On my MacBook I have tried this many times: after opening a product with a band ‘Unw_Phase…HH’, click File->Export->CSV, then in the box give a name and tell SNAP to write to which folder, I get a file with ‘.csv’ extension. In this file the name of the 1st column of data is ‘featureID’ with integer value, the 2nd column is float values for another band in the product, and the 3rd column is float values of ‘Unw_Phase…HH’ band. I think I did nothing wrong, but can not get the plain text file I want: for all pixels in a band the 1st column is the longitude of the pixel, the 2nd column is the latitude of the pixel, and the 3rd column is the 'Unw_Phase…" at that pixel. Can you give me more help? Thank you!
I also find another function: ‘Raster->Export->Extract Pixel Values’ in the SNAP on my MacBook. For the ‘Unw…HH’ band in a product, the dialogue box does not show the ‘Unw…HH’ in the title row but ‘Latitude Longitude DateTime’, and I can not add the column name I want. Also It looks I need an expression to convert the un-wrapped phase to LOSD, yet where are the wavelength and incident angles? It looks they are not stored in the product. I can not find the relative material in the online tutorials for these post-processing steps. Please help! Thank you!
I still don’t get why you want a csv file.
If you tick to the BEAM-DIMAP format during processing no metadata gets lost. There you will find incidence angle and wavelength.
The Pixel Extraction Tool is used to find pixel values for a list of coordinates. If I understand you correctly, that’s not what you are looking for.
You could write a script to reformat the csv file to look as you want it. Also, you could do a band subset and only select lat, lon, and Unw…HH, so you only have these three left (you’d still have to live with the featureid column when expoting to csv).
After clicking the ‘Unw…HH’ band in a geocoded product, it shows a color contour graph in another view box. I want to add lon-lat coordinate axis and legend to this graph then it can be pasted to a Word file for a report. Because I can not find how to do this in SNAP’s tutorial material and its help, I want to export the data of this graph in a plain text file ( lon, lat, UnW in each column ) that other software can read and plot what I desire. Can you help on
(1) how to add lon-lat coordinate axis and legend to a graph?
(2) if SNAP is still not able to do this, how to export the desired data of a graph to a plain text file?
Sorry for not declaring my problem clearly because I just enter this geoscience field and may use some words and terms not commonly used this field.
An additional question, after the InSAR processing how to convert the un-wrapped phase to LOS displacement? I do not know where the exact wavelength value is and where the incident angle values stored.
If you want to export raster values with lat/long coordinates, create a rectangle by digitizing it at the area of your interest, select it and right-click > export transect pixels:
This allows you to write a txt file which you can easily open in Excel. If you select “export tie points” it contains la/lon bands.
I think I have imported the un-wrapped phase to SNAP correctly, and menu item ‘Radar>Interferometric>Product>Phase to Displacement’ does not work as you said. Thank you for the help! Here are still questions:
(1) In the picture of your ‘Band Maths Expression Editor’ window, in the ‘Expression’ box I think ‘3.11’ is the ‘wavelength in cm’. Where does you find this value? In the page of ‘Properties’ of my items( right click the product or band name ) no wavelength shown, even after opening the original downloaded image package.
(2) In my ‘Band Maths Expression Editor’ window, in the ‘Data source’ box there is no ‘incident_angle’ band shown as that in your picture. I think it should be somewhere in the product to be used in later processing. Where can I find it?
(3) I think ‘digitizing it at the area of your interest’ means selecting a subset for the interest area on the image. But ‘right-click’ on a subset or on a whole image, the item ‘Export Transect Pixels’ in the menu is always disabled( grayed ). What is wrong here?
I just find what ‘create a rectangle …’ means: draw the ‘purple’ rectangle by the drawing tools as shown in your picture. Yet after clicking ‘Export Transect Pixels’ it always shows an error ‘There is no transect defined in the selected band.’ What is wrong?
Regarding the incident_angle grid - did you check the box with tie-point-grids as I indicated with the red marks? If you imported and processed your SAR dara correctly, the should have been preserved in the product.
For digitizing vectors, note two things. First, you can save your product after digitization (File > save product) to make sure the changes are applied. Second, you have to select the digitized box with your normal mouse cursor (not the digitizing cursor) so it is marked with a yellow frame and white squares as in my example.
(1) Thank you for the table about the wavelength! Since each band is a range of frequency, the real frequency for a particular image data I downloaded should be stored somewhere in the package. Can you give me a hint how to get it? Picking up a value of frequency in the range will have the processing result not accurate enough.
(2) I find why there is no ‘incident_angle’ data in the ‘Data sources’ box when doing ‘Band Maths’: when I did ‘Radar>Ceometric>Terrain Correction>Range-Doppler Terrain Correction’ I did not check the box of ‘Local incidence angle’ and ‘Projected local incident angle’. Yet another 2 questions: (a) which angle( ‘local’ or ‘projected local’ ) should I use for ‘Band Maths’( any difference between an ‘incident’ angle and an ‘incidence’ angle )? (b) There is no ‘incident_angle’ data in the ‘Tie-Point Grids’ folder but ‘localIncidenceAngle’ and ‘projectedLocalIncidenceAngle’ in the ‘Bands’ folder. Is their data structure compatible?
(3) After drawing a rectangle on the image, how to remove it? There is no such an icon in the tool bar to do this.
you still didn’t tell us what sensor you are using Would help us to help you.
Yes, you can derive the exact wavelength from the frequency of your sensor http://www.wavelengthcalculator.com/
I find why there is no ‘incident_angle’ data in the ‘Data sources’ box when doing ‘Band Maths’: when I did ‘Radar>Ceometric>Terrain Correction>Range-Doppler Terrain Correction’ I did not check the box of ‘Local incidence angle’ and ‘Projected local incident angle’. Yet another 2 questions: (a) which angle( ‘local’ or ‘projected local’ ) should I use for ‘Band Maths’( any difference between an ‘incident’ angle and an ‘incidence’ angle )? (b) There is no ‘incident_angle’ data in the ‘Tie-Point Grids’ folder but ‘localIncidenceAngle’ and ‘projectedLocalIncidenceAngle’ in the ‘Bands’ folder. Is their data structure compatible?
Don’t apply Terrain Correction before the very last step. You don’t need your data in real geometries before and as you say, you lose much information. You can later terrain correct your displacement image. For the calculation of displacement, the global incidence angle (without topography) is used. You need the local incidence angeles (with topography) for other things, calibration for example. So those of the TC module don’t actually matter to you.
(3) After drawing a rectangle on the image, how to remove it? There is no such an icon in the tool bar to do this.
Just select it with the white mouse and hit the delete button of your keyboard.
Thank you for the response! Since I have many of image set downloaded, I think it is not polite to ask you to calculate or find the wavelength of each of them for me, but to know how to calculate it myself or where it is stored in the downloaded package for each sensor( ‘type’ of images ). For example the ‘type’ of my images includes ALOS1CEOS, ALOS PALSAR, TSX, RADARset2 and S1A, I can not know the exact wavelength from their property page, how to find it or calculate it by some other parameters in the package?
Thank you for the method of finding wavelength! Another problem: When selecting a subset from an S1A image ( it is so large, ~8GB ) it says “The functionality you have chosen is not supported for products with bands of different size. … Do you want to resample the product now?” After I clicking ‘Yes’ button and using the default parameters it says “SNAP-Error: Type: OperationException, Message: Cannot construct DataBuffer.”
(1) It looks the buffer is not large enough. Since my MacBook has 16GB RAM and more than 250GB free space, what is wrong in SNAP?
(2) As I checked on all recent S1A images downloaded, each has 6 bands of data and three( IW1, IW2 and IW3 ) 2 of them are in same size( one VH and one VV ). I do not know what role each band of data is in the processing( may get more precise result? ), yet I think using one VH and one VV should be OK, right?
So I want to delete other 4 bands, any potential problems after doing this?
You first have to deburst your data before the subset or just download the corresponding GRD product instead of SLC. It is smaller in file size and already debursted. You can’t use GRDs for interferometry however.
Follow the “TOPS Interferometry Tutorial” I have done interferometry for 2 S1A-TOPS images. Since the size of image file is large that needs long processing time and I want to do for more files, I try to use graph builder. Yet the menu-item( operator ) ‘S1 TOPS Coregistration’ can not be selected but 3 sub-menu-items. I think “Double-Difference-Interferogram” is what I want. Unfortunately it shows a red error message “split product is expected”. What does this mean? How to split an image for doing interferometry?
Follow the “TOPS Interferometry Tutorial” I have done interferometry for 2 S1A-TOPS images( date difference is 12 days ) . Then follow the step 8 to step 11 in the “Interferometry Tutorial” I have got LOS displacement. Then I have used “BandMath” with ‘incident angle’ convert the LOS displacement to the real displacement. Yet image of it automatically becomes in different gray levels, not in color. How to make it in color again?
I did not export “raster values with lat/long coordinates” into a plain text file for a while, and now have problem to do it. After drawing a rectangle on the image of geocoded un-wrapped phase, sometimes it shows a purple rectangle with 6 place holders around it yet sometimes it only shows yellow rectangle without place holder. If the later case it says “Transect pixels cannot be exported: There is no transect defined in the selected band.” Can you help where is the problem? Since I just patched 5 updates of SNAP, is the problem related?
Would help us to help you.
