ocean.Stanford.EDU

| home | syllabus | lecture notes |

GP141 - Satellite Remote Sensing of the Oceans

- - Lab Notes: More SeaDAS Functions- -




Even More SeaDAS!

For more specific information about these programs also see the SeaDAS Seadisp Program-Level Documentation at the SeaDAS homepage.


Data for Blotches/Histograms/etc...:

199707h09ma-gdm.hdf.Z  
199712h09ma-gdm.hdf.Z
Or your own favorite image

Either they are stored in your directory from earlier Lab work or just copy from :

/data8/gp235/lab_data/

Load using seadas:

SeaDAS Main Menu, -> 'Display'
Choose Pixel and Line sampling rate to be 8 each.
Click the Sea surface temperature button

Working with Blotches
[ Displayed Image Window -> Functions -> Annotate/blotch ]

Often you do not want to do statistics on an image as a whole (see histograms), but want to focus on a smaller area of interest, or want to compare 2 or more 'provinces' or 'zones' in the image. This can be done with the Annotate/Blotch function (find it under the Functions menu of the Display window).

With Annotate/Blotch you can draw blotches of various shapes; freestyle (polygon), ellipse and rectangular. Also note that you can add text or arrows on your image. Might be useful when you want to point out some interesting features on your fully processed image.

Example: Suppose you want to analyze the area of higher temperature in the middle of the Gulf of Mexico (see image). This feature might be an eddy or something. Using the Anntotate/Blotch function you can mark this area easily with the freestyle object (see figure). Just click and hold and then outline the feature on the image. The blotch is defined as the area marked by the color you choose (red is the default). So you will have to Fill the area Solid [ by clicking next to the 'Solid' button, in the 'Fill' category], otherwise SeaDAS thinks that the blotch is only the colored outline itself.

You can change the shape or move the blotch by draggin the white squares at each corner, or the middle square, resp. Managing blotches is a little bit tricky sometimes. It's easy to delete them accidentally. If you see the white squares, the blotch will be deleted if you select another object (Rectangular, Ellipse, etc.). Save them by pressing the Save button.

More blotches can be drawn as well on the same image. If the area of interest is not contiguous (ie., cannot be drawn with one blotch), you want to use the same color. If you want to have two separate areas of interest use a different color for the second blotch.

Often you want to do the same analysis on more then just this image. So you have to save the current blotches to an object file so you can load it later on another image. Saving to file is different than just pressing the save button on the Annotate window. (The latter save function is only to prevent that the blotch you have just drawn won't be deleted/replace when you want to add another one). The key is when using saved (to file) blotches that the new image should have the same projection and size as the original one. If not, the blotches will be misaligned. To save the blotches as an 'object', select File --> Save As. To load an earlier saved blotch select Load instead of Save As.

One can also add text using the Annotate/blotch function. For this check the circle next to 'ABC', Type the desired text in the text window; then click on the image to add it. The size, alignment, and font can be changed from the menu. Also, the text can me moved using the white squares. The text can be saved in the same way as the blotches.

Histograms

[ Displayed Image Window -> Functions -> Histogram ]

Some simple statistics of your image can be done by using the histogram function. Of the display window of interest select 'histogram'. A new window will pop up. Select 'blotch' instead of 'Full Image'. Select the Data Min and Max carefully. Whole numbers is easiest to understand. The number of bins (intervals) shouldn't be too high or too low. Some experimenting might be necessary. If for example the maximum value (say temperature) is 28.3500 degrees, select 30 as your max. As min you can use 0 or the real min (rounded). If you make the bin-size 1 (degree) the resulting histogram will be easy to analyse (so you will have the number of pixels which have a temperature of 0-1, 1-2, 2-3, ..., 29-30). Click on Plot and the result pops up (see figure). In the legend you can read the min, max, median, mean, etc. You can output (save) the histogram either as an image (eg. postscript or .png, by selecting Output on the histogram plot) or as data by saving it to a text (ascii) file (select Output Data on the Histogram Settings window). The ascii file you can download/ftp to your PC/Mac for further analysis, plotting in Excel for example.

The default blotch color is red. If you made more blotches, or drew them in a different color, select the appropriate color. The Non-Blotch selection will render a histogram of the full image excluding the blotch. Check the generated plot (and output data file) carefully if the number of points selected (shown at the bottom of the histogram plot) seems logical. Also, when you output data to a text file, the data of the last plotted histogram will be saved. For example, if you change the settings after you plotted a histogram, and click on Output Data the result will not be those of the changed settings but of the original histogram. So be sure to plot it first before you output the data.

Scatter Plots
[  Seadas Main Menu -> Utilities->Data visualization-> Scatter Plots ]

To see if there is any relationship/correlation between two parameters displayed in two different bands you can perform a scatterplot analysis. You can have the histogram of the input image show on the side as well. After you have selected 'Image 1' and 'Image 2' you might have to modify the Data and/or Axis range somewhat to exclude land/bad/missing data (see figure).Also you might want to choose the 'Blotched area' instead of 'Full Image'. For this, one has to load the saved blotches first from the annotate/blotch window.

Bathymetry Maps

[ Seadas Main Menu -> Utilities->Data visualization-> Generate Bathymetry Image ]

It is easy in SeaDAS to generate a bathymetry map with the same projection/size as your processed satellite image. Select 'Generate Bathymetry Image' under Functions (Seadisp Main Menu). Select the band number (see Band List Selection window) from where you want to retrieve the lat/lon limits, size, projection, etc., and click 'Okay' (see figure). The newly generated Bathymetry Image will show up as a new band in the Band List Selection window. Change to different color scheme if requried, in order to understand. Also, adding the color bar might help.

Profiles

[ Displayed Image Window -> Functions -> Rline ]

With the Rline function you can read and plot the pixel data along a line segment in the satellite image. When you start Rline from the image of interest a new window will pop up, as well as a '+' shaped cursor on the image. You can move the cursor by dragging the 'normal' cursor in the black box ('virtual mousepad') in the Rline settings window. To mark the starting point, either right click or click 'Add a new point'. Then move to your endpoint and right click again. A line will be drawn between the two points (see figure). More points can be added as well.

Then click on 'Plot line segment data' to generate a plot of the pixel values along that line. You might want to change the plot setup first by selecting 'Line segement data plot setup' (see figure).

FTP data to your directory: NCFTP and wget
Quickly 'ftp' one file to your home directory: wget. For example, on NASA's oceancolor website, in the Level-3 Standard Mapped Images browser, right click on the link to a file, and select 'Copy link location...'. On the ocean.stanford.edu server, you type: wget and paste the link. E.g.: wget http://oceancolor.gsfc.nasa.gov/cgi/getfile/A20080172008024.L3m_8D_CHLO_4.bz2 Then it will automatically downloard right away to your current directory.

| top | home | syllabus | lecture notes |