Hosted by "1PLs Agency"

Home > 6. Motion Vectors > 6.5 GDP/ASCII

6.5  Surface Current Vectors from the Global Drifter Program: ASCII Spreadsheets Option

1.  The above ASCII data files (whichever ones you want) should be downloaded to DATA > OCEAN > GDP.  You can use the file names used below, or the original filenames.  They all have months names to identify them correctly.  But please use the latest version of the dataset.
2. Run Saga. 
  • Use FILE > GRID > LOAD to load the 0.5-degree dummy grid listed above
  • Then use FILE > TABLE > LOAD to load any months you want to work with right now.  Here you can see 4 of them (JAN, APR, JUL, OCT).
3.  Open any table to see its contents (from the GDP webpage):
  • Lat (1x317): latitude (degrees), negative = south

  • Lon (1x720): longitude (degrees), negative = west.

  • Ubar (317x720): mean eastward speed (m/s) vs. Lat/Lon.

  • Vbar (317x720): mean northward speed (m/s)

  • SSTbar (317x720): mean sea surface temperature (degrees C)

  • eUbar, eVbar, eSSTbar: Not included after Version 2.06

  • N (317x720): Number of drifter-days per square degree


4.  To work with the data, we need to convert the tables to point shapes.  Select TOOLS > SHAPES/POINTS > CONVERT TABLE TO POINTS.
  • TABLE - Select the month you want, e.g. January
  • X - Select LONGITUDE
  • Y- Select LATITUDE
  • Z - Leave it alone to select all the other fields

Then click OK.

5.  Now you have your first point shape, as you see here.  Check to make sure.
6.  Here you can use ADD TO MAP to see if it resembles a global map.  Seems to be ok.
7.  To fix the map and make it easier to see, make these settings:
  • OUTLINE - Unchecked
  • VALUE RANGE - ATTRIBUTE > SSTBAR (just as an example)

Now click APPLY at the bottom.

8.  You should see something like this.  It indicates you're working with good data and they are behaving as expected.

Remember, these are thousands of actual surface measurements, not satellite data.

9.  And here we've zoomed into the Liberia area to confirm that the original GDP data are a regular array of points.
10.  Now to work with these data to see the current vectors, we need to grid the point shapes.
11.  Select TOOLS > GRID-GRIDDING > INVERSE DISTANCE WEIGHTED (the gridding algorithm we frequently use in Saga):
  • POINTS - Select the new point shape (January in this case)
  • ATTRIBUTE - Select Ubar (the east-west component)
  • TARGET GRID - Select GRID (means we will specify a dummy grid in the next step)
  • SEARCH RANGE - Local
  • MAXIMUM SEARCH DISTANCE - 1 (degree) which should be enough for a 0.5-degree array
  • NUMBER OF POINTS - All points
  • DISTANCE WEIGHTING - Inverse distance to a power
  • POWER - 2

Of course you are invited to experiment with these settings and use anything you like better.

Now click OK to continue.

12.  Now you're asked to specify the dummy grid for the gridding.  Select the 0.5-degree dummy grid.

Make certain that GRID is set to CREATE, or you might lose something you want to keep.

Now click OK.

13.  Now a new grid appears ("Ubar") in the same system as the dummy grid.
14.  You can use ADD TO MAP to see the grid.  The strong eastward current along the coast of Cote d'Ivoire and the westward flow in the equatorial area are both expected features.

The LEGEND confirms the data values must be m/sec.

15.  Use the same method to grid the Vbar data.  It is again extremely important you set GRID to CREATE, so you don't over-write the previous grid.
16.  And here is the Vbar grid.  Typically, north-south components are much weaker than east-west, as you see here.
17.  Check to make sure that you have these 2 new grids in the proper system.
18.  Now that we have the U and V grids, we can make current vectors.
  • GRID SYSTEM - The dummy system with the new grids
  • X COMPONENT - Ubar
  • Y COMPONENT - Vbar
  • STEP - Distance between vectors in the source grids; usually begin with 1; increase if arrows are too crowded
  • SIZE > MINIMUM - If not sure, then select zero to be safe
  • SIZE > MAXIMUM - Select 1000 to be safe; expenence with Saga shows this item behaves quite erratically from dataset to dataset
  • AGGREGATION - Mean value; if multiple values are given for a single location
  • STYLE - Cell-centered arrow

Now click OK to make the vectors.

20.  A new line shape appears with a double name indicating the sources of the components.
21.  And here you can see the vectors by using ADD TO MAP.
22.  Here are some adjustments you can make in the properties panel:
  • STEP - Set to 2 to double the distance between the arrows
  • MAXIMUM - Reduce to 400

Click APPLY in the bottom of the panel.

23.  Now the vectors are not so crowded, and much easier to understand.
24.  Here are some more adjustments you can make to improve appearance:
  • TYPE - Graduated colors
  • ATTRIBUTE - LEN (sets the palette colors to reflect the "length" of the vector arrows)

Click APPLY in the bottom of the panel.

25.  These are the best vectors you can do in Saga, with 2 kinds of visualization of the currents speeds:  arrow size and arrow color.
26.  If you want to use these vectors in another system, Integrated Data Viewer (IDV) for example, right-click on the line shape and select SAVE AS.
27.  Navigate to PRODUCTS > SAGA > VECTORS and save the shape with the name current_surface_liberia_jan_gdp_ver2.06_saga.shp

28.  And here, in IDV, is the same shape, added through DATA CHOOSERS > FILES.  Note that this object is a true vector, not a pixilated image, so you can zoom in or out without any distortion.

29.  Also in Saga, you can use IMPORT/EXPORT-GDAL/OGR > EXPORT SHAPES TO KML to create vector shapes for Google Earth that work the same way.
30.  If, though, you want to display very large areas or even the entire globe, then the initial display might look a little strange.  If you want to deal with these vector fields, then go to the older exercise, 6.3