is in use. To configure proxy support for VISAT you have to change the properties of the Java Virtual Machine. In the %BEAM-HOME%/bin directory is file named visat.lax. Open this file in an editor and add the following properties and adjust the values to your needs.

  • http.proxyHost=poxy.dummy.com
  • http.proxyPort=8080
  • http.nonProxyHosts=*.dummy.com|localhost

http.proxyHost and http.proxyPort indicate the proxy server and port that will be used. http.nonProxyHosts indicates the hosts which should be connected directly and not through the proxy server. Multiple hosts can be seperated by a |, and in addition you can use the wildcard character (*).

If I convert a product (e.g. MERIS L2) to HDF5 and read it in Matlab, the tie-point grids have now a different resolution as the other bands. Why?

The tie-point grids are interpolated to the scene resolution in order to directly display and analyse band pixel values with values from tie-point grids in VISAT.
If you would like to export tie-point grids with the full scene resolution, you can use the VISAT band arithmetic tool to create new "real" bands out of the tie-point grids, and then export the data product.

I have noticed that the tie point grids are image dependent and not equal in terms of localization and orientation. Why?

Like the pixel grids of the MERIS bands, tie point grids are also provided in satellite coordinates (as seen from the sensor) but subsampled.
No reprojection or rectification is applied to the scenes. The lat/lon values are given on the WGS-84 ellipsoid. The geo-coding of MERIS FR accuracy is better than 150 meter if and only if the observed scene surface matches the WGS-84 ellipsoid well (no mountains). See also BEAM help for the Orthorectification tool.

Here are two methods used to co-locate two FR scenes with the best accuracy possible up to now in VISAT using orthorectification with the
help of a digital elevation model.

Method A produces two separate outputs using orthorectification and direct co-location:

  1. Select first FR scene in the products view
  2. Goto Main Menu/Tools/Orthorectification
  3. Select projection and adjust output parameters so that the second scene fits into it as well
  4. Enter name of output product, e.g. SCENE-PROJ-1
  5. Klick OK --> You get a new product called SCENE-PROJ-1
  6. Select second scene in the products view
  7. Goto Main Menu/Tools/Orthorectification again
  8. Select projection and this time select "co-locate with" and select SCENE-PROJ-1
  9. Enter name of output product, e.g. SCENE-PROJ-2
  10. Klick OK --> You get another new product called SCENE-PROJ-2 which is co-located with SCENE-PROJ-1.

Method B produces a single output product by mosaicing two or more input scenes:

  1. Goto Main Menu/Tools/Mosaic Processor
  2. In the I/O parameters tab, add your two FR scenes
  3. In the product definition tab, select 'orthorectify' and adjust scene bounds
  4. In the processing parameters tab, add the desired band(s) e.g. "refl_5" and valid-mask(s), e.g. "NOT l2_flags.CLOUD"
  5. Press run --> You get a single output product containing the arithmetic averages of the two input products. SCENE-PROJ-1 and SCENE-PROJ-2 are now co-located.

When projecting data products the projection is only applied to bands, but tie-point grids are removed. Why is the projection not applied to tie-point grids, too?

Because of the data-explosion this would cause. You can always create a new band from a tie-point grid by using the band arithmetic tool. This will then get projected as well.

  • No labels